g~~~~~~~~~~ I PON | | i5 m~ Januay20 Health, Nutrition, and Population Series This series is produced by the Health, Nutrition, and Population Family (IFlP) oif the World Bank's Human Developmenlt Network. It provides a vehicle for publislhing polished material on the Banuk's work in the I-INP Sector, for consolidating previous mfomlal publica- tions, and for improvinig the standard for quahty control, peer review, and dissemination of high-quality analytical work. The series focuses on publicauions that expand oaLr knowledge of -TINP policy and strategic issues that can improve outcomes for the poor and protect vulnerable populations against the impoverishing effects of illness. Best practice examples of both global and regional relevance are presented through thematic reviews, analytical work, and case studies. The Editor in C hief of the series is Alexander S. Preker. Other members of the 'Editorial Committee are Mukesh Chawla, Mariam Claeson, Shantayanan Devarajan, Gilles Dussault, A. Edward El_-mendorf, Arniun H. Fidler, Charles C. Griffin, Jeffrey S. Hainnier, Peter F. Heywood, Prabhat Jha, Gerard Martin La Forgia, Jack Langenbruinner, Ruth Levine, Maureen Lewis, Samuel S. Lieberman, Benjan-un Loevinsohni, Elizabeth Lule, Akiko Maeda, Judith Snavely AlcGuire, Milla McLachlan, Tlhomas W. Merrick, Philip Musgrove, Ok Pannenborg, Oscar Picazo, Mead Over, Juan Rovira, George Sclheber, and Adam Wagstaff. Innovations in Health Service Delivery The Corporatization of Public Hospitals Alexander S. Preker April Harding Editors THE WORLD BANK Washington, D.C. © 2003 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone: 202-473-1000 Internet wwwworldbank.org E-mail: feedback@worldbank.org All rights reserved. First prntngJanuary 2003 1 2 3 4 05 04 03 The findings, interpretations, and conclusions expressed herein are those of the authors and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denommations, and other information shown on any map in this work do not imply any )udgment on the part of the World Bank con- cerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete mformation to the Copyrnght Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other quernes on rights and licenses, including subsidiary rights, should be ad- dressed to the Office of the Publisher, World Bank, 1818 H Street, NW, Washing- ton, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. ISBN 0-8213-4494-3 Library of Congress Cataloging-in-Publication data has been applied for. Contents Foreword xi Sir Alan Langlands, FRSE Foreword xiii James Christopher Lovelace Acknowledgments xvii Acronyms and Abbreviations xix Introduction 1 Part 1: Overview of Marketizing Organizational Reforms 21 1 A Conceptual Framework for the Organizational Reforms of Hospitals 23 April Harding and Alexander S. Preker 2 Implementing Organizational Reforms to Hospitals in the Public Sector 79 Chris Ham and Loraine Hawkins 3 Evaluating the Impact of Organizational Reforms in Hospitals 105 Mead Over and Naoko Watanabe 111 iv o Innovations in Health Service Delivery Appendix 3A. Selected Indicators 152 Appendix 3B. World Bank Hospital Reform Projects 165 Part 2: Cross-Country Views 167 4 Reviewing the Case Studies: Tentative Lessons and Hypotheses for Further Testing 169 Loraine Hawkins and Chris Ham 5 The Missing Link? Hospital Reform in Transition Economnies 207 Melitta Jakab, Alexander S. Preker, and April Harding 6 Contracting Public Health Care Services in Latin America 239 Alexandre Abrantes Part 3: Case Studies 263 7 Betwixt and Between: Autonomization and Centralization of U.K. Hospitals 265 Chris Ham Appendix 7A. National Health Service and Community Act 1990 297 8 Maladjustments in the Corporatization Model: Hospital Reform in New Zealand 305 Graham Scott, Lynne McKenzie, and James Webster 9 Autonomous Hospitals Become a Commercial Network: Hospital Rationalization inVictoria, Australia 345 Simon Corden 10 Autonomizing a Hospital System: Corporate Control by Central Authorities in Hong Kong 391 Winnie C. Yip and William C. Hsiao Contents v 11 Corporatization of a Single Facility: Reforming the Malaysian National Heart Institute 425 Rozita Halina Tun Hussein, Syed AI-Junid, Soe Nyunt- U, Yahaya Baba, and Willy De Geyndt 12 Attacking Hospital Performance on Two Fronts: Network Corporatization and Financing Reforms in Singapore 451 Kai Hong Phua 13 Autonomous Structures-with Incomplete Autonomy: Unusual Hospital Reform in Tunisia 485 Hedi Achouri and Eva jarawan 14 Autonomization in Indonesia: The Wrong Path to Reduce Hospital Expenditures 511 Samuel S. Lieberman and Ali Alkatiin 15 From Plans to Actions: Hospital Refonn in Ecuador 533 Patricio V Marquez, Fernando Sacoto, and Marta Molares-Halberg Glossary of Terms 549 Bibliographies 559 About the Authors 589 Index 597 Tables 2.1 Tools for Planning and Managing Change 99 3.1 Hospital Performance Indicators, by Dimension and Production Stage 128 3.2 Technical Efficiency Indicators 129 3.3 Allocative Efficiency Indicators 131 3.4 Quality Indicators 133 vi o Innovations in Health Service Delivery 3.5 Equity Indicators 135 3.6 Hospital Autonomy in Ghana, India, Kenya, and Zimbabwe 136 3.7 Methodological Approaches to Assessing Reform's Effects on Hospital Performance 138 3.8 Performance Dimensions by Stage of Production Process, Two Sample Surveys 143 3A.1 Selected Indicators, Organizational Reform 152 3A.2 Selected Indicators, Hospital Management 155 3A.3 Selected Indicators, Hospital Performance 158 3B World Bank Hospital Reform Projects 165 4.1 Applying the Conceptual Framework 170 4.2 Spectrum of Case Studies and Their Market Exposure 174 4.3 Preliminary Assessment of the Success of Organizational Reforms 177 4.4 Success and Failure in Planning and Managing Implementation 195 5.1 Ownership and Legal Organizational Status of Hospitals 214 5.2 Decision Rights on Labor Input, Selected Countries 216 5.3 Internal Incentive Environment during Transition 229 6.1 Self-Managed Hospital Autonomy in Argentina 244 6.2 Self-Managed Hospital Autonomy in Argentina, Chile, and Uruguay 250 6.3 Performance Indicators in Hospital Management Contracts in Argentina, Chile, and Uruguay 253 7.1 Comparison of the Regimes of NHS Trusts and Directly Managed Units 275 8.1 Hospital Powers Pre- and Post-1993 Reform 308 8.2 Area Health Boards: Aggregate Trends in Costs and Deficits 325 8.3 Crown Health Enterprises: Aggregate Trends in Costs and Deficits 326 8.4 Users' Views on New Zealand's Health Systems 332 9.1 Timeline of Policy Changes 348 Contents * vii 9.2 Acute Health Care Services Access Indicators 359 9.3 Merging Acute with Extended Care Providers 367 10.1 Health Expenditures in 1986-87 and 1996-97 392 10.2 Characteristics of Governance in Hospitals before Corporatization 406 10.3a Characteristics of Governance after Corporatization: Government and the Hospital Authority 406 10.3b Characteristics of Governance after Corporatization: Hospital Authority Head Office and Hospitals 407 10.4 Hospital Authority Attrition Rates 412 10.5 Patient Satisfaction with Selected Aspects of Hospital Authority Services, 1997 414 10.6 What Do Public Hospitals Most Need to Improve? 415 10.7 Trends in Recurrent Cost per Unit of Output 417 10.8 Hospital Authority Trends in Outputs 419 11.1 Socioeconomic and Demographic Indicators, Selected Asian Countries, 1996 427 11.2 Number and Type of Health Service Facilities in Rural and Urban Malaysia, 1996 429 11.3 Operating Statistics for the National Heart Institute, 1993-97 439 11.4 National Heart Institute Operating Revenue, by Source of Financing, 1993-97 441 11.5 National Heart Institute Operating Expenses, FY 1992-93 to 1996-97 442 11.6 National Heart Institute Performance Indicators, 1993-97 444 11.7 National Heart Institute's Efficiency and Productivity Gains from Corporatization 444 11.8 Total Revenue and Expenses per Inpatient-Day, 1993-97 144 12.1 Average Cost of One-Day Stay in Hospital, April 1989 465 12.2 Average Cost of "A" Class Hospitalization 465 12.3 Average Cost of One-Day Stay in Hospital, December 1989 466 viii ° Innovations in Health Service Delivery 13.1 Operating Costs of Public Health Facilities, by Source of Financing 494 13.2 Health Care Spending 499 13.3 Changes in EPS Operating Budgets 505 14.1 Class III Bed Charges before and after Swadana 518 14.2 Actual Costs and User Charges 519 14.3 Increase in Sample Hospital Revenues, by Year 520 14.4 Hospital Revenues and Subsidies, by Year 521 14.5 Beds in Sample Hospitals, by Year 522 14.6 Total Hospital Expenditures per Bed, by Year 523 14.7 Bed Occupancy Rates in Sample Hospitals 524 14.8 Trends in Average Incentive Package, by Staff Category 524 14.9 Occupancy Rates in MOH-Owned Hospitals 525 15.1 Fee Schedule for Selected Procedures 541 Iigueus5 1.1 Key Determinants of Changes in Organizational Behavior 44 1.2 Putting It All Together 49 1.3 Forces Determining the Level of Competition 51 1.4 Complementary Systemic Measures for Organizational Reforms to Work 69 2.1 A Framework for Health Policy Analysis 82 3.1 Framework for Evaluating Hospital Management Reform 114 4.1 Authorizing Environment for Effective Implementation 196 5.1 Determinants of Hospital Behavior 209 5.2 The Hospital Environment in the Socialist Era 211 5.3 The Hospital Environment during Transition 212 7.1 The Structure of the NHS in England, 1996 271 7.2 Source of Trusts' Financial Failures 284 12.1 Cost-Recovery Ratio, Restructured Hospitals, 1990-97 470 12.2 Hospital Revenues, 1990-97 471 12.3 Hospital Expenditures, 1990-97 471 12.4 Physician/Nonphysician Ratios, 1989-97 472 Contents * ix 12.5 Admissions, Public and Private Hospitals, 1989-98 473 12.6 Admission Rates, Public and Private Hospitals, 1990-98 474 12.7 Average Length of Stay, Singapore General Hospital, 1990-97 475 12.8 Bed Occupancy, by Ward Class, Singapore General Hospital, 1990-97 475 12.9 Bed Occupancy Rate, by Ward Class, Public Sector Hospitals, Various Years 476 Boxes 1.1 Definitions: The Players, the Rules, and the Objects of the Game 25 1.2 Influence Activities 33 1.3 High-Powered Ownership Incentives 35 1.4 Incentive Environments from Public to Private 42 1.5 Internal Markets versus Performance Budgeting: What's the Difference? 50 1.6 Market Environment: Structure and Competition 56 1.7 Good Governance: What Is It and Why So Little in Public Hospitals? 59 1.8 Key Lessons from Other Sectors 65 3.1 Dimensions of Organizational Reform 118 3.2 Dimensions of Hospital Behavior 124 6.1 From Direct Public Hospital Budget Allocation to Nonbudgetary Revenue 246 7.1 Efficiency Initiatives in the 1980s 268 7.2 The Financial Regime of NHS Trusts 278 7.3 The Dissolution of Trusts 288 7.4 Decision Rights 295 9.1 The Victorian Health System 346 9.2 The Changing Governance of Public Hospitals in Victoria 351 9.3 Introducing Service Quality Improvements in a Network Environment 365 9.4 Equity Issues Faced by Network Boards 372 x o Innovations in Health Service Delivery 10.1 Hospital Authority Relations with Individual Hospitals 398 12.1 Timeline of Health Care Reform in Singapore 454 12.2 Assessing the Finance Reforms 478 14.1 Ward Classes and Internal Cross-Subsidies in Indonesian Public Hospitals 516 15.1 Decentralized Health System in Tena Canton 537 15.2 Socioeconomnic Classification of Health Care Users in Ecuador 542 15.3 A Public HMO: Community System for Integral Health in Colinas del Norte 545 Foreword There is a shift in currency in many health care systems around the world. We are moving away from the old pattern of primary, sec- ondary, and tertiary provision toward integrated networks of care. The old professional hierarchies and established approaches to edu- cation and training are making way for a new currency of knowledge, skills, access, information, and teamwork. Many countries are also seeking to redefine the ethos of professional practice and public ser- vice in ways that will keep the trust of the people, calling into ques- tion traditional forms of professional and public accountabilty. As ever, change is endemic in all dimensions and at every level of our health care systems as policymakers and practitioners juggle with rising public expectations, the breathtaking speed of advances in sci- ence and technology, changes in the burden of disease, and the re- sulting imbalances between supply and demand. As media scrutiny in- creases, people are making the link between the competence of governments and the performance of domestic public services, ensur- ing that political interest in health systems is more intense than ever. Amid all this change, some things remain constant. Many coun- tries are as committed as ever to achieving three key results in their health care systems: Equity: improving the health of the population as a whole and re- ducing variations in health status by targeting resources where needs are greatest, xi xi, ° Innovations in Health Service Delivery Efficiency: providing patients with treatment and care that is both ef- fective and a good value, and Responsiveness: meeting the needs and wishes of individual patients and users. And despite the trends toward integrated networks of care, hospi- tals of all shapes and sizes remain central to the provision of good health care, striving to maintain the quality and cost-effectiveness of their services and relying on the skill, motivation, and sheer pro- fessionalism of their staff to deliver these services, often in difficult circumstances. The question of how best to run our hospitals has been a subject of intense interest for decades with a strong focus over the past 15 years on what Alex Preker and April Harding call "marketizing re- forms." In this important book, they seek to understand the design, implementation, and impact of these reforms in a number of differ- ent countries. By establishing a powerful conceptual framework, tak- ing care in their analysis, and drawing on the experience and insights of the authors of the case studies, they cast light on the impact of marketizing reforms and identify some of the barriers to the effective implementation of change. In a world where we often seem to learn and relearn the same lessons over and over again, this book and the companion volumes to be published will provide a valuable source of advice to policymakers and practitioners as they work to make things better. Sir Alan Langlands, FRSE Principal and Vice Chancellor University of Dundee, United Kingdom, and Former Chief Executive National Health Service Foreword From 1992 until 1995 I had the privilege of being Director-General of Health in New Zealand. Graham Scott, Lynne McKenzie, and James Webster write an interesting case study of the New Zealand reforms in chapter 8 of this volume. The authors point to a panoply of factors associated with the reforms that conspired to undermine their success, but to me it boiled down to five main lessons. First, it is very difficult to remove the partisan political process and the role of government from a corporatized model, and the no- tional separation between the ownership interest and the public pol- icy interest therefore blur, whatever the apparent firewalls. It is bet- ter to recognize this explicitly and deal with it in a transparent manner. The owners regularly engaged with the boards and man- agers of their newly formed public health corporations, the Crown Health Enterprises (CHEs), oftentimes to avoid political conse- quences of the CHEs' intended decisions. Second, corporatization lifts the veil of hidden subsidies off a pub- lic health system. For example, balance sheets had to be created, as- sets valued, depreciation considered, and liabilities, including con- tingent liabilities, had to be accounted for. This all required explicit financing. Further, the model suggested relatively easy access to, and exit from, the health system as service providers, and for CHEs the ability to exit from unprofitable services and dispose of poor- performing assets. This proved far harder in practice. Conspiring with lesson one, the expedient thing to do was to minimally take these xlii xiv o Innovatnons in Health Service Delivery things into account and try to avoid creating the impression that the reformed system cost more than the unreformed one. Third, many features of the model were either not introduced, or abandoned at an early stage. For example, contestability was seen to be an important element of the New Zealand reforms on both the purchaser and the provider side of the reforms. Contestable pur- chasing was abandoned before the reforms were formally intro- duced, largely due to legitimate technical considerations but also due to a concern that critics of the reforms would characterize this as an "Americanization" of the health system. On the provider side, efforts to move in this direction were somewhat half-hearted. Similarly, the CHEs were intended to be taxable, an effort to level the playing field, but early on this was surrendered to political considerations. No one decision mattered so much as taken collectively they served to undermine the integrity of the model, and gave encouragement to the critics of the reform process. Fourth, related to the second point, due to concerns of general fis- cal restraint, and the potential of an admission that the reformed structures were more costly than the system they replaced, there was a real reluctance to adequately finance the public health purchasers for the volume of services they felt necessary to meet the rising ex- pectations and health needs of New Zealanders. Indeed, a contrary decision was taken to finance health services by explicitly allowing CHEs to run significant deficits (and thus run down their balance sheets). This was in part due to a belief that there were efficiencies to be obtained (and there were), but beyond that it was clear to those of us involved that there was a significant structural deficit as well and not financing this only served to undermine the purchaser- provider split (since the price and/or quantum of services purchasers needed to buy was greater than the price they could pay). This, in turn, led to a fifth mistake: reacting to problems by "drip feeding" additional funds into the system, thus creating an atmos- phere of perpetual crisis, followed by continuous funding adjustments. I differ with the authors in some of the interpretation of details, particularly in one specific respect. Ministers and their advisers (in- cluding me and Graham Scott, who was secretary of the treasury for Foreword * xv part of this time) considered many of these issues, both in the design and subsequently at each of the many decision steps along the way. There were some in the reform process who consistently underesti- mated the consequences of these incremental decisions-and, in truth, their consequences are much easier to see in hindsight. But in my opinion, as the political heat turned up, it became irresistible for the government not to tinker to reduce the heat-understandable but, in this case, fatal. This lesson, the impact of the full political economy on the reform process, overshadows the many important, if somewhat arcane, specifics of the reform process. It is a lesson ap- plicable well beyond the "Land of the Long White Cloud." The subsequent consolidation of purchasers into a single national authority eliminated whatever vestiges of competitive (or at least comparative) purchasing, and with the election of a new Labour and Alliance Government (both parties opposed the reforms from the outset), the circle was completed with a new set of reforms, which look ever so much like the reforms Labour introduced back in the 1980s. None of this makes me think the possibilities of corporatizing public health assets are any less real, just more difficult to achieve. Finally, I note that there were many successful and enduring as- pects of the health reforms in the 1990s, from new laws on health privacy and health information, to a new New Zealand Health In- formation Service and the introduction of minimum national data sets, unique identifiers, pioneering work on contracting and pur- chasing of services, important work done by a National Advisory Commission on Health and Disability Services, and many others, too numerous to mention here. I hope you find this volume of interest and of value. James Christopher Lovelace Director, Health, Nutrition, and Population and Chair, Health, Nutrition, and Population Board The World Bank, and Former Director-General of Health, New Zealand Acknowledgments The authors are grateful for reviews provided by Simon Blair, Philip Davis, Jerry La Forgia, and Christopher Lovelace. Invaluable in- sights from thematic reviews and country case studies were provided by Alexandre Abrantes, Hedi Achouri, Ali Alkatiri, Yahaya Baba, Mukesh Chawla, Simon Corden, Willy De Geyndt, Ramesh Govin- daraj, Christopher Ham, Loraine Hawkins, Judith Healy, William C. Hsiao, Rozita Halina Tun Hussein, MelittaJakab, Syed AlJunid, Eva Jarawan, Patricio V Marquez, Samuel S. Lieberman, Martin McKee, Lynne McKenzie, Soe Nyunt-U, Mead Over, Hong Phua, Laura Rose, Graham Scott, Naoko Watanabe, James Webster, and Winnie C. Yip. Kathleen A. Lynch provided the text editing. Deirdre Ruffino provided the copyediting, and Sithie Mowlana assisted with word- processing. xvii Acronyms and Abbreviations AHBs area health boards ARCI accident rehabilitation compensation insurance CCMAU Crown Company Monitoring and Advisory Unit CEE Central Eastern Europe CEO chief executive officer CHEs Crown Health Enterprises CPF Central Provident Fund DHS Department of Human Services DRGs diagnosis-related groups EFL external financing limit FSU Former Soviet Union GMS government medical store HA Hospital Authority HCS Health Corporation of Singapore HGC hospital governing committee HKSAR Hong Kong Special Administrative Region HWB Health and Welfare Branch MHCN Municipal Integrated Health Care Network MHD Medical and Health Department MHPB Metropolitan Hospitals Planning Board MIS management information systems MOH Ministry of Health MPH Ministry of Public Health NEP New Economic Policy XiX xx ° Innovations in Health Service Delivery NGO nongovernmental organization NHI National Heart Institute NHP National Health Plan NHS National Health Services NTUC National Trades Union Congress NUTH National University Hospital OECD Organisation for Economic Co-operation and Development PCU project coordination unit PFI Private Finance Initiative RHA Regional Health Administration RHAs Regional Health Authorities SGH Singapore General Hospital SMPHs self-managed public hospitals SOE state-owned enterprise TPP total purchasing project WHO World Health Organization Introduction Despite much attention and emphasis on primary care as a first point of contact for patients, hospitals in most countries remain an impor- tant source of critical health care services, providing both basic and advanced care for the population. Hospitals are often the provider of last resort for the critically ill and poor. Yet hospitals also comprise the largest expenditure category of the health system of both indus- trial and developing countries. As a result, although their critical role as an integral part of the health system is well recognized, hospitals are often the target of health sector reforms aimed at efficiency, eq- uity, and quality improvements and more systemic reforms in fi- nancing and the health care delivery system. This volume provides some insights about recent trends in the re- form of public hospitals, with an emphasis on organizational changes such as increased management autonomy, corporatization, and pri- vatization. The material presented tries to answer three questions: (a) what problems did this type of reform try to address; (b) what are the core elements of their design, implementation, and evaluation; and, (c) is there any evidence that this type of reform is successful in ad- dressing problems for which they were intended? Why Look at Autonomization, Corporatization, and Privatization of Public Hospitals? Decentralization, which dominated much of the discussions on structural reforms in the public sector during the 1960s and 1970s in 2 a Innovations in Health Service Delivery industrial countries, arrived in the developing world during the 1980s. By that time, Western Europe had turned its attention to improving the performance of government-owned services through organizational reforms of the service providers themselves. This included altering the incentive regime that managers within the or- ganizations were exposed to and changing the external policy envi- ronment, governance structures, funding arrangements, and com- petitive pressures. Reforms of this type, which are now commonplace throughout the world in the infrastructure, telecommunications, and transportation sectors, have included: (a) increasing the management autonomy of the organization (autonomization); (b) transforming the hierarchical bureaucracy into parastatal corporations that are exposed to market- like pressures (corporatization); and (c) outright divestiture of the or- ganizations from the public sector (privatization). They are often re- ferred to as "new public management" or "marketizing" reforms. Influenced by the lessons learned from the problems and reforms tried in other sectors, many health care policymakers concluded that the performance problems of public hospitals were similarly grounded in the rigidity of hierarchical bureaucracies, the lack of control by managers over day-to-day operations of their facilities, and the absence of performance-based incentives. Having success- fully applied new public management techniques and marketizing re- forms in other sectors, it was a natural step for policymakers in some countries to consider applying similar reforms to the health sector. Initially, the reform of choice was to give hospitals some degree of management autonomy. Limited success with this type of reform in some settings led policymakers to go a step farther by transforming some of their state-owned hospitals into public corporations. The path-breaking reforms of this type, which occurred through the creation of Hospital Trusts in the United Kingdom and Crown Health Enterprises in New Zealand, drew worldwide interest. Soon many developing economies such as Hong Kong, Malaysia, Indone- sia, Tunisia, and Argentina were attempting similar reforms. Often they were accompanied by parallel reforms in the overall health pol- Introducuon * 3 icy framework, provider payment system, and competitive market environment. The debates surrounding these reforms have been lengthy, heated, and rarely enriched by evidence gleaned from rigorous eval- uation of experiences. Much of this debate centered on whether in- dependent hospitals can play a positive role in a well-functioning health system. Polemics over this issue obscure the reality that in many industrial countries that have traditionally paid for health ser- vices through social insurance, inpatient services have always been provided through a mixture of public, semiautonomous parastatal, nongovernmental, and private hospitals. Since the hospital sectors in all the European countries that use such mixed delivery systems are part of socially responsible health care systems, it is clear that independent hospitals can indeed play a productive role in a well-functioning health system. Hence, the end- point of these reforms is not really at issue. What is at issue is, when does it make sense to move from an integrated public system to a sys- tem with independent hospitals, and how should it be done? What kinds of changes and improvements can these reforms bring? How can policymakers assure that these improvements will actually be forthcoming? How can they move from a rigid, integrated delivery system, with hierarchical control of hospitals, to a better performing system that relies on indirect mechanisms to guide substantially more independent service providers? The research presented in this volume was motivated by a desire to assemble available information to try to answer this question. Scope of the Volume This volume examines marketizing organizational reforms, that is, reforms that relied on the combination of increased independence, and market-based performance pressures. WVhile privatization is such a reform, we will not examine that reform in this volume, concen- trating instead on autonomization and corporatization reforms, which 4 o Innovations in Health Service Delivery keep the hospitals in the public sector. This volume includes cases from both industrial and developing countries. A core concern of these reforms is hospital governance, and this volume looks closely at this issue. For the sake of clarity and comparability, we have cho- sen to focus on the governance issues between the relevant ministry or ministry officials and the hospital. In his thesis, Klaus Pugner refers to this as the level of "secondary governance." Hence, we will not be examining in any detail the political governance issues related to how the representative bodies (parliaments) are holding ministries accountable for health sector and hospital performance. Neither will we look into issues related to governance within the hospital (how managers are holding departments accountable). Need for More Research and Analysis This volume is a first attempt to conduct a systematic, if subjective, review of autonomization and corporatization of public hospitals. It should not be considered a definitive statement on this subject but an opportunity to crystallize key questions about objectives, design, im- plementation, and evaluation of such reforms. It highlights several important areas for further investigation: o The institutional and contextual requirements for and constraints to "marketizing" organizational reforms (e.g., what works at dif- ferent income levels, stages of health systems development, cul- tural settings, market environments) o A direct comparison of autonomized, corporatized, and privatized units to see which reform creates a more workable hospital system in different contexts o Policy options for reforming public hospitals in situations of ex- treme government failure (is there any evidence that improved management of integrated hierarchical systems do better in this context than if governments were to introduce organizational re- forms, complex as they may be?) Introduction * 5 * The nature of the parallel reforms in resource governance, re- source allocation/purchasing arrangements, and market environ- ment that are needed for successful reform * Ways to achieve more rigorous and ongoing monitoring and eval- uation of the reforms to ensure that policymakers will use the les- sons learned and that these will be available to countries that have not yet ventured down the organizational reform path. Research the World Bank is now undertaking in these areas will be published in companion volumes to this first volume on Innovations in Health Service Delivery: The Corporatization of Putblic Hospitals. Roadmap for the Volume A roadmap of this volume follows. Part 1 provides a conceptual framework for understanding organizational reforms, their design, implementation challenges, and monitoring and evaluation of impact based on a global review of the literature. Part 2 provides a cross- cutting analysis of empirical data from a global review of selected countries and two regional assessments of marketizing reforms in the hospital sector-Eastern Europe and Latin America. Part 3 presents the results from eight case studies. Part 1-Conceptual Framework By reviewing the literature on institutional economics and organiza- tional theory in its application to a wide range of sectors, chapter 1 identifies five areas in the incentive regime of hospitals and three critical factors in the external environment that need to be carefully coordinated during organizational reforms. The cross-country analysis and case studies show that countries encounter problems when parts of the hospital's incentive regime-such as decision rights, market exposure, residual claimant status, accountability arrangements, and explicit policies and reimbursement of social functions-are unbalanced or "out of sync" with each other. 6 a Innovations m Health Servce Delivery For example, even with extensive decision rights over areas such as strategic planning, financial management, and procurement, hos- pital managers will find that efficiency and productivity improve- ments elude them if they do not also have control over labor. Many countries yielded to the demands of powerful stakeholders and did not transfer control over labor to the reformed hospitals. But even consistency among the changes in the hospital's incen- tive regime under these five categories is often not enough. Success depends equally on underpinning these reforms with broad changes in the external policy environment to ensure the hospital's payment system and market environment work to promote improvements. For example, hospital reforms designed to take advantage of effi- ciency gains prompted by competition will not work if the factor markets (e.g., pharmaceuticals, labor, or custodial services) or prod- uct markets (e.g., hospital services) are allowed to manifest monopo- listic behavior. Likewise, parallel changes in funding arrangements and the provider payment system are often a particularly critical ele- ment of marketizing organizational reforms. Even with extensive changes in decision rights and accountability arrangements, the behavior of hospital managers is unlikely to change significantly if hospital funding continues to rely on histori- cal patterns with soft budget constraints. And reforms that create hard budget constraints are likely to impair equity unless parallel fi- nancing reforms ensure that the subsidies for poor patients are in line with the unit cost of their treatment. Chapter 2 highlights the importance of the political economy, context, and process dimensions of reforms like the autonomization and corporatization of hospitals. Even a well-designed reform will fail without the political consensus to implement it, or if strong vested stakeholders such as the medical profession or labor unions are not brought on board, or if the political cycle is too short and subsequent governments reverse or dilute the reform policies. Com- promises on labor reforms and political interference with decision rights and accountability arrangements were among the most dam- aging of such compromises observed in many countries that at- tempted but failed to achieve organizational reforms. Introduction * 7 Powerful medical groups that see their incomes or clinical auton- omy threatened by the reforms can easily frustrate implementation. Hospital managers that see their informal discretion over the hospi- tal reduced through new accountability mechanisms may equally re- sist the reforms. Wmning support from the medical profession and hospital management was therefore critical to the more successful reforms. Lack of such support contributed to failure. Similarly, vocal groups in society that have deep-rooted, "anti- market" value systems can poison the reform environment, especially in the case of complex reforms aimed at efficiency gains without eas- ily identifiable "short-term wins." Even without being elected, op- position parties that seize the opportunity to exploit such negative reactions can pressure policymakers into compromises that weaken the needed coherence and complementarities among the different reform elements. The context in which reform takes place can sometimes be as im- portant as the design of the reform itself. Many public sector reforms in the late 1980s and early 1990s, including the autonomization and corporatization of hospitals, were motivated by similar problems of poor performance from public institutions during fiscal crises. Yet the context for reforms in the health sector was often very different. The country context also varies greatly. For example, the environ- ment for private sector activities, especially in terms of contracting and regulatory enforcement, directly influences the feasibility of corporatization reforms. Hence, it is particularly important to exam- ine what is already going on in the private sector as an important predictor of what will be the outcome of hospital corporatization reforms. For example, country "X" might have a high level of informality, corruption, weak contract/company law, a low general "rule of law," and no organized public purchasing of services (often the case in low-income countries). Country "Y" might have a small informal sector, little corruption, strong contract/company law, a high "rule of law," organized public purchasing, and other "quasi-public" organi- zations already in existence. Even if both countries had the same pol- icy objectives and everything else was the same in the rest of their 8 o Innovations in Health Service Delivery health systems, the feasible and advisable design for each of them would have to be very different. The speed of reform and extent of changes are also significant. The "big bang" approach runs the risk of outstripping a country's implementation capacity. The "incremental approach" runs the risk of going off course over time. For example, if country "Y" had a strong central government, parallel public sector reforms, and sig- nificant institutional capacity in the health sector, it might be able to jump directly into hospital corporatization under a big-bang reform. Country "X" most likely could not. The chances of success in coun- try "X" might have been enhanced by a more incremental reforrn process. This might consist of first passing through a learning phase, using more limited hospital autonomy or piloting a limited number of hospitals. Provider payment reforms could slowly progress from funding based on inputs/historical resource use, to global bud- gets based on inputs, to partial funding/bonuses based on some per- formance indicators (benchmarking), to funding tied to outputs, to noncompetitive purchasing of outputs, to competitive/selective purchasing. Chapter 3 demonstrates just how difficult it is to monitor and evaluate the impact of reforms in a complex, multiproduct organiza- tion like a hospital. Surprisingly, none of the case-study countries de- veloped an evaluation strategy in advance. In the United Kingdom, a deliberate attempt was even made to avoid examining performance during the first phase of its introduction for fear that evidence of fail- ure early in the experimentation period might bolster opposition to the reform. In every case, the limited evaluation carried out later lacked a clear baseline and focused only on target hospitals, missing the opportunity for comparison, using nonreformed hospitals as a control. The limited evaluation that was possible highlighted several im- portant problem areas. First, reform objectives often were not clearly stated, making it difficult afterward to assess successes and failures in achieving the objectives. Second, the main reform levers-altering the incentive regime of the organization and the external environment-must lead to Introduction * 9 changes in the behavior of providers and patients before their net ef- fect shows up in final impact indicators in terms of health outcomes, efficiency, equity, and quality. Changes in the behavior of providers and patients provide proxies for impact. Third, without explicitly anticipating some of the potentially neg- ative consequences of the reform (such as the financial burden of user fees on the poor) and introducing mitigating policies (such as subsidies or exemptions), the corporatization of public hospitals runs a high risk of being associated with some serious health, efficiency, equity, and quality trade-offs. Many of these negative consequences could be avoided by looking at the impact of the corporatization of hospitals on the overall structure of the health system and making compensating adjustments in its structure. Finally, many of the desired effects of organizational reforms- such as efficiency gains and improvement in consumer responsive- ness-are subtle and, hence, not easily perceived by the public. Lit- tle credit will be given for valid efficiency gains in managing hospital resources and activities if the public sees long waiting lists, decreased availability of drugs in dispensaries, and overworked and rude staff. In all the countries examined, too little attention was given to ensur- ing high-visibility "quick wins" that would increase public confi- dence in the reform process. Part 2-Cross-Cutting Analysis Part 2 of the volume includes a cross-cutting analysis of the global experience on organizational reforms and two regional cases studies (Central Europe and the former Soviet republics, and the Southern Cone of Latin America). Chapter 4 presents a cross-cutting analysis of the design and im- plementation of the experience of marketizing organizational re- forms based on data drawn from the two regional reviews and eight case studies. The first section assesses the hospital reforms according to their effectiveness in achieving stated objectives, how well they stayed on track, and their overall coherence. Against these criteria, the authors find that Singapore, Hong Kong, Malaysia, and Tunisia 10 ° Innovations in Health Service Delivery were "more successful." They find the United Kingdom refonn "partially successful;" New Zealand and Indonesia are assessed as "less successful." The authors then apply the Gill Walt political economy frame- work to characterize the common elements among each group of reformers. Clear patterns emerge in terms of the success in deal- ing with implementation challenges. The more successfully imple- mented reforms dealt more pro-actively with the powerfil stake- holders that are usually mobilized by these reforms, including public sector unions and professional associations. Refonmers in these countries also dealt more effectively with the significant political de- mands these complex reforms generated. Finally, the more success- ful reforms included efforts to address the changed, and increased, demands on hospital management. Given the dearth of serious monitoring and evaluation associated with marketizing organizational reforms, the authors are tentative in outlining a number of hypotheses explaining the success or lack thereof. Helpfully, they also outline some of the most successful "mitigating strategies" applied to address common problems con- fronted in implementing these reforms. Organizational changes in the hospital sector have been a com- mon component of health system reforms throughout Central Eu- rope and the former Soviet republics during the 1990s. Chapter 5 se- lectively reviews these refonms, focusing notably on the countries where reforms have progressed farthest. Health systems in Central Europe and the former Soviet republics have been slow at introducing marketizing organizational reforms in the hospital sector. This is puzzling since the theme of moving hos- pitals at arm's length from the core public bureaucracy and subject- ing them to market pressures is fully congruent with the economic and public sector reforms in transition economies which aim to reduce the role of the state and emphasize individual self-reliance. Instead, decentralization of ownership and output-based payment mechanisms predominated the reform agenda of service delivery in the region. As the authors in chapter 5 discuss, the changes did not "add up." Introduction * 11 The governance arrangements that emerged from hospital devo- lution themselves were inconsistent and problematic-with local governments struggling to deal with these, and many other, new roles. Some social insurance funds did take tentative steps toward active, output-based funding, but, owing to the weak governance arrangements, found the hospitals generally unresponsive. The authors conclude that most of these countries will find it nec- essary to revisit hospital governance in the near future, if desired per- formance improvements are be achieved. Hospital organizational reforms were also common in the South- ern Cone of South America during the 1990s. Chapter 6 reviews the reform experiences in Argentina, Chile, and Uruguay. In Argentina, the national government was committed to these re- forms, and it attempted to motivate the provinces (as hospital owners) throughout the country to "buy in" to the reform model. The Argen- tinean model was designed to implement autonomy at the level of the individual hospital. In contrast, Chile opted for a vertically integrated network model in autonomnizing its regional health administrations. Uruguay, after failing to introduce a nationwide reform, pushed for- ward with four pilot hospital autonomization programs. Although all three countries sought to implement modest auton- omization reforms in their hospital sector, they nevertheless in- tended to create indirect accountability mechanisms that are often associated with substantially greater hospital autonomy. In addition to creating boards in the reformed hospitals, all three countries tried to establish contracts with hospital managers in an attempt to en- hance management focus on performance. This regional case study underscores the implementation chal- lenges that almost always emerge during organizational reforms, es- pecially in countries where labor interests hold significant power. With a few exceptions in some Argentinean provinces, political and institutional problems substantially blocked reforms in all three countries. The relative success in Argentina illustrates the impor- tance of support during implementation at the hospital level where the required organizational changes are significant but where oppo- sition from vested interests is also the greatest. 12 o Innovations in Health Service Delivery Part 3-The Case Studies The United Kingdom introduced several reforms that decentralized and regionalized the hospital sector during the 1970s and 1980s. But it was not until the early 1990s, after a major organizational reform and privatization of public infrastructure and utilities, that the U.K. government decided to apply in earnest a marketizing organizational reform model and to introduce marketlike pressures on the hospital sector (chapter 7). Although the original corporatization model for the health sector was inspired by the managed care movement in California and initially piloted in Sweden, it was its introduction in the United Kingdom that sparked a global fascination with this type of reform. At the time that the United Kingdom introduced its reforms, hos- pitals functioned as virtual government departments. The hospitals that the government selected for the first wave of reform were in- tended to take on a radically different governance structure, includ- ing establishment of independent legal status (Trust), greater control over the employment and management of staff, and many other im- portant decision rights that had previously been centrally adminis- tered. Newly created hospital boards were to be modeled on com- mercial boards and to provide oversight of the Trusts' management and operation without day-to-day government intervention. And new purchasers were to generate performance pressures in their se- lective purchasing, hence exposing the Trusts to considerable market pressures. In practice, a number of these elements were left out during im- plementation, and the real forces for accountability continued to be exercised administratively through the Department of Health and the NHS Management Executive. Paradoxically, instead of increasing hospital autonomy, the reforms ended up increasing the influence of central authorities over the hospital sector. The U.K. reforms, there- fore, underscore the distinction between formal governance struc- tures and governance practices. VVhile it is clearly important to put in place organizational arrangements that support desired governance processes, this change is not sufficient in itself. Structures that are de- Introduction * 13 veloped to support enhanced autonomy can end up serving as a vehi- cle for more centralized administration. Beyond this mismatch be- tween governance practices and structures, a broad range of other factors that are discussed in the volume also contributed to the dis- appointing results associated with the U.K. reforms. In 1993, New Zealand became the second industrial country to implement a hospital organizational reform modeled on structural reforms originating in the state-owned enterprise sector (chapter 8). Similar to the enterprises that remained state-owned, the reforms were intended to expose hospitals to market pressures while keeping ultimate control in public hands. As in the case of the United Kingdom, the New Zealand hospitals were converted to legally independent entities (Crown Health En- terprises, CHEs), with associated changes in decision rights and accountability mechanisms. In contrast to the Umted Kingdom, however, New Zealand's public hospitals already had substantial day- to-day autonomy from the central government. As in the United Kingdom, in New Zealand the 1993 reforms increased instead of de- creasing the hospitals' reliance on direct accountability mechamsms applied by the central government. Despite this paradoxical nature of the reform model, and an im- plementation process that vacillated back and forth in response to political pressures, New Zealand witnessed improvement in some performance indicators (allocative efficiency, transparency about costs, and enhanced equity in access). On the whole, however, the re- forms were not viewed as successful. As a result, in 2000, a new gov- ernment substantially reversed the reforms. This lack of success is largely attributed to fundamental alterations made to the financial regime of the reform model as well as weaknesses in the implemen- tation process. The 1995 hospital reforms in the state of Victoria in Australia were driven by a desire to increase efficiency and the recognition that this would require substantial rationalization. This reform is pre- sented in chapter 9. Instead of having a government-driven rationalization plan, the reforms were designed to enable this process to occur in a decentral- 14 o Innovations in Health Servce Delivery ized manner. Thus, the reforms integrated groups of metropolitan hospitals (and subsidiary providers) into several networks, which could then compete with each other. As in the case of New Zealand, the hospitals in Victoria were al- ready fairly autonomous. Hence, the reforms did not focus on en- hancing autonomy but on introducing more corporate-like opera- tions at the network level. Many of the desired improvements that took place, including rationalization, resulted from the combined influence of the hospital reforms and a diagnostic-related group (DRG), performance-based provider payment system. The driving force behind the rationalization that took place appeared to be the organizational reforms that set up the network hospital structure rather than changes in any individual hospital's decision rights or other incentives. In 1991, Hong Kong policymakers believed the biggest problems in their hospitals related to rigidity and lack of management expertise. They designed their reforms to address these issues (chap- ter 10). New incentives were not a central element of the organi- zational reform introduced in Hong Kong, perhaps because the government apparatus generally performed well. The Hong Kong reform was not designed to rely on markets or marketlike pressures to enhance performance. Instead, policymakers created a single new corporatized Hospital Authority that was granted significant au- tonomy and enhanced administrative accountability arrangements. The reform integrated all public and publicly funded hospitals, constituting almost 90 percent of beds, into this newly created au- tonomous legal entity. The Hospital Authority was encouraged to undertake managerial and structural changes that would make it function like a corporation. The reforms gave the Hospital Author- ity a great deal of day-to-day freedom, relying on annual perfor- mance targets for accountability. The Hong Kong reform was relatively successful on a number of fronts, but of mixed success in improving quality. Consumer respon- siveness and queues in particular remain issues. Accountability relied almost entirely on the effectiveness of the performance measurement system, since there was no other source of performance pressure such as output-related payment, hierarchical control, or consumer Introduciion * 15 choice. Evidence to date shows that this system, though improving, still falls short in making the Hospital Authority truly accountable for performance. As yet there appears to be no penalty for failing to meet performance targets. In 1992, Malaysia reformed its newly built National Heart Insti- tute using a corporatization model that had been applied to other state-owned enterprises in that country (chapter 11). As in many other countries, difficulties arose during implementation, when the original design of the reform was scaled back in a number of areas. The resulting model had some elements that were more reminiscent of enhanced hospital autonomy than the more complete corporati- zation originally envisaged. But, since the reimbursement system was not designed to fund specific services or services for targeted indi- viduals, the reform went as far as it could, given this constraint, to- ward the establishment of marketlike incentives and performance pressures. Not surprisingly, as a result of the funding system, the Heart In- stitute did shift toward providing more services to private patients who could pay for their treatment. To make up for these structural problems, the Malaysian authorities made some provisions to deal with the resulting negative impact on equity. These provisions in- cluded: funding coverage for a portion of needy patients, ensuring that other hospitals continued to provide cardiac services to the poor, and mobilization of additional funding to cover losses associated with services to the poor. Initiated in 1985, the Singapore hospital reforms were the first to combine autonomy with reliance on market-based performance pressures (chapter 12). As in Australia, the Singaporean reform was implemented in a group of hospitals or "network," rather than in in- dividual hospitals. Unlike in Australia, however, the model did not envision competition among public entities, since the group of hos- pitals integrated into the network constituted most of the public hos- pital system. Singapore's experience is instructive in illustrating the strong re- liance of marketizing organizational reforms on a complementary fi- nancing system to create the needed incentives for productivity as well as accountability. In addition to its hospital reform, Singapore 16 ° Innovations in Health Service Delivery simultaneously undertook far-reaching reforms in its system for financing health care. This resulted in Singapore's unique system of Medical Savings Accounts that allows individuals to generate per- formance pressures on participating hospitals through consumer choice, while retaining protection against financial risk and con- straints on overall expenditure. In the early 1990s, Tunisia undertook a multifaceted hospital re- form of its 22 teaching hospitals (chapter 13). The reform is viewed largely as a success in the country and has proven to be sustainable. The Tunisian program is notable for pursuing changes on the tech- nical, managerial, and organizational fronts simultaneously. The organizational changes were, however, relatively modest, with the endpoint arrangements falling closer to a budgetary entity than an autonomous entity. As observed in some of the other case studies, the reform para- doxically increased the administrative influence of central authorities over hospitals. The Tunisian reform, therefore, once again under- scored the distinction between formal governance structures and ac- tual governance practices. While it is clearly important to put in place organizational arrangements that support desired governance processes-it is not sufficient. In Tunisia, as in the United Kingdom, structures developed to support enhanced autonomy, ended up serv- ing as a vehicle for centralization. Marketing organizational reform of hospitals is a complex and challenging means of addressing problems in the sector. Conse- quently, it is rarely done unless there is a strong driving force. Seri- ous fiscal problems are one of the most common motivating factors. The hospital reforms in Indonesia were clearly driven by the fiscal crises of the late 1980s and early 1990s, and the resulting desire to reallocate budgetary expenditure from hospitals to facilities deliver- ing ambulatory care (chapter 14). The government opted for a version of autonomization applied to individual hospitals rather than a full-blown corporatization model. Given the focus on reducing expenditure, it is somewhat surprising that the Indonesian reforms did not directly deal with labor man- agement. This constrained the eventual efficiency gains that were Introduction * 17 possible through the reform. Instead, reformed hospitals were en- couraged to earn more private revenue in an attempt to decrease the needed budgetary support. While there were indications of im- provements in both efficiency and quality, provisions to protect the poor against fees and cost increases were weak and often not imple- mented. In the final analysis, the reforms did not enable the govern- ment to decrease its hospital funding, and are therefore not viewed as a success. Building on previous efforts, the government that took office in August 1998 in Ecuador began to attack the most critical bottlenecks in the Ecuador health care system (chapter 15). First, constitutional changes opened a window of opportunity for pushing forward a modest reform agenda. This included: (a) a strengthening in the pol- icymaking and regulatory role of the central Ministry of Health; (b) deconcentration or delegation of administrative and financial func- tion to peripheral branches of the public bureaucracy; and (c) decen- tralization of a range of political, economic, administrative, and fi- nancial functions. Although not all the reforms envisaged were implemented be- cause of further changes in the government in January 2000, several changes in the incentive regime of the hospital system are now well under way. These include: (a) changes in the organizational and gov- ernance structure of public hospitals; (b) greater decision rights by hospital mangers over planning, financial management, cost recov- ery policies, financing capital investments (civil works, equipment maintenance), and some aspects of human resources management (training, performance incentives, and career development); (c) in- troduction/increase in some user fees; and (d) safeguards to protect poor households from the negative impact of user fees such as dis- counts and exemptions. Finally, participating hospitals now use "shadow prices" to gradually acquaint managers with an output- based financing system. It is too early to know the full impact of these reforms on the per- formance of the Ecuador hospital sector. As in other countries around the world, success will depend as much on the politics of im- plementation as it will on the techmcal soundness of the design. 18 a Innovations m Health Service Delivery A lFew Notewortliy Themes in Condlusion In conclusion, two major lessons learned are worth highlighting from the topical chapters, regional reviews, and case studies: o If reforms are too complex to fully design ex ante, go with the broad-brush or blueprint and be prepared to adapt as you go along-while maintaining overall policy coherence. o Organizational and marketizing reforms are systemic in nature- they cannot be introduced in the hospital sector alone without parallel reforms in other parts of the health system. The Broad-Brush Approach As would be expected, the case studies demonstrated that trial-and- error experimentation has marked the early generation of marketiz- ing organizational reforms of hospitals. There are both striking suc- cesses and dismal failures. Policymakers poorly understood many of the lessons learned through similar reforms in sectors such as trans- port, infrastructure, and telecommunication; other lessons did not directly apply to the health sector. For example, much has been learned about the need for special techniques like performance benchmarking and long-term contracts when public firms are transformed into entities that end up with a natural monopoly. This is especially true in the highly specialized hospital sector and in rural areas, where overheated competition among overlapping units may be undesirable since it can lead to a wasteful duplication of capacity and an expensive "medical arms race" for the latest technology. Yet lessons from other sectors in this regard do not apply directly to a multiproduct organization such as the hospital, where output and health-outcome performance indicators are much more difficult to define and monitor than, for example, kilowatt consumption in the energy sector. Much more refined instruments are needed to guide the behavior of substantially independent hospitals, such as case-mix adjusted payments that are data-intensive and require so- Introduction * 19 phisticated patient records and accounting systems. Much new learn- ing has, therefore, had to take place as the principles of marketizing organizational reforms were applied to the health sector. Parallel Reforms This volume also shows that the organizational reform of hospitals is a multidimensional reform that requires coherent changes in a num- ber of critical factors, not just appointing a management board or placing the hospital within the remit of company law. Reforms in- troduced in isolation in the hospital sector almost always engendered an incoherent policy framework and had many adverse effects on other parts of the health sector. Alexander S. Preker April Harding Chief Economist Senior Economist Health, Nutrition, and Population Health, Nutrition, and Population Human Development Department Human Development Department World Bank World Bank Part 1 Overview of Marketizing Organizational Reforms CHAPTER 1 A Conceptual Framework for the Organizational Reforms of Hospitals April Harding and Alexander S. Preker Throughout the world, governments are reassessing their role in health service delivery. They are being forced to do so by growing pressures, including cost escalation and increasing user dissatisfac- tion with services. In public hospital systems, the problems are typi- cal problems associated with publicly run services: inefficiency, both technical and allocative; low productivity; unresponsiveness to users (patients); waste; and, sometimes, fraud and corruption. Excessive in- fluence or even domination of health services by provider organiza- tion and health workers is often recognized as an important part of the problem. With increasing frequency, autonomization and corporatization are being considered and applied to improve performance of publicly run health services, similar to recent innovations in organizational reform elsewhere in the public sector. So far, however, little of this recent experience has been pulled to- gether in a way that is readily accessible to policymakers and sector experts. This volume attempts to do so. To support this review, we outline the relevant theoretical literature and experiences with orga- nizational reforms both within and outside the health sector. We hope this volume will help practitioners understand: 23 24 o Innovations in Health Service Delivery o What the reforms consist of o Which problems they attempt to address o Why they are structured the way they are o Why their designers think they will resolve certain problems o What lessons can be gleaned from these experiences. We use the term organizational to distinguish reforms that alter the structure of hospitals and their relations with other parts of the system from reforms that try to improve hospital performance within the existing parameters of the system (box 1.1). There are three types of organizational reforms that share a common element of "marketization." These are autonomization, corporatization, and privatization. All three involve reducing direct government control over the hospitals in question and exposing them more to the mar- ket or marketlike incentives.1 This volume covers only autonomiza- tion and corporatization, which apply these marketlike incentives while maintaining public ownership. Privatization will be reviewed in a separate volume. The Need for Reform of Health Servoces During the past 50 years, many low- and middle-income countries have established publicly funded health care systems, with services produced by a vertically integrated bureaucracy in the public sector. These systems were structured this way primarily as a response to market failures and inspired by Western systems such as the New Zealand and British National Health Services. Often with the help of donors, health sector policies focused on expanding the underlying human resources and physical infrastruc- ture (clinics, diagnostic facilities, laboratories, and hospitals). Sys- tems were developed to supply drugs and medical equipment and to train staff. Worldwide, the number of hospital beds rose between 1960 and 1990 from 5 million to 17 million, more than doubling A Conceptual Framework for the Organizational Reforms of Hospitals o25 Box 1.1 Definitions: The Players, the Rules, and the Objects of the-Game The key distinctions between organizations, institutions, and interventions parallel those between the players, the rules, and the objects of a game. Organizations are the players-the way the people are structured or organized (e.g., hospitals, clinics, pharmacies, and public health programs). Institutions are the rules (formal and infonnal customs) of the game or activities-the hunanly devised and socially shared con- straints that shape human interaction and the mechanisms by which these rules are enforced. Interventions are the objects of the game or activities (e.g., clinical interventions, public health interventions, and intersectoral action). In common usage, the terms organization and institution are often used in-terchangeably. Physical facilities such as hospitals and insurance agencies are often called institutions, and management capacity is often referred to as institutional capacity. For example, management training, information systems, and changes in human resources policy are often described as ways to strengthen institutional capacity, even though none of these measures addresses the underlying problems in rules and informal customs (the institutional problems) that may be at the root of dysfunctional behavior of service providers. To avoid misunderstanding, this volume refers to rules, rather than using the term institution. Governance is the relationship between the organization and its owner(s). Ownership (public or private) provides the right to make decisions on the use of an asset and the right to the income that remains after all fixed obligations are met. Sources: S. E. Berryman, N. Boyle, F. Golladay, M. Holmes, P. Keefer, and K. Sigrist, "Assessing Institutional Capabili- ties," Draft Working Papers, World Bank, Washington, D.C., 1997; World Health Organization, World Health Report 2000: Peifor-mance of Health Systems (Geneva: WHO, 2000). 26 o Innovations in Health Service Delivery supply per head. Parallel to this development, the number of doctors increased more than fivefold, from 1.2 million to 6.2 million.2 These input-focused strategies have contributed to many suc- cesses, including improved equity and access to health care for mil- lions of people as well as the control of communicable diseases and other public health activities that respond well to direct government involvement. Despite these impressive accomplishments, however, increasingly serious problems are apparent in publicly operated health services. Many of these problems originate in the delivery sys- tem, and many of them have parallels in infrastructure and other seg- ments of the public sector. Publicly delivered health services, like other public services, are plagued by critical problems, among them technical inefficiency. Re- sources within facilities are used poorly, often very poorly.3 At the systemic level, allocative efficiency is a severe problem-with re- sources often flowing disproportionately to urban, curative, and hospital-based care.4 The public sector is often lax in tracking the cost of services. In the health sector, such inattention to costs reduces the ability to identify and deliver cost-effective services. Although equity is a key rationale for public delivery, distribution of resources in public systems is rarely targeted toward the people who need them most.5 Social services delivered by public providers are notoriously unresponsive and unaccountable to users. Stories abound of poor staff treatment of patients in government health fa- cilities. Quality is often questionable-both clinical and consumer quality. Equipment is often faulty or broken. Changing Views on tie Role of the State "The world is changing, and with it our ideas about the state's role in economic and social development."6 Reforms in the organization of health service delivery are indicative of fundamental changes in views about the appropriate role of the state in the economy. Many factors have contributed to a better understanding of the shortcomings of an overextended state role in the economy: the collapse of centrally A Conceptual Framework for the Organizational Reforms of Hospitals * 27 planned economies; the fiscal crises in welfare states of the advanced industrial economies; and the Asian crisis-calling into question the "miracle" of sustainable state-led growth of the East Asian "tigers." In developing countries, overextended governments try to do too much with too few resources and little capability. Concurrently, they often fail to ensure provision of the most fundamental social goods such as basic health and education, property rights, and roads. The growing consensus is that the path to greater state effectiveness and rapid development lies in matching the government's role to its ca- pabilities, while making better use of the private sector in areas where the state has no comparative advantage.7 Government divestiture of commercial activities has yielded the first and easiest gains on the road to focused and effective govern- ment. Cement production, for instance, is an activity in which state production has not been found to provide any advantage over the pri- vate sector. The widespread success of this "first wave" of organiza- tional reforms showed how much a society could gain just by getting government out of the business of producing commercially viable goods and services. Although many governments still hold back their countries' growth through involvement in producing commercial goods, no serious evidence or analysis supports this policy. Building on successful transformation of commercial companies, governments throughout the world have begun to apply these re- forms to their public utility services in a "second wave" of reform. Redefining the state's role in delivering infrastructure services has been a more difficult path to navigate. Although the state definitely has some role to play here, prevailing wisdom on what that role is has changed. In these sectors, the long-held view was that because of a "natural monopoly," these services had to stay in public hands-to capture scale economies and curtail exploitation of monopoly power. Tech- nological change and institutional innovation have made it possible to diversify production and service-delivery arrangements. This has led to huge improvements in efficiency, quality, and responsiveness of services, not to mention lightening fiscal burdens previously asso- ciated with operating these services as public sector monopolies.8 28 o Innovations in Health Service Delivery These experiences in privatizing or commercializing (usually through corporatization) the organization of infrastructure services outside the social sectors have revealed the previously large hidden costs of public sector monopolies in these areas of the economy.9 At this point, few will argue in favor of government operation of infra- structure services or other commercially viable areas of the economy. Now a "third wave" of reforms is in evidence as many countries experiment with applying marketizing reforms to social services (health, education, and pensions).'0 In the health sector, policymak- ers are struggling to apply and amend these reforms to address the many problems in publicly delivered health services, while pursuing social protection and equity. In addition to being influenced by re- forms outside the sector, they are also supported by developments on the analytical front. Theoretical advances of many types-from neoclassical econom- ics to the economics of organizations-have contributed to the new perspectives on the organization of service delivery. Neoclassical Economics The neoclassical paradigm lays out the potential sources of market failure. The rationale for public ownership has been its effectiveness as a tool for pursuing social objectives in the presence of market fail- ures. This belief is based on a simple view of the relation between ownership and control. Privately owned companies are generally conceived of as profit maximizers-since by maximizing profits they maximize benefits to their shareholders (or owners). Sometimes maximizing shareholder benefits is not seen as maximizing benefits to society as a whole. Broadly speaking, shareholders' and society's interests can clash when competitive solutions do not exist (natural monopoly) or when they exist but are not efficient (because of externalities, the nature of public goods, or asymmetry in the information available to each party). In health, this reasoning supports public intervention to ad- dress market failures with regard to both equity and efficiency. Based on these insights, public ownership has been used as a tool to get the A Conceptual Framework for the Organizational Reforms of Hospitals * 29 organization to replace the narrow interests of owners with the wider interests embodied in the state, to pursue social goals as assiduously as private benefit. Economics of Organizations Although neoclassical economics delineates potential sources of mar- ket failure, the theory is silent on the critical issues of structuring in- stitutional solutions. The mechanism by which public ownership is supposed to maximize social goals is nowhere precisely defined. Neoclassical economics is essentially "institution free."' 1 Recently, this vacuum has been filled by new analytical tools for understanding the impact of different ownership and governance arrangements. Much progress has been made in identifying the key factors caus- ing wide variations in organizational performance. The most rele- vant developments for organizational reform in health come from principal-agent theory, transaction cost economics, and property rights and public choice theory. These fields are often grouped to- gether under the title "economics of organizations"-and all deal with considerations of information, motivation, innovation, and the implications for how productive activity can best be organized. The traditional rationale for public ownership was based on a sim- plistic model of individual behavior-presupposing that the objec- tives of the government and of the managers of public organizations were identical. Policymakers assumed that managers told to pursue the public interest would be able to determine what that meant and would have the incentives to do it. In practice, the vagueness of the objectives and the difficulty of precisely determining and monitoring output has proved "inimical to the efficient management of the [sectors] concerned."'2 Public own- ership removed the opportunistic profit maximizer-but civil ser- vants or politicians have turned out to be no "high custodian of pub- lic interest." Instead, they have often pursued their own private benefits. 13 The economics of organization has shed light on the issue of how best to structure organizations, assuming that individuals, at least in part, are pursuing their own self-interest. This analysis added 30 o Innovations in Health Service Delivery to the general understanding of the sources of the problems that occur with incentives in publicly run services. Agency Theory Agency theory highlights the need to reconcile divergent interests among individuals under conditions of widespread uncertainty and uneven access to information. The key relationship, as modeled, oc- curs between a principal and an agent. The principal needs the efforts and expertise of the agent but has only limited ability to mon- itor the agent's actions or evaluate whether the final outcome is satisfactory. The agency literature surveys the range of contracts such as pay- ment and monitoring arrangements observed in the economy as attempts to align incentives and to reward cooperation between self- interested but interdependent individuals.'4 The need to align in- centives pervades the health sector: the relationship between patient and physician is a classic case of the principal-agent structure. Physi- cians and hospital managers also have divergent interests and differ- ent competencies, yet they need one another. Most important for our review is the principal-agent relationship between the government owner and hospital management. Governments, like firms, must design evaluation and reward mech- anisms to obtain high-quality performance regardless of whether they contract with employees or with outside providers/suppliers. Several studies have generalized the agency insight from the employment context to the full range of relationships that make up the firm-now conceptualized as a nexus of many contracts.15 This conceptualiza- tion has increased the understanding of ownership and governance by clarifying the relationship between the firm's managers and sup- pliers of capital, equity shareholders as well as bond debt holders. By illuminating critical elements of relations between owners and firms, this analysis has also improved understanding of the governance re- lations between governments and public service providers. The rise and predominance of the modern corporation is attrib- uted to its successful governance structure. This structure allows A Conceptual Framework for the Organizational Reforms of Hospitals * 31 professional managers to be assigned decision rights and performance incentives, although they bear little financial risk. Risk is borne by di- versified investors, who need not assume control. This analytical framework helped identify and explain some of the performance differences of organizations with different governance structures. These insights, about what governance mechanisms work in which situations, have enabled governance improvements, both in public and private organizations. Progress was made in particular in understanding the impact of monitoring and accountability mecha- nisms on incentives. Transaction Cost Economics Transaction cost economics (TCE) has focused attention on the dis- tinctive features of activities organized within an organization, versus those organized through market interactions. TCE looks at questions such as: Why does a company buy some inputs, rather than producing them in-house? Why does the direc- tor of a departnent decide to hfre someone to undertake certain tasks, rather than buying the services from a company or individual? In answering these questions, TCE scholars have identified the gen- eral advantages of internal organization. They have found that inter- nal organization is used when the exchange or activity is simply too comphcated to contract for. Because of the cognitive limits of eco- nomic agents, their willingness to pursue self-interest, and the un- foreseeable changes in the environment, every contract, even the most detailed, is inherently incomplete. None can fully anticipate and accommodate the differing interests of the negotiating parties. For certain types of activities, contracts won't be sufficient, rather the interaction needs to be supported by organizational means. Integrat- ing activities inside a single organization enables better responses to unforeseen events and adjudication of the problems they create. TCE analysis sheds the most light on firm boundaries and the con- ditions under which activities are best arranged within a hierarchy instead of through interactions in a market with suppliers or other contractors. 32 o Innovations in Health Service Delivery As noted above, vertically integrated organizations arise when activities can't be organized through market contracting. Vertical in- tegration permits the details of future relations between suppliers (including employees), producers, and distributors to remain un- specified. Differences can be adjudicated as events unfold. Vertical integration (or unified ownership) pools the risks and rewards of the organization's activities and can facilitate information sharing, the pursuit of innovation, and a culture of cooperation. Despite these positive features, vertical integration suffers from characteristic weaknesses as a mechanism of governance. Notably, it substitutes low-powered incentives, like salaried employment, for the markets' high-powered incentives of profit and loss-thus re- ducing dramatically incentives for productivity. Incentives weaken as people capture less and less of the gains of their own efforts when re- wards and losses are spread throughout the organization. Another often-observed problem with vertical integration is the proliferation of influence activities (box 1.2). Despite its focus on the contracting problems that motivate internal organization, transaction cost eco- nomics views vertical integration as the governance mechanism of last resort. In most contexts, contracts, contractual networks, virtual integration, franchising, or establishing concessions are observed to outperform unified ownership arrangements. Governance arrangements are evaluated by comparing the pat- terns of costs generated for planning, adapting, and monitoring pro- duction and exchange.16 Unlike public organizations, private firms have the flexibility (indeed the requirement) to adjust their gover- nance structure to changes in the market environment-making them good targets for identifying "better practices" for governance arrangements. The TCE insights about the problems with internal organization have been particularly influential in analysis of public sector opera- tion. Performance problems within large firms are similar in many ways to those seen in the public sector. Recent public sector reforms have been strongly influenced by the TCE literature. The emphasis in the 1980s in many OECD countries on contracting out services, for example, was motivated at least in part by the better understand- A Conceptual Framework for the Orgamzational Reforms of Hospitals 0 33 Box 1.2 Influence Activities An important issue related to moral hazard and the structure of organizations is influence activities and the associated costs, known as influence costs.a Recent analysis has shed much light on the propensity of publicly owned service delivery organi- zations to capture inordinate portions of the sector budget, as well as on their ability to influence sector policy to their benefit-often at the expense of public interest. In the health sector, provider organizations expend effort to affect decisions regarding the distribution of resources or other benefits among providers to their advantage. These influence activities occur in all organizations, but counter- vailing forces are particularly weak in public service delivery systems-and influence costs are one of the most significant costs of centralized control. Evidence of such influence ac- tivities is seen in public utilities where monopolies are often maintained to protect low-productivity, state-owned enter- prises from competition from more efficient producers. In the health sector, the tendency to allocate resources to ter- tiary and curative care at the expense of primary, preventa- tive, and public health is likewise attributed to "capture." The costs of these activities include both the losses asso- ciated with poor resource allocation decisions and the loss associated with efforts to capture the rents. These costs can be reduced when no decisionmaker has the authority to make decisions that service providers can easily influence. This condition can sometimes be brought about by creating legal or other boundaries between the policymaker, the fun- der, and the service provider unit. Many organizational re- forms have attempted to diminish these activities. Examples include privatization of utilities as well as reforms separating the policymaker from the payer and from the provider in public service delivery. a. Milgrom and Roberts, "Bargaining Costs, Influence Costs, and the Organization of Economic Activity." 34 0 Innovations in Health Service Delivery ing of the incentive problems associated with operating these ser- vices in-house. TCE analysis has also shed light on the structural roots of the problems with internally organized (publicly operated) health services, laying the groundwork for the recent wave of orga- nizational reforms in the health sector. Property Rights Theory Property rights theory looks at the same incentive issues from a slightly different perspective. Since private ownership appears to have strong positive incentives for efficiency, property rights theo- rists have attempted to find out why. Explanations have focused on two issues: the possession of residual decision rights and the alloca- tion of residual returns.17 Residual rights of control are the rights to make any decisions r e- garding an asset's use that are not explicitly contracted by law or as- signed to another by contract. An asset owner usually holds these rights-although the owner or the law may allocate many rights to others.18 The notion of ownership as residual control is relatively clear for a simple asset like a car. It becomes more complicated when applied to an organization such as a firm. Large organizations bun- dle together many assets, leading frequently to ambiguous decision rights. For example, do a firm's directors have the right to accept a takeover offer without soliciting competing bids? In addition to residual decision rights, an owner holds the rights to residual revenues from assets. That means the owner has the right to whatever revenue remains after all funds have been collected and all debts, expenses, and other contractual obligations have been paid. Just as the allocation of residual control can be vague and ambiguous in the case of firms (because rights of control over different kinds of decisions may be poorly specified or may lie with various parties), so can the notion of residual returns be unclear. One problem is that recipients of residual returns may vary with the circumstances (box 1.3). A firm that cannot pay its debts may have to pay lenders an increasing share of its earnings. The lenders become residual claimant. Firms may pay bonuses, increase workers' A Conceptual Framework for the Organizational Reforms of Hospitals o 35 Box 1.3 High-Powered Ownership Incentives Suppose a transaction involves several people supplying la- bor, materials, and other physical inputs. If all but one per- son have contracted to receive fixed amounts, there is only one residual claimant. In that case, maximizing the value re- ceived by the residual claimant is the same as maximizing the total value received by all parties. If the residual claimant also has residual control, just by pursuing his own interests and maximizing his own returns the claimant will be led to make efficient decisions. The combination of residual con- trol and residual claims provides strong incentives and ca- pacity for an owner to maintain and increase an asset's value. Firms often attempt to reproduce these high-powered in- centives by allocating residual claims in the form of bonuses or shares to key decisionmakers in their firm.a Misalignment of residual rights and returns causes serious problems. The residual claimant to the returns from state- owned enterprises is the public pLLrse, but the residual deci- sionmakers are the enterprise manager, the workers, and the bureaucrats in the supervising ministry. None of these has any great personal stake in the value of the enterprise. The resulting low productivity is well documented. Another ex- ample of misalignment comes from the U.S. savings and loan (S&L) industry. The individuals who had the right to control the S&L's investment also had the right to keep a portion of any profits earned but were not obligated to make good on losses. That combination of rights and obli- gations created an incentive for risk taking and fraud that was not effectively countered by other devices during most of the 1980s'b These fields of analysis have improved understanding of the institutional sources of government failure. The frame- work has been used to design organizational reforms that (Box continues on the following page) 36 o Innovations in Health Service Delivery Bt 1] .3 (conNinued) seek to allocate to the holders of critical information the au- thority to make relevant decisions and the financial incen- tive to do so (in the form of residual claims on the outcome of the decision). a. Milgrom and Roberts, Economics of Organization and Man- agement, p. 291. b. Ibid., p. 292. pay, and promote more workers into higher ranking, higher paying jobs when performance is up. Thus, some of the workers share in the firm's residual returns. The pairing of residual returns and residual control creates most of the powerful incentive effects of ownership. These effects are very powerful because decisionmakers bear the full financial impact of their choices. Political Choice Theory Political choice theory has strongly influenced organizational re- forms. A central tenet of public choice is that all human behavior is dominated by self-interest. Individuals are viewed as rational utility maximizers. Public choice theorists apply this model to understand how individuals will react to a range of institutional settings and their intrinsic incentive structures. They also study collective action prob- lems, problems that arise when the pursuit of individual interests produces suboptimal outcomes for the collectivity. This field focuses on the self-interested behavior of politicians, interest groups, and bureaucrats and studies the implications for ef- fective government and the scope of government. Bureaucrats, at- tempting to maximnize their budgets, will acquire an increasing share of national income. As a result, the state will grow well beyond the size needed to deliver core functions. Much has been learned from this work about how powerful interest groups are able to capture re- A Conceptual Framework for the Organizational Reforms of Hospitals * 37 sources. This analysis has also shed light on the nature of institu- tional rigidities that reduce economic growth.19 Because of the focus of this work on opportunism and public administration, many public choice theorists support a conservative political agenda (minimizing the role of the state). However, this field has also generated insights on ways of structuring governance and administration in organiza- tions to minimize opportunities for this destructive behavior. These insights on incentives, contracting, and governance have influenced recent reforms in health service delivery. Options for Reforming Health Care Delivery Systems Marketizing organizational reforms such as autonomization and cor- poratization are usually initiated to address problems that publicly run health services have with efficiency, productivity, quality, and client responsiveness.20 However, these reforms are not the only methods used to solve them. Management or technological reforms as well as reforms in funding or payment arrangements are also com- monly used. Technological Reforms Perhaps the most common reforms undertaken to improve hospital performance are focused on enhancing the technological capacity of the hospitals. The prevalence of broken medical equipment and primitive or nonexistent management information systems (MIS) makes this an obvious place to start an effort to improve the opera- tion of public hospitals. The common public sector tendency to allow wages to crowd out capital costs makes such problems en- demic. Reforms usually consist of targeted investment in medical equipment, or the computers, databases, and software needed to op- erate an effective MIS. The skills required to operate the equipment or systems are often lacking. Therefore, technological reforms often include a capacity development component, supporting training for relevant staff. 38 o Innovations in Health Service Delivery These reforms have been very important in improving public hos- pitals in both developing and industrial countries. The rapid changes in medical equipment and information systems technology make these initiatives necessary. However, in many cases, these "techno- logical solutions" have not achieved the desired impact. Follow-up visits to the hospitals find that much of the equipment is not prop- erly maintained, and often not in operation. New MIS systems may be generating valuable information, but often neither clinicians nor managers are using it. In most cases, it is evident that these invest- ments need to be accompanied by additional attention to incentives. Management Reforms Many attempts to address problems in publicly run health care de- livery systems have been made through management reforms.21 These reforms have included efforts to strengthen the managerial expertise of health sector managers-both through staff training and through changes in recruitment policies to attract managerial skills.22 Commonly, these efforts are accompanied by improvements to in- formation systems to facilitate effective decisionmaking. Clinical directorates have been created in some systems, and benchmark- ing of departmental performance has been introduced.23 Many of these efforts are part of the growing trend toward apply- ing "best-practice" management techniques from private companies to reform public hospitals. Frequently, attempts are made to intro- duce business-process reengineering, patient-focused care, or quality- improvement techniques.24 However, implementation of these new management practices has been inhibited by the public sector con- text in which public provider organizations operate. Private organizations have introduced recruitment and compensa- tion policies, based on the best personnel management techniques for finding and motivating high performers. Civil service constraints have blocked or undermined attempts to apply these methods to public hospital systems. A critical barrier to applying best-practice principles from the private sector is the broad lack of control that public sector managers have over factors of production, especially A Conceptual Framework for the Organizational Reforms of Hospitals * 39 labor. Thus, although methods for reinvigorating private organi- zations have sometimes been successfully transferred to public hos- pitals and systems, more often than not, the common constraints generated by public sector control structures have frustrated these attempts.25 Indeed, attempts to apply private sector management principles to public delivery of health services have added momen- tum to the organizational reforms discussed below. Funding and Payment Reforms Reform of public hospitals' funding and payment arrangements is another common approach to solving problems of productivity, effi- ciency, and responsiveness. These payment reforms usually alter the structure of payments to tighten the link betwveen resource allocation and delivery of specific outputs. Retrospective fee-for-service, per diem, or case-based payments are examples of such changes. Some reforms try to encourage efficiency by shifting expenditure risk onto providers via capitated payments or prospective global budgets. Different structural changes are made in funding and payment systems to address concerms about clinical or consumer quality or re- sponsiveness to users. These payment reforms usually tighten the link between resource allocation and user or payer selection. Exam- ples include limited or fully competitive contracting with providers, fund holding with patient selection, and demand subsidies (health vouchers to be used with providers or insurers). None of these instruments is perfect. Each helps to achieve one goal at the expense of others. Systems that improve productivity en- courage supplier-induced demand. Systems that better contain costs usually encourage shirking and low productivity. The incentives cre- ated under each payment structure can be powerful and often create some degree of overshoot that must be addressed. Most systems are not fully understood, nor are measures to compensate for overshoot or known disadvantages. This often requires a mix of multiple pay- ment structures so that the positive incentives of one element of the payment counterbalance the negative features of another. An exam- ple is the frequent combination of capitation elements with fee-for- service in areas where productivity is especially important. 40 o Innovations in Health Service Delivery For payment system reforms to achieve their objectives, evidence strongly suggests that other reforms must also take place to encour- age or enable providers to respond to the new incentives. As dis- cussed below, organizational reforms are complements to payment reforms, not substitutes. Neither works on its own. A similar con- clusion applies to management reforms. Although management and funding reforms may be sorely needed to improve the performance of health care delivery systems, by themselves their results have been limited. They did not hit the roots of the problem: poor incentives inherent in the organization of public health service delivery. This realization has prompted the reforms reviewed in the remainder of this volume. The Nakwre of Markelizing Organizafiondl Reforms The growing awareness of the structural nature of public service de- livery problems has led policymakers in some countries to make or- ganizational reform a core component of health sector reform. These changes are designed to improve the incentive environment by altering the distribution of decision-making control, revenue rights-and hence risk-among participants in the health sector. Organizational reforms come in many different forms. Some focus on changing the mapping of functions across agencies, for instance, creating health insurance agencies that collect premiums and pur- chase health services. Others focus on endowing providers with fund-holding or purchasing authority, thus integrating funding with service provision. Decentralization is another common organiza- tional reform in the health sector, a reform that shifts decision- making control, and often revenue rights and responsibilities, from central to lower level government agencies. In this volume, we focus our attention on marketizing organiza- tional reforms, that is, organizational reforms that shift decision- making control to provider organizations and attempt to expose them to market or marketlike pressures to improve performance. These re- forms also attempt to create new incentives and accountability mech- A Conceptual Framework for the Organizational Reforms of Hospitals * 41 anisms to encourage management to use its autonomy to improve the facility's performance These reforms may be categorized under three headings: autonomization, corporatization, and privatization.26 Regarding terminology, unlike other sectors such as infrastruc- ture, all health reform modalities include continued funding, con- tracting, or purchasing by the government. Therefore, the three re- form modalities reviewed are often grouped together as separation of provider-payer reforms. In some cases, reform on the funding side of moving from budgeting to contracting is emphasized-hence the title contracting reforms may be used. Quasi-mnarkets, internal markets, and regulated competition are other terms used to describe these re- forms within the public sector. Traditional public hospitals and clinics operate as part of the inte- grated govermment structure, usually as a form of budgetary organi- zation (government department). The reforms applied to such or- ganizations vary in magnitude, depending on how far from public toward private the organization is moved (box 1.4). The structure of health sector reforms being discussed and imple- mented is strongly influenced by new public management, a set of principles for structuring public sector activities that has gained great currency in the industrial world, especially in the anglophone coun- tries.27 These reforms are also influenced by similar reforms in gov- ernment-run infrastructure companies and other public enterprises. Determinants of Hospital Behavior Organizational Structure Determines Hospital Behavior The model developed here encompasses the key structural elements of these marketizing organizational reforms. There are clearly other important structural features of hospital systems; however, this model captures the components that these reforms have most fre- quently been aiming to change (figure 1.1). While this model does overemphasize certain features of hospital governance and manage- ment, it enables us to disentangle the components of the reforms and to connect them to provider incentives, as well as to identify critical 42 o Innovations in Health Service Delivery Bon 1.4 fincenlive lEnvironmenls rom Public o £Privve One way to illustrate the differences between reform mo- dalities (autonomization, corporatization, privatization) is to view the possible options for structuring service delivery as a spectrum of incentive environments within which the tasks of government can be performed.a The civil or core public service lies at the center (usually constitutional con- trol bodies, line ministries), where the activities of the staff are highly determined. Job tenure is strong. Oncentive Envoonnments / McI kets/pi vote /secior - Broader public sector Core public sector B Budget units A Autonomous n C Corporatized units P Privotized units The broader public sector is distinguished by the rela- tive flexibility of its financial management regime and by the greater freedom allowed managers in recruitment and promotion. This may include special purpose agencies, autonomous agencies, and, on the outer limits, state-owned enterprises. Beyond the public sector lies the domain of the market and civil society. Services may be delivered by for- profit, nonprofit, or community organizations. The incen- tives for efficient production are higher moving outward, where service delivery is often better than at the core. A Conceptual Framework for the Organizational Reforms of Hospitals ° 43 Many refonrs throughout the world have sought to move service delivery away from the center of the circle to more arm's-length contracts with public and private organizations. However, there are constraints on moving delivery outward related to the nature of the outputs and the existence of mechanisms for public sector management of their delivery. Increased autonomy-moving from the center of the circle to the outer limits-requires accountability mechanisms not tied to direct control. Controls such as contracts take con- siderable capacity to write and enforce, especially for health services, where outputs and outcomes are difficult to specify. How far countries may go in pushing activities to incen- tive environments in the outer circles depends on the nature of the outputs (services) and the capacity to create account- ability for public objectives through indirect mechanisms such as regulation and contracting. a. Adapted from N. Manning, "Unbtundling the State: Au- tonomous Agencies and Service Delivery," World Bank, Washington, D.C., 1998, processed. interconnections. The reform modalities (autonomization, corpora- tization) are then characterized according to these elements. The literature reveals that three systemic features have typically been targeted in these reforms. These factors have been selected be- cause of their strong and direct influence on the incentive regime, and hence behavior, of publicly run health service providers: o Governance:28 the relationship between an owner(s), in our cases, government or government agencies, and the organization (health care providers) o Market environment: the level and nature of competitive pressures in the multiple markets within which the providers operate, in- cluding the full range of input and output markets 44 o Innovations in Health Service Delivery Figure 1.1 Key Determinants of Changes in Organizational Behavior Government E E ¢ / Governance 5 , E E / '., ~Hospital / ' : E i E \ ° E o g | Mair *Mketoxposure / J a _ o * Accountabiigyhrtrsaemernts ° . ~ X Explicitsocal functions f oFundmng arrangements: the structure of the flows from the funder payer to the provider, including the degree of formality, specificity, and so forth, regarding the related responsibilities or deliverables. There is no doubt that these three factors exert a powerful influ- ence on the behavior of hospitals as well as the management and staff within them. However, because of their complexity and connected- ness, it is often hard to tease apart the different sources of influence. Within our framework, the impact of changes in the governance and funding arrangements and the market environment is determined through the combined influence on five critical determinants of the incentives that the hospital faces. We refer to these elements as the hospital organizational structure: o Allocation of decision rights o Distribution of residual claims o Degree of market exposure A Conceptual Framework for the Orgamzational Reforms of Hospitals * 45 * Structure of accountability mechanisms * Provisions for socialfiunctions. These elements are further elaborated below. Decision rights. All the reforms discussed in this volume attempted to influence hospital and managerial behavior by expanding their auton- omy, or rights to make decisions. Thus, each reform can be charac- terized by the magnitude of control shifted from the hierarchy, or su- pervising agency, to the hospital. Critical decision rights transferred to management may include control over inputs, labor, scope of ac- tivities, financial management, clinical management and nonclinical administration, strategic management (formulation of institutional objectives), market strategy, sales, and the production process. Residual claims. The additional autonomy given to the hospitals and managers doesn't directly motivate them to use their added discre- tion productively. In the reforms discussed below, additional deci- sion-making control was usually complemented by giving managers and staff a material interest in resources. This interest was generated by allowing "leftover" resources to remain in the hospital, rather than turned over to the treasury or local government. As James Q. Wilson, a famous scholar of public sector management once asked, "Why scrimp and save if you cannot keep the results of your frugal- ity?"29 Therefore, a distinctive feature of the reforms is the degree to which the public purse ceases to be the residual claimant on revenue flows. The wisdom of linking these two factors is underscored by property rights theory, which emphasizes the importance of aligning revenue flows and decision rights appropriately to bring about the right decisions. Market exposure. What distinguishes autonomization and corporati- zation from other organizational reforms, such as decentralization, is their reliance to some degree on market or marketlike incentives. 46 o Innovatons in Health Service Delivery That is, all these reforrns moved hospitals toward at least some re- liance on earning revenue under market conditions instead of rely- ing only on budget allocation. Again, as we'll see, the means for doing so varied, but usually the reforms increased the importance of patient or other payer revenue based on choice. After the reforms, it was intended, if not actualized, that hospitals would have to work harder to deliver services that either patients, insurers, or other pay- ers would choose to use. To characterize this element of the change in organizational structure, it is best to identify the proportion of services the hospital delivers to customers with choice. The first two components of the reform model, decision rights and residual claims, compel managers to focus on financial viability. Thus, knowing which strategies generate the most revenue after the reform becomes critical. If improving service quality and attracting more patients is the best way to get revenue, that strategy will be pursued. If serving a larger number of healthier or otherwise low- cost patients is the easiest way to increase income, then these efforts will also figure into hospital behavior. If political lobbying or ex- tracting monopoly rents is the best way, these strategies will be pur- sued. In Part 3 of this volume, we see that in many countries, this component of the reform model went awry. Accountability. The reforms are also characterized by the degree to which accountability for achieving objectives is generated through hierarchical supervision of the organization versus rules and other indirect mechanisms such as regulations, contracting, or boards. As these reforms delegate some decision rights to the hospitals, the government's ability to assert direct accountability (through the hi- erarchy) diminishes. Thus, reform plans included a range of ac- countability mechanisms that would work in the new enviromnent. Most reforms relied as least partially on market pressures to create accountability, as markets were perceived to render a nonpolitical, nonarbitrary evaluation of performance, at least its economic per- formance.30 In some cases, where the capacity of government fund- ing authorities was higher, efforts were made to move toward pur- chasing with these funds. These countries intended to rely on this A Conceptual Framework for the Organizational Reforms of Hospitals * 47 purchasing (and the contracting and monitoring process) to generate accountability. As we know, both market failures and social values prevent health markets from delivering the full range of sectoral objectives. Thus, rules and regulations regarding the operation of these organizations constitute an additional form of accountability mechanism. Strength- ening these indirect mechanisms is a fourth critical element of orga- nizational reforms that reduce the use of traditional, hierarchical ac- countability mechanisms. Social functions. The final critical factor determining the impact of these reforms is what was done with regard to social functions. As au- tonomization and corporatization often compel hospital manage- ment to focus more on financial viability, management is likely to de- crease output of services that do not at least pay for themselves. The financial bottom line thus undermines the ability to cross-subsidize certain services internally.3' The marketizing organizational reforms discussed in this volume deal with this factor by creating alternative, more explicit mechanisms to ensure the continued delivery of such services. These mechanisms include explicit funding, demand-side subsidies, and regulation and development of insurance-changes that were often overlooked or not implemented in the cases described in this book. Complementary reforms that can protect nonefficiency objectives such as access for vulnerable groups and quality are further discussed below. External Environment The changes made to decision rights, residual claims, market expo- sure, and the like, characterize how things "feel" for the hospitals and management after the reforms. However, policymakers' discussions almost never take place at this rather academic level. The focus of these discussions, and the object of policymakers' actions, are three factors that are somewhat external to the hospitals: governance arrangements, funding arrangements, and the market environment. The changes in these arrangements undertaken in the reform take 48 o Innovations in Health Service Delivery place in the external environment. The policymakers' actions re- garding these elements of the external enviromnent determine what happens with regard to decision rights, residual claims, and so forth, which in turn determine the incentives for the hospital after the reforms. Reform debates often focus primarily on hospital governance arrangements. However, the governance structure alone does not determine the organizational structure or incentives. Rather, gover- nance arrangements combine with the other two features to fully determine the important elements of the hospitals' organizational structure after the reforms. For example, while governance changes may give the hospital the right to keep leftover revenue, you have to look at the funding arrangements to understand what incentives re- lated to revenue generation the hospital actually faces. Also, the gov- ernance changes may expose a hospital to the market for services, but you have to know whether there are any other hospitals that could provide the same services to know whether or not competition will emerge to exert pressure on the hospital. Hence the market exposure element of the organizational struc- ture captures these combined features by asking what portion of the hospitals' customers or payers have choice. To be a meaningful framework for understanding hospital incentives, components of the organizational structure must elaborate on the actual key conditions that influence hospital behavior, not just the formal provisions as recorded in the legislation. Governance. Structural changes in governance, the relationship be- tween the organization and its owner(s), influence the incentive regime in different ways. Governance arrangements vary substantially in the amount of autonomy given managers, the mechanisms used to generate new incentives, and the provisions for accountability. Funding or payment arrangements. The relation between payment arrangements and the new organizational structure directly influ- ences hospitals' incentives. The governance structure and the pay- ment system jointly determine three of the key determinants of A Conceptual Framework for the Organizational Reforms of Hospitals * 49 Figure 1.2 Putting It All Together Budgetary Autonomized7Corporotized Privatized uit Unit unt umt Decision Vertical a h n e rights hierarchyauno ,,, U Morket n Direct a exposure budgetallocaton o Residual Public a _- Er o o claimant purse D-w ° ° Account- Direct V R aos I -o obility hierarchical contral Social I Unspecifiedd functions and unfunded mandate i provider behavior: distribution of residual claims, provisions for so- cial functions, and market exposure (figure 1.2). Most organizational reforms endow the hospitals with formal claims to residual revenue in different categories, but the structure of the payment system direcdy determines whether this claim has any real meaning or incentive effect. If, for example, services are priced below cost, there will be no residual to claim. The relation of costs to the price-setting and capital-charging fonnula in the payment sys- tem thus becomes a critical determinant of incentives. The crucial factor is whether the provider's marginal cost-saving effort generates revenue flows that the provider can keep. Socialfunctions. As hospital managers start to cost out their activities, the payment system (sometimes combined with price setting or reg- ulation) will determine which services cover their costs. They will re- duce internal cross-subsidization where possible. If hospitals have been playing a substantial safety net role by generating funds from some services to cover costs of services delivered to the needy, the payment system will need to take this into account. The payment system will determine the degree to which unfunded mandates based on internal cross-subsidization become explicit and finded (box 1.5). 50 o Innovatons in Health Service Delivery Box 1.5 Inlerna3 Markeft versuxs Performance Budgehng: Wlhs tihe Difference? Two common funding arrangements used for autonomized/ corporatized providers are performance budgeting and pur- chasing in internal markets. These funding systems differ in three critical areas: specificity, comzpetition, and risk. Perforntance budgeting is a general agreement for funding in exchange for delivering certain services or products- where the funding level is tied to explicit performance re- sults and quality indicators (utilization, average length of stay, staffing ratios, infection rates). Purchasing via internal mzarkets usually entails much greater specificity on what is being purchased (rather than funded). Hence, revenues are tied to output more directly, although other performance indicators such as quality are also contracted for. In inter- nal markets, purchasers subject their choice of provider to competition. The specificity of output contracted for, combined with competition, reduces the provider's ability to directly pass through all costs and cost increases in their reimbursement. This enables greater shifting of risk for delivery cost to the provider. While it strengthens the incentive for efficient delivery, it also encourages cream skimming for low-cost patients. Market exposure. Market exposure is perhaps the most obvious be- havioral determinant of the reform model influenced by the funding arrangements. Funding arrangements influence market exposure, by determining the relative importance of different payers (govern- ment, insurers, individuals). Some of the reforms we are examining, for example, increased hospitals' reliance on private payers, some- times insurers, often out-of-pocket. Funding arrangements also in- fluence market exposure by determining the degree to which gov- A Conceptual Framework for the Organizatonal Reforms of Hospitals * 51 emnment payers act strategically to exert pressure through their pur- chasing and funding decisions. Market environment. The market environment most directly influ- ences market exposure. As noted above, governance and funding arrangements may be altered to compel a hospital to focus on gen- erating revenue through delivery of services. If the hospital is a mo- nopolist in its catchment area, however, it is not in any meaningful way exposed to a market. Thus, we have to look at the market envi- ronment to determine if a reform increased a hospital's exposure to a market. In some of these reforms, competition was intended to be one source of accountability. Hence, the market environment also partly determines whether this indirect accountability mechanism was actually established (figure 1.3). Figure 1.3 Forces Determining the Level of Competition Threat of I new entrants E t t rd Industry 41y Bargaining F ockeying for Bargaining power of position power of suppliers among current customers competitors v Threat of * subshtute A products or f services 52 o Innovations in Health Service Delivery Reform Mlodallies Hospitals' incentives are thus shaped by these five elements of the organizational structure after the reforms: decision rights, residual claims, market exposure, accountability, and social functions. How, then, do these features fit into different organizational reform mo- dalities: from budgetary organizations (the least autonomous and least exposed to the market), through autonomized and corporatized organizations, to privatized organizations (the most removed from government control and most exposed to market forces)? Budgetary Organizations Let us begin with a budgetary unit such as a hospital run as a gov- ernment department, where the hospital manager is essentially an administrator. The government's hierarchy of officials and rules con- trols all strategic issues and determines most day-to-day decisions re- lated to production and delivery of services-from staff mix, staff size, and salaries to services offered, technology used, and account- ing and financial management methods. Usually, the government determines the hospital's revenue through a direct budget allocation, based on historical norms. Other revenues are controlled as well, since the government also controls services rendered, patients served, and permissible copayments. Any "excess revenues" belong to the public sector-and must either be returned to a superior agency or spent as directed. Any "excess losses" are covered by the public purse. In this sense, the public sec- tor is the residual claimant of the hospital operating as a budgetary unit.32 The government's objectives in running the hospitals closely re- semble sector objectives and are often unrecorded and unmonitor- able. The social functions performed by the hospital are not distin- guished from its other activities-nor are they funded separately. Bureaucrats in the hierarchy are responsible for monitoring hospi- tal and managerial performance, usually tied to input and financial control. A Conceptual Framework for the Organizational Reforms of Hospitals * 53 Autonomized Organizations Dissatisfaction with budgetary organizations' weak performance has led to various approaches to reform. Many of the most serious effi- ciency and quality problems have been tracked to management's per- vasive lack of control over resources (especially labor) and produc- tion (service delivery). Autonomization focuses on "making managers manage"-by shifting much of the day-to-day decision-making con- trol from the hierarchy to management. Increased scope for generating revenue tied to service delivery often accompanies these changes. This may be achieved by moving toward funding via performance-related payments, by allowing pay- ing patients to be served, or by allowing copayments to be charged. Only if revenue can be retained do additional revenue opportunities motivate. Therefore, autonomization reforms increase an organiza- tion's scope for retaining revenue. Often, this is partially achieved by moving from a line item to a global budget, whereby savings in one service or budget area can be shifted to another. In this sense, the hospital or clinic becomes a partial residual claimant on certain sav- ings from cost cutting or other improvements. Accountability arrangements still generally come from hierarchi- cal supervision but with more clearly specified and narrowed objec- tives focused on economic and financial performance. An agreement between the government and hospital management may specify monitorable performance targets and responsibilities for performing social functions. A board may be created to exercise supervisory con- trol, thus mimicking private sector governance structures. Autononmization in the health sector has led to a wide variety of arrangements and differences in managerial autonomy. Most gov- ernments have been unwilling or unable to transfer control over labor, recruitment, salaries, staff mix, and the like and have instead left employees in the civil service, employed directly by the health ministry. In some cases, the organization has been legally established as a new form of government agency-which serves to define the new governance arrangements, secure the changes made, and per- suade management that the changes are irreversible. Accountability 54 e Innovations in Health Service Delivery arrangements have taken many forms-but all of them make some attempt to specify performance requirements in advance and to monitor their achievement.33 Sometimes these performance requirements are recorded in a framework agreement or performance contract. This mechanism is in- tended to narrow and clarify the organization's objectives as well as to lay out formal criteria for evaluating management. Occasionally, a board of directors is created to implement the process of monitoring managerial performance and to depoliticize decisionmaking. As mentioned, these reforms are often accompanied by a move to global budgeting or performance-related payments, which leaves some effi- ciency gains in the hospital. Corporatized Organizations Corporatization reforms have evolved from efforts to mimic the struc- ture and efficiency of private corporations while ensuring continued emphasis on social objectives through public ownership. Under corporatization, provisions for managerial autonomy are stronger than under autonomization, giving managers virtually com- plete control over all inputs and issues related to service delivery. The organization is often legally established as an independent entity, making the transfer of control more durable than under au- tonomization. A corporatized entity's status includes a hard budget constraint or financial bottom line-which makes the organization fully accountable for its financial performance. In case of insolvency, liquidation is at least theoretically possible. Management's greater latitude is complemented by market pressures as an important source of incentives, crucially including some element of competi- tion or contestability. These market incentives come from the combination of a hard budget constraint and increases in the retained portion of revenue from sales (instead of budget allocation) and discretion over its use. The corporatized hospital is thus much more a residual claimant than the autonomized one-in that it can retain excess revenues but must also bear any losses. Accountability mechanisms usually take A Conceptual Framework for the Organizauonal Reforms of Hospitals * 55 three forms: direct hierarchical control (or ownership accountabil- ity), funding/payment, and regulatory accountability. The ownership accountability that remains is usually manifested via a board. The hospital's responsibilities are usually narrowed to cover primarily economic targets-as part of the effort to mimic the effective gover- nance structures associated with private corporations (box 1.6). However, this emphasis on economic performance necessitates alternative arrangements for ensuring continued delivery of social functions (services previously cross-subsidized). Under corporatiza- tion, access to these services is usually pursued through purchasing, insurance regulation, demand-side financing, or mandates that apply to all organizations, not only to public facilities. In practice, when a hospital is corporatized, it is often established as a private corporation, though still publicly owned.34 The account- ability mechanisms are anchored in the board, and often include some form of corporate plan, a binding agreement between the hos- pital (and its board) and the supervisory agency. The plan contains fi- nancial performance targets such as profit or rate of return on assets or equity, dividends, and reinvestment policy. These targets usually require the hospital to earn sufficient commercial returns to justify long-term asset retention within the organization and to pay com- mercial dividends from those returns. The reliance on accountability from market pressures to earn rev- enue has necessitated the establishment of a functioning framework for direct payment or transfers to reimburse the hospital for the costs of pursuing noncommercial objectives. Instead of forcing the hos- pitals to deliver services below cost to the poor, for example, an ap- propriate subsidy may be delivered to either the patient or the hos- pital. Later we will explore some important issues associated with the complementary funding or purchasing reforms needed to support marketizing organizational reforms of hospitals. In a corporatized hospital, directors (board members) usually have absolute responsibility for the hospital's performance and are fully accountable to the responsible minister. They are sometimes re- sponsible for bringing the hospital's operation into conformity with world best practice (with necessary modifications for noncommercial 56 o Innovations in Health Service Delivery Box 1.6 Market [Environment Srucure aend CompaiNon The structure of the market to which the reformed hospi- tals are exposed vitally influences their behavior-as it di- rectly determines which revenue-generating strategies make sense. A central argument for exposing providers to market forces is that competitive forces in a functioning market lead to a more efficient allocation of resources than command economy or nonmarket solutions (figure 1.3). In a wide range of sectors, enhanced competition has led to increases in productivity, efficiency, quality, and innovation. In the health sector, however, market structure is prob- lematic in two respects. First, little or no competidon may emerge-reducing pressures on the provider to deliver "value for money" to maximize profits. Alternatively (or in addi- tion), competition may emerge, but it may be dysfunctional. Some health services, especially tertiary and quaternary, exhibit scale economies in production-which relieves in- cumbent hospitals of pressure from new entrants. Geo- graphic monopoly over certain services may leave buyers with little leverage to negotiate with service providers. Even for services where monopoly power is not an issue, providers may still capture market share or maximize prof- its through various forms of distortionary behavior. For ex- ample, medical treatment is to a large extent a "bundled" good where the seller (doctor) guides patients' consumption decisions-which hospital to choose for surgery, which lab for diagnostic services, and so on. Thus, the provider's in- formation advantage can be parleyed into control over a lu- crative referral chain. Doctors may "forward-integrate" into diagnostic labs or pharmacies and steer their patients to- ward consumption where the referring physician has a fi- nancial stake. Hospitals may "backward-integrate" by creat- ing strong links with doctors, thereby insulating themselves from competition for patients. Medical professionals are fre- A Conceptual Framework for the Organizational Reforms of Hospitals o 57 quently able to create cartels, limiting competitive pressures that strengthen the influence of patients and purchasers. Since patients and payers know less than providers about the true value or cost of health services, providers can creamn- skim, selecting healthy patients who cost less to treat than other patients. Thus, providers can increase their profits not by delivering better services to capture market share or cut- ting costs but by selecting more profitable patients. To maximize their profits in a competitive market, firns use whichever method makes sense in that environment. In a healthy market environment, they will try to capture mar- ket share from their competitors by pleasing customers more, maximize profits by reducing costs through efficiency gains, and expand their product lines through imitation or innovation. Wherever possible, however, they will try to exploit or construct advantages. Where they succeed, market-generated pressures for efficiency may be weak. Distortionary features of health service markets often enable providers to counter the bargaining power of suppliers, pa- tients, or purchasers; ward off the threats posed by new en- trants and imitation products; and control a large share of the relevant market. Ensuring the existence of healthy com- petition is thus a critical element of the incentive regime created under organizational reform. Where competition isn't achieved, regulation plays an even more important role in reining in opportunistic behavior. government requirements). Reviews, including comparison with this benchmark, are included in corporate plans.35 Privatized Organizations Privatization is the most extreme version of marketizing organi- zational reform. This reform entails transferring a public hospital to private ownership, either a for-profit or nonprofit organization. 58 o Innovations in Health Service Delivery Nonprofit privatization is conceptually distinct from for-profit privatization. Privatization removes the hospital from all direct control of the hi- erarchy of government officials or public sector rules. The organiza- tion is thus fully independent of the hierarchy, although the manage- ment is likely to be constrained by the new owners or trustees. All incentives come from opportunities to earn revenue, and they are strong, since private owners or shareholders (in the case of for-profits) are the residual claimants on extra revenues, now called "profits."36 These two forces drive this model's high-incentive features-complete market exposure to earn revenue and strong owner motivation to cap- ture revenues and monitor management. The owners of a privatized hospital have at their disposal the full range of institutions developed in private corporations to ensure good governance or monitor relations between the owner and man- ager.37 Dissatisfied owners can voice their views-through their se- lection of board members or, more commonly, by divesting. Stock or dividend performance can alert owners or boards to poor man- agement performance. For managers, the job market creates pres- sures to perform well to maintain their employment, reputations, and employability. Anticipating problems in dealing with profit-maximizing provid- ers, many countries are exploring nonprofit privatization as an alter- native. This consists of transferring or converting a public hospital to a nonprofit, which differs in its governance from a for-profit entity. Because ownership is private, the hierarchy does not directly control the hospital in any way. In some countries, however, regulatory re- quirements to maintain nonprofit status, and hence subsidy eligibil- ity, mean that the government retains certain "control" rights. Governance of privatized nonprofits conceptually resembles that of corporatized hospitals more than privatized for-profits for two reasons (box 1.7). First, through nonprofit regulation, governments exercise strong but indirect and voluntary controls over them. Sec- ond, there are no private residual claimants on leftover revenues. In a privatized hospital, market pressures complement perfor- mance pressures from owners. Owners want profits, and succeeding A Conceptual Framework for the Organizational Reforms of Hospitals ° 59 Box 1.7 Good Governance: What Is It and Why So Litfle in Public Hospitals? Governance is commonly defined as the relationship between an organization and its owner(s). Good governance is said to exist when managers closely pursue the owner's objectives or when "principal-agent" problems have been minimized. Governance is not usually problematic in small businesses or organizations where owners can directly observe and evaluate managerial performance. From studying successful, large private organizations, experts have identified these key ingredients for good governance: e Objectives. Narrow, clear, nonconflicting objectives of own- ers translate into narrow, clear, and measurable criteria for management performance. In private (for-profit) com- panies, managers can be monitored easily because owners have two clear objectives-maximize profits and maximize share price. Both criteria are observable and measurable. o Supervtsory structure. Responsibility for supervising man- agement is vested in an effective, professional body (e.g., board of directors) that itself has clear responsibilities and accountabilities. Competitive environment. Competition in the product, la- bor, supply, and capital markets promotes managerial ef- ficiency by forcing the adoption of the most efficient pro- duction arrangements to stay competitive and capture market share. Competition in the product market allows owners to compare performance of the firm (and man- agement) with other firms and diminishes monopoly rents, which management might misallocate, or use to hide weak performance. Ability to monitor performance, combined with a competitive managerial labor market, allows owners to compare company managers' perfor- mance and motivate managers through rewards and job (Box contniues on the following page) 60 0 Innovatnons in Health Service Delivery BoJ 1.7 (coninued) security. Accounting standards and well-functioning mar- ket institutions such as stock markets drastically reduce the costs of monitoring management. Profits from one company can be easily compared with similar companies in the sector. Share prices can be easily followed. Why do public hospitals (budgetary organizations) have had governance? o Fuzzy or coniflicting objectives. Hospital goals are not well de- fined and may conflict. Hospital goals are not differenti- ated from sectoral goals (may include delivery of quality health services, efficient use of government resources, pov- erty alleviation [equity], and delivery of "social" goods). o Weak supervisory structuire. Accountability mechanisms are weak and input-control focused. Objectives are not usu- ally translated into narrow, clear performance criteria for management. Often, there is no effective structure for monitoring managerial performance. Politicians and bu- reaucrats involved with supervision have room to pursue their own (nonhealth-related) agendas-such as jobs or sinecures for loyal supporters. o Poor information environment. Even with managerial per- formance criteria, monitoring may be impeded by lack of competition or other external institutions (like equity or debt markets) that generate information about relative performance. How do the reform modalities of autonomization, corporatization, and privatization address the governance problems of public hospitals? o Objectives. These reforms are designed to address gover- nance problems by narrowing the range of objectives for A Conceptual Framework for the Organizational Reforms of Hospitals 0 61 which managers are accountable. The objectives are trans- lated into measurable performance criteria. * Supervisoiy structure. Organizational reforms often include the creation of a professional organization (agency or board) vested with responsibility for monitoring achieve- ment of performiance targets. Frequently individuals are recruited on technical or professional bases. The objec- tives are usually narrowed to focus on economic efficieng- which is more easily monitored than other objectives. However, this requires the development of alternative mechanisms to pursue other sector (social) objectives. * Competitive environment. Organizational reforms some- times include provisions for product market competition or benchmarking to help the government-owner judge managerial performance. Capital funds may be allocated on a competitive basis to encourage accountability in fi- nancing improvements and repaying debt. Management employment and salary may be tied to performance. What are the biggest problems in trying to inmprove governance through organizational reforms? * Continued politicization of decisionmaking and opaque inter- ventions. Old habits of informal intervention by "owners" in hospital operations are usually perpetuated by failure to establish an oversight structure that ensures accounta- bility for the narrowed range of goals or by failure to de- velop or ensure the use of other mechanisms to achieve key sector goals (e.g., related to access and equity). * Failure to identify andfence off "social" goods. Governments often have a hard time clarifying which "social" services they want delivered and targeting subsidies effectively. Hospitals can end up relying on internal cross-subsidiza- tion. Management may then try to excuse poor economic performance or failure to meet targets by citing ad hoc (Box continues on tliefollowing page) 62 o Innovations in Health Service Delivery IZox 1.7 (conlinued) interventions, unfinded mandates, and their associated costs. This reduces the supervisor's ability to hold the hospital accountable. Why these failures? o Internal stakeholders disagree. Defining narrow objectives is hard in health because multiple interests in government may well disagree on what the key objectives are or ought to be. Government-owners may have many health care objectives but no sense of priorities. o Clear objecttves and priorities reveal trade-offs. Specifying objectives and priorities can make explicit what the state will and will not fund. This is often a politically costly action. o Challenging new tasks for bureaucrats. Creating alternative mechanisms' to pursue other sector objectives (besides organizational efficiency) is hard because it forces gov- ernments to engage in more complex activities (like contracting, purchasing, and regulation). Under an in- tegrated public system (budgetary organizations), gov- ernments can pursue sectoral objectives through implicit understandings that they will transfer a certain amount of resources and the hospitals will provide all takers with services in some form. Under an organizationally re- formed system, the government would have to identify which services would be delivered to the poor (for exam- ple), and purchase (or sometimes mandate) and monitor their delivery. o Bureaucrats prefer direct control and discretion. Under indi- rect accountability mechanisms, discretion is intended to be constrained. Politicians and bureaucrats, however, usually prefer ad hoc direct interventions with fewer con- straints on their relations with hospitals. Not constrain- A Conceptual Framework for the Organizational Reforms of Hospitals 0 63 ing these interventions has created many problems in hospital reforms. Govermments trying to improve governance through au- tonomization or corporatization will need to enhance their capacity to develop and implement sector policy through in- direct mechanisms such as contracting and regulation. They must create structures for administering the new accounta- bility arrangements-and for restraining ad hoc interven- tion by politicians and bureaucrats. Good Governance a. These mjechaisms inldeacuntability mc ecais s Clearly deine goays t o Independent oNarrow scope o Professional i mhievable targeyt g eransparent Hospial that dce st exposure . o itaptl market P Suppries m crket o Labor market a. These mechanisms include accountability mechanisins stemming from funding/paynient arrangements and regula- tory mechanisms. in the market is the only way to generate them-competing with other hospitals to increase sales, attract patients, undertaking changes that decrease costs, and so on. Privatization requires even more systemic reforms than corporati- zation to complement the hospital reforms-to ensure that social objectives such as access, equity, and clinical effectiveness are not 64 o Innovations in Health Service Delivery sacrificed in the name of efficiency and consumer quality. Reforms to increase regulatory capacity and to establish effective purchasing arrangements are particularly crucial. Resuls and Lessons fromn Ougside the H3eaGt Seclor Relevant experience from other sectors provides useful insights. These insights are especially valuable since the reforms have only r e- cently been applied in health, and they haven't been as rigorously evaluated as in other sectors (box 1.8). Autonomization Autonomization has improved performance in some public agencies that produce unspecifiable or unsaleable outputs such as policy ad- vice or policy implementation, but it requires fairly sophisticated in- stitutional arrangements to work.38 With production and service delivery organizations outside the health sector, the results of autonomization reforms have been dis- appointing. In many cases, management has been given too little control over production to enable or encourage a response to the newly created rewards for performance. In other cases, substantial autonomy was given, but accountability arrangements were ineffec- tive. A comprehensive review of performance contracts throughout the world found that their influence was weak and often negative.39 The key reasons cited were: o The informational advantages of managers over government offi- cials enabled them to negotiate contracts that did not require high performance (i.e., they were able to maintain organizational slack). o Contracts rarely rewarded or penalized managers or staff. o Governments often reneged on their promises to the organiza- tion's management by formally or informally retaking control, thereby interfering with management decisions. A Conceptual Framework for the Organizational Reforms of Hospitals 0 65 Box 1.8 Key Lessons from Other Sectors • Organizational reform must address labor issues directly and up front. Leaving labor "for later" is at the root of many reforms that have failed to have an impact on effi- ciency anld performance problems. Institutional innovations can allow organizational reform to be applied to new areas. Unbundling can enable mar- ket forces to be brought to bear in many areas previously thought to be natural monopolies. For example, conces- sions can create competition "for the market" when com- petition in the market is impossible. o Intermediate reforms of autonornization and corporatiza- tion are more institutionally intensive-they imply a more sophisticated role for government-because government must use indirect disciplining forces instead of simpler instruments of direct control (as in a budgetary organiza- tion) or market forces (as in a privatized organization). o A supportive external competitive environment must ac- company reforms in internal incentives. If this environ- ment is not present, systemic reforms may have to be implemented before organizational reforms can work. In practice, autonomization outside the health sector usually failed to introduce durable changes in incentives, either because the re- forms were not fully implemented or because they lacked internal coherence. Frequently, alternative accountability arrangements were never realized, encouraging reversion to previous mechanisms or creating a vacuum. For these reasons, corporatization and privatiza- tion have become the preferred types of organizational reform for commercial and infrastructure enterprises.40 Corporatization Corporatization has also had mixed results in commercial and infra- structure enterprises. In some cases, performance improved, but im- 66 o Innovations in Health Service Delivery provements were not sustained.41 In other cases, failure to implement key aspects of the model led to poor results. In every case, effective corporatization appears to have entailed intense organizational re- form of the enterprise, one necessitating a sustained, complex, and politically challenging role for government agencies and officials.42 The main problems connected with corporatization are rooted in the failure to depoliticize decisionmaking in a sustainable way. First, the board or management has rarely been made responsible for ful- filling a sufficiently narrow and clear set of objectives. Second, fi- nancial accountability has not been created; managers are simply better informed about costs and turnover. This problem is severe in enterprises that remain responsible for delivering some goods or services without payment. This unfunded mandate, by blocking comparison with providers that do not carry that burden, often gives management a convenient excuse for poor performance. Few governments have truly exposed corporatized enterprises to competition. Instead of limiting funding to what would be commer- cially obtainable, they have continued to give troubled enterprises capital injections. This has substantially reduced market pressure. Rarely have governments succeeded in removing systemic privileges for corporatized enterprises. Beneficial regimes for price setting, capital allocation, purchasing, and tax provisions all have prevented a leveling of the playing field. In most cases, market forces have not been allowed to play their full role in creating accountability. While a lack of competition has muted pressure to improve per- formance, corporatized enterprises have also suffered from contin- ued constraints related to public sector ownership. Rules limiting in- dividual access to loans are one such constraint. Management often mentions as a key problem its inability to raise capital for service expansion and capital renovation based on a business plan or project viability. This limitation is also used to explain failure to improve, hence diminishing accountability. Pnivatization Privatization of commercial and infrastructure enterprises has gen- erally led to good results: improved productivity, wages, and, for A Conceptual Framework for the Organizational Reforms of Hospitals * 67 government, tax revenue. Subsidies could be cut, thereby reducing the fiscal drain associated with the enterprises while under public control. In infrastructure, quality and availability of services has gen- erally improved.43 Competition faced by enterprises is a key determinant of perform- ance, perhaps even more so than privatization itself. However, even in markets where competition is weak, the performance of privatized enterprises has improved.44 Privatized entities, however, do not make special provision for the poor. In enterprises producing services of so- cial import, additional steps have been needed (i.e., direct subsidies, price regulation) to ensure that commercially run providers continue services to the poor or other needy users. Well-designed organiza- tional reforms have included changes in both funding and regulation and have been successful in addressing these potential conflicts. A Coherent Approach to Organizational Reform and Altering Incentives Based on our discussion thus far, one important issue emerges re- lated to the design of organizational reforms: the internal coherence of the reform package and its fit with the institutional and market environment. Internal Consistency of Reform Package Both the theoretical literature and experience in applying these re- forms to other sectors point to critical linkages among the reform elements. Different parts of the organizational structure must be aligned with each other. For instance, managers who have incentives to cut costs must also have the ability to alter the use of the key cost drivers, including labor. External Consistency of Reform Package In addition to internal consistency, the components of the external environment must fit together. The governance arrangements cre- 68 0 Innovations in Health Service Delivery ated must complement the funding arrangements. Both funding and governance arrangements must be congruent with the market envi- ronment. It is extremely challenging, however, to harness the market environment to reform objectives. In most cases, this factor was taken as a given, and reform packages focused on governance and funding arrangements. Using market structure. Under these reforms, markets are relied upon to create pressures for improved performance. Vhile health services present unique challenges, the generic prerequisites for functioning markets must be fulfilled in any event. Providers must compete with each other. Contracts must be enforceable. The legal and judicial framework to review anticompetitive practices and enforce con- sumer protection will play a heightened role. Some weaknesses can be addressed with the instruments discussed below. However, the government's capacity to generate these conditions changes only slowly. Hence successful reforms will either take them as given, or rely on only incremental improvements. Fit with the institutional environment. Besides matching each other, the key elements of the organizational reform package must also be consistent with the more general institutional environment. For ex- ample, reforms must explicitly take into account the capacity of pub- lic organizations, especially those that will be directly involved in the reforms, and in new tasks created by the reforms. Since these reforms will reduce the government's direct control over provider organizations, indirect control mechanisms become doubly important. The soundness of the government's capacity re- garding budget management, procurement (contracting), regulating, monitoring, and auditing will play a crucial role in encouraging the reformed hospitals to respond productively to their new freedom. Reform packages may thus need to include measures to strengthen the public sector functions identified as most critical to the new in- centive and monitoring regime. A Conceptual Framework for the Organizational Reforms of Hospitals o 69 Complementary Reforms and Sector Objectives Marketizing organizational reforms move health services closer to a market environment and encourage providers to focus on their fi- nancial viability. When a government department performs service delivery, sector objectives can be pursued through direct control of the providers. However, when providers are encouraged to respond to market incentives, critical sector objectives may be threatened, in- cluding quality, cost control, and access for the poor. In these areas, the government must make up for lack of direct control through en- hanced dissemination of information and the creation of an effective regulatory and contracting framework for health services. The fur- ther one goes toward a market-based incentive environment, the more important these complementary reforms become (figure 1.4). Access and equity. As providers become increasingly concerned about their financial viability, they will be more reluctant to provide unre- munerated (or not fully remunerated) care. Internal cross-subsidiza- Figure 1.4 Complementary Systemic Measures for Organizational Reforms to Work X Budgetary A Autonomized Corporatized Privatized unit unit unit ~~~~~~~unit Cti ~~~~~~~~~~~~Ifomto E a) E a) E - 0 Contracting 70 o Innovations in Health Service Delivery tion among patients or services will cease to be a viable mechanism for ensuring availability of important services or access for the needy. Problems of equity and access will emerge, unless the government addresses these problems through other mechanisms. Where once the government pursued universal access to services through direct control of delivery, it may now have to contract for delivery of these services, and public servants will have to develop new skills for contracting or purchasing. This new way of operating may also provide a strong instrument for cost control through a na- tional global cap on expenditures. Where the government itself is not the sole or primary funder of health services, effective regulation of purchasers or insurers is needed, in addition to whatever contracting the government does. Government contracting, subsidization, or mandates on coverage for services delivered to the poor is critical-especially if the reformed hospitals and clinics have previously been playing a safety net role. As organizational reforms increasingly shift risk for the cost of services onto providers, the usual tendency to engage in cream skim- ming will emerge. Thus, regulation and monitoring of such practices is a critical reform element. Quality and cost escalation. Questions about quality can arise from the principal-agent structure of the doctor-patient relationship. Patients must rely on doctors to make clinical and therapeutic decisions on their behalf. However, the doctor may have different objectives from the patient's. Doctors may prescribe one treatment over another be- cause they can earn more money from it, even if it costs the patient more or may not be the most effective method of treating the pa- tient's condition. Profit maximization is a strong motivating force for bad as well as good. It can encourage service providers to behave opportunistically, to take advantage of their information asymmetry vis-a-vis patients and purchasers, and to skimp on unobservable quality features of care. Effective mechanisms to enable monitoring of quality are re- quired to ensure that the providers are motivated to supply high- quality and clinically effective services. Prescriptive quality regula- A Conceptual Framework for the Organizational Reforms of Hospitals * 71 tions can be complemented by information to support patient over- sight (as individuals or through consumer groups) and by effective government contracting for quality services. Market forces can put pressure on profit-maximizing providers to give quality care and keep prices down. But a large and cohesive group of providers has market power and can work to block compe- tition. To head off anticompetitive behavior, pro-market regulations and regulatory capacity must also be put in place. Moral hazard problems also occur in the doctor-patient relation- ship because a third party (government or insurer) often pays for health care, leaving neither doctor nor patient with any incentive for cost economy. Thus, additional cost-escalation pressures are likely to occur as providers move to market incentives. A critical element of marketizing organizational reform is thus to ensure that government contracting or purchasing is sufficiendy effective to put pressure on providers to deliver quality services-both in terms of responsiveness to patients and clinical effectiveness. A discussion of effective regulation of various forms of health in- surance goes beyond the scope of this book. But one thing is certain: organizational reforms to make government providers efficient and responsive will not work if their customers/payers do not insist that they deliver low-cost, high-quality services. Nonprofit regzulation. Converting public facilities into nonprofit or- ganizations will create a separate set of regulatory challenges related to the unique nature of nonprofit organizations in providing services for the "public good." To support the delivery of social services, gov- ernments commonly grant financial benefits such as tax exemptions and tax deductibility of contributions to nonprofit organizations. Ef- fective targeting is crucial to ensure that these forgone revenues are well spent. Such support should go only to nongovernmental insti- tutions whose primary activities center on delivering care benefiting the population. Verification of the social benefit nature of nongovernmental activ- ities to qualify for preferential treatment is a critical feature of an effective regulatory framework for nonprofit organizations. In addi- 72 o Innovations in Health Service Delivery tion, transferring public assets to the nonprofit sector usually entails a concessionary price, based on the understanding that these assets will continue to be used for the public good.45 Thus, the government must create regulations, as well as monitoring and enforcement ca- pacity, to ensure that new owners do not dissipate the value of these assets for private gain.46 For some goods and services, there are few serious problems re- lated to market failure, including most factor inputs (other than la- bor), medical goods and supplies, and nonprescription drugs. In these cases, the only complementary reforns needed may be to improve in- formation disclosure to help purchasers make informed choices. Conclusions What can readers take away from this discussion? Organizational re- forms are complex, and key elements are interconnected. Internally, the elements of the governance reforms must fit together if the desired high-powered incentives are to develop. Externally, these in- centives must be complemented and directed by pressures and con- straints from government funding arrangements as well as market forces.47 In reviewing country experiences with these reforms, readers should look closely to see how well they meet these consistency requirements. Did the reformers get the "dial settings" right? Notes 1. Although the hospital is the organizational unit examined in this volume, much of the discussion applies generically to a wide range of provider organizations. 2. There are major variations in the distribution of these resources across the world and within countries, with the greatest concentra- tion in richer countries and urban areas, Central Europe, and the former Soviet Republics. A Conceptual Framework for the Organizational Reforms of Hospitals * 73 3. M. A. Lewis, G. M. La Forgia, and M. B. Sulvetta, "Measuring Public Hospital Costs: Empirical Evidence from the Dominican Re- public," Social Science Medicine 43 (2): 221-34 (1996). 4. S. Bennett, "Promoting the Private Sector: A Review of Devel- oping Country Trends," Health Policy and Planning 7 (2): 97-110 (1992). 5. D. Gwatkin, "Poverty and Equity and Health in the Develop- ing World: An Overview," World Bank, Washington, D.C., 1997; processed. 6. World Bank, World Development Report 1997: The State in a Changing World (New York: Oxford University Press, 1997). 7. Ibid., chap. 3, pp. 41-60. 8. Examples of institutional innovations include unbundling of competitive from monopolistic components of a previously vertically integrated industry, which has allowed competition to replace reg- ulation in many areas, and tendering of concessions, which has brought competitive pressures to bear even on networks-by auc- tioning off the right to run the network for a period of time, creat- ing periodic competition for the market. 9. D.J. Donaldson, "Privatization: Principles and Practice," Lessons of Experience Senes, no. 1, International Finance Corporation, Wash- ington, D.C., 1995. 10. S. Mathur and G. Torres, The Third Wave of Privatization: Pri- vatization of Social Sectors in Developing Countries (Washington, D.C.: World Bank, 1996). 11. J. Robinson, "Physician-Hospital Integration and Economic Theory of the Firm," Medical Care Research and Review 54 (1): 3-24; discussion 25-31 (March 1997). 12. OECD, Regulatory Reform, Privatisation and Competition Policy (Paris: OECD, 1992), p. 17. 74 o Innovatons in Health Service Delivery 13. Ibid. 14. D. E. Sappington, "Incentives in Principal Agent Relation- ships," Journal of Economic Perspectives 5 (2): 45-66 (1991); cited in Robinson, "Physician-Hospital Integration and Economic Theory of the Firm." 15. E. F. Fama, "Agency Problems and the Theory of the Firm," Journal of Political Economy 88 (2): 288-307 (1980); M. C. Jensen and W H. Meckling, "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure," Journal of Financial Economics 3: 305-60 (1976). 16. Two useful references include: 0. Williamson, The Economic In- stitutions of Capitalism: Firns, Markets and Relational Contracting (New York: Free Press, 1985); and 0. Williamson, "Transaction Cost Eco- nomics," chap. 3 in R. Schmalensee and R. Willig, eds., Handbook of Industnal Economics (New York: North-Holland, 1989). 17. P. Milgrom and J. Roberts, Economics of Organization and Man- agement, chap. 9 (Englewood Cliffs, NJ.: Prentice-Hall, 1992). 18. For example, a person may own a house but not have the right to occupy it if he has leased it out. He may own a car but not have the right to transfer it freely if he has a loan secured by the car. 19. M. Olson, The Rise and Decline of Nations: Economic Growth, Stagflation and Social Rigidities (New Haven, Conn.: Yale University Press, 1982). 20. Allocative efficiency in this context refers to cost-effective use of public resources. In reform programs that emphasized this ob- jective, the organizational reform delinking the funder from the provider was viewed as an instrument for breaking the "provider cap- ture" inherent in systems allocating resources to inputs (hospitals, doctors) rather than population or services. 21. Here we refer only to management reforms within existing or- ganizational structures (nonstructural) to avoid confusion with the A Conceptual Framework for the Organizational Reforms of Hospitals * 75 structural/organizational reformns we are reviewing in this volume, many of which obviously affect management. 22. Changing administrators into managers: The emphasis on managerial skills is indicative of a trend to hold those in control of public service organizations accountable for outputs or outcomes rather than for administering services in an acceptable fashion. 23. R. Saltman and J. Figueras, European Health Care Reform: Analysis of Current Strategies (Copenhagen: World Health Organiza- tion, 1997), pp. 213-14. 24. D. Osborne and T Gaebler, Reinventing Government (New York: Plume, 1993). 25. Chapter 7 in J. Q. Wilson, Bureaucracy: What Government Agencies Do and Why They Do It (New York: HarperCollins, 1991), which contrasts the context within which managers manage in pub- lic as opposed to private bureaucracies. 26. This volume only covers autonomization and corporatization, that is, marketizing organizational reforms that maintain public own- ership. Privatization fits within the same conceptual framework, but also involves a transfer to private ownership, which brings in a num- ber of additional issues that are beyond the scope of this material. 27. J-E. Lane, ed., Public Sector Reform: Rationale, Trends, Problems (London: Sage Publications, 1997). 28. This definition refers to "organizational governance." There is another form of governance relevant for health services, and that is "sectoral governance." Sectoral governance refers to the structure and activities undertaken to monitor and enhance the performance of the overall health sector, and hence includes such activities as reg- ulation, information dissemination, and mandates. 29. Wilson, Bureaucracy. 30. Ibid., p. 117. 76 o Innovations in Health Service Delivery 31. This movement away from internal, implicit cross-subsidiza- tion is often an explicit objective of organizational reforms in other sectors. Some scholars, notably M. Pauly, believe it should also in health services. See M. Pauly, "Health Systems Ownership: Can Regulation Preserve Community Benefits?" Frontiers of Health Ser- vices Management 12 (3): 3-34; discussion 51-2 (spring 1996). 32. This description does not address the well-documented cases of retention of copayments or informal payments by employees or management-which makes them also residual claimants, though informally. 33. It is useful to distinguish between ownership accountability ar- rangements, which are tied to the governance relations between the government and the organization (i.e., corporate plan, performance contract), and funding or payment accountability arrangements, which are generated by the structure of the payment or funding mechanisms. 34. Reforms that transform hospitals into public corporations- corporations governed by public rather than private, company law- are also used in some cases. 35. Benchmarking is also an important tool for purchasers to help them set reimbursement by diminishing the informational advantage of hospital management through comparison with similar institutions. 36. In the case of nonprofit organizations, the need to sustain themselves over time requires that they usually cover at least their operational costs, even if donations are forthcoming to cover capital costs. 37. For an overview of the key ingredients for good governance in private companies, see I. Millstein, "The Corporation as an Owner- ship and Management Structure" (paper written for the World Bank Private Sector Development Forum, Baltimore, Md., 1998). 38. N. Manning and D. Parker, "Nationalization, Privatization and Agency Status within Government: Testing for the Importance of Ownership," in P. Jackson and C. Price, eds., Privatisation and Reg- ulation (London: Longman Group, 1994). A Conceptual Framework for the Organizational Reforns of Hospitals * 77 39. World Bank, Bureaucrats in Business: The Economics and Politics of Government Ownership, World Bank Policy Research Report (New York: Oxford University Press, 1995). Performance contracts in state- owned enterprises are almost always utilized in conjunction with enhanced autonomy. 40. Critical factors to consider in determining which services can be shifted to the high-incentive environments of corporatized or pri- vatized entities, and under what conditions. See N. Girishankar, Re- forniing Institutions for Service Delivery, World Bank Policy Research Working Paper 2039, World Bank, Policy Research Department, Washington, D.C., 1999; R. Gregory, "Accountability, Responsibil- ity and Corruption: Managing the Public Production Process," in J. Boston, ed., The State under Contract (Wellington, Mass.: Bridget Williams Books, 1995). 41. I. Duncan and A. Bollard, Corporatization and Privatization: Lessonsffrom New Zealand (Auckland, N.Z.: Oxford University Press, 1992). 42. A. Schick, "No Leapfrogging" (presentation made at the World Bank, 1998); from A. Schick, "Why Most Developing Countries Should Not Try New Zealand's Reforms," World Bank Research Ob- server 13: 123-31 (February 1998). 43. A. Galal, L. Jones, P. Tandoon, and I. Vogelsang, Welfare Con- sequences of Selling Public Enterprises: An Empirical Analysis (New York: Oxford Umversity Press, 1995). 44. N. Boubakri and J-C. Cosset, "The Financial and Operating Performance of Newly Privatized Firms: Evidence from Developing Countries," Journal of Finance 53: 108 1-1110 (June 1998). 45. J. Casagrande, "Social Sector Privatization in Mongolia: Con- cepts Learned through Experience," Private Sector Development Department, World Bank, Washington, D.C., 1999; processed. 46. A "nondistribution" constraint is a critical component of non- profit organizational rules-blocking distribution of residuals to any individuals. 78 o Innovations in Health Service Delivery 47. For a more detailed discussion on the public and private roles in the health sector, see A. Harding and A. S. Preker, eds., Private Participation in Health Services, Washington, D.C.: World Bank, 2003; and A. S. Preker, A. Harding, and P. Travis, "Make or Buy Decisions in the Production of Health Care Goods and Services: New Insights from Institutional Economics and Organizational Theory," Bulletin of the World Health Organization, 78 (6): 779-90 (May 2000). CHAPTER 2 Implementing Organizational Reforms to Hospitals in the Public Sector Chris Ham and Loraine Hawkins The "slip between cup and lip"-the frequent gap between policy- makers' intentions and the results of health care delivery reforms- has prompted extensive analysis of the way policy becomes practice.' From these studies have come advice and recommendations on ways of implementing policies more effectively, which is relevant to the reformns of the hospital sector described in this volume.2 In parallel, analysts, focusing on the role of actors in agencies responsible for implementation, have highlighted the difficulties of delivering poli- cies on the ground. Through negotiation and adaptation during im- plementation, policies are often refashioned, leading to outcomes different from those intended at the outset, according to these ana- lysts. From this perspective, the failure to implement policies is less surprising than the fact that they have any impact at all, considering the hurdles to be surmounted in translating policymakers' intentions into action. Accumulated research into implementation has progressively chal- lenged the linear model of the policy process that informed much of the early thinking about health care reform. Instead of assuming a se- quential relationship between policymaking and implementation and seeing the challenge as one of making this process as smooth as pos- sible, policy analysts have studied action and behavior within public 79 80 o Innovations in Health Service Delivery agencies and the influence of front-line staff or "street-level bureau- crats."3 In doing so, they have highlighted the discretion allowed staff and their agencies as an important source of implementation "deficit." The issue then became not so much how to design policies to en- sure effective implementation but rather how to change the behavior of street-level bureaucrats in the desired action by understanding their motivation and values. By viewing implementation as a bar- gaining process between actors over time, the focus has shifted to implementation as evolution and to policymaking as a learning exer- cise. In this way, researchers who adopted a "bottom-up" perspective threw into sharp relief the weaknesses of the "top-down" approach to policymaking, attracting increasing support from students of the pol- icy process. Many of the hospital reforms described in this volume were caught between such top-down and bottom-up pressures. In reality, policy analysts from both traditions made a significant contribution to our understanding of the complexity of implementa- tion issues, although each offered only a partial interpretation of the phenomena studied. Increasingly we recognize that studies of the implementation of public policies need to draw on the insights of both the top-down and bottom-up approaches by focusing on the range of factors that influence implementation and the role of actors and institutions at different levels. Just as policymakers at the top or center of government can influence policy outcomes, so, too, man- agers and professionals at the bottom or periphery of government shape, modify, and sometimes initiate developments. Implemen- tation studies must pay attention to both sets of actors as well as the history and context because rarely are policies implemented in circumstances where there are no established interests. Achieving change usually depends on overcoming potential resistance from beneficiaries of the status quo. The relative neglect of implementation issues applies as much to studies of health policy as to other areas of public policy. In the U.S context, Marmor comments: The tendency to overlook implementation issues is not surpris- mg, given how difficult, and unglamorous, it is to figure out the hinplementing Organizational Reforms to Hospitals * 81 nuts and bolts of real programs-and how much more enticing it is for politicians and policy analysts to bandy about big ideas. The implementation of policy is less visible and less dramatic than the framing of policy-and often, frankly, more arcane. The neglect of implementation issues is more than a simple in- tellectual mistake; it may be a rational response to the fact that our political system confers more rewards for the shrewd de- ployment of symbols and generalized arguments than it does for detailed realistic analysis and forecasting.4 One approach to the implementation of health care reform is il- lustrated in figure 2.1. Starting from the view that policymaking and implementation are inextricably linked, this approach organizes the factors affecting implementation into four categories: context, pro- cess, actors, and content. Walt, one of the originators of this ap- proach, emphasizes that, although these factors can be separated for analytical purposes, in practice they are interrelated.5 She argues that the content of policy and its design, issues so frequently emphasized by policy analysts, are often less important in accounting for imple- mentation success or failure than the context, process, and actors. That is why Walt's framework gives prominence to analysis of the context of implementation, the processes involved in carrying policy into action, and the actors affected. There are strong similarities between Walt's framework and Reich's analysis, which focuses on health care reform in developing countries. Reich sets out a model of applied political analysis for re- form, drawing on the insights gained from studies of the politics of policymaking to indicate how implementation problems can be an- ticipated and overcome.6 The more general point here concerns the importance of bringing a political economy perspective to bear on the study of implementa- tion. The term political economy describes a range of perspectives.7 From our point of view, however, its significance is in indicating the need to examine the interaction of economics and politics, both in theory and in practice. In relation to theory, this means drawing on the insights of the new institutional economics (see chapter 1) as well as the analysis of political scientists who focus on the role of institu- tions and pressure groups in the policy process.8 In relation to prac- 82 0 Innovations in Health Service Delivery igjure 2.0 A Framework for Health Policy Analysis COMTlEXT / -As ~Individuals\ / ~~~-As group mebers \ COMTEMT Source Walt and Gibson, "Reforming the Health Sector in Developing Countries tice, it entails attending to the economic changes that impinge on policy development and the political response to these changes. Macroeconomic analysis of these factors, in turn, needs to be related to the use of both incentives to promote implementation and politi- cal directives to encourage change. Faced with evidence of failure in the health sector by both market and government, a political econ- omy perspective enables us to undertake a more complete and real- istic assessment of the factors encroaching on health care reforms, including those centered on the corporatization of public hospitals described in this volume. Conlexl of Reform The context of reform in the public hospital sector during the past few years also encompasses historical, economic, political, and insti- tutional factors. History and Inherited Commitment History, at its simplest, is important in establishing a series of prece- dents and commitments that reformers have to confront. Policies Implementing Organizational Reforms to Hospitals * 83 and organizations with a long history may be more difficult to re- form than those of recent origin, unless exceptional circumstances such as war or major economic change create conditions that thaw or shatter established arrangements. This means that marginal change is both more feasible and more common than fundamental change. Moreover, because policymakers rarely confronted a blank sheet of paper, what already existed shaped and constrained innovation. In some circumstances, past policies themselves may stimulate change, as when weaknesses in original designs force a reappraisal of current arrangements or when the financial consequences of these arrange- ments become unsustainable. In Wildavsky's formulation, policy may be its own cause in a cycle of development that, in some respects, be- comes self-perpetuating.9 Macroeconomic Situation In relation to the economy, many countries have sought to constrain public expenditure since the oil crisis of the 1970s. A broadly based program of public sector reform has ensued, in which the set of ideas referred to as the new public management has influenced the poli- cies pursued.'0 These ideas have stimulated various efforts to reform health services. The effects have been seen in the epidemic of health care reform in the industrial world during that period, including a re- assessment of the role of the state in health care and measures to en- hance systemic efficiency and responsiveness."1 There are strong parallels between such reforms in state-owned enterprises and those needed in the hospital sector of the health system. Interest in corporatization has been integral to this process as pol- icymakers have searched for new ways of tackling perceived weak- nesses in service delivery. In Central and East European countries, the move from centrally planned to market economies since the end of the cold war has influenced not only industrial ownership and control but also the financing and delivery of health services and other forms of social insurance. 12 In relation to developing countries, Bennett and colleagues have shown how economic recession serves as the catalyst for health care reform, with the most sweeping reform 84 0 Innovations in Health Service Delivery proposals in countries that had encountered the most severe reces- sion.13 The macroeconomic environment in this way influences health policy both in relation to the availability of resources and in forcing a reassessment of established delivery mechanisms. Political Debate on Role of the State Changes in the economy have been instrumental in shifting the terms of political debate. The post-World War II consensus on the expansion of the welfare state has broken down as politicians of the Center-Right have questioned the size of the public sector and the balance between collective and individual responsibility in the face of increasing evidence of government failure. Incremental policy de- velopment has therefore become less common, and more radical op- tions have been proposed. This trend has found expression in some quarters in the argument that, as part of a fundamental reappraisal of government intervention in health care, the role of the state should be limited to providing a safety net for citizens unable to make their own provision. Less radical have been moves to encourage an expan- sion of private financing and provision alongside that of the public sector and to make health care more businesslike by importing ideas from successful companies. Yet political debate has fluctuated as enthusiasm for private fi- nancing and service provision has been superseded by recognition of the weaknesses in health care markets, including managed or quasi- markets. As a consequence, the state continues to fund around three- quarters of total health care expenditures in industrial countries, and there has been no significant retreat from the commitment to basic social goals like ensuring access to necessary medical care and pursu- ing equity through government regulation of the health sector. This commitment has an important bearing on the implementation of corporatization reforms in the hospital sector in that the authoriz- ing environment may not give actors the support needed to pursue these reforms.14 Implementation is particularly problematic in these circumstances. lmplementing Organizational Reforms to Hospitals * 85 The Institutional Environment Political debate about the role of governments and markets in health care has been rehearsed in an environment in which a number of in- stitutions stand to gain or lose from change. On balance, the inertia of these institutions has been more evident than their enthusiasm for change, illustrating Machiavelli's aphorism that there is nothing more difficult to carry out, nor more doubtful of success, nor more dangerous to handle, than to initiate a new order of things. For the reformer has enemies in all those who profit by the old order, and only lukewarm defenders in all those who would profit by the new order. 2 5 Pierson, in his comparative analysis of welfare reform in the United Kingdom and the United States, provides a practical illustra- tion of the impact of history on policymaking.16 As Pierson notes, welfare reform under Prime Minister Margaret Thatcher and Pres- ident Ronald Reagan delivered much less than promised, partly because past policies generated support for services from both recip- ients and the agencies delivering these services. The cost to politi- cians who challenge existing arrangements is often considerable, and retrenchment therefore usually occurs indirectly rather than through frontal attack. In this way, institutions act as a barrier to change, rein- forcing the argument of analysts who assert that an institutional per- spective is central to understanding the dynamics of policymaking. Dominant Values The values that inform health policies and the respective roles of the state and the market cut across economic, political, and institutional factors. Dominant values may be deeply entrenched and, like estab- lished institutions, can act to protect the status quo and constrain the set of policy options placed on the agenda and the manner of imple- mentation. Only at times of real or perceived crisis will these values be questioned and the opportunity for radical refortn created. The nature of these values and the policy response will vary from country 86 o Innovations in Health Service Delivery to country. Also, the implementation of policies to reform health services will depend on the degree of consistency or fit between the values embedded in these policies and those held by actors in the sys- tem. As an example, corporatization may be viewed quite differently in health care than in state-owned industries because of the attach- ment citizens feel to public health care. In view of the importance of values, politicians may obfuscate while pursuing changes that might be unpopular or while seeking to implement their poLicies in stages.17 In implementation, it is the interplay between the factors influ- encing context that is important. Global economic trends have dif- ferent effects in different systems, depending on the response of politicians and the role of institutions and dominant values. Al- though certain general tendencies can be discerned, the relationship between context and policy is contingent and variable, and the impact on public policy of economic recession or political realign- ment is unpredictable. The radical response of the Center-Right to changes in domestic and international economies that occurred in the 1980s in New Zealand, the United States, and the United King- dom was not emulated in West European countries where social democratic values remained resilient. Actors both in the poLitical sys- tems and in the health care system mediate the effects of context and put their own imprint on the reform agenda. Acors Many different actors can influence policy implementation: politi- cians and bureaucrats, managers and health care professionals, the public and patients, public sector health workers, and stakeholders. Politicians and Bureaucrats The most important actors-politicians and bureaucrats-hold posi- tions of power within government. By definition, many policies orig- inate among these actors, and their determination and consistency of purpose (the "political will" to produce change) are likely to be cru- Implementing Organizational Reforms to Hospitals * 87 cial in ensuring effective implementation. In this context, govern- ment is plural and not singular, and differences between these actors (e.g., between politicians in governing coalitions, politicians and bu- reaucrats, health ministries and finance ministries) can make policy implementation more difficult. Changes in the party or parties in power will also have a bearing on implementation, in extreme cases leading to a complete about-face. Research into policymaking has highlighted the influential role of bureaucrats at certain points'8 and of politicians at others,'9 suggesting that the contribution of each needs to be considered. Managers and Health Care Professionals Managers and health care professionals responsible for implementa- tion of policies at a lower level also exert influence. Indeed, many policy analysts believe these actors hold the key to understanding the dynamics of power and decisionmaking in health care. Managers and professionals express their influence pardy through their repre- sentative bodies and partly in their day-to-day work as street-level bureaucrats in the health care system. The way lower level managers and professionals interpret policymakers' intentions can have a sig- nificant bearing on what happens in practice whenever loose policy definitions leave room for interpretation. This helps to explain why policymakers may pay special attention to anticipating potential resistance from managers and professionals, including the adoption of "divide and rule" strategies.20 For example, in recent health care reforms in the United Kingdom, financial in- centives were offered to encourage family doctors to hold down budgets and to persuade managers to volunteer their hospitals to be- come self-governing organizations. These incentives avoided the im- pression that the reforms were being imposed and allowed for im- plementation to be phased in. The United Kingdom's experience during this period demonstrated that seeing managers and profes- sionals as a cohesive group would be a mistake. There are divisions within each category, and the balance of power has also shifted be- tween managers and professionals. Indeed, health care reforms may 88 o Innovations in Health Service Delivery deliberately seek to strengthen the role of one group vis-a-vis the other, as in the steps taken in a number of systems to hold profes- sionals accountable to managers for their performance. Managers are not only involved in the implementation of politi- cally initiated reforms but may also themselves be a source of inno- vation. This point is at the heart of the bottom-up perspectives on implementation. It is emphasized by Moore, who challenges the pas- sive view of public sector managers conveyed in traditional accounts of public administration.21 In its place, Moore offers an interpreta- tion in which managers are more akin to their entrepreneurial coun- terparts in the private sector. According to this perspective, strategic managers in the public sector are involved in managing downward to ensure the implementation of change, and managing upward to se- cure support from the authorizing environment and political leaders. Managers are at their most effective when these different elements are aligned. Thus, implementation problems may arise in case of mismatch between the authorizing environment, the endorsement provided by politicians, and the capacity of institutions to deliver change. Building strategic management capacity in the public sector is therefore a precondition for effective implementation. Public and Patients The public and patients are another key group of actors. As the clients and beneficiaries of the services under reform, their attitude toward the new policies will be one of the considerations in the minds of politicians promoting change. To the extent that the public supports existing services, public opposition may rule out change as a serious option. Equally, public dissatisfaction with existing arrange- ments may create unexpected pressures for reforms. Either way, the influence of the public and patients should not be exaggerated, considering evidence from a number of different sys- tems that these actors are more often than not the "repressed" inter- ests, whose views are rarely articulated strongly in the health policy debate. This applies to developing countries as much as industrial countries. Bennett and colleagues show how, in developing coun- tries, too little effort to inform the community about the rationale Implementing Organizational Reforms to Hospitals * 89 of reforms has led to misunderstanding among those affected by change.22 Public Sector Health Workers Bennett and colleagues also note the influence of public sector health workers in the reform process. In the countries studied, these work- ers could be powerful opponents of reform, and often the power of public sector unions had to be weakened to overcome opposition. Politicians unwilling to challenge the power of public sector health workers and their trade unions faced formidable obstacles in imple- menting their policies.23 This finding reinforces Pierson's conclu- sions about the obstacles encountered in welfare reform in the de- veloped world. Structural Interests and Interest Groups Beyond these actors, many other structural interests, interest groups, or stakeholders may influence implementation. Their actual role depends on the precise nature of the reforms being pursued and the extent to which their interests are affected. Structural interests in health care who may be systematically advantaged or disadvantaged in the policy process must be distinguished from interest groups who are active on specific issues. Alford notes that debate about options for change and the ability to implement these options will be affected by the relative influence of three types of structural interests: professional monopolists (in his terms, the dominant interests), corporate rationalizers (the challeng- ing interests), and the community population (the repressed inter- ests).24 Echoing the argument of analysts of the use of power, this perspective helps to explain how, in the health care sector, the reform process often resembles "dynamics without change" through the ability of dominant interests to rule certain issues off the agenda or at least to preserve the existing pattern of benefits even in the face of change. The influence of both structural interests and interest groups varies over time as well as between issues, and politicians constantly weigh the electoral consequences of their actions even if the com- 90 - Innovations in Health Service Delivery munity is not well represented in the policymaking process. A thor- ough analysis of the role of stakeholders is therefore necessary in un- derstanding the influence of different actors in either facilitating or hindering implementation.25 Process The implementation process encompasses the nature of the political system, the relationship between policymakers and those responsible for implementation, and the approach taken to reform. Majority Governments versus Coalitions Political systems come in many shapes and sizes, but there is an im- portant distinction between systems based on coalitions and those in which the majority party forms the government. In majority govern- ment systems, politicians can usually drive through change-and quickly. Coalition governments need to build agreement, which not only slows the implementation process but may also result in limited change. The United Kingdom's experience in the 1980s and 1990s illustrates the power available to governments with majorities and their ability to introduce changes even in the face of strong opposi- tion (see chapters 7 and 8). The experience of the Netherlands shows that, in systems based on coalitions, reform proceeds in a more mea- sured fashion and may not move forward at all at certain points as coalition partners change and politicians adjust direction. However, the time spent negotiating change in coalitions, by securing the key actors' commitment, may ease implementation, just as the rapid pace of policy development in majority governments may be followed by implementation difficulties as the direction of change is modified to accommodate the views of stakeholders. Federal versus Unitary Systems Federal systems share some characteristics of systems based on coali- tions, as in the United States, where the separation of powers be- Implementing Organizational Reforms to Hospitals * 91 tween the executive and legislature contributed to the defeat of the reforms proposed by the Clinton plan. Federal systems also illustrate the importance of the relationship between policymakers and the implementers. The division of au- thority between federal and state governments affects the ability to implement policies determined at the federal level. This also applies in unitary systems where national, regional, and local governments play a part in the administration of health services. The exercise of discretion by lower level agencies may hinder the implementation of national policies-but may also be a source of innovation and ex- perimentation. In federal systems, for example, federal agencies may pick up and generalize innovation at the state level. Extent of Decentralization The spread of innovation from state to federal level illustrates the way policy may originate at the bottom of political systems rather than the top. In Sweden, county councils have used their consider- able freedom to shape the development of health services to test dif- ferent approaches to health care reform.26 Divergences of approach of this kind may result from differences in political control of agen- cies at different levels or from administrative discretion. Whatever the reason, decentralization is likely to make national policies harder to implement, although the existence of a strong value system may help to promote consistency in approach between decentralized agencies. Again, this is evident in Sweden, where a long-standing po- litical consensus on the role of the state and the provision of public services has helped to ensure a similarity of policy development de- spite control of the county councils by different parties. Although the precise path of reform has varied regionally, variations have stayed within fairly narrow bounds. One manifestation of decentralization in the health sector is the role of private sector organizations in the funding and delivery of health services. In some European countries, for example, govern- ment's role is to ensure access and equity by regulating these organ- izations' activities and financing their work instead of assuming own- 92 0 Innovations in Health Service Delivery ership of health care facilities. Where this happens, policymakers cannot exercise control as directly as they can in countries with na- tional health systems. Working with and through a variety of self- governing institutions, policymakers have to seek change through bargaining and negotiation as well as through legislation. In these corporatist systems, since government is but one actor among many, arguments about implementation deficits are particularly evident.27 Tuohy makes a related point comparing the dynamics of health care reform in the United States, Canada, and the United King- dom.28 She shows that each system has its own logic, based on the relationship between the state and the medical profession and the relative importance of markets, hierarchies, and collegiality as mech- anisms of social control. Whereas in the United Kingdom the exis- tence of hierarchical and collegial networks tempered the impact of internal market reforms and preserved considerable autonomy for the medical profession, in the United States the dynamics of market- driven reform led to widespread change, including a decline in the influence of the medical profession as for-profit health plans emerged. Canada experienced greater stability than either the United Kingdom or the United States. The changes that did occur were negotiated and mediated through established arrangements linking Canada's pro- vincial governments and medical associations and, as in the United Kingdom, maintaining the medical profession's autonomy. Approacbes to Reform The approach to reform covers a spectrum ranging from incremen- tal at one extreme to "big bang" at the other.29 Big bang, the term coined to describe the Thatcher Government's approach to the reform of the United Kingdom National Health Service (NHS), reflected the governing politicians' conviction that fundamental, systemwide change was needed. A similar approach to reform was taken in New Zealand (see chapter 8, this volume). Big bang is usually associated with majority governments, and incremen- talism with coalition governments, although the tendency for change to occur at the margins is well established in all political systems. Implementing Organizational Reforms to Hospitals * 93 Other approaches include bottom-up reform, as in Sweden through the actions of county councils to initiate change, and "reform with- out reform," as in the United States where actors within the system have stimulated change in the wake of the failure of the Clinton plan.30 The relationship between these approaches and the imple- mentation of policy is complex, with some evidence that big bang reform may be more effective than incrementalism in delivering in- tended change, but with the associated risk that change may be more difficult to sustain after policymakers in key roles move on. Pace of Change The pace of change is part of the approach to reform. In some cir- cumstances, politicians may move rapidly to introduce new policies; in others, changes may be phased in over a period of time. Pace depends partly on the length of time elected governments have to implement their policies and partly on the way deemed most effec- tive for countering potential opposition. The existence of majority or coalition governments also influences pace. The Netherlands is an example of relatively slow change. The United Kingdom illustrates the opposite, though with a commitment to the progressive exten- sion of reforms annually.31 The Thatcher Government's commit- ment to move quickly was underscored by its rejection of arguments that its policies should be introduced on a pilot basis before wide- spread implementation. There is some evidence from the United Kingdom and elsewhere that reforms introduced quickly run into difficulty and may have to be amended in the process of implementation.32 However, reforms introduced relatively slowly may fail to deliver what is expected or may be assimilated by existing institutions to frustrate the intentions of reformers. These features of the policy process again lend support to the ar- gument of researchers who maintain that political institutions have an important bearing both on the implementation of policies in in- dividual systems and on differences in policy development between systems.33 Immergut highlights the existence of veto points in the 94 o Innovations m Health Service Delivery political process to explain variations between countries in policy outcomes. In her view, differences in the design of political systems are more important than the power of pressure groups in accounting for the pursuit of policy changes in some systems but not in oth- ers. Together with Tuohy's analysis of the impact of relationships between political institutions and the medical profession on policy outcomes, this analysis underlines the influence of process on imple- mentation. From an implementation perspective, the importance of Immergut's and Tuohy's work is in indicating that, while the impact of structural interests in the health sector poses similar challenges in different systems, the way these challenges are resolved depends on the institutional context. Content The content of policies, and the attention given to implementation arrangements as part of policy design, constitutes the fourth element of the implementation framework. Content is influenced by speci- ficity of objectives, internal consistency and coherence, institutional capacity to implement change, and arrangements for managing im- plementation. Specficity of Objectives One component of content relates to the specificity of policy objec- tives. In some cases, implementation follows a broad statement of in- tent; in others, policymakers set out their intentions in detail. The Clinton plan contained specific goals; the Thatcher reforms resided on a broad statement of intent. One analysis of the Thatcher reforms argues that the emphasis on a broad vision rather than a detailed blueprint was a key part of a novel macroimplementation strategy that contributed to relative success during implementation.34 By ex- tension, the overly prescriptive approach to the U.S. reforms may have contributed to their failure. Differences in legal traditions affect the degree to which change must be codified in advance. The Anglo- Implementing Organizational Reforms to Hospitals * 95 Saxon permissive approach allows more room for basing reforms on framework legislation than do systems where the detail of change has to be specified at the outset. Internal Consistency and Coherence Content also encompasses the internal consistency and coherence of policy. Both are related to the detail in which the different compo- nents of the proposed reforms are specified and to the relationship of these components to each other within the reform package. These is- sues are particularly important in systemwide change where the risks of implementation problems are much greater than in change affect- ing only single institutions or parts of the system. An illustration is the implementation of organizational reforms such as corporatiza- tion, where a complex range of factors need to be addressed (see chapter 1, this volume). The incentive environment affecting corpo- ratization indicates that attention needs to be given to decision rights, residual claimant status, market exposure, accountability arrange- ments, and social functions if reforms are to work as intended. If these elements are not adequately synchronized, problems are likely to arise. Policy coherence may be affected by the bargaining and ne- gotiation processes used in coalition governments, processes that may result in, for example, compromises and inconsistencies. As an example, in public health care systems, management auton- omy needs to be matched by appropriate forms of accountability. Giv- ing hospitals freedom over decisionmaking without accountability may frustrate the achievement of systemwide objectives. Similarly, exposing hospitals to competition without providing explicitly for the discharge of social functions may result in adverse selection. Even if these issues are addressed at the design stage and the "dial settings" are aligned appropriately, implementation deficits may occur if poli- cyTnakers depart from their plans during implementation or if the context of reform is hostile. In addition, actors whose interests are ad- versely affected may seek to delay or divert implementation. Our political economy perspective emphasizes the interaction be- tween the incentive environment and the political processes involved 96 o Innovations in Health Service Delivery in health care reform in explaining outcomes, particularly the fre- quent gap between intentions and results. Institutional Capacity to Implement Change Judgments about content also need to be informed by an assessment of institutional capacity to deliver the proposed reforms. In this con- text, capacity refers both to the availability of trained staff and to the provision of infrastructure support such as information systems. These issues are especially important in developing countries where evidence abounds of the dearth of managers and others with the skills needed to implement reform.35 Capacity building is also an issue in developed countries where policy involves a shift from hos- pitals managed as budgetary units in bureaucratic hierarchies, to cor- poratized entities. If the new regime retains the old staff, it may need training and development to handle new responsibilities. All will need training-the employees who deal with purchasing and market regulation as well as the managers of hospitals and other provider units. Arrngements for AAlnaging Implemenlaion Inasmuch as reforms are often politically driven, resolving all these issues in advance will be unusual. That is one reason implementation arrangements need attention in advance. In particular, significant ac- tors must be told about and won over to the intended reforms. Units dedicated solely to managing the implementation process, unen- cumbered with other responsibilities, must be formed. Potential re- sistance to reform may also be overcome if implementation proceeds in stages, with effort focused initially on sites likely to be supportive and to demonstrate success that others will want to emulate. Oiling the wheels of change by allocating additional funds to support im- plementation and to meet transition costs is another component of reform, especially if it overcomes stakeholder doubts or resistance. In this context, Moore's analysis of the role of public sector man- agers in bringing about strategic change contains some pointers rel- Implementing Organizational Reforms to Hospitals * 97 evant to health care reform. As Moore notes, these managers work both inside their organizations and outside to ensure support and to negotiate terms of accountability with the authorizing environ- ment.36 Working within their organizations, managers exert pres- sure to achieve change but also provide their colleagues with reas- surance and support. Instead of following a detailed plan, they work from a general sense of what they are trying to do, improvising and innovating in the process. Managers also judge the pace of change in relation to prospective opposition or support and their organiza- tions' capacity to respond. Using Mintzberg's formulation, managers pursue "emergent strategies" that enable change to be crafted and shaped in the light of experience.37 Against this background, the analysis by Ferlie and colleagues of the implementation strategy for reforming the United Kingdom's health service offers a point of reference for discussion.38 Against a background of sustained political and management commitment over a number of years, the strategy involved: * Following a broad vision rather than a detailed blueprint at the be- ginning of the process * Providing a visible focus of central leadership within the Depart- ment of Health, both at the ministerial level and overall project management level, to drive through key changes; using tried and tested external advisers (e.g., Sir Roy Griffiths) * Creating new intermediate tiers in "greenfield sites" (the Out- posts) seen as uncontaminated by the old regional cultures; setting clear targets for intermediary tiers in relating to the fundholder status of trusts and general practitioners (GPs) * Using a proactive communications policy to "sell" the reforms * Parachuting key personnel from the center into high-profile lo- calities (e.g., Guy's Hospital; energizing and using such allies as resources) * Sponsoring a program of development projects that could quickly be held up as role models nationally; building up coalitions locally and support networks centrally (e.g., first wave trust chairs) 98 o Innovations in Health Service Delivery o Identifying and intervening in "receptive" sites (e.g., first wave trusts) before moving on and diffusing the intervention to less re- ceptive contexts o Establishing early successes (first wave trusts and GP fundholders; forming district and family health services authorities) that helped signal that there was no going back.39 In reality, what Ferlie and colleagues characterize as a "novel macro-implementation strategy" at the time resembled what Ham has described as "policymaking on the hoof," or the government's "making it up as it went along."40 Nevertheless, their observation that this approach appears to have been relatively effective, even al- lowing for a strong element of retrospective rationalization, indicates how a model of the policy process that goes beyond the simplistic framework of top-down models holds out lessons for countries about to embark on reform. Analysis of other countries' experiences of the kind summarized in later chapters provides examples of alternative implementation strategies, thereby adding to our understanding of the range of choices available to policymakers. Effecdive Omplementafion By bringing these different strands together, we can identify a num- ber of factors that affect the implementation process (table 2.1). In interpreting these factors, it is important to avoid falling into the trap of the early rational and linear policy-process models and inferring that implementation will occur smoothly if these conditions are met. The reality is that the process of stakeholder analysis and analysis of financial, technical, and managerial resources is inherently complex and difficult. No recipe for managing political change can guarantee success. Even in the best circumstances, there may be a gap between intention and action, with emergent strategies that are as vulnerable to being blown off course as are detailed blueprints. Table 2.1 should be read with this in mind. Implementing Organizational Reforms to Hospitals * 99 Table 2.1 Tools for Planning and Managing Change AREA OF ANALYSIS PLANNING ACTION Analyze the ease with Analyze, and adiust when possible, conditions for which policy change can facilitating change (one implementing agency, clear be implemented. goals, one oblective, simple technical features, marginal change, short duration; visible benefits; clear costs) Make values underlying Identify macro- and microlevel values underlying policy policy explicit decisions If values conflict with policy, support will have to be mobilized, costs, minimized Do stakeholder analysis Review interest groups (and individuals) likely to resist or promote change in policy at national and institutional levels Plan how to mobilize support by consensus building or rallying coalitions of support Analyze financial, Consider costs and benefits of external funds. Assess technical, and managerial "rent-seeking" behavior. Review salary levels, incentives resources available to change behavior, need for training, new information systems, or other resources, inducements and sanctions Build strategic Involve planners and managers in research and analysis implementation process of ways to execute policy. Identify networks of supporters for policy chonge, clinical champions Manage uncertainty Promote public awareness campaigns Institute mechanisms for consultation, monitoring, "fine- tuning." Source Walt, "A Framework for Discussion There are similarities between the approach set out in table 2.1 and Reich's work on health sector reform.41 Reich directs our atten- tion to policy content, the actors or players (in his terms) involved, and their stances. He also highlights the importance of analyzing the opportunities and obstacles facing reformers and the strategies avail- able to them for overcoming these obstacles. Another consideration that might be added is the importance of monitoring and evaluating the impact of policy. Monitoring and evaluation contribute to policy learning. They are essential when implementation is viewed as evo- lution, as we do. Returning to the beginning of this discussion, we stress that poli- cymaking and implementation must be seen as part of the same 100 ° Innovations in Health Service Delivery process. In circumstances of bounded rationality, institutional com- plexity, and conflict over values, policymakers have no alternative but to adjust course during the implementation process, seeing the de- livery of change as an iterative experience in which feedback from earlier initiatives informs subsequent developments. This applies as much to the technical design of policy (e.g., institutional arrange- ments and payment mechanisms) as to the polincs of implementa- tion (e.g., anticipating and overcoming stakeholder opposition). Im- plementation is more likely to be effective if the inherently messy nature of the policy process is explicitly acknowledged. Notes 1. J. Pressman and A. Wildavsky, Implementation (Berkeley: Uni- versity of California Press, 1973). 2. B. Hogwood and L. Gunn, Policy Analysisfor the Real World (Ox- ford: Oxford University Press, 1984). 3. M. Lipsky, Street-Level Bureaucracy (New York: Russell Sage, 1980). 4. T. R. Marmor, Understanding Health Care Reform (New Haven, Conn.: Yale University Press, 1994), p. 217. 5. G. Walt, "Implementing Health Care Reform: A Framewvork for Discussion," in R. Saltman, J. Figueras, and C. Sakellarides, eds., Critical Challenges for Health Care Reform in Europe (Buckingham, U.K.: Open University Press, 1998). 6. M. Reich, "Applied Political Analysis for Health Policy Re- form," Current Issues in Plublic Health 2: 186-91 (1996). 7. M. Reich, "The Political Economy of Health Transitions in the Third World," in L. C. Chen and N. Ware, eds., Health and Social Change in International Perspective (Boston, Mass.: Harvard School of Public Health, 1994). Implementing Organizational Reforms to Hospitals * 101 8. J. March and J. Olsen, Rediscovering Institutions: The Organiza- tional Basis of Politics (New York: Free Press, 1989); P. Pierson, Dis- mantling the Welfare State? (Cambridge: Cambridge University Press, 1994). 9. A. Wildavsky, Speaking Truth to Power: The Art and Craft of Pol- icy Analysis (Boston, Mass.: Little, Brown, 1979). 10. C. Hood, "A Public Management for All Seasons," Public Ad- ministration 6: 3-19 (1991). 11. OECD (Organisation for Economic Co-operation and Devel- opment), The Reform of Health Care: A Comparative Analysis of Seven OECD Countries (Paris: OECD, 1992); OECD, The Reform of Health Care Systems: A Review of Seventeen OECD Countries (Paris: OECD, 1994); C. Ham, Health Care Reform: Learningfrom International Ex- perience (Buckingham, U.K.: Open University Press, 1997); R. Salt- man and J. Figueras, European Health Care Reform: Analysis of Current Issues (Copenhagen: World Health Organization, 1997). 12. A. S. Preker and R. G. A. Feachem, Market Mechanisms and the Health Sector in Central and Eastern Europe, World Bank Technical Paper 293, Washington, D.C., 1996; and N. Barr, ed., Labor Markets and Social Policy in Central and Eastern Europe (Oxford: World Bank/ Oxford University Press, 1994). 13. S. Bennett, S. Russell, and A. Mills, "The Changing Role of Government in Adjusting Economies," Paper 3: Health Sector (summary paper presented at a workshop at the University of Birm- ingham, March 28-30, 1999). 14. M. Moore, Creating Public Value (Cambridge, Mass.: Harvard University Press, 1995). 15. N. Machiavelli, The Prince and the Discourses (New York: Ran- dom House, 1950), p. 21. 16. Pierson, Dismantling the Welfare State? 102 o Innovations in Health Service Delivery 17. Ibid. 18. H. Heclo, Modern Social Politics in Britain and Sweden (New Haven, Conn.: Yale University Press, 1974); R. Klein, The New Poli- tics of the NHS, 3d ed. (London: Longman, 1995). 19. J. W Kingdon, Agenda, Alternatives and Public Policies, 2d ed. (New York: HarperCollins, 1995). 20. Pierson, Dismantling the Welfare State? 21. Moore, Creating Publhc Value. 22. Bennett, Russell, and Mills, "The Changing Role of Govern- ment." 23. Ibid. 24. R. Alford, Health Care Politics (Chicago: University of Chicago Press, 1975). 25. Reich, "Applied Political Analysis for Health Policy Reform." 26. A. Arnell, "The Monopolistic Integrated Model and Health Care Reform: the Swedish Experience," Health Policy 37: 19-33 (1996). 27. T E. D. van der Grinten, Scope for Policy: Essence, Operation and Reform of the Policy System of Dutch Health Care (Rotterdam: Erasmus University, 1996). 28. C. H. Tuohy, "Dynamics of a Changing Health Sphere: The United States, Britain and Canada," Health Affairs 18 (3): 114-34 (1999). 29. Klein, The New Politics of the NHS. 30. Ham, Health Care Reform: Learningfrom International Erperience. 31. C. Ham and M. Brommels, "Health Care Reform in the Netherlands, Sweden and the United Kingdom," Health Affairs 14: 106-19 (1994). Implementing Organizational Reforms to Hospitals * 103 32. Bennett, Russell, and Mills, "The Changing Role of Govern- ment." 33. E. Immergut, Health Politics. Interests and Institutions in Western Europe (Cambridge: Cambridge University Press, 1992). 34. E. Ferlie, L. Ashburner, L. Fitzgerald, and A. Pettigrew, The New Publtc Management in Action (Oxford: Oxford University Press, 1996). 35. Bennett, Russell, and Mills, "The Changing Role of Govern- ment." 36. Moore, Creating Public Value. 37. H. Mintzberg, "Crafting Strategy," Harvard Business Review 65(4): 66-75 (1987). 38. Ferlie, Ashburner, Fitzgerald, and Pettigrew, The New Public Management in Action. 39. Ibid., pp. 230-3 1. 40. C. Ham, "Where Now for the NHS Reforms?" BMJ 309: 351-52 (1994). 41. Reich, "Applied Political Analysis for Health Policy Reform." CHAPTER 3 Evaluating the Impact of Organizational Reforms in Hospitals Mead Over and Naoko Watanabe As evidenced by the studies presented in this volume, many govern- ments around the world are attempting to reform the management of state-owned hospitals, and others are considering such action. The political alliance supporting reform typically unites proponents of fiscal responsibility, who are searching for ways to reduce govern- ment deficits, with health sector advocates, who hope that reform will improve the quality of the health care these institutions provide. However, reforms have opponents as well as supporters. Unless hos- pital organizational reform demonstrably achieves at least a portion of its intended effects, those among the general public who might benefit from those achievements will lose interest in the reform, and opponents of the process will gain control. The case for hospital organizational reform varies from country to country, but a stylized version of that case might be expressed as fol- lows. Publicly owned and operated hospitals are generating large and growing budget deficits that the government (or a group within the government) has decided it can no longer afford. Privatization of the hospitals, though a possibility, would be politically costly for three reasons. First, the general population would perceive privatization as government's reneging on its past assurances of free or low-cost health care. Second, many public hospitals serve the entire national health care system in teaching or referral roles, and privatization 105 106 o Innovations in Health Service Delivery might jeopardize this service. Third, privatization would typically be accompanied by firing large numbers of hospital staff, a potent polit- ical interest group. A middle course, which governments hope can either avoid these pitfalls or deflect criticism regarding them, is to retain public ownership of the hospital but introduce organizational reforms that induce the hospital to operate more efficiently, while continuing to meet most of its perceived public duties. In this book, this middle course is termed corporatization of the hospital. To succeed, hospital organizational reform needs sustained politi- cal support from the general public that is strong enough to resist pressures from interest groups. The political support recipe will vary from situation to situation, but reformers who provide their sup- porters with information demonstrating the program's achievements will be more successful at sustaining support than those who do not.1 To ensure the constant flow of information demonstrating the suc- cess of reform or to diagnose problems, the reform process should include monitoring and evaluation mechanisms among its core func- tions. The four politically sensitive aspects of reform highlighted above-financial performance, access to care, fulfillment of public duties other than direct patient care, and personnel management- should be among the dimensions measured by the monitoring and evaluation system. We do not assume that organizational reform of public hospitals will improve every dimension of their performance. One kind of reform may improve one set of indicators, while another kind may improve a different set. Some kinds of reform may fail to improve any dimension of hospital performance-or may impair performance. We therefore propose indicators for judging whether hospital organizational reform affects hospital performance and, if it does, for understanding what each dimension of reform contributes to the overall result. Analysts interested in evaluating hospital reform must be able to distinguish between changes in performance due to the reform related to the hospital itself and changes caused by other shifts in the hospital's environment. They must characterize both the reform itself and the hospital performance that it is supposed to improve. Tracking the management-reform process inside the hos- Evaluating the Impact of Organizational Reforms in Hospitals * 107 pital is a useful way of learning the most from the reform, in lieu of ideal measures of performance. Therefore, we develop four related sets of indicators. One set describes the hospital's environment, a second set characterizes the organizational reform intervention, a third set depicts the response by hospital managers to the re- form, and the fourth set measures the reform's impact on hospital performance. Corporatization as a Middle Path The health sector is not the only sector that has been attempting to improve the management of government-owned firms through organizational reform as an alternative to privatization. Bureaucrats in Business (B in B), a recent World Bank book, considers whether governments can induce state-owned enterprises (SOEs) to improve their efficiency by exposing them to increased competition in their product and factor markets and by weaning them from financial pro- tection via "hard" budgets and other disciplinary controls.2 These two strategies are also emphasized in ra Millstein's recent applica- tions of lessons from corporate governance in the private sector.3 Both strategies are also likely to be important components of any successful effort to improve the efficiency of state-owned hospitals, especially in countries that have sufficient demand and enough trained medical manpower to sustain a thriving, high-quality private hospital sector. Competition and financial discipline cannot change hospital per- formance, however, unless managers have sufficient independence and authority to alter hospital behavior. For other sectors, B in B concludes, "[M]any countries do not give SOE managers the power to react to competition with appropriate measures, such as laying-off excess workers, seeking cheaper suppliers, stopping money-losing services or searching for new markets."4 Therefore, understanding the progress of any hospital organizational reform effort means observing not only hospital performance (as measured by changes in inputs, processes, outputs, and outcomes) but also the environment 108 0 Innovations in Health Service Delivery within which hospital managers make their decisions. Bureaucrats in Business identifies as important four aspects of that environment: o The external economic conditions that determine the demand for the hospital's services and its supply of important factor inputs (e.g., medical manpower, drugs, equipment) o The country's policy regime, which influences the ease with which hospitals can enter or exit a market and many other aspects of doing business o The type of relationship government has with the hospital man- agers-Is it an arm's-length relationship where government keeps its commitments both to reward and to punish? Does it grant autonomy to hospital managers both de jure and, especially, de facto? o The degree of monitoring by local stakeholders, who exercise "voice" to influence the hospital's performance. Like the present volume, B in B used case-study methods to eval- uate the impact of organizational reforms on a small sample of SOEs. Of the 12 enterprises studied, B in B found that only three showed a substantial improvement in total factor productivity after gov- ernment introduced formal performance contracts for the firms' managers. Of the companies, six continued their past unproductive performance and "three performed substantially worse under con- tracts than before."5 The authors concluded that the performance contracts had failed primarily because government systematically "reneged on written promises, sometimes every year."6 If we were to apply the conclusions from Bureaucrats in Business to the hospital sector of developing countries, we would not advocate corporatization to improve hospital efficiency. However, there are several reasons to avoid such a generalization. First, the B in B find- ings were based on a small sample of firms, none in the health sec- tor. Second, the authors have set aside the arguably more favorable outcome of subjecting managers to performance contracts in China: "[S]ome studies have found that the greater autonomy allowed under Evaluating the Impact of Organizational Reforms in Hospitals * 109 the contracts contributed to an improvement in total factor produc- tivity" in Chinese SOEs.7 The authors point to the close monitoring by concerned local stakeholders that characterizes China's "township and village enterprises" as possibly responsible for the much faster growth of this subset of China's publicly owned enterprises. Fur- thermore, Millstein reports without elaboration that "a number of countries such as New Zealand, the United Kingdom, Chile, Swe- den, and the [Republic ofl Korea have tried, with some success, to reform SOEs by introducing internal and external incentives."8 And a study of Korean public enterprises reports their generally positive results with performance contracts.9 In light of this account of experience in other sectors, do specific institutional and political features of the hospital sector in develop- ing countries militate for or against the success of their hospital orga- nizational reform efforts? On the positive side, public hospitals in some countries are likely to be subject to close monitoring by con- cerned local stakeholders such as contributed to the success of the Chinese town and village enterprises. On the negative side, govern- ments are likely to find the imposition of hard budget constraints and binding managerial rewards and punishments even more politically difficult in the hospital sector than they have been in other sectors, partly because the public considers that state-owned hospitals serve socialfunctio77s and should not be subject to the profit-and-loss calcu- lations of private business. Recognizing and Protecting the Social Functions Organizational reform is likely to change the mix of hospital services, with potential to either improve or worsen hospital performance rel- ative to social objectives. Exposed to competition in its product and factor markets, any SOE will produce more of the products and ser- vices for which customers are willing to pay more than their cost of production, and less of those that are less in demand relative to their cost. In the case of an SOE producing electricity or telephone ser- vice, which are essentially private goods, the output mix resulting 110 ° Innovations in Health Service Delivery from increased responsiveness to market incentives is likely to be socially beneficial. However, the markets for the products of state-owned hospitals are distorted by two sets of forces that destroy the presumption that changes in service mix resulting from increased competition in those markets will necessarily increase overall welfare. First, third-party payers subsidize the cost of individual services to different degrees that may not reflect the relative social valuation of those services. Second, some of the outputs of a state-owned hospital serve objec- tives beyond those of the individual patient who receives the services. We refer to these broader objectives as the social functions of the hospital. 10 Social functions that a state-owned hospital may pursue include the production of "public goods" and goods with "positive external- ities," which economic theory predicts would be underproduced under perfect competition." Examples of public goods include teaching and research, laboratory support to epidemiological sur- veillance, and health education services to patients and the commu- nity, all of which produce or disseminate information. Examples of goods with positive externalities include immunization programs, family planning programs, and treatment of communicable diseases. Free or highly subsidized care at state-owned hospitals serves the additional social function of removing barriers to access by the poor- est and of providing a de facto health care safety net for people who lack catastrophic health insurance. 12 Sometimes public hospitals sup- port other social programs. For example, Punjabi provincial hospi- tals house the local offices of Pakistan's social welfare program. Millstein suggests that reforms of SOEs are likely to have a greater chance of success when social functions can be transferred "from the enterprise to municipal or central government authori- ties. "13 As part of hospital organizational reform, it will be useful to attempt to unbundle hospital products into private and social func- tions, assigning the reorganized hospital the responsibility for pro- ducing the private functions and transferring social functions to gov- ernment authorities. For example, the social welfare offices housed by the Pakistani state-owned hospitals might logically be moved to Evaluating the Impact of Organizational Reforns in Hospitals * 111 other quarters. I-Towever, to a greater degree than for SOEs like elec- trical or telephone utilities, a hospital's public and private products are often inextricably linked. Thus, although teaching surgery is arguably a social function of a hospital and delivering surgical ser- vices to patients a private function, setting up a separate institution to teach but never perform surgery would be difficult and damaging to both the teaching function and care provided. Because the output of a typical publicly owned hospital is sub- stantially a social function, two implications for the evaluation of hospital organizational reform become evident.14 On one hand, this feature of hospitals makes management contracts much more diffi- cult to design and monitor. Govermnent supervisors must not only measure the hospital's total factor productivity (e.g., quality-adjusted patients served per index unit of inputs) but also track the hospital's performance in granting access to the poor, in serving as a national center of excellence in health care, in controlling infectious disease outbreaks, and the like. On the other hand, the extent to which this public role is a joint one with the hospital's private role argues a strong case in favor of parallel reforms durmg privatization of such organizations and suggests that the benchmark for such organiza- tional reforms is not only profitability. Research Scope The subject of organizational reform has broad applicability in the health sector. It could apply to government-owned clinics, to health insurance functions, to government pharmaceutical supply oper- ations, or even to medical schools. This chapter focuses on the government-owned hospital. We view the hospital as an organizational unit. Although for some purposes, such as the analysis of the quality of care, the various med- ical and support services should be separately analyzed, this chapter's overarching purpose of evaluating the reform process does not per- mit full treatment of the details involved in evaluatmg any individual hospital service. Thus, in designing an evaluation, the analyst is 112 o Innovations in Health Service Delivery advised to seek technical guidance from local specialists in the specific medical and support services most affected by the specific reform. Information requirements for routinely monitoring an enterprise, whether private or government-owned, are less demanding than those for evaluating the success of a policy experiment. The frame- work and indicators presented here are intended primarily for use in evaluating an initial experiment in hospital organizational reform over a limited trial period. After that, the government and public will decide whether-or in what form-the experiment should be gener- alized to other state-owned hospitals. Since hospital organizational reform is expensive and time-consuming, it will be in the interest of the public to gather more information about the reform process and its impacts than would be necessary or desirable in the context of more routine monitoring. Adapting this framework for use in rou- tine monitoring by hospital managers or by government regulators is a large task, one requiring careful consideration of the specific con- text and the costs and benefits to users (inside or outside the hospi- tal) of each item of information. This task deserves separate study.'5 Although we believe that formal, rigorous evaluation of hospital organizational reform is useful for sustaining political support for the reform process, we do not attempt to marshal evidence in support of this conjecture. Other authors better grounded in political economy may address this issue as the body of accumulated evidence from around the world expands. Concepal [Framework The framework proposed here can be used to analyze the impact of hospital organizational reform on hospital performance in low- and middle-income countries. The need is not for a complex model of hospital or firm behavior but rather for the simplest possible concep- tual framework that highlights the areas where measurement will be most useful and identifies the categories of variables to be measured. In the tradition of social experimentation and program evalua- tion, organizational reform can be thought of as an intervention Evaluating the Impact of Organizational Reforms in Hospitals * 113 imposed on the hospital within a certain policy environment in the expectation of a behavioral response by hospital managers and staff, which leads to improved hospitalperformance.16 Figure 3.1 presents a conceptual diagram of the presumed causal relationship from the organizational reform intervention through changes in the hospital production process to a response in the form of changed hospital performance. The evaluation of a hospital organizational reform can be broken into four parts, corresponding roughly to the four parts of figure 3.1: * Taking account of other outside determinants of behavior that constitute part of the reform environment * Characterizing the organizational reform (intervention) * Describing the resulting change in hospital behavior (response) * Measuring the impact of the reform on hospital performance from society's perspective (impact). Each of these four parts of the evaluation is necessary to see whether a reform has any effect on hospital performance and why performance is better in some dimensions than in others. Only if policymakers learn these lessons from the reform can they confi- dently proceed to apply similar reforms to other hospitals. Environment: Taking Account of Background Determinants of Hospital Behavior The outcome of hospital reform depends on background or exoge- nous features of the hospital's external environment as well as char- acteristics of the specific intervention. The dividing line between dimensions of the environment and dimensions of the interven- tion is somewhat arbitrary, because policymakers have great lati- tude to choose which features of the hospital's circumstances they will change as part of the organizational reform. Three dimensions of the hospital's circumstances, though usually excluded from ex- plicit attention in a hospital organization reform effort, need to be -c 2: ~5D . E~~~~~~~~~~c o~~~~~~~~~ '1, ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ 0)~~~~~~~~~~~~~~ C: 0~ ~ ~ ~~~~~ 0 E~~~~~~~~ oEoE 0) ~~~~~O m ) 0~~~~~~~~~~ 0~~~~~~~~~~~~~ E~~~~~~~~- 0 0~~~~~~~~~~~0 ~~~~~~~ 1140 C Evaluating the Impact of Organizational Reforms in Hospitals * 115 considered: market environment, finance and payment systems, and government oversight. These dimensions were described in chap- ter 1 as the "three systemnc factors" that jointly determine the hos- pital's incentive regime. Market environment. For any level of demand for hospital services in a catchment area, the impact of reform on a specific hospital depends on the market environment. Suppose that managers of the corpora- tized hospital raise user fees without improving quality of care. The degree to which patients respond to this kind of reform by taking their business elsewhere depends on the proximity, price, and qual- ity of available substitutes. Thus, comparing the impact of price increases on two hospitals without observing each hospital's position in its own health care market vis-a-vis alternative providers is likely to lead the analyst astray. Part of the data collected on any hospital reform project should be the size of the local catchment area of patients, the total number of outpatient visits and inpatient bed-days demanded of all hospitals in the catchment area, and each reforming hospital's market share of these totals. Similar care should be taken to record features of supply and demand in the markets for person- nel and other inputs. Finance and payment systems. The four options for hospital financing are direct financing from the government budget, contracts between the government or social health insurance agency and the hospital, contracts with private insurers or employers, and direct payment by consumers from their own pockets. In many countries, the govern- ment's most urgent motive for hospital reform is to shift the financ- ing system from the first option toward one or more of the other three alternatives. To the extent that the source of financing shifts to patients' pockets, the reform shifts not only the cost but also the risk inherent in health care expenditures from the government to private citizens. In exchange for bearing a larger burden of cost and risk, patients who can choose among several hospitals will be able to put market pressure on those hospitals to improve the quality of their services. The net effects of these changes on the welfare of con- 116 o Innovations in Health Service Delivery sumers are likely to vary by income class. While bearing a larger share of cost and risk hurts the poor more than the rich, shifting the purchasing power to patients sometimes benefits the poor more than the rich-as in situations where the supposedly "free" medical care is rationed out to individuals with "connections" or those able to make informal side-payments. The availability of third-party payment mechanisms (including any of the first three types of hospital financing mentioned above) will reduce the sensitivity of demand for care to the price received by the hospital and increase the demand for care at any given hospital fee schedule. Reducing price sensitivity may be a critical element in permitting hospital managers to respond to corporatization by rais- ing user fees on certain services without losing revenue. But, in some cases, third-party payment mechanisms may incorporate preferential arrangements either for or against the corporatized hospital. Thus, information must be gathered on the nature of hospital financing mechanisms that prevail in the market served by a corporatized hos- pital, including the share of costs financed by each method and any changes occurring in these shares, to hully understand the effect of reform on utilization and revenue. Government oversight. Managers of a state-owned hospital, or indeed any enterprise, are held accountable to several different constituen- cies for their own performance and for that of the enterprise. As in chapter 1, box 1. 1, of this volume, we use the term governance to refer to the set of formal and informal institutions that collectively deter- mine how and to whom managers and staff are accountable. Since we are focusing on evaluation of the organizational reform, we must distinguish between aspects of the governance structure that remain constant as part of the reform environment and aspects that are explicitly changed by the reform. Although each organizational reform is distinctive, reforms often make substantial changes in internal reporting mechanisms while holding constant the external government regulatory structure. Therefore, for our purposes, we separate the discussion of governance into two pieces. Here we use the term government oversight to refer to the external government Evaluating the Impact of Organizational Reforms in Hospitals * 117 regulatory structure. Later, in discussing intervention, we use the rubric accountability to discuss internal reporting mechanisms (includ- ing the presence or absence of a board of directors) that organiza- tional reforms often change. Government oversight could be exercised by a formal government regulatory body charged with hospital regulation or by direct report- ing by hospital managers to ministry of health bureaucrats.'7 The degree to which hospital managers are accountable to a regulatory board or to government officials affects the success of any organiza- tional reform. Regulatory bodies from line ministries to independent regulatory boards are notorious for often being "captured" by the industry they regulate. Therefore, indicators of the quality of gov- ernment oversight should include measures of the regulatory bodies' degree of capture. Dimensions of Organizational Reform: Characterizing the Intervention If the managers of public sector hospitals could be placed into envi- ronments that emulate perfecdy competitive product, factor, debt, and asset markets, there would be no need to collect data on hospi- tal performance. Pareto optimal performance would theoretically be guaranteed.'8 However, our discussion of corporatization as a mid- dle road between traditional public sector management practices and complete privatization made clear that accurate emulation of a com- petitive planned market is a difficult-to-attain goal for organizations whose products cannot be cleanly unbundled into private and social components. Such reforms may be undesirable if not done in such a way that the negative aspects of the reform are mitigated. Other imperfections contaminate every country's market for hospital ser- vices but are more severe in low- and middle-income countries. They include monopsonistic factor markets, asymmetrically distributed information, and a large, minimally efficient scale of production (in comparison to the local demand for sophisticated tertiary hospital care). Thus, the analyst should not expect a "good" hospital organi- zational reform initiative to achieve the highest scores on all five 118 o Innovations in Health Service Dehvery dimensions of organizational reform listed on the left in figure 3.1 and in box 3.1. The point of measuring these dimensions is to char- acterize the degree to which the reform process emulates the condi- tions of a perfectly competitive enterprise in order to understand how much hospital performance can be improved without privatiza- tion and at what cost.19 A caveat merits mention: no matter the fanfare with which policy changes are announced, hospital reform is not always implemented as planned. Whether hospital reform works is likely to depend not only on the formal granting of autonomy in these dimensions but also on whether hospital managers themselves feel autonomous. For example, in the education sector, King and associates have found that educational outcomes are unaffected by the granting of de jure autonomy but improve significantly when school administrators attain actual autonomy.20 Analysts are therefore urged to measure actual autonomy on the five dimensions listed in box 3.1 to charac- terize organizational reform. Bon 3. Dimensions of Organizaliona Refom Managerial decision rights with respect to o Strategic management and the formulation of institu- tional objectives o Procurement o Financial management o Human resource management o Administration (clinical and nonclinical) O Innovation and learning Market exposure o Information asymmetries. Disclose results of independent monitoring of hospital performance to reduce asymme- try of information between hospital and consumers (in the product market) and between hospital and owners (in the equity market). Evaluating the Impact of Organizational Reforms in Hospitals -119 * Product miarket. Enable freedom of entry for competing private hospitals. o Labor market. Eliminate or reduce civil service con- straints on personnel policies. Residual claimant status e Impose a hard budget constraint. * Allow hospital to raise capital by seLling shares to, or borrowing from, private investors. Subject hospital to bankruptcy laws that limit private and public shareholder risk. Accountability of managers and staff o Improve voice. On hospital board of directors, include representatives from private nonhospital sector, insur- ance industry, local community and nongovernmental organizations. • Require independent audits and publish results in clear, understandable language. o Require formation of one or more quality-of-care com- mittees to monitor work of all physicians delivering care in hospital. Social functions/public goods * Identify public and private goods/services in perfor- mance contract. o Unbundle private from public goods and, where possi- ble, transfer public goods to government entities. o Subsidize or regulate remaining social functions. Sources: Authors' construction based on Millstein, "The Corporation as an Ownership and Management Struc- ture"; World Bank, Bureaucrats in Business; and Shirley, "Performance Contracts." 120 o Innovations in Health Service Delivery Grant decision rights to managers. Managerial autonomy is the most obvious feature of hospital organizational reform. As discussed above, the relationship between the government and hospital man- agers plays a critical role in all organizational reform efforts. Each reform can be characterized by the magnitude of control shifted from the hierarchy, or supervising agency, to the hospital. Critical decision rights transferred to management may include control over inputs, labor, scope of activities, financial management, clinical and nonclinical administration, strategic management (formulation of institutional objectives), market strategy, and sales. Expose managers to markets. The third key element of the high-pow- ered incentives sought in these reforms involves the extent to which hospitals earn their revenue in the market instead of relying on direct budget allocation. The question is, "How much of its services does the hospital deliver or sell to earn its own revenue?" Hospital ser- vices identified as largely private goods may be deregulated, reducing barriers to entry that might prevent private firms from competing in these markets. Organizational reform can further be characterized by the degree to which it exposes the corporatized hospital to mar- ket forces in the input (factor) and equity markets. Entry barriers to the medical or nursing professions, other than professional school tuition or professional qualifications, may be reduced, as might any advantages to employment by the corporatized hospital relative to employment by its private sector competitors. Some reforms (i.e., in New Zealand) have compelled reformed hospitals to obtain invest- ment funds on the same terms as private hospitals. Impose residual claimant status. Financial discipline is notoriously lax in SOEs of all sorts, and hospitals are no exception. Managers are unlikely to make the best of any autonomy they are granted unless they face a hard budget constraint from their owners. Some reform- ers have therefore attempted to create "hard" budgets by establish- ing new and clearer rules that are mutually understood in advance of the budget period for funding deficits and taxing surpluses. Others Evaluating the Impact of Organizational Reforms in Hospitals * 121 have created categories of services or users whose payments are allowed to be kept in the hospital. These types of changes are in- tended largely to give hospitals residual claims on revenue-to moti- vate their interest in enhancing efficiency and lowering costs. Improve accountability of managers and staff The reforms are also char- acterized by the degree to which accountability for achieving objec- tives is based on hierarchical supervision of the organization versus rules, regulations, or contracting. As decision rights are delegated to the organization, the government's ability to assert direct accounta- bility (through the hierarchy) diminishes. Partially, accountability is intended to come from market pressures, seen as generating a non- political, nonarbitrary evaluation of organizational performance, at least its economic performance.21 If the government is a purchaser, accountability will also be pursued via the contracting and monitor- ing process. Many reform efforts have created supervisory boards to allow for monitoring of hospital and management performance, while still providing for day-to-day autonomy for managers. Some reforms have tried to support the creation of internal accountability by requiring the formation of internal quality-of-care committees in reformed hospitals. Rules and regulations regarding the hospital operation constitute an alternative form of accountability mechanism. Therefore creating or strengthening such mechanisms is a fourth critical element of organizational reforms that reduce the use of traditional, hierarchi- cal accountability mechanisms. Define and protect socialfunctions. The final critical factor characteriz- ing these organizational reforms is the degree to which "social func- tions" delivered by the hospital shift from being implicit and unfunded to specified and directly funded.22 As reforms motivate the hospital to focus more closely on financial viability, management will move to decrease output of services that do not at least pay for them- selves. The financial bottom line thus undermines the ability to cross-subsidize certain services internally.23 Organizational reforms 122 o Innovations in Health Service Delivery must therefore create alternative mechanisms to ensure that previ- ously cross-subsidized services continue to be delivered (via explicit funding, demand-side subsidies, insurance regulation). Defining such social functions requires a concerted attempt to dis- tinguish hospital services that are purely private goods from those that have a social function and to define and subsidize the hospital's social functions. To the extent this effort succeeds, the chances of achieving clear improvements in the hospital's performance are greatly increased. Of course, reasonable people can disagree regard- ing the degree to which an individual hospital service can be consid- ered a social function. Some would argue that all care for noninfec- tious disease is a "private good," because it benefits primarily the recipient, and anyone who does not pay for it can be inexpensively barred from consuming it. Others would counter that treatment of chronic disease among the poor fills the social function of providing a safety net for the most destitute. Still others argue that access to all health care services is a human right, which the government cannot morally deny any citizen. Despite these differences of view, a working consensus may be possible regarding a ranking of hospital services from least to most private. Organizational reform designers may use such a ranking to transfer services agreed to be primarily social functions to govern- ment health care services, regulate their production, or compart- mentalize them within an administratively separate part of the hos- pital (thus insulating the rest of the hospital from their production costs). Useful indicators for characterizing this dimension of the inter- vention are whether or not the issue of protecting social functions has been addressed and, for each identified social function, the nature and degree of protection implemented. Dimensions of Hospital Behavior: Measuring the Response Organizational behavior can be described as the set of formal and informal administrative rules and procedures for selecting, deploy- ing, and supervising resources in the most efficient way to achieve Evaluating the Impact of Organizational Reforms in Hospitals * 123 institutional objectives.24 Given the degree of de facto autonomy granted to the organization, the key to successful hospital operations lies in the response of hospital managers to these directed changes. Specifically, response would be measured by investigating whether hospital managers exercise their new independence by experiment- ing actively and purposefully with innovative policies and practices in the areas of finance, marketing, human resources, procurement, business management strategy, and medical management strategy (box 3.2). This more active managerial involvement has substantial costs both in managerial time and in the form of occasional managerial errors. A key hypothesis of the organizational reform movement is that the improved efficiency resulting from this experimentation will reap benefits that exceed these costs. Finance. As explained earlier, governments are often compelled to undertake hospital reform to address inefficiency, insolvency, and financial mismanagement. In most cases, the hospital's financial health is therefore a foremost concern of hospital managers. Brad- ford and Tiscornia argue that hospital managers need not be fi- nancial experts.25 A good financial manager must simply be able to communicate with financial specialists, asking meaningful questions that lead to sound financial policy. The fundamental questions to be posed would encompass such topics as financial statements, capital assets, profit, debt, pricing and payment, and long-term financial management. Formulatng and asking the right questions will enable managers to select and subsequently monitor financial policy. Marketing. Kaplan and Norton argue that corporate managers have relied too much on traditional financial measures.26 These measures reveal past accomplishments but cannot foretell the organization's future directions. Therefore, a balanced evaluation framework has to include measures of dimensions that can serve as leading indicators. One of these dimensions, conventionally known as marketing, mea- sures business operations from a customer's perspective, with a view toward assessing long-term potential for profitability and growth. 124 ° Innovations in Health Service Delivery Box 3.2 Dimensions of Hospital BehavEor Finance o Financial statements o Capital assets o Profit o Debt o Pricing and payment o Long-term financial management Marketing o Customer acquisition o Customer satisfaction o Customer retention o Customer profitability Human resources o Employee selection o Employee retention o Employee satisfaction o Employee productivity Procurement o Selection of equipment and supplies o Productivity of equipment and supplies Business management strategy o Defining the vision o Communicating and learning o Business planning o Feedback, innovation, and learning Medical management strategy o Functional status and well-being o Physiological and biomedical measurement o Cost-effectiveness of health care delivery o Patient satisfaction with care Sources: Kaplan and Norton, "The Balanced Scorecard"; Heskett, Sasser, and Schlesinger, The Service Profit Chain; Tarlov and Colley, "Clinical Outcome Management"; Zan- der, "Critical Pathways." Evaluatng the Impact of Organizational Reforms in Hospitals * 125 Heskett and associates assert that various components of customer attitudes and behavior are connected like a chain linked to business performance.27 Profit and growth are linked to customer loyalty, which results from customer satisfaction. Customer satisfaction is determined by quality and price. For hospitals, customer attitudes will be affected by what they hear from their physicians, relatives, friends, and neighbors. Hospital reform should ideally improve all of these indicators. Human resources. This chain of customer attitude and behavior can feed into another chain in the human resources dimension. Heskett and associates call the whole system the service-profit chain.28 In this service-profit chain of causation, the service value is created by employee productivity, which is derived from employee loyalty. Loy- alty is in turn the product of employee satisfaction, which results from the internal quality of work life. The service-profit chain grows in importance as a business becomes more service-oriented since, by nature, hospital services require direct interaction between care providers and their patients. Satisfied employees are more likely than disgruntled employees to give patients good service and to do it more pleasantly, therefore providing utility directly to the customers. Although producing satisfied employees is not the objective, fail- ure to satisfy physicians and other hospital workers can create polit- ical obstacles to hospital organizational reform. Thus, ensuring that the employees who stay with the hospital during the reform are the ones who can contribute to and gain from the reform will help pro- tect the reform process from political backlash. Employees forced out of the hospital or otherwise disadvantaged by the reform may oppose the reform for personal reasons. Some- times reform deprives an individual of opportunities to profit ille- gally from selling drugs or other public property. In other cases, reform might simply force workers to exert more effort or be pres- ent a larger part of the work day. Protecting a well-performing reform from misguided or misleading criticism is one reason for thorough, timely evaluation of organizational reform, using methods described here. To the extent that the evaluation reveals problems in 126 o Innovations in Health Service Delivery the reform process that support reform opponents' claims, reform and public health are both best served by revealing these problems and discussing their solutions in public. Procurement. Procurement refers to purchasing procedures for hospi- tal equipment and medical and nonmedical supplies. These inputs constitute a large part of factor inputs required for hospital produc- tion processes, especially in developing countries where labor is rel- atively cheap. A cost analysis of hospitals in a group of developing countries by Barnum and Kutzin suggests that the total recurrent cost of drugs and other nonlabor costs ranges from 21 percent (Ni- geria) to 78 percent (China).29 A recurrent-capital ratio averaging 0.20 for another group of developing countries highlights the rela- tive size of the nonlabor component of hospital cost. When hospital management does not face a hard budget, lack of accountability often leads to irrational investment and purchasing decisions. Effective management would address this issue and ensure a sound decision- making process by incorporating the procurement dimension into its evaluation norm. Business management strategy. Hospital management affects all the stages in the hospital production process that influence hospital per- formance. Kaplan and Norton argue that linking an organization's long-term strategy with its short-term actions is the key to success- ful organizational reform.30 As the first step of such strategic man- agement, the senior managers and the hospital board define the organization's mission or, in their terms, its "vision." For practical relevance, the vision must be closely tied to specific objectives and measures endorsed by senior managers. The long-term plan should be defined with input from the marketing department regarding consumer attitudes. It should state the hospital's objectives over the next few years, including its product mix, projected patient mix, tar- get communities, and service quality. The second step is communicating and linking, which refers to dissemination of the mission set by senior managers to every level of the organization to ensure that departmental and individual goals are not limited to short-term financial goals. The next process Evaluating the Impact of Orgamzational Reforms in Hospitals * 127 is business planning, which consists of setting priorities, based on the organization's performance goals, among potentially conflict- ing reform programs so as to allocate scarce resources in the most efficient way. These three processes set the context within which management can define and subsequently monitor managerial and administrative procedures for managing human, physical, and finan- cial resources. The final step is feedback and learning, which enables strategic learning, based on the review of departmental and individ- ual performance. Medical management strategy. In contrast with business management strategy, medical management strategy concerns hospital behavior at the level of direct patient care.31 Its main purpose is to improve clin- ical standards and practice patterns in order to achieve better health outcomes, with cost control a secondary but important concern. Indicators of the quality of medical management can include whether or not the work of individual physicians is reviewed by a quality-control committee of their peers and whether or not the hos- pital is developing and applying a set of recommended "clinical path- ways" for specific, frequently encountered sets of presenting condi- tions.32 Once clinical pathways have been adopted, the quality of medical management can be judged by the degree to which physi- cians apply those pathways in their practice. Dimensions of Hospital Performance: Measuring the Impact The purpose of evaluating hospital corporatization is to determine whether hospital performance improves as a result of the reform process. We propose to value the changes produced by the reform process in four dimensions, which appear at the right in figure 3.1: technical efficiency, allocative efficiency, quality, and equity. To get a good idea of the success or failure of reform and the reasons for those results, we suggest that the analyst examine at least one indicator on each of these dimensions. The indicators themselves can be classified by the dimension of hospital performance to which they contribute and by the stage of the production process where they are observed. Table 3.1 presents 128 o Innovations in Health Service Delivery sbSle 3.1 Hospital Performance Indicators, by Dimension and Production Stage DIMENSIONS OF HOSPITAL PERFORMANCE STAGES OF HOSPITAL PRODUCTION PROCESS TECHNICAL EFFICIENCY ALLOCATIVE EFFICIENCY QUAUTY EQUITY Input Process Appendix 3A contains the complete table Output and outcome a conceptual table of this sort. The columns of the table are the four dimensions of hospital performance (figure 3.1, right-hand side). The rows of the table are the three stages of the production process where measurements can be made. In principle, every indicator of hospital performance can be located in a cell in this table. In reality, locating a specific indicator is sometimes somewhat arbitrary, since the distinctions between dimensions or stages are sometimes fuzzy. In the interest of clarity of exposition, we have categorized all of the indicators we have found in the literature or proposed ourselves into one of the 15 cells defined by table 3.1. Appendix 3A reports the complete classification. Here, we define the dimensions in more detail and discuss examples of useful indicators in each dimension. This section presents the four dimensions of performance intro- duced on the right side of figure 3.1 as the columns of table 3.1: efficiency (technical and allocative), quality, and equity. Since the manifest inefficiency of government-owned and -operated hospitals is one of the most important motivations for hospital reform, track- ing the impact of reform on efficiency is a high priority. However, opponents of hospital reform frequently point to reductions in quality or equity or increased corruption as the side effects of reform that are so detrimental to public well-being as to negate any purported efficiency improvements. Thus, the complete story of the impact of hospital reform can be told only with measures on all four dimensions. Technical efficiency. A production process is technically efficient if it wastes nothing. If a hospital's existing inputs can be reorganized to produce more output with no more resources, the hospital is not operating efficiently. Alternatively, if the existing flow of outputs Evaluating the Impact of Organizational Reforms in Hospitals * 129 could be maintained with fewer or less costly inputs, the hospital is not efficient.33 Technical efficiency can thus be measured by any of a list of indices of hospital output per unit of hospital input (table 3.2). In outpatient clinics, patients seen per clinic-hour is a handy measure. In inpatient wards, average length of stay and bed occu- pancy rate are useful and frequently used measures of technical efficiency. Table 3.2 Technical Efficiency Indicators IINPUT PROCESS FINAL Monetary inputs Capacity utilization Unit cost * Total revenue * Case-mix-adjusted bed * Case-mix-adjusted cost * Total expenditure utilization (average per outpatient visit length of stay, occu- pancy rate, turnover rate) * Expenditures on staff * Case-mix-ad1usted * Case-mix-adjusted cost and drugs capacity utilization of surgical intervention of (other) medical equipment Physical inputs * Case-mix-adjusted cost * Medical staff Inumber of per inpatient case qualified medical staff, percentage absenteeism of medical staff) * Availability and state of Labor productivity Finance medical equipment and * Outpatient visits per * Net operating supplies (percentage of physician per day balance per patient essentials available and * Inpatient cases per * Net operating usable) physician per day balance per bed-day Service mix * Total hospital com- pensation (R) o Marginal cost (MC) o Ri/Ri=MCi/MCj (see Appendix 3A) Health outcome o Perinatal mortality rate • Case fatality rate Note To provide guidance to evaluators under time or budget constraints, the bold type flags indicators that are more difficult or expensive to collect, while the easier-to-collect indi- cators appear in roman type We refer to these indicators, respectively, as comprehensive and rapid As implied by these designations, we believe that the quality of an evaluation will suffer if only rapid indicators are collected 130 ° Innovations m Health Service Delivery In principle, these measures of technical efficiency apply only if the quality of the output (e.g., health outcomes, consumer satisfac- tion) is held constant. To the extent technical efficiency in units of quality-adjusted outputs can be measured, the quality dimension of hospital performance would not need separate attention but could be subsumed within technical efficiency.34 However, because the measurement of "quality-adjusted" output is in fact quite difficult, we follow the practice of other authors and separate quality into a second dimension. Allocative efficiency. While technical efficiency is a matter of achieving the maximum output with any given mix of inputs, allocative efficiency means producing the "correct" mix of outputs using the "correct" mix of inputs. Technical efficiency is doing things right, while alloca- tive efficiency is doing the right things. Earlier we said that society views some hospital services as social functions that deserve protection from marketplace forces. Measur- ing the allocative efficiency dimension of hospital performance requires identifying which hospital services provide more social functions than others (table 3.3). The criteria for selecting these services could be derived from public economics (i.e., services that are public goods or produce positive externalities deserve protec- tion), but government decisionmakers may sometimes identify a hos- pital service as a social function for other more political or subjective reasons. However the hospital's social functions are identified, the allocative efficiency dimension of hospital performance can then be judged by measuring the production of both social function and other services over time. If the hospital's production of social func- tions slips after reorganization, either relative to other hospital out- puts or relative to the population to be served, there is cause for alarm on this dimension. Setting aside the problem of protecting social functions, two spe- cial features of the market for hospital services can impede the achievement of allocative efficiency. First, third-party payment pro- visions may elicit overuse of one service and underuse of another because the ratio of the prices patients pay for the two services bears Evaluating the Impact of Organizational Reforms in Hospitals * 131 Table 3.3 Allocative Efficiency Indicators INPUT PROCESS FINAL Physical input mix Internal rationing mechanism Fees, copayments, and other * Physician/nonphysicial * Price mechanism incentives used to: medical staff * Nonprice mechanism * Limit moral hazard with * Staff/medical supplies respect to nonsocial funcfions Input price ratio * Limit excess demand * Wage of physician/non- * Limit bypossing of PHF physician medical staff * Encourage use of social * Wage of physician in funcfons. public sector/private Reword to provider used to: sector * Improve quolity of care * Encourage production of social funcfion. Service mix Fees out of pocket (P) Total hospital compensation (RI Pi / Pj c Ri / Rj (see Appendix 3A) Number of pafients bypass- ing primary health care PHF, Primary health facility Note See table 3 2 no relation to their relative costs. An example is when patients bypass nearby primary care facilities to go directly to the hospital, because the hospital's more costly care is cheaper to the patient. Overuse of health care in response to third-party payment provisions is called moral hazard. Organizational reform can address it by instituting nominal charges for the more costly service. For example, some reforms introduce "bypass fees" charged to patients who come to the hospital without having been referred by a primary care center. The second pervasive feature of markets for hospital care is asym- metric information between patient and provider. The fact that the patient knows so much less about medical care than the provider, exacerbated by the infrequency with which most patients purchase hospital care, makes the patient incapable of judging the clinical dimension of the quality of the care received. The problem of asym- metric information produces allocative inefficiency, which can be 132 o Innovanons in Health Service Delivery remedied by a combination of regulating providers and disclosing information about quality to the public. Gauging allocative efficiency in earlier stages of the production process is even more difficult than judging it at the final product stage. Harris pioneered the development of techniques to analyze the internal allocation of resources in hospitals, techniques that could be used to examine the impact of hospital reform on internal allocation.35 One allocation issue facing developing countries is what internal mechanisms need to be developed to eradicate the wide- spread practice of requesting side payments from patients for inputs like sheets or soap. Collecting almost any data on these internal allo- cation processes would be useful to track what happens to them dur- ing the reform process. At the input stage of the production process, the most important issue relating to allocative efficiency is the relative compensation received by different categories of personnel. Data on wages paid by the hospital before and after reform must be collected and compared not only across categories within the hospital but also with other comparable employers outside the hospital setting. When these wage ratios do not accurately reflect the relative values of categories of personnel to the hospital production process or are higher or lower than wages paid for comparable work outside the hospital, every area of hospital performance is threatened. Quality. Because of information asymmetry, tracking its impact on the quality of care is an essential aspect of measuring the impact of organizational reform. There is substantial discussion in the litera- ture about how to measure quality of care. Some of the disagree- ments among authors stem from a failure to distinguish measures by the stage of the production process. To avoid these debates, table 3.4 suggests indicators of quality at all three stages of production. The analyst is free to select measures from this menu, keeping in mind that measures of quality at the final stage of production will be more compelling to policymakers or to the public than measures taken from earlier stages. The simple availability of specific drugs has been used as an index of the quality of care in rural primary health care facilities.36 Since it Evaluating the Impact of Organizational Reforms in Hospitals * 133 Table 3.4 Quality Indicators IINPUT PROCESS FINAL Number of qualified Percentage of treatment Health outcome medical staff according tcodefined * Mortality rates adjusted protocol by severity Availability and state Availability of patient * Rate of adverse outcome of medical equipment history for selected severity- (percentage of essentials adjusted conditions available and usable) Pattern of drug * Rate of hospital-caused administration infection liatrogenic Availability and state of disease) medical supplies (percent- * Rate of postoperative age of essentials available infection and usablel * Rate of emergency read- mission within two weeks of discharge * Rate of return to operating theater for same condition * Frequency of changed diagnoses for outpotients * Patient satisfaction Note See table 3 2 cannot capture how effectively drugs are used when they are avail- able, this index is imperfect. In the absence of other measures, how- ever, knowing that some facilities have basic drugs in stock, while others do not, is informative. Drug availability is one possible index of qualhty at the input stage of the health care production process. In the hospital context, such input-stage indexes can be supplemented by process measures of quality such as an index of how well the hos- pital maintains patient records from one visit to the next.37 In a more difficult but more revealing measure of process quality, an expert physician would observe patient treatment and judge the adequacy of treatment protocols. Holland and other authors have emphasized comparison of treatment procedures to accepted standards for qual- ity care.38 The most useful measures of quality include measures of outcome from the final stage of the production process-such as rates of adverse outcome from specific procedures, adjusted for the severity of presenting cases or rates of iatrogenic diseases such as staphylo- ,zoccus infections in hospital wards.39 Abedian and other authors 134 o Innovations in Health Service Delivery have stressed the importance of meeting customers' requirements as a component of high-quality health care.40 Client satisfaction can be measured with exit surveys, which can also be used to reveal the true cost to the patient of hospital care-including side payments. Equity. In a recent review of evidence from case studies of hospital autonomy in five developing countries, Govindaraj and Chawla found that, despite some measurable improvements in technical effi- ciency, "equity and access issues have either worsened or not im- prove[ed] after autonomy."41 Thus, measuring the degree to which reform affects access of the poor to hospital care is a principal task of any evaluation of hospital reform results. An analysis of the equity impact of reform can begin by examin- ing the impact of the reform on the mix of patients to see if either the percentage or the number of poor patients served has fallen.42 This measure of equity at the input stage of the production process can be supplemented by reexamining any of the measures of the quality of care at the process or final product stages to see the distri- bution of the quality index by the patient's income class (table 3.5). For example, are drugs less available to poorer patients than prior to the reform? Or are poorer patients less satisfied with their care? Analyses of the equity impact should not be restricted to the hos- pital undergoing reform but should look at the reform's impact on access by the poor to health care in general. Sometimes the poor are discouraged by higher fees from presenting minor complaints at a hospital's outpatient clinic, but find equally good care in less expen- sive primary health care facilities outside the hospital. In this case, the reduced percentage of poor patients in the hospital's patient mix might be interpreted as an improvement in allocative efficiency, with little offsetting reduction in quality or access. Although collecting data on health care utilization of the entire local poor population before and after hospital reform is expensive and time-consuming, the analyst who wants to measure the impact of reform on equity should collaborate with local statistical offices and poverty programs to identify a more population-based measure of access to care than would be available from the hospital's patients alone. Evaluating the Impact of Organizational Reforms in Hospitals * 135 Table 3.5 Equity Indicators INPUT PROCESS FINAL Rationing of services by Use of medical services Percentage of unfulfilled price * Mean physician visit by needs by socioeconomic Exemption socioeconomic index index Government expenditure * Mean number of use of Health outcome by per patient by socioeco- selected services by socioeconomic index nomic index socioeconomic index Use relative to needs by Financial burden index Rationing of services by socioeconomic index * Ratio of out-of-pocket time * Use-disability ratio health expenditure to * Appointment waiting time * Symptom-response ratio food expenditure by * Office waiting time socioeconomic index Geographical barrier Out-of-pocket expenditure * Poverty gap before and after out-of-pocket Prepayment or insurance health expenditure by Perepaymen socioeconomic index coverage Note See table 3.2 Another way in which superficial analysis might suggest a reform to be inequitable, when its net effect is in fact equity-enhancing, involves the reform's effect on the practice of requesting side pay- ments from patients. The existence of such side payments in many government hospitals and the fact that poorer patients are rarely exempted from their payment means that the hospital's apparently pro-poor pricing policy is much less equitable than it seems. A hos- pital reform that raises official prices but eliminates side payments may appear to raise the price of care to the poor when it actually reduces the net cost of care for that group. Exit polls designed pri- marily to measure patient satisfaction can also capture the effect of reform on the net price to the patient and thus allow inferences about the true impact of the reform on prices paid by the poor. The k.ey issue is the extent to which the reform exposes lower income groups to an increasing financial burden of illness. Trrade-offs between the Dimensions The fact that hospital reforn can be evaluated only by examining its effect on four distinct dimensions of performance implies that a 136 o Innovations in Health Service Delivery reform might do well on some dimensions but poorly on others. What if, as in the cases examined by Govindaraj and Chawla, effi- ciency improves while equity worsens (table 3.6)?43 When hospital reform improves some measures of performance but worsens others, policymakers will ask whether a somewhat dif- ferent reform design could have achieved the observed gains with fewer of the observed costs. Collecting data on every stage of the production process in all these dimensions may give health sector decisionmakers the information they need to answer this question, or will at least suggest alterations in reform design that may achieve the desired gains with fewer offsetting performance reductions. In the last analysis, the policymakers themselves must decide, with input from all the appropriate constituencies, which combination of performance goals for public hospitals best meets the nation's needs. Learning from HospiCal Reform Experience Hospital reorganization is an expensive and time-consuming pro- cess. The direct costs of reform consist of the cost of designing, im- plementing, and monitoring the reform process and the costs of the dislocation that change entails. The choice to reform govern- ment hospital management also implies a choice to postpone explo- Table 306 Hospital Autonomy in Ghana, India, Kenya, and Zimbabwe ADVERSE SOME SUBSTANTIAL IMPACT NO CHANGE IMPROVEMENT IMPROVEMENT Efficiency n a. Zimbabwe, India, n a Ghana Kenya Quality of care n.a Kenya, India, n a. and public Zimbabwe Ghana satisfaction Equity Zimbabwe, India, n.a. n.a Ghana Kenya n.a. Not applicable Source. Govindaral and Chowla, Recent Experiences with Hospitol Autonomy. Evaluating the Impact of Organizational Reforms in Hospitals * 137 ration of other avenues toward improved hospital efficiency. By choosing not to pursue more direct remedies such as restructuring the government health care bureaucracy or hospital privatization, the government and the public incur a substantial opportunity cost in the form of the forgone benefits from those other approaches. In exchange for these direct and indirect costs, it is imperative that the nation learn as much as possible from the hospital corporatization experience. Those who countenance health care reform without suf- ficient monitoring and evaluation to guide and draw lessons from the reform process are courting chaos in the "reformed" system and con- sequent financial and political disasters. Experimental Design and Its Alternatives Learning is not easy. Hospital reforms often take place in a complex and dynamic environment. Any of many variables in that environ- ment can affect the reform process, for good or ill. To have any hope of learning whether hospital reform is working and, more important, which features of the attempted reform are having positive, and which negative, effects on the reform process, the question that must be answered is, "What would have happened in this hospital if orga- nizational reform had not been instituted?" For this, as many as pos- sible of the outside influences on hospital performance that are not themselves part of the reform process must be "held constant" or "controlled for." The rows in table 3.7 represent three different approaches to con- structing the counterfactual, the story of what would have happened without the reform. The three approaches are not mutually exclu- sive. Indeed, combining all three tells the most comprehensive and accurate story. The reflexive comparison adopts the implicit assump- tion that the hospital's behavior and performance prior to the inter- vention would have continued unchanged, so that any differences are due to the reform. The control group approach compares behavior and performance among reforming hospitals to behavior among hos- pitals presumed to be unaffected by the reform, assuming that any observed differences are due to the reform. Randomly assigning 138 0 Innovations in Health Service Delivery TaIb4e 3.7 Methodological Approaches to Assessing Reform's Effects on Hospital Performance TIMEFRAME OF ANALYSIS STATIC DYNAMIC CROSS-SECTION RETROSPECTIVE PROSPECTIVE -a Reflexive comparison A B C ,, Compare reforming hospitals to -a themselves prior to reform. = Control group D E F 0 u Compare reforming hospitals to -E nonreforming ones. 0) .' Statistical control G H I u Exploit variation in intervention mix among reforming hospitals 0 v to control statistically. Note Cell A is empty by definition Source Authors' construction individual hospitals to the control or intervention group is the most powerful way to designate the control group, and the results will be most convincing to policymakers. When the detailed nature of the reform (as described by the five dimensions at the left in figure 3.1) differs from hospital to hospital, this variation can be statistically analyzed to estimate the effect of each dimension of reform on measures of behavior and performance. This statistical approach relies on the assumption that measures of the most influential reform characteristics and of the exogenous envi- ronment are included in the available data or can be gathered using the survey instruments discussed below. Provided that these mea- sures are available, such analysis is a useful supplement to either of the other two approaches. The columns in table 3.7 represent the three possible timeframes for analysis. Once more, the three possibilities are not mutually exclusive. Indeed, the most powerful analysis can be performed by combining all three, by collecting data on several hospitals that describe intervention, results, and performance in the past and pro- spectively as the intervention is implemented. Evaluating the Impact of Organizational Reforms in Hospitals * 139 Cross-section analysis. One way to control for outside influences is to observe at least two hospitals, one that has gone through reform and one that has not. This technique is of most use if, except for the reform process, the two hospitals are as much alike as possible. For example, they might both be government-owned facilities with sim- ilar patient populations, similar medical staff, similar local monitor- ing, and similar exposure to competition from the private sector, and therefore similar demand for their services. Given sufficient similar- ity on such gross characteristics, the performance of the two hospi- tals can then be compared, using any of the measures listed above, and the observed differences can be attributed to the organizational reform process. This cross-section approach to analyzing the impact of reform is most effective if the analyst can form two groups, one of hospitals that have undergone reform and another of those that have not. In this way, the idiosyncrasies of any individual hospital are less likely to dominate the findings, and the true impact of the reform is most likely to emerge by inference. Retrospective measurement. One disadvantage of the cross-section approach is that the current differences between hospitals that have reformed and those that have not may have existed prior to the reform process. By restricting data gathering to the current period, after the reforms have been launched or are complete, the analyst runs the danger of misinterpreting association as causation. For example, if the reformed hospitals have lower morbidity rates among newborn babies than do the nonreformed hospitals, it is tempting to conclude that the organizational reform lowered these morbidity rates, when in fact the rates may have differed even before the reform. This would be the case if the government had decided which hospitals to reform on the basis of indicators of health care quality, allowing only the higher quality hospitals to undergo corporatization. Gathering retrospective data on each of the reforming and the nonreforming hospitals is a relatively low-cost remedy for this defect of the cross-section approach. If enough such data are avail- 140 o Innovations in Health Service Delivery able and of reasonable reliability, the analyst will be able to distin- guish between pre- and postreform differences between the two groups of hospitals. Such a distinction enables the analyst to argue more persuasively that a current observed difference is due to the reform rather than to a preexisting difference between the groups of hospitals. Prospective measurement with control: the ideal. Unfortunately, retro- spective data are not usually adequate to capture every feature of hos- pital performance or every feature of the hospital's environment that would allow solid analysis of the hospital reform's impact. In fact, the absence of strong data management is sometimes cited as one of the reasons to undertake hospital reform. The alternative is to collect high-quality, comparable data prospectively on two groups of hospi- tals, some that will reform and some that will not. Prospective data collection by itself does not, however, suffice to dramatically improve the persuasiveness of any lessons learned from the analysis of a country's hospital reform. To evaluate all of the important aspects of the organizational reform, first, enough data on the right indicators have to be collected. We argue that four groups of indicators are necessary on the four parts of figure 3.1 (environ- ment, intervention, response, and impact). But it is impossible to col- lect data on every aspect of hospital performance. Some variables are expensive to collect; for example, outcome data that require large samples and lengthy patient monitoring after a hospital visit. And some data are simply unobservable, for example, the political forces that determine how hard a hospital's budget constraint is. The most rigorous approach to controlling properly for all these other sources of variation in hospital performance is to employ randomized trial techniques, which are well known to the medical research commu- nity but infrequently applied to health care reform. Hospitals eligi- ble to undergo corporatization should be randomly assigned (by ran- dom drawing or lottery), either to a group that will be subjected to reform and intensive monitoring or to a control group that will only be monitored. Evaluating the Impact of Organizational Reforms in Hospitals * 141 Combining random assignment of hospital reform with prospec- tive data collection offers the best chance of learning exactly why hospital reform works out as it does in the hospitals within a health system. However, in some national contexts, prospective data collec- tion and random assignment are sometimes unfeasible. Even in instances where less rigorous study design leaves uncertainty about some of the conclusions, it will always be better for a country to learn as much as possible from its reform effort. Since analytical time and resources for data collection are expensive, many governments must be satisfied with less rigor, in the interest of learning as much as pos- sible right away. The case studies in this book all represent compromises between the desire for the maximum possible rigor and the exigencies of the actual data and reform situation in specific countries. The authors point out the uncertainties that are due to inadequate data and research design. Based on these results, each reader can decide whether, in a few countries over the coming years, investing the sub- stantial resources required for more rigorous studies, such as pro- spective randornized controlled trials, might be worthwhile. Alternative Measurement Instrments When a set of indicators is defined based on what has to be mea- sured, the analyst has to decide on the method to collect the data. The selection of an appropriate measurement instrument may be based on time, cost, and the degree of comprehensiveness associated with each alternative. In general, however, each instrument presents advantages and disadvantages, and usually trade-offs among the above selection criteria. Therefore, the analyst must identify priori- ties and assess which instruments best reflect them. Regular information systems. We use the term regular information system to refer to information routinely and regularly gathered by health sector or hospital personnel. The information is likely to be facility- or personnel-based quantitative data and used mainly to monitor 142 o Innovations in Health Service Delivery inputs and outputs. The analyst can use these data to produce statis- tical indicators (e.g., the number of admissions by demographic group) or combine them to estimate composite indicators (e.g., aver- age length of stay). Collecting such basic data is relatively inexpensive and fast. However, this type of information is unlikely to support the analysis of causal relationships between organizational reform and hospital performance that are the goal of evaluation studies. Surveys. Surveys gather data less frequently than do the regular information systems, but they typically gather more useful data. Sometimes a survey or a survey component can be fashioned to respond to the needs of the evaluation exercise. One of the easiest forms a survey can take is a simple checklist. This type of question- naire is relatively quick and easy to design, administer, and process. It is also convenient and timely when the focus of analysis is presence or absence of particular inputs and outputs. However, checklists are not the best way to gather qualitative data or to gather quantitative numerical data. A formal survey is a more costly and time-consuming alternative. It can be quantitative or qualitative and can focus on diverse issues, including financial, quality, and health status issues. Though more complicated to construct, apply, and interpret than the other two alternatives, a formal survey is more useful for evaluation, because it provides more detailed measures of many critical indicators, espe- cially on outputs and outcomes. Surveys also help attribute any observed changes in performance to specific policy innovations. For several years, health facility surveys have been included among the community surveys administered as part of the Living Standards Measurement Survey program in the World Bank's Devel- opment Research Group. Since these surveys were intendecl to gather community-level data necessary to understand household consumption and health-seeking behavior, not to track the organiza- tional reform of hospitals, they are not ideal for the present purposes. Nevertheless, they form a good starting point for constructing a practical hospital survey. Two of the surveys available from the Liv- ing Standards Measurement Survey unit of the Bank are the Health Evaluating the Impact of Organizational Reforms in Hospitals * 143 Facility Survey from the Jamaican Survey of Living Conditions,44 and the Model Facility Survey.45 As can be seen in table 3.8, the questionnaires include many indi- cators of hospital performance that focus on inputs and are therefore relatively easy to collect, but fewer that capture the process or the output stage of production. In developing survey instruments for a specific evaluation exercise, the analyst may build on these examples by adding questions to capture selected outcome indicators, creating additional sections devoted to other performance dimensions such as equity and organizational performance, and perhaps deleting ques- tions that collect duplicate information. There is a relationship between the type of data collected and the timeframe of the analysis discussed above. Retrospective measure- ment is often limited to the use of the regular information systems, because such data are often the last resort of an evaluation conducted after the change of interest has taken place. In contrast, the prospec- tive measurement approach affords a chance to design and adminis- ter special purpose surveys to capture the specific changes that the reform is supposed to produce. Table 3.8 Performance Dimensions by Stage of Production Process, Two Sample Surveys DIMENSIONS OF HOSPITAL PERFORMANCE TECHNICAL ALLOCATIVE EFFICIENCY EFFICIENCY QUAUTY EQUITY Stages of Inputs 1, III, IV-1, IV-1 1, 111, IV-1, 11-2, V hospital V, VI, VII-l, D V, VI, VII-1, production VII-2, VIII VII-2, VilI process A, B, D, E, F A, B, D, E, F Processes A, C IX-1, IX-2, IX-3, IX-4 A, F, G, H, 1, J Outputs and C 1, 11-1, III, V outcomes Vl-I, Vil, IX C, F Note Numbers in the table denote sections and charts (or question numbers) in Survey 1 Letters denote sections of Survey 2 144 ° Innovations in Health Service Delivery Use of existing population-based information. Evaluations of hospital reform will often benefit from the use of preexisting, population- based data sources available from official or unofficial institutional sources. Because this type of information is unlikely to focus on hos- pital organizational reform issues, it cannot substitute for more direct measures of intervention, response, and performance. How- ever, the analyst may need to turn to such information, especially in calculating the indicators of utilization rates by socioeconomic group that determine the equity with which hospital services are provided. Recommendations and Condusions The evaluation framework proposed here could be used to judge whether hospital organizational reform affects hospital performance and, if so, to understand each reform dimension's contribution to the overall result. Because hospital organizational reform is a complex procedure applied to a complex organization, understanding how and why any reform succeeds or fails will always present a challenge. The most useful evaluations will have two characteristics: o They will gather and analyze data on reforming hospitals that track over time changes in the hospitals' environment, the nature of the reform, the managerial response to the reform, and the associated changes in hospital performance. o They will gather data on matched hospitals that are not simulta- neously undergoing the same reform. Evaluations that do either of these without the other might be useful. However, they will be subject to criticism by policymakers as well as by other health systems researchers who will argue that the observed associations between reform and performance are due to coincidence and do not reflect a causal relationship from one to the other. Evaluating the Impact of Organizational Refonns in Hospitals * 145 For completeness and to offer a large number of options, we have included several ways to measure each of 18 relevant dimensions, where in many cases 1 or 2 would suffice. But even collecting data on only a fraction of the variables suggested would be expensive and time-consuming, especially in developing countries where statistical services may be weak or overburdened. Furthermore, national gov- ernments may be reluctant to invest in such data-collection efforts, especially since a substantial part of the benefits of any lessons learned would accrue to the international community beyond the borders of the country collecting the data. These are arguments for bilateral and multilateral donors to bear a portion, perhaps a large portion, of the cost of such evaluations. Unless such evaluations are performed, however, the questions raised at the beginning of this chapter regarding the link between various possible reforms and social benefits will remain shrouded in mystery. Notes 1. M. Young, J. van der Gaag, and D. de Ferranti, "Learning from Early Childcare and Education Programs," in R. Picciotto and E. Wiesner, eds., Evaluation and Development: The Institutional Dimension (New Brunswick, NJ.: Transaction, 1998). 2. World Bank, Bureaucrats in Business: The Economics and Politics of Government Ownership, World Bank Policy Research Report (New York: Oxford University Press, 1995). 3. I. Millstein, M. Albert, A. Cadbury, R. Denham, D. Feddersen, and N. Tateisi, Corporate Governance: Improving Competitiveness and Access to Capital in Global Markets: A Report to the OECD by the Busi- ness Sector Advisory Group on Corporate Governance. (Paris: Organisa- tion for Economic Co-operation and Development, 1998); I. Mill- stein, "The Corporation as an Ownership and Management Structure" (paper written for the World Bank's "Private Sector Development Forum," Baltimore, September 16-18, 1998). 146 o Innovations in Health Service Delivery 4. World Bank, Bureaucrats in Business, p. 93. 5. Ibid., p. 7. 6. Ibid., p. 130. 7. Ibid., box 3.2, p. 131; M. Shirley, "Why Performance Con- tracts for State-Owned Enterprises Haven't Worked," World Bank, Finance, Private Sector, and Infrastructure Network, Viewpoint, no. 150, August 1998; and M. Shirley, "Performance Contracts: A Tool For Improving Public Services?" in Picciotto and Wiesner, Evalua- tion and Development. 8. Millstein, "The Corporation as an Ownership and Manage- ment Structure." 9. Song Dae Hee, "New Public Enterprise Policy and Efficiency Improvement," Korea Development Institute Working Paper 8811, as cited in L. Jones, "Comments," in Picciotto and Wiesner, Evalua- tion and Development. 10. Following Millstein, "The Corporation as an Ownership and Management Structure." 11. The established criteria for a private good are rivalry and excludability. Hospital-based health care treatment is typically both rivalrous (treatrnent consumed by one patient is not available to another) and excludable (treatment can be inexpensively denied to patients who do not pay for it) and therefore qualifies as a pure pri- vate good. However, treatment of infectious diseases, though a pri- vate good, fulfills a social as well as a private function because of the positive externalities or "spillover effects" on people who would have caught the disease from the untreated patient. 12. The hospital's availability to provide care for catastrophic ill- ness or injury in a poor country can be considered a second-best remedy for failure in the insurance market, which could be viewed as a social function of the hospital. D. Filmer, J. Hammer, and L. Pritchett, Health Policy in Poor Countries: Weak Links in the Chain, Policy Research Working Paper 1874, World Bank, Development Evaluating the Impact of Organizational Reforms m Hospitals * 147 Research Group, Policy Research Department, Poverty and Human Resources, Washington, D.C., 1998. 13. Millstein, "The Corporation as an Ownership and Manage- ment Structure," p. 6. 14. For a detailed discussion of the implications of the character- istics of various health care products for organizational reform efforts, see chap. 1, this volume, and N. Girishankar, "Reforming Institutions for Service Delivery," Policy Research Working Paper 2039, World Bank, Policy Research Department, Washington, D.C., 1999. 15. For a discussion of the importance of routine monitoring to improving the management of the public sector, with examples from Australia, Canada, Chile, Indonesia, and Zimbabwe, see K. Mackay, ed., "Public Sector Performance-The Critical Role of Evaluation" (selected proceedings from a seminar on the experience in develop- ing national evaluation systems, World Bank Operations Evaluation Department, Washington, D.C., April 1998). Appendix 3B lists World Bank projects that include hospital reform components. The project documents for several of these projects include proposals for monitoring the hospital reform process supported by the project. 16. Readers acquainted with the economics of industrial organi- zation will hear an echo of the classic "structure, conduct, perfor- mance" paradigm. 17. Mead Over was a member of such a regulatory board, the Western Massachusetts Health Planning Council, from 1978 to 1981. 18. The welfare theorems of neoclassical economics prove that given the conditions of perfect competition, markets will distribute goods and services efficiently such that no one could be made bet- ter off without making someone worse off, a situation that is defined as "Pareto optimal." The consequences of corporatiza- tion for the poor would still require monitoring. See, for example, D. Salvatore, Microeconomics: Theoiy and Applications (New York: Macmillan, 1986). 148 o Innovations in Health Service Delivery 19. See chap. 1, this volume, and Girishankar, "Reforming Insti- tutions for Service Delivery," for detailed taxonomies of health care services on the dimensions of "contestability," "measurability," and "information asymmetry." This taxonomy helps in the design of a reform to predict which services are likely to be more efficiently pro- vided in a market and in which hierarchical system. 20. E. M. King and B. Ozler, "What's Decentralization Got to Do with Learning? The Case of Nicaragua's School Autonomy Reform," Working Paper 9 on Impact Evaluation of Education Reforms. World Bank, Development Economics Research Group, Washington, D.C., April 1998. 21. Ibid., p. 117. 22. By social functions, we mean services or products delivered to recipients at a price less than cost-either for distributional purposes or due to externalities in consumption. 23. This movement away from internal, implicit cross-subsidiza- tion is often an explicit objective of organizational reforms in other sectors. Some scholars, notably M. Pauly, believe it should also in health services. See M. Pauly, "Health Systems Ownership: Can Regulation Preserve Community Benefits," Frontiers of Health Ser- vices Management 12 (3): 3-34 (spring 1996). 24. J. R. Gordon, A Diagnostic Approach to Organizational Behavior (Newton, Mass.: Allyn and Bacon, 1987); P. Hersey and K. H. Blan- chard, Management of Organizational Behavior Utilizing Human Re- sources (Englewood Cliffs, NJ., 1993); and R. Numerof, The Practice of Management for Health Care Professionals (New York: AMACOM [Division of American Management Associations], 1982). 25. C. K. Bradford and J. E Tiscornia, Monitoring the Hospital's Financial Health: A Guide for Trustees (Chicago: American Hospital Association, 1987). 26. R. Kaplan and D. Norton, "The Balanced Scorecard: Mea- sures That Drive Performance," Harvard Business Review, Harvard Evaluating the Impact of Organizational Reforms in Hospitals * 149 Business Review on Measuring Corporate Performance (Boston: Harvard Business School Press, 1992). 27. J. L. Heskett, W E. Sasser, Jr. and L. A. Schlesinger, The Ser- vice Profit Chain: How Leading Companies Link Profit and Growth to Loyalty, Satisfaction, and Value (New York: Free Press, 1997). 28. Ibid. 29. H. Barnum and J. Kutzin, Public Hospitals in Developing Coun- tries: Resource Use, Cost, Financing (Baltimore, Md.: The Johns Hop- kins University Press, 1993). 30. Kaplan and Norton, "The Balanced Scorecard." 31. For substantial discussions regarding tools for assessing med- ical management, see A. R. Tarlov and R. V Colley, "Clinical Out- come Management," in M. M. Melum and M. K. Sinioris, Total Quality Management: The Health Care Pioneeri-s (Chicago: American Hospital Publishing, American Hospital Association, 1992); and R. N. Zander, "Critical Pathways," in Melum and Sinioris, Total Quality Management. 32. Clinical pathways, along with the "integrated package," are a widely accepted disease management approach. Disease manage- ment aims to improve effectiveness of care and cost effectiveness and involves shifting away from more expensive inpatient and acute care to areas such as preventive and ongoing care, health promotion and education, and outpatient care. The evaluation of disease manage- ment approaches is discussed in L. Reeder, "Anatomy of Disease Management Program," Nursing Management 30 (4): 41-45 (1999); and in D. J. Hunter and G. Fairfield, "Disease Management (Man- aged Care, Part 3)," British MedicalJournal 314 (7099): 50-53 (1997). 3 3. In the analysis of technical efficiency outside the health sector, the qualifier "at constant quality" would be added to each of the pre- vious two sentences. Because of the importance and difficulty of measuring the quality of health care, we treat it as a separate dimen- sion of performance, discussed below. 150 o Innovations in Health Service Delivery 34. For proponents of this view, see R. Evans, Strained Mercy: The Economics of Canadian Health Care (Toronto: Butterworths, 1984); and W. De Geyndt, Managing the Quality of Health Care in Develop- ing Countnes, World Bank Technical Paper 258, Washington, D.C., 1995. De Geyndt expresses concern that there is too much emphasis on efficiency at the expense of quality, and asserts, to the contrary, that "[i]mproving quality would also improve efficiency. There are growing concerns with the inefficiency of providing medical care of doubtful efficacy." Emphasizing that low-quality care leads to com- plaints, bad reputation, lost referrals, legal procedures, and loss of staff morale, all of which harmn efficiency, similar concerns are expressed by J. Ovretveit, Health Service Quality: An Introduction to Methods for Health Services (London: Blackwell Science, 1992). 35. J. Harris, "The Internal Organization of Hospitals: Some Eco- nomic Implications," Bellyournal of Economics 8: 467-82 (1977). 36. See, for example, G. Mwabu, M. Ainsworth, and A. Nyamete, Quality of Medical Care and Choice Medical Treatment in Kenya: An Empirical Analysis, Technical Working Paper 9, World Bank, Africa Technical Department, Hurnan Resources and Poverty Division, Washington, D.C., 1993; and H. Alderman and V. Levy, "Household Responses to Public Health Service: Cost and Quality Tradeoffs," World Bank Observer (International) 11: 3-22 (1996). 37. W R. Scott and A. B. Flood, "Conceptual and Methodologi- cal Issues in Measuring the Quality of Care in Hospitals," in A. B. Flood and W R. Scott, eds., Hospital Structure and Performnance (Bal- timore, Md.: The Johns Hopkins University Press, 1987). 38. W Holland, ed., Evaluation of Health Care (Oxford, U.K.: Oxford University Press, 1983). 39. R. Evans, Strained Mercy. 40. I. Abedian, B. Strachan, and T Ajam, Transformation in Action: Budgeting for Health Service Delivery (Cape Town: University of Cape Town Press, 1998). Evaluatng the Impact of Organizational Reforms in Hospitals * 151 41. R. Govindaraj and M. Chawla, Recent Experiences with Hospital Autonomy in Developing Countries-What Can We Learn? Data for Decision Making Project, Department of Population and Interna- tional Health, Harvard School of Public Health, Boston, Mass., 1996. 42. A fall in the percentage of poor patients may not be worrisome provided that the number of poor patients stays the same or increases. 43. Govindaraj and Chawla, Recent Experiences with Hospital Auton- omy, "p. 19. 44. The health facility of the Jamaican Survey of Living Condi- tions was designed by Bank staff Paul Glewwe and Margaret Grosh in collaboration with the Jamaican statistical authorities. 45. See P. Gertler, E. Rose, and P. Glewwe, "Health Module," in P. Glewwe and M. Grosh, eds., Designing Household Survey Question- naires for Developing Countries: Lessons fiom Ten Years of Living Stan- dard Measurement Experience (New York: Oxford University Press, 2000). z >E 0 <0 000 o- 0, .-Z - o U 2 a 0 ~~~~0 ~ ~ o El E ED E o - o 0o co Z - Zr 0 U 0 t E E o E o o 0 0 o0 -0 0-0 - o ~~~~~~0 0 0 0r O ~ ~~~~ >, E - V 9C, D-, D3 -2 0~~~~~~~~~ P ~~~~c_O co co co >\ Q3 "ao j wQ Z W o < D 0 0 0 m 0 0 -0 - Q ffi 0 O t- 0 E - COC0C0 _ ~~ ~~~_ 000ai- 0)c0 .20 0.2 -0.2' 0,,, -~~~~)a C 0) - 0 . c -a 2- U .C~~~~~~-Ž'~ 0 (C 0 Et. 0 C= >.a -z 00 a a- _ 0 200.., C ~~or3 EE 0E Z >~C C0Co 50 E la E 0 CE) La- :D > 0 0~~~~~~~~~~~~ CV 0 0 - N -? 00 5.C C t V Z ~0 Ž u to & o 8o ?. 2a- 0 : 0 .9 <0U2 C. D 0 LoE e a-a a 152 0 0 0~~~~~~~~~~~~~~~0 0~~~~~~ ~~~r_C~~~ z z z z o Z @ ooo f° U u jJ u 1w u I"UU u L>I (D a) uf f f f ) f Yo | 0-° ° cf > c I, -u s o-' Z t (D 0 0~~~~~~~~~~~~~~~~ 0 -C~~~~~~~~~~~~~~~ C: u~~~~~~~~ ° 0 -° > a) -o _C -C3 24, _CZ , ,u , o E O O E e~~'O i- E oX > 0 ,> 1° 0 Mu -u --oC 4 M Q l3 m oE 0~~~~~~ u o .o, o - % -o o E E J a -° '2.o j ° > ° U E o o E X o o 0a g ° E0> 0 0~~~~~~~~~~~~~~~~~~~ o za a -c -- c i ! U o 0-5 ZZS ? Z1 03 U V 010 U) Z0 U r- 0),2 -0~~~~~~~~~~~~~~~~~~~~~~ -o > I~0- 0 -o 0 0 '- ~~~~,0K,W ~~0- C U ~~~~0 E 0 -~~ 0 - -46 K0 ~~- ~ ~ U~~EU j~~-~~ I _- 0 2~~~~~~~~~~~~~~ 0) 0~~~ ~~' -0~~~~ 0- 0 .2;~~U OC- o CO~~~~~~~~~~~~~~ 0~~~~~~~~~ 1=z _ E C E 1W2 0~~~~~~~~r3 0 ~~0E 0 E V) u ~ ~ ~ ) 0 - -flo -u ~ 00)UU O)C 0 C~~ ~g~~-2 -~15 <0 -u 0> 0> 0 _' o y E 0-~~~~ a-~~~~-0 - ~V o U z ~ ~ Ua E ,)°i S oW°- so) oE -o -~~~~~~ ~ c co.- co E- _s U, O c C u a0 ,N.O '- ' o Q °, , 0 , 0 , 0C: CL 0 E ~~30 5~ c~~.~-~500 _)0 g it °- aD a) , C4 0 aI £ 0~~~~ 0 j,7! ~~E z 00) -O.o BcE" - 0 a >~~~~ ~~ 0 0 0 0 E~~~~). a)a Z- r- a-aa N E~~~ 4) CL0 ~ 0- -L 0 0. N 0 0 r a) 0 ~~ ~~ ~ ~~ a a ) -c z a aa 0E 00 -5 5 z -C,L ~ CO 0 ~~~ C 00)0 154 0) 00Q oE -0 Qi) ol~ ~ ~ ~ ~ ~~~~~~~~~~~~~- CC) g)~~~~~~~~~~~~~~ .00 CO C 2 ~~~~~~~~~~~~~-~~~Cl u o o 0o c o c o oc oc U5 Z~~~~~~~~~~~~~~~~~~~ U C~~0 U, CO~~~~~~~~~~C CO ~0 - (D >- C - -L Oa 0 2 c CL 0 a C - o .2- :0- 2 r C CO 0000 aC 0 o ~~~~~~~~ ~ 0-0 00 0 -00) Q- = 0 -~~~0 , OC - a~~~~~~~ C ) - 5 0 C-~~~~~~~~~~CO0 -Ed) u c - > >UOa . 0~a C 0 cc, E- -~~~ g~~~~2-oo j~ D 2-9) 0 o~~~~~OCO~~~~2 -oo ~~~~~~~-2 co 0 c 0 CO~~~~~0) C 0 W ~~~~~~0 0 C ) t Y 30 Za x 0C - E~~~~~ e UL) c . 0~~~~~~~~~5 0 ~~~~20 LiU)E U) w 0- CO' ~~~~~~~ 0 ' C0 ' 0' - C~~~~~~~~~~-7 a- -o C C z C 20 ~ ~ ~ C~~ a- - a) U ~~~U> 'D00 C D- 2 2Z > 0) D- - u 0) 0 0 0 >~ - 2 o c o E a E 0a > 0, 0- -0 u ~ 0> 0 U.. 0a) I) 4 9 C 0D u D a 0 XOU E a)' - 0~ < C >. E0oD-0 0 E 0 .0 0) ) 0 0o 0, 0 0 ~~~~~~~~,~~- 0) u 0 o-0~~~~~~~~- o 0~~~0,0 >.~ 0 E 'c 0 _ 0, C " E v 0 E E 0) 0) a E~~~~~~~~ 0 a E E a) CE U ,DC 4)0 ", 040C D C 0 2 2 E0a 0C2 0- 2Q C C 0a0~~~~~0~~ O~~ Et 0 0,,oa l0 C 0)~~0 E C~~~~ 0 C~~~~0) C 0 0 Z' C CU E ' 0 C~~~~0 0 0 - E0 0 a CU EC~ Z 0 U 0 C0 a- a- 0 ~~~~~ C E~~~~~~~ n EE2 -0 0 0 0 0 J JoZ0 uj L 156 C -0 _ l z -o Rl O' ~0 ' CE2 - 0 0 0 -c. Z 0, 0 1 -0 ol .0~0 -0' a -0 -C E >1~~C 0 0' Q--2 a ol > 0' 2o C C u -C 0~O-C, -0 5D0 00 c 0~~~~~~~~~~~~~~~~~~~ C) 0 2 2 o.9 c - ~~~0 o E ~u0 0 "'7 C u 0 .~~~~~~~~~C 0 0 ~~~~C .~~~~~ to Qo~~0 m- -o '4 a 0 .9 0~~~~~ .fl~~ a 0 0 c C00- - 4,-2 C) 0) C) ',toC 0~~~~~~~~~~~~~4 o0)=" (D 0 a) 0 E ~~~~u C) E --~~~~~~E 0D 0 C a 0 -0 -10 E c o a-C 0 -~~ -a >C o -2 0- c D-2 a u 0 E o,0 a) ~ -o o~~~~~~2 C~~~~ -o .9 5 Cs C 2 > E 0000 ~~~~~~ >.fl.i ~~~ C) > Eo - o ;a-0, 0 0 C C)>'o~~~~0C U ~~~~0 0C 0 O 5E C >~~~~~~ C 2 .C = 2> E E nC a 0 0 u 0 tC C>0 O0E~~ 0 C)D CC- 0- EL C - C EŽ 0 0 c ~ C C .2 0~~~~ 4 C"' a 0 o CCL0 .9 L Cco 0 '40C)a E'~~Ž' C- E "' 157 AL .2 0' 0 0 z - 0 0 - 0 a) - 2 o~~~~~~~~~~~~~~v 0 < -0 U U) o 0 >~~~~~~~~~~a- 0E _ - _ oo~~~~~~~~~~- *5o -~~~~ ~~~~~ ~~~~~~~~ 0 ,E 0-. 0)~~~~~~~~~~~~~~~~~~~~~~C o -~~~~~~~~~~~~~~o F7 K, c- 21~- - 0, ol~~~0 0 : 0, L CL~~~~~~~~~~~ C > fz: ~~~~0 ~~~~~~~~ - c C E r D a 0 000 0 0 0 0)w 4 a)0 20~~~2- aaa a -a a 0 00~~~~~~- >0~~~~~~~~~~~~~ LUE a ~~~~0 LU Y~~~~~~~~E U) - Er - u EYE < 2 - CM 0 0 D-~~~~0. n_ 0 0 I 0 0L E 2 -0O0 ., E o-o-c - 0 ~~~~~~~~~~~~' -'-C Du o ~~~~~~~ ~~~~ > 0 C - U.0 00 2 u - 0 >,0I--L -0 - E ....o o 00 - 0.0 0 U 0 0 -0O 0 158 0 0'~ ' ou 0 J ~~~~0) ~~~~~~~~0 O' ~~~~~~~~~~E 0, Q)~~~~~~~~~~~~~~~~~~~~~0 -o =Es Lo o- 00 -0 O _ U) DI) 0 - O 0- °' O _ I~ Y> - u t - E D0 0>~ Co <( E vM , 4) a) ) a- Q O G 00: M 0 0)0 C 0N c>. 2 D D ui O N_- U CY C) @ ) X z 0 0 ~~~~***o * :D-O C 0 0' 0C Q- 0 o N U~~o 'r~~~~~~~~~~~~~~' ~~~v0~~v OC d0 0 ui Ec) -0~~~~~~~~ (2 00 0)) 00)) 0 0- a 0a ~ ~ ~ ~ _ 'U Sy~~~~~~~~~~~L C) -C)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~2 0 0- C U. 0) - 0 )a 0 c >~l '- 0CL p~~~ 0) -~ ( 0)0) 0 00 u 0I 0E o vU U U, u~~~~~~~~ _D 0) -D 00 0) co0 o0 0 0 2~0 *.E SE ao 2 di d) d) d) 0)00 0 0 0 0 oZZ LI 159 z 0 0 o 00,~~~~~~~0 -o C "4 c 4 -o N - 00 0'~~~~~~~O > C~~~~~~~~~C a)0 0a) 6- 0 6 oo )a)a >> 0 0 0) -~~~~~~ 00 0 ~~~~0 0 LU 0 -00 0) c 0 0 .> 0~~~~~~ 0 E ? 0~~~~ 2 ~~- C E 0 2 .2_ 00 0 0 -o U~~~~~~~~ 0) >0. 0 C E ~~~~~~~~~40 E ~~~~ E .2.~E(D O 0 E~ . 0 -j0 0u 0 - E~~~~~~~~~~~~~~~~~~~) --E ¼ ~~a cc 0E 0 E -14 0 0 ~ 00 ~, O0 EE > oLA 0 -c~~3 '4. 0) - ,00 )D- D I to ~C -Co 4 )& .0Q- .2 3-~00 g{ I 0 2&o >' 0-g g O2 r (a0 '4 0Y 0E o x 0 >w 0 v 160 0)~~~~~~0 0. ol 0 ol o,~~~~~~~~~~~~~~~~0 ~0) 0' - ~ c ~ 0 .2 - - o 2 2. . - 0 aC U (D 0 (I 4 t, *) E 0) ) 0 r 0 0 0 a ao 0 ~o ol 0 5l -0 0 1E~~~~~~0) c~~c '- a '- o, ~0O, a 1 - 0-2~~~~~ Z -~~~~~~ -R< LU 0 O- Cl 0'. oo2FZ -, 0'~~~ 00 0o, 0- C -o 0 o a 0 C cC OM~ 0) 00c 4) 0 E~o 00 wo 00 io 0 E ) 0)0o "O 0) 0) 0) 0) 0) 0) 0 0 a 0 a -a0 > > > > > >> -o ao -a .2 p p p~ -2 .2 La < < < < < < (D ~~~~0 ID U E 0 ID 0)0 - .~~~~2 0 0) 0 0 u 0 0o 0~~~~~~~~~~~~~~~~~~~~~~~~~0 E --~ - - -E r: ( > 0 0-00 -o CU _ 0 C_ 0 0) E ~~~~~p ) 0 .20E.2-~~~~U ,, 0 C E u~~5 LU 0 0~~~~'- E0 E Ei 0.> 00 - 161 0z 0 0) C) 0~ 0: 2 D ~~~~~00 00 0 a0 a~ 0~ 0 0 a~~ 0 00 a 0 0 0)>0 > 000P)R U > > > > > >> 2 2 2 22 P2 22 00 u tiC 0) O 0). . 00 o ~~ ~~ ~~0 0 0 0-0 E2- 0 -0P C 0 0> -C o0) 0j( 0 0- -- c > 0 0~~~~c E >- - 0) 0) ) N0 0 I0 2r - -0.. o22E Q 0 0_ ) 0- - r- ( 0) 0- E0E' 0-E-2 - (DC' 0 - 0 0 0 0U0. :, 0 o 0 ) oN 0)0 C 0 00~~~~u 00 6 2 c. - > CL 0) ug V C CL) 0 O 0~' 0 0 . o o ' o0 E0 , E 0 E < o ~ -o 5 0) -0 !Ea SE M 2a -u2 0j -E0> I ~~~ 00-0O'..-~~~~Oo. 0) a)-hz < ) LUaID < 0 o ~~~~~~~~~~~~~ ~~~~~~~~~ ~~~0 0 U Z a 0 00 0 00.. 00000 0 -.00 000 o 162 co 0~~~~~~~~~~~~~~~ 0, 0, 2~~~0 0-a - 00 00 -0 0' 00N u 0 ~ ~ ~~~~ -0 ' 0 Z_r 0~~~~~~~~~~~~~~~~ -2 0 3 2 L 0 0 -0 a1 0 co~ 0 0 000 0~~~~~~~~~~~0 -l C Y :0 0~~~~~0 0 0' 0 co C1 N 0 No-o C > >>> - C 0' C 0'- u u '0~~~~ 0'~~~~~~~~ 0' C C C0'~~~~~~~~~~4 - 0 0 0 to-~~0 u) >. 0 >-~0 ': o~~~~~~~~~~~ 0 0 ~~~~~~0 . ~~5 oo~~~u -0 C - to to 0 ~ ~ ~ ~ ~ ~~~0 00- a) 0 00 LU 0~~~~~~~~~ o- E 0- 0>o 6 0 V~~~~~~- 0 X (D (D 0) p ~ ~ C 0 E~~~~> -t -o oW V X U u -o c -00 o -0 -Z 00 C 0 E to2 a to E 0 E0 *- 0 E 0 0 o to0 C 0 0- C00O : t -E -0 -E 0 P' x 0 ) E o - ~~~~~> . 0- a - 2 a) > , 0 * Q.C 0 0 CL 163 z1 0 01~~ C) -o~~~~- 0' a0 C)~~~~~~~~~4 0~~ 0o -aO- Q) a) a 4 ) C)0 C) C ID 0 a) a, UD .2 2 0 ~ ~ 0 C 0~~Q CU D O) (D - E -0 o o- 0 C)UC)E x n 0) ~ ~ ~ ~ aC 2 G.) )a r - U ej C 0t00 164 - Evaluating the Impact of Organizational Reforms in Hospitals * 165 Appendix 3B. World Bank Hospital Reform Projects I DATE COUNTRY PROJECT TITLE' 06/26/98 Panama Health Sector Reform Pilot Project 05/20/98 Ecuador Health Services Modernization Prolect 05/11/98 Nicaragua Health Sector Modernization Project 02/13/98 Tunisia Health Sector Loan Project 10/21/97 Guinea-Bissau National Health Development Program Project 05/16/96 Russian Federation Medical Equipment Project 05/09/96 Morocco Social Priorities Program Basic Health Project 02/20/96 India Second State Health Systems Development Project 02/06/96 Sierra Leone Integrated Health Sector Investment Project 11/07/95 Mozambique Health Sector Recovery Program Project 07/05/95 Argentina Provincial Health Sector Development Project 01/26/95 Croatia Health Project 01/19/95 Estonia Health Project 12/01/94 Lebanon Health Sector Rehabilitation Project 11/09/94 Venezuela, R.B de Health Services Reform Project 11/08/94 Albania Health Services Rehabilitation Project 11/02/94 India Andhra Pradesh First Referral Health System Project 09/21/94 China Comprehensive Maternal and Child Health Project 11/24/93 Nicaragua Health Sector Reform Project 03/01/93 Hungary Health Services and Management Project 02/01/93 Jordan Health Management Project 06/01/92 Ecuador Second Social Development Health and Nutrition Project 06/01/92 Korea, Rep of Public Hospital Modernization Project 11/01/91 Chile Technical Assistance and Hospital Rehabilitation Project 04/30/91 Zimbabwe Second Family Health Project 04/24/91 Korea, Rep. of Health Technology Project 02/01/91 Tunisia Hospital Restructuring Support Project 05/07/90 Yemen, Rep of Health Sector Development Proiect 03/01/89 Indonesia Third Health Project 02/01/89 Mozambique Health and Nutrition Project 05/01/84 Brazil Second Health Project 11/01/83 India Third Population Project 01/01/83 Indonesia Provincial Health Project 05/01/81 Tunisia Health and Population Project 04/01/71 Trinidad and Tobago Population Project 06/01/70 Jamaica Population Project a. These projects have hospital reform components or address hospital-related issues Part 2 Cross-Country Views Reviewing the Case Studies: Tentative Lessons and Hypotheses for Further Testing Loraine Hawkins and Chris Ham During the design or implementation phases of reforms, problems can arise. In a number of our country/economy case studies, key de- sign features were altered when difficulties were encountered during implementation. Conversely, poor design contributed to implementa- tion difficulties in some cases. We use the frameworks set out in chap- ters 1 and 2 as a basis for a preliminary assessment of the success-or otherwise-of the design and implementation of our case studies of marketizing organizational reforms. Table 4.1 develops a fuller frame- work for evaluating organizational reforms of health care delivery. We do not to attempt to make a rigorous evaluation using this frame- work, but to draw some tentative lessons from the case studies that might be viewed as hypotheses for the further testing of factors that are important for successful reform and as strategies for reducing the risk of implementation failure. Diversily in Design The case studies are diverse to the extent that the reforms changed the organizational structure, market environment, and funding 169 4 to> o QX C~~~~~~~~~~ E .0. . \ ~ 0U _O -O ~ ~ ~~~~- N ,/ Ste ardship a) ORGANIZATIONAL Ag //^. 0C-0 .0 Co 0 . - 0, -= >1 >c C 0 Z c ~ C O - ..001-o C O ~~0 0 0~0 o 0O o 0 u - - E 'D~E 00 0. - (D c 'E~ C 0 < 0E o- 0E >0)~ -0 -0 0- O0a,000 9OuC 00 t *U 00 E L- C ,, - - D C8 ~ 0 r0 0~oo~ 0 C ~~~~~~~C E Eor- 00 0 2 02Oo C0,o2.2.C0 -va CL 0 - D 0 m x o ~E o0 C3~ 0 o U o co ') Z ~ o 0U) ui) co Z i 0 m ., 0 0 < I 0 r" 'A0 .2 (I) 0 .~~~~~~~~~~~~C 4 0 0 o U - 4 C 2 ~~~~~~~~~~ OC~~~~ 0 C 0 Eo o2 0 0 2 o r/ C -00~~~~ C 0rC C , a N ~~~~~~~~~C 0 -~~~~~~~~~~~~~~~~ _0,D0 0 0 C 0 CU 0 2 CO' 0) 0 0 0 0~~~~~~~~~~~ - 0C -,0 2 < a ' C0 c 0)D D 0 -C (D~~0 0. ID 0~~~~~~~~~~~> C, >>. . a 0.- C.- 0 ~~~~~~~~~~~O CC D 0 J- .0 o0.ud1 -t-40, 0E4, dr0 r- -r Z0).~~~~~E~~5) *EaC~~~U) 0 C3 9) 0 ~~~~~~~~~~~~~~~~~~~0 Cl0 0 C C' Do 0 0 2 0) -0 0 0) 0 6 ~ C - -C E00 U0 ' 0) 0 0 0 C 0r_ C00 C N a E 0 u 2 U~ 9 0 D a -o oe The Missing Link? Hospital Reform in Transition Economies * 215 formerly monolithic state bureaucracy and removed key functions from the MOH. This has required massive institutional adjustments to move away from the former command-and-control approach to governing the health sector and to redefine its new roles and func- tions. The MOH, the new social insurance organizations, and local governments had to build new capacity and expertise in contracting, performance monitoring, formulating health sector strategy, and reg- ulating the health sector. Changes in Hospital Organizational Structure during Transition In contrast to the systematic changes in the hospitals' environment, changes in their organizational structure have been minimal, ad hoc in nature, and often side effects of other reforms instead of explicit. Below, we track changes in hospital organizational structure in terms of hospital autonomy, market exposure, residual claimant sta- tus, accountability structures, and social functions. Autonomy Autonomy is the extent of the decision rights that hospitals have over various aspects of the service production process. Decision rights over six aspects are now reviewed: labor input, capital input, other inputs, output level and mix, pricing to purchaser, and management processes. Decision rights over labor input. In the communist era, physicians were salaried employees of the state. Overall staffing and salary levels were centrally planned and regulated, including the number of employees, appointments, remuneration levels, and hiring and firing. This meant that hospitals had littie autonomy over personnel decisions. During transition, central planning of human resource capacity has been deemphasized, and certain decision rights over employment of staff have been transferred to hospitals (table 5.2). In most coun- tries, the state has ceased to be the direct employer of physicians, and hospitals contract directly with their employees. In theory, decision a 0 c 00 0 0 Ž ° 2C ° ° 2 0° E C - E E - E 0 C3) 0 O E _ _ cOE * E E ^ E E _ ° E E > E C _ 5 'o O , o o 2 ° ° 2°,u 0 u a o~~ ~ X C °2 oc oo - o° ' 0 o ) 0 - 0) oo O. 0 > z1~~~~, ~ ~ ~ -00 -0 0 000 o (0 - 9 9 O >1 '= o C O 0 0 0 0 a~~~~~~~~~~~3 2U -0- 0 -=~~~~~~~~~~~ _ 0o 0 00 0 > O o C 0 0 0 ~=s 4> 0 O 0 O a 0 0 a- 5 .- .- . 21 0 O~ 0 0 c 0 ~~~~~~.2 ~~~~~~ ~ 2.2, V a) .2 0c 0O . 0 0 .0 0, ' 4 20 2 20 2 2.2~~~~ ~~ ) 2 2 -O:)E E -: ' 2D E>0 _ L-Ž 0) 07, 02 o0 .-2 0 C 00 0 00 ~~~a0 0 0 0onC CC C CC ~~CL L - C C 0- C -c- 0u0 0 ~ ~0 Ž0- C ) ) 0 4 C) C) Ci ) (D UC) 0 .22 .2 .2 .2 .2. .2.2.2 u M I~0 0 0 (D 000 0) 0 0 a) 0 (D 4)~~ 00 0- 0 . 22 2 Ur 2)0 a ~ a) .2 0 0 Ca a)C 0 ~0 0 0 C 0) . 0 CO0 0 0 000 ) 2 0 ~~~2 N .2~~~ 0 E V ~ 0< U U C) I ~2 Q.) cle 216 The Missing Link) Hospital Reform in Transition Economies * 217 rights over hiring, firing, and remuneration have been transferred to hospital directors. In practice, however, this autonomy is constrained by rigid labor markets, political pressures, and financial constraints. * Rigid labor markets. Employment regulations-whether general em- ployment act or civil service act-create inflexible labor markets by making firing and hiring decisions costly and by making remuneration difficult to differentiate. For example, in Hungary and Poland, hospi- tal staff members are civil servants, and the rigorous civil service act af- fords them significant protection such as open-ended contracts and substantial severance payment. II Although Romania has no civil ser- vice act, the general employment act is just as inflexible in requiring open-ended contracts.12 Furthermore, even though physicians are employed directly by hospitals, their payment in nearly all countries is subject to the act regulating remuneration of all public sector em- ployees-even in countries without special statusAegislation for civil servants. The Baltic states are an exception where physicians have been entirely taken off the civil service salary scale. 13 * Political pressure. Most hospital director positions are allocated through a political process, based on party affiliation, local political interests, and personal networks (table 5.2). This suggests that the actions of hospital directors will be influenced by the interests of politicians who appoint them. This limits the range of unpopular measures directors are willing to take for the sake of greater effi- ciency. The scope of hospital managers' autonomy is also limited by powerful health sector labor unions, which negotiate fiercely with government over salary increases. These negotiations define across- the-board pay for all staff that is binding for all hospitals. This leaves little room for managers to reward individual performance on a dif- ferentiated basis. * Financial constraints. With significant decline in hospital budgets, many hospitals have problems funding severance pay and other ben- efits required upon dismissal. 218 o Innovations in Health Service Delivery Decision rights over assets and capital input. In contrast to labor inputs, hospitals do not have decision rights over physical assets and capital investment decisions. Decision rights over assets are vested with the owners: central or local governments. Thus, although hospitals have incentives to downsize facilities to become more efficient, they do not have decision rights to do so. Local government owners, on the other hand, may have decision rights, but for political reasons have no incentives to sell their assets. This misalignment of incentives and decision rights is at the heart of the difficulties experienced with downsizing the hospital sector. Even in Estonia, where hospitals can form trusts under foundation law and own assets, by-laws restrict a trust's autonomy to liquidate assets. If a trust decides to divest its physical assets, the money goes back to the founding owners. Not only do hospitals have no decision rights over the sale of as- sets, they also have no instruments to ensure that their assets retain their value. The new provider payment mechanisms do not contain depreciation costs; capital investments are still financed from general tax revenues in most countries. Allocations of capital investment con- tinue to resemble the communist central planning process. Rational medium-term planning criteria are not applied, and decisions are often ad hoc and determined by personal and political networks. Medium-term planning is more likely to occur where international donor assistance is the major source of financing for capital invest- ment and where donors require medium-term plans (as in Albania).'4 Exceptions are the Czech Republic, where insurance payments to hospitals contain depreciation allowance, and Croatia, where the Health Insurance Institute procures and distributes all equipment.15 Capital expenditures in all transition economies have significandy declined and generally are below replacement rate. As a result, the condition of physical assets has considerably deteriorated. Despite decentralization of ownership, municipalities are not explicitly re- quired to maintain the value of their assets. Furthermore, their abil- ity to do so is constrained by their lack of revenue rights. In other words, transfer of ownership-and with it the implicit responsibility to finance asset maintenance and new investments-has not been The Missing Link) Hospital Reform in Transition Economies * 219 matched with revenue-raising authority. This implies a contradictory arrangement for maintaining asset value. Decision rights over other inputs. Nearly every country has extended full autonomy to hospitals for purchasing other inputs (nonlabor, noncap- ital) such as pharmaceuticals and appliances. This is a big departure from the pretransition era, with its MOH-centralized procurement of pharmaceuticals and other supplies. To regulate procurement processes, most countries have passed public laws that define hospital procurement practices and attempt to introduce transparency. Little is known about the ways hospitals have used their increased autonomy in terms of pharmaceutical procurement and manage- ment. However, anecdotal evidence suggests consistent problems across the region. In Hungary and Croatia, hospital managers often complain that physicians have no incentives to economize on drugs because the more drugs patients get, the more likely they are to re- ward their physicians with gratuities. Fraud and corruption are often mentioned, too-in the form of staff selling the hospital's drugs and pocketing the money. Decision rights over output mix and level. Within their budgets, hospi- tals have full autonomy to determine output mix and level. Under communism, they already had this autonomy, since central planning was input-, not output-, oriented. With the purchaser-provider split, health insurance organizations are supposed to shift the emphasis from specifying inputs to specifying outputs in terms of both mix and level in their purchase agreements. However, most countries have not been successful in moving away from the inherited central planning approach and continue to focus on inputs. In Hungary, contracts between the Health Insurance Fund and hospitals still specify inputs, including the number of hospital beds, number of staff, physician hours, and types of hospital departments. Outputs are mentioned only as aggregate service categories of inpatient or out- patient care. This is all the more puzzling since payment from insur- ance funds is formally related to outputs, not inputs. In Romania, for 220 o Innovations in Health Service Delivery example, the District Health Insurance House signs an agreement with individual hospitals without specifying outputs and without any binding legal force. This lack of focus on outputs in the contracting process is a generic problem across the region. An exception is the Czech Republic where contracts between in- surers and hospitals specify the volume and type of services, reim- bursement method, data provision requirements, termination condi- tions, and validity period. This is based on an overall list of services, the Schedule of Procedures, which specifies 5,000 procedures. Hos- pitals can decide which services to provide, but are reimbursed only for those in the contract. Decision rights over pricing to organized purchasers. By and large, hospi- tals have litde autonomy over payments, which are exogenous to hos- pitals and uniform for the entire country. This is the case in Croatia, Estonia, Georgia, Hungary, Lithuania, and Romania. Prices are set not by the payer but by the Ministry of Health or an appointed com- mittee, made up of various interests (typically physicians). In Estonia, regional sick funds pay hospitals by a combination of bed-days and fee for service, based on a price list generated by the Health Care Ser- vices and Investigations Price Committee, housed in the Ministry of Health. Although hospitals are theoretically allowed to price services 25 percent below list, they rarely do. In Hungary, relative DRG weights are estimated and updated by Gyogyinfok, an MOH institute. A committee of MOH-appointed physicians, where medical special- ties bargain and lobby, makes the final approval. In Georgia, prices of services in the state benefit package are set by the Committee on Med- ical Standards under the MOH and approved by Parliament. In a few countries, payment rates are set by negotiation between payers and hospitals, which gives hospitals some influence over pricing. This group of countries includes the Czech Republic, Georgia, and Poland. Decision rights over management processes. An interesting issue is how the transfer of decision rights to hospitals has affected decision- making processes and management practices within hospitals. In most countries, hospital directors and department heads wield considerable The Missing Link? Hospital Reform in Transition Economies * 221 power, using management instruments inherited from the past.'6 For example, in Romania, one observer noted that hospital directors are like "feudal lords" with full authority over many processes.'7 Because hospital directors are appointed in a political process (municipal and/or national), the incumbent director has the political establish- ment's support, which often motivates his actions and protects him from farther scrutiny by staff, patients, or representatives of the local government or community at large. Some countries attempted to enhance managerial professionalism and transparency by creating a management team to run the hospi- tal. In Hungary, initially a three-member management team was ap- pointed with a general director, a nursing director, and a finance director, but this system was quickly abolished because, as a decision- making mechanism, it proved ineffective. Other initiatives to improve internal processes include increasing participation of hospital managers and physicians in management training programs. Some independent schools of public health and health service management have been established (the Czech Re- public, Hungary, Poland), but numerous courses, diploma programs, and professional networks are calling attention to the importance of improving internal management practices. Market Exposure Market exposure determines to what extent hospitals are at risk for their performance: whether they lose revenue as they treat fewer pa- tients and, conversely, whether they gain revenue as they treat more patients. The level of hospitals' exposure to the disciplining force of the market is jointly determined by the provider payment mecha- nism and the level of direct out-of-pocket charges from patients (chapter 1, this volume). In CEE and FSU, hospitals'formal market exposure through direct out-of-pocket payments is low because few countries have introduced official copayments. Nevertheless, the widespread practice of informal gratuity payments creates significant effective market exposure. During the communist era, all health care, including hospitaliza- tion, was free of charge for patients. During the transition, a few 222 o Innovations in Health Service Delivery countries have defined user charges for hospital use. For example, in Croatia, patients are charged 15 percent hotel costs and a fixed flat amount, both centrally set. In Latvia, copayment levels are centrally determined as part of the health insurance benefit package. In Geor- gia, hospitals are allowed to charge copayments for services in the municipal benefit package. The proposed copayment levels must be submitted to the MOH annually for approval. The price list must be posted in a visible place in the hospital. In most cases, however, user fees affect marginal areas of hospital admissions. For example, in Hungary, hospitals can charge patients who arrive without appropri- ate referral and/or insurance coverage and can set the price. Formal user charges, however, and their (potential) impact on hos- pital performance have to be evaluated in light of the informal pay- ment practice. Lewis defines informal payments as "payments to indi- vidual and institutional providers in-kind or in cash that are outside official payment channels, or are purchases meant to be covered by the health care system."18 Formal user charges are "crowded out" by the practice of "informal" payments. This trend jeopardizes the fi- nancial health of hospitals, since official copayments contribute to the overall hospital budget, whereas informal payments are retained by physicians and other staff members. As a result, medical equipment upgrades, innovations requiring up-front investment, cost-effective medical protocols, raising of nursing standards, and other elements of a functioning health care system lack appropriate funding. Ministries of Finance and international donors often argue that the introduction of official copayments will automatically drive out informal gratuities. Current experience shows, however, that with low physician salaries and in the absence of enforcement, physicians waive official copayments in exchange for lower informal "gratu- ities"-a win-win situation for physicians and patients, but hospitals as institutions lose out. Informnal user fees make physicians directly accountable to patients and allow patients to obtain higher quality services than they would be able to purchase officially. In this sense, informal user charges cre- ate direct incentives for physicians to improve the responsiveness of their services. However, informal payments create many distortions. The Missing Link? Hospital Reform in Transition Economies * 223 For one, the purported responsiveness improvement takes place only for patients who can pay. Further, out-of-pocket payments-formal or informal-restrict access for people who cannot pay, and payment levels are usually quite arbitrary. In this sense, informal payment is a less desirable form of out-of-pocket payment than formal payment in that patients cannot be protected from the financial loss resulting from illness. Finally, the more widespread the practice, the less doc- tors are interested and invested in the reform of the public health care system because they have the best of both worlds: their own pri- vate business run within the safety of the public system.19 The relative weight of informal payment to public funding differs markedly in the systems we reviewed. The highest proportion in our sample was in Georgia-around 70 to 80 percent of hospital rev- enues derived from informal payments. This is in marked contrast with countries like Croatia, Hungary, and the Baltics, where esti- mates suggest that informal payments do not exceed 10 to 20 percent of total hospital revenues.20 Hospitals' market exposure, and the re- sulting expectation for their behavior, varied with the relative weight of public to private payments. The impact of informal payments on hospital and physician behavior is expected to be much greater in the region's low-income countries, where public financing collapsed, than in the early-reform, higher income countries. Residual Claimant Status The organization's residual claimant status reflects its degree of financial responsibility-both its ability to keep savings and its re- sponsibility for financial losses (debt). A hospital's residual claimant status is a key incentive to generate savings and efficiency gains. In the communist era, the central budget was the residual claimant: the MOH took back and reallocated any resources that hospitals did not spend. Since hospitals had no residual claims on revenue flows, cou- pled with input-based, line-item budgets (often with more than 30 line items), they had no incentives to generate savings and efficiency gains. During transition, the public purse has ceased to be a residual claimant as new payment mechanisms have been introduced. As most 224 o Innovations in Health Service Delivery countries are moving toward output-based payment systems, hospitals are automatically becoming residual claimants.21 Moreover, hospitals are increasingly able to generate and keep their own revenue (in addi- tion to the purchaser or central budget) through four main mechanisms: charging copayments, renting out facilities, collecting donations, and offering "corporate services" to private companies (e.g., screening). The other aspect of the residual claimant status is the hardness of the budget. Hospitals are not held liable for their deficits as most countries have not been successful in enforcing hard budgets for hos- pitals. This has weakened incentives to achieve savings and efficiency gains. Hospitals in nearly all countries have run up debt, and nearly every administration has responded by a centrally arranged bailout, repeatedly in many cases. Debt was accumulated to different parties, most typically to pharmaceutical companies that had not been fully privatized (e.g., Albania), to utility companies (Croatia, the Czech Republic), and physicians (Albania, Georgia). In Hungary, hospitals accumulated debt every year from 1995. Initially, few hospital direc- tors were replaced, and hospitals runmng losses continued to receive interest-free loans from central budget resources that were not paid back. To stop the process, the public finance act was amended in 1998, making owners explicitly responsible for hospitals' financial losses, and bankruptcy commissioners have been appointed to over- see problem hospitals. The Czech Republic has been an exception in its handling of hos- pital debt in line with its more market-oriented health sector strat- egy. Only two hospitals were offered interest-free loans from the state-owned Consolidation Bank; the others were required to repay loans from their own future savings. The Baltics are also an excep- tion in that hospitals have not incurred any debt, primarily because inpatient care expenditures in real terms have increased. Whether this financial discipline will be maintained is questionable, as the health sector is being subjected to tighter budgets. Accountability In the socialist era, accountability was ensured by hierarchical, direct administrative control, exercised by the Ministry of Health or its The Missing Link? Hospital Reform in Transition Economies * 225 regional offices. This control consisted of, for example, financial inspections to ensure that resources were spent according to the budget line items, and the ability to hire and fire staff. Thus, ac- countability, like other aspects of socialist health systems, focused heavily on inputs. With health sector reform, the MOH has been divesting its func- tions, including its powers of direct supervision and control. How- ever, the new organizations that assume these functions (local gov- ernments and social insurance organizations) have been unable to develop appropriate accountability arrangements. This has created an accountability vacuum in the region. Accountability to owners. As hospital owners, municipalities lack the in- centives, instruments, and capacity to hold their hospitals account- able for their performance. This is the case both in terms of financial performance and service quality. The idea behind decentralized ownership was that responsibility for service delivery would be trans- ferred closer to the people who, through the local electoral process, would assert their expectations for hospital services. This seems to be working in most countries. But the problem lies with the inability of local governments to hold hospitals accountable for their perfor- mance. Since local governments do not finance health services and usually have htde control over capital investments, they lack the in- struments to influence hospital behavior. Thus, their response to the complaints of their local electorate consists of putting the blame for unsatisfactory services on the central government and on purchasers for not providing adequate funding for hospitals. For example, in Hungary, as hospitals are independent legal enti- ties and receive their budget from the Health Insurance Fund, the municipalities, as owners, have no legal right to supervise and mon- itor internal hospital processes. Hospitals refuse to allow local gov- ernments to look into their activities and account books. This prob- lem became acute in 1999 when municipalities were made legally responsible for hospitals' financial losses. Some countries have been experimenting with improving hospital accountability to owners by creating governance boards (the Czech 226 O Innovations m Health Service Delivery Republic, Estonia, Latvia). Whether these boards can create a mean- ingful link between the organization and its owner is questionable. In the Czech Republic, for instance, although the minister of health ap- pointed board members, hospital directors were also asked to nomi- nate people. Furthermore, the boards work on the basis of overall im- pressions, not hard data provided by hospitals. In Latvia, the boards of joint stock-company hospitals consist of only three people.22 Accountability to purchasers. Similar issues arise regarding accountabil- ity to funders. Ample evidence underscores the need to improve ac- countability to purchasers: all countries that moved to performance- based financing are encountering fraud in performance reporting. Although social insurance organizations could theoretically rely on contracts as new accountability mechanisms, contracts are not used as instruments. The contracting process is not performance-oriented: performance measures, targets, and benchmarks are not relied on and there is no selective contracting with providers. In most countries, purchasers are required to contract with all publicly owned facilities. Hungary is an example of the weak use of contracts as purchasing and accountability instruments. Contracts specify the contracted hospital's capacity in terms of number of beds and physician-hours provided but remain silent on volume, service mix, and quality even though payments to hospitals are based on their output (DRG basis). Even the more market-oriented reformers have shied away from relying on performance pressure from purchasers. In the Czech Re- public, insurers were initially required to contract with any hospitals that applied for a contract. Since 1995, selective contracting has been allowed in theory. In practice, however, contracting decisions are not made by insurers but by a committee of health insurance fund rep- resentatives, the MOH, the Chamber of Physicians, and the Hospi- tal Association. No contracts have been withdrawn or refused, and only some marginal shifting of services has occurred. In Romania, how new legislation translates into practice will be interesting to see because it explicitly allows selective contracting and attempts to en- sure sector neutrality by allowing insurance houses to contract with private as well as public providers. Hungary had difficulty imple- The Missing Link? Hospital Reform in Transition Economies * 227 menting sector neutrality for lack of any clear guidelines on con- tracting with private sector providers. Such inconsistencies result from the difficulty of moving from central planning of inputs to proactive purchasing, based on outputs and performance targets. This requires institutional adjustments in rules and regulations, capacity building, and change in public sector culture and norms. Accountability to the Ministry of Health. Quality assurance and mini- mum standards are in the early stages and, although they have be- come catchwords, most countries are unsure how to use such instru- ments. In Hungary, physicians defined minimum standards and, in the first inspection, most hospitals and departments did not meet them. Since bringing all of them up to the level of the defined stan- dards (for equipment, access to laboratories, and the like) would have been too costly, the issue has been temporarily taken off the agenda. Accountability to patients. Structures for accountability to patients are lacking, since there are no accessible procedures for formal patient complaints. The exception again is the Czech Republic where deal- ing with the increased number of patient complaints is one of the main new tasks of the Minstry of Health. Since none of these coun- tries has created systems where funds flow in response to patient de- mand, they have funding systems that do not increase accountability to patients. Social Functions Social functions in the communist era were unfunded and implicit. These social functions included hospitalizing nonmedical cases such as dependent elderly people. With increased financial pressure, these traditional functions are gradually disappearing, but unfunded and implicit social functions are appearing in new forms. For example, while health insurance is compulsory in most countries, many peo- ple fall through the social safety net. Although their number may be small, there is no information on what happens to them. In the 228 0 Innovations in Health Service Delivery Czech Republic, hospitals incur the cost of treating people without insurance. In Romania, the recent insurance legislation assumes that everyone is covered, but this is not the case. Hospitals in lower in- come countries bear the cost of social functions in a different way: hospital budgets are often delayed for months, and physicians con- tinue to work without pay. In some sense, in these consistently underfunded systems, all hospital services have become social ser- vices-with volumes unspecified and services delivered at the dis- cretion of the staff. Summary of Organizational Structure As table 5.3 attempts to illustrate, hospitals in the reviewed CEE and FSU countries no longer function as direct budgetary units of the core public sector bureaucracy, as some change occurred in the five key elements of their organizational structure. At the same time, the current organizational structure of these hospitals cannot be clearly labeled, for lack of consistency in their five key organiza- tional features. The resulting inconsistency in their overall incen- tive regime has contributed to their limited success in improving their performance. Although hospital autonomy has increased in various decision areas, additional regulations and/or political pressures have limited the de- cision rights of hospital managers in practice. Social functions have not changed much either, as hospitals continue to provide unfunded social functions. In contrast, residual claimant status, market exposure, and accountability structures have undergone significant changes. The public purse has ceased to be the residual claimant of savings, unspent allocations, and efficiency gains, at least under the new payment mechanisms. The increase in informal payments has created some el- ements of market exposure, though in a distorted form. But perhaps the most striking feature of the current organiza- tional structure is the lack of effective accountability mechanisms. As the nearly empty accountability row in table 5.3 suggests, most countries do not enforce direct accountability through the hierar- chy or through explicit regulations and contracting. The changes in .2 . E (D -0 :C - 0 0 C~~~~~ C - E 0 ~~~~~~~~~~~~~~~~0C U 4 0 ~ ~~~~ CX -0 OC)~~( C C _ -~~~~~~~~~~~~~~~~L aC O 0 ~~~~~~~~ -CC) 0~~~~~~~~ - 0) ~~~~~~~0-0i 04 CN C14-a ~ )o C -02 04 5D 0~~ c 0~~~~~~~~~~~~~~~~~~~~ CN a~~~~~~~~~~~~~~~ U-C - .-CNa -~~~u N C --~~~~~(~~~< N N 2' a"Z, Z Li 0)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - ____ ____ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ "t -V)C C ~~~~~~~~~~~~~~~~~~~~~~LU -LU C a) - C)C P E C)- J C~~~~~~~~~~~~~~~~~uC LU a_E .C S C)0C 0CC N o.0- 2 C N~~~~~~~~~~~ C ) 0 CU d C QC C) 00~~~~~~~~~~~~~ )2 E pE 0)~~~~~~~~~~~~~~~~~~~ .2 C C) a 2 - 0U'Ž (J0 -,C)C 0) a :5! C) U' 0 -~~~~~~~~~~~~~ _ 0~~~~~~~~~~~~~ 2 U 0 ~~~~~~~~~~~~~U 0 0o < < - 230 ° Innovations in Health Service Delivery the external environment removed several decision rights from the Ministry of Health, including its powers of direct supervision and control. The new organizations that received these decision rights have been unable or unmotivated to develop appropriate accounta- bility arrangements. Dmplicolions for Mospi5ao Behavior and IPermonnance The changed external environment put new pressures on hospitals to adapt their behavior. Reduced real budgets and new provider pay- ment mechanisms were expected to trigger efficiency gains, while de- centralized governance was expected to improve responsiveness to community needs and user expectations. However, because the orga- nizational structure was inconsistent with the external environment, incentives for behavioral change have been weak and contradictory, and potential synergies have been lost. The incoherence between internal and external incentive envi- ronments has had three manifestations. First, decentralization blurred the governance structures. The operational meaning of local government ownership and governance has remained unclear in so- cial insurance-funded systems. In terms of funding, local govern- ments have played a small role, as hospitals' operating expenditures come from social insurance receipts, and capital investment allo- cations come mainly from the central budget and only to a limited extent from local government budgets. Thus, in terms of financial responsibility, local governments are limited in their ability to influ- ence hospitals' strategic development. The legal status of hospitals under local government ownership is a further complication. When social insurance organizations were established and required to contract with hospitals, hospitals needed to be granted legally independent status to sign binding contracts and to act as legally recognizable contracting partners. At the begin- ning of the transition, they did not have this independence as bud- getary units of the core government bureaucracy. This dilemma over the meaning and enforceability of contracts between two public bod- The Missing Link? Hospital Reform in Transinon Economies * 231 ies provided the opportunity to systematically rethink hospitals' or- ganizational form and governance structure. Nevertheless, few coun- tries gave attention to this issue, and most did not develop new laws regarding hospital governance. Instead, the issue was quickly ad- dressed under existing laws designed for the governance of general nonprofit organizations and state-owned companies. As a result, hos- pitals were turned into extrabudgetary funds, nonprofit institutions, and state-owned enterprises, regardless of whether these forms and existing regulations were appropriate for the health sector. The second manifestation of incoherent external incentives and organizational structure is the misalignmnent between incentives and decision rights. Decision rights were not transferred to the organiza- tional level that would benefit from introducing behavioral change: hospitals had incentives to change their behavior but not the instru- ments. Anticipation of greater (technical) efficiency was based on the expectation that hospitals would respond to niew payment incentives and financial pressure by reducing excess physical and human resource capacity. These two inputs are the most significant for savings and efficiency gains. Personnel expenditures make up 60 percent of all health care costs in the region's middle-income countries, while phys- ical capacity determines most fixed costs. Since hospitals had no deci- sion rights over physical capacity and were limited in their decision rights over human resources, the expectation that hospitals would re- duce their inputs to produce the same level of output was unfounded. In contrast to hospitals, local governments do have decision rights over physical assets, but they lack incentives to divest or manage them well. Closure of a local hospital appears to the local electorate as a loss of community assets and as a local government failure to ful- fill its legal (sometimes constitutional) mandate to deliver services. As a result, old patterns of focusing on inputs persist, despite changes in external structures and increasing financial pressure. Finally, hospitals do not incur costs for not adjusting their opera- tion and performance. The interaction between the lack of market exposure and the lack of accountability structures created an incen- tive environment that did not penalize poor hospital behavior and performance. Repeated government bailouts of debt-ridden hospi- 232 o Innovations in Health Service Delivery tals, lack of monitoring of financial and nonfinancial performance, lack of reporting requirements, and the absence of any exit threat meant that hospitals could continue their previous patterns of be- havior without paying any penalty. They were rewarded, however, for improved performance in that they could keep savings and effi- ciency gains. As a result, behavioral improvement depended on the hospital managers' drive and entrepreneurial spirit. Conclusions Coherence with Other Reforms Based on this review, we looked only at the extent to which organi- zational reform in the hospital sector of the higher income countries has been consistent with other reforms introduced during the tran- sition. As mentioned, the main health sector reform elements in these countries included the establishment of social insurance with performance-based provider payment arrangements and explicit con- tracting with decentralized providers. The rationale for introducing payment reforms before organizational reforms was that hospitals would respond to the new provider payment incentives by becoming more efficient and would curtail their use of inputs. However, the current organizational structure of many hospitals does not allow them to respond optimally to the incentives introduced through the funding reforms. Some restrain (limited decision rights) while others undermine hospitals' ability to respond appropriately to the new pro- vider payment mechanisms (soft budgets, market exposure through informal payments, lack of accountability). The consistency of hospitals' current organizational structure with local government ownership is a more complex issue. Local govern- ments' role as owner is not well defined, and their control over hos- pital operations is limited. They have a key function and opportunity, however, in terms of their decision rights over their hospitals' physi- cal assets. As various (unsuccessful) attempts to close hospitals have demonstrated, local governments value these assets for reasons other The Missing Lmk) Hospital Reform in fransition Economies * 233 than the hospitals' performance and may oppose organizational re- form that would reduce their control over these assets. A key issue here is that the more assets a local government has, the bigger its run- ning budget must be to maintain them. Since money translates indi- recdy into political power, the incentive to downsize is small under the current organizational arrangements. Coherence with Capacity to Implement Change In chapter 1 of this volume, the authors point out that organizational reforms are often complex and require a sophisticated policymaking and implementation capacity to design and carry through. While ed- ucation levels in CEE and FSU are high, policymakers and staff often lack the relevant experience, training, and background for the type of reforms needed during hospital corporatization. Further- more, the last decade has left many CEE and FSU policymakers and citizens with reform fatigue and skepticism about new initiatives. Thus, a challenge to structural reform is not greeted with much enthusiasm. Nevertheless, the promise of new opportunities could appeal to the medical profession, which feels that it has not fully ben- efited from the transition to a market economy. We offered a number of hypotheses to explain the lack of attention to the organizational structure of hospitals in the CEE and FSU countries. We also suggested that the failure to address these issues has slowed or blocked efficiency gains. Though organizational re- forms are complex and challenging to design and implement, it is clear that many of the countries in this region will find them a neces- sary component of the next phase of their health sector development. The countries where health financing and provider payment re- forms are well under way (e.g., Croatia, the Czech Republic, Hun- gary) are likely to find themselves considering such reforms in the near future. More groundwork is probably needed in other countries (e.g., Albania, Bulgaria, Romania). More moderate hospital auton- omy may be a way for these countries to ease into a progressive re- form process for their hospitals. In low-income countries, where the priority is to mobilize resources for the sector and to rebuild pre- 234 o Innovatons in Health Service Delivery payment plans, complex organizational reforms are probably ill ad- vised in the near future. For countries contemplating hospital organizational reform, the key issues to address include aligning incentives and decision rights, enforcing hard budgets, and introducing new accountability mecha- nisms through more effective use of contracts, quality assurance, and performance monitoring. This will necessitate clarifying the ex- pected future ownership role of local governments and completing the transition to a more proactive health-service purchasing model. Finally, reorienting the capacities of the Ministries of Health to ful- fill a strategy-setting stewardship function is essential for the success of any kind of health reform in the region. Notes 1. Throughout this text, we use transition to mean the process of moving from a centrally planned, largely state-owned economy to a market-based economy. By transition economies, we mean all countries of the former communist block in CEE and FSU. By transition pe- riod, we mean the time from 1990 to the present. 2. T. Ensor, "Health System Reform in Former Socialist Coun- tries of Europe," International Journal of Health Planning and Man- agement 8: 169-87 (1993); R. B. Saltman and J. Figueras, European Health Care Reform: Analysis of Current Strategies (Copenhagen: WHO Regional Office for Europe, 1997); J. Klugman and G. Schieber, "A Survey of Health Reform in Central Asia," in R. Feachem, Z. M. Henscher, and L. Rose, eds., "Implementing Health Sector Reform in Central Asia," EDI Learning Resources Series, World Bank, Wash- ington, D.C., 1997; E. Goldstein, A. S. Preker, 0. Adeyi, and G. Chellaraj, "Trends in Health Status, Service and Finance: The Tran- sition in Central and Eastern Europe," World Bank Technical Paper 341, Social Challenges of Transition Series, Washington, D.C., 1996; and M. Jakab, A. S. Preker, and A. Harding, "Linking Organizational Structure to the External Environment: Experiences from Hospital Reform in Transition Economies," in M. McKee and J. Healy, eds., The Missing Link) Hospital Reform in Transition Economies * 235 Hospitals in a Changing Europe (Buckingham, U.K.: Open University Press, 2002), pp. 170-202. 3. T Ho, "Hospital Restructuring in Central Eastern Europe and the Former Soviet Union," in M. McKee and J. Healy, eds., Hospi- tals in a Changing Europe (Buckingham, U.K.: Open University Press, 2002); Goldstein et al., "Trends in Health Status, Service and Finance"; V Staines, "A Health Sector Strategy for the Europe and Central Asia Region," Health, Nutrition and Population Series, World Bank, Human Development Network, Washington, D.C., 1999. 4. The 11 countries reviewed include: Albania, Croatia, the Czech Republic, Estonia, Georgia, Hungary, Kazakhstan, Latvia, Lithua- nia, Poland, and Romania. 5. W Hsiao, "Risk and Incentive: The Tale of Modem Provider Payment Systems" (paper presented at the International Conference on Health System Reform, Taipei, Taiwan, China, 1997); A. Maynard and K. Bloor, "Payment, Contracting, and Regulation of Providers," Flagship Course Material on Health Sector Reform and Sustainable Financing, World Bank, Washington, D.C., 1999; M. M. Wiley, "Hospital Financing and Payment Systems: A Review of Selected OECD Countries," Report for the Organization for Economic Co- operation and Development, Paris, 1995; and M. M. Wiley, "Hospi- tal Financing Reform and Case-Mix Measurement: An International Review," Health Care Financing Review 13 (4): 119-33 (summer 1992). 6. Ho, "Hospital Restructuring in Central Eastern Europe and the Former Soviet Union." 7. A. S. Preker and R. Feachem, "Market Mechanisms and the Health Sector in Central and Eastern Europe," World Bank Techni- cal Paper 293, Washington, D.C., 1995. 8. Ho, "Hospital Restructuring in Central Eastern Europe and the Former Soviet Union." 9. J. Langenbrunner and M. Wiley, "Paying the Hospital: Pay- ment Policies and Reforms in Eastern Europe and the Former So- 236 o Innovations in Health Service Delivery viet Union," in M. McKee, ed., An Appropriate Role for the Hospital (Buckingham, U.K.: Open University Press, 2002). 10. S. Schiavo-Campo, ed., Institutional Change and the Public Sec- tor in Transitional Economies, World Bank Discussion Paper 241, Washington, D.C., 1994; Saltman and Figueras, European Health Care Reform. 11. Observatory (European Observatory on Health Care Sys- tems), Health Care Systems in Transition: Hungary. http://www.who. dk/observatory/hits/hits.htm, 1999; and Health Care Systems in Tran- sition: Poland. http:// www.who.dk/observatory/hits/hits.htm, 1999. 12. Observatory, Health Care Systens in Transition: Romania. http:// www.who.dk/observatory/hits/hits.htm, 1996. 13. Observatory, Health Care Systems in Transition: Estonia. http:// www.who.dk/observatory/hits/hits.htm, 1996; Health Care Systems in Transition: Latvia. http://www.who.dk/observatory/hits/hits.htm, 1996; Health Care Systems in Transition: Lithuania. http://www.who. dk/observatory/hits/hits.htm, 1996. 14. Personal communication, Besim Nuri, Health Economist, World Bank Resident Mission, Tirana, Albania, July 1999. 15. Observatory, Health Care Systems in Transition: Croatia. http:// www.who.dk/observatory/hits/hits.htm, 1999; and Health Care Sys- tems in Transition: Czech Republic. http://www.who.dk/observatory/ hits/hits.htm, 1996; World Bank, "Health Policy Note: Croatia," World Bank, Europe and Central Asia Department, Human Devel- opment Unit, Washington, D.C., 1999. 16. Historically, control and accountability over the use of inputs in accordance with the plan were relatively strong-but there was virtually no scrutiny or accountability for the operation of the hospi- tal, the quality of care, or the delivery of outputs. 17. Personal communication, Silviu Radulescu, Health Econo- mist, World Bank Resident Mission, Bucharest, Romania, July 1999. The Missing LinkO Hospital Reforn in Transition Econornies * 237 18. M. Lewis, "Informal Health Payments in Eastern Europe and the Former Soviet Union: Issues, Trends, and Policy Implications," in E. Mossalios and A. Maresso, eds., Funding Health Care: Options in Europe (Buckingham, U.K., Free Press, 2000), p. 184. 19. Ibid. 20. Ibid.; A. S. Preker, M. Jakab; and M. Schneider, "Health Fi- nancing Reform in Transition Economies," in Mossalios and Ma- resso, Funding Health Care: Options in Europe, pp. 80-108. 21. Langenbrunner and Wiley, "Paying the Hospital." 22. Given the general lack of functioning accountability struc- tures for hospitals throughout the region, research on the potential and actual effectiveness of boards in enhancing accountability is needed. CHAPTER 6 Contracting Public Health Care Services in Latin America Alexandre Abrantes Throughout the 1990s, several Latin American governments-in- cluding Argentina, Brazil, Chile, Colombia, Costa Rica, Mexico, Panama, and Uruguay-have undertaken reforms to the organiza- tion of and payment for the delivery of hospital services. Both the public and governments have sought change. The public has felt that public hospitals offer too little choice, do not respond to patients' needs, and often deliver poor-quality care. Governments face rising expenditures, with no corresponding growth in productivity, effi- ciency, quality-or public satisfaction. Traditionally, public hospitals and ambulatory networks in Latin America have cared for the poor and have been used by the middle classes for care that could result in catastrophic expenditures. Ensur- ing adequate governance of public hospitals has been difficult be- cause they have had ill-defined and conflicting objectives and weak and input-focused accountability mechanisms. As a result, many public hospitals have been inefficient and their service quality, poor. Under tight fiscal constraints and public pressure, Latin American governments have emulated the "marketizing" organizational re- forms undertaken in several OECD countries. This chapter reviews the early experiences of Argentina, Chile, and Uruguay with this type of reform, and attempts to shed light on the varied levels of success with implementation. Though it is too 239 240 ° Innovations in Health Service Delivery early to evaluate the impact of these reforms, information about the evolution of some production and efficiency indicators are presented wherever available. Case Studies The conceptual framework developed in chapter 1 is utilized to make cross-country comparisons easier. Argentina, Chile, and Uruguay were among the first Latin Ainer- ican countries to engage in the organizational reforms of hospitals. Because they are geographically and culturally close, significant ex- change of experiences took place, as illustrated by the cases de- scribed. Although they embarked on reforms with the same overall objective, reform progressed further in Argentina than in Chile or Uruguay. This chapter explores why implementation was more suc- cessful in one country than in the others. Argentina The organizational reforms initiated by the federal government in 1993 were a response to the following issues: o Increasing demand for hospital services by the uninsured, due to an increasing number of unemployed and the consequent reduc- tions in coverage by social insurance;1 o Additional demand by beneficiaries of poor or poorly managed so- cial insurance funds (obras sociales), who preferred the public hos- pitals to low-quality/copayment private services; o Lack of reimbursement of hospitals by social insurance funds for services rendered, thus creating a subsidy to the social insurance funds from the public purse-and straining further the resources available to ensure access and quality for uninsured patients; o Highly centralized, rigid hospital management. Despite previous decentralization efforts (from the federal to the provincial level), Contracting Public Health Care Services in Latin Arnerica * 241 hospital management remained very centralized and rigid. This rigidity was believed to be a major determinant of the perfor- mance problems, especially problems related to changing health services to meet new/different demands, allocation of resources, management of personnel, purchasing, contracting, governance, and the like. Reforms There were a few experimental efforts with organizational reform as far back as 1966. The National Pediatric Hospital Dr. Juan Garra- han and the Clinics Hospital Jose de San Martin are examples of these early attempts and will be described below. But the strong leg- islative push for far-reaching hospital reform did not come until the 1990s. In 1993, the Federal Government of Argentina initiated several changes to these problems, including organizational reform of the public hospitals, increasing decision rights of hospital administrators, exposing public hospitals to some degree of market risk, allowing hospitals to claim locally raised revenue and establishing new ac- countability systems.2 Decision rights. Self-managed public hospitals (SMPHs) would control their operative plan and annual budget and could propose new ser- vices and programs, enter into contract agreements with health in- surance carriers, contract-out services, and designate, promote, and transfer personnel within an agreed structure according to their juris- diction. All these rights and responsibilities were to be performed at the hospital level, thus removed from the Ministry of Health. Market exposure. SMPHs would start to raise an additional portion of their income from the delivery of services to insured patients.3 The hospital would continue to receive the bulk of its income from the provincial budget. In addition, the law envisioned a shift from historical-based budgeting of inputs to funding of outputs delivered to public patients (through the establishment of provincial health 242 ° Innovations in Health Service Delivery insurance funds)-so that patients' choices as to where they would go for treatment would bring resources with them. In other words, money was intended to reimburse those services actually demanded, instead of what was "offered." Residual claimant. SMPHs would be allowed to collect fees from in- surers and through the workers compensation scheme and charge copayments to patients with the ability to pay. This revenue was to be divided, according to a provincial law, among personnel, invest- ments, and operation, to motivate their performance and hence pro- duction and efficiency gains. Accountability. Given the planned delegation of some substantial de- cision rights, it was necessary to create accountability mechanisms that would work even with reduced influence over the day-to-day operation of the hospitals. To this end, the reform included the fol- lowing provisions: o The functioning of a board of directors (consejo de administraci6n), which would supervise the general performance of management and the hospital overall o The requirement that each reformed hospital would have to meet minimum targets for production, efficiency, and quality, which the program would establish for each category o The requirement that each hospital would have to approve the pe- riodic evaluation of efficiency control and quality that the author- ity defined. Socialfunctions. SMPHs would continue to serve primarily the poor and uninsured. As noted above, the intention was to move toward the establish- ment of provincial health funds-which would reimburse hospitals for services provided to public patients. No mechanism was outlined for funding of research or training (the decree mentions that the hos- pital would continue to receive its budget, in part for these activities), Contracting Public Health Care Services in Latin America * 243 though SMAPHs were clearly intended to continue to provide these services. Because public hospitals belong mainly to provinces and some municipalities, the federally legislated reform design was imple- mented quite differendy in various jurisdictions.4 In such provinces as Entre Rios, Formosa, La Rioja, San Luis, Santiago del Estero, and Tucumin most hospitals remained in essence organizations; in prov- inces such as Salta, public hospitals have been almost corporatized. Corrientes, Catamarca, La Pampa, Mendoza, and Rio Negro have made their most important hospitals autonomous organizations. In the area of management control, most self-managed hospital admin- istrators in these provinces gained good control over nonlabor in- puts. They can move or promote staff within a set framework and promote voluntary retirement, although they cannot contract with or dismiss staff. They can rent facilities or subcontract services but cannot acquire or sell assets. Though professing a desire to move to demand- and production-based budgets, these provinces have kept their historically based budgeting systems.5 All have promoted cost recovery for services provided to mandatory and voluntary health in- surance beneficiaries and have allowed hospitals to retain up to 80 percent of these revenues to distribute among investment, mainte- nance, and staff performance bonuses. All have set administrative and technical boards and have appointed general managers, but the composition of such boards and the rules to nominate boards and managers still make them dependent on the political process and cycle. Most have developed management contracts templates, in- cluding performance indicators, but only Rio Negro actually got these agreements with its self-managed hospitals, and tied the distri- bution of staff performance bonuses to meeting agreed targets in successive quarters. Finally, social functions are either not mentioned or not funded. The provinces of Buenos Aires, Entre Rios, and Santa Fe have passed similar legislation but have been slow or inconsistent in implementation. Salta and San Juan have passed the most far-reaching hospital re- form legislation. Under the new rules, self-managed hospitals would become what we could classify as corporatized organizations.6 The o y X X U X L - -- 2z z xi z z z z xU ux z z u ZZWZZZZL ZZ 52=o vov>Ovo>0 U oooU oo V o D u( o E E D E D zzz: zOz z E zc 0 O O -D g -D D- C, -D - 0) C~~~~~~~~~~~~~- M~~~~~~ O Z j @ o E E ,° E 2440 0 0 Z 2~~~~~ 0 r_ z = 00 0~ 00 .2 4) od0d 0 CY) C) 0~~~~~~~~~~~ 0 0 0 Cd) 0~~E 0 -o 4C 0 CO p E 70 244 Contracting Public Health Care Services in Latin America * 245 new legislation: (1) legally establishes these hospitals as public trusts, operating under both private and public law, in which they become the full trustees of all assets and personnel; (2) sets a mechanism to switch, over a period of three years, from historical budgets to demand-driven budgets (box 6.1); (3) creates independent hospital managers, with fixed-term contracts, accountable to a hospital board and independent from the political cycle; (4) provides for manage- ment contracts that include performance indicators covering pro- duction, efficiency, and quality of care (e.g., targets for number of consultations, discharges, surgeries, emergency room visits); (5) gives the hospital full claim to the revenue generated at the hospital and power to incur debt (after authorization by the Provincial Ministry of Health); and (6) makes explicit, and funds, some of its social functions. The extent of decision rights delegated to management varies widely within the autonomous organization category. Alvarez and Prieto developed a highly precise hospital autonomy index, com- posed of six categories of decision rights.7 Their index allows greater differentiation within the category "autonomous," and is being ap- plied to monitor the evolution of reformed hospitals. None of these reforms has been in place long enough to measure its impact on process or output. Information is available only on the revenue generated by cost recovery. Invoicing, though still generally poor, is improving. Collection rates are low, despite the automatic collection system. Revenues from cost recovery are typically around 10 percent of the hospital budget but can be as low as 3 percent or as high as 3 0 percent at tertiary hospitals such as Dr. Juan Garrahan and .Jose de San Martin.' These revenues have become important sources of financing investment and maintenance. They also offer staff strong financial incentives to adopt new policies and standards of care (in some hospitals, these nonsalary bonuses can amount to 50 percent of the doctor's total income). Chile Chile's public hospital system was run as government departments prior to the refomi, though at the regional level. Chile opted to im- 246 o Innovations in Health Service Delivery Ban 6.1 From 13cge Publics smpW Bvu dgssD Mo$aftn ~o Monbudge0ary Revenua Chile and Argentina chose two different methods to help pub- lic hospitals move from historical budget allocations based on inputs, to demand-driven revenues based on outputs. In Chile, the new health service cost-accounting system, based on services provided, was introduced in 1996. It gen- erated an output-based "shadow budget," which reflected what the hospital would have earned if all of its revenues came from services provided to persons insured privately or by Fonasa. In some hospitals the shadow budget was larger than the historical one, while in others the reverse was true. Over a period of years, that gap was supposed to be elimi- nated by freezing budgets for regional health services that received more than they would have earned for services ren- dered and by allowing budget growth for other services that would have earned them more had the new system been in place. As the gap approached zero, health services would be switched from the input-based historical budget to the demand- and production-based budget. Although the shadow budgets have been operating for four years, no hospital or health service has yet been switched to the new budgeting mode, and the gap has been allowed to grow, instead of being forced to shrink. In Salta province, Argentina, historical budget allocations are to be reduced over three years, starting in January 1999, by 10 percent, 50 percent, and 90 percent, respectively. The revenue gap will be offset by charging the provincial gov- ernment for services provided to the uninsured poor. Hos- pitals will charge the provincial government on the same fee schedule it uses to charge the obras sociales and private insur- ers, and will be reimbursed 10 percent of charges in the first year, 50 percent in the second year, and 90 percent after the third year. By then public hospitals will have reached about 90 percent of market exposure. Contracting Public Health Care Services in Latin America * 247 plement the reform as the level of the regional health administration (RHA). In this sense the reform resembles somewhat the Victoria, Hong Kong, or Singapore models of establishing a group of health providers as an autonomous network, with subsequent government supervision focused on the headquarters rather than the hospitals.9 Reform Implementation The Ministry of Health has established management contracts with RHAs, which can function as an indirect accountability mechanism for autonomous hospitals. In Chile, however, these contracts resem- ble ministerial planning documents more than business contracts, and many regional health service administrations and some hospitals have become more autonomous organizations. The reform model is fairly "mild." Nevertheless, the RHAs have become somewhat more autonomous organizations. Decision rights. RHAs have been granted significant control over nonlabor inputs (i.e., procurement) but have little influence over se- lection, hiring, moving, promoting, or dismissing staff. Also, cover- age and the scope of activities delivered continue to be set by the MOH, and the manager cannot dispose of any assets or revenue gen- erated in the region or hospital. Before the reform, all these decisions were centralized at the MOH. Market exposure. While the MOH does contract with some RHAs, most continue to be financed primarily through direct budget alloca- tions, based on inputs and historical trends. Labor continued to be funded direcdy from the budget. They may charge private health in- surance carriers and the National Health Fund for services provided to their beneficiaries and can charge copayments to all but the poor- est uninsured patients. In reality, however, relatively litde is collected from either source. In the past few years, the govermnent has set up a shadow budget system based on: payment per production for most services rendered by the public hospitals and ambulatory clinics; diagnosis-related groups for about 25 clinical conditions, 11 of them 248 o Innovations in Health Service Delivery needing complex medical interventions that result in catastrophic health care expenditures; and capitation for general ambulatory care provided by municipal health services. The shadow system would allow private insurers to compensate public hospitals for services pro- vided to its beneficiaries and make the public money follow the pa- tient. However, no decision has been made to implement the system. While hospitals know how much they would receive if exposed to the market, they still receive direct budget transfers, as the transition to the new budget system has not yet taken place. Residual claim on revenue. RHAs have no claim on any revenues they earn by charging private health insurers, or from collecting copay- ments. Cost-recovery revenues are about 10 percent of total public health expenditures. This should not come as a surprise, as hospitals and health services are not allowed to keep any of the revenue gener- ated and therefore have no incentive to correctly identify insurance beneficiaries, or to collect from patients. Patients themselves have no incentive to identify themselves as insured, as that would require them to pay both a deductible and a copayment, but they pay nothing if they are not identified insured patients. In case of a deficit, the pub- lic purse is ultimately responsible, meaning that the ministry covers the difference between the planned and actual expenditures. Hence, the administrations also are not residual claimants on deficits. Accountability. RHA management personnel have remained civil ser- vants, directly appointed and supervised by the government. Their activity is framed by the "management agreements." However, they continue to be directly regulated by the MOH.10 The agreement, which resembles a planning-by-objectives document, is quite differ- ent from a typical hospital management agreement. In 1995, the standard agreement included 55 objectives and indicators, cover- ing mostly processes to implement government public health poli- cies and regulations. These stipulations include some of the usual health service production, efficiency, or quality-of-care indicators (e.g., shorten waiting lists, improve use of surgical units, reduce ab- Contracting Public Health Care Services in Latin America * 249 senteeism, adopt quality-of-care standards resulting from evidence- based medicine). This planning instrument is for the most part pro- duced centrally, at the MOH. Regional health service or hospital managers are "consulted" but have relatively little negotiating power in the agreement process. There is no incentive mechanism to facil- itate meeting the objectives or targets-financial or other-and in the case of noncompliance, no penalty is imposed. Several health service managers have defied the MOH by refusing to sign the pro- posed agreements, and none of them was disciplined or removed. In- vestrnent in the management information systems has been insuf- ficient to allow adequate monitoring of the proposed targets and indicators. Social finction. Social responsibilities of the RHAs are laid out in great detail in the management agreement, but the annual health service budget does not specifically address their funding. It is too soon to judge the impact of this mild variant of network autonomization on the public health system's production and effi- ciency. Some broad analysis seems to suggest the following: produc- tion has increased between 2 percent and 4 percent annually; since the implementation of agreements, fiscal delinquency among the re- gional health services has decreased to between 15 percent and 30 percent, compared to 100 percent before the agreements; and sys- temic indebtedness has continued to grow, despite the introduction of management agreements." No causal relationship can be in- ferred, however, because public health expenditure financing grew steadily even as management agreements were being introduced. This could explain the increase in health care utilization and the de- crease in health service deficits. Uruguay Most public hospitals in Uruguay are budgetary organizations, run as MOH departrnents. The government of Uruguay chose to pilot a reform model established within a "broader" interpretation of the existing public hospital legal framework. This arrangement empow- z 0 Z u a) 0 C 0 0) a L (D u a- 0 -0_-o 0) o x x , C - o z LU LLi 2 LLi .2 D- (D -0 o a E d) c, Z5 u E o u a 2 0 (QI> - 0 0 C a) Q. -YO 0 -O E E E c 0 S 0 o- a) 0) > O u = m u 0 Q. R D a 0 L 0 d) - (UD ti Ru u 0616-0 -,pcr-E O.-C Omooc 0 -0 z E 0 E E Z c .Z C) -0 0 -0 co a i2 4) 83 0) 0) d) -0 > > 21- > -O D a) -E 4) _O r- 0 c (D -0 C) ID 0 'r- CL > C: z - 2 -00) 6 -00 -;; c .- 0) d) (D O C) 2 0 -0 2 x u 0 0 r- E C f (D x 0 -O u- c c CL 'c E E Z 0- Z u 6-2 6 (D 'P E -U 7 ou 0 CL, r- -O 0 (D 0) E > 0 0 E .> u 'c 2 oo) a 0 ce 0 C) z 0 -0 U) -0 -0 0- >. co 0. O > = 0 a) a c -0 0 a) O U m t E -0 u O- E o T -0 0) 2 Z E 0 E -V 0 D > -0 ci 0 0) P 0) 0 u 0 -C -0 -0 u -0 -0 d) 0 O A: -2 -F , >, , c 0 _co, .2 -o E O -0u u -0 -0 0 < 0 E -0 -O a)> 0 2 0)0 0 O r- - a 0 OL 0 E .4 0 z, U- 0 g ~~~ ~~- -oE - C -0_ao > 0 C ~~~~~~~32o~~~C 2 0- 0~~~~~~~~~~ C00 U0 0D-- o . U 0 0tF 0) U 2 - 0 - 0 C) 0U 0~~~~~~~~ 2- 0 0C 0 T-~~~~~~~~~~~~~ 0 0 0 02 0) U~~~~~~~~ C c Z 0E '0) C') Z O-C_0 - 0- I-~~~Uo ~ a)~~~~~5 0 o 0C U *U: ~0 0 0) 0 c 00 )10) E0 a Q- - 0- '1)C 22E ~~~~~~~. -0~~~~~~~~~~~~C -0~~~~~~~~~~~~ D0 0(0 D ~~~ U ~~~~~ a.~ -EC -0 E L 0 9 N~ C ~30 .2 2*0 -o- 5 a)~ C-0 ~~~ *0 o= 00 0~ 0 p0 E a o0 E~ - U i> I 00 00 . a- 9-~oo E U0 U -2 0 CO v C C3 C C N O 00( I 0 - C 0 . 2 - o *Eoo ~~ *-0C_e 0 o~O C~0 .2 o.0= .0 E~ . 0 .72 0 0 00 0. 0.0 o0 0 0 > 0 . 0 2c 0 Eo o O- 0 0 a- cl.-o E 0) 0-0 0- 00-~~0 C)0a0 0 C) -0 0 E 2 >~~~~0) 0-E~ 200E a - c 0 -0 ~ ~ ~ > - 0) 0~ ~ ~ ~~- . 0 ~~~~0) 0 .0 -0 0~~C D 0 4) U)UO C 0.C C - C f C 2 5V 000 -a 0 254 Contracting Public Health Care Services in Latin America * 255 run as government departments. In Argentina, reforms have gone farther than elsewhere, but even there, the reforms described have been implemented in no more than 50 out of more than a thousand public hospitals. The halting implementation can partially be explained by the con- text in which the reforms were introduced and the process by which the reform was implemented. Context As noted above, the reforms initiated in the three countries were similar in nature. In some regards, the institutional structure of the health systems was similar also. However, in terms of the institu- tional and political context, there were large differences. In Argentina, the federal reform legislation was passed without major political opposition during President Menem's first administra- tion (1991-95) in the context of a sweeping reform of the state's role. Privatization of public enterprises and utilities and other market- oriented reforms introduced at that time led Argentina to strong and sustained economic growth over the 1990s. Thus, the general con- text was relatively favorable. However, the hospital reforms were more difficult to implement than other public sector reforms because public hospitals were owned by subnational (mainly provincial) gov- ernments. The degree of interest in and commitment to these reforms at the provincial level varied greatly.14 Even provinces com- mitted to the reforms found themselves held back by looming fiscal crises requiring increased budgetary scrutiny and fiscal discipline. Thus, both attention and funds were relatively unavailable. Provin- cial treasury officials appeared reluctant to loosen financial control, perhaps due to uncertainty regarding the controls included in the re- form design. In Chile and Uruguay, legislation for hospital autonomy reform was even more difficult to pass and implement. One reason for the difficulty may be that the reform was introduced by Center-Left coalition governments, which rely heavily on the support of labor and public employee unions. In addition, health reform was pro- 256 o Innovations in Health Service Delivery posed as an isolated sector reform, not as part of a sweeping and broad range of other public sector reforms, as was the case in Ar- gentina. Professionals, provider associations, and the political Left all fought reform on the following grounds: cost recovery would reduce access by the poor; revenue rights would lead to a two-tier system, as public hospitals would have incentives to give paying patients pref- erential service; hospital autonomy and competition would destroy health care networks and provide an incentive for excess supply and duplication, leading to a disjointed and costly system; public hospi- tals would unfairly compete with the private sector; and hospital au- tonomy was a first step to the privatization of public hospitals and would lead to massive loss of public health care jobs. In the end, both governments declined to do political battle with the medical profes- sion or with the private hospital associations. Implementation Process Of the three case study countries, Argentina did the most to manage the process of change, by reeducating key opinion makers and pro- fessional staff, facilitating change at pilot hospitals, providing ade- quate framework legislation, and-most important of all-offering some incentives to get the reform implemented. This may partly ex- plain the progress in Argentina relative to its neighbors. Organizational reform of public hospitals imposes significant changes in public hospitals' overall objectives and way of doing busi- ness. In those hospitals, more power is in the hands of front-line service providers (doctors) and patients than in the hands of admin- istrators or managers. Therefore, early on in the process of change, we would expect to need strong communication and education pro- grams to create a good environment for success. Once sufficient con- sensus was obtained, we would also expect some coercive measures to change the status and some persuasive incentives to ensure com- pliance in the medium term. Early in Argentina's reform process, the government promoted supportive education activities such as studies, study hours, work- shops, seminars, and training programs. These activities were spread Contracting Public Health Care Services in Latin America * 257 throughout the country, and included both the health policy network and the provincial hospital and health insurance networks. This was very important for establishing an enabling environment among key opinion makers and preparing professionals for the complex changes envisaged. The government also contracted intensive organizational develop- ment exercises for key pilot hospitals in the city of Buenos Aires and in the provinces of Buenos Aires and Mendoza, as well as lighter organizational development exercises in several other provinces. Under these contracts, pilot hospitals received a team of resident hospital-administration consultants who worked with hospital man- agement to define the hospital's mission and objectives, create its medium-range development plan, and coordinate the contribution of a series of short-term consultants who came for periods of two to three weeks. The short-term consultants helped the hospital set up key administrative services such as cost accounting, personnel man- agement, pharmacy and supply management, bed management, con- sultation scheduling, clinical management, and customer services. This was very important for assisting hospital management and key professionals in accepting the public hospitals' change in objectives and procedures. At the same time, the government established an umbrella legal framework through a national Self-Managed Public Hospital Law that included a provision for the automatic collection of unpaid insurance bills through the internal revenue service. This law was an essential incentive for provinces to introduce the organi- zation changes in their hospitals, since the federal government did not have authority over public hospitals. At the hospital level, most provinces did little reeducation but did pass the relevant provin- cial decrees and laws-providing some momentum to make things move forward. They also created persuasive incentives for long-term implementation-the capacity to retain revenues generated by bill- ing insurance carriers and charging copayments, and to distribute the revenue as performance bonuses. Chile also undertook many studies and educational programs. Most activities were at the ministry or managerial level, however, with limited penetration at the service delivery level, where the key 258 o Innovanons in Health Service Delivery decisionmakers are and where it might have made a greater impact. Also, there was no significant investment in the organizational de- velopment of the new regional health service administrations or spe- cific training of the respective managers. Finally, the government was not able to get the key legal acts passed in the legislature. In Uruguay, implementation was particularly problematic. The government did get the enabling legislation. However, passage was not accompanied by support or development at the level of hospital management. Although the government did promote studies and ed- ucational activities, most of them were concentrated within the cen- tral government apparatus. Neither was any initial investment made in hospitals' organizational development that might have facilitated the introduction of reform in the pilot hospitals, as in Argentina. Without enough public debate or consensus building among key stakeholders and without strong support even at the hospital level, the government passed the hospital autonomy decree. Two weeks later, the government had to withdraw it under a barrage of criticism from the provincial health care corporations, which feared that the reform meant privatization in disguise and that public hospitals would compete with their establishments. Condcusions Like many countries with integrated public hospital systems, Ar- gentina, Chile, and Uruguay undertook initiatives to reform their hospitals in the 1990s. Also following this worldwide trend, they did so by reducing direct government control and granting public hos- pitals and ambulatory care networks increased autonomy, while keeping them within the public sector. Thus far, the reforms have not been widely implemented. Most public hospitals are still run as government departments. Because of the weak implementation of organizational hospital re- form in the Southern Cone, it is impossible to draw conclusions on the impact of such reforms. Only a few public hospitals in each coun- Contracting Public Health Care Services in Latin America * 259 try have acquired critical decision rights related to labor. Market ex- posure, including financing systems in which the money "follows the patient," is experimental in Argentina and still a "shadow" budgeting system in Chile, five years after implementation began. Public hospitals in Argentina and Uruguay can now claim rev- enues obtained by recovering the costs of insured patient care. In Chile, hospitals formally have this right, but have little incentive to recover costs, because they would have to turn the extra revenue over to the treasury. Pro forma hospital boards and councils are now widespread. How- ever, they tend to function more in an advisory capacity.15 In addition, the board makeup and appointment process leaves them highly de- pendent on politics. Only in Argentina can we find a few examples of hospital managers who have significant autonomy. In all three coun- tries, management contracts are still experimental. Their content and emphasis on performance indicators is varied (table 6.3). In many instances, the contracts appear to be more form than substance.'6 In Argentina, in some instances, the distribution of staff bonuses is tied to continued compliance with performance indicators, but no explicit penalties have ever been imposed for noncompliance. The social functions of public hospitals, though increasingly rec- ognized, are rarely funded explicitly. However, there is increased interest in implementing funding reforms that would allow such pay- ments. Chile is the only country to propose a different budgeting system to finance "uninsurable goods," namely, typical public health interventions. Argentina is experimenting with independent financ- ing for training and medical education as "social functions of public hospitals." Public hospital autonomy and contracting have been implemented to different degrees and at different paces in the three Southern Cone countries. These differences stem from differences in societal and political contexts, the nature of the changes themselves, the management of the change process, and support and opposition from key stakeholders. 260 o Innovations in Health Service Dehvery NWes 1. Approximately 60 percent of the population is not insured. 2. Most of the public hospital system in Argentina is owned by provincial and municipal governments. 3. It became mandatory that health insurers pay public hospitals for care provided to their beneficiaries, and an automatic collection mechanism was established for obtaining payments from delinquent insurers. 4. Government of Argentina, Law 17.102/66. 5. In fact, none of the provinces has established output-based funding or purchasing. 6. See chapter 1 for a detailed definition of what constitutes au- tonomized and corporatized hospitals in the conceptual framework of this volume. 7. A. Sojo, "Reformas de gesti6n en la salud puiblica en Chile," CEPAL Review 3: 127-45 (August 1996). 8. Strictly speaking, neither of these hospitals was reformed under this program; they were both established earlier with a more au- tonomous governance structure. 9. See chapters 9, 10, and 12 of this volume. 10. Government of Chile, Ministerio de Salud, "Desarollo e im- pacto de la implementaci6n de compromisos de gesti6n en el sistema puiblico de salud de Chile," Santiago, November 1998. 1 1. Ibid. 12. Government of Uruguay, Ministerio de Salud Puiblica, Ad- ministraci6n de Servicios de Salud del Estado, "Compromiso de gesti6n, antecedentes, anexos, marco normativo," Montevideo, Oc- tober 1998; processed. Contracting Public Health Care Services in Latin America * 261 13. G. Walt and L. Gilson, "Reforming the Health Sector in De- veloping Countries: The Central Role of Policy Analysis," Health Policy 9: 353-70 (1994). 14. For more detailed information on implementation at the provincial level, see A. Harding and G. Prieto, "The Public Hospi- tal System in Buenos Aires Province: Current Situation," Health Systems Development Thematic Group Case Study, World Bank, Washington, D.C., 2001. 15. Ibid. 16. Ibid. Part 3 Case Studies CHAPTER 7 Betwixt and Between: Autonomization and Centralization of U.K. Hospitals Chris Ham Changes in the global economy, stemming from the oil shocks of the 1970s, and the rise to prominence of politicians of the New Right prompted increased questioning of the efficiency and responsiveness of public services. Responding to macroeconomic challenges, gov- ernments in the industrial world acted to control public spending. As part of this process, monetarist theories challenged the dominant or- thodoxy represented by Keynesianism, most obviously in the United States and the United Kingdom. The public was also increasingly re- luctant to pay taxes, and the ensuing "fiscal crisis" marked the start of a fundamental reappraisal of the relationship between the public and private spheres. Inspired in part by the writings of public choice theorists and other critics of the growth of big government, leaders such as Ronald Rea- gan and Margaret Thatcher sought to roll back the state and reinvig- orate the private sector. In the United Kingdom, this was manifested in policies to privatize nationalized industries and state-owned enter- prises. New Right thinking was also reflected in measures designed to strengthen the performance of public services encompassed in the emergence of the "new public management."' Through a combina- tion of market-oriented and managerial strategies, politicians sought 265 266 o Innovations in Healti Service Delivery to tackle problems in the funding and delivery of public services and in the performance of government itself. The failure of government agencies to deliver services that were efficient and responsive was acknowledged by politicians of the Center and Left as well as those on the Right.2 An important strand in the thinking behind the new public management came from management theorists and researchers analyzing trends in organi- zational structure and culture in both the public and private sectors. Ferlie and colleagues describe how later manifestations of the new public management accommodated this perspective.3 Of particular note in this context is the work of Osborne and Gaebler and Ran- son and Stewart, which articulates an approach of learning from the private sector while recognizing the distinctive features of public services.4 Underlying these developments is the argument that the market failures that generated government intervention in the economy and in social policy have given rise to government failures. Although the policy response was perceived to be outright privatization through the sale of state assets in areas where competitive markets existed (e.g., the steel, coal, and airline industries), analysts across the polit- ical spectrum acknowledged that this was not the only, or even al- ways the most appropriate, way of addressing government failures. Accordingly, in other cases alternative approaches were pursued, as in the privatization of utilities in the United Kingdom, where sup- pliers had a monopoly or near monopoly, coupled with new forms of regulation designed to avoid abuse of this position. Commenting on the U.K. experience with privatization, Foster and Plowden note that two-thirds of formerly state-owned industries have been transferred to the private sector. They add that this ap- proach is based on the principle that "whenever bodies are separated from government[,] either competition or regulation must be estab- lished to stimulate greater efficiency."5 Where continuing public ownership was considered necessary, as in the health sector, policies were pursued to make the public sector more businesslike and to adopt successful practices from private sec- tor organizations. An eclectic mixture of policy instruments resulted, Autonomization and Centralization of U.K. Hospitals * 267 some designed in advance and others that emerged in practice. 'Walsh, analyzing the approach to the reform of public services in the United Kingdom, notes in particular the use of competitive tender- ing and the introduction of internal markets.6 Impact on the U.K. Health Sector 'W;hat was the impact of these developments on the health sector in the United Kingdom? And how was the National Health Service (NES) affected, given its position as a centrally planned and admin- istered public service, in many ways the archetype of the large, bu- r eaucratic agencies targeted by critics of government failure? In addressing these questions, we should keep in mind that the NJ-IS became the principal source of health care for the population of the United Kingdom after its establishment in 1948. Funded mainly through taxation and delivered almost entirely through gov- ernment-owned agencies, the NHS provided both universal and Comprehensive services that, with limited exceptions, were free at the point of use. The creation of the NHS contributed to the achieve- mnent of population health outcomes comparable with those of other developed countries at a relatively low rate of expenditure (around 6 percent of gross domestic product in the mid- 1 990s). A small private health care sector continued to play a part alongside the NHS, en- abling some patients to avoid long waits for NHS treatment. Successive governments introduced modifications in the organi- zation of the NHS in the first three decades of its existence. How- ever, not until the election of the Thatcher Government in 1979 'were more fundamental options for reform considered. The main emphasis initially was on policies designed to increase efficiency (box 7.1). Among these policies, the most significant was the Griffiths in- quiry into NHS management, which reported in 1983.7 The report was highly critical of management arrangements in the NHS and recommended that the system of consensus management through multidisciplinary teams should be replaced by a clearly defined gen- eral management function. 268 o Innovations in Health Service Delivery Box 7.A [f;ciency WiOV3S in h 19800 Five initiatives during the 1980s illustrated the Thatcher Government's approach to increasing efficiency. First, a re- quirement was placed on health authorities to generate an- nual efficiency savings. Second, a series of scrutinies was conducted into areas such as transport services, recruitment advertising, and the use of staff accommodation. Third, a set of performance indicators was developed to enable health authorities to compare their performance with achieve- ments elsewhere. Fourth, competitive tendering was intro- duced for catering, cleaning, and laundry services. And fifth, an income-generation drive was launched. This included generating income from private patients treated in NHS hospitals and allowing health authorities to set their own charges under the 1988 Health and Medicines Act. The Thatcher Government welcomed the Griffiths report and ac- cepted all of the main recommendations. General managers were appointed, beginning in 1984, although only a small proportion came from outside the NHS. Consistent with the approach proposed by Griffiths, managers were appointed on short-term contracts and could earn performance-related pay after a formal appraisal process. Through the resource-management initiative, steps were taken to involve hospital doctors in management and to devolve budgets to clinical directorates within hospitals. The Health Services Super- visory Board and the NHS Management Board were established, and soon afterward the NHS Management Executive was created as the "head office" for the NHS within the Department of Health. The Griffiths report served as a stimulus to improve performance and created the conditions in which the changes in the 1989 white paper, Workingfor Patients, could be introduced.8 The policies pur- sued in this period were a dress rehearsal for the radical plans set out Autonomizauon and Centralization of U.K. Hospitals * 269 in the white paper, plans whose implementation was to have rever- berations not only in the United Kingdom but also farther afield. NHS Trusts Reform Design Workingfor Patients was the outcome of a ministerial review set up by Margaret Thatcher in 1988. The white paper, published a year after the review began, was an attempt to address problems that had arisen in the NHS at the end of the 1980s. Foremost among these was a perception that the NHS was underfunded, despite the impact of the Thatcher Government's various efficiency initiatives. The challenge for the government was to apply its radical reforming instincts in an area of social policy where the influence of staff groups was strong and where the public placed a high value on the service provided. After considering and rejecting options for changing the way the NHS was funded, the review focused mainly on the delivery of health services. Drawing on ideas outlined by the American econo- must, Alain Enthoven, ministers proposed to introduce competition between hospitals and other providers. Competition was to be achieved by autonomizing hospitals and introducing marketlike in- centives through changes to the funding arrangements within the NHS. The core proposals in the government's plans were to: * Convert hospitals to self-governing NHS trusts, to manage ser- vice provision * Transform health authorities into buyers of services * Also have "fundholding" groups of general practitioners who pur- chase hospital services for their patients * Use contracts to provide the link between purchasers and providers. "Money should follow patients" was a key idea contained within the reforms. It was intended to overcome the perverse incentive of the previous budgeting arrangements in which hospitals were in ef- 270 o Innovations in Health Service Delivery fect penalized if they increased productivity because their income was fixed at the beginning of each year, but their expenditure rose in line with activity. The reforms were also intended to increase effi- ciency and responsiveness by forcing providers to compete for in- come from purchasers. The establishment of self-governing NTS trusts was the main in- stitutional change on the provider side of the NHS. On the pur- chaser side, the new arrangements centered on health authorities and general practitioner (GP) fundholders. Initially, health authorities in England were responsible for buying health services for populations of around 300,000 on average, although subsequent mergers be- tween authorities increased the size of the population served.9 GP fundholders were allocated a budget to purchase a limited range of services for their patients, and these budgets were deducted from the allocations of the relevant health authorities. Fundholders continued to work alongside other general practitioners who preferred not to take responsibility for a budget and who instead advised health au- thorities on where services should be purchased. A start was made in implementing these changes in 1991, and the number of NHS trusts and fundholders increased each year. By 1996, the organizational transformation of the NHS, initiated by Working for Patients, was effectively completed in England with the establishment of almost 450 trusts and 100 health authorities. By that date, fundholding covered about half the population. Alternative models of findholding evolved in parallel, including networks of practices known as "multifunds" and a number of pilot "total pur- chasing projects" (TPPs), in which general practitioners purchased all services. In addition, commissioning groups of general practition- ers, often involving fundholders and nonfundholders, were estab- lished in many places to advise health authorities on purchasing services. Thus, purchasing arrangements grew in variety and com- plexity. The resulting structure of the NHS is illustrated in figure 7.1. Among the many innovations to result from the reforms, the es- tablishment of NHTS trusts to manage hospitals and other health care providers was particularly important. A wide variety of services came Autonomization and Centralization of U.K. Hospitals * 271 Figure 7.1 The Structure of the NHS in England, 1996 l NHS l Executive LocalRein authorities Refiona Health /NS trusts authorities /450 GPFHs (15,000 GPs) Commissioning / roups (7,000 GPs) 80 100 (150PPs) Multifunds I1,500 GP ) (2,600 GPs) Accountability Contract together to form trusts. Some were responsible only for large acute hospitals, in line with the expectations set out in Workingfor Patients and other early guidance. Others managed community and mental health services as the guidance was gradually relaxed to allow other providers to seek trust status. Yet others formed "whole district" trusts, bringing together the full range of acute and community ser- vices within a single organization. Responsibility for running ambu- lance services was also vested in trusts. 272 o Innovations in Health Servce Delivery The decision to set up NHS trusts was based on two considera- tions. First, the architects of NHS reforms perceived that devolving responsibility for management to hospitals and other health care providers would be beneficial. Under the pre-1991 arrangements, health authorities carried this responsibility, and it was argued that this served to delay decisionmaking and to militate against local ownership of and identification with service provision by requiring decisions to be referred "up the line." In this respect, the establish- ment of NHS trusts harked back to an earlier period in the history of the NHS (1948-74), when teaching hospitals had their own boards of governors. Those involved in writing the white paper drew on this model in framing their proposals. Second, and more radically, removing hospitals and other providers from the management control of health authorities was seen as essential to enable health authorities to assume their new purchasing role within the NHS market. At one level, this was because health authorities would no longer carry ultimate respon- sibility for the direct management of services and could therefore concentrate on assessing the population's health needs and commnis- sioning services appropriate to these needs. At another, it was a pre- condition for health authorities to be able to place contracts with providers of their choice rather than simply those under their own management. To this extent, creating a genuine separation of pur- chaser and provider roles was an important step on the road to se- lective contracting within the NHIS. In explaining the plans set out in the white paper, government spokespersons emphasized that trusts remained part of the NHS. They were established as separate legal entities (though still public bodies) under the NHS and Community Care Act of 1990, and their powers and duties were contained within that statute (for ease of ref- erence, the relevant parts of the act are reproduced in Appendix 7A). Each trust was placed under an obligation to balance its budget, tak- ing one financial year with another. The general powers of trusts included: (a) to acquire and dispose of land or other property; Autonomization and Centralization of U.K. Hospitals * 273 (b) to enter into such contracts as seem to the trust to be ap- propriate; (c) to accept gifts of money, land or other property, including money, land or other property to be held on trust, either for the general or any specific purposes of the NIHS Trust or for all or any purposes relating to the health service; and (d) to employ staffon such terms as the trust thinks fit.'0 The position of trusts as public bodies within the NHS is rein- forced by the provisions of the 1990 Act, which gives the secretary of state powers to supervise trusts and to require them to comply with guidance or directions. They also remain subject to legislation af- fecting NHS facilities, including the Hospital Complaints Procedure Act of 1985, the Data Protection Act of 1984, and the Access to Health Records Act of 1990, as well as European Union law. Taken together, this meant that trusts operated within a highly regulated environment. The rationale behind trusts was set out in Workingfor Patients and also in the working paper, published shortly afterward, which stated: The Government is committed to devolving decision-making in the National Health Service to local operational level in order to make hospitals more responsive to the needs of their patients, to secure local commitment and to achieve greater value for money. The next logical step in the process of extending local responsibility is to enable NI-IS hospitals to achieve self-gov- erning status. Self-governing hospitals will remain firmly within the NIS and there will be safeguards to ensure that essential local services continue to be provided locally. But they will have far more freedom to take their own decisions on the matters which affect them most without detailed supervision from above. This new development wIll give patients more choice, produce a better quality service and encourage other hospitals to do even better in order to compete.1I As this working paper and subsequent guidance made clear, the reform model designed for the trusts was best understood in com- parison with that of hospitals and units directly managed by health authorities, particularly in their greater level of operational inde- 274 o Innovations in Health Service Delivery pendence. Table 7.1, drawn from guidance issued in 1990, summa- rizes the main points of similarity and difference. The Political Context of Implementation Before discussing the impact of the reforms and the evolution of NHS trusts, a little more detail may be helpful on the context in which these changes were implemented. Working for Patients re- ceived a hostile reception from the main professional groups in the NHS who expressed concern at the commercialization of health care implied by the government's plans and who also feared a hidden agenda that would lead to the privatization of service provision. Min- isters consistently denied that this was their intention, arguing that their aim was to strengthen and modernize the NHS, not undermine it. They stressed their commitment to maintaining a health service where treatment was based on need, not ability to pay. Despite their reassurances, many NiHS employees distrusted the government's motives, which meant that the changes were taken forward with lit- tle enthusiasm in many quarters. In implementing Working for Patients, ministers started from a broad framework, and much of the policy and organizational detail of their plans was missing at the outset. This reflected the speed with which the white paper was produced and the concern about meeting the prime minister's one-year deadline for publication. Further guid- ance was forthcoming in a series of working papers, published soon after the white paper was issued, but to a considerable degree, mak- ing the government's plans work on the ground was left to NHS staff. This approach has variously been described as ministers "mak- ing it up as they went along" and as an "emergent strategy, "12 but the effect was the same. Key elements of policy were worked out only in the course of implementation, and by no means was implementation informed by a coherent view of how the reforms were intended to work in practice or where they would eventually lead. Moreover, the language shifted away from explicit discussion of markets to a con- cern that purchasers and providers should seek to build long-term partnership relationships. Autononmzation and Centralization of U.K. Hospitals * 275 Table 7.1 Comparison of the Regimes of NHS Trusts and Directly Managed Units IISSUE NHS TRUST DIRECTLY MANAGED UNIT Management Each trust is run by its own DHA is responsible for unit board of directors The trust is UGM is responsible for day-to- free to determine its own man- day management Internal agement structure Senior pro- management arrongements are fessional staff members must subiect to DHA approval be involved in management Accountability Each trust board is directly Each unit is accountable to its accountable to secretary of managing DHA DHA is state via NHSME accountable to RHA, and RHA is accountable to secretary of state via NHSME Funding Each trust's income derives Each unit's income derives largely from contacts with largely from contracts with health authorities, GP fund- health authorities, GP fund- holders, and private sector holders, and private sector Services Each trust is free to determine Range and extent of services range and extent of its services, offered by a unit are deter- except that where a service mined by managing DHA must be provided locally, a trust can be obliged to provide it if it is the only unit able to do so Each trust provides services it Each unit provides services it is is contracted to provide contracted to provide Trust is not required to consult Closures and changes of use Community Health Council on are subject to formal consulta- closures or changes of use tion with Community Health Council Employment of staff Each trust sets its own staffing Each unit determines its own structure and levels It employs staffing structure, but DHA or a/l its own staff, including RHA employ its staff consultants Each trust is free to determine Pay and other terms and con- pay and other terms and con- ditions of employment of nearly ditions of employment of all all staff members are sublect staff Staff members transfer- to review body or National ring to trust retain their existing Whitley Council agreements or terms and conditions of service departmental determination until changes are negotiated (Table continues on the following page) 276 o Innovations in Health Service Delivery Taoie 7.1 (continued) ISSUE NHS TRUST DIRECTLY MANAGED UNIT Each Trust was placed under Managing DHA has statutory Financial duties an obligation to balance its duty to balance its budget budget, taking one financial each year UGM has manage- year with another It is also rial imperative to ensure unit required to achieve a 6 per- breaks even. The unit has to cent return on assets and keep pay capital charges equivalent within agreed EFL to depreciation and 6 percent interest on fixed assets Each Trust prices to cover run- Each unit prices to cover run- Prices ning costs, depreciation, and ning costs and capital charges return on assets (depreciation and interest) Trusts may retain surpluses. Units cannot retain surpluses Surpluses Each Trust is free to borrow Units have no power to Borrowing within agreed EFL. borrow. Trusts do not insure for clinical Units do not insure for clinical Insurance negligence They may insure negligence or other insurable for other insurable risks. risks, with certain limited exceptions Each Trust owns assets It is Each unit's assets are owned Ownership of assets generally free to acquire and by the secretary of state or dispose of assets and retain health authority. Acquisition proceeds from any sales. and disposal of assets are sub- lect to control and regulation by Department of Health, RHA, and managing DHA. Retention of sales proceeds is subject to decisions by RHA. Each Trust makes case for cap- Each unit makes case for capi- Capital ital development to NHSME tal developments to DHA or and funds agreed program RHA. Funding depends on allo- from own resources or by bor- cations from regional capital rowing within agreed EFL program Note: DHA district health authority; EFl external financing limit, NHSME NHS Management Executive, RHA regional health authority, UGM unit general monager. Source Adopted from Department of Health, NHS Trusts A Working Guide (London: Her Majesty's Stationery Office, 1990), Annex A, pp 29-30 Autonomization and Centralization of U.K. Hospitals * 277 Table 7.1 indicates that the key decision rights transferred to NHS trusts fell into three main areas: financial, managerial, and personnel. The government emphasized that its plans would enable trusts to borrow money more easily, make decisions without recourse to higher tiers of management, and employ staff on terms and condi- tions determined locally, not centrally. These freedoms were designed to bring benefits for staff and patients and to foster local owner- ship and pride in the provision of health services. Put another way, the aim was to create a stronger set of incentives to improve performance than had existed hitherto. And, although the decision to seek NHS trust status was a matter for the staff involved to resolve, the gov- ernment made clear its expectations that this would become the pre- ferred model. At an early stage in the implementation of the reforms, it became clear that some of the freedom available to NHS trusts was more hy- pothetical than real. This applied most obviously to the financial regime under which trusts operated. The details of this regime were complex and were only fully understood by those closely involved in their operation. For the purpose of this chapter, the essential points are summarized in box 7.2. The main financial duties of trusts when they were established were threefold: * To balance its budget, taking one financial year with another * To earn a 6 percent return on their assets * To keep within the external financing limit (EFL) set each year. The last of these duties was particularly important. The concept of external financing limits had been applied by the government in other areas (such as the public utilities) and it was therefore a well- established element of public sector reform. Each trust agreed on an EFL with the regional office of the NHS Management Executive an- nually, and this acted as a form of expenditure control. Regional of- fices worked within expenditure totals agreed between the Treasury and the Department of Health and subsequently distributed between the eight English health regions. 278 o Innovations in Health Service Delivery Mont 7.2 Th1 Fi:nanil R e gf lHt0 Tnusb Trusts were established with the Treasury as the "share- holder." Each trust owned its assets and the value of these assets was matched by an originating capital debt owed by the trust to the Treasury. The originating capital debt was made up of interest-bearing debt and public dividend capi- tal. The payments made by trusts on their debt were recy- cled to health authorities to ensure that purchasers had ad.. equate resources to pay the prices charged by trusts, including the 6 percent return on assets required under the financial regime. The establishment of trusts in this way opened up the opportunity of more radical change, includ- ing the privatization of services, at a later date. One of the aims of these arrangements was to move away from the previous system in which NHS capital was a "free good" by introducing greater awareness of the costs of cap- ital expenditure. It was also intended to provide an incentive to trusts to sell land and property that was surplus to re- quirements as this would reduce their interest payments as well as generating resources for development. Although in some cases entrepreneurial managers seized the flexibilities available to them in precisely the manner that had been an- ticipated by the architects of the reforms, more often there was reluctance to dispose of assets because of anticipated opposition from staff and the public. Together with the bu- reaucracy associated with these arrangements, this under- mined the system of charging for capital. Under these arrangements, a trust could be given an EFL that was positive, negative, or zero. These were defined as follows: A positive EFL is set where the NHSME has agreed capital spending for a Trust which exceeds internally generated re- Autonomizaton and Centrahzation of U.K. Hospitals * 279 sources, resulting in the Trust needing to borrow or reduce in- vestments in order to finance its spendmg programme. An EFL of zero is set where the agreed spending programme equals internally generated resources. A negative EFL is set where the agreed spending programme is less than internally generated resources. In these circum- stances a Trust is not able to use all its retained surplus or de- preciation on capital expenditure. Some or all of these resources must be used to repay loans or be invested. The investments made are normally held by the Trust and may be available to fi- nance future planned spending."3 What this meant was that trusts with a negative EFL were only al- lowed to use internally generated resources up to the limit agreed on with the regional office. Although these resources were retained by the trust, the incentive to sell surplus assets to fund new develop- ments was weakened. Trusts were further constrained by guidance that required them to borrow on the best terms available. Effectively, this ruled out raising capital on the private markets as the rates available from the Treasury were lower than those available from private sources. Only with the introduction of the private finance initiative (see below) did the aver- sion to the use of private capital change. And with the availability of funds from the Treasury for capital developments subject to the out- come of the annual public expenditure negotiations and the macro- economic objectives of the government, bureaucratic bargaining and bidding on the basis of business cases exercised a strong influence on the allocation of capital resources. This was reinforced by the De- partment of Health and Treasury retaining control over capital proj- ects of any scale through the need to secure their approval of these projects before they could proceed. Yet another constraint within the financial regime was the re- quirement that the prices charged by trusts be based on their costs in order to enable them to earn a 6 percent return on their assets. The only exception was in relation to contracts with the private sector where trusts were advised to price their services at the level the mar- ket would bear. Department of Health guidance also stated that 280 o Innovations in Health Service Delivery trusts should not plan to cross-subsidize their services, nor use mar- ginal costing unless spare capacity arose during the year. In this re- spect, they were expected to follow the same pricing regime as di- rectly managed units; trust status therefore offered few advantages in comparison with these units. Although empirical evidence indicated that trusts did not always follow this guidance,14 central regulation of pricing was symptomatic of the way in which the market was regulated. There was much greater potential freedom in relation to person- nel where trusts were no longer constrained by national pay rates and terms of service but were able to negotiate locally with their staff. This was linked to the movement of consultant contracts (i.e., med- ical specialists) to trusts. A number of trusts used this freedom to in- troduce innovative forms of employment but in general they were slow to depart from existing arrangements. One of the reasons was that NHS staff retained their employment rights and contracts when they transferred to trusts. Changes could be made as a result of ne- gotiation between trusts and staff but could not be imposed. Most trusts chose to make changes at the margins, for example, in the ap- pointment of new staff, and avoided rapid or radical departures from past practice for existing staff. This was due less to the resistance of staff and the veto power of trade unions (which had been seriously weakened by the Thatcher Government's reform of labor laws) than to the size of the NHS work force and the scale of the work involved in renegotiating employment contracts across the board. It also re- flected the culture of the "NIS family" and the reluctance of man- agement to challenge established practices. What this meant was that the management freedom trusts had was more significant than changes in the financial regime or personnel policy. Under the NHS and Community Care Act of 1990, each trust was run by a board of directors. This comprised a chairman ap- pointed by the secretary of state for health, up to five other nonex- ecutive directors (two appointed by the regional health authority and the remainder by the secretary of state), and an equal number of ex- ecutive directors, including usually a chief executive, a finance direc- tor, a medical director, and a nursing director. The trust board was Autonomization and Centralization of U.K. Hospitals * 281 responsible for determining the overall policies of the trust, moni- toring the execution of these policies, and maintaining the trust's fi- nancial viability. Freed from the direct management control of health authorities, trusts ran their own affairs within the framework of the contracts they negotiated with purchasers and the accounta- bility arrangements described earlier. As research has shown, those involved in trusts used their management freedom to make a num- ber of changes in their organizations.15 The remuneration of chairmen and nonexecutive directors was determined nationally. Each trust was responsible for deciding the terms of the contract offered to chief executives, and salaries varied depending on the size and complexity of the organization. Con- tinuing the trend initiated by the Griffiths general management re- forms, chief executives were usually appointed on short-term (but renewable) contracts and were eligible for performance-related pay. Senior managers were expected to deliver on the objectives set for them or face termination of their contracts. Market Exposure As time went on, the internal market in the NHS became in reality a managed market in which politicians were reluctant to allow com- petitive forces free rein. In the first year of implementation, this re- luctance was reflected in guidance that emphasized the need to achieve a "smooth takeoff' by ensuring a "steady state" in the service market-hence at least delaying the development of any market ex- posure or competitive forces. The fact that 1991-92 was an election year heightened the sensitivities of politicians and helped to explain their reluctance to allow the market to operate in an unregulated fashion. Purchasers were permitted greater freedom to make changes in subsequent years but still had to give advance notice of plans to move services and contracts to avoid harmful disruption. This meant that although trusts were required to earn their income, their market exposure was limited. Outside London, the regulatory framework for the market devel- oped piecemeal. Only in 1994 was national guidance issued.16 By 282 a Innovatons in Health Service Delivery that stage, many of the behavior patterns that were to shape the re- lationship between purchasers and providers had become well estab- lished.17 Indeed, one purpose of the guidance was to draw together some of the lessons learned in the first phase of the reforms and to provide a framework for their further development. In so doing, the guidance started from the position that government regulation should promote competition. It then noted that the presumption should be against intervention in the market unless necessaiy to counter anticompetitive behavior such as collusion between trusts. Four areas were examined in detail in the guidance: provider mergers, purchaser mergers, managing change where providers are in difficulty, and collusion. Rules were set out for handling these situations. The guidance cautioned against provider mergers that would result in the emergence of local monopolies and stressed that purchaser mergers should proceed only where the views of patients and general practitioners could be heard effectively. On the issue of providers in difficulty, the guidance stressed that change should be planned and carried out with minimal disruption. In this case, health authorities and trusts were reminded of the need for public consul- tation and support of regional offices. In this area at least, the guid- ance underlined the extent to which the market remained regulated. The excess capacity in London made it the most likely place for competition to emerge and generate performance pressures. Per- versely, this is where central control was most evident, following a government-sponsored review under Professor Sir Bernard Tomlin- son to advise on how the effects of the market in London could be handled to properly plan change. Subsequently, purchasers were re- quired to place contracts in a way that was consistent with govern- ment policy. An example that arose in 1993 concerned the Camden and Islington Health Authority, instructed not to move its contracts to trusts offering lower prices. This health authority had to leave the contracts with University College Hospital so that the rationaliza- tion of services taking place in that part of London could proceed ac- cording to plan. In contrast to the market for hospital services, market exposure did emerge in the market for managerial labor. Chief executives hired Autonomization and Centralization of U.K. Hospitals * 283 after the reforms could have their contracts terminated-an outcome that occurred with reasonable frequency. In some cases, chief execu- tives (and occasionally chairmen) were forced to resign after losing the confidence of medical staff. Though a relatively rare occurrence, the possibility of forced resignations served to encourage the rest and brought market pressures to bear in the managerial labor market. The Perforrmance of Trusts Recognizing these tensions, what does research into the reforms tell us about the performance of trusts? Hamblin has summarized the re- sults of the studies that have been conducted.18 As he and other re- searchers emphasized, drawing conclusions from the work that has been done is difficult because study designs were not always rigor- ous, the decision to seek trust status was voluntary, and differences in performance may therefore be attributable to the decision of certain types of providers to go down this route in the early stages of imple- mentation. Nor is it easy to disentangle the effects of trust status from other changes occurring at the same time. Accepting these cautions, Bartlett and Le Grand drew on an analysis by the Newchurch consultants to show that most first wave trusts either met or improved on their initial financial performance targets.19 This level of performance was not sustained. For example, the annual audit of NHS accounts, published by the Comptroller and Auditor General, drew attention to the financial difficulties trusts experienced in 1995-96 and expressed concern at the number of trusts that had failed in their financial duties. A similar picture emerged in 1996-97 when 148 out of 429 trusts failed in one or more of their financial duties after adjusting for technical factors. Figure 7.2 identifies the source of these failures. Bartlett and Le Grand's own research into the costs of hospitals involved in applications for trust status and other hospitals found lower unit costs in trusts than nontrusts, especially in the first wave. This led them to conclude: [TMhe first wave trust applicants were a self-selecting group of hospitals which had lower average unit costs, especially in the 284 o Innovations in Health Service Delivery Nuigre 7.2 Source of Trusts' Financial Failures External financing limit t 0 (~~~~~7 Failures) Break even Target rate of return (133 failures) Source. Comptroller and Auditor General, NHS (England) Summarised Accounts 1996-97 (London: Stationery Office, 1998) nonclinical departments.... Any research showing improved performance of the first wave trust applicants relative to DMUs [directly managed units] . . . must therefore be treated cautiously, as these are likely to have been those hospitals which were al- ready perfonming better than others under the old system.20 They also noted that higher cost providers were likely to be at- tracted to trust status over time, which would make sustaining im- proved performance in trusts difficult. They attributed this poor prognosis to the fact that differences in performance between second wave trusts and nontrusts were related to factors such as size and case mix instead of intrinsically better managerial performance. More recent research by Soderlund and colleagues has challenged the view that first wave trusts had lower costs prior to the reforms.21 After adjusting for case mix, they found that the early trusts started out less productive than directly managed units. Costs decreased signifi- cantly with the change from directly managed to trust status. Over the Autonomization and Centralization of U.K. Hospitals * 285 first three years of the reforms, real productivity gains were achieved and, in the case of second and third wave trusts, the largest gains were in the first year of gaining trust status. By the end of the period stud- ied, there were no significant differences in productivity between trust waves or between trusts and directly managed hospitals. Soderlund and colleagues found that competition between hospi- tals had no significant effect on productivity. This conclusion con- flicted with other research that found: [C]osts were higher in more competitive markets at the start of the market, but decreased rapidly after the start of the reforms in competitive areas. Hospitals with few competitors, on the other hand, had lower costs on average in 1991/92, but made lit- tle by way of cost savings during the study period.22 Of particular interest is Csaba and colleagues' finding that trusts were more responsive than directly managed units to competition.23 In a separate study, Propper found some evidence that the degree of competitiveness was related to the prices charged by trusts, lending support to the argument that the market, where it existed, was hav- ing the effect anticipated.24 Other evidence on the comparative performance of trusts comes from Smee's analysis of NHS activity data.25 In relation to the num- ber of patients treated, Smee showed that trusts outperformed directly managed hospitals in the first two years. First wave trusts performed particularly well in undertaking a high proportion of day-case surgery but this did not apply to the second wave. First wave trusts also performed better than directly managed hospitals in achieving the government's waiting time targets, although early data from second wave trusts showed that they performed less well on this indicator than directly managed hospitals. A similar picture emerged from comparisons of the total number of patients waiting for treatment. In relation to other areas of performance-for example, the qual- ity of care provided by trusts-the available research evidence is scanty. As Hamblin emphasized in his review, analysts therefore have to rely on anecdote and indirect research.26 This shows that it is hard 286 o Innovations in Health Service Delivery to prove that trust status was direcdy responsible for improvements in the quality of services. There is no evidence that trusts increased patient choice or that trusts became more accountable to their local populations. Indeed, examples of quality failures in this period (e.g., in relation to cancer-screening programs) could be seen as evidence that trust status and competition had adverse effects on patient care in some areas. Against this, the management freedoms available to trusts released a good deal of energy in many places, helping to bring about improvements in the environment of care and in organization of services. The Dynamics of Trusts In explaining the performance of trusts, two factors should be em- phasized relating to the reasons for establishing trusts. First, the ten- dency of ministers and civil servants to intervene in a range of areas eroded the operational freedom envisaged for trusts at the outset. Smee has observed: [M]inisters and the center are finding it difficult to reconcile de- volved accountability with the demand for detailed monitoring created by parliamentary and media interest in operational is- sues. In consequence, the center is drawn into a whole range of issues, from hospital catering standards to the freedom of speech of hospital staff that it once expected to leave to the dis- cretion of local management. The dilemma is that without sub- stantial operating freedom, Trust management cannot be ex- pected to produce a better performance than the old directly managed units, but that with such freedom there is bound to be a diversity of behaviors and performance. The existence of out- liers is then seen-by the press, auditors and politicians-as a cause for central regulation.27 Put more colloquially, ministers risked being damned if they did intervene and damned if they didn't. The result was an uncomfort- able compromise between genuine devolution and centralized direc- tion-a practical illustration of the tensions involved in the oxy- moron otherwise known as the "managed market." Autonomization and Centralization of U K. Hospitals * 287 Second, the extent to which the conditions for competition existed varied across the NHS. While there was evidence that many acute care hospitals were in a competitive environment, some trusts en- joyed a near or total monopoly.28 Moreover, even where a market existed, purchasers were often reluctant to use their leverage to improve performance. This was illustrated by health authorities' ex- tensive use of block contracts, which offered little advantage over the global hospital budgets they replaced and meant that, in practice, se- lective contracting rarely took place. GP fundholders were more in- clined to use cost-and-volume and cost-per-case contracts and to "shop around" and get the best deal for their patients, but they, too, were often loyal to their local NHS trusts; in many cases, they were not inclined to move to alternative providers. Adding the require- ment that trusts fulfill their social functions by treating emergencies, whatever their source, further attenuated market incentives. The lack of a tradition of purchasing in the NHS and the com- plexities involved in being an active, intelligent commissioner rather than a passive payer, meant that the countervailing force needed to stimulate changes in provider behavior was not always present at the outset. This deficit was compounded by the emergence of a "com- mercial-in-confidence" culture in which trusts were reluctant to dis- close information about their costs and quality unless required to do so under the terms of their contracts with purchasers. Because the Department of Health did not support purchasers by requiring in- formation disclosure, the resulting asymmetries meant that pur- chasers were often negotiating from a position of weakness, not strength. To the extent that providers' performance only partially lived up to the reform architects' expectations, one important con- tributing factor was that the envisioned accountability regime never became fully functional, especially with regard to NHS purchasing and weaknesses in the regulatory framework. Rarely was a trust dissolved. The most notable exception, the winding up of the Anglian Harbours Trust in 1997, was widely (though erroneously) reported as the first trust to go "bankrupt" (er- roneous because trusts as public bodies cannot go bankrupt). The 288 o Innovations in Health Service Delivery event received so much attention precisely because it was an unusual case (box 7.3). In other circumstances, market tensions and market pressures were handled through established administrative mechanisms in- stead of allowing trusts to go out of business. Regional offices played their part in these mechanisms, for example, by brokering cash sur- pluses and deficits between different areas in circumstances where funds were sufficient to overcome short-term problems and by work- ing with trusts to tackle financial difficulties. Financial difficulties were identified as part of the quarterly reporting arrangement be- BIox 7.3 The DissoluNon of Truss The Anglian Harbours Trust, established in 1991, provided a range of community health services, including mental health and learning disability services and community hos- pitals. It was a small trust with an income of £28 million (US$45 million) and a staff of just under 1,000. In its early years, it was relatively successful. Later, it had difficulties ,with its mental health services, and the two health authori- ties buying these services decided to take their contract to an adjacent specialist mental health provider. This raised a question about the viability of the remaining services run by the trust. Again, the health authorities decided to remove these services from Anglian Harbours and to invite bids from other trusts. The Anglian Harbours' staff was trans- ferred to the new providers, with the exception of 63 mem- bers who were made redundant. The trust ceased to exist on September 30, 1997. Source: A Barrick and P. Balcombe, Anglian Harbours-The First NHS Trust Dissolution (n.p.: NHS Executive, 1998). Autonomization and Centralization of U.K. Hospitals * 289 tween trusts and regional offices, and underlying recurrent problems were dealt with by a recovery plan agreed between the trust and the regional office. Progress against the recovery plan was monitored through monthly reporting, regular meetings with trust senior man- agers, and the trust's achieving specific milestones. Regional offices were also involved in arbitrating contract disputes. Because contracts were not legal documents, disagreements between purchasers and providers were resolved informally instead of through the courts.29 Trust boards themselves attached great importance to maintaining their organizations' viability and expanding and developing their ser- vices. Yet given the slow development of the NIS purchasing func- tion and the constraints on public expenditure for health services, many trusts looked to the private sector to enhance their income. Trusts were in fact required to seek income, not just to manage ex- penditure (the main requirement on hospital managers before the in- troduction of the reforms). As a result, trusts running acute care hos- pital services worked to supplement the income they received from NHS purchasers by attracting additional resources from private pa- tients, using the freedoms created by the 1988 Health and Medicines Act (box 7.1). This often involved creating or upgrading dedicated NHS pay-bed units within trusts and putting them into competi- tion with private hospitals. Ironically, NHS trusts became major providers of services to private patients, thereby posing a threat to private hospitals, which had expected to become beneficiaries of NHS reforms by attracting patients and income from NHS purchasers. The blurring of the distinction between the public and private sectors was accentuated by the application of the private finance ini- tiative (PFI) within the NTHS. PFI was launched in 1992 as a way of attracting private finance to pay for public sector capital projects across government as a whole. Initially, PFI was of marginal impor- tance in the NHS, confined mainly to small projects such as car parks and incinerators. However, under a change of rules published in 1995, NHS trusts wishing to undertake major capital programs were required to seek private financing. The rules enabled private in- volvement to extend not only to new building costs but also to the 290 o Innovations in Health Service Delivery provision of associated services such as catering, cleaning, and some clinical services. PFI entailed the coming together of consortia com- prising financial institutions, construction companies, and facilities managers to bid for NHS projects. Whereas originally PFI was con- sidered a supplement to treasury funding, over time it became a sub- stitute as the NHS capital program was cut back. After the 1997 gen- eral election, the Labour Government made clear its continuing support for PFI and moved quickly to ensure that projects at an ad- vanced stage of planning were taken forward. Emerging Lessons Researchers and independent analysts have offered a variety of judg- ments on the overall impact of the marketizing organizational re- forms that took place in the United Kingdom. The most compre- hensive early assessment detected relatively few changes in the first stages of implementation and argued that more time was needed to reach an informed judgment.30 A more positive assessment was made by the Organisation for Economic Co-operation and Development (OECD), which found much to commend in the changes that had been introduced, pointing to encouraging early results from the per- formance of NHS trusts as an example.3' However, Bloor and May- nard challenged these conclusions, pointing to the inadequacies of the evidence on which they were based.32 In a separate review, May- nard and Bloor argued that the success of the reforms had been mixed, a view echoed by Klein in his assessment.33 A systematic re- view of the research evidence on different aspects of the reforms con- cluded that, overall, little change (either positive or negative) can be detected.34 A common point in many of these analyses is the diffi- culty of separating the effects of the reforms from the effects of in- creases in NHS funding and other changes in government policy to- ward the NHS occurring at the same time. My own reading of the evidence and reflections on the experience of working with a wide range of NHS agencies throughout this pe- riod indicates that, while measurable change may have been limited, Autonomization and Centralization of U.K. Hospitals * 291 the balance of power within the NHS has shifted over time.35 Most notably, the influence of trusts has been increasingly challenged, which has led to increased attention to public health and primary care. It has also been associated with emerging interest in evidence- based medicine and health-technology assessment. This has resulted in a questioning of traditional patterns of resource allocation and pri- ority setting and of the old system of "planning by decibels," in which acute care providers won the biggest share of resources. These findings are supported by the research of Ferlie and colleagues, which also noted changes in the balance of power within the NHS and changes in culture, particularly in relation to the increasing in- fluence of GPs vis-a-vis hospital doctors, the rise of managerialism, and the beginnings of a shift from management by hierarchy to man- agement by contract.36 Of particular importance in this process was the increasing influ- ence of both health authorities and fundholders as purchasers. As noted, the slow and uneven development of purchasing helps to ac- count for the evidence of relatively limited change in trust perform- ance in the early stages of the reforms. Although progress continued to be variable, an increasing number of reports pointed to examples of innovations in purchasing, including the use of more sophisticated contracts and relational contracts.37 This suggested that the coun- tervailing force that was needed to challenge the power of trusts was at last beginning to achieve results. A more general point follows, namely, that bringing about change in health care takes time, and at- tempts to evaluate reform programs have to be sensitive to this. Alongside these largely positive effects, the reforms have also had some undesirable consequences. Most notably, a range of anecdotal and other evidence has it that equity has been undermined by the ability of GP ftmdholders to obtain quicker access to hospitals for their patients, regardless of clinical need.38 Concerns were also wide- spread that the reforms had increased transaction costs as the num- ber of managers employed increased to deal with the complex process of contract negotiation and monitoring. In response, plans to streamline the structure of the NHS and to simplify contracting arrangements were implemented. Targets were also set for reducing 292 a Innovations in Health Service Delivery management costs in both trusts and health authorities. An efficiency scrutiny that reported in 1996 set out a range of proposals for cutting back on paperwork and regulation, emphasizing the need to move toward long-term contracts in the internal market.39 Going beyond these assessments, I have suggested that the man- agement of the NHS should be guided by the following principles: o A commitment to a genuine and statutory separation of purchaser and provider roles o Development of devolved management of providers to ensure that decisions are made as close to the patient as possible o Support for the continuation of an independent commissioning or purchasing authority able to base planning on health needs o Recognition that the purchaser/provider system should be used to ensure accountability for the use of resources o Acknowledgment that the purchaser/provider system should en- courage collaborative arrangements in which purchasers and pro- viders work together on a long-term basis o A commitment to contestability rather than competition as a way of stimulating improvements in performance and providing in- centives for efficiency o Support for different models of GP involvement in commission- ing, including accountable fundholding o A commitment to reduce transaction costs by moving away from the annual contracting cycle to long-term contracting relation- ships.40 In this context, the term contestability is used to indicate the lower- ing of barriers to entry for other providers, which opened up the pos- sibility of switching providers if necessary. The starting point of such an approach is that efficiency and quality gains are best achieved by purchasers and providers working together to deal with issues of Autonomization and Centralization of U.K. Hospitals * 293 common concern. Only if this fails will purchasers consider moving contracts to alternative providers. However, this threat is ever pres- ent, and the knowledge that purchasers may move contracts and resources acts as a stimulus to providers to improve performance. The emphasis on contestability recognizes the need for a planned approach to be combined with marketlike incentives to stimulate providers to use resources for the benefit of service users. These in- centives include the use of comparative information on performance as well as the possibility to switch resources to alternative suppliers. It was felt that this would lead to "competition for the market" rather than "in the market," thereby avoiding expensive duplication in ser- vices, as was known to be effective in other sectors with a natural monopoly. An Assessment Drawing these strands together suggests that in the United King- dom, the failure of the organizational reforms to live up to the ex- pectations of its architects can be attributed to three factors: * Many trusts were reluctant to use their resources fully, and Trea- sury and the Department of Health continued to exercise hierar- chical oversight, undermining the independence and responsibil- ity of the trusts. * The external environment (both funding arrangements and mar- ket environment) did not generate sufficient competitive pres- sures. * The institutional context of reform and political pressures mili- tated against the use of the market in practice. Treasury and the Department of Health continued to exercise oversight, especially in relation to the financial decisions. Only in the area of management freedom did trusts have the powers anticipated 294 o Innovanons in Health Service Delivery at the outset, but the extent to which they used these powers to bring about improvements in efficiency and responsiveness depended on the attitude and abilities of trust boards and their senior managers. This meant that the nature and pace of change varied widely and re- lied more on management capacity than market stimulus. With both decision rights and accountability arrangements constrained by hier- archical controls, the potential benefits of trust status remained largely unfulfilled. The environment in which trusts operated did not always give rise to the conditions in which competition could emerge. Despite scope for competition in parts of the NHS, neither fundholders nor (espe- cially) health authorities showed great enthusiasm for using their leverage to improve performance by moving contracts and resources between providers. This inertia was reinforced by the constraints politicians placed on purchasers in the initial stages and the time taken to develop an effective purchasing function. Purchasers them- selves were not accountable for their performance, so they in turn had weak incentives to press trusts for improvements. And, trusts' "commercial in confidence" attitude to the use of information did not help. This situation might have been addressed by developing an explicit regulatory framework for the internal market, but the framework was introduced belatedly and did not deal directly with information disclosure. The reluctance to allow trusts to fail and to manage change through traditional administrative mechanisms per- petuated established working arrangements and demonstrated the difficulty of altering a firmly entrenched culture. Market exposure was therefore limited, and command-and-control mechanisms con- tinued to prevail. The institutional context in which the reforms were introduced, and the political influence that was brought to bear, also militated against the use of the market in practice. This is arguably of the greatest importance in explaining the failure of the internal market to live up to expectations. With the benefit of hindsight, it is not dif- ficult to explain why. To overturn a 40-year history of planning and regulation is no easy task. Competition might have had an impact if politicians had had the courage of their own (and their predecessors') Autonomization and Centralization of U.K. Hospitals ° 295 convictions and used regulation to encourage the market to develop. This did not happen. An assessment of the U.K. reforms using the framework described in chapter 1 is summarized in box 7.4. Box 7.4 Decision Rights The main freedoms gained were in the area of day-to-day decisionmaking. Trusts no longer had to seek permission from higher authorities and enjoyed more management au- tonomy than directly managed units. The financial freedom trusts had was constrained by Treasury rules on borrowing and the use of internally generated fumds, and by rules that prices should be based on costs. The freedom trusts enjoyed as employers to determine pay rates and contracts with staff was greater than their financial freedom, but was con- strained both by the staff's retention of its existing rights and contracts when it transferred to trusts, and by the require- ments of the specialist medical bodies. Residual Claims Trusts retained surpluses, but their use of these surpluses was governed by the operation of external financing limnts (see text). Accountability Trusts were managerially accountable to the Secretary of State for Health via regional offices of the NHS Manage- ment Executive. They were accoumtable to the public through annual reports, which were discussed at a public meeting. Trusts were accountable to purchasers through the contracts they negotiated for the delivery of services. (Box continues on the follmoing page) 296 ° Innovations in Health Service Delivery Ban 7.4 (coninud) Market aposure The scope for competition varied from place to place. Where a market existed, the extent to which competition occurred depended on the willingness of purchasers to engage in selective contracting and the willingness of gov- ernment to allow this to happen and to accept the conse- quences. GP fundholders were more inclined than health authorities to contract selectively with trusts but their im- pact was constrained by the limited range of services they purchased. Health authorities were discouraged from mak- ing major changes in their contractual arrangements with trusts in order to avoid destabilising the system. Social Functions Trusts continued to provide services to all patients and there was no evidence of discrimination against high cost service users. There were claims that inequity had widened as a re- sult of GP fundholders using their purchasing power to gain faster access to services irrespective of clinical need, though the extent to which this happened is unknown. Conclusions There is no simple way of summarizing the experience of the United Kingdom with reforming the provision of health services. If the re- forms have not fully delivered on the promises that accompanied their launch, the reasons include the contradictions inherent in man- aged markets, the failure to give sufficient attention to developing the purchasing function, and the time needed to make radical changes of this kind work in practice. Whether the result is a glass half full or half empty depends on one's perspective. Autonomization and Centralization of U.K. Hospitals * 297 These conclusions emphasize the clash between the logic of eco- nomic theories and the reality of political life. At root, the U.K. ex- perience demonstrates not just the gap between intention and im- plementation, nor even the inertia built into existing institutions. Rather, it indicates that even well-crafted policies, filtered into im- peccably designed agencies, will fail if politicians judge that the risks in implementing these policies exceed the benefits. What is therefore needed is an approach that acknowledges the political economy of health care reforrn given the inseparability of economic and political considcerations. Appendix 7A National Health Service and Community Care Act 1990 Schedule 2 Specific duties 6. (1) An NHS trust shall carry out effectively, efficiently and economically the functions for the time being conferred on it by an order under section 5(1) of this Act and by the pro- visions of this Schedule and, with respect to the exercise of the powers conferred by section 5(10) of this Act and para- graphs 10 to 15 below, shall comply with any directions given to it by the Secretary of State, whether of a general or a particular nature. (2) An NTIS trust shall comply with any directions given to it by the Secretary of State with respect to all or any of the following matters- (a) the qualifications of persons who may be employed as officers of the trust; (b) the employment, for the purpose of performing func- tions specified in the direction, of officers having qualfica- tions or experience of a descnption so specified; (c) the manner in which officers of the trust are to be ap- pointed; 298 o Innovations in Health Service Delivery (d) prohibiting or restricting the disposal of, or of any in- terest in, any asset which, at the time the direction is given, the Secretary of State reasonably considers to have a value in excess of such sum as may be specified in an order under sec- tion 5(1) of this Act and in respect of which the Secretary of State considers that the interests of the National Health Ser- vice require that the asset should not be disposed of; (e) compliance with guidance or directions given (by circu- lar or otherwise) to health authorities, or particular descrip- tions of health authorities; and (f) the implementation of awards relating to the distinction or merit of medical practitioners or dental practitioners or any class or classes of such practitioners. 7. (1) For each accountng year an NHS trust shall prepare and send to the Secretary of State an annual report in such form as may be determned by the Secretary of State. (2) At such time or times as may be prescribed, an NHS trust shall hold a public meeting at which its audited ac- counts and annual report and any report on the accounts made pursuant to subsection (3) of section 15 of the Local Government Finance Act 1982 shall be presented. (3) In such circumstances and at such time or times as may be prescribed, an NHS trust shall hold a public meet- ing at which such document as may be prescribed shall be presented. 8. An NHS trust shall furnish to the Secretary of State such re- ports, returns and other information, including information as to its forward planning, as, and in such form as, he may require. 9. (1) An NHS trust shall be liable to pay: (a) to the chairman and any nonexecutive director of the trust remuneration of an amount determined by the Secre- tary of State, not exceeding such amount as may be approved by the Treasury; Autonomization and Centralization of U.K. Hospitals * 299 (b) to the chairman and any nonexecutive director of the trust such travelling and other allowances as may be deter- mined by the Secretary of State with the approval of the Treasury; (c) to any member of a committee or sub-committee of the trust who is not also a director such travelling and other al- lowances as may be so determined. (2) If an NHS trust so determines in the case of a person who is or has been a chairman of the trust, the trust shall be liable to pay such pension, allowances or gratuities to or in respect of him as may be determined by the Secretary of State with the approval of the Treasury. (3) Different determinations may be made under sub-para- graph (1) or sub-paragraph (2) above in relation to different cases or descriptions of cases. Specific powers 10. In addition to carrying out its other functions, an NHS trust may, as the provider, enter into NHS contracts. 11. An NTIS trust may undertake and commission research and make available staff and provide facilities for research by other persons. 12. An NHS trust may- (a) provide training for persons employed or likely to be employed by the trust or otherwise in the provision of ser- vices under the principal Act; and (b) make facilities and staff available in connection with training by a university or any other body providing training in connection with the health service. 13. An NHS trust may enter into arrangements for the carrying out, on such terms as seem to the trust to be appropriate, of any of its functions jointly with any Regional, District or 300 0 Innovations in Health Service Delivery Special Health Authority, with another NTHS trust or with any other body or individual. 14. According to the nature of its functions, an NHS trust may make accommodation or services or both available for pa- tients who give undertakings (or for whom undertakings are given) to pay, in respect of the accommodation or services (or both) such charges as the trust may determine. 15. For the purpose of making additional income available in order better to perform its functions, an NHS trust shall have the powers specified in section 7(2) of the Health and Medicines Act 1998 (extension of powers of Secretary of State for financing the Health Service). General powers 16. (1) Subject to Schedule 3 to this Act, an NHS trust shall have power to do anything which appears to it to be neces- sary or expedient for the purpose of or in connection with the discharge of its functions, including in particular power: (a) to acquire and dispose of land and other property (b) to enter into such contracts as seem to the trust to be appropriate (c) to accept gifts of money, land or other property, includ- mg money, land or other property to be held on trust, either for the general or any specific purposes of the NTS trust or for all or any purposes relating to the health service; and (d) to employ staff on such terms as the trust thinks fit. (2) The reference in sub-paragraph (1) (c) above to specific pur- poses of the NHS trust includes a reference to the purposes of a specific hospital or other establishment or facility which is owned and managed by the trust. 17. (1) Without prejudice to the generality of paragraph 16 above, for or in respect of such of its employees as it may de- termine, an NHS trust may make such arrangements for Autonomization and Centralization of U.K. Hospitals * 301 providing pensions, allowances or gratuities as it may deter- mine; and such arrangements may include the establishment and administration, by the trust or otherwise, of one or more pension schemes. (2) The reference in sub-paragraph (1) above to pensions, allowances or gratuities to or m respect of employees of an NIHS trust includes a reference to pensions, allowances or gratuities by way of compensation to or in respect of any of the trust's employees who suffer loss of office or employ- ment or loss or diminution of emoluments. Acknowledgments. Several people have made helpful contributions to the writing of this chapter. I would especially like to thank the following who either commented on an earlier draft or made suggestions that have influenced my thinking Alike Bid- dle, Alan Langlands, Julian Le Grand, James Raftery, Ray Robmson, Clive Smee, and John Wyn Owen. I would also like to thank the staff of the World Bank who have been involved in the project of which this paper forms a part, most particularly Alex Preker, but also Gail Richardson, Loraine Hawkins, and April Harding. At the Health Services Management Centre, Brigit Ayling and her colleagues m the library helped me track down references, and Anne van der Salm turned my rough drafts into polished final products. Notes 1. C. Hood, "A Public Management for All Seasons," Public Ad- ministration 69: 3-19 (spring 1991). 2. S. Hall and M. Jacques, eds., New Times (Lawrence and Wis- hart: London, 1989). 3. E. Ferlie, L. Ashburner, L. Fitzgerald, and A. Pettigrew, The New Public Management in Action (Oxford: Oxford University Press, 1996). 4. D. Osborne and T. Gaebler, Reinventing Government (Reading, Mass.: Addison-Wesley, 1992); S. Ranson and J. Stewart, Manage- ment for the Public Domain: Enabling the Learning Society (London: Macmillan, 1994). 302 o Innovatons in Health Service Delivery 5. C. D. Foster and F.J. Plowden, The State under Stress (Bucking- ham, U.K.: Open University Press, 1996), p. 84. 6. K. Walsh, Public Services and Market Mechanisms (London: Macmillan, 1995). 7. NHS Management Inquiry ["Griffiths Report"] (London: De- partment of Health and Social Security, 1983). 8. Secretary of State for Health, Working for Patients (London: Her Majesty's Stationery Office, 1989). 9. This chapter draws on experience in the NHS in England. Similar trends were evident in the rest of the United Kingdom, al- though there were some differences in the organization of the NHS in those countries. 10. The NHS and Community Care Act of 1990, Schedule 2. 11. Department of Health, Self-Governing Hospitals, Working Paper 1 (London: Her Majesty's Stationery Office, 1989), p. 3. 12. C. Ham, "Where Now for the NI-IS Reforms?" BMJ 309: 351-52 (1994). 13. Department of Health, NHS Trusts: A Working Guide (Lon- don: Her Majesty's Stationery Office, 1990), p. 15. 14. C. Propper, "Agency and Incentives in the NHS Internal Mar- ket," Social Science and Medicine 40 (12): 1683-90 (1995). 15. C. Pollitt, J. Birchall, and K. Putman, Decentralising Public Service Management (London: Macmillan, 1998). 16. Department of Health, The Operation of the NHS Internal Mar- ket, London, 1994. 17. Ibid. 18. R. Hamblin, "Trusts," inJ. Le Grand, N. Mays, and J, Mulli- gan, eds., Learning from the NHS Internal Market (London: King's Fund Institute, 1998). 19. W. Bartlett and J. Le Grand, "The Performance of Trusts," in R. Robinson and J. Le Grand, eds., Evaluating the NHS Re- Autonomization and Centralization of U.K. Hospitals * 303 forms (London: King's Fund Institute, 1994); Newchurch, The Third Newchurch Guide to NHS Trusts (London: Newchurch, 1993). 20. Bartlett and Le Grand, "The Performance of Trusts," pp. 62-63; Newchurch, The Third Newchurch Guide to NHS Trusts. 21. N. Soderlund, I. Csaba, A. Gray, R. Mine, and J. Raferty, "Im- pact of the NHS Reforms on English Hospital Productivity: An Analysis of the First Three Years," BMY 315: 1126-29 (1997). 22. By I. Csaba, A. Gray, R. Milne, and J. Raftery, as reported in C. Propper and N. Soderlund, "Competition in the NHS Internal Market: An Overview of its Effects on Hospital Prices and Costs," Health Economics 7: 193 (1998). 23. Ibid. 24. C. Propper, "Market Structure and Prices: The Response of Hospitals in the U.K. National Health Service to Competition," Journal of Public Economics 61: 307-35 (1996). 25. C. Smee, "Self-Governing Trusts and GP Fundholders: The British Experience," in R. Saltman and C. von Otter, eds., Imple- menting Planned Markets in Health Care (Buckingham, U.K.: Open University Press, 1995). 26. R. Hamblin, "Trusts," in Le Grand, Mays, and Mulligan, Learning from the NHS Internal Market. 27. Smee, "Self-Governing Trusts and GP Fundholders," p. 190. 28. J. Appleby, P. Smith, W Ranade, V Little, and R. Robinson, "Monitoring Managed Competition," in Robinson and Le Grand, Evaluating the NHS Reforms; C. Propper, "Regulatory Reform of the NHS Internal Market," Health Economics 4: 77-83 (1995). 29. J. McHale, D. Hughes, and L. Griffiths, "Conceptualising Contractual Disputes in the National Health Service Internal Mar- ket," in S. Deakin and J. Michie, eds., Contracts, Co-operation and Competition (Oxford: Oxford: University Press, 1997). 30. Robinson and Le Grand, Evaluating the NHS Reforms. 31. OECD (Organisation for Economic Co-operation and Devel- opment), OECD Economic Surveys: United Kngdom 1994, Paris, 1994. 304 0 Innovations in Health Service Delivery 32. K. Bloor and A. Maynard, "An Outsider's View of the NHS Reforms," BMJ7 309: 352-53 (1994). 33. A. Maynard and K. Bloor, "Introducing a Market to the United Kingdom's National Health Service," New England journal of Medicine 334: 604-608 (1996); R. Klein, "Big Bang Health Care Re- form-Does It Work? The Case of Britain's 1991 National Health Service Reforms," Milbank Quarterly 73: 299-337 (1995). 34. Le Grand, Mays, and Mulligan, Learning from the NHS Inter- nal Market. 35. C. Ham, Management and Competition in the NHS (Oxford: Radcliffe Medical Press, 1997). 36. Ferlie, Ashbumer, Fitzgerald, and Pettigrew, The New Public Management in Action. 37. J. Raftery, R. Robinson, and J. Mulligan, "Contracting in the NHS Quasi-Market," Health Economics 5: 353-62 (1996); L. Locock and D. Welsh, Pulling the Right Levers (Oxford: NHS Executive An- glia and Oxford, 1994); and R. Flynn, G. Williams, and S. Pickard, Markets and Networks: Contracting in Community Health Services (Buckingham, U.K.: Open University Press, 1996). 38. N. Goodwin, "GP Fundholding," in Le Grand, Mays, and Mulligan, Learning from the NHS Internal Market. 39. NHS Executive, Seeing the Wood, Sparing the Trees (Leeds, U.K.: NHS Executive, 1996). 40. C. Ham, Public, Private or Community? What Nextfor the NHS? (London: DEMOS, 1996). CHAPTER 8 Maladjustments in the Corporatization Model: Hospital Reform in New Zealand Graham Scott, Lynne McKenzie, and James Webster This chapter, written in 1999, considers the corporate modelfor hospitals used in 1993-1999. New Zealand, a country about the same size as the United King- dom, spends NZ$8.0 billion a year on health care for its 3.6 million people. Publicly funded expenditure accounts for 77.1 percent of this amount.1 Private expenditure is mostly out-of-pocket spending by consumers on services (16.4 percent), on private health insurance (6.2 percent), and on charitable organizations (0.3 percent). Al- though private insurance accounts for a relatively small proportion of expenditure, 3 7 percent of the population has some private insur- ance coverage. New Zealand spends 7.6 percent of GDP on health, just below the OECD average of 7.7 percent.2 Government funding covers a broad range of services: hospital- based care (emergency department, medical and surgical inpatient; outpatient, and hospital-based community services); disability sup- port and mental health services (residential, specialist services, and community-based services for people with age-related, physical, sen- sory, or psychiatric disabilities); public health functions; primary 305 306 0 Innovations in Health Service Delivery health care services; maternity care; other primary care-referred ser- vices (prescription pharmaceuticals, laboratory and radiology testing); accident-related health care, rehabilitation services, and income- protection insurance. The government agency responsible for purchasing health ser- vices spends NZ$6.5 (US$3.2) billion a year, half of it on services provided by Crown Health Enterprises (CHEs), the government- owned hospital and related service providers.3 CHEs dominate the hospital sector, providing most of the acute and elective care in the tertiary and secondary sectors from hospital-based care and special- ist services to community care and public health functions. Private hospitals focus on delivering nonacute services such as outpatient services and elective surgery. Nongovernmental organizations (NGOs) provide general pri- mary health care services, some specialist services, long-term care for the elderly and disabled, laboratory services, and services targeted to the indigenous Maori population. Since the reforms of 1993, the pri- vate sector has enlarged its role in delivering community-based dis- ability support, mental health, and maternity services. Accident-related services, including income-protection insurance, are funded and managed separately through the government-owned Accident Rehabilitation Compensation Insurance (ARCI). In July 1, 1999, the government enabled employers to purchase insurance from private companies to cover workplace accidents, however the government elected later in 1999 quickly nationalized that function. For more than a dozen years, the government has been continu- ally remodeling its health system. From the mid- to late 1980s, hos- pitals were formed into area health boards (AHBs), a set of 14 re- gionally based institutions, governed by a board with a majority of locally elected members. In 1993, the AHBs were reformulated into 23 Crown Health Enterprises as part of the organizational reform discussed in this chapter.4 Four government-owned purchasing agen- cies, regional health authorities (RHAs), were established in 1993. They were merged in 1997 into one purchasing agency, the Health Funding Authority (HFA). Under the AHB structure, hospitals were quite autonomous from the central government until the 1993 reforms. Paradoxically, these Malad)ustments in the Corporatization Model * 307 reforms led to the use of more direct accountability mechanisms by the government as their shareholder, while establishing an arm's- length contractual relationship with the purchasing organizations. To understand the policies behind the reforms and the changes in the sector's performance, we have examined official documents and reports. We have also relied heavily on information provided in in- terviews of people involved in the health system before, during, and after the reforms.5 Reforms: 1980s to 1993 In 1983, the government enacted legislation to create regional pub- lic sector agencies, area health boards, which managed public hospi- tal services, community health services, and public health programs. AHBs were not responsible for most primary health care, which was provided mainly by private general practitioners, subsidized on a fee- for-service basis by the Department of Health (table 8.1). Though highly autonomous, AHBs made little use of their free- dom to contract out services. Decisionmaking by the elected board members was sometimes diverted from long-term rational service goals toward short-term political objectives. As a consequence, capi- tal maintenance and investment spending sometimes took low prior- ity to immediate service needs. By some reports, individual board members intervened so deeply in organizational management that they undermined management's authority. By the end of the 1980s, the AHBs' financial problems were apparent. Some large AHBs had borrowed extensively from the private sector, which perceived an im- plicit government guarantee. This posed fiscal risks to the central government since it lacked commensurate power to monitor and control the decentralized AHBs. A change of government in 1990 brought the next set of health re- forms, which had their genesis in a 1991 government white paper.6 This remodeling of the system created four regional health authori- ties-government-owned purchasers that received funding for a wide range of services. Initially, these public purchasers were to compete with private sector "alternative health care plans." However, this idea 308 o Innovations in Health Service Delivery TeiUo 8.I Hospital Powers Pre- and Post-1993 Reform CROWN HEALTH ENTERPRISES ITEM AREA HEALTH BOARDS (1983-931 (FORMED IN 1993) Objectives 0 To promote, protect, and con- o To provide health and disability serve public health and to services provide health services o To provide those services in o To provide for effective coordi- accordance with Statement of nation of planning, provision, Intent and any contract entered and evaluation of health ser- into by it while operating as a vices between public, privote, successful and efficient business and voluntary sectors o To exhibit sense of social o To establish and maintain responsibility by considering appropriate balance in provi- interests of community in which sion and use of resources for it operates health protection, health pro- o To uphold ethical standards motion, health education, and expected of health or disability treatment services (Section 9 service providers Area Health Boord Act 1983) - To be "good employer" o To be as successful and efficient as comparable businesses not owned by Crown (Section 11 Health and Disability Services Act 1993, abbreviated) Boards and Majority of board members All board members appointed by accountability elected Minority appointed minister of finance and CHEs by minister of health Accountable to regional health Accountability through elections, authorities (RHAs) for service open board meetings, and to delivery and outputs and to mm- minister empowered to dismiss ister for management of people boards and assets Monitoring Department of Health monitored CCMAU monitored "ownership" performance against contract. matters. RHAs monitored service Monitoring by voters delivery against contract Residual Government, as owner, paid for Same as for AHBs claimant losses despite a statement in the law that AHBs were not part of Crown Gains tended to be spent on more services and higher salaries so public and employees were residual holders of gains. Market Government monopoly for acute Natural geographic monopoly exposure care with small number of pri- with small number of private vate hospitals not in significant hospitals not in significant competition competition. Twenty-three CHEs Maladjustments in the Corporatization Model * 309 Table 8.1 (continued) CROWN HEALTH ENTERPRISES ITEM AREA HEALTH BOARDS (1983-93) (FORMED IN 1993) created from 14 AHBs Competition law appiies Increased competition in long- term and community care Capital finance Depreciation included in overall Depreciation included in RHA capitated funding revenue Equity injections and Borrowed from private sector loans from government Also borrowed from private sector Management AHB general managers had full Same as for AHBs but outputs decision rights delegations for these matters were specified in much * hiring/firing except for determining out- greater detail in contracts by * pay puts, setting user charges, and RHAs, also, the State Services * capex working with the State Ser- Commission was less involved * supplies vices Commission on national in collective employment con- * contracting out labor collective agreements tracts, having only a consulta- * outputs Freedom to determine outputs tion role * prices was being reduced by con- tracts with minister of health Government set prices Provider Capitated payment based on Ninety percent of funding from payment size and characteristics of RHAs was based on contracts population covered by AHB to deliver specified volumes district. Small amount of of services Remaining fund- bequests and gifts Investment ing came from user part income Adjustment for cross- charges, donations, invest- boundary patient flows, ment income, and other based on service volume government purchasers 0 Rationing Undertaken principally by RHAs acting under government decisions AHBs, except where specified policy, which included some and agreed to by minister of specific directions on services health through contract to be purchased Social functions Unspecified mondates for clini- Increased specification and cal teaching, research, ser- explicit payment for these vices for other government mandates deportments, services for nonresidents a Other government purchasers included the Public Health Commission and Accident Rehabili- tation Compensation Insurance Sources Area Health Board Act 1983, Health and Disability Services Act 1993, authors' knowl- edge of policies and practices. 310 o Innovations in Health Service Delivery was abandoned, and the RHAs operated as public sector regional monopolies. Other new institutions included a Core Services Committee, set up to advise on the basic package of services to be funded by the pub- lic health system, and a special purpose commission, established to oversee public health policy and purchasing functions. The public hospitals were established as government-owned companies, incorporated under New Zealand company law. Minis- ters held the shares and appointed the board members. The model of government-owned companies for hospitals was an adaptation of the earlier reform of government-trading activities into state-owned enterprises (SOEs). This influenced the choice of business struc- tures, objectives, and personnel for the CHEs. The SOE reforms had been so productive that many advisers implementing hospital reform expected similarly high efficiency gains from the hospitals, including improved services, shortened waiting lists, and reduced costs.7 The wider context for these changes included a major exercise in developing policies to sharpen social expenditure targeting across all categories, including health spending. Copayments were introduced, and community services cards were issued to low-income people to reduce or negate their copayments. Major Changes: Area Healti hirds b RHAs and CHIEs The 1993 changes focused primarily on removing service planning and resource allocation from area health boards. These functions were placed in the RHAs as independent purchasing bodies, but politicians exercised considerable influence over purchasing deci- sions.8 The 1993 reforms did not significantly increase hospital au- tonomy and, as finally implemented, actually increased centralization of the health system's national accountability. In practice, CHE de- cisionmaking was limited. For example: o Decisions to exit major services and reconfigure facilities were sometimes reversed by ministers. Maladjustments in the Corporatization Model * 311 * Investments in major capital items had to be approved by minis- ters through business plans or separate business cases for large investments. * Investments in new lines of business were sometimes discouraged; for example, politicians intervened in the offshore business devel- opments of one CHE.9 * Strong stances in politically uncomfortable stand-offs with doc- tors were discouraged, for fear of strikes. The financing of CHEs and AHBs differed. Service payments to AHBs were based on capitation under a broad contract to the minis- ter of health. CHEs were expected to provide senrices specified by the RHAs under contracts that increasingly paid CHEs according to their output. AHB decisionmaking was influenced by broad policy directions from the government and specific requirements from their commu- nities, reflected through the decisions of locally elected boards. Under the 1993 system, the central government and its agencies decided many matters, including RHA pricing policies, CHE capi- tal expenditures, and service rationalization in politically sensitive areas. The monitoring regime involved less direct public scrutiny and influence over hospitals' decisions. Local communities no longer elected board members and, unlike AHBs, the board meetings of CHEs and RHAs were closed to the public. The consultation obli- gations on the 14 AHBs moved principally to the four RHAs rather than the 23 CHEs, possibly leaving communities feeling they had less ready access to decisionmakers. RHAs monitored the CHEs mainly on compliance with service delivery contracts. The Crown Company Monitoring and Advisory Unit (CCMAU) monitored the ownership interest on behalf of the shareholding ministers.10 The Ministry of Health monitored the RHAs on behalf of the minister of health. Under the former system, the Department of Health held all the monitoring functions on the area health boards. 312 o Innovations in Health Service Delivery nmplementaion during lie 1993-96 Period New Zealand's health reforms took place in the context of a broad public sector reform, part of a drive to strengthen public agencies' accountability to the national government. Corporatization and pri- vatization were part of this reform agenda, invoked wherever normal service or capital market mechanisms were considered capable of providing direct accountability to consumers and shareholders. When corporatization was applied to the health system, strong so- cietal values came to the fore, favoring universal health care as a largely free public service. The public did not accept the premise that an organization with commercial objectives could be given appropri- ate incentives to provide high-quality health care to all. The 1993-96 period was a testing time for the sector. The policy framework for the reforms was marked by continual ad hoc changes that diluted the original design of the reform. First, elements of com- petition were removed, and no arrangements were developed to deal with the resulting bilateral monopolies engaged in the purchasing and provision sides of the sector. Tensions resulted that were exacer- bated by poor information on service levels, quality and costs, and funding issues. The following sections discuss the formal arrange- ments, the market, financial aspects, and the implementation of the reforms during the 1993-96 reform period. Legal Structure, Governance, and Accountability Arrangements for CHEs The structural and accountability arrangements for CHEs were more attuned to a competitive model, with an arm's-length relation- ship to the government, than the model that developed in practice. Legalframe7vork. As noted, CHEs were government-owned compa- nies set up under the Health and Disabilities Service Act of 1993, with the ministers of health and finance as the only shareholders.11 They were subject to the Companies Act, which made directors per- sonally liable for their decisions regarding the company's financial Maladjustments in the Corporatzation Model * 313 viability. In reality, most CHEs held letters of financial support from their shareholding ministers.12 Boards thus faced no financial risk. Revisions in the legislation in 1998 altered CHEs' objectives, re- ducing the emphasis on efficiency and specifying that they should operate as not-for-profit organizations. Decision rights. At first sight, the CHE boards appeared to have wide discretionary powers over the governance and management of their organizations. They could hire employees on individual or collective contracts in accordance with New Zealand's general labor laws and raise private sector debt. Less tightly monitored on their financial sit- uation than AHBs, they also had more freedom to manage their cash. Although CHE boards seemed to have a loose rein, the govern- ment made many significant decisions affecting CHEs such as deter- mining plans for capital and service development. CHE boards could resist informal directions from ministers and the monitoring agen- cies on the grounds that the Companies Act gave them business management responsibility. On occasion, informal directions raised questions about whether the parties giving the directions could be deemed directors of the CHE. Pressure put on the boards by minis- ters was not always transparent. In practice, boards that persistently ignored ministers' wishes risked losing their members, but this did not always deter the boards. Reputational risk from dismissal from the board was small, partly because many board members resigned, with or without pressure from the government, without any loss of reputation. Some resigna- tions were provoked in disputes with the government over funding for board members' community hospitals, but the reasons for other departures often were not clear. Accountability arrangements. The accountability arrangements for the CHEs had four main components: board appointments, the state- ment of intent, their business plans, and their purchase contracts. * Board appointments. The ministers engaged the CHE chair and board members, usually by a simple letter of engagement, with no 314 o Innovations in Health Service Delivery explicit statement of expectations. It was an accountability lever with some potential to be used more effectively. o Statement of intent. The statement of intent, prepared under the Public Finance Act, was submitted to the minister and tabled in Par- liament. The general purpose of these statements, required for all Crown entities, was to inform Parliament about the entity's main purposes and direction. CHEs did not always consider these state- ments important documents, a partial reflection of weaknesses in performance monitoring of these documents by the parliamentary select committees and ministers. o Business plan. Each CHE agreed with the ministers on its business plan. The plan contained strategic objectives, projected capital ex- penditures, planned service changes, changes in asset configurations, projected revenues, projected costs, and financial forecasts. Business plans were used as the basis for agreeing on deficit funding from the Crown and equity funding for capital development. The business plan carried much more weight with CHE boards and management than the statement of intent or the purchase contract and became the main tool for controlling the CHEs. Failure to meet business plan fiscal targets usually resulted in problems for CHEs, and sometimes in the removal of the chairs and the CEOs. o Purchase contracts. The purchase contracts, initially with the local RHA and later the Health Funding Authority, accounted for 90 per- cent of CHEs' revenue.1 3 These agreements were legally based con- tracts for the delivery of specified services at agreed prices. Purchase contracts held little weight with CHEs initially. When contract breaches occurred, the parties did not take strong action. Though separate legal entities, these contracts were de facto internal government contracts. The threat of enforcement lacked credibility because of reluctance to go to court. Centrally imposed political so- lutions were the usual method of resolution. In serious disputes be- tween RHAs and CHEs, officials or ministers intervened, compli- cating the contractual relationship. Their involvement reflected the Maladjustments in the Corporatization Model * 315 lack of solid information about the contended issues, which is needed for a rational and transparent approach. On occasion, ministers or central agencies put pressure on the parties to sign contracts that left major issues unresolved or contained unrealistic provisions. External Environment: Market Structure and Purchasing Arrangements The CHEs were set up around 24-hour acute care facilities. This was done to encourage competition between CHEs in services where competition was practicable, given the geographical catchment areas for acute services.14 Under these configurations, however, some CHEs were not clinically and financially viable. This problem was known when they were set up. The rationalizations and mergers that CHE designers expected did not materialize. 15 Apart from one mer- ger, the remaining 22 CHEs retained similar configurations, and po- tentially unviable configurations persisted. Nor did CHEs cooperate well in planning, managing, and developing services. No capital market or clear administrative process impelled take- overs or mergers, which would have required detailed ministerial in- volvement. The Crown Company Monitoring and Advisory Unit had little success encouraging CHEs to consider mergers. Instead, the unit appointed common board members to CHEs that could benefit from joint service rationalizations. Geographical monopoly over 24-hour acute care services pre- vented competition for their delivery. Ministers discouraged some CHEs from looking for business in the catchment areas of other CHEs. Private providers won bids from CHEs in some localities but, to prevent purchasers from threatening the public hospitals' viability by selectively shifting services to the private sector, formal proce- dures were imposed.16 Some remarkable expansions in private delivery occurred in con- tinuing care, disability support, and mental health services in the de- institutionalization program (only a small part of all CHE services). Particularly with continuing care and disability support services, 316 o Innovations in Health Service Delivery there was a ready private market. The RHAs purchased these ser- vices competitively but consumers, when allowed a choice, chose pri- vate facilities. As a result, many CHEs exited these services. Little competitive purchasing of elective surgery occurred; less than 5 percent for the most active RHA in this area. Although the private hospital market for elective surgery was established, difficul- ties arose. There were concerns about clinical viability if CHEs lost too many surgical cases. Rarely was the private market cheaper than the CHEs. Private hospitals believed CHEs were underpricing and shifting costs onto their government-funded deficits. Some doctors in public hospitals resisted competition and did not want to release waiting list details to the successful bidder, nor were they always eager to execute contracts won by the CHE they worked for. Some consumers did not wish to move out of their local area for treatment. The resistance of doctors and consumers created tensions that jan- gled political nerves. The bilateral monopoly between the government's purchasers and providers meant that prices were difficult to determine, and some- times contracts settled, if they settled at all, on the side of the party with the strongest bargaining power. RHAs sometimes used their dominant position to offer "take-it-or-leave-it" deals, backed by their statutory powers to impose their terms.17 In a commercial environment, such bilateral monopolies recognize that failure to agree on workable arrangements with each other would result in the loss of mutual gains. The parties might be expected to move toward one of two scenarios: a long-term contract, specifying how the parties would distribute surpluses and risks and behave under future unforeseen circumstances requiring contract adjustments; or vertical integration, so that the tools of governance could be used to manage transactions under a common maximization objective. The CHEs, in their environment of poor accountability, instead used po- litical and other pressures such as the media to extract gains. Competition. New Zealand's competition laws and supporting organ- izations have been little concerned with health sector issues. Prob- lems have arisen with the market definition, a prerequisite to discov- Maladjustments in the Corporatization Model * 317 ery of monopolistic abuse in the market. Monopoly issues have been addressed piecemeal and dealt with inconsistently in the contracting processes. For example, some contracts have restricted a provider's market share. Separating the CHE contracts into contracts for services and con- tracts for facilities proved to be one way to improve the market. This enhanced the purchasers' ability to target changes directly to the clinical decisionmakers and to design contracts that provided incen- tives for them. This approach is operated for maternity services where CHEs held contracts to provide facilities, while other organ- izations had contracts to provide clinical services.18 Purchasing arrangements. The complexity of the CHEs' outputs cre- ated measurement and contracting problems. RHAs could not spec- ify services and prices in sufficient detail to allow CHEs to calculate whether their deficits were due to inefficiencies, faulty measure- ments, pricing peculiarities, or a combination of these factors. The initial RHA/CHE contracts specified service levels in very broad terms. As these service levels were refined into detailed measurement units (e.g., diagnostic-related groups for hospital inpatient services), prices and volumes were adjusted accordingly. Each CHE's revenues shifted from year to year with changes in purchase unit definitions and prices. Revenue swings also occurred as CHEs improved their information systems. Many services not previously captured by in- formation systems were recorded. These revenue shifts undermined the credibility of the contracts, the contracting process, and the par- ties, and made planning difficult for all. All these problems created an environment of conflict, which the government attempted to manage by ad hoc interventions instead of analysis and action to deal with the underlying issues. Noteworthy government interventions in the contracting arrangements were the setting of revenue and service levels for CHEs in 1993-94 and the central agencies' allocation of overall revenue for CHEs in 1994-95. Both activities were based on rudimentary service-cost estimates. CHEs resisted signing contracts. In an environment with poor in- formation and no realistic options for alternative providers, CHEs 318 o Innovations in Health Service Delivery could refuse to sign contracts with few consequences. Commonly, in any one year, half the CHEs operated without formal service and revenue contracts. Funding and Financial Structures Publicly funded health care will always generate financial tension- the system's built-in budget constraints will run head on into grow- ing demand and expectations from the community, as expressed through the political system. However, more financial tension than expected accompanied the 1993 New Zealand health care reforms for several reasons. Purchaserfinding. Ministers and their advisers had unrealistic expec- tations of the efficiencies that could be achieved and thought they would allow a match between funding and demand for services. Ex- tracting efficiency gains in an environment of static or declining real revenue early in the reform proved difficult. Historically, many services were funded by open reimbursements, and RHAs made steady but slow progress in negotiating new arrangements for capping budgets and sharing financial risk with providers. For example, the movement of primary medical services from fee-for-service to capitation was slowed by resistance from en- trenched provider interests. Although the RHAs tightened manage- ment control over expenditure in some demand-driven areas, these measures had little impact on services driven by clinical decisions. The pressure on fee-for-service budgets encouraged RHAs to resist CHEs'requests for price and service increases.19 Financial environment. The reforms were implemented in a period of financial stringency. In 1988-89, when most area health boards were created, real per capita funding for health was NZ$1,391. This was re- duced to NZ$1,292 the following year, and did not reach the 1988-89 level again until 1996-97.20 By 1992-93, hospital boards, then AHBs, had operated for a decade with the same revenue.21 In 1992-93, AHBs ran an aggregate deficit of NZ$66.1 million (-2.5 percent of costs).22 In 1993-94, the CHEs' first year of operations, the deficit was NZ$175 million (-6.6 percent of CHE costs). Maladjustments in the Corporatization Model * 319 The reforms' transition costs were not financed in their initial years and showed up largely as higher CHE deficits. In the first year of the reforms, the RHAs' revenue was lower than the cost of AHB services the previous year and lower than the CHEs' projected costs for 1993-94. The RHAs' costs were covered by reducing CHE rev- enue by 2 percent. CHEs were expected to absorb the set-up costs the first year, along with the costs of inflation-while maintaining the previous year's output and range of services. Many AHBs were believed to have run down their assets and un- derstated their depreciation expenses to improve their short-term re- ported financial condition. When the CHEs were formed, asset val- ues were written down and debt was written off.23 The likely reasons for these large write-downs of assets were the unrealistically low de- preciation rates on AHB assets and a high degree of redundancy and obsolescence; many facilities were overdesigned in relation to the services they were delivering.24 The government established CHEs with an opening debt position based on debt-equity ratios typical in commercial companies, and provided further loans to finance cash deficits, restructuring costs, and any urgent capital expenditure.25 This was part of a policy regime to simulate corporate financial structures so that CHEs had to con- front the opportunity cost of capital in a way that encouraged them to focus on the value of the business as a whole. The CHEs were to refinance this debt from private sources over time. To encourage this shift, CHEs were charged above-market interest rates. Shifting their borrowing from the government to the private sector was to have put CHEs on the same footing as private firms, and private lenders were expected to assume a monitoring role. In July 1993, the aggregate CHE debt held by the government stood at NZ$289.7 million. By June 1997, this debt had decreased very little, to NZ$280.1 million, while borrowings from other sources amounted to NZ$305.5 mil- lion. The total debt-to-equity ratio was 68 percent.26 A pivotal requirement for private debt to be an effective monitor- ing influence on CHEs was that the government not guarantee the debt. CHEs were required to place a disclaimer in debt contracts to emphasize this. In practice, private lenders perceived CHEs as gov- ernment-guaranteed, regardless of disclaimers, inasmuch as govern- 320 - Innovations in Health Service Delivery ment was not thought likely to allow a CHE to fail in case of finan- cial disaster. Although many CHEs were in serious financial trouble, none failed to service its creditors. In these circumstances, the value of monitoring by private financial institutions would be unlikely to exceed the extra cost of private debt above the government's lower borrowing cost. If the government wished, it could have used its bor- rowing powers to secure cheaper funds and lend these to CHEs at lower rates than the artificially high rates used to encourage CHEs to move financing to the private sector. Though expected to earn a normal return on equity, CHEs rarely did so until the 1998-99 financial year. From 1993 to 1997, no more than three CHEs ran a surplus in any one year.27 CHEs did not act like typical commercial businesses, often delivering services beyond their funding. CHEs that cut off services upon reaching contracted levels were, on occasion, directed by the government to provide them.28 CHEs were also restrained from exiting from unprofitable services,29 which contributed to the deficits most CHEs experienced. This account of the structuring of the CHE finances reveals many problems. Not all these problems should be seen as a negative aspect of the reforms-in fact, quite the reverse. The CHEs' financial and accounting systems had not previously followed the standards re- quired under general company law and "generally accepted account- ing practice." Introducing these requirements brought to the surface long-concealed problems in the hospitals' underlying financial con- ditions, just as corporatization had in government-owned enterprises. The proper financial values of obsolete, run-down, and underutilized capital stock began to be revealed. This step was vital to put hospitals on a sound footing. Government's response to a difficult financial situation and un- derdeveloped management systems was piecemeal. Constant adjust- ments in funding to address specific problems undermined incen- tives for financial discipline. Holding back on capital maintenance and development and on staff training helped bridge the financial gap, but only in the short term.30 Financial pressures manifested themselves in different places at different times, wherever the weakest point was located. Due to Maladjustments in the Corporatnzation Model * 321 widespread unrealistic expectations, hospital finances can be seen, with hindsight, to have been out of equilibrium from the beginning. The Reform Process A number of agencies were involved in implementing the reforms in the 1993-96 period. No one agency or person had overall control of the implementation phase below the ministerial level. Policy and im- plementation of purchaser and provider reforms were managed sep- arately. As a result, inconsistent and confusing policy messages exac- erbated inherent tendencies for conflict between RHAs and CHEs. The reforms were implemented in a top-down, "big bang" manner. An 1 8-month implementation period was planned, after the release of the government paper in 1991, for the disestablishment of old entities and the creation of new ones. RHAs were formed in October 1992 and given 8 months to set themselves up to take on purchasing from 23 new CHEs as well as other providers. At the same time, the 23 CHEs were being set up. Purchasers and providers were struggling with poor information, inadequate systems, and the need to hire and rapidly train staff for the new roles. Finding suitable people for the management and board positions was difficult, given the small pool of available talent, the many open slots, and the timing of recruitment. The policy framework in the government's paper relied on the creation of purchasers, providers, and monitoring agencies that could quickly comprehend and start to perform their roles. Acquir- ing the necessary skills and information took the institutions a long time and, in some areas, they were still not well developed by the next restructuring in 1997. The expectations of rapid gains in health service planning and delivery turned out to be unreasonable in terms of the time and resources needed to build new organizations. Stakeholders and Communication The instigators of the New Zealand health reforms were politicians, managers, and civil servants. Change was resisted by public sector health unions and the Medical Association, which perceived no ben- efit and some threat from the reforms. Parliamentary opposition par- 322 o Innovations in Health Service Delivery ties of the Left supported these groups. Health reform thus became a party-political issue-with attendant risk of policy reversal in a change of government.31 The views of the general public, patients, and health sector suppliers were muted in the policy debate and im- plementation process. To explain reform to the public, a special communications unit was established, public meetings were held around the country, and mass media and leaflet drops were used. Nonetheless, many health sector participants interviewed for this report did not consider the communication effort a success. The campaign tried to sell reform instead of first explaining the problems reforms were intended to ad- dress. The public was not very aware of the previous system's prob- lems or their structural origins. Hence, the reforms were hard to promote as a solution to such popular concerns as waiting lists for nonurgent surgery. Particularly unpalatable were the increases in pa- tient copayments that accompanied the reforms and contributed to an impression that the focus on increasing financial restraints and re- ducing the state's role in financing health care would impair delivery of quality health care to the public. In a tough budget speech to the Parliament, the minister of fi- nance announced the reforms. In an environment of fiscal crisis and spending cuts, many people saw the health reforms as cost-cutting measures for which they would receive few benefits. Senior politicians gave no strong public impression of unified leadership. Many government politicians, other than the ministers directly involved, did not fully understand the reforms. Major changes in the original policy made the messages harder to explain. CDhanges in Policy and Shrucwres afer 11996 After a change of government in 1996 to a coalition that stressed so- cial spending, decisionmaking was further centralized. The scope for provider competition was reduced, and CHEs were given nonprofit status. The parliamentary Opposition parties continued to call for more extensive policy reversals. Maladjustments in the Corporatization Model * 323 The Coalition Government's health policy led to the merger of the four RHAs into the Health Funding Authority. At an operational level, two developments affecting CHEs are of interest: national contracting and strengthened accountability for the Health Funding Authority. National Contracting The Health Funding Authority (HFA) standardized the measure- ment units used across New Zealand so that prices and services could be compared across providers. The prices were set at a flat payment per unit across all CHEs, with limited variations to account for the costs of tertiary services and diseconomies of scale in small hospitals. Essentially, a national approach replaced 23 separate bilateral nego- tiations on price. This pricing approach involved calculating the efficiencies CHEs could generate from their current operating costs. These efficiencies were estimated using Data Envelopment Analysis, which modeled the production frontier for specific services. This work resulted in price increases that eliminated the deficits of 10 CHEs. HFA Accountability The new system introduced a much stronger accountability frame- work for the Health Funding Authority, based on a strategic business plan developed with key stakeholders. The plan underpinned the formal external accountability requirements and was integrated with the management plans and internal accountabilities. The plan in- cluded a "Service Coverage Document" that set out the services to be provided from public funds. The framework described the outputs of the Health Funding Authority such as contracting for services, mon- itoring and payment, priority setting, and policy advice on service mix. The costs of these outputs were to be increasingly identified and refined. Money for the Health Funding Authority's outputs was ap- propriated separately from its purchasing fund. This new framework was expected to markedly lift the quality of planning and manage- ment and increase transparency. 324 o Innovations in Health Service Delivery Assessment of &he Impad of Organizational Reforms in New Zealand How did these reforms affect technical efficiency, allocative effi- ciency, clinical outcomes, consumer satisfaction, and equity of ac- cess? No comprehensive examination has been done in these terms. One study that compared CHE and AHB performances concluded: CHEs are large, complex organizations with a high public pro- file and which operate in a dynamic external environment. An- alyzing relative and absolute perforrnance in this context is a difficult task. It is dangerous and unfair to form superficial con- clusions about either individual or aggregate CHE perfor- mance. Reasonable conclusions can only be drawn on the basis of detailed analysis and having due regard to the performance of all components of the institutional framework within which CHEs operate.32 Our comments can thus be only tentative. Because of a lack of sys- tematic evidence, our judgments involve an unavoidable element of subjectivity. Technical Efficiency An examination of hospital board and AHB technical efficiency over a decade concluded that this component of performance did not change significantly during the reform period.33 CHEs had to make only small gains to better their predecessors, but cost and output data throw little light on the differences between AHBs and CHEs. They were recorded neither fully nor consistently and therefore cannot be compared across structural modalities.34 From 1989-90 to their final year, 1992-93, AHBs moved from a surplus of NZ$41 million to a deficit of NZ$66 million. The fluctu- ations in cost changes mirrored revenue changes for two years, with a large swing from increases to decreases. In their final year, costs in- creased while revenue decreased (table 8.2). Maladjustments in the Corporatization Model * 325 Table 8.2 Area Health Boards Aggregate Trends in Costs and Deficits IAREA HEALTH BOARDS 1989-90 1990-91 1991-92 1992-93 Deficit/surplus (NZ$ million) 41 -2 -12 -66 Costs (NZ$ million) 2,456 2,626 2,571 2,628 Deficits as percentage of costs 1 7 -O 08 -0 5 -2 5 Percentage change in costs over previous year 6 9 -2 1 2 2 Percentage change in revenue over the previous year 5 1 -2 5 -0 3 Percentage change in consumer price index 2 8 1 1 3 Note. CPI figures are for all groups of products and services, with the index based on the June 1999 quarter The CPI for each CHE financial year is taken as the percentage change for the June quarter over the June quarter of the previous year Statistics Department Tables, Wellington, 1999 Source Deloitte Touche Tohmatsu, "Trends in Area Health Board/CHE Performance 1989/90-1997/98," Report for CCMAU, Wellington, 1996, CPI figures from Statistics Department Tables, Wellington, 1990-94 CHEs ended their first year of operations in 1993-94 with deficits of NZ$175 million and continued with deficits of up to NZ$201 mil- lion but achieved a reduction, to NZ$39 million, in 1998-99.35 The provisional figures for the first six months of 1999-00 showed a small surplus (table 8.3).36 In 1993-94, the government mandated the revenue levels. Rev- enue fell; costs continued to increase. The government centrally set aggregate CHE revenue levels again in 1994-95. Costs grew at 5.0 percent and revenue at 4.6 percent over the previous year. In 1995-96 and 1996-97, CHEs and the four RHAs had slightly more scope to negotiate contracts. During these years, costs grew around 5 percent a year; revenue increased between 4.6 percent and 6.8 per- cent a year. After undertaking an efficiency analysis of the CHEs in 1998, the government agreed to increase the Health Funding Authority's fund- ing by NZ$129.5 million to pay price increases to the CHEs. The CHEs' revenue increased 8.2 percent between 1997-98 and 1998-99, due partly to price increases and partly to differences in service levels and mixes. The trend in cost increases slowed, showing 326 o Innovations in Health Service Delivery Talbe 8.3 Crown Health Enterprises. Aggregate Trends in Costs and Deficits CROWN HEALTH ENTERPRISES 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99 Deficits (NZ$ million) -175 -194 -159 -201 -176 -39 Costs (NZ$ million) 2,657 2,790 2,932 3,102 3,180 3,291 Deficits as percentage of costs -6 6 -7 0 -5 4 -6.5 -5 5 -1 2 Percentage increase in costs over previous year 1.1 5 0 5 1 5 8 2 5 3 5 Percentage increase in revenue over the previous year -2.7 4.6 6.8 4.6 3.6 8 2 Percentoge change in consumer price index I 1 4.6 2.0 1.1 1.7 -0 4 Source New Zealand, Department of Statishcs Public Health Financial Statistics-Annual Operating Statements, 1998-99 figures are provisional, CPI tables June 1999. cost increases over the previous year of 2.5 percent for 1997-98 and 3.5 percent for 1998-99. To understand why costs increased more slowly would require detailed analysis. One causal factor may have come from the HFAs work with CHEs to carry out extensive effi- ciency studies of their services. This work was the basis for imple- menting a national purchasing framework that included national prices for services. AHB and CHE cost trends should be considered in relation to general cost trends. AHB costs fluctuated above and below the per- centage increase in the consumer price index. After 1992-93, CHE costs rose above the consumer price index in increasing amounts, reaching a difference of 4.7 percent in 1996-97. In 1997-98, the gap closed to a difference of 0.8 percent but again widened in 1998-99, to 3.9 percent. Cost trends should also be considered against output and com- plexity. This is difficult because of poor information on noninpatient services and changes in the recording of inpatient and day-patient services. There is a paucity of useful data to make comparisons over time. This is due in part to the variable purchasing frameworks used by the four RHAs before the creation of a national purchasing framework and the constant adaptations in these frameworks.37 Maladjustments in the Corporatization Model * 327 Firm baselines on output are hard to find, and rarely can quality data, collected against baselines, be located. At a high level of aggre- gation, CHEs increased hospital discharges by 12.6 percent between 1994-95 and 1996-97, an average increase of 4 percent each year.38 Total outpatient attendances increased by 3 percent between 1994-95 and 1995-96, but dropped back down to 1994-95 levels in 1996-97.39 Hospital discharges and outpatient services accounted for 52 percent of CHE revenue. For the inpatient services measured by discharges, a tentative observation can be made that workloads increased at a rate close to the increase in costs. This leaves output levels for about half of CHE services unexplained and, in these circumstances, no firm conclusions can be drawn. Under the AHBs, the average length of stay fell 18 percent be- tween 1989-90 and 1992-93 and by 13 percent between 1993-94 and 1996-97 under CHEs. CHEs continued the trend toward re- ducing average stays, but no comment can be made about the ade- quacy of the reduction, partly because of the increasing difficulty of reducing the average length of stay as it progressively shortens.40 Day surgery increased by 61 percent between 1989-90 and 1992-93 under AHBs and by 17 percent between 1993-94 and 1996-97 under CHEs. The same limitations on drawing conclusions from the available data also apply to day surgery.4" In terms of equity and quality, useful data on baselines and the measurement of gains are lacking. The RHAs did not have well-developed systems to spec- ify, measure, and monitor these factors. Subjectively, views have been expressed that gains were made in the governance and management systems in CHEs.42 Business cases were now expected for capital developments, and cost benefit analy- ses were required. Some of the individuals interviewed thought the CHEs understood their services and costs better than AHBs did and operated better management systems. The quality of information was considered to have improved. One CEO said that the costs of domestic services such as catering and cleaning were reduced but that the more difficult gains, associated with changing clinical prac- tice, were not widely achieved. 328 o Innovations m Health Service Delivery CHEs exited services where they could not compete with other providers-mainly residential care for the elderly and community services for mental health and disability clients. This movement began under the AHBs, but they could not have achieved the amount of change that occurred. This judgment is based on our experience in dealing with powerful resistance from hospitals, worried patients, and hostile communities when consulting them about moving ser- vices to nonhospital providers. AHBs would have been more exposed to these pressures than RHAs and CHEs, due to the influence of elected boards. CHEs lost a significant portion of the residential and community care business, but with the wisdom of hindsight, by 1999 they had little or no desire to be involved in these areas. Barriers to achieving gains in CHEs. The CHE chief executive officers interviewed for this report commented on the barriers they faced in improving efficiency. These included: o The power of the doctors to resist changes was significant.43 They were not easily replaced, and some had objectives that aligned with their private practices rather than the CHE's objectives- they had conflicting obligations to their professional bodies, pa- tients, and the CHE. Many managers were new to the health sec- tor and could not assess the demands and arguments of clinical staff. o Deficit financing gave CHE boards an excuse when costs ex- ceeded revenue and reduced the pressure to make hard decisions. o CHE boards were often reluctant to reduce services when they ex- ceeded the volumes set by the RHAs. o Whether deficits stemmed from inefficiencies or problems with RHA pricing, which varied around the country, was not clear. Prices were not linked to value, and the variations were not expli- cable on this basis. o Staff resisted changes, and managing change took considerable ef- fort. Sometimes retaining the status quo was easier. Maladjustments in the Corporatization Model * 329 * The public resisted changes, particularly around the centralization of services. Changes strained the CHEs' scarce management resources. * Politicians did not always support changes and at times inter- vened. * The RHAs did not provide strong incentives for change. Their contracting requirements were unsophisticated and their moni- toring, weak. Other factors affecting the ability of CHEs to make gains were their limited scope to expand their revenue, the requirement to maintain the same range and level of services in the initial year of the reforms, active political resistance to rationalizing services, and prob- lems with the balance sheet set-ups. CHEs debated balance sheet is- sues with central agencies over a two-year period. Some CHEs ex- pressed concern that inaccurate valuations placed them on an unfair footing compared with other CHEs. Allocative Efficiency There is no empirical analysis of the effects of the reforms on al- locative efficiency, although a range of policies was introduced to further this objective: * "Single-pipe" financing of primary, secondary, and tertiary care to allow resources to be shifted between these levels of care within an overall global budget limit * Allocation of funds to regions (RHAs) according to a needs- weighted, population-based formula * Specification of priorities for health improvement (such as Maori health) and of public health targets (such as immunization and cervical screening coverage targets) * Introduction of evidence-based clinical guidelines for a range of common conditions and services 330 o Innovations in Health Service Delivery o Introduction of booking systems, based on clinical guidelines for elective surgery o Reference pricing for prescription medicines within therapeutic groups, based on evidence of clinical value and cost-effectiveness. Some impacts of these initiatives on resource allocation can be identified, including: increased spending on Maori health programs, shifts in pharmaceutical expenditure, reallocation of rural health ex- penditure from inappropriate hospital services to community health services, and reduced regional disparities in expenditure. However, these positive changes were overshadowed by other factors. RHA boards were consumed with the problems of dealing with powerful provider interests that assumed, from experience, that public money would be available to them as a de facto monopoly service provider. Some ClEs also believed, from experience, that the government could not permit them to suffer serious financial loss and would res- cue them from bankruptcy. While the allocation methods of the RHAs and, later, the Health Funding Authority had considerable scope for development, work was done on using transparent and analytical approaches, including "program budgeting and marginal analysis." Formal rationing crite- ria, and a decision-making process for choosing services to fund, was part of the Health Funding Authority's early work. The biggest step in this direction was the introduction of a system for specifying max- imum waiting times for elective surgery, based on clinical assessment. Clinical Outcomes The link between the new system and its impact on clinical outcomes and health status has not been assessed. This would be difficult, con- sidering the many factors influencing health status besides the ef- ficiency and effectiveness of the organizational or institutional arrangements. In the areas of population health gain, it is also too early to iden- tify the results. It is reasonable to expect good results from contract- ing systems focused on targeting services to disadvantaged sectors of Maladjustments in the Corporatization Model * 331 the population and using local networks to reach people who have been underserved by the traditional system. Consumer Responsiveness and Satisfaction No systematic information on consumer responsiveness and satisfac- tion is available. Many community groups that were provided with resources and skills to deliver various primary care services say the reforms made new opportunities and services available to them. Pur- chase contracts did not emphasize consumer surveys, perhaps in part because rationing issues dominated dialogue with communities. Although the public no longer had access to locally elected rep- resentatives on health boards, they had a collection of substitutes. Government appointees on boards had to be responsive to their communities. Also, the central government established several special-purpose health policy and monitoring organizations for Maori health, mental health, and general health policy. A commis- sioner was set up to conduct inquires into service failures. Many local governments became involved in health service planning issues. In 1998 and 1999, several surveys of consumers and citizens were done. The results of a fragmentary study based on a telephone sur- vey of 1,000 people can be seen in table 8.4. Other surveys have had similar results.44 Respondents to a Ministry of Health survey said big changes were needed in the system (50 per- cent) and hospital waiting lists were a problem (47 percent). An HFA survey found that 55 percent of respondents rated the government's performance poor but that 75 percent of those who had used health services were satisfied. This pattern also appeared in the Ministry of Health survey. There is clearly a gap between what people think the services will deliver and what they do deliver, with dissatisfaction about the system in the abstract but greater satisfaction in practice. Concerns over the health sector have featured in national politics. Many of the debates have been about rationing issues such as ration- alizing small hospitals, waiting times for surgery, and limiting care in a few high-profile cases. A National Business Review poll before the November 1999 election showed 20 percent rating health as the 332 o Innovations in Health Service Delivery Talbife 8.4 Users' Views on New Zealand's Health Systems STATEMENT -AGREE- (PERCENr Some things are good but fundamental change is needed. 57 The system has so much wrong that it needs complete rebuilding. 32 The system works well and only needs minor change 9 The most important issues are waiting times for surgery and government funding levels.o 29b I cannot get care for financial reasons 25 Difficult to get care when needed 18 I am worried that I will not get advanced care if seriously ill 38 Note. This table reports the results of a telephone survey of 1,000 people as part of an International Health Policy Survey The study compares the results from Australia, Canada, New Zealand, the United Kingdom, and the United States a. The survey took place during the introduction of the guoranteed waiting times policy, and many people, whose need was seen as insufficient to be on a waiting list, had lust been referred back to their doctors. b. For both Source International Health Policy Survey (New York Commonwealth Fund), 1998 most important issue deciding their vote, ranking second behind the economy (25 percent). Equity of Access The reformed system had strong incentives toward equity of access in a number of dimensions. The Health Funding Authority, as a na- tional organization, had to establish national access criteria and made progress in several areas. The work on waiting times for surgery is a prime example, but there were other significant moves to even out inconsistencies that arose by having the four RHAs. A major prob- lem under the previous system was that the politics of the AHBs had sustained capacity for surgery in provincial areas beyond what could be justified by national criteria. The RHAs and the Health Funding Authority made considerable changes in this regard, reducing or eliminating purchasing of inappropriate services in smaller hospitals. Conclusions While there is a lack of comprehensive analysis on the successes and problems of New Zealand's health reforms, three problems stand Maladjustments in the Corporauzauon Model * 333 out. The reforms lacked a coherent policy framework due to last- minute changes in design. They were hampered by considerable funding problems that underpinned some of the implementation problems. Implementation was poor in many respects, with unsatis- factory ad hoc patches being applied to the system. Problems with policy, finances, and implementation are easy to identify, but changes in the quantity, quality, or efficiency of the CHEs are less easy to pin- point, given the lack of comparative data. In these circumstances, we must be cautious about drawing conclusions and lessons for other countries. The following points emerge from the material presented in this chapter. Other countries attempting such reforms should base their design on a comprehensive understanding of the situation in the health sys- tem as a whole, even if implementation processes are focused on par- ticular areas. In other words, systemwide thinking is imperative in countries like New Zealand, whose governments set ambitious and far-reaching goals for health policy. Implementation should be se- quenced and geared to practical capabilities. Ideally, reform should be driven by a policy framework that iden- tifies government's role, sets priorities, recognizes generally what in- stitutions are required to deliver, and establishes the necessary regu- latory environment. As health care is a complex system, this can be a process of discovery, retaining the basic framework of principles but adjusting the details as experience adds information on successes and failures. Reform objectives and barriers to reform will vary from country to country. Characterizing health reform internationally as having similar policy with local variations (as is the case with privatization policy) is likely to be very misleading. The details are the telling points. Superficially, New Zealand's reforms look like an application of its previous corporatization policy, but the details show that they departed from the critical principles for the success of that policy. Also, they did not entail moving from centrally managed to more autonomous hospitals. Though originally designed as a decentral- ization of health management, with the policy modifications, they became a recentralization in many respects. The AHBs already had considerable autonomy. The reforms focused mainly on moving ra- 334 o Innovations in Health Service Delivery tioning decisions away from hospital management to four indepen- dent government purchasing agencies, and later to one. There was a shift from funding hospitals based on populations served to ser- vice purchasing by government purchasers that received funding ac- cording to a broad demographic formula, applied to the population served by each purchaser. This amounted to increased central con- trol through: o The purchasing role of the RHAs and the Health Funding Au- thority o The influence of the monitoring agency and shareholding minis- ters on the CHEs' business plans and capital spending o The influence of the minister of health by setting goals for the RHAs and Health Funding Authority and advising these agencies of preferences. In New Zealand, wider objectives were rapidly subsumed by the goal of increasing efficiency, or achieving financial viability, in the public hospital sector. This was mainly in response to the financial problems built in when the reforms began. Countries seeking effi- ciency increases in a government-owned hospital sector may wish to note the following points: o Top-down restructuring should not be expected to produce the same gains in health as in other sectors. This probably stems from the predominance in the health sector of powerful provider and consumer interests and high political sensitivity to the life-and- death issues of health. The people who control the services at the customer end are not happy to have their processes reengineered from the top. o Fiscal pressures early in the reforms can make implementation difficult. Realistic assumptions have to be made about any effi- ciency gains that can be expected. o Staff members will resist change and can be in a powerful position to do so if they are in scarce supply and if they hold vital informa- Maladjustments in the Corporatization Model * 335 tion that management cannot access or properly analyze. Medical monopolies increase the power of clinicians to resist change when supply is restricted. * In the eyes of the public, it is difficult for managers and boards to bridge the credibility gap between themselves and the clinical pro- fessionals. Changes involving clinicians have to be well managed. * Boards (where they exist), management, and staff may line up with community wishes for increased services and ignore a government budget constraint. * The performance measurement system should be sufficiendy ac- curate to judge performance. Benchmarking can be a useful tool. Where services are hard to define, cost, and price, holding hospi- tals accountable for meeting their budgets and service require- ments will be problematic. * Performance commitments between government agencies can be weak unless they are credible and well founded initially and the parties' incentives are aligned. * The public may resist changes that seem rational to policymakers, such as closing clinically unsafe services in rural areas and central- izing them at another location. Changes have to be worked through with communities. * Politicians often have difficulty with detailed changes that are the consequence of their own broader policies and intervene because health is so politically sensitive. Countries attempting such reforms should also try to balance re- sources with the obligation to provide services at an aggregate level. Expectations about efficiency gains should be realistic. The financ- ing problems built in at the beginning of New Zealand's reforms re- sulted in a series of poorly conceived and managed interventions. These included ad hoc revenue injections to purchasers, equity in- jections for CHEs, waivers of CHE directors' exposure to liability, and centrally imposed solutions to contracting deadlocks. 336 ° Innovations in Health Service Delivery The politics of change must be understood and planned for. New Zealand's reforms were weak in managing this aspect. Any substan- tial health sector reform will most likely extend beyond the life of the government that starts it. Securing an accord may be possible across political parties on critical matters such as the government's role in the health sector and broad directions for the sector. New Zealand seems particularly prone to launching regular structural changes in the health system with the attendant potential damage to capability. Greater acceptance across the major political parties about what works and what is contentious would engender steadier progress. The Ministry of Health is developing a medium-term sector strat- egy, but this work is in its infancy. This could provide a backdrop for assessing proposed changes and may improve the quality of decision- making. Effective cooperation between government agencies on pol- icy development could also lift the quality of policymaking activities. Countries attempting such reforms should also try to link the pol- icy design to implementation realities. Policies should not demand a level of complexity or expertise that cannot be delivered in the time- frames being considered. This happened to New Zealand's policy of allowing private alternative health care plans to compete with the government's purchasing authorities. Policies have to recognize what capacity exists, what is required, and what can be achieved. When possible, policy design and implementation should identify the key sources of influence in the current system and address ways to engage them in change. If this is impossible, ways to manage their dissatisfaction should be identified. Problems associated with health sector monopolies, natural or otherwise, should be recognized in making changes to health sys- tems. Monopolies can be in the form of government roles in policy- making, purchasing, and provision. They can also exist in the private sector supply side (e.g., some clinical groups are monopolies). Mo- nopolization is usually extensive in health systems, particularly in small countries. The point at which performance would be enhanced by addressing monopolistic situations depends on circumstances in each case. Sometimes the cure can be worse than the disease. Gov- ernments should be careful not to make a bad situation worse Maladjustments in the Corporatization Model * 337 through poorly conceived and implemented regulatory interven- tions. Contrary to its original intentions, the New Zealand govern- ment managed to entrench in the system some of the monopoly characteristics of its own providers and purchasers through succes- sive policy modifications that were prompted by its dissatisfaction with performance. New Zealand's experience demonstrates the difficulties of using central requirements and monitoring to steer performance. It also shows the complexity of marrying a model that emphasizes choice or competition to a system where the government takes responsibility for a comprehensive national health service. The internal markets needed to do this are difficult to operate. If government-owned hospitals are part of a country's health care model, the government probably has social objectives for them. The extent of their autonomy has to be systematically integrated with re- quirements that they adhere to government policy. The absence of simple financial discipline can result in the worst of both worlds- financial losses and unresponsiveness to government policies. There should be sectorwide systems to manage the government's interests as the hospitals' owner. Which system is best depends on the details of the situation. New Zealand's experience suggests that "halfway houses" between rational, disciplined central controls and decentralized systems, backed by strong incentives, can be a source of confusion and performance failure. New Zealand's experience shows how difficult it is to build the organizations and policy and management frameworks that ensure high performance from inde- pendent agencies responsible for multiple and often conflicting goals. This quest must continue, however, if the marriage of central policies and principles with local control and initiative is to succeed. It requires an advanced practice of public management techniques that is likely to be difficult for many developing countries to achieve. For example, contracts between a national purchasing agency and in- tegrated care organizations are technically complex, and their man- agement demands extensive statistical information and risk analysis. A system where government dominates the roles of the purchaser and provider needs rules on the disclosure of information and proto- 338 o Innovations in Health Servce Delivery cols for the agencies to work together on critical issues. In this situ- ation, the normal conventions of commercial life regarding confi- dentiality of information are irrelevant. New Zealand has had prob- lems with purchasers having poor information on services and costs and with CHEs planning their capital and service developments in isolation from purchasers. The accountability frameworks developed for government-con- trolled agencies should align internal and external accountabilities. The Heath Funding Authority used a strategic business-planning process to drive planning and management activities. This process underpinned the external and internal accountabilities at every level. It showed great promise as a way to significantly improve the man- agement of public resources. Finally, the impact of reform in the hospital sector cannot be judged by looking only at the hospitals' legal, policy, and manage- ment frameworks. New Zealand's story shows that hospital perfor- mance was very much affected by the structures and operations of the purchasing function and the influences of the political system. No simple associations can be drawn either about hospital perfor- mance and the degree of decentralization or about forms of gover- nance. Only by taking a systemwide view can the agencies' behaviors, achievements, and failures begin to be understood. PosiscripH-September 2002 This chapter was written in 1999. The November 1999 elections re- sulted in the replacement of the architects of the 1993 reforms, the national government, with a Labour-Alliance coalition. The new govermment disbanded the health purchaser (Health Funding Au- thority) and set up district health boards to replace the CHEs. These boards have the joint functions of purchasing and providing services, much as did the area health boards of the 1980s. This restructuring is the fifth significant restructuring in the health sector in 15 years. New Zealand has gone from elected hospi- tal boards, to 14 elected area health boards, to 4 purchasers and 23 CHEs providing hospital services, to 1 purchaser and 22 CHEs. Maladjustments in the Corporatizatnon Model * 339 New Zealand now has a conglomerate Ministry of Health and 21 district health boards. Some fiscal impacts are emerging with a col- lective deficit for the district health boards of over NZ$200M being recorded for the 2001/02 year, despite increases in the level of gov- ernment funding.45 Vhile aggregate fiscal impacts can be assessed, there is a lack of comprehensive analytical information on the in- tended costs and benefits of these latest reforms. Also the analysis of previous reforms has been insufficient in many respects. Most new systems did not have time to develop before the next wave of reform. There is an opportunity now to benchmark what is provided for what cost under the current system and, at least for the CHE inpa- tient services in the latter years of the existence of the CHEs, what was provided for what cost under the previous system. Consideration could also be given to levels of public and consumer satisfaction and health status indicators for different populations. This work could provide some critical information for New Zealand and other coun- tries in assessing the value of alternative health systems in the future. Authors' note: The views expressed are the views of the authors, except where at- tributed to others. While care has been taken in the preparation of this chapter, the authors are not responsible for the results of any act or omission done or omitted in reliance in whole or in part on this report, nor for any error or omission from the chapter. Acknowledgments. The authors would like to acknowledge the helpful comments on an earlier draft from Loraine Hawkins and Chns Lovelace of the World Bank and the editorial assistance provided by Loraine Hawkins. Helpful comments were also provided by Kathy Spencer from the Office of the Minister of Health in New Zealand. The authors also wish to thank Mani Maniparathy for assistance with the financial information on Crown Health Enterprises. Notes 1. Ministry of Health, Health Expenditure Trends in New Zealand 1980-1998, Wellington, 1999. The figures quoted are for 1997-98. 2. Ibid., p. 44. Note that there are problems with comparing health expenditure between countries, and these figures must be used with caution. 340 o Innovations in Health Service Dehvery 3. These are 1999-2000 figures, based on the Health Funding Authority's Funding Agreement with the government. Note that Crown Health Enterprises was renamed Hospital and Health Ser- vices in a 1998 amendment to legislation, but to avoid confusion, we refer only to Crown Health Enterprises. Crown Health Enterprises is a group of companies owned by the state; the government minis- ters are the shareholders. 4. "The Crown" refers to the state, which exercises its ownership through the government in the form of the "Executive." 5. Interviews were held with: CHE CEOs (3); a CHE board mem- ber (1); former AHB CEOs (2); a former minister of health (1); heads and members of implementation agencies (5), including reform agencies (National Interim Provider Board, Health Reforms Direc- torate, Communications Unit); taskforce members (2); an academic involved in health sector (1); a director general of health and a for- mer director general of health (2); a senior manager in the former Department of Health (1); Treasury staff (2); former SSC staff in- volved in the reforms (2); a former CCMAU staff member (1); de- partment of prime minister and cabinet staff and former staff (2); managers of private hospitals and health insurance companies (3); former AHB board members (2); former RHA board members (2); Health Funding Authority board members (2); and consultants in- volved in the reforms (2). Note: Some people appear in more than one category. 6. Minister of Health, "Your Health and the Public Health," Wel- lington, 1991. 7. Ibid. 8. There were numerous instances of ministers and their min- istries influencing purchasing decisions. For example, in the early years, RHAs had to comply with detailed policy guidelines on pur- chasing. Ministers set aggregate revenue levels for hospital services in the first two years. In 1998, the minister of health issued a plan for hospital services throughout the country. Maladjustments in the Corporatizafion Model * 341 9. The Otago CHE was planning to provide services for Saudi Arabia but was prevented from proceeding by the government. 10. The ownership interest covers interest in the organization's vi- ability. It comprises the management of physical and human assets, the robustness of vital systems such as the financial, planning, risk management systems, and return on capital. 11. Between 1994 and 1997, CHE directors were appointed by and reported to the munisters of CHEs and finance. The minister of CHEs was separate from the minister of health, to whom the RHAs reported, reinforcing the purchaser-provider split. From 1997 on, the ministers of health and finance held the power to appoint and control the CHE directors, and the minister of health exercised those powers alone in relation to the Health Funding Authority. 12. Controller and Auditor General, "Second Report for 1998," Wellington, 1998, p. 23. 13. Other funding is sourced from Accident Rehabilitation Com- pensation Insurance, private donations, supply of services to private entities, interest, and user fees. Controller and Auditor General, "Second Report for 1998," p. 26. 14. National Interim Provider Board, Providing Better Health Care for New Zealanders: Report to the Government and to the Public, Welling- ton, 1992. 15. Officials involved in the policy development expressed the view that CHE boards were expected to initiate takeovers and merg- ers. Interviews, July to September 1998. 16. Formal procedures include requirements to demonstrate net gain over 10 years and no danger to the viability of remaining services. 17. The Health and Disability Services Act empowered the RHAs to issue terms and conditions deemed to be accepted by a provider if the provider accepted payment. 342 o Innovations in Health Service Delivery 18. Separating the service and facility components might have promoted competition for services by permitting groups of providers to access essential facilities, thereby lowering entry barriers. 19. The development of integrated care organizations with capi- tated budgets might have strengthened incentives to select the best care within financial limits. 20. Ministry of Health, Health Expenditure Trends in Ne7w Zealand 1980-97, p. 12. 21. Ministry of Health, Ne-w Zealand Hospital Sector Performance, Wellington, 1993, p. 18. Note that consumer price index deflated an- nual percentage changes in revenue varied from -5.4 percent to 16.2 percent in any one year, while the consumer price index deflated an- nual average percentage change for the decade was 0. 22. Deficit and cost calculations include depreciation and interest costs but exclude extraordinaries. Deloitte Touche Tohmatsu, "Trends in Area Health Board/CHE Perfornance 1989/90-1997/98," report for CCMAU, Wellington, 1996. 23. New valuations of CHE fixed assets led to a write-off of NZ$993 million (30 percent of CHE assets). Further adjustments in 1993-94 led to another NZ$457 million write-off, a combined total write-off of 47 percent of AHB asset value at the start of 1992-93. There was also a write-off of NZ$460 million in debt. 24. Deloitte Touche Tohmatsu, "Trends in Area Health Board/ CHE Performance." 25. L. Hawkins, "CHE Shareholder and Banker Roles and Link- ages to the Role of the Funder and of Contracts." Report prepared for the Central Regional Health Authority, Wellington, 1997. 26. Controller and Auditor-General, "Second Report for 1998." 27. Ibid., p. 17. 28. An example is the Hutt Valley CHE, which ceased elective surgery for a short period in 1995-96 before being told to resume these services, despite exceeding their contracted volumes. Malad)ustments in the Corporauzation Model * 343 29. An example of this is the mtroduction by the minister of health of "exit protocols," requiring CHIEs to provide six month's exit notice. 30. The CHEs advised the Health Funding Authority in the 1998-99 contract round that they had extensive hidden costs in the form of deferred maintenance and capital development. One CEO estimated this at more than $1 billion for all CHEs. In interviews, views were expressed that investment in staff training and develop- ment was insufficient. 31. The 1999 Labour-Alliance Government created district health boards to replace the CHEs and the Health Funding Authority- thereby combining the purchasing and providing roles, similar to the AHBs. 32. Deloitte Touche Tohmatsu, "Trends in Area Health Board/ CHE Performance," p. 7. 33. Ministry of Health, New Zealand Hospital Sector Performance 1983-1992, Wellington, 1993. This analysis was based on the limited information available at the time. 34. Reports reviewed included the Ministry of Health, Purchasing for Your Health reports, Wellington, and Deloitte Touche Tohmatsu, "Trends in Area Health Board/CHE Performance." 35. Provisional figures; audited deficit likely to be NZ$56 million. 36. Department of Statistics, Public Health Financial Statistics: Quarterly Results for December 1999, Wellington, 1999. 37. Further studies could consider output levels achieved related to CHE costs for 1998-99 and subsequent years. The "data envel- opment analysis" by the Health Funding Authority and CHEs could provide useful information on service volumes, costs, and efficiency levels. The 1999 Labour-Alliance Government set up a new system of funding based on population characteristics. Tracking relation- ships between costs and outputs will become increasingly difficult in the future. 38. Ministry of Health, Purchasing for Your Health 1996-97, Wellington, 1998. Discharge numbers have been weighted to adjust 344 o Innovations in Health Service Delivery for case mix and complexity of treatments. These figures are not equivalent to raw throughput. 39. Ministry of Health, Health Expenditure Trends in Nezw Zealand 1980-97. 40. Data from Ministry of Health, Purchasing for Your Health 1996-97, Wellington, 1998. 41. Ibid. 42. C. Clayton, "Review of Managed Care." Report for central RHA. Wellington, 1996. 43. For an analysis of the authority of doctors in New Zealand hospitals, see R. Vaithianathan, "The Failure of Corporatization: Public Hospitals in New Zealand," Agenda 6 (4): 325-38 (1999). 44. Ministry of Health, "1998 Survey, Health Funding Authority," Wellington, 1998. 45. The hospital and health service provider function of the dis- trict health boards recorded an annual provisional operating deficit of $217.0 million for the year ended June 30, 2002. Provisional re- sults published by Statistics New Zealand, Public Health Financial Statistics, June Quarter 2002. Autonomous Hospitals Become a Commercial Network: Hospital Rationalization in Victoria, Australia Simon Corden In 1995, Victoria's Kennett Government restructured metropolitan health care services, combining 32 independent, publicly owned and operated hospitals in Melbourne into 7 health care networks: 6 re- gionally based and 1 specialist network. Three legally independent public hospitals operated by the Catholic Church were also incorpo- rated into the networks for planning and funding purposes.1 Net- works changed the way health services were delivered in Melbourne, capital of the state of Victoria and Australia's second largest city. Acute care hospitals were combined with other acute care hospitals and with non-acute-care providers. Governance arrangements simi- lar to those in commercial organizations were introduced for metro- politan hospitals.2 Networks were part of a larger package of changes designed to im- prove the efficiency and effectiveness of the delivery of government health services. The other institutional changes were specific to health care-in the two years before the creation of networks, the state health budget had been cut significantly and output-based/case-mix funding of inpatient acute services introduced (box 9.1). Other 345 346 - Innovations in Health Service Delivery Bon 9.1 The vicior;iw Healb Syslem Health services in Australia are delivered by both the public and private sectors under complex Commonwealth/state funding arrangements. State governments are responsible for providing public hospital services, mental health programs, home and community care, and health promotion. The Commonwealth government shares funding responsibility for most state services and subsidizes most private general practice and specialist medical and surgical services. Local governments provide some community health services. By the mid-1990s, metropolitan Melbourne had 35 hospi- tals, including 6 tertiary acute care (or teaching) hospitals to serve 3 million people. Almost a third of the state's acute care hospital activity occurred within 5 kilometers of the city cen- ter, while services in the growing outer suburbs were under strain. Moving or reconfiguring services had proved very dif- ficult, although during the 1980s a Labor Government had relocated 2 hospitals from the inner city, despite strong re- sistance. Other changes during the 1980s included moving the delivery of general acute psychiatric services from spe- cialist institutions to general hospitals ("mainstreaming"). In the early 1990s, two influential reports commissioned by the government concluded that Victoria's hospitals were much more expensive to operate than their counterparts in New South Wales and Queensland. In 1993-94, the Ken- nett Government introduced output-based (or case-mix) funding of acute care hospital services and large budget cuts to restore the state's finances. By 1995-96, Victorian inpa- tient costs were among the lowest of all states. Sources: MHPB, Developing Melbou7ne ' Hospital Network: In- terim Phase 1 Report, Melbourne, April 1995; D. Nicholls, Independent Review of Victoria's Public Sector Finances, Victo- rian Govermnent, Melbourne, 1992; Victorian Commission of Audit, Report of the Victorian Commission of Audit: Volume Two, Melbourne, May 1993. Autonomous Hospitals Become a Commercial Network * 347 changes were either more general (e.g., changes in industrial rela- tions) or occurred at the same time as the creation of the networks (e.g., government commitment to competitive tendering of interme- diate services and whole institutions). The introduction of networks (in conjunction with the other de- velopments) changed health care managers' incentives. Further, they changed the ability of managers to act on these new incentives. We describe these changes but do not try to evaluate or quantify their ef- fects individually. Because no independent analysis had been done on network introduction and the quantitative data were poor, we relied primarily on documents published by networks and the Department of Human Services (DHS), and on discussions with departmental and network staff, board members, academics, clinicians, and advo- cacy groups. An informed, though subjective, perspective on the effect of the changes in publicly delivered health care was obtained through a structured questionnaire, completed by 13 current board members from across the networks who had previously served on hospital boards of management.3 This survey provided useful, though still anecdotal, information. Objectives and Timelines of Policy Changes The Kennett Government set out to restructure Melbourne's health care services to improve: * Quality, with closer client focus overall and improved continuity of care * Efficiency, partly by delivering services in more cost-effective settings * Accessibility, particularly for people in the middle and outer sub- urbs of Melbourne.4 In 1995, an independent Metropolitan Hospitals Planning Board (the "Planning Board") was created to determine the number and 348 o Innovations in Health Service Delivery TaIOo 9.11 Timeline of Policy Changes DATE EVENT October 1992 Kennett Government is elected with large malority in both houses of Parliament May 1993 Victorian Commission of Audit recommends rationalization of hospital services through joint or network arrangements between hospitals wherever it is cost-effective. June 1993 Case-mix funding is introduced for acute services provided by pub- lic hospitals, and real recurrent funding is cut February 1995 Newly formed Planning Board calls for submissions from interested parties. April 1995 Planning Board releases Interim Phase I Report Bonus payments are introduced for hospitals meeting specified emergency department targets June 1995 Health Service (Metropolitan Hospitals) Bill enables implementa- tion of Planning Board's final recommendations passed by State Parliament Planning Board releases final Phase 1 report, recommending six geographically based and one specialist women's and children's network. August 1995 Networks are formed by administrative order, aggregating exist- ing hospitals December 1995 Planning Board releases Phase 2 Report July 1996 Government merges two geographically based networks Decision is made to competitively tender for design, finance, con- struction, and operation of two new tertiary public hospitals October 1996 Government releases A Healthier Future. A Plon for Metropolitan Health Core Services, outlining a 10-year plan for capital works and proposed hospital closures and reconfiguration of services November 1997 Malor tertiary hospital is removed from one network, preparatory to contracting out its reconstruction and operation, and remainder of the network is merged with another network Sources Victorian Commission of Audit, Report of the Victorian Commission of Audit Vol- ume Two, Melbourne, May 1993, Victorian Auditor General's Office, Acute Health Ser- vices under Case Mix A Case of Mixed Priorities Melbourne, May 1998, MHPB (Metropolitan Hospitals Planning Board), Taking Melbourne's Health Care Networks into the 21st Century Phase 2 Report, Melbourne, December 1995, MHPB, Developing Mel- bourne's Hospital Network Phose 1 Report, Melbourne, June 1995, MHPB, Developing Melbourne's Hospital Network Interim Phase I Report, Melbourne, April 1995, DHS, A Healthier Future A Plan for Metropolitan Health Core Services, Melbourne, 1996, and NWH, North Western Health Annual Report 1997-98, Melbourne, 1998 composition of networks (table 9.1). When setting up the Planning Board, the Kennett Government noted: The existing pattern of hospital locations, functions, and gover- nance is largely a result of historical and political evolution rather Autonomous Hospitals Become a Commercial Network * 349 than systematic planning. Individual Victorian public hospitals are separate bodies established by ad-hoc and localized initiatives over the years, each governed by a Board of Management.5 Most network operations were concentrated around metropolitan Melbourne. The catchment area for all but one nonspecialist net- work extended from inner Melbourne (where most of the tertiary/ teaching acute hospitals are located) to the outer suburbs (where there are mostly smaller hospitals and other nonacute care institu- tions). Patients could use any network's providers. Most rural hospi- tals remained stand-alone institutions with local boards (aged and acute care institutions were merged in three regional centers). The DHS remained responsible for meeting the state's health needs, effectively functioning as a monopoly purchaser of services, with the networks providing a specified quantity of services (or out- puts). Each network, though monitored by the DHS, was responsi- ble for its own financial viability. Design Issues The key design issues discussed below are the governance arrange- ments, the provider payment systems, the market environment, and the arrangements for community input. WVhile each of these issues is examined in turn, it is clear that each element affected the others. Governance Arrangements With network introduction (and changes in funding, monitoring, and other arrangements) the new boards were given clearer objec- tives, more specific performance measures, and minimum targets. These directives were intended to bring a more commercial ap- proach to decisionmaking, slmilar to changes applied to many Aus- tralian public enterprises in the 1980s. However, boards and their managers were not given a single mea- surable goal such as maximizing the organization's long-term wealth. Publicly operated health care providers are inevitably different from corporatized public enterprises, which sell their services commer- 350 o Innovations in Health Service Delivery cially, in that health care providers produce mainly "community service obligations." These obligations require providers to deliver services that could not or would not be offered commercially. Con- sequendy, the networks still faced multiple, potentially conflicting, objectives related to enhancing efficiency, quality, and access. The changing role of independent boards. Victorian hospitals had been incorporated and independendy run for more than a century (box 9.2). This system had generally served the community well but had its critics. In an editorial on the release of the Planning Board's In- tenim Phase 1 Report, The Age newspaper noted that the hospitals had become "self-serving and did not adapt well to the changing shape and needs of the city they were supposed to serve ... more remote from the community they are supposed to serve."6 With network introduction, three key changes in governance arrangements were designed to encourage a more responsive and strategic approach to health care delivery. First, boards had an ex- plicit governance (or oversight) role rather than a managerial role. Second, the expectation that boards would meet agreed budgets was made explicit. Third, board members were selected for their skills and decision-making expertise, rather than primarily to represent different stakeholders. The explicit oversight role of the network boards was reflected in the differences between the parts of the legislation applying to the metropolitan networks and those applying to other, mainly country, public hospitals. Under the Health Services Act, public hospitals had "Boards of Management" to "oversee and manage the hospital" and to ensure compliance of the hospital's services with the Act and "the objects of the hospital."7 In contrast, the title of network boards as "Boards of Directors" (the usual tide in Australia for a private corporation's board) was intended to reflect their different role. The related legislative provi- sions were also intended to bring a more commercial approach to governance, including the first specific requirement to budget for long-term financial viability. This was reflected in the introduction Autonomous Hospitals Become a Commercial Network o 351 Box 9.2 The Changing Governance of Public Hospitals in Victoria Victoria's first hospitals were set up by the colonial govern- ment. They took care of convicts and the niilitary, for whom the government was considered responsible. Free settlers took care of their own health. The government encouraged the creation of charitable institutions for the poor, funded by subscription. However, hospitals were partially government-subsidized philanthropic enterprises, seeking and receiving land and some financial help from the government. Hospitals also re- lied on the voluntary labor of honorary medical officers and their students. Hospitals had a major teaching role and, working with the university medical schools and the spe- cialists' colleges, pardy organized hospital inpatient and outpatient services to help junior doctors gain experience. The hospital boards (or governing committees) and hon- orary medical officers were elected by contributors, who performed a similar function to shareholders of a private company. In 1864, the state government passed an Act for Hospitals and Charitable Institutions to allow hospitals to incorporate and thus engage in activities in the name of the organization. Until 1988, the regulatory regime for hospi- tals was part of the broader regime to regulate charities rather than government bodies. Even today, they are legally separate from the government, unlike the Department of Human Services. Public hospitals have become more closely linked to the government, both financially and institutionally. By the late 1950s, donations accounted for only 2 percent of income, as rising costs led to more government funding. The Hospitals and Charities Act of 1958 enabled the Charities Board to (Box contines on the following page) 352 o Innovations in Health Service Delivery Bon 9.2 (clinued) recommend amalgamation of subsidized institutions, and the Health Services Act of 1988 made their transition to statutory corporate bodies almost complete, with board ap- pointments subject to the recommendation of the state's health minister. Sources: E. Adeney, "The Invisible Reconstruction of the Victorian Metropolitan Hospital System," Journal of Law and Medicine 4 (3): 267-75 (February 1997); V. Lin and S. Duckett, "Structural Interests and Organizational Di- mensions of Health System Reform, in H. Gardner, ed., Health Policy in Australia (Melbourne: Oxford University Press, 1997); MHPB, Taking Melbourne's Health Care Net- works into the 21st Centu?y: Phase 2 Report, Melbourne, De- cember 1995. of payments for network board members. The 1995 amendments to the Health Services Act of 1988 stated that, instead of managing the network (made up of several campuses of what were previously inde- pendent hospitals), the functions of the board of a network were: (a) to establish the objects of the [network]; (b) to establish the organizational structure of the [network]; (c) to appoint a person to fill a vacancy in the position of chief executive officer; (d) to appoint senior management in consultation with the chief executive officer; (e) to oversee the management of the [network] by the chief ex- ecutive officer; (f) to develop a business plan for the operation of the [network]; (g) to develop plans, strategies and budgets to ensure the provi- sion of health services by the [network] and the long-term financial viability of the [network]; Autonomous Hospitals Become a Commercial Network * 353 (h) to monitor the performance of the [network]; and (i) to monitor the performance of the chief executive officer of the [network].8 The Planning Board suggested that, if network boards were to perform these duties, their membership should consist of people with a mix of skills, including finance, law, management, human re- source management, and research/academia. "Health care industry (nonstaff representation)" was also sought, but not specifically those with clinical expertise.9 The first group of network board members and chief executive of- ficers was selected differently from most of their predecessors on hospital boards, who were often nominated by the existing board or hospital management. From a group identified by using a commer- cial search contract based on criteria developed by the Planning Board, the government selected and approached candidates about the positions.'0 The act allowed the government complete discretion in the appointments and reappointments. The Planning Board stressed that the new boards should play a strategic role. From some anecdotal evidence, this occurred. All but one network board member who completed the questionnaire thought board focus on strategic issues had increased, and all be- lieved that focus had increased on alternative care settings and on shifting services to the most appropriate locations. They saw these changes as positive developments. Clarity about responsibilities. The governance arrangements made the health minister effectively the networks' sole shareholder (on behalf of government) and their key client (recognizing patients as the ulti- mate consumers). The minister could direct both network and rural hospital boards, intervening, if necessary, in politically sensitive or strategic matters (e.g, industrial relations). This relationship was reflected in the DHS's close monitoring of each network's financial viability.1" The Planning Board argued that network creation required a new relationship between providers and the DHS, based on a "clear sep- 354 o Innovations in Health Service Delivery aration" of the DHS from hospital management and of the roles of purchaser and provider of health services. "Boards need to be given clearly defined independence for operational and strategic decision," the Planning Board argued.12 The legislation creating networks gave board members some statutory independence by allowing their removal only if they were mentally or physically incompetent, convicted of a crime, not at- tending board meetings, or bankrupt. No similar legislative con- straints applied to members of the rural and former metropolitan hospital boards (or, as the government noted, to board members of state-owned business corporations). Most constraints were removed by amendments to the legislation in 1998, with the government arguing: "[A]s a rule acts confer a gen- eral power to remove directors or members of boards of public bod- ies. This is appropriate to ensure adequate accountability by direc- tors to government and the community."13 The Parliamentary Opposition focused on the effect on board in- dependence, noting that removal powers went beyond merely dis- missing a member for poor performance of duties. "[T]he legislation could be used to force [network] boards simply to toe the political line, rather than providing the best service for patients, which is what their jobs are meant to be."14 Twelve of the network board members surveyed thought that de- partmental influence over board decisions had been reduced or was unchanged since the creation of networks. Five thought the level of indirect political influence had increased; three thought it had de- clined. Compared to other hospital reforms, it appears that the Vic- toria reforms were relatively successful at reducing the politicization of decisions related to hospitals. Accountabilities were affected by the nature of the Health Service Agreements between networks and departments.15 The agreements did not specify the case mix of services to be provided, but still pre- clude networks' reconfiguring their services unilaterally. DHS's Public Hospitals Policy and Funding Guidelines 1999-2000 stated that funding is to be based on continuance of "the current range of ser- vices" and that "the planning implications" of any "significant change Autonomous Hospitals Become a Commercial Network * 355 in the range or scope of services ... must be discussed and agreed with DHS."'16 In addition, the DHS continued to oversee networks' decisions in two key areas: capital investment and terms and conditions of staff. * Capital investment. The network boards' latitude for making strate- gic capital decisions, or even to substitute capital for labor or other inputs, remained somewhat constrained. The prices paid for outputs through annual funding (largely uniform across hospitals) were not intended to cover depreciation or any return on assets.17 Instead, separate streams of funding to cover capital costs, and annual capital grants, treated as capital injections, were provided to cover depreci- ation expenses. To borrow money, networks needed the approval of the health minister and the state treasurer. According to the Inner and Eastern Health Care Network, approval, particularly from the treasurer, was difficult to obtain.18 There were some moves to incorporate capital into the output- based funding arrangements. Grants to replace plant and equipment became largely output-based. Decisions on plant and equipment, an estimated half of the total depreciation expense, were made at the network level and increased the scope to substitute between capital and other inputs. Grants for major capital works, including replacing buildings and expanding capacity, remained submission-based and were evaluated centrally by the government. But, as Lin and Duckett noted: [T]he question remains unresolved about the balance between autonomous investment decisions by the hospitals [or networks], either in capital infrastructure or service re-configuration, and the macro resource allocation and planning decisions by gov- ernment on behalf of the community. 19 * Staff conditions. In line with the trend across the Australian econ- omy away from industry-wide bargaining, netwvorks negotiated staff terms and conditions, as the Kennett Government had devolved this responsibility from the DHS to hospitals in 1994. The DHS re- mained somewhat at arm's length from the detailed negotiations but 356 o Innovations in Health Service Delivery vetted agreements, before signing, for consistency with the Kennett Government's wages policy and ability to cover any resulting unit cost increases. Some individuals involved with the reforms have ar- gued that the government continued to exert significant implicit pressure over labor negotiations. Overall, achieving clear accountability between a network board and the DHS posed challenges. Still, 11 of the 13 network board members who responded to the questionnaire considered the role of the board and individual members clearer since the network cre- ation, and 10 thought this development positive. Provider Payment System DHS funding agreements were with each network, not individual hospitals. The nature of the funding system for health care providers and the role of their boards have been closely related. Each hospital used to negotiate with the DHS, and board members' political influ- ence was important in the competition for government funding. Walker noted that, historically, the most successful hospitals in this competition "had the best relationship between their committees of management, the government and the charitable trusts."20 Similarly, Swerissen and Duckett noted, "Traditionally, the basis for public hospital funding from state governments has been obscure: a hospi- tal's budget reflecting a combination of history, negotiating skill, po- litical influence and luck.",21 However, this began to change in the 1980s, as governments sought to create incentives for improved performance within the fund- ing arrangements. Specifiing output. In the mid-1980s, Health Service Agreements with each hospital were introduced, replacing detailed input controls (staffing size and composition determined by the department) with a single global budget and some broadly specified expectations of hos- pitals. But, as Parliament's Economic and Budget Review Commit- tee noted: "Health Service Agreements do not provide an explicit link between output and funding. . . . no reward for improving Autonomous Hospitals Become a Commercial Network * 357 efficiency ... no funding mechanism to ensure that [more efficient] organizations ... are rewarded for that efficiency.22 Overall, the committee found large variations in the costs of sim- ilar hospitals and "little tangible evidence to indicate that [hospital service agreements] have tackled the problems of discrepancies in hospital performance ... [they] have not achieved a significant move from historical patterns of funding."23 The basis of funding negotiations (and with it the importance of board members' political influence) changed in 1993 when the Ken- nett Government introduced a more transparent and objective sys- tem of output (or case mix) based on funding. Under this system, acute care inpatients were classified into diagnosis-related groups (DRGs), with standard weights (essentially prices) assigned to each DRG. Hospitals and networks were paid for the number of cases treated in each DRG. This approach was extended to the outpatient services of the network's major metropolitan hospitals in July 1997. In April 1995, the DHS introduced waiting-time targets and per- formance bonuses with the Emergency Services Enhancement Pro- gram. A 1999 study reported a sustained reduction in recorded wait- ing times for emergency treatment and occasions of ambulance bypass of emergency departments since 1995.24 The study empha- sized the importance of the incentives in the funding mechanisms but did not examine the influence (if any) of network establishment on performance. Less complex output measures were used for rehabilitation, aged care, and mental health. The development work had led to an ac- cepted means of objectively categorizing the expected cost of each case. Large variations in the actual cost of cases with the same diag- nosis frustrated attempts to develop a basis for setting an agreed price. As a result, simple output measures were used, including the number of separations, bed-days, registered clients, and client contacts. The introduction of case-mix funding for acute care services cre- ated strong incentives to contain unit costs and to produce a speci- fied quantity of outputs. The government's funding formula pro- gressively linked funding directly to output. Funding for acute care services represented about three-quarters of network revenues and, 358 o Innovations in Health Service Delivery under the 1999-2000 guidelines, about 95 percent of that was closely related to output.25 Changes in performance monitoring. The Planning Board envisaged that the new governance regime would decrease focus on detailed monitoring of inputs and processes and increase focus on outcomes. But, in practice, the DHS kept a close watch on the networks. The Victorian Auditor General's Office noted: [T]he Department actively monitors hospital financial perfor- mance through a framework which includes: o monthly electronic reporting to the Department of financial and performance data by each network and hospital; o analysis by the Department of defined qualitative and quanti- tative indicators against benchmarks; and o regular discussions between the Department and individual hospitals on budget and perfornance issues.26 Two key questions relate to changed monitoring under the net- work system: the extent of changes and the causes. o Changes. Was DHS monitoring of providers streamlined or merely changed in form? There seems broad agreement that, since the 1980s, the DHS gradually moved from detailed controls of inputs (while still actively monitoring financial aggregates) to more active monitoring of output performance measures (table 9.2). o Causes. To what extent do changes in the intensity of monitoring reflect factors other than network introduction (e.g., the implemen- tation and refinement of case-mix funding and increased knowledge about measuring performance)? There seems plenty of evidence that other factors were more important. In addition to increasing efficiency, another important objective of case-mix funding was to develop a system free from centralized bu- reaucratic control, and thereby promote hospital autonomy.27 More- over, at a national level, there has been a major focus on improving performance information in health, although progress has been slow (as indeed many argue it has been in other countries). Autonomous Hospitals Become a Commercial Network * 359 Table 9.2 Acute Health Care Services Access Indicators PERFORMANCE iNDICATORS REPORTING Elective surgerya Proportion of Category 1 patients admitted within 30 days D, AR, M, Proportion of Category 2 patients admitted within 90 days D, AR, M Average waiting times for Category 2 patients on the waiting list at year's end D, AR, M Average waiting times for Category 3 patients on the waiting list at year's end D, AR, M Emergency Deportment Performonce Proportion of patients needing resuscitation receiving immediate attention D, AR, M, H Proportion of emergency patients receiving attention within 10 minutes D, AR, M, H Proportion of urgent patients receiving attention within 30 minutes D, AR, M, H Patients staying in the Emergency Department for more than 12 hours before admission to an inpatient bed (number) D, AR, M, H Ambulance bypass (periods of two hours or less, number of occasions) D, AR, M, H D Monthly reporting to the department, AR Reported in network annual reports, M Required under the Medicare agreement on Commonwealth/state funding of health care, H Reported quarterly in Hospitol Services Report, Department of Human Services a Category 1 and 2 elective surgery patients are those for whom admission within 30 and 90 days, respectively, is considered clinically desirable Category 3 patients have a condition that causes minimal pain, dysfunction, or disability, and is unlikely to deteriorate quickly Sources DHS, Financia/ Management Act Annual Reporting Requirements Pub/ic Hospi- tals 1998/99, Melbourne, 1999, S Duckett, L Hunter, and A. Rassaby, Health Services Policy Review Discussion Paper, Department of Human Services, Melbourne, March 1999 The finding arrangements between state and Commonwealth governments also affected monitoring at the hospital level. The Vic- torian government needed information from hospitals to qualify in 1995-96 for almost AUD$50 million (US$38 million) bonus pay- ments from the Commonwealth. The Victorian government had to demonstrate that it was meeting targets for inpatient activity (based on case mix), outpatient activity, and waiting-time targets for elective surgery and emergency departments. An additional AUD$3.5 mil- lion was made available for further data development.28 Quality indicators were still relatively undeveloped, while access indicators included waiting times for elective surgery and emergency 360 o Innovations in Health Service Delivery treatment, by category of patient. The indicators developed focused on surgery waiting times and emergency care and as such were not entirely consistent with the Planning Board's emphasis on increasing the focus on chronic and other nonacute care. Work on encouraging a focus on quality has not been confined to Victoria. The Australian Health Ministers' Advisory Council Task- force on Quality in Australian Health Care noted in 1996: Boards of Directors of health care facilities and their managers must receive the same kind of encouragement to ensure safe, high quality care provision as they currently receive to ensure responsible financial management. If managers are to be ac- countable for safety and quality there is a reciprocal responsibil- ity for clinicians to be accountable for efficiency.29 This is a major challenge. Nearly four years after this task force report, anecdotal evidence suggested that most network boards were still more comfortable addressing financial and efficiency-through- put issues than quality of care. The focus of the external monitoring regime, including broad performance indicators and targets, against which they are held accountable, may have reinforced this dilemma. Re7vards and sanctions linked to performance. Health Service Agreements between the DHS and networks formalized the agreed annual output levels and prices, in conjunction with published funding guidelines. In addition, a complex web of other institutional arrangements also affects the incentives for networks and independent hospitals to improve performance. The institutions involved include the legal system and insurance companies (in cases of medical malpractice), the media (whose scrutiny is helped by freedom of information leg- islation), and universities and professional associations (particularly teaching hospitals). Views differ about how hard a budget constraint there was for higher cost networks or independent hospitals under Health Service Agreements. The constraint was clearly harder under case mix than it was under block funding, although the government could still pro- Autonomous Hospitals Become a Commercial Network * 361 vide assistance through the non-case-mix components of normal funding. In fact, the Kennett Government provided "restructuring packages" for two networks and three rural hospitals, although these packages came at the cost of public disclosure of what the govern- ment considered unsatisfactory performance, increased oversight by the DHS in hospital decisions, and possible senior management or board changes.30 Plans to address the problems were reviewed by external consultants from an international accounting firm before additional grants (essentially an equity injection) or short-term, interest-free loans were provided. Whether the same performance pressures are faced by publicly run and nongovernmental providers was still debated. Dr. Clive Wellington, chief executive officer of St. Vincent's, a public tertiary teaching hospital run by the Catholic Church, argued that his non- government hospital had a much harder budget constraint than did network hospitals. It thus had greater incentives to improve financial performance: We have no choice but to achieve the results. It is much more likely a Government would stand by if we faced closure because it can be argued that a Catholic church hospital must manage themselves, and if they and their board didn't do it well enough to keep the hospital viable, well then the Government is just going to have to shift services to nearby publicly-run hospitals where closed wards exist.31 Eleven of the 13 network board members surveyed considered that, since network introduction, changes in funding arrangements had led to tighter Health Service Agreements. Ten board members considered that tighter agreements had increased pressure on boards to perform and that it was a positive development. Nine of the 13 network board members surveyed also considered that rewards and sanctions in Health Service Agreements had become more powerful since network introduction. Views differed about whether this was good or bad. 362 o Innovations in Health Service Delivery Market Environment In the decade before networks were created, bipartisan support had been growing for the view that the structure of Victoria's health sys- tem could be improved. In 1992, the Victorian Parliament's Eco- nomic and Budget Review Committee noted that: Health Service Agreements provide a framework for tackling ef- ficiency within hospitals, but not for allocating resources effi- ciently between hospitals. [The Committee] finds .. . substan- tial scope for efficiency gains through: hospital amalgamations; hospital closures; hospital use-conversions; and specialisation in health services provided by hospitals.32 The Planning Board aimed to enhance both the degree of coop- eration (largely within networks) and competition (largely between networks) among Melbourne's health care institutions. Strategic alliances. The Planning Board considered two approaches to improving integration in service delivery: strategic alliances between independent, publicly run organizations, and amalgamations. It re- jected the first option because these affiliations were "unlikely to hold firm when hard decisions must be made about the location or rationalizing of services. Common governance of hospitals is a more effective mechanism to achieve the objective of high quality, patient focussed services and provides greater certainty that these benefits will be attained."33 However, the board noted that partnerships that could benefit both parties, as well as the patients served by the institutions, were more relevant when describing the links between the government and nongovernment hospital sectors. These partnerships could be structured as strategic alliances, affiliations, or joint ventures. Such relationships existed between the networks and preexisting denomi- national hospitals, hospices, and other facilities and were envisaged for the future privately run public hospitals (to be awarded by com- petitive tender). Autonomous Hospitals Become a Commercial Network * 363 The Planning Board's views on the fragility of alliances were borne out by the breakdown of the relationship between the Inner and Eastern Health Care Network and St. Vincent's Hospital. After the merger of the Inner (which contained St. Vincent's) and Eastern (which contained the Alfred Hospital) Health Care Networks, the new entity had two tertiary teaching hospitals. The network's plan- ning for both hospitals envisaged "some resources will be shifted over time to build up services in the outer east" with explicit tar- gets for an increase in the proportion of services being delivered more than 8 kilometers outside the central business district from 32 to 41 percent.34 Moving these resources heightened strains in the relationship-St. Vincent's thought it was being asked to bear an un- fair proportion of the burden. When St. Vmcent's and Inner and Eastern found themselves part of two different consortia competing for a major tender, St. Vincent's sought to end the alliance and deal directly with the DHS for planning purposes. St. Vincent's CEO had argued that the fragility of alliances is part of their strength, as they will persist only if all parties continue to benefit from the activity. He thought this explained the continuation of the many cooperative relationships with the various campuses and clinical units within the Inner and Eastern Health Care Network after St. Vincent's separation from the network. Merging acute care providers. Acute care providers were merged to re- duce duplication and help develop more responsive services, partly by greater sharing of clinical and nonclinical services. Each network had several acute care hospitals, typically one tertiary or teaching hospital, and several smaller "community" hospitals. The Planning Board's preliminary analysis of hospital cost struc- tures suggested that some small and poorly located community hos- pitals were not viable as acute care institutions.35 It was expected the networks would be forced to close or reconfigure a number of these institutions. In the four years since their establishment, the networks closed nine hospitals and reconfigured the operations of several more, but there are no published estimates of any cost savings from these 364 0 Innovatons in Health Service Delivery changes. However, the Inner and Eastern Health Care Network es- timated that it could produce twice as many inpatient services as it was currently delivering at Burwood and District Hospital with the same funds by closing this institution and moving the activity to the nearby Box Hill Hospital.36 In addition, surplus assets were sold, and the proceeds retained by networks to allow them to deliver more services within their existing budgets. Neither the legislation nor the DHS prescribed how each networkc should capture the benefits of merging acute care providers, and dif- ferent networks have adopted different approaches. The Southern Health Care Network developed networkwide clinical programs. Inner and Eastern developed more along campus-based lines, but or- ganized some cooperative activity such as quality-improvement ini- tiatives (box 9.3). The Western Health Care Network implemented networkwide clinical programs but, after merging with North East- ern, introduced more campus-based clinical management structures.37 The Planning Board saw merging services as a way to raise both efficiency and quality. It cited work from the United States that sug- gested that higher volumes at each site led to better outcomes for some procedures. The introduction of network boards was seen as more likely to reduce the barriers to mergers or reconfiguration and thus lead to improvements in quality through specialization and in- creased volume. Case-mix funding, with its extra pressure to constrain unit costs, created strong incentives to consolidate multiple sites within a net- work to capture economies of scale. However, because case mix di- rectly funded outputs rather than outcomes, there were no financial incentives to consolidate sites if higher volume improved quality of care instead of reducing unit cost. Merging acute and nonacute providers. The Planning Board envisaged that, ideally, each network would deliver acute, aged, psychiatric, pal- liative, rehabilitative, and home-based services. Mainstreaming specialist (particularly psychiatric) services with more general services and integrating acute inpatient care with non- acute community services commenced in the late 1980s and early Autonomous Hospitals Become a Commercial Network ° 365 Box 9.3 Introducing Service Quality Improvements in a Network Environment The Planning Board proposed that each network establish a quality assurance subcommittee of the board. The roles of these boards have included analyzing performance informa- tion and overseeing the implementation of improvement strategies. Some quality improvement activities may be easier to es- tablish when campuses are part of a network, rather than stand-alone institutions. Claimed benefits include: o Improved ability to develop and share performance infor- mation by facilitating agreement on common data stan- dards and disclosure o Increased incentives and ability to share insights across campuses o Increased efficiency in resource use through joint develop- ment of better information and imiprovement approaches o Enhanced incentives to share knowledge about how to undertake quality accreditation through mentoring pro- grams or exchanging documentation o Expanded scope to reduce quality lapses resulting from lack of coordination between different modes of care. However, it is possible networks' more centralized or standardized approaches may have inhibited tailoring qual- ity improvement strategies to the needs of particular sites and inhibit experimentation with different approaches. Sources: J. Wllkinson, Board Member, Inner and Eastern Health Care Network (IEHCN), Melbourne, personal communication, August 16, 1999; G. Ryan, chief of clinical services, IEHCN, Melbourne, personal communication, August 26, 1999. 366 o Innovations in Health Service Delivery 1990s, predating networks. This process had general support among providers, but the possible diversion of resources to acute care ser- vices or to a more medical than social work approach caused some concern.38 The establishment of Victoria's networks continued the mainstreaming and integration of services. In its Interim Phase I Report, the Planning Board recommended in- tegrating extended care providers (aged and palliative care services) within the hospital networks (table 9.3). As a result, some smaller acute care hospitals (not considered viable as general hospitals) were reconfigured as specialist aged or palliative care institutions within networks. Network management may be more supportive of non-hospital- based care (or service substitution) than hospital-based management. Asked by the Victorian Auditor General's Office whether network hospitals could achieve further efficiency savings through service substitution, four of the six network chief executives (but less than half of hospital chief executives) said they could.39 This response seems consistent with the hope that networks would focus less on particular sites and more on services. Initiatives to improve continuity of care have included encourag- ing better discharge planning and strengthened links between the network's campuses and general practitioners in the community. A number of people interviewed for this study said that improving the efficiency and quality of acute care had been a higher priority for the Kennett Government and networks than improving links be- tween services in the short term. This concern was reflected in atten- tion to issues such as the design of funding mechanisms and changes in the delivery of services. However, respondents thought the diffi- cult task of improving links was receiving more attention than in the past. Twelve of the 13 board members surveyed considered that, since network introduction, focus had sharpened on care settings other than hospital inpatient care, changing the care mix away from acute, and shifting services to the most appropriate locations. Market exposure. Victoria's government-controlled hospitals have tra- ditionally competed among themselves and with some denomina- tional hospitals for funding to provide services to public patients. Autonomous Hospitals Become a Commercial Network * 367 Table 9.3 Merging Acute with Extended Care Providers ADVANTAGES DISADVANTAGES * Potential for integrating acute and * Loss of focus on subacute and nonacute postacute services care programs * Improved care of the aged in both * Possible increase in operating costs and acute care hospitals and extended reduced patient access to the collocated care centers services since most patients are disabled * Potential for collocating services and * Loss of autonomy and service identity reduced future capital investment within the community * Shared expertise * Weakened links with the community * Streamlined discharge planning and * Increased costs as expensive practices transitions to rehabilitation associated with acute care hospitals spill over into aged care programs * Reduced duplication, shared * Possible diversion of funds intended for infrastructure costs, and economies aged care programs to support acute of scale care programs * Improved access to technology * Diseconomies of scale * Potential for increased distance between senior management and the communities they serve Source MHPB, Developing Melbourne's Hospital Network Interim Phase 1 Report, Melbourne, April 1995, p 54 Much of this competition occurred in the political domain. They also competed with denominational and for-profit private hospitals for privately insured and self-funded patients. Private hospitals could not generally compete to provide government-funded services to public patients. Postnetwork competition increased in four ways: between net- works for market share; through the use of tendering for the al- location of clinical services statewide and cross-networks; through increased use of competitive tendering within networks, and through the use of tendering for the operation of public hospitals. * Between networks for market share. Competition for market share had been largely limited to some funding for output growth. Net- works lacked strong incentives to compete for public patients (who have some choice of hospital), as funding did not follow individual patients and most networks covered a specific geographical area. Under the case mix-based funding arrangements, the DHS allocated funding to each network for a specified number of acute care inpa- 368 o Innovations in Health Service Delivery tients (adjusted for complexity). Networks could compete to fund ad- ditional output (only about 2 percent of the state total). Those that do not reach their funded output faced the prospect of reduced fund- ing that year and lower allocations the next year. In the absence of strong direct competition for market share, the DHS used yardstick competition to boost performance, mainly monthly comparisons of networks' performance against a standard set of fi- nancial and performance indicators (table 9. 1).40 In addition, networks had to publish a standard set of performance data in their annual reports, and the DHS published comparative data.41 The media, academics, or community groups have given lit- tle scrutiny to annual report data, despite wide reporting of the wait- ing time for elective surgery and emergency department services published by the DHS. 0 Tendering statewide and cross-network clinical services. Networks bid to deliver some new and restructured statewide and cross-network clin- ical services. For example, contracts to deliver statewide services for infectious diseases and HIV/AIDS were awarded in March 1996 to what were then the Western Health Care Network and the Eastern Health Care Network (the Alfred Hospital).42 St. Vincent's won the contract for lithotripsy services in one region.43 Other services the Planning Board had envisaged as being awarded by competitive ten- der were instead allocated by expert review, which arguably decreased the level of competition and increased the degree of politicization of decisions. Competitive tendering for the allocation of cross-network and statewide services was used less than the Planning Board envisaged. The number of providers who could also furnish necessary related services can be small (e.g., for major trauma). Also, in some cases, a convenient location was important for the main client group (e.g., for HIV/AIDS, in the part of Melbourne where many gay men live). Groups representing some patients (notably, AIDS/HJV patients) have also argued that moving services at the end of a fixed contract would adversely affect the quality of care, where a collaborative rela- tionship with providers was vital.44 Autonomous Hospitals Become a Commercial Network * 369 Competitive tendering within networks. Within-network competition was driven partly by the Kennett Government's implementation of competitive neutrality under Australia's national competition policy and partly by budget pressures. Competition policy aimed to facili- tate "fair" competition between private and public providers of final goods and intermediate inputs. The Kennett Government identified some business activities within hospitals that could be provided privately and that conse- quently had to undergo internal review and appropriate internal cost- ing. These services included noncinical services such as car parks, computing, laundry services, engineering, cleaning, and catering, and such clinical services as medical imaging, pathology, pharmacy, allied health services, and general practitioner services.45 Networks could choose to open other services to competitive tendering. Some ser- vices have been extensively tendered-37 percent of cleaning and catering in the larger hospitals have been contracted out.46 The cost savings from outsourcing are undoubtedly one of the major achieve- ments of the reforms. Network managers, one step further removed from direct service delivery, may have been more enthusiastic than hospital managers about competitive tendering. Almost every network board member surveyed considered that, since networks were established, board in- terest had increased in competitive tendering of both clinical and nonclinical services, although this may just have reflected increasing cost pressures. Most board members thought that heightened inter- est in competitive tendering of services was a positive development. * Tendering public hospital operation. The Kennett Government was moving toward a mixed system, where corporatized networks coexist with competitively tendered and privately built and operated public hospitals in the metropolitan area. After first tendering for financing, construction, and operation of two rural hospitals, the Kennett Gov- ernment began a process of competitive tenders for two more com- plicated tertiary teaching hospitals in the metropolitan area and one less complex outer suburban hospital. 370 o Innovations m Health Service Delivery Community Input The move to networks reduced the number of people involved in hospitals at board level. Before network introduction, the 32 nonde- nominational public hospitals each had 12 or more board members (about 400 people, allowing for those on more than one board). Af- terward there were seven network boards of directors, each with no more than 9 members. The community representatives on the earlier hospital boards were sources of information about community views and prefer- ences. They also served as spokespersons when seeking donations and volunteers from the community, links to philanthropic organiza- tions, advocates when seeking political support, and sources of advice to hospital management on business issues. The Planning Board focused on the first two roles, noting the im- portance of maintaining the sense of community ownership of ser- vices. It received submissions on the importance of community input into service development and suggested that individual campuses re- tain local advisory boards and nurture hospital auxiliaries (box 9.4). The board also argued for the retention of individual hospital iden- tities, each with its community of interest and local affiliations. The legislation that created networks required each to set up an advisory committee of community representatives nominated by the board. The Health Issues Centre maintained that this had not been a satisfactory substitute for board representation because of "the ab- sence of any legislative timeframes for the establishment of these committees, guidelines for their terms of reference and membership, or any arrangements for monitoring/reporting to government on the extent to which the recommendations of such committees are ac- cepted by the Networks."47 The Inner and Eastern and Peninsula networks set up network ad- visory committees. However, the representatives of four patient advocacy groups contacted for this study considered the network superstructure irrelevant to most consumers whose contact is with individual hospitals or services. This is reflected in the strategy of networks such as North Western Health Care: Autonomous Hospitals Become a Commercial Network * 371 Each hospital and program area within the Network is respon- sible for understanding the needs of the community it serves and maintaining and developing working relationships with health service providers in their local area. Local responsibility recognizes the differences that exist between the communities who use our services. The Network catchment spans 11 local government areas and the needs and preferences of the com- munities change within these areas. The best way to ensure that the services meet the needs of these communities is to allow the development of local strategies by the staff closest to the partic- ular community.48 Similarly, the Inner and Eastern Health Care Network considered its advisory committee only one conduit for community input, the main ones being feedback received by individual hospitals or clinical services. The responsiveness of clinicians and managers to consumers' con- cerns varied across networks and campuses. Some advocacy groups suggest the Alfred Hospital became more responsive, partly due to its experience after assuming responsibility for HIV/AIDS patients, an exceptionally articulate and motivated consumer group.49 Developing satisfactory consultation mechanisms nonetheless re- mains a challenge. On the Alfred's HIV Care Committee, the hospi- tal's main liaison group with its HIV patient representatives, the ex- ternal evaluation of the IHIV/AIDS service noted that effectiveness ratings vary from "an effective vehicle for the voicing of concerns and lobbying for action" to "slow to respond to change, with a ten- dency to focus on day-to-day issues rather than servicewide or strate- gic issues."50 A related issue is the level of community in-kind and financial sup- port for health care services. When the networks were created, the parliamentary opposition argued the changes had the potential to "destroy a lot of [beneficial] community input and enthusiastic, com- petitive fundraising."51 Board members held divergent views on whether board-level atten- tion to findraising issues had increased or decreased, and whether any such change had been positive or negative. But the way community 372 o Innovations in Health Service Delivery Box 9.4 Equity Issues rFcced by Nelowrk -rdls DHS funded a nongovermnental organization, the Health Issues Centre, a community advocacy group, to coordinate a submission to inform the Planning Board on issues asso- ciated with consumer and community consultation. They organized a forum where Dr. Joanne Wilkinson, a health care academic and network board member, noted: Network Boards would have to deal with questions of equity in deterniining who is most worthy of receiving services. In an environment of financial restraint, Network Boards would have to make choices about the relative merits of: o Powerful and loud lobby groups o The unemployed or severely disabled, dependent and dis- enfranchised, the elderly, or the young o Those who require many and expensive tertiary services or those who woutld benefit from less costly continuing supportive care/health promotion and maintenance o Those who are close and more visible or those who are at some distance from the provider agencies, as m the case of statewide services or existing usage patterns where some organizations agree to provide specialist services to those outside their immediate region o Those of the dominant culture or those with cultural difference.a The Health Issues Centre has noted the challenge of rep- resenting the diverse and competing needs of various groups served by large and complex health care institutions. It did not support "the appointment of representatives linked to only single-issue groups but rather the appoint- ment of well informed and committed advocates of health care consumers with all their interests" on network commu- nity advisory committees.b a. Health Issues Centre, Networking Hospitals: Consumers and Communities, Melbourne, 1995, pp. 17-18. b. Ibid., Appendix D, p. 3. Autonomous Hospitals Become a Commercial Network * 373 views are incorporated into decisionmaking had clearly changed. Ten of the 13 network board members surveyed considered that lobby groups now have less influence over board decisions than in the past; the other 3 thought it was unchanged. Overall, there seems little ev- idence that the move away from an explicit community representa- tive role for boards has hindered the development of new and more focused ways to gain community input at the individual hospital level. Implementation Network introduction followed other changes introduced early in the Kennett Government's term, including budget cuts and changes in funding arrangements. It also built on changes implemented by previous governments that clarified the relationship between the DHS and independent hospitals. These other developments-for example, comparative reporting for hospitals, based on case mix and Health Service Agreements that more clearly specified hospitals' ob- jectives and relaxed input controls-were important precursors to the changes of the 1990s. (Context I;n the two years before networks were established, the Kennett Gov- ernment marshaled evidence that the Victorian health system needed fundamental changes. These efforts were critical in developing pop- ular and political support for the hospital reforms.52 The Kennett Government had gone into the previous election with a policy to in- troduce area health authorities. These bodies would have had re- sponsibility (thus a service-purchasing role) for hospital and other state government-provided services for a defined population. The 1993 Victorian Commission of Audit provided a comparative analy- sis, suggesting that Victorian hospitals could be much more cost- effective, but cast doubt on the efficacy of a new layer of purchasing authorities. Noting its concerns about the "complexity of any mas- sive organizational change of this kind," the commission suggested 374 o Innovations in Health Service Delivery that "it may be appropriate to trial one or two [area health authori- ties] for the whole state before embedding an extensive new state- wide area health management structure."53 However, the Commis- sion of Audit supported network arrangements between hospitals to capture economies of scale in functions such as engineering, cater- ing, pathology, computer services, and finance. Process Economic issues of economies of scale and scope played a major role in the design of networks, particularly the number of networks, their composition, and the choice of clinical services to be delivered statewide from only a few locations. Further, the calculations were aimed at establishing scope for competition in the Melbourne hos- pital market. However, the political economy of this sector also heavily influenced the design and the institutional structures. One participant in the change process described the initial challenge as "turning clay into putty" so that the capital could be better used by reconfiguring existing sites and closing sites, where network boards considered it necessary. Sequencing. As described in table 9.1, the newly elected Kennett Gov- ernment initiated a sequence of key changes within the health sector: o Budget cuts were imposed and case mix-based funding was intro- duced. o The creation of networks was announced, the combinations of providers to be decided by the Planning Board. o A capital works program was announced, including publicly fi- nanced and built projects, and competitive tendering of construc- tion, financing, and operation of five new acute care hospitals. Sequencing can have a major effect on the outcome of major change. At two levels, it is relevant to network introduction: o The changes in the level and structure of funding occurred before, and separately from, network introduction. Autonomous Hospitals Become a Commercial Network * 375 Before recommending any specific grouping, the Planning Board built in-principle support for the concept of networks and the need to close some institutions. The budget cuts of 10 percent over the two-year period 1993-94 and 1994-95 (on top of a 4 percent cut in 1992-93) preceded the move to networks. Coinciding with budget cuts, the introduction of case-mix funding had wide-ranging repercussions on the hospital system. This system of funding highlighted the less economically vi- able institutions and increased acceptance of a need to reconsider the structure of services across sites. Performance data on a case-mix basis had been introduced several years before but without any link to recurrent funding. Major resistance was expected to any specific grouping of hospi- tals and the closure of any institutions. After receiving the board's In- terim Phase 1 Report, the Kennett Government introduced and passed legislation necessary to implement the specific recommendations on hospital groupings contained in the final Phase 1 report. The interim report was released for public comment with options for the key rec- ommendations. This meant that the legislative instruments were in place before the government received the final report. The final rec- ommendations could be implemented without additional legislation. The main opponents to the proposal to merge the management of acute care institutions were the medical and nonmedical staff and the existing boards. The Kennett Government announced that the Plan- ning Board's role was to determine the number and composition of the networks, not whether they should be created. This approach forced the opponents to decide whether to devote their energy to achieving the best combination of providers, and thereby reduce the likelihood of closure, or to focus on opposing any change. The Planning Board was not specifically asked to make recom- mendations on hospital closures. It suggested some hospitals should be downgraded as services were moved to the outer suburbs. How- ever, it judged that the final decision on the future of individual sites was better made by incoming network boards in consultation with government. 376 0 Innovations in Health Service Dehvery Speed of change. The administrative process of aggregating existing in- stitutions happened fast. The amendments to the Health Services Act that enabled the mergers to occur also prevented review of the merger decisions by the Supreme Court (the highest state court). This en- sured that, once the Kennett Government announced which combi- nations of institutions would be merged, the interest groups that had been consulted in the process could not frustrate its decisions.54 The Kennett Government's resolve in other politically charged areas of policy outside health may have affected resistance to network creation. Two years before setting up hospital networks, the govern- ment had merged local governments and closed a number of small urban schools in the face of fierce organized resistance. Indeed, in opposing the bill to create hospital networks, the parliamentary Op- position described the changes as applying "the principles of local government amalgamation to our health care system."55 The government went for a "big-bang" approach to the changes, simultaneously abolishing all the metropolitan hospital boards and establishing the networks. In contrast, the parliamentary Opposition advocated implementation "on a trial basis: done in one region first, monitored, assessed and, if proved to be successful, applied more generally" but accepted some advantages in what it saw as similar to a regional health board model.56 The Kennett Government already had experience with rapid pol- icy implementation with the introduction of case-mix funding. Both political sides supported the introduction of this funding tool, but the previous government had intended to phase it in gradually over several years. Almost immediately, the incoming government intro- duced case-mix funding and, although it was refined over time, did not experience major implementation problems. In contrast, the Kennett Government used phased implementa- tion in other areas of health, particularly with the introduction of competitive tendering and contracting of public hospitals. It first tendered less complex and smaller rural institutions, drawing on some of the expertise developed tendering prison construction, fi- nancing, and operation. The officials involved then moved on to ten- dering more complex metropolitan tertiary hospitals. Autonomous Hospitals Become a Commercial Network * 377 Transparency. The Planning Board's public processes elicited about 400 responses from various interested parties (compared with about 100 received during the consultation phase before the introduction of case-mix funding).57 The use of a public process and an independent board to develop options may have facilitated some acceptance of the changes. For ex- ample, The Age newspaper, which had often been critical of the Ken- nett Government on health issues, was convinced the changes were appropriate: "a long-overdue attempt to bring Victoria's hospital sys- tem under control and up to date."58 Preliminary Assessment Networks were created with three key objectives: to improve effi- ciency, quality, and accessibility.59 Lack of available performance in- formation and the number of confounding factors make it difficult to assess pre- and postnetwork performance changes, let alone attribute change to specific factors. The Planning Board's recommended in- dependent evaluation of the networks after three years did not occur.60 When the Kennett Government announced the metropoli- tan health care services plan in October 1996, it foreshadowed the fact that it and the networks would evaluate implementation within five years.61 The 20-odd network board members who had previously served on hospital management boards were able to make before-and-after comparisons, although these are subjective. The responses of 13 of these people to the structured questionnaire provided some insights into the changes. Efficiency Unit costs have declined since networks were introduced. Real unit costs per case mix-adjusted inpatient in the network hospitals fell about 5 percent in the two years after the introduction of the net- works.62 The dispersion of costs across networks appeared to have grown. The average unit costs of the highest-cost network were an 378 o Innovations in Health Service Delivery estimated 7 percent higher than those of the lowest-cost network (compared with 5 percent in the last prenetwork year). Lower costs could have been due to technical change rather than network introduction or case-mix funding. Comparisons of costs across the Australian states and territories provide some control (each had access to similar new technology). In the last prenetwork year, Victoria's reported unit costs were 7 percent higher than South Australia's, the lowest-cost jurisdiction. In the next three years, Vic- toria had the lowest reported unit costs.63 Data for 1998-99 are not yet available. However, these data are subject to caveats, in part be- cause of differences across jurisdictions in how services are struc- tured. The Victorian data cover nonmetropolitan and nonnetwork hospitals. Observers such as Davidson suggested that networks represent an extra layer of overheads rather than a source of efficiency, and the policy of the then-Opposition Labor Party at the September 1999 state election was to abolish networks as a cost-saving measure.64 However, the rationale for networks was that certain activities cost less or achieve better outcomes at network level-either substituted for activities that could be centralized in the DHS such as finance functions, planning, and some monitoring, or devolved as shared services at the hospital level. The claim that networks add an un- necessary layer is difficult to assess in the absence of a detailed cost- benefit study. Most observers interviewed for this study said that the postnet- work reconfiguration of services, and any resulting cost savings, could not have been achieved as readily with the site-based boards. They argued that having provider networks made commitments to retain access to services more credible, and the ability to move staff within the organization eased the changes. A few observers disagreed because the acute care hospitals that were closed or reconfigured to nonacute care roles were all smaller, less powerful institutions. They said grouping into networks did not result in closure or downgrading of any tertiary teaching hospitals. Although the Planning Board earmarked the Royal Melbourne and Alfred Hospitals for reduced roles as resources were moved to the Autonomous Hospitals Become a Commercial Network * 379 outer suburbs, the Kennett Government subsequently allocated to them new acute care services (trauma and HIV/AIDS). The behavior of denominational and for-profit hospital operators suggests that merging stand-alone institutions or sharing services may confer some benefits. The Sisters of Charity have developed a regional structure and have also formed a joint venture with the Sis- ters of Mercy to bring together their respective private hospitals, al- though their motivation appears to be enhancing their ability to at- tract new business as much as cutting costs.65 In the 1990s, for-profit hospitals have merged into large groups, at the expense of indepen- dent, stand-alone, for-profit hospitals, but how much is driven by cost saving is difficult to determine. Most network board members who had previously served on hos- pital management boards thought that network introduction and case-mix funding contributed significantly to efficiency improve- ments, whereas competitive tendering and industrial relations changes had not had a major impact. Factors that may have improved effi- ciency include the movement of some services between campuses and closure of some smaller campuses. Other observers have argued that case-mix funding, and subsequent reductions in unit prices, have been more important than networks in driving efficiency gains. Some anecdotal evidence suggests that more rigorous rationing decisions are now being made at the point of admission. However, this seems to have occurred largely because demand growth ex- ceeded funded supply. Quality How much the quality of care changed during the 1990s is unclear- and less clear, the cause. The Victorian Auditor General's Office's performance audit of case-mix implementation concluded that the quality of care had suffered from the introduction of case mix and from funding cuts. However, the DHS vigorously disputed this con- clusion on the grounds that the analysis was based almost entirely on a questionnaire of managers, clinicians, nurses, and allied health pro- fessionals instead of objective measures of a decline in care.66 Any contribution from network introduction was not analyzed. 380 o Innovations in Health Service Delivery It is possible that lower costs were achieved through quality re- ductions instead of through real efficiency increases. Some quality initiatives were undertaken at network level, although the approach differed across networks (box 9.3). However, data for assessing the network effect on quality are scarce.67 One possible service-quality indicator, the proportion of separations resulting in unplanned re- admission to hospital, is an imperfect measure, in part because of im- precise definitions. Based on aggregated departmental data for all networks, this does not seem to have changed since the introduction of networks (or since the budget cuts and introduction of case-mix funding in 1993-94). Published data from the Inner and Eastern Health Care Network show a decline from 6.45 percent of separa- tions in 1995 to 2.86 percent in 1999.68 Data for other networks are not published. Another possible quality indicator is the number of hospitals ac- credited by the Australian Council on Health Standards. Since net- works were created, more hospitals have been accredited. However, the Kennett Government gave all hospitals incentives to obtain ac- creditation and was moving toward mandatory accreditation for all public acute care hospitals.69 Most of the network board members surveyed considered that network introduction and technological change have both con- tributed significantly to quality improvements, but that unit price reductions have adversely affected quality. Views differed about whether boards are giving issues of quality improvement more or less scrutiny. However, one board member noted that the boards of stand-alone hospitals sometimes get involved in analyzing specific problems, but he doubted that this was the best way to address broader quality issues. Accessibility Postnetwork accessibility, in terms of geographic access to services and timeliness, may have improved in some areas, although again it is difficult to assess how much to attribute to the establishment of networks and how much to policy and funding decisions that could Autonomous Hospitals Become a Commercial Network * 381 have been implemented without networks. The Planning Board envisaged that improved access to services in the middle and outer suburbs would be achieved by moving services and establishing "hub-and-spoke" services for specialist care such as eye treatments, palliative care, and renal and cancer services. Most questionnaire respondents considered that network intro- duction, capital injections for new capital works, and technical change contributed significantly to accessibility improvements. Central specialist hub sites, supporting accessible spoke sites, were devel- oped for some eye, ear, cancer, renal, and obstetric treatments, and more patients were being treated at outer suburban sites.70 Some de- lays have been associated with the establishment of some new sites, particularly the competitive tendered hospitals, but most delays oc- curred at the departmental level. Data on timely access are difficult to interpret. The number of "semiurgent" cases, whose wait for elective surgery in public hospi- tals exceeded the "ideal" (90 days), declined after network introduc- tion, but increased subsequently. In the early 1990s, many "urgent" cases had to wait more than 30 days, but the recorded number of longer waiting times had already been reduced to nearly zero before network introduction. More than four years later, Victorian hospitals were treating nearly 40,000 more patients per quarter (a 20 percent increase)-at lower unit costs after adjusting for case mix. However, output and funding have not kept up with demand, and total waiting lists increased by 5,200 (20 percent) to about 34,000 in the Septem- ber 1999 quarter.7' This reflected more the level of output pur- chased by the DHS than individual network performance. Conclusions According to some commentators, most policy changes in the health sector are incremental because of complex interactions within the health system, powerful interests, and the need for caution because of the potentially damaging implications of inappropriate policies. 382 o Innovanons in Health Servce Delivery Network introduction built on earlier gradual development of more specific arrangements with quasi-independent institutions. It complemented fundamental changes in the funding arrangements that encouraged a greater focus on outputs and a more systematic ap- proach to improving Victoria's hospital system. Four broad observations emerge from this case study: o Victoria has adopted a multipronged approach to improving the efficiency and effectiveness of the health care system, and the sep- arate contribution of any one element cannot be assessed. o A single independent board responsible for several sites (and at arms' length from the party political process) may be more effec- tive in closing sites and reconfiguring services than either a cen- tral department or boards responsible for each site. The political economy of health reform presents many challenges, and the de- sign of institutional structures needs to consider these issues as well as considerations such as economies of scope or scale and the potential for competition. The networks rapidly closed or recon- figured individual institutions without major disruption of services or public outcry. o Output- (or outcome-) based funding arrangements and effective performance monitoring underpin arms'-length relationships and improved corporate governance. The introduction of case-mix funding for acute inpatient care and emergency department ser- vices strengthened incentives for hospitals to find ways of deliv- ering services better, including site closures and service recon- figuration, among many other strategies. o Community representation on management boards is only one means of creating incentives for providers to focus on consumer needs. Identifying individuals who can adequately represent the disparate needs of consumers using large service providers can be difficult. Engaging the community in an advisory capacity at the point of delivery might be more beneficial. Autonomous Hospitals Become a Commercial Network * 383 Questions remain about the robustness and stability of changes of this kind, including the extent to which results depend on nonstruc- tural factors, such as the qualities of the individuals appointed to boards and the willingness of governments to remain at arms' length. The answers to these questions will become apparent only over time, but ongoing institutional changes, including those introduced by a new government in early 2000 (which were largely refinements of the network model), complicate the picture. However, analysis of the experience with similar reforms to the governance of other govern- ment activities may suggest some answers. Notes 1. Public hospitals provide free treatment and accommodation to eligible admitted patients who elect to be treated as public patients. They also provide free services to nonadmitted patients and may provide (and charge for) treatment and accommodation services to private patients. They can be run by governments, not-for-profit entities, and, more recently, for-profit organizations. Steering Com- mittee for the Review of Commonwealth/State Service Provision, Report on Government Services 1999, Canberra, wwwpc.gov.au/service/ gspindex.html (accessed 2 August 1999). 2. The number and composition of networks changed over time; four networks were merged into two and one tertiary hospital was re- moved from a network to form an additional network in preparation for being competitively tendered. There was a change of govern- ment after this case study was prepared. The new government is re- viewing the role of networks, with a view to increasing community participation and having more, but smaller, networks. 3. Twenty-one people who had served on the boards of both net- works and hospitals were identified. Sixteen individuals could be con- tacted and 13 agreed to participate. 384 o Innovatons in Health Service Delivery 4. Metropolitan Hospitals Planning Board (MHPB), Taking Mel- bourne's Health Care Networks into the 21st Century: Phase 2 Report, Melbourne, December 1995, p. 1. 5. MHPB, Developing Melbourne's Hospital Network: Phase 1 Report, Melbourne,June 1995, p. 81. 6. The Age, May 2, 1995, p. 13. 7. Health Services Act of 1988, Section 33. 8. Networks are called "metropolitan hospitals" in the Health Ser- vices Act of 1988. The functions of the board are outlined in Section 40D. 9. MHPB, Phase 1 Report, p. 57. 10. Five of the in-itial seven network chief executive officers were from Victorian hospitals. 11. Victorian Auditor General's Office, Acute Health Services under Case Mix: A Case of Mixed Priorities, Melbourne, May 1998, p. 215. 12. MHPB, Phase 1 Report, p. 73. 13. Victoria, Legislative Assembly, Hansard, September 3, 1998, p. 204. 14. Victoria, Legislative Assembly, Hansard, October 8,1998, p.42 1. 15. Service agreements are not legally binding contracts, partly be- cause networks, as public statutory bodies, have limited ability to de- cline or negotiate terms. Inner and Eastern Health Care Network (IEHCN), Response to the Health Services Policy Review Discussion Paper, East Melbourne, 1999. 16. Department of Human Services Victoria (DHS), Public Hospi- tals Policy and Funding Guidelines 1999-2000, Melbourne, 1999, p. 16. 17. The Victorian Government introduced a capital user charge for 1998-99 to encourage improved resource allocation, but this charge was not imposed on networks or other health care providers. Autonomous Hospitals Become a Commercial Network * 385 Department of Treasury and Finance (DTF), The Capital Assets Charge in 1998-99, Melbourne, 1998. 18. IEHCN, Response to the Health Services Policy Review Discussion Paper, EastMelbourne, 1999. 19. V Lin and S. Duckett, "Structural Interests and Organizational Dimensions of Health System Reform, in H. Gardner, ed., Health Policy in Australia (Melbourne: Oxford University Press, 1997), p. 59. 20. C. Walker, "The Emergence of the Hospital System in Mel- bourne: 1846 to 1975" (Ph.D. diss., La Trobe University, Melbourne, 1994), pp. 304-305. 21. H. Swerissen and S. Duckett, "Health Policy and Financing," in Gardner, Health Policy in Australia, p. 21. 22. Economic and Budget Review Committee, Hospital Services in Victoria: Efficiency and Effectiveness of Health Service Agreements, Par- liament of Victoria, Melbourne, August 1992, p. 17. 23. Parliament's Economic and Budget Review Committee, Effi- ciency and Effectiveness of Health Service Agreements, p. 11. 24. P. Cameron, M. Kennedy, and J. McNeil, "The Effects of Bonus Payments on Emergency Service Performance in Victoria," Medical,Journal of Australia 171 (6): 243-46 (September 1999). 25. DHS, Policy and Funding Guidelines. The proportion closely linked to output had progressively grown, from 57 percent in 1993- 93, and then to 66 percent in 1996-97. SCRCSSP, Reforms in Gov- ernment Service Provision, Canberra, 1997, p. 49. 26. Victorian Auditor General's Office, Report of Ministerial Port- folios, Melbourne, May 1999, p. 72. 27. SCRCSSP, Reforms in Government Service Provision, Canberra, 1997. 28. S. Duckett, "Economics of Hospital Care," in G. Mooney and R. Scotton, eds., Economics and Australian Health Policy (Sydney: Allen and Unwin, 1998), p. 103. 386 o Innovations in Health Service Delivery 29. Taskforce on Quality in Australian Health Care (TQAHC), The Final Report of the Taskforce on Quality in Australian Health Care, Aus- tralian Health Ministers' Advisory Council, Canberra, June 1996, p. 17. 30. Victorian Auditor General's Office, Acute Health Services under Case Mix: A Case of Mixed Priorities, pp. 210-12. 31. Dr. Clive Wellington, chief executive officer, St. Vincent's, Melbourne, personal communication, September 7, 1999. 32. Victorian Parliament's Economic and Budget Review Commit- tee, Efficiency and Effectiveness of Health Service Agreements, pp. 14-15. 33. Ibid. 34. DHS, A Healthier Future: A Plan for Metropolitan Health Care Services, Melbourne, 1996, pp. 27-29. 35. J. Fahrer, Hospital Costs in Melbourne: An Econometric Analysis, Metropolitan Hospitals Planning Board, Melbourne, 1995. 36. Simon Blair, chief executive officer, Inner and Eastern Health Care, Melbourne, personal communication, September 2, 1999. 37. Western Health Care Network (WHCN), Western Health Care Network Annual Report 1996-97, Melbourne, 1997; North Western Health (NWVI, North Western Health Annual Report 1997-98, Mel- bourne, 1998; Southern Health Care Network (SHCN), Annual Re- port 1997, Melbourne, 1997, and Annual Report 1998, Melbourne, 1998. 38. See, for example, W Healy and H. Varney, "Mainstreaming and Integrating Psychiatric Services: The Victorian Experience," Australian Journal of Social Issues 30 (2): 179-94. (May 1995). 39. Victorian Auditor General's Office, Acute Health Services under Case Mix: A Case of Mixed Priorities, p. 203. Autonomous Hospitals Become a Commercial Network * 387 40. Yardstick competition is particularly helpful in cases where it is difficult to get sufficient competition going in the market, as is com- mon with hospitals. 41. See, for example, DHS, Hospital Services Report: December 1999. 42. WHCN, Annual Report 1996-97; Health Outcomes Interna- tional (HOI), Evaluation of the Relocated HIV/AIDS and Infectious Dis- eases Services, Department of Human Services, Melbourne, 1999. 43. St. Vincent's Hospital, Annual Report 1997, Melbourne, 1997, p. 12. 44. HOI, Relocated HIV/AIDS and Infectious Diseases Services, De- partment of Human Services, Melbourne, 1999. 45. Department of Premier and Cabinet (DPC), Competitive Neu- trality: A Statement of Victorian Government Policy, Melbourne, 1996, p. 21. 46. Duckett, Hunter, and Rassaby, Health Services Policy Review, p. 94. 47. Health Issues Centre, "Health Services Policy Review Re- sponse," Melbourne, July 1999; processed, p. 5. 48. NWIH, North Western Health Annual Report 1997-98, Mel- bourne, 1998, p. 33. 49. Victorian Aid Council/Gay Men's Health Centre (1999), p. 5; HOI, Relocated HIV/AIDS and Infectious Diseases Services, pp. 24-25. 50. H01, Relocated HIV/AIDS and Infectious Diseases Services, p. 25. 51. Victoria, Legislative Council, Hansard, June 7, 1995, p. 1265. 52. It is interesting to contrast the Victoria and New Zealand re- forms on this score. The New Zealand public never seemed to have 388 o Innovations in Health Service Delivery been convinced of a need for serious hospital reforms-a fact that made implementation even more challenging than it would have been otherwise. 53. Victorian Commission of Audit, Report of the Victorian Com- mission ofAudit: Volume Two, Melbourne, May 1993, p. 133. 54. E. Adeney, "The Invisible Reconstruction of the Victorian Metropolitan Hospital System," Journal of Law and Medicine 4 (3): 267-75 (February 1997). 55. Victoria, Legislative Council, Hansard, June 7, 1995, p. 1264. 56. Victoria, Legislative Council, Hansard, June 7, 1995, p. 1275. 57. Victorian Auditor General's Office, Acute Health Services under Case Mix: A Case of Mixed Priorities, p. 25. 58. The Age newspaper, May 2, 1995, p. 13. 59. MHPB, Phase 2 Report, p. 1. 60. MHPB, Phase 2 Report, p. 5. 61. DHS, A Plan for Metropolitan Health Care Services. 62. Unpublished DHS data, 1999. 63. Based on unpublished DHS data for 1997-98; SCRCSSP, Re- port on Government Service Provision 1997, Canberra, 1997; SCRC- SSP, Reforms in Government Service Provision, Canberra, 1997; and SCRCSSP, Report on Government Services 1999, Canberra, 1999; and Australian Institute of Health and Welfare (AIHW), Australian Hos- pital Statistics 1997-98, Canberra, 1999. 64. See, for example, K. Davidson, "Why Our Public Hospitals Are So Sick," The Age newspaper, September 6, 1999, p. 17. 65. St. Vmcent's Hospital, Annual Report 1998, Melbourne, 1998. 66. Victorian Auditor General's Office, Acute Health Services under Case Mix: A Case of Mixed Priorities. Autonomous Hospitals Become a Commercial Network * 389 67. Timeliness of treatment for elective surgery, which, in some respects, is a dimension of quality, is considered under accessibility. 68. IEHCN, Progress to Date, East Melbourne, August 1999. 69. SCRCSSP, Report on Government Services 1999. 70. IEHCN, Progress to Date, East Melbourne, August 1999. 71. DH&CS, Hospital Services Report: December 1995; and DHS, Hospital Services Report: December 1999. CHAPTER 10 Autonomizing a Hospital System: Corporate Control by Central Authorities in Hong Kong Winnie C. Yip and William C. Hsiao In 1991, Hong Kong transformed its public hospital network into a single, statutory, nonprofit public corporation, the Hospital Author- ity (HA). The newly created authority became an autonomous fund- ing and management agency for both government-owned and non- profit hospitals. The reform resembled the other programs described in this volume, in that the I-HA received many important decision rights previously held by the government bureaucracy. In contrast to the other reforms, however, Hong Kong did not decentralize control. Rather, the creation of the HA changed the nature of the centralized control, transforming it to resemble somewhat the corporate head- quarters for the hospital network. Despite the HA's resemblance to a corporate structure, the relations between the government and the public hospital network remained essentially administrative in nature. The nature of the mechanisms for accountability for the HA and the resulting incentives continued to rely on the assessment of government officials. Direct public funding continued to provide the lion's share of the budget, and was not tied to actual delivery of services. Hence, incentives for performance of the I-IA were not connected to market exposure. 391 392 o Innovations in Health Service Delivery ,ObEe l0. Health Expenditures in 1986-87 and 1996-97 (total health expenditures, millions of U.S. dollars) EXPENDITURE 1986-87 1996-97 Total health expenditures 1 ,600.8 7,498.3 Public 1,027 1 4,029.1 Private 573.7 3,469.2 Total health expenditure as percentage of GDP 3 7 4.6 Total health expenditure per capita 289 7 1,188.1 Source- W Yip, R. Rannan-Eliya, and W. C. Hsiao, Health Core Expenditure Estimates for Hong Kong: The Domestic Health Accounts, Special Report 1 to the Hong Kong Govern- ment, January 1999 Health status in Hong Kong compares well with other advanced economies, despite lower health spending. In 1996-97, Hong Kong spent less than 5 percent of its gross domestic product (GDP) on health, US$1,200 per capita (table 10.1). Health care is financed through a combination of public (54 percent) and private (46 percent) sources. Public health expenditures are financed from general tax rev- enues, and most private health expenditures are financed through out-of-pocket household expenditures. Privately purchased health in- surance and employer-provided medical benefits pay for the rest. The Hospital Sectior before Reforms In the mid-1980s, hospital care was provided by 44 hospitals with a total of 24,600 hospital beds, most of them in the public sector. This translated into 4.5 beds per 1,000 inhabitants. Prereform, three types of health care facilities provided inpatient care: government hospitals (46 percent of all beds), nonprofit hospitals (43 percent), and private hospitals (11 percent).' Government Hospitals In the mid-1980s, the government operated 13 hospitals, ranging in size from 15 beds to more than 1,900, and numerous small health centers, usually with fewer than 25 beds. Most of these facilities were Autonomizing a Hospital System * 393 general hospitals. Hospital beds were classified as beds in general wards and beds in a small percentage of private wards. Patients in general wards had no choice of physician. Treatment costs were charged at a flat daily rate, subsidized at around 97 percent of treat- ment costs. Patients in private wards were either university doctors' private patients (who paid between 100 percent and 150 percent of actual costs) or civil servants (who paid a reduced rate). Nonprofit Hospitals Private nonprofit hospitals were established by charitable and volun- tary organizations. In the niid-1980s, there were about 20 such hos- pitals ranging in size from 80 beds to more than 1,500 beds. Origi- nally, nonprofit hospitals received only small government subsidies to complement their own funding but, by the mid-1980s, most of their operating budgets came from the government.2 Similarly, the government covered most of their capital expenditures, contributing between 5 percent and 10 percent of the total capital cost. Private, For-profit Hospitals In 1980, there were 11 private hospitals, ranging in size from 80 beds to more than 500. Private hospitals cover their own operating and capital costs, and hence their patient fees were (and still are) almost 10 to 40 times higher than in the public sector. Overall governance of the hospital sector was divided between the Health and Welfare Branch (HWB) and the Medical and Health De- partment (MHD). The HWB was in charge of setting overall health sector policy, and the MHD was the executive arm in charge of run- ning the public health care delivery system. By the mid-1980s, the MHB had grown into one of the largest bureaucracies in the Hong Kong public administration system, employing 24,330 people, 13 percent of the civil service. Its two main functions were to directly manage the government hospitals' 21,337 beds and to supervise nonprofit hospitals.3 Since the nonprofit hospitals had come to play an integral role in providing services to public patients, they were considered part of the public hospital network. 394 o Innovations in Health Service Delivery Issues Confronting the Hospill Sector in Ihe 1980s Three interrelated problems contributed to widespread dissatisfac- tion with inpatient care: overcrowding of government hospitals, dis- parity in conditions among the three types of hospitals, and ineffec- tive and fragmented hospital governance. Overcrowding The size and distribution of the hospital sector was not in line with the increasing needs and expectations of the populace. Rising de- mand was driven predominantly by large-scale demographic changes in the postwar period. The population grew by 63 percent, from 3.17 million inhabitants in 1961 to 5.18 million in 1981, and this growth was spread unevenly across the region.4 The New Ter- ritories, where many new immigrants settled, experienced the fastest growth, but hospital facilities were concentrated in Hong Kong Island and Kowloon. In 1984, the distribution of hospital beds per 1,000 inhabitants was, respectively, 5.63, 4.49, and 3.78 in Hong Kong Island, Kowloon, and the New Territories. These trends resulted in overcrowding, long waiting lines, and heavy workloads in public inpatient care facilities.5 Regional government hospitals had the highest occupancy rate-an average of 92 percent between 1980 and 1985, peaking at 100 percent in 1982-unparal- leled in the nonprofit or private for-profit sectors.6 To ease bed shortages, canvas camp beds were frequently placed temporarily in hospital corridors. Government versus Nonprofit Hospitals The second main source of dissatisfaction originated in the disparity in conditions and funding among the different types of hospitals. There was a divide between the government-run and nonprofit hos- pitals. Disgruntled physicians were constantly moving out of the public and into the private, for-profit sector because of poor working conditions, lack of training opportunities, and little hope of promo- Autonomizing a Hospital System * 395 tion.7 The public hospital network itself was split between the non- profit and the government-run hospitals-which were better funded. Physicians and other staff at government-run facilities were em- ployed as civil servants and received pay and fringe benefits such as housing, vacation, sick pay, training subsidies, and pensions. Staff in nonprofit hospitals did not enjoy the same benefits and working con- ditions, even though their facilities were also publicly funded. As a result, nonprofit hospitals had a hard time hiring and retaining well- qualified professional staff. Hospital Sector Governance The third issue was related to the public hospital sector's dual gover- nance structure, which undermined overall health policy coordination and planning. The Medical and Health Department and the Health and Welfare Branch were more rivals than collaborators, competing for greater control in the sector.8 As a result, coordination was poor between the policy design and implementation processes. The situa- tion was described by K. L. Thong, former MHD director: This illogical separation of responsibilities between the policy branch and its executive department is as frustrating to the De- partment concerned as it is untenable to the Branch since the Branch officers are placed in the unenviable position of having to deal with and even make pronouncements and decisions on policy matters on professional subjects of which they have no knowledge or expertise ... [A]dministrators of the medical and Health Service are no longer in a position to formulate policies and to take the vital and necessary actions for the smooth run- ning and improvement of the services without intervention from both administrative and political quarters.9 The lack of an effective governance structure also contributed to ongoing problems with financial discipline, and many hospitals ran deficits. Running a deficit was sometimes viewed as a means to en- sure a bigger budget allocation for the next financial year. 396 o Innovations in Health Service Delivery Reform Design Based on the recommendation of the Scott consultancy report, Hong Kong decided to create a single body-the Hospital Au- thority-responsible for coordinating hospital care across the en- tire public network, including both government and nonprofit hospitals.10 Organizational reform of public hospitals in Hong Kong was heavily complemented by a transformation of hospital management. External Environment As noted in chapter 1, hospital organizational reforms consist of ac- tions on three fronts: governance (relationship between public hos- pitals and the government), the extent to which public hospitals are exposed to market forces and the market environment, and the incentives embedded in the funding/payment mechanisms. The arrangements in these three areas joindy determine the incentive regime and hence the behavior of public hospitals under organiza- tional reforrn. Hong Kong's reform entailed limited (or no) change in the last two factors. The establishment of the Hospital Authority shifted the locus of decisionmaking from the government to the HA, which in turn established a centralized planning mechanism to man- age and control individual hospitals. Here we focus on the broad changes in the determinants of in- centive regimes. Specific changes in relation to the different incen- tive regimes are described later in more detail. The Hospital Authority was created by an Ordinance of the Leg- islative Council in July 1990. It was legally constituted as a nonprofit public corporation responsible for managing public hospitals on behalf of the government. The Hospital Authority formally took over responsibility for all public (government and nonprofit"1) hos- pitals on December 1, 1991, including 38 institutions and more than 37,000 staff members.12 Autonomizmg a Hospital System * 397 The Hospital Authority's objectives were set forth in the Ordinance: The Authority shall... manage and develop the public hospital system in ways which are conducive to achieving the following objectives- (i) to use hospital beds, staff, equipment and other resources efficiently to provide hospital services of the highest possible standard within the resources obtainable; (ii) to improve the efficiency of hospital services by developing appropnate management structures, systems, and performance measures; (iii) to improve the environment in public hospitals to meet the needs of patients; (iv) to attract, motivate, and retain qualified staff; (v) to encourage public participation in the operation of the public hospital system; and (vi) to ensure accountability to the public for the management and control of the public hospital system. Governance. Accountability of the HA to the government is exercised through a board, as well as the hospital governing committees (HGCs). The board, supported by its functional committees, is re- sponsible for guiding and overseeing the chief executive and his ex- ecutive team in formulating and implementing policy strategy. The chief executive and his team carry out executive functions at the head office; the hospital chief executives and their management teams carry out their fimctions at the hospitals. Payment/funding mechanism. The reform design provided for the HA to have the freedom to set prices, which would have given them ad- ditional potential income from user fees, and shifted the balance to- ward private payment. However, due to strong opposition, the gov- ernment backed away from this idea. Hence, the method and source 398 o Innovations in Health Service Delivery of funding between the government and the public hospital network remain unchanged. The Hospital Authority continued to receive 97 percent of its funding directly from the government budget. Market environment. The market environment within which the HA was established exhibited little competition. The reform in fact insti- tutionalized the monopolistic structure of the hospital market, since it established the Hospital Authority with 89 percent of total beds. Competition between public and private hospitals is limited. Nor does any competitive measure such as contracting or competitive bidding introduce competitive pressures into the market. Prior to the reform, competition had emerged in some input markets, in par- ticular for skilled medical staff. The establishment of the entire pub- lic network as a single organization (and monopolistic purchaser) reduced the competition in the markets for inputs. The Incentive Regime The governance and funding arrangements combined with the mar- ket environment to create the incentive regime faced by the HA upon its creation. The elements of this regime are discussed below. Within the HA, governance and incentives for individual hospitals also changed. This is discussed separately in box 10.1. Bon 10.1 Hlospital Aulwriity Pleltions wvih IndMdual alospials In addition to the organizational changes in the relation- ship between the government and the public hospital net- work as a whole (in the form of the Hospital Authority [HA.), there were equally critical changes occurring in the relations between the individual hospitals and their new owner-supervisor. Compared with the previous direct ad- ministration of government hospitals, the HA was granted considerable autonomy in its operation and decision making. Yet individual hospitals saw their autonomy decline with the Autonomizing a Hospital System ° 399 rules imposed by the HA head office, in particular through the annual planning process. Tables 10.2 and 10.3a and b summarize the change in the governance structure before and after the reform of the public hospital network. As corporations go, the HA is quite centralized-the head office retains direct control over many aspects of op- eration (service and manpower planning, service coordina- tion, operational organization, monitoring, and evaluation). Within the corporate structure, however, the HA head office has allocated some decision rightv to hospital manage- ment. For example, a hospital chief executive is allowed to makes appointrnents to all positions except senior manage- ment posts reporting directly to him.a However, hospitals are constrained in their hiring and firing decisions by the annual planning process, which establishes the service plans and the associated manpower by rank at each hospital, as agreed with the HA head office. The HA head office must endorse any increases in the agreed number of senior posi- tions.6 For more junior positions, hospitals have greater flexibility and discretion, as long as they stay within the agreed overall budget. With the exception of consultant and nurse specialist positions, creation of a new permanent position does not require the HA head office approval as long as the numbers agreed in the manpower plan are not exceeded. To ensure quality professional staff, the HA has established minimum standards anld job requirements for all ranks. Decisions about capital acquisition are jointly deter- mined with the HA head office during the annual planning process. Upon agreement between individual hospitals and the HA head office, the latter submits proposals to the gov- ernment for approval. The individual hospitals, like the HA itself, are residual claimants on only a minuscule amount of revenue-they (Box continues on thefollowing page) 400 o Innovations in Health Service Delivery Box 1O.1 (coninued) retain a portion of the user fees they levy on a small number of services. Individual hospitals face little market exposure, since the major share of funds is still allocated according to a baseline budget. Accountability mechanisms within the HA are, naturally, administrative and hierarchical in nature. The hospital chief executive is accountable to the Hospital Governing Commnit- tee (HGC) and the HA chief executive. The HA head office holds regular meetings between deputy directors, cluster managers, and hospitals, where hospital chief executives re- port on key results and performance targets agreed in the an- nual plan and service agreements. The hospital chief execu- tive and the senior management team initiate hospital-level planning. Senior management identifies parameters, which are a combination of prior agreement with the HA head of- fice and the hospitals' own initiatives. Planning must adhere to the broad product lists and programs previously agreed and defined between the hospitals and the HA head office. Before the establishment of the Hospital Authority, hos- pitals focused almost exclusively on their own activities and services, without viewing themselves in the context of the entire public system or the greater community. There was less inclination to work within the confines of the annual budget through increased efficiency or to strive for broader societal goals. Although such management and control mechanisms represent improvements from the prereform era, in prac- tice, effectiveness of the different monitoring mechanisms is largely constrained by availability of data and information. To hold each level of management accountable, reliable, standardized, and timely evidence is needed on the extent to which performance targets, service outputs, and quality Autonomizing a Hospital System 0 401 standards have been met. The current structure puts the onus on subordinates to report up the hierarchy. Financial data at the end of each year are clearly black or red, but the many gray areas in performance and quality outcomes require an effective means of monitoring. This is not to imply that hospitals attempt to deceive their supervising bodies with inflated outcomes, but to show the difficulties of measuring these intangibles by one party, let alone each and every hospital. A related point is the apparent absence of penalties if hospitals do not measure up. a. Hospital chief executives are selected by a board made up of members of the HGCs and the chief executive of the Hospital Authority; they are appointed by the HA chief ex- ecutive. Approval of the HGC is required' for the appoint- ment of senior management posts directly reporting to the hospital chief executive. Hospital Authority, Hzman Resource Policies Manual, chapter Cl, July 26, 1994. b. For example consultant, chief pharmacist, senior clini- cal psychologist, senior medical officers. Decision rights. A shift in decision rights (the magnitude of control or- ganizations have over various aspects of their production process) from the government to the HA was a central part of the reform. The HA gained considerable control over various aspects of the pro- duction process, including:'3 o Strategic management. In contrast to the prereform era, the locus of decision rights in public hospitals' strategic planning and manage- ment has been clearly transferred from the government to the Hos- pital Authority. The HA takes a major role in formulating institu- tional visions and objectives. One of the HNAs foremost tasks is to formulate the corporate plan, setting out long-term strategies and providing guidance in planning services. 402 o Innovations in Health Service Delivery The corporate vision, as developed by senior management at the HA head office and as endorsed by the HA Board and the Health and Welfare Branch, is as follows: "The Hospital Authority will collabo- rate with other health care providers and caregivers in the commu- nity to create a seamless health care environment which will maxi- mize health care benefits and meet community expectations."'4 0 Outputs and scope of activities. Corporate directions and strategies, as stated in the corporate plan, are operationalized through the HA planning process, the main planning framework. The HA service product list, which spells out the hospitals' volume and mix of new program initiatives for the coming year, is formulated with input from the government, the board, the hospital governing committees, the specialist service coordinating committees, and the community. Input from hospitals is also sought through communication with hos- pital senior management and front-line staff so as to align values and priorities in the definition of service products. With the list of product items and quantum for each in hand, the HA head office works with hospital chief executives in drawing up the plans for individual hospitals. During this process, the quantum might be adjusted to reflect budget constraints, overall service re- quirements, and local hospitals' aspirations. Through this process, resources are allocated to the various services for each hospital, and the HA service product list is finalized. o Inputs. Extensive decision rights over labor were transferred to the HA, including the right to hire and fire its own staff and set the terms of employment. Appointment of principal officers, including the chief executive, needs the approval of the chief executive of the Hong Kong Special Administrative Region (HKSAR). In terms of firing, the HKSAR chief executive has approval authority to dismiss the chief executive; the chief executive, to dismiss staff members of the HA head office and hospital chief executives; and hospital chief exec- utives, to dismiss hospital staff members. Previously, when all hospi- tal staff members were civil servants, hiring and firing decisions were Autonomizing a Hospital System * 403 done through the civil service, according to uniform criteria (i.e., no differential criteria for physicians). In terms of capital input, the government still allocates land. The HA submits proposals to the government for the amount required. Proposals for capital projects over HK$15 million (US$2 million) are first submitted to the Health and Welfare Branch for considera- tion. 15 Those endorsed by the HWB are then considered under the government's Capital Works Reserve Funds allocation exercise and the Legislative Council. Projects between HK$100,000 and HK$15 million (US$13,333 and US$2 million, respectively) are funded under the budget of the government's Architectural Services Department. The Hospital Authority initiates each capital project proposal for government funding support. Items below HK$100,000 are consid- ered recurrent items. Rules for procurement are clearly laid out in the HospitalAuthority Procurement and Matenals Management Manual. The HA received expanded authority to procure pharmaceutical products. Though bulk supply contracts continue to be arranged through the government supplies departrnent, for items not included in the bulk contract arrangements, individual institutions purchase drugs to meet their needs. * Financial management. As noted above, the funding arrangements were not altered. Hence, decision rights related to raising additional resources were not granted the HA. The Hospital Authority has no control over pricing, as the HWB must approve setting and changing user fees. Only under exceptional circumstances can the Hospital Au- thority borrow from the government to meet its obligations as spec- ified by the Ordinance. Should that occur, the Hospital Authority is subject to directions given by the secretary for health and welfare. Government finding continued to be administratively allocated. Each year, the Hospital Authority is required to submit its recurrent funding requirements to the Health and Welfare Branch. The re- current budget is based on the previous year's budget, adjusted for expected change in the service pattern, recurrent costs of new capi- tal projects, and the overall government budgetary situation.16 404 o Innovanons in Health Service Delivery Residual claims. The Hospital Authority became the residual claimant on only a tiny amount of hospital income; it is allowed to keep a re- serve of up to 5 percent of its budgeted annual recurrent expendi- tures.17 Any excess reverts to the government. Though the Hospital Authority may keep part of the additional revenue from new fees, these new fees can only be implemented under government ap- proval. Clearly, access to these funds as a motivation for improved performance is not part of the Hong Kong reform model, as, to date, the government allows the Hospital Authority to keep part of the revenue on only one or two items. Market exposure. The reform did not include structural changes to in- crease market exposure for hospitals in the public network. Because the Hospital Authority earns and retains so little of its revenue (only 3 percent of income) from payments for services, its market exposure remained low. The 97 percent of its income that comes from the government budget is not allocated in a strategic form, nor is com- petition supported; thus these funds do not expose the HA to any marketlike pressure either. Accountability. The Hong Kong hospital reforms created a board for the HA. Boards serve to hold organizations accountable in a more in- direct fashion than direct supervisory intervention on a day-to-day basis. Hence, in this regard, the reforms did move toward creating an accountability regime more congruent with a truly autonomous or- ganization. In most other respects, however, the accountability re- mained hierarchical and administrative in nature, relying on the di- rect reporting of subordinates rather than incentives generated by market exposure or payment mechanisms. o The Board. The reform provided for the HA to be accountable to the Health and Welfare Branch and the community through the board. The board is comprised of community leaders and govern- ment representatives. The HKSAR chief executive appoints its chairmnan. The board delegates authority to the chief executive of the HA head office to carry out the HA mission, goals, objectives, strate- Autononiizing a Hospital System * 405 gies, and purposes. The board monitors the chief executive's and the HA's performance through senior executive team reports on annual plan targets, progress reports with financial information, and an an- nual report with audited accounts. A Annual plan. Congruent with its administrative function, HA ac- countability revolves around measuring performance against the an- nual plan rather than being driven by incentives induced by market exposure or payment methods. Performance indicators relate to in- puts such as number of beds, implementation of clinical protocols, developing cost information system, rather than outcome or output. * Financial accountability. The HA is required to keep within its al- lotted budget from year to year, in keeping with its responsibility to hold the public hospital system budgetary targets. To exceed budget and secure additional funds, the Hospital Authority needs the Leg- islative Council's approval. Social functions. Marketizing organizational reforms usually require that provisions for social functions are made more explicit, to ensure that increased emphasis on financial performance doesn't undermine delivery of services to the poor and other needy patients.18 Since the incentives in the Hong Kong reforms did not rely on market expo- sure or financial performance, it was not necessary to change the mechanisms for ensuring delivery of social functions. Complementary Reforms In keeping with the administrative nature of the reforms, significant emphasis was placed on improving how things worked inside the hospitals. Management reform in particular was an important com- plement to the organizational reform in Hong Kong. A key structure put into place during the initial phase of the reform was the New Management Initiative. The main objectives were to establish a proper management infrastructure and the requisite systems in hos- pitals to facilitate service delivery, strengthen management capability 406 o Innovations in Health Service Delivery in hospitals, cultivate the necessary organizational culture and values consistent with modem management concepts, and achieve improve- ments in fulfillment of the Hospital Authority's mission.19 (See tables 10.2, 10.3a, and 10.3b.) The New Management Initiative was implemented in hospitals in three phases over three years. They included: o Clinical management teams, comprising the chiefs of service, department operations managers, ward managers, and the busi- ness and administrative support staff o Strategic planning o Business planning, setting out annual operational targets based on longer term strategies identified during the strategic planning process 7abie ¶0.2 Characteristics of Governance in Hospitals before Corporatization UKE CENTRAL tIKE FULLY BUDGET UNIT PRIVATIZED UNIT CHARACTERISTIC 1 2 3 4 Decision rights Accountability arrangements Residual claimant status M Market exposure Social functions Tabile 10.3a Characteristics of Governance after Corporatization: Government and the Hospital Authority LIKE CENTRAL UKE FULLY BUDGET UNIT PRIVATIZED UNIT CHARACTERISTIC 1 2 3 4 Decision rights M Accountability arrangements M Residual claimant status M Market exposure M Social functions M Autonomizing a Hospital System * 407 Table 10.3b Characteristics of Governance after Corporatizationr Hospital Authority Head Office and Hospitals LIKE CENTRAL - LIKE FUiLY BUDGET UNIT V7 PRvATIZED UNIT CHARACTERISTIC 1 2 -3 4 Decision rights 6 Accountability arrangements Residual claimant status Market exposure Social functions v * New management structure, comprising the hospital chief execu- tive and general managers for finance, nursing, clinical services, allied health services, and administrative services * Staff development review * Continual quality improvement * Operating systems, including new technologies and systems in support services, financial management, and management infor- mation (e.g., standardized accounting systems, internal audits, human resource, and payroll system). Implementafion of Reforms The reform process extended over five years from commissioning the Scott Report to public consultation, legislation, and the actual implementation. Context Hospital reform in Hong Kong focused on changing the structure of hospital administration and supervision, rather than opting for a move to create areas of market incentives and performance pres- sures. Why did Hong Kong choose this direction when countries with similar public sector traditions and less-entrenched laissez-faire economies (the United Kingdom and New Zealand) decided to im- 408 o Innovatons in Health Service Delivery plement more radical changes in their publicly run health sectors? Hong Kong embarked on this reform process earlier than other countries. The idea of establishing an independent hospital author- ity crystallized in the mid-1980s when ideas for public sector man- agement incorporating market structures and pressures were begin- ning to take hold, but application of these reforms to health services was relatively novel and untested. Therefore, these did not enter into the discussions in Hong Kong at the time of the debate. At the time, it was strongly felt that the public hospital sector's most egregious problems were the lack of management expertise and the rigidity of the system's structure, which hampered its ability to adapt to the changing environment. Moreover, poor coordination across facilities and institutional inefficiencies were the main con- tributory factors to a disorganized public hospital system.20 In addi- tion, incentives to perform for individuals and organizations were not seen as particularly problematic. Consequently, organizational reform of an administrative nature was seen as the solution.21 Actors Although the general concept of the reforms within the public hos- pital sector was accepted, not all the actors in the system supported specific elements of the proposed reforms. In March 1986, the Ex- ecutive Council opened a public debate that focused on the overall framework, the financial costs, and benefits of setting up the Hospi- tal Authority, integrating government and publicly funded nonprofit hospitals, and the costs of transferring staff into non-civil-service terms of employment. Witthin the government, the Medical and Health Department and the Health and Welfare Branch supported the Scott Report recom- mendations and the idea of organizational reform.22 Both parties had expected that it would allow them greater control over the hospital sector. In contrast, the Finance and Civil Service Branches were con- cerned about the financial implications of the proposed changes. In particular, the Finance Branch feared the cost increase implied by the new terms of service for staff in the reformed hospitals under the new authority, and insisted that the switch be cost-neutral. Autonornizing a Hospital System * 409 Medical staff members in government hospitals opposed the re- forms for fear of losing civil service benefits. The Medical and Health Department did its own review and received generally neg- ative feedback. By July 1986, 10,000 hospital staff members had signed a petition opposing the report recommendations. They formed a joint council of 35 hospital staff associations and strongly opposed removing staff from civil service as well as the report, which they thought "failed to submit concrete proposals which would im- prove ward services and meet the basic needs of patients. "23 Medical staff in nonprofit hospitals, however, supported the pro- posed changes. Primarily, their objective was to secure more re- sources while maintaining a high degree of autonomy. On this basis, they welcomed the proposed elimunation of inequities in employ- ment conditions and status. Yet they were not enthusiastic about the idea of a new independent supervisory body that might control re- sources and internal management more stringently than the Medical and Health Department. Boards of nonprofit hospitals were con- cerned about losing their charitable or voluntary identities. The community at large maintained a low profile during the de- bate and held mixed opinions. The public-as-patient sought relief from overcrowding and declining service standards, while the pub- lic-as-taxpayer was anxious lest changes and improvements in the health care system entail increased fees, higher taxes, or reduced ac- cess. Opponents were generally concerned that the government would withdraw from its responsibility of providing medical services. At the end of the day, the government was able to persuade the pub- lic that the reforms would not mean reducing the scope of govern- ment responsibility, which somewhat lessened opposition to reforms. Process W'hile the reform process followed a top-down approach, the long duration provided plenty of opportunity for the government to con- sider various points of view and reach compromises. At the same time, however, disagreements did not deter the government from proceeding with its plan. After the public consultation process, seri- ous disagreements remained with government medical staff, but ini- 410 o Innovanons in Health Service Delivery tial steps to establish the Hospital Authority were taken. The Provi- sional Hospital Authority was established in 1988 and was chargecl with developing a detailed implementation plan based on the Scott Report's recommendations, outlining strategies for merging staff of the nonprofit and government hospitals, and constructing a legisla- tive framework for the proposed Hospital Authority. Its three main tasks were related to addressing concerns over employment condi- tions, level of cost recovery, and staffing for the administration. The dispute over employment conditions was the main stumbling block in the process of establishing the Hospital Authority. More than 22,000 employees in civil service status under the Hospital Ser- vices Department were meant to switch to Hospital Authority em- ployment with new terms and conditions. Unions demanded better payment conditions against the insistence of the Finance Branch that the switch be cost-neutral. Since the proposed benefits meant more equitable salaries and benefits for nonprofit hospital staff, cost-neu- trality could have been achieved only at the expense of the benefits of Medical and Health Department medical staff. Finally, after a pro- tracted bargaining process that lasted beyond passing the HA Ordi- nance, the government agreed to a compromise. The compromise entailed a benefit package comparable with civil service terms and a grandfather clause to allow Hospital Services Department employees a choice of either accepting the new HA terms or continuing em- ployment with the civil service. Staff members were allowed three years to decide whether to transfer to HA terms and conditions or remain on civil service and nonprofit terms. Ninety percent of non- profit hospital staff opted for new terms while only 24 percent of for- mer Medical and Health Department staff were willing to give up civil service benefits. At the end of the three years, only about 2 5 per- cent of HA employees remained on civil service status. The second issue was related to the level of cost recovery. Origi- nally, increased cost recovery had been an integral part of the Hos- pital Authority's financial plan, but these efforts were shelved when public pressure groups strongly protested fee increases. In response, the government decided that the Hospital Authority would not be permitted to set fees and charges. Any user fee changes had to be Autonomizing a Hospital System * 411 cleared with the Health and Welfare Branch. Thus, the Hospital Au- thority became almost entirely reliant on government financing. The final concern was related to HA staffing issues in recruiting acceptable personnel for executive positions at the head office and as hospital chief executives. It was decided to offer attractive remuner- ation packages to recruit and retain qualified staff with managerial acumen.24 Assessment of the Impact Assessing the impact of organizational changes on hospitals is a chal- lenging task. First, in many instances, there is lack of time-series data to make pre- and postreform comparisons. Second, the identification of a valid control group is not always feasible to tease out the pure ef- fect of organizational changes from other developments affecting the health sector. Third, reform measures usually take time to change the behavior of key players, thus our assessment is necessarily pre- liminary. Given these caveats, we attempt to marshal available indi- cations of performance change. Several aspects of hospital behavior and performance will be ex- amined. The first two relate to addressing problems specific to the pre-HA hospital sector, namely, to reduce overcrowding and address provider dissatisfaction. Our findings suggest that the Hospital Au- thority has made large strides in meeting both objectives. The sec- ond two performance measures include quality and efficiency im- provements-the ultimate objectives of hospital reform in all countries. Our findings suggest that certain aspects of quality have improved while others remain to be addressed. In terms of efficiency gains, evidence is still inconclusive. The HA has expanded its own direct procurement and now pro- cures about 30 percent of its own pharmaceutical needs. Coordination among hospitals seems to have improved as well. Before the establishment of the Hospital Authority, hospitals focused almost exclusively on their own activities and services, without view- ing themselves in the context of the entire public system or the 412 ° Innovations in Health Service Delivery greater community. There was less inclination to work within the confines of the annual budget through increased efficiency or to strive for broader societal goals. Human Resource Issues One of the main problems before the Hospital Authority's creation was the public sector's difficulty in keeping good doctors in the sys- tem. Beyond pay and working conditions, medical staff morale is an important contributor to effective, efficient, and high-quality health services. The Hospital Authority has placed great emphasis on creat- ing improved conditions for medical staff. As part of the management initiatives, staff development review was implemented to promote manager-staff communication and to manage performance. At the same time, sizable resources were injected to construct new facilities and renovate existing ones to improve the working enviromnent. For example, all hospitals are now air-conditioned. In addition, the HA has been able to use its ability to offer attractive remuneration pack- ages to get needed management skills. There is now indication that HA staff satisfaction has become better: attrition rates have been de- clining since the early 1990s (table 10.4). Financial Discipline Financial discipline in Hong Kong's health sector has undergone considerable improvement since the establishment of the Hospital Ta;lOe 1 0.4 Hospital Authority Attrition Rates (percent) PRACTITIONER 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 Doctors 8.7 7 2 6 8 6.7 7.0 4.6 Nurses 84 82 10.9 93 86 57 Allied health 8 6 8.1 10 3 9.1 6.1 4.8 Hospital Authority, overall 10.3 12.3 13 8 8 2 7 2 5 4 Source Hospital Authority. Autonomizing a Hospital System * 413 Authority, which has so far adhered to balanced budgets. To enforce financial discipline at the hospital level, the HA head office uses its centralized powers to continually monitor hospital-level financial performance. Should hospitals foresee a shortfall due to unforeseen circumstances, they are expected to make the HA head office aware of the situation as soon as possible and plan, as agreed, to manage the shortfall. This could mean additional resources released from HA head office reserves or service reduction to contain or eliminate the shortfall (other hospitals may take up additional services or capacity). This is a big improvement over the prereform era problems with fi- nancial discipline. Accountability Accountability in the system as a whole appears to have improved. However, problems remain as well. In particular, effectiveness of the different monitoring mechanisms is largely constrained by the avail- ability of data and information. To hold each level of management accountable, reliable, standardized, and timely evidence is needed on the extent to which performance targets, service outputs, and quality standards have been met. The current structure puts the onus on subordinates to report up the hierarchy. Financial data at the end of each year are clearly black or red, but the many gray areas in perfor- mance and quality outcomes require an effective means of monitor- ing. This is not to imply that hospitals attempt to deceive their su- pervising bodies with inflated outcomes, but to show the difficulties of measuring these intangibles by one party, let alone each and every hospital. A related point is the apparent absence of penalties if hos- pitals do not measure up. Overcrowding and Access Overcrowding in major public hospitals has significantly declined. Occupancy rates in selected major hospitals declined to below 80 percent from the previously extremely high rates, and camp beds have been less commonly seen. Of course these improvements can only be partially attributed to the reforms, since the number of beds 414 0 Innovations in Health Service Delivery available has also increased. The opening of additional public sector beds raised the ratio of total beds per 1,000 population from 4.4 in 1990 to 4.7 in 1996. Quality An assessment of the specific aspects of quality change reveals the following conclusions: the Hospital Authority has succeeded in im- proving certain aspects of service quality, but improvements in other quality-related aspects have not yet fully materialized (tables 10.5 and 10.6). Hospital environment is the highest-rated quality compo- nent of patient assessments. However, improved facilities could be attributed to the increased funding that accompanied corporatiza- tion rather than organizational change. There is general consensus that interpersonal aspects of medical care, including attitude of med- ical staff, responsiveness to patients, and communications with rela- tives, have improved, in part because of the more patient-centered focus. Waiting time remains a major source of dissatisfaction among Tabe 10.5 Patient Satisfaction with Selected Aspects of Hospital Authority Services, 1997 SATISFIED DISSATISFIED SERVICE/FACILITY (PERCENT) (PERCENTI Physical environment 71.4 4.7 Medical facilities 70.8 4 1 Quality of medical care 67 6 6.7 Attitude of nonmedical staff 65.5 8.0 Attitude of medical staff 64.7 10 8 Waiting time for accident and emergency treatment 43.0 36 3 Booking time for operations 34 7 43.1 Queuing time for SOPD appointment 34 6 50 4 Waiting time for SOPD treatment 33.8 48.9 Quality of food 40 7 24 0 SOPD Specialist outpatient department. Note: Sample consisted of Hospital Authority service users, which may lead to upward bias in results Source Hospital Authority, Patient Satisfaction Survey, April 17, 1998 Autonomizing a Hospital System * 415 Table 10.6 What Do Public Hospitals Most Need to Improve? 1995 1996 1997 SERVICE/FACILITY (PERCENT) (PERCENT) (PERCENT) Queuing time for the specialist outpatient clinic 24 8 24 3 22 5 Waiting time for the specialist outpatient clinic 14 6 14 3 10 6 Waiting time for A&E services 6 2 8.5 12 6 Attitude of medical staff 10 5 7 2 8 9 Quality of medical care 4 0 6 8 7 2 Physical environment 3 7 3 9 4 1 Queuing time for operation 3 7 3 3 3 9 Medical facilities 5 4 3 3 3 9 Attitude of nonmedical staff 1 6 1 5 2 4 Quality of food 2.1 33 2 2 Don't know/hard to say 23.4 23 7 21.6 1000 1000 1000 A&E Accident and emergency Source: Hospital Authority, Patient Satisfaction Survey, April 1 7, 1998 patients. Half the respondents were dissatisfied with the length of waiting for appointments, and 23 percent and 11 percent, respec- tively, of HA users indicated that queuing and waiting times needed to be improved (table 10.6). Although waiting time for first atten- dance in HA specialist outpatient clinics declined initially, it has since risen: from 8 weeks in 1992-93 to 11.4 weeks in 1996-97. The vol- ume of specialist outpatient visits has increased over the years, but this can only partially explain the rise in waiting time. Based on patient focus-group meetings, lack of physician choice in public facilities and lack of interface between public-private and pri- mary-tertiary care are further sources of user dissatisfaction. Despite the Hospital Authority's ongoing efforts to create a "seamless health care environment," evidence is by far less optimistic. For example, when patients are discharged from a hospital, they are referred to general outpatient clinics where they often see different providers on each visit; moreover, their records may not routinely follow them. Additional issues arise when patients are referred for care or follow- up across sectors, largely because of a serious lack of communication 416 o Innovatons in Health Service Delivery or information exchange between hospital-based and community- based practitioners. Patient feedback is another area of quality that has yet to show im- provement. In the HA Patient Satisfaction Survey, only 3 5 percent of the patients think that the Hospital Authority "cares about patients' concerns" or "responds to patients' needs." Over time, evaluations of these dimensions have not improved. The Hospital Authority has set up a hierarchical complaint system: complaints and appeals are orig- inally filed in the hospitals through the HA head office or through offices of Legislative Council members and are then referred to the Public Complaints Committee. However, decisions regarding ac- tions to be taken are largely dominated by medical professionals and HA staff. For example, the committee's membership consists of the chairman (who has to be a member of the Hospital Authority), and two other HA members. Patients interviewed by the study team ex- pressed a lack of confidence in the complaint process. Most com- plaints and appeals were found to be "unsubstantiated," and patients saw little point in going to the effort of lodging a formal complaint.25 In 1996-97, for instance, out of the 1,735 complaints filed at hospi- tal level, only 21 were referred to the HA head office's complaints committee, and only 1 was substantiated. Productive Efficiency According to international standards, changes in productive effi- ciency should be measured by cost per episode of illness, adjusting for case mix. Since the Hospital Authority has not collected histori- cal data, it was necessary to rely on proxy data such as cost per bed- day or discharge, which gave ambiguous results.26 The cost per visit decreased for specialist outpatient care and emergency services but increased for general outpatient care between 1992-93 and 1996-97 (table 10.7). For inpatient services, while the cost per discharge fell, the cost per bed-day increased. The weighted average of cost per output for all services declined when discharges were used as an out- put measure for inpatient services, but increased when bed-days were used as the output measure. Autonomizing a Hospital System * 417 Table 10.7 Trends in Recurrent Cost per Unit of Output (Real 1990 = 1 00) OVERALL ANNUAL COST PER BED-DAY/ATTENDANCE0 1992-93 1993-94 1994-95 1995-96 199697 GROWT INPATIENT (1) Cost per discharge Acute care hospitalsb 8,792 8,532 8,266 7,885 8,061 - Year-to-year growth (percent) - -3 0 -3 1 -4 6 2 2 -2 2 General and extended hospitals 11,734 12,483 13,664 14,928 15,809 - Year-to-year growih (percent) - 6 4 9 5 9 3 5 9 7 7 Psychiatric hospitals 111,822 111,709 111,387 120,210 118,286 - Year-to-year growth (percent) - -0 1 -0 3 7 9 -1 6 1 4 Overoll 10,042 9,859 9,602 9,279 9,481 - Year-to-year growth (percent) - -1 8 -2 6 -3 4 2 2 -1 4 (2) Cost per bed-day Acute care hospitalsb 1,585 1,596 1,639 1,621 1,643 - Year-to-year growth (percent) - 0 7 2 7 -1 1 1 4 0 9 General and extended hospitals 699 724 764 796 844 - Year-to-year growth (percent) - 3 6 5 6 4 1 6 0 4 8 Psychiatric hospitals 414 440 430 456 472 - Year-to-year growth (percent) - 6 1 -2 2 5 9 3 5 3 3 Overall 1,092 1,112 1,160 1,182 1,222 - Year-to-year growth (percent) - 1 9 4 3 1 9 3 4 2 9 OUTPATIENT Specialist outpatient (cost/attendance) 329 313 297 282 299 - Year-to-year growth (percent) - -4 9 -5 1 -5 1 6 0 -2 4 Accident and emergency (cost/attendance) 385 390 343 317 300 Year-to-year growth ipercent) - 1 3 -12 1 -7 6 -5 4 -6 0 General outpatient (cost/attendance) 108 120 123 114 123 - Year-to-year growth (percent) 11 1 2 5 -7 3 7 9 3 3 Community nursing (cost/attendance) - - 267 251 221 - Year-to-year growth (percent) - - - -6 0 -12 0 n a Outpatient overall growth (weighted, using cost per discharge for inpatient- without community nursing) -1 8 Outpatient overall growth (weighted, using cost per bed-day for inpatient- without community nursing) 1 6 -Not applicable, n a Not available a Cost per bed-day/attendance is calculated based on the method used for fees and charges purpos- es, which includes staff cost, drugs and consumables, depreciation, services provided by government departments, and allocated share of overhead and supporting services b Mentally hondicapped beds are deducted from acute care hospitals and included in the calculation for psychiatric hospitals c Too few years Source Compiled by authors based on data provided by the Hospital Authority 418 o Innovations in Health Service Delivery We examined the trends in output and input separately. Table 10.8 shows the rate of increase in output disaggregated by different HA services. An overall output growth indicator-generated by weight- ing the relative share of services in total HA expenditures-suggests that output grew by 4.9 percent (using the total number of hospital days) or 7.5 percent (using the number of discharges for inpatient ser- vices). In terms of input, the annual average growth rate amounted to 3.7 percent for hospital beds, 5.2 percent for HA employees, and 10.6 percent for capital expenditures, if construction of new facilities is not taken into account (in real terms). Since growth in inpatient admissions is made possible with new bed openings, the 10.6 percent growth in capital expenditures underestimates the resources input to produce the hospital admissions. Among HA employees, the great- est growth rate was among administrative and management staff, as a result of modernized management techniques introduced after the Hospital Authority's formation. Another staff category that experi- enced a high growth rate was consultant physicians. Such trends are consistent with several hypotheses. First, previ- ously long stays27 have been reduced or replaced by day surgery, leading to cost reduction per discharge, but cost per bed-day in- creased, since patients who cannot be shifted to the ambulatory or community settings are more severely ill. This could be an improve- ment in efficiency. Second, the reduction in lengths of stay is achieved at the expense of premature discharge, leading to medical complications and higher readmission rates. Third, the increase in cost per day is due to changes in labor mix or labor and capital mix with no commensurate increase in output. For example, while the number of physicians grew by 6.2 percent between 1992-93 and 1997-98, the number of consultants grew by almost 14 percent. Fourth, the increase in bed-days is due to quality improvement. These hypotheses, of course, are not mutually exclusive. To verify these hypotheses, data on case mix, readmission rates, cost per bed-day for day surgery versus inpatient stay, quality changes, and the like are required. Data made available to us do not allow dif- ferentiation among these hypotheses; hence, whether production ef- ficiency has improved in the Hospital Authority remains ambiguous. 0' N C~~~ w 0' c 0c N 0 N 0g 's O m > 0 '0 . 0 l 'o so ( Cl) sn V > z) >t v N S ts( -_ O K, S No x, , -'0 2~~~~~C 'O i V Cl 0, °, - - S s C4 Y CN o E & & Y Y Y E o O~~~~0 °, ° Vioz°p°° zo N CN4 N N N C - 4 9 04) 0 'O L M l o O ~~ ~~ N CN co 0 O m 0 0 N 0 ol o co U' C ) N 0 'q co 0,C - N - ('4: - N, .2~ Ln CN N co) a, L o ' '0 ('I N N - '0 0 '0 '~~~~t NO C Ne N v No 0 CIL~~~~~~~~~~~~~~~~~~0 o,- 6 Q - 0 0 C (N "' C., C - 0I.Lo IC fl '00 N C E" 0 co C14~N - -0 o~~~~~~~~~~~~~~~~~~ coa a,a a, a, a,) ol 0 C C C4 C 0 0 0,~~~~C - 0 Ep Ep "'2 a, 0 00 0 _0 o __ mZ>Z)I> O<- I> . 66 419C C 420 0 Innovations in Health Service Delivery Besides, costs presented in table 10.7 reflect only recurrent expendi- tures. Capital investment in technology and information systems that allow for substitution of more day surgery and elimination of unnec- essary length of stay per discharge are not included.28 Con&usions Like the other reforms in this volume, Hong Kong's hospital reforms involved organizational change. That is, the reforms entailed struc- tural changes that shifted rights, responsibilities, and functions around the system.29 And, like the other reforms, they created autonomy by transferring many decisions away from the government bureaucracy to a subordinate unit. But this is where the similarities end. Perhaps because their public sector functions at a relatively high level, incentives for performance were not viewed as a core problem in the Hong Kong public hospital network. Thus, while the reforms did delegate much autonomy to the HA, they did not directly address the issue of the incentives that would determine what the HA would do with its autonomy. Evidently, it was intended that accountability would continue to be derived from line authority combined with a new focus on meeting performance targets. This somewhat discor- dant structure appears to have been effective on a number of fronts. Combined with a direct and sustained initiative to improve manage- ment, the organizational reform succeeded in improving some as- pects of the public hospital system. The HAs strategy of mimicking the corporate model in their op- eration and relations with subordinate hospitals seems to have re- lieved the public hospital network of some of its more egregious problems. Relative to the previous supervisory structure, these mechanisms have proven effective in improving the performance of public hospitals. In particular, the annual planning process makes production much more objective and results-oriented and makes some explicit linkages to inputs and outputs. Some lessons can be gleaned from Hong Kong's reform experi- ence. First of all, in addition to structural reform, management re- Autonomizing a Hospital System * 421 form appears to have been crucial to Hong Kong's success. When or- ganizational changes give hospital management new decision rights and responsibilities, sound management is essential throughout the transition phase and beyond. Significant financial and human re- sources have been needed to develop and implement management structures and processes for annual planning, continual quality im- provement, information systems, managerial recruitment, staff de- velopment, and standardized accounting systems. Second, as decision rights were delegated to the Hospital Author- ity, the government's ability to assert direct accountability dimin- ished, while the new performance monitoring system seems not to have fully compensated for this. The existing structure and mech- anism for accotmtability rely solely on reporting by subordinate units-the Hospital Authority to the Health and Welfare Branch, or the individual hospitals to the HA head office-on performance measured against the annual plan, with few incentives imposed by market pressures or incentive-based payment mechanisms. The ef- fectiveness of this structure is further constrained by measurable out- come and data availability. During the initial set-up period, most available indicators focused on outputs such as utilization measures instead of health outcomes and efficiency. Although efforts are now being made to develop measures of clinical outcomes, how such in- formation is being collected, used, and fed back into the planning process is unclear. Finally, Hong Kong's experience underscores the peril of hospital reforms designed without adequately taking into consideration wider health system policy and performance. The design of the Hong Kong reform was (hospital) sector-specific and ignored the relation- ship with other parts of the health system, namely, private hospitals and primary care (85 percent of primary care is delivered privately). Thus, the reform further institutionalized the segmentation of the health system as a whole. The reform was not accompanied by any financing reform. While the cost of services at public hospitals re- mains highly subsidized and the quality and scope of services have improved under the reform, the Hospital Authority experienced a rapid surge in demand. Eight years into its operation, the Hong 422 - Innovations in Health Service Delivery Kong government's burden to finance health care services is growing. The delivery system is highly fragmented, with limited coordination between the public and private sectors or between the tertiary and primary/community sectors, despite the Hospital Authority's mission to create a seamless health care system. Organizational reform of the public hospital sector must always be considered in relation to other key players and components of the entire health system. Notes 1. R. Gauld and D. Gould, "The Hong Kong Healthcare System" (Otago, New Zealand: University of Otago Press, 2000). 2. In Hong Kong, private, nonprofit hospitals are referred to as "subvented" due to their reliance on govermnent funding. 3. A. B. L. Cheung, "The Politics of Administrative Reforms in Hong Kong: Corporatization of Public Services During the 1980s," (Ph.D. diss., London School of Economics, 1995), chap. 4, p. 71. 4. Hospital Authority, Statistical Report 1995/96, Statistical Infor- mation Section, Hospital Authority, 1996. 5. Cheung, "Politics of Administrative Reforms," p. 72; R. Hutcheon, Bedside Manner: Hospital and Health Care in Hong Kong (Hong Kong: Chinese University Press, 1999), chap. 2, p. 42. 6. W D. Scott, The Delivery of Medical Services in Hospitals. A Re- port for the Hong Kong Government (Hong Kong: Coopers & Ly- brand, 1985). 7. Hutcheon, Bedside Manner, p. 34. 8. Cheung, "Politics of Administrative Reforms," pp. 78-80. 9. Ibid., p. 78. 10. Scott, The Delivery of Medical Services in Hospitals. 11. While the subvented hospitals retained ownership of build- ings, their management and control was transferred to the Hospital Autonomizing a Hospital System - 423 Authority, and from then on they received funding on the same basis as government-owned hospitals. 12. Hospital Authority, Annual Report, 1991/92, Hong Kong, 1992. 13. Other complementary management reform initiatives are dis- cussed in more detail in W. Yip and W Hsiao, "Organizational Re- form in the Public Hospital Sector: The Case of Hong Kong," Har- vard University, Cambridge, Mass., 1999, processed. 14. The derivative strategies are: (1) developing outcome-focused health care to maximize health benefits and meet community ex- pectations; (2) creating seamless health care by restructuring and reorganizing medical services in collaboration with other providers and caregivers in the community; (3) involving the community as partners in health in the decision-making process; (4) cultivating organization transformation and development through a multidis- ciplinary team approach to holistic patient care and continual qual- ity improvement; (5) promoting corporate infrastructure develop- ment and innovation to support service improvement. 15. Each proposal would include building works costs, funding for furniture and equipment, and recurrent consequences. 16. The administrative costs of the Hospital Authority's head of- fice are also included in the overall budget, based on existing head of- fice salaries and other costs as well as planned new administrative costs that have to be agreed on with the HIWB. Hospital Authority, PC-P53, Overview of Government Funding of Hospital Authority Ser- vices; P647: Budget Allocation 1998/99. 17. Based on the draft Memorandum of Administrative Agree- ment between the government and the HA. 18. See chapter 1, this volume, for a further discussion of social functions and organizational reforms. 19. Hospital Authority, Initial Plan to Sutstain the Organization Transformation Following the Current Phase of the NMI Effot, " Hong Kong, p. 1. 20. Hutcheon, Bedside Manner, p. 41. 424 o Innovations in Health Service Delivery 21. Experience and problems encountered in implementation are documented in Yip and Hsiao, "Organizational Reform in the Pub- lic Hospital Sector." 22. Scott, The Delivery of Medical Services in Hospitals. 23. South China Morning Post, March 27, 1986, no page. 24. For a more detailed account of the implementation process and problems encountered, see Yip and Hsiao, "Organizational Re- form in the Public Hospital Sector." 25. Submission from the Alliance for Patients' Mutual Help Or- ganization, 1999. 26. Output is measured by admissions, bed-days, and visits. 27. The average length of stay fell from 9.2 to 7.2 days for general specialties between 1992-93 and 1996-97, and 12.3 to 10.1 days for all specialties (general and psychiatry, mentally handicapped, and infirmary). 28. Government funding to the Hospital Authority consists of three block grants: one for all recurrent expenditure that HA can de- ploy flexibly, retaining any unspent funds in its reserve up to a max- imum 5 percent of its budgeted expenditure, and two capital grants specifically for plant and equipment and information technology. 29. See chapter 1, this volume, for further elaboration on the dis- tinction between organizational and other types of hospital reforms. CHAPTER 11 Corporatization of a Single Facility: Reformling the Malaysian National Heart Institute Rozita Halina Tun Hussein, Syed Al-junid, Soe Nyunt- U, Yahaya Baba, and Willy De Geyndt In September of 1992, the Government of Malaysia corporatized its national referral cardiac center, the National Heart Institute (NHI). Unlike the other cases in this volume, the Malaysian reform was ap- plied to a single hospital, one that had been recently established. Fol- lowing successful application in the state-owned enterprise sector, corporatization was the chosen model to reform the hospital sector as well-where cost escalation was seriously straining the budget. Perhaps because of the benchmark of public enterprise reform, the government opted for the relatively far-reaching reform of corpora- tization-extending management considerable autonomy, enshrined in independent legal status. Revenue was retained in the facility; ser- vice lines were largely decided upon there as well. Accountability was provided for indirectly, via a board of directors. Social functions were partially made explicit, though a good deal of cross-subsidization continued to be required to cover the cost of services provided to poor patients. All in all, the Malaysian reform represents one of the rare cases of hospital corporatization in a developing country. 425 426 o Innovations in Healti Service Delivery As it appears to have been relatively successful, it merits thorough scrutiny by others contemplating a sirmilar path. The one notable proviso is that the reform has proven successful only as applied to a single, newvly established hospital, which makes it hard to draw conclu- sions for policymakers who are considering applying the reform to a larger number of facilities. Context and Background Since independence in 1957, Malaysia has made impressive gains in delivering quality health care services to its diverse and diffuse pop- ulation (21.7 million). Malaysians' life expectancy has increased dra- matically (table 11.1).1 With improved health care and increasing numbers of children surviving to adulthood, people 59 years or older will make up more than 13 percent of the population by 2025. Malaysia's changing demographic and health profiles have im- portant implications for the quality, cost, and scope of health care services. Chronic illnesses are already the chief medical problem, one that puts different demands than in the past on hospitals and doc- tors.2 Doctors need expensive, sophisticated, imported medical tech- nologies for diagnosis and treatment. Recently, infectious diseases such as tuberculosis-especially drug-resistant strains-and malaria have reemerged, and HIV/AIDS poses additional challenges.3 The Malaysian government has always viewed the development of the health sector as an integral part of the country's socioeconomic development. As part of longstanding efforts to alleviate poverty, the government has targeted the poor and underprivileged in delivering health care and other public services. The strategy of ensuring equity and access to health care for all originated in the Rural Health Ser- vices (RHS) initiative of the 1950s to provide health care to 75 per- cent of the population in rural villages. The Ministry of Health (MOH), with other social sector agencies, undertook massive construction projects to establish the physical in- frastructure for its first rural health services. Next came services in Corporatization of a Single Facility * 427 Table 11.1 Socioeconomic and Demographic Indicators, Selected Asian Countries, 1996 INDICATOR MALAYSIA INDONESIA THAILAND REP OF KOREA JAPAN Population (millions) 21 197 60 46 126 GNP per capita (U S dollars) 4,370 1,080 2,960 10,610 40,950 Health expenditure (percentage of GDP)a Total - 1 5 5 3 54 70 Public -1 4 0 7 1 4 1 8 5 5 Private - 0 8 3 9 3 6 1 5 Births (per 1,000 population) 27 23 17 15 10 Deaths (per 1,000 population) 5 8 7 6 7 Age dependency ratio 0 7 0 6 0 5 0 4 0 4 Annual growth rate (percent)b 1 6 1 3 0 6 0 7 0 1 Life expectancy at birth (years) Male 70 63 67 68 77 Female 74 67 74 76 83 Infant deaths (per 1 000 live births) 11 49 35 9 4 Access to sanitation (percentage of population)c 91 51 70 100 85 Access to health core (percentage of population)d 88 43 59 100 100 Physicians (per 1,000 population)d 0 4 0 2 0 2 1 2 1 8 Total fertility rate (per woman) 3 4 2 6 1 8 1 7 1 4 -Not available a 1990-95 b 1996-2010 c 1995 d 1994 Source World Development Indicators, 1998, World Bank, Washington, D C ,1998 cities and towns. Expanding distribution of health care facilities con- tinually improved geographical access. Economic access was assured when people who could not afford to pay were given care free of charge. Because of these accomplishments, other developing coun- tries look to Malaysia's public health care system as a model. The Sixth National Plan (1991-95) further improved the distri- bution of facilities. Efforts were made to provide affordable health care, enhance the quality of curative and preventive programs, and train more manpower for the health sector. The Seventh Plan (1996-2000), which brought in reforms including the corporatiza- tion and privatization of public facilities, should be seen in this con- text of continued commitment to social policy. 428 o Innovations in Health Service Delivery Financing and Organization of the Health Sector Public sector During this period, the MOH became one of Malaysia's largest ministries, employing some 84,000 people, more than 10 per- cent of the civil service work force, and managing a budget of US$1.21 billion, equivalent to 1.22 percent of GDP in 1996. In 1996, total health sector expenditure was estimated at 3.73 percent of GDP, compared with the 5.0 percent recommended for develop- ing countries by the World Health Organization (WHO).4 A pyramidal organizational structure at the national, state, and dis- trict levels ensures a hierarchy of accountability and responsibility. Specific responsibilities were delegated to officers at each level of the hierarchy. Technical and nontechnical managerial functions were in- tegrated, and formal lines of communication were established among the three levels. Private sector Recent estimates show that the private sector's role in Malaysia's health care industry is large and growing. In 1995, the number of private general practitioner clinics was estimated at 5,000 nationwide.5 The number of private hospitals has grown steadily from 50 in 1980 to an estimated 241 today. Most private facilities are urban-based and profit-driven, concentrating on high-return cura- tive care with some preventive care (e.g., immunization against child- hood diseases). Since Malaysia does not have a national insurance program, private services are paid for out of pocket or via third-party payment for patients with employment benefits or private medical insurance. The Malaysian Medical Association has established a nonbinding fee schedule for the private sector. The nominal fees in the public sector act as a counterbalance to curb rampant fee escala- tion in the private sector. Nonetheless, health care costs throughout Malaysia have risen steadily (table 11.2). Escalating costs are associated with corporate investment and the increasing availability of sophisticated health care services.6 Almost 70 percent of Malaysia's pharmaceuticals are imported. In 1998, Par- liament passed the Private Healthcare Facilities and Services Bill. This comprehensive bill was one of the first explicit regulatory at- Corporatization of a Single Facility * 429 Table 1 1.2 Number and Type of Health Service Facilities in Rural and Urban Malaysia, 1996 INDICATOR PUBLIC PRIVATE TOTAL Midwife clinics/community clinics 1,998 n a 1,998 Rural health centers 609 n a n a Hospitals 123 241 364 Hospital beds 35,881 8,138 44,019 Beds (per 1,000 population) 1 7 0 4 2 1 n a Not applicable Source Information and Documentation System, Ministry of Health, 1996 tempts by a developing-country government to introduce checks and balances on the private sector. The population's total out-of-pocket health expenditure is esti- mated at US$1.52 billion (1.53 percent of GDP in 1996). Most of this goes for private treatment (61.6 percent of the total), primarily ambulatory care (62 percent of the total). Spending on hospitaliza- tion, the next largest category, amounts to only 15 percent of the total, as 80 percent of all hospitalizations occurs in the highly subsi- dized government facilities. The Underlying Pressures for Reform Even before the 1997 Asian economic crisis, the rising cost of health care made reform an important issue. It was feared that unchecked cost escalation in the public sector could jeopardize Malaysia's im- pressive accomplishments in providing quality, low-cost health care to all segments of its population. Malaysia hoped to build upon that system by downsizing and corporatizing state-owned hospitals, by encouraging the development of private hospitals and clinics, and by moving basic care from hospitals to local clinics to improve effi- ciency and reduce subsidies from middle- and upper income citizens. Health sector reform. Policymakers recognized that, to meet the chal- lenges of the 21st century, Malaysia's health system had to be re- structured. Its public and private components had to be integrated 430 o Innovanons in Health Service Delivery into a coordinated system. It was felt that health financing had to be further focused on equity and social objectives. Delivery, whether public or private, had to be decided by service quality and efficiency. The government has long demonstrated a commitment to health sector reform. Health financing strategy. The government recognized that reform of health care financing had to underpin the future health sector. As the biggest provider of health care, the MOH had to play a pivotal role in designing the new system, guiding and being guided by the much- needed restructuring and reorganization of public sector health de- livery. It could not afford to lose the initiative to the private sector. Health delivery reform. To meet society's demands, the government also recognized that the health care delivery system had to be up- graded. The MOH has embarked on selective upgrading of such services as hospital support and bulk distribution of drugs and phar- maceuticals within the MOH through the privatized Govermnent Medical Store. Upgrading the ambulance service through privatiza- tion is planned. For medical and nursing care at MOH hospitals, the ministry is emphasizing corporatization rather than privatization in order to maintain public ownership. Role of government and health regulation. The Seventh Malaysia Plan assigned the MOH an enlarged regulatory and enforcement role and a reduced role as a direct provider of health services. Functionally in its new role-protecting consumers and safeguarding the public in- terest-the MOH was to emphasize regulating prices and tariffs and enforcing service standards and quality.7 Corporatization and Privatization in the Broader Public Sector Discussion regarding corporatization and privatization in Malaysia goes back to the First Outline Perspective Plan (OPPI) for 1971-90, which set the broad socioeconomic framework for achieving the ob- Corporatization of a Single Facility * 431 jectives of Malaysia's New Economic Policy (NEP). The govern- ment implemented NEP by increasing the public sector's participa- tion in the economy. As a result, public expenditure rose eightfold between the Second and Fourth Five-Year Plans, putting a heavy burden on the government's financial resources and widening the public sector budget deficit. Policy adjustments during the 1980s attempted to prevent the deficit from worsening. Hopes were high that greater private sector development would drive economic recovery. Privatization of many government services brought a new dynamism to the economy, im- proving efficiency. Pnvatization was also supposed to reduce the public sector's financial burden by having the private sector take over services traditionally delivered by the public sector. The government adopted the "Malaysia Incorporated" concept in 1983 as a frame- work for tightening cooperation between the public and the private sectors. The Second Outline Perspective Plan (OPP2) of 1991-2000 continued this strategy by opening new opportunities for the private sector to participate in the country's economic recovery. Corporatization and Privatization in the Health Sector The marketizing reforms in the health sector occurred against this socioeconomic backdrop. As mentioned earlier, the private health sector had already been growing, first in cities and large metropoli- tan areas, then in smaller towns, partly through active encourage- ment by the government. In the early 1990s, the private health sector, especially private hos- pitals, grew rapidly in the main urban areas. At the same time, the govermnent initiated a few projects to corporatize or privatize some aspects of the public health services, in line with the overall policies of the 2020 Vision and Mission for Health and broader economic policy. To date, three major public services have been either priva- tized or corporatized: * The govermment corporatized the newly established National Heart Institute in 1992. 432 o Innovations in Health Service Delivery o The Government Medical Store (GMS) was privatized in 1994. The service is now provided by Remedi Pharmaceuticals (M) Sdn. Bhd., a member company of the United Engineers Malaysia Group. o The government privatized many hospital support services through contracting out, including clinical waste management service, cleaning service, linen and laundry service, facility engineering maintenance service, and biomedical engineering maintenance service. Three different companies, under concessions, began to provide these services in January 1997. University teaching hospitals were included under the corporati- zation of tertiary educational institutions. As described in chapter I of this volume, corporatization, unlike privatization, is often selected as the reform modality because governments hope to free hospitals of rigid bureaucratic and administrative controls while retaining the ultimate control through public ownership. The idea is to promote responsiveness to communities and patients through the increased flexibility and efficiency of operating within a competitive open mar- ket system. These arrangements are expected to enhance quality of care and service while stimulating innovation and creativity. For ex- ample, NHI is expected to play a substantial role in driving down the cost of cardiac care, particularly for routine procedures. Corporatizatioon of the taional Heart Institute The National Heart Institute (NHI) is a 286-bed specialized cardiac hospital on a campus of 60,300 square meters. It houses eight wards, an intensive care unit, an intensive coronary care unit, and a 24-hour emergency department with an observation unit. It has four cardio- vascular laboratories for invasive and interventional cardiology pro- cedures and five operating theaters for open-heart, closed-heart, and thoracic surgeries. It was intended to be financially self-sustaining in its operation as well as to meet future expansion needs. All support Corporatization of a Single Facility * 433 services are provided at NHI by various departments under the Medical Division. The hospital provides outpatient treatment for referral patients, pre- and post-operative consultation for cardiothoracic patients, and follow-up on cardiology cases. Outpatients may also come in for noninvasive investigations and consultations. Patients with unstable angina and myocardial infarction are treated as inpatients. Diagnos- tic and therapeutic invasive procedures are part of the inpatient ser- vices. NHI pledges to deliver medical care, conduct research, and provide education in keeping with the highest standards of ethics and quality; to provide comprehensive and compassionate patient care in cardiothoracic medicine; and to excel in attention to patients. Structural Rfrnm of NHI NHI started operations in July 1992 as part of Hospital Kuala Lumpur and was corporatized on September 1 the same year. As the first public entity in the health sector to be corporatized, prevailing government policies (developed in application to the state-owned enterprise sector) heavily influenced the nature of the objectives, namely: obtaining private sector efficiency gains, tapping private re- sources, and freeing up the government's financial resources for other important needs. Among the many reasons for corporatizing NHI-, one notable ob- jective was to improve the retention of highly trained and specialized personnel within the public health care sector. Many doctors, nurses, and paramedics had been leaving for the private sector where work- ing conditions were more attractive and the pay higher. Many of these staff members were initially trained under government schol- arships or in highly subsidized public institutions. This exodus of skilled personnel from the public system caused concern about the quality of care offered to patients at government facilities, most of them poor and disadvantaged. Legal status and ownership. During the broader public sector reforms in Malaysia, a public holding company was established under the 434 o Innovations in Health Service Delivery Treasury to hold equity in corporatized companies. Upon corporati- zation, the NHI became one of its holdings. Decision rights. Corporate status gave NEI increased management autonomy in the areas of service delivery, personnel, and procure- ment. NTI is completely responsible for which services are provided to its customers and how these services are provided. For example, NHI has more freedom in using its financial re- sources than noncorporatized public hospitals. The chief executive of- ficer can authorize purchases of medical equipment up to RM50,000 (US$13,000). Purchases between RM50,000 and RM500,000 must be authorized by the Management Committee, and those above RM500,000 (US$131,500) need approval by the Board of Directors. Construction financing must also be endorsed by the Board of Direc- tors. In a public hospital, major expenditures need approval by the MOH and, in some instances, by the Ministry of Finance (MOF). NIH needs MOF permission only for expansion plans or purchases above RMv1 5 million (US$3.8 million). Heads of the various N1h departments also handle a large amount of autonomy and influence the overall running of the organization. Specialists can develop and implement new plans and strategies for their departments with relative ease. Submission of proposals and ideas for major changes only have to go as far the NHI Board of Di- rectors, which makes decisions fairly quickly. Other decision rights are still constrained. For example, the Board of Directors has more limited autonomy over human resource deci- sions such as hiring, firing, and paying workers, including starting salary for new workers. The board must obtain MOF approval to change staff salaries or pay bonuses and dividends, as well as to change any charges or fees. The MOF, in consultation with the MOH, sets service fees for both public and private paying patients. Patients in third-class wards are charged at cost; those in second- and first-class wards or rooms pay cost plus 5 percent and cost plus 10 percent, respectively. Any re- vision of this fee schedule is subject to government approval. The MOF approved NHI's first fee revision, in September 1995, and the Corporatization of a Single Facility* 435 fees were set inclusive of a dividend rate for its shareholders. NHI is in the process of negotiating new fees. Impetus for this effort stems from the MOH, eager to consider packaging more routine proce- dures on a case-mix or diagnosis-related grouping-evidence of some conflict in their roles as both payer and provider. Accountability arrangements. To exercise oversight over the relatively independent hospital, the reform established a nine-member Board of Directors, comprising senior officials from the Ministries of Fi- nance and Health, the private sector, three representatives from NHI (the chief executive officer and the two executive directors) and the chairman of the board. The MOF appoints board members, with advice from the MOH. Through this arrangement, direct control over service delivery in NHI was removed from the MOH, although it can still influence management of the hospital as one of the repre- sentatives on its Board of Directors. Board members need not nec- essarily have a medical background; they are seen as resource persons to help infuse NHI with a businesslike culture. The board is NHI's final decision-making body. The Board of Directors recommends the appointment of the chief executive officer (CEO), with MOF approval.8 The medical, admin- istration, and finance divisions are under the CEO's direct man- agement. A Management Committee, Audit Committee, and Pro- fessional Advisory Committee report to the Board of Directors. Recently, a new post of medical director was created to limit man- agement responsibility of the professional aspects of care, while the CEO concentrates on administration and finance. An external pri- vate firm (Ernst & Young) conducts the internal audit. Residual claimant status and market exposure. NIh derives revenue from private and public sources: private patients and private compa- nies paying for their employees as part of their employment benefits; and the government for treating patients who are civil servants and their dependents, government pensioners, and the poor. The government purchases services from NHI for patients for whom the government assumes responsibility, and the hospital bills 436 o Innovations in Health Service Delivery the MOH. The federal government has not assumed responsibility for the health care of employees of statutory bodies and local au- thorities since May 1995 or, since October 1997, for state govern- ment employees. The employers of these three groups pay NTI di- rectly for services provided. NHI has been billing the government on a fee-for-service basis. The growing workload and complexity of cases managed has resulted in steady annual increases in expenditure for these services provided to the government, although recently there seems to be some stabilization in this respect. Market environment. Although NTIi controls a large market share, it does not have monopoly power in the markets for most of its ser- vices. There are several additional government-owned cardiotho- racic centers. These include centers in Johor Bahru Hospital and Penang Hospital, and two centers are planned for the East Malaysia states of Sabah and Sarawak. In addition, some private hospitals offer cardiothoracic care. Nevertheless, NHI does possess monopoly power in some markets as the main center of excellence for more so- phisticated and complex services such as cardiac transplants, pedi- atric cardiac care, and conduction disorders. Social jfunctions. Unlike the other components of the reform, the so- cial functions were only partially dealt with. The NHM continues to be responsible in a general way for ensuring access to poor patients as well as those who can pay. Services to some poor patients are cov- ered via government transfers, but many others must be covered via cross-subsidies from fee-paying patients. Clearly, NHI does not have complete financial control over its affairs like a private entity. Its fi- nancial status can swing dramatically depending on the government's willingness to pay the cost of "social responsibility" services that the hospital provides for the poor. These swings necessitate unplanned- for changes in payment of salaries to its staff, maintenance, as well as service improvements. Dependence on substantial government financing to pay for ser- vices and 100 percent government ownership, therefore, impose cer- tain limitations on NHI's management autonomy and decision- Corporatnzatnoin of a Single Facility * 437 making rights. The hospital could reduce its dependence on such government financing by increasing reliance on fee-paying private patients. However, this would not be consistent with its implicit re- quirements to provide necessary care to all segments of the popula- tion, irrespective of socioeconomic status. The absence of a national health financing mechanism covering the whole population means that the NHI remains dependent on direct government subsidies and cross-subsidies to cover its social responsibilities. By reducing the overall financial control and management autonomy over its corpo- rate affairs, the financing mechanisms undermine the impact of the reform. The corporatization of NHI was initially intended as a pilot for similar reforms in other parts of the health sector. But the Treasury recognized that, until a new health-financing system is in place, cor- poratization of other health institutions had to proceed cautiously so as not to undermine the social protection for the poor, provided by directly operated health facilities. Issues Arising during Implementation A number of issues encountered during the corporatization of NTI in Malaysia are worth highlighting. Decision rights over human resource issues. Although the hospital itself was newly established in 1992, the staff was to come from the cardi- ology and cardiothoracic units of Kuala Lumpur General Hospital. The transfer of civil servants to the corporatized hospital was prob- lematic. Staff, especially the more senior members, worried about their service benefits (e.g., pensions and gratuities) after leaving gov- ernment service to join NHI. To ensure that NHI would have enough workers to start its service while observing the rights en- shrined in civil service legislation, the govermment developed a pack- age of options for the relevant staff. Government employees were given three options: Option A-resign from government service by taking early retirement and joining NTI; Option B-be seconded to NHI for two years while retaining civil service benefits; or Option 438 o Innovations m Health Service Delivery C-reject the offer to join as NHI staff and remain in the civil service. In September 1992, of the 315 staffs members offered the three options, 298 selected Option A; 1 chose Option B, and 16 preferred Option C. Thus, 95 percent of the staff opted to join NHI, mainly because NHI's salaries were higher than government's and, after a short probationary period, job security was good. The new jobs also offered retirement benefits at age 55, the pension depending on years of service years with NHI. Of NI-'s initial staff of 299, 13 were doctors. An active recruitment program during the first year of op- erations nearly doubled the staff to 459 by the end of 1993. Recently, there has been a move to hire employees on renewable contracts to increase flexibility in managing human resources. A key challenge encountered during the reform was to change management culture from that of a government civil service to that of an efficient corporation. The staff had to shed the civil servant image, often perceived as inefficient and unfriendly, and learn to work in a dynamic, consumer-sensitive environment. Another chal- lenge was to ensure that health staff and specialists remain in NHI and not leave for the more lucrative private sector. NaI- met this challenge by improving pay and work conditions compared with other public hospitals. Since November 1996, a small professional fee has been levied on private paying patients in first- and second- class wards. These fees are pooled and divided equally among doc- tors of the same specialty. Clinicians are also given opportunities for specialized training. Such human resource policies have not only succeeded in retaining the workers in NIS but also in increasing the total number of staff-by 30 percent between 1992-93 and 1996-97. Changing the scope of services and utilization patterns. Over the five-year period examined, invasive and noninvasive cardiology procedures in- creased by 78.2 percent and 111.4 percent, respectively. Of the 10,750 cardiothoracic surgeries performed at NHI, 68.8 percent were open-heart and 27.4 percent closed-heart surgeries; the rest were thoracic surgeries. The most common noninvasive cardiology procedures were echocardiography and electrocardiography, to- Corporatizanon of a Single Facility * 439 gether accounting for 75.3 percent of all noninvasive procedures. Coronary angiography was the main invasive cardiology procedure done in NHI; nearly half of the workload was associated with this procedure. Coronary artery bypass was the most common open- heart surgical procedure, and most closed-heart surgeries were car- ried out for patent ductus arteriosus. In 1997, NTI started the coun- try's only heart transplant program, and the first human heart transplant was performed on December 18, 1997. The number of patients using NIHI's services also increased steadily (table 11.3) during the same five-year period. The number of outpatient visits increased by 57.9 percent, and the number of ad- missions and patient-days increased by 62.5 percent and 38.6 per- cent, respectively. NHI functioned with 207 beds until 1996 when 4 intensive care unit (ICU) beds were added. In 1998, the bed com- plement increased with the addition of 26 beds in first class and 2 pe- diatric ICU beds. Even with more beds available, the increasing pa- tient load resulted in a higher bed occupancy rate (from 58 percent in 1992-93 to 76 percent in 1995-96). The mean length of stay was cut by more than a day (from 7.9 to 6.8 days). These results are even more impressive, considering that the case mix changed, with more sophisticated procedures performed. Changes in patient funding profiles. Patients treated in NHI fall into three groups, based on source of financing: civil servants and their dependents or pensioners; low-income patients; and private patients. The government pays medical bills for the first two groups. Patients Table 1 1.3 Operating Statistics for the National Heart Institute, 1 993-97 AVERAGE FISCAL NUMBER BED OCCUPANCY LENGTH OF NUMBER OF NUMBER OF NUMBER OF YEAR OF BEDS RATE (PERCENT) STAY (DAYS) PATIENT-DAYS ADMISSIONS OUTPATIENTS 1992-93 207 58.2 8 0 39,548 4,982 35,611 1993-94 207 61 9 76 46,151 6,125 39,265 1994-95 207 70 9 6 9 53,475 7,797 45,312 1995-96 207 75 6 7 1 57,255 8,081 50,442 1996-97 211 71 9 6 8 54,846 8,098 56,262 Source National Heart Institute 440 o Innovations in Health Service Dehvery who cannot afford to pay are certified as poor by medical social workers. For example, when NHI began operations in 1992-93, 81.4 percent of inpatients discharged were government-sponsored, and only 18.6 percent were private patients. Over the five-year period, the proportion of private patients increased to 30.2 percent. The same trend is observed for outpatient attendance. The proportion of the poor using inpatient and outpatient services decreased 7.8 per- cent and 6.5 percent, respectively, and use by private patients in- creased 11.6 percent and 9 percent, respectively. The decrease in the number of civil servants and poor using ser- vices has raised concerns that corporatization has led NHI to scrimp on its social responsibilities. However, a more thorough ex- amination leads to an ambiguous conclusion. First, in 1995 the cen- tral government stopped sponsoring employees working in local authorities and statutory bodies as well as all school-age children. While subordinate bodies are now responsible for these costs, it is not clear that their financing is as reliable-which may account for some of the decrease in demand for NHI services. Second, since 1995, two new MOH cardiothoracic units have undoubtedly taken some government-sponsored patients away from NHI. Third, the MOH now carries out more stringent audit checks to ensure that all necessary documentation is submitted before making payment for the poor. Because of these stringent checks, some of patients whose bills are submitted by NHI as poor patients are held back until full documentation is available. Lack of hard budget constraints. Although service charges are based on a government-approved costing formula, the open-ended, fee-for- service billing arrangement has been financially advantageous to NHI. However, whether the unit cost applied is measured optimally is questionable. Its salary structure, lower than in private hospitals, allows NHI to remain competitive by offering services for many pro- cedures for less than private hospitals. Partially as a result of these factors, operating revenue increased 108 percent, from RM41.7 million in 1992-93 to RM86.8 million in 1996-97 (table 11.4). Most of the increase in income came from Corporatizanon of a Single Facility * 441 government-subsidized patients, who comprise about three quarters of the total. Government's contribution increased from RM31.3 mil- lion in 1992-93 to RM65.6 million in 1996-97. Payment from pri- vate patients increased from RM10.4 million in 1992-93 to RM21.2 million in 1996-97. The sharp increase in the government's contribution has led to a criticism that allowing NTI to claim whatever expenses it incurred for public-sponsored patients does not encourage efficiency. Some of the increase, however, is from an expanded workload. Over the five- year period examined, the workload for outpatient and inpatient ser- vices grew by 58 percent and 63.9 percent, respectively. In the future, NHI may lose its competitive edge if it does not move toward bundling services and charging fees for service pack- ages, particularly when competing for contracts with large funders such as governments, managed-care organizations, or large compa- nies paying for their employees. Such packages are being offered by a private facility in Sarawak. Furthermore, balancing the bottom line with NHI's social responsibilities has not been easy. Clients have in the past been billed only after interventions were carried out. As a re- sult of private patients' reneging on their obligations, NHI has in- curred a cumulative debt of RM 1 million. Management is now at- tempting to recover some of these outstanding debts, and private patients are being asked to pay a deposit on admission or produce a guarantee letter from their employers. Table 11.4 National Heart Institute Operating Revenue, by Source of Financing, 1993-97 GOVERNMENT I CIVIL SERVANTS AND POOR PRIVATE PAYING TOTAL FISCAL YEAR DEPENDENTS (PERCENT] (PERCENT) (PERCENT) (MILLIONS OP RINGGIT) 1992-93 49 6 255 24 9 41 7 1993-94 54 6 25 6 198 59.0 1994-95 55 3 25 7 19 0 769 1995-96 53 0 26 9 201 770 1996-97 51 4 24 2 24 4 868 Source National Heart Institute 442 0 Innovations in Health Service Delivery Operating expenses increased from RM30.8 million in 1992-93 to RM77.3 million in 1996-97. Manpower and supplies were the major expense items throughout the five-year period; the proportion of ex- penses for manpower remains almost constant while expenses for supplies increased from 28.6 percent to 40.1 percent (table 11.5). The increase in expenditures for supplies is attributed to an increase in the range and intensity of specialized procedures being conducted by specialists who are returning from overseas training with greater expertise and a desire to undertake more aggressive and invasive car- diology procedures and pediatric cardiac surgery. Unclear residual claimant status over capital investments. Capital invest- ment is still borne by the government: building the hospital cost RM151 million (US$39 million), and the government allocated RM15 million the first year to start the operation. Since then, NH1 has been self-financing with payments from both private patients and the government to cover government-sponsored patients. Every year operating revenue has exceeded operating expenditures. Annually, NIH is expected to pay a minimum of 10 percent of the invested share capital as a dividend to the Ministry of Finance. The MOF derives the invested share capital of RM71.6 million from the TIaoe 1 l .5 National Heart Institute Operating Expenses, FY 1992-93 to 1996-97 (millions of RM [percent]) FISCAL MAINTENANCE/ UTIUTIES/ YEAR STAFF SUPPLIES DEPRECIATION REPAIRS COMMUNICATIONS OTHER TOTAL 1992-93 11 4 8.8 6.2 0.4 1 5 2.5 30.8 [37 0] [28.6] [20 1] [1 3] [4 9] [8 1] 1993-94 165 13.4 6.5 07 1.6 24 411 [40 1] [32.6] [15 8] [1 7] [39] [58] 1994-95 22 2 20.3 7.3 2 3 1.8 3 3 57.2 [38.8] [35 5] [12 8] [4 0] [3 1] [5 8] 1995-96 24.5 24 3 7.3 2 5 1.9 4.4 64.9 [37 8] [37 4] [11 2] [3.9] [2 9] [6.8] 1996-97 29.4 31 0 77 3 1 2.0 4.1 773 [38 0] [40 1] [10 0] [4.0] [2 6] [5.3] RM Malaysian ringgit Source National Heart Institute Corporatizanon of a Single Facility * 443 RM52.7 million of fixed assets, the RM15 million launching capital, and the additional working capital of RM3.9 million. The building and land remain in government ownership. At the end of the 1997 fis- cal year, the MOF had recouped about 30 percent of its invested share capital. No decision has been made on what to do when the original investment is paid back fully. Profits are subject to a 28 percent tax. Assessment of Impact Since its corporatization on September 1, 1992, no formal evaluation has been carried out to assess the impact of this organizational reform in delivering health services in Malaysia. Yet a better understanding about the positive and negative aspects of this type of reform could provide an important basis for further similar reforms of health ser- vices in the public sector. The following provides a first attempt at such an evaluation for NTI's managerial performance, financial re- sults, and clinical outcomes. It could form the basis for more in debt analysis in the future. Efficiency Gains A number of efficiency and productivity gains that can be docu- mented during NHI's first five years of operations under a corporate model can be attributed to the reform process (tables 11.6 and 11.7). As acknowledged by the Malaysian director-general of health, simi- lar changes and progress would not have been achieved had NHI re- mained in its original budgetary form. For ease of presentation, these efficiency gains are presented using a structure-process-output framework (table 11.7). The ratio of revenue to expenditure is decreasing, suggesting that NHI may be providing more services at cost and striving to remain a competitive player in an open market (table 11.8). During the first fee revision, most fees were lowered. The Privatization and Corporati- zation Sector of the Economic Planning Unit (EPU) is of the opin- ion that NTHI has become a benchmark for other facilities offering 444 o Innovations in Health Service Delivery TaIle 11 1 .6 National Heart Institute Performance Indicators, 1 993-97 (annual percentage change) BED OCCUPANCY AVERAGE PATIENT OUTPATIENT- YEAR STAFF INPATIENTS RATE LENGTH OF STAY DAYS VISITS 1994 8 1 22.9 6.4 -5.1 16.7 10.3 1995 87 273 14.5 -93 159 154 1996 4 1 3 6 6 6 +4.4 7.1 11 3 1997 5.7 02 -49 -42 -42 11 5 Source. National Heart Institute 7ablIe 11 1.7 National Heart Institute's Efficiency and Productivity Gains from Corporatization Structure o Governance by Board of Directors; chief executive officer recommended by board o Internal auditing performed by an external private accounting firm o Increased decision-making rights Process o Decision-making power over personnel matters o Authority to do own procurement of equipment and supplies o Authority to renovate and expand physical plant o Authority to retain all fees collected Output o More sophisticated clinical services provided, new technologies and techniques available more quickly under own decisionmaking authority and resources o Higher hospital bed occupancy rate and lower average length of stay; more efficient use of physical infrastructure o Increased revenue per patient/day o Constant number of staff but more patients treated, using more difficult procedures Table M1I .8 Total Revenue and Expenses per Inpatient-Day, 1993-97 YEAR REVENUE (RM) EXPENSES (RM) RATIO REVENUE/EXPENSES 1993 1,054 779 1 35 1994 1,278 890 1 44 1995 1,438 1,070 1.34 1996 1,345 1,134 119 1997 1,583 1,409 1.12 RM Malaysian ringgit. Source National Heart Institute Corporatization of a Single Facility * 445 similar services in the market and has decelerated the hitherto un- checked rapid rise in cardiac care fees. Revenue collected from private patients (25 percent of total rev- enue) is much higher than the corresponding rate in the traditional public hospitals (5 to 10 percent of the operating budget at a gov- ernment hospital; 5 percent of the total recurrent MOH budget).9 Patients refuse to pay fees in a public hospital because of the tradi- tion of free care in a public institution. The increased emphasis on collecting fees by NHI-I seems to have broken that connection in peo- ple's minds, sending a message to the public that payment for ser- vices is expected and required for population groups not covered by subsidies. Since the regional economic crisis started, many private hospitals have had fewer patients and substantial losses. NHI has been rela- tively spared from this drop in revenues partly because it is a highly specialized, tertiary care institution, with much of its work based on referral of complicated cases from all over the country. Equity Goals The Malaysian culture stresses equity of access to health care and ed- ucation. For this reason, the impact of the reforms on poverty is monitored closely and continually to ensure that no needy patient is refused services from NHI. A Patient Assessment Unit, housed within NEI and staffed by MOH personnel, evaluates the socioeco- nomic status of patients requesting government financial assistance. Though not strictly adhered to, guidelines have been established by the MOH for the number of poor who have to be allowed admission to the various ward classes. Reductions in the percentage of poor pa- tients receiving care have to be explained. Despite these guidelines, the number of poor patients treated by NHI, both as inpatients and outpatients, decreased over the five-year period following the reform. But the amount of money paid by the government for treating poor patients remained stable for the same time period (table 11.4). This may indicate that more complicated and more critical cases are being treated at NHI, while others go to 446 o Innovations in Health Service Delivery hospitals closer to the patients' residence (1IHlI has trained staff from hospitals in other parts of the country in cardiology and cardiotho- racic procedures). In the long run, out-of-pocket payment for critical care such as that provided by NHI is not a sustainable or socially desirable sys- tem. Corporatization reforms of tertiary care facilities such as that in Malaysia therefore need to be accompanied by parallel funding re- forms such as the introduction of a more comprehensive national health insurance (see chapter 1, this volume). Quality There has been no detailed assessment of the impact of the reforms on the quality of patient care provided by NHI. Ongoing consumer satisfaction studies by in-house staff members indicate a high satis- faction rate among patients, but the NHI managers feel that the study has limitations as guidance for quality-improvement efforts. A new patient-satisfaction questionnaire is being used and monitored by NMI's Quality Assurance Unit. The unit also monitors 12 quality assurance indicators, including such measures as outpatient waiting time and mortality. Despite indications that NTH specialists would like to assess the clinical outcomes of their services in greater depth, no such efforts have yet been undertaken. But NTI's Quality Coun- cil plans to initiate medical and third-party audits. Conclusions During the first years after corporatization, NHI has achieved its basic objectives: providing services to manage major heart conditions in Malaysia and substantially reducing dependence on foreign heart centers. The case-mix and socioeconomic background of patients in- dicates that the hospital has succeeded in balancing its role as a cor- poratized hospital with its social obligations. During these crucial initial years, NIHI has become the main provider of cardiovascular services in Malaysia. Its corporatized sta- Corporatization of a Single Facility 447 tus has enabled it to position itself within a relatively short period as a leading national and international medical care center. This status has allowed it to attract and retain dedicated and experienced pro- ductive health care workers. It is developing alliances with other hos- pitals to "export" staff expertise, both nationally and internationally. Similarly, both NHI and the government plan to actively promote the hospital to attract more clientele from neighboring countries. The hospital already receives patients from Brunei, Singapore, and Indonesia. To be truly self-sustaining, the hospital cannot continue to rely only on the limited funding provided by government subsidies for lower income patients. The perception is that the current 70-30 ratio of government-sponsored to private patients should increase to a 50-50 ratio. Achieving this ratio without adverse effects on equity will depend on the parallel introduction of more comprehensive health insurance. N-lHI is clearly using its flexibility to expand the scope and increase the quality of the operations. It is undertaking additional research and training. For example, it has already increased its role in research by conducting clinical trials. Staff members are also encouraged to engage in operational research related to clinical management and organizational aspects of health services and to strengthen their vis- ibility and reputation by publishing this work and participating in conferences and other training events. To help fund these activities, a foundation has been set up to manage donations from charitable organizations and individuals. Moreover, NHI has established links with such internationally renowned heart centers as the Mayo Clinic to stimulate collaborative research. In the case of training, NEI has initially concentrated on its own staff and some staff from other public university teaching hospitals and two private facilities. The hospital trains not only specialists but also paramedics such as their own nurses. In meeting its social obligations, NHI has initiated and financed a Heart Transplant Fund to provide assistance to needy patients who become transplant recipients. It is also working with the Malaysian 448 o Innovations in Health Service Delivery Heart Foundation (Yayasan Jantung Malaysia) in an "Adopt a Heart Transplant Patient" campaign to help transplant recipients with the high cost of their medical care. Acknowledgments: The authors would like to gratefully acknowledge the follow- ing individuals for their invaluable contribution to this case study: Tan Sri Dato' Dr. Abu Bakar Suleiman, director-general of health, Malaysia, Dato' Dr. Nor'aini bt Abu Bakar, CEO, National Heart Institute (NHI); Dato' Dr. Yahya Awang, head of the Cardiothoracic Department, NHI; Dato' Dr. Robayaah Zambahan, head of the Cardiology Department, NHI; Mr. Jamal Salhmin, deputy secretary-general of health (finance), MOH; Ms. Khor Lee Hian, Finance Division Manager, NPI; Mr. Noordin Mohd. Idrus, Human Resource Manager, NHI, Mr. Yusof Abd. Rah- man, Economic Plannng Unit (Privansation and Corporatization Sector), Pnme Mnuster's Department. The authors also thank Suzanne McLees of Bethesda, Maryland, for her assistance with this chapter. Notes 1. A. B. Suleiman, "A Script-Looking at Malaysia's Impending Healthcare Crisis," Healthcare Asia (First Quarter): 1-12 (1998). 2. The top five killers are cardiovascular ailments (20 percent), cancer (10 percent), cerebrovascular disorders (7.8 percent), motor vehicle accidents (6.5 percent), and septicemia (6.3 percent), accord- ing to the Ministry of Health (MOH) Annual Report, 1996. 3. Nearly 23,000 cases of HIV infection were recorded by the MOH between 1986 and September 1997. 4. R. H. Hussein and S. Nyunt-U, "Estimating Out-of-Pocket Health Care Expenditure in Malaysia" (paper presented at the Eighth Annual Public Health Forum, Reforming Health Sectors, London, 1998). 5. A. B. Suleiman, "The Approach of the Ministry of Health in Malaysia in Further Upgrading Quality in Health Services" (paper delivered at EDI(UNFPA/ICOMP/World Bank Seminar on Qual- ity of Reproductive Health Care as the Way Forward, Kuching, Malaysia, November 14-17, 1995). Corporatnzation of a Single Facility * 449 6. Economist Intelligence Unit, "Profile-Malaysian Healthcare: The Slow Road to Privatization," Healthcare Asia (Third Quarter): 67-76 (1998). 7. A. B. Suleiman, "Health Care Issues in Malaysia: A Review of Government Policy Priorities" (paper presented at the Economist Conferences, Health Care Systems in Asia-Challenges and Oppor- tunities, Singapore, January 21-22, 1997). 8. Dato' Dr. Nor'aini bt Abu Bakar has managed NHI since its in- ception, and her initial three-year contract has been renewed twice. A physician with a postgraduate degree in hospital administration, she served as a hospital director in Johor Bahru Hospital. 9. A. B. Suleiman, "Health Challenges Facing Malaysia." CHAPTER 12 Attacking Hospital Performance on Two Fronts: Network Corporatization and Financing Reforms in Singapore Kai Hong Phua From 1985 through 1998,1 Singapore implemented a far-reaching reform to many of its public hospitals. This reform followed closely the reform modality characterized in this volume as corporatization. The reform was undertaken in conjunction with equally extensive reforms to the hospital financing system. Singapore's ability to tackle organizational reform on two fronts was enabled by the high capacity of its public administration, as well as a political system that is rela- tively conducive to structural reforms. The hospital reforms are seen as relatively successful and are in- teresting on their own. However, it is not possible to analyze them separately from the financing reforms. This chapter therefore reviews the reforms together, distinguishing where possible the likely source of positive results. 451 452 o Innovations in Health Service Delivery Confext and Background Singapore's 24 hospitals have a total of 10,500 hospital beds, a ratio of 3.5 beds per 1,000 population. Eighty percent of those beds are in public hospitals whose bed complements range from about 200 beds to 2,500 beds. Most private hospitals are small, with 60 to 500 beds each. Public hospitals set the standard for medical care and hospital charges. Of the public hospitals, five are acute care general hospitals. The others specialize in areas such as obstetrics and gynecology, psy- chiatry, and infectious diseases. The public general hospitals provide multidisciplinary inpatient and specialist outpatient services and a 24-hour accident and emergency service. There are also specialist institutes for cancer, heart, eye, and skin diseases and the neuro- sciences. Tertiary specialist care (cardiology, renal medicine, hema- tology, neurology, oncology, radiotherapy, plastic and reconstructive surgery, pediatric surgery, neurosurgery, cardiothoracic, and trans- plant surgery) is centralized in two of the larger general hospitals, the Singapore General Hospital (SGH) and the National University Hospital (NUH). Private hospitals have similar specialist disciplines and comparable facilities. The government has also introduced low- cost community hospitals for intermediate health care for the conva- lescent sick and aged who do not require the more expensive care of the acute care general hospitals. As in many East Asian health systems, Singaporean public hospi- tals operate an internal system of cross-subsidization via tiered pric- ing.2 Individuals admitted to public hospitals choose their level of ac- commodations and, in so doing, a price regime. The more amenities they want, the more they pay. Thirty percent of the accommodations are private (one-bed) or semiprivate (four-bed) rooms, the most ex- pensive. These patients thus subsidize the remaining patients, who pay much reduced prices. Medical care in public hospitals is similar for all types of accommodations. With an average occupancy rate of about 80 percent, the available beds are well utilized. Attacking Hospital Performance on Two Fronts * 453 Institutional Reforms and Implementafion Issues In 1982, the government embarked on a plan to restructure the health care system in Singapore, principally through an alternative financing method, based on the Central Provident Fund (CPF) con- cept of compulsory savings (box 12.1). The idea of using CPF sav- ings for medical care had been mooted in the 1970s when the gov- ernment was looking at various options to finance the increasing costs of public services. Not until the appointment of a new minister for health in 1981, however, did the policy take shape in the form of a major national health planning exercise. The National Health Plan The National Health Plan (NHP), established in February 1983, in- troduced Medisave, a compulsory saving plan for medical coverage. The intent of the NHP was to change the direction of the Singapore system and avoid the pitfalls and problems of comprehensive health care systems elsewhere. The declared objectives of the NHP were: * To secure a healthy, fit, and productive population through active disease prevention and promotion of healthy lifestyles * To improve the health system's cost-efficiency * To meet a rapidly agmg population's growing demand for health care.3 The rationale for the NHP was that Singapore's health care sys- tem must "stand the test of time as the demand for hospital care will go up while the anticipated tax revenue may be expected to go down in relative tenns."4 The solution was seen to lie in a personal savings plan like Medi- save, with its philosophy of rewarding individuals for staying well. The plan also detached the quality of the health care system from complete dependence on the tax base, and hence from the vagaries of economic cycles. These objectives were consistent with traditional 454 o Innovations in Health Service Delivery Bon l2.M TimeDeine of Hlei CoaD Rerim M SingaPre 1982 March Medisave plan announced to Parliament by Minister of Health. 1983 February Blue Paper on the National Health Plan released. August Medisave approved by Parliament. 1984 April 1 Medisave implemented in every government hospital. May Plan to restructure the governance of government hospitals announced. 1985 June New National University Hospital (NUIH) incorporated as a subsidiary company under a government-owned holding company. Medisave introduced as a pilot in the first restructured goverment hospital NUH. 1986 January 1 Medisave expanded to include approved private hospitals. 1987 April Health Corporation of Singapore (HCS) formed to manage all restructured government hospitals. 1990 July 1 Medishield, a catastrophic insurance plan, implemented. 1992 January Parliament passed the Medical Endowment Act to introduce Medifund. February Report of Review Committee on National Health Policies accepted by government; Ministerial Committee on Health Policies formed to implement recommendations. 1993 April 1 Medifund, a health care plan for the poor, implemented. Attacking Hospital Perfornance on Two Fronts * 455 Singaporean values of self-reliance and strong family ties, promoted as the primary support for care of the sick and aged. Two common sayings in Singapore encapsulate these values-"Save for a rainy day" and "Charity begins at home."5 Medisave accounts, as they finally evolved, allow holders to with- draw money to pay their own and their family members' hospitaliza- tion costs, within certain limits. This additional source of funds was to enable increased private payment for individuals and their fami- lies. Medisave has, in effect, become an efficient and convenient method for collecting hospital bills. As prices rise to reflect increas- ing costs, it could also recover the costs of major development proj- ects in restructuring the government health sector. Individuals are encouraged to take responsibility for their own health by saving for expected future medical expenses. To reinforce this sense of personal responsibility, the health care system is built upon three health care financing programs that help people pay for medical expenses: Medisave, Medishield, and Medifimd. These three programs, discussed in more detail below, were designed to create a largely self-funded health care system, requiring people to look first to personal and famuly resources for health care and to rely on the government only after depleting their own resources. Medical Savings Singapore's health care financing reforms developed in three stages: Medisave, the medical savings component, came first (1984); then Medishield, a catastrophic insurance plan (1990); and Medifund, a medical endowment plan for the poor (1993). Medisave. Employees and employers each contribute 20 percent of the employee's wages to the Central Provident Fund, a national so- cial security and pension fund based on savings. From these funds, 6 percent of the employee's wages is deposited each year into the em- ployee's Medisave accournt until age 34. The percentage deposited increases to 7 percent between ages 35 and 44 and 8 percent at age 45 until retirement or until savings reach the ceiling of S$20,000 (1998 dollars). 456 o Innovations in Health Service Delivery Medisave has been modified several times, based on the experi- ence acquired. Initially, account holders could use their Medisave ac- counts to pay the full charges of a hospital stay in lower priced wards but only partial charges for the more expensive rooms. Now, almost all categories of hospital charges are covered, but within maximum daily limits and with the proviso that Medisave accounts used for higher priced rooms may not be overdrawn. Medishield. While Medisave account balances will usually cover hos- pital expenses, they are not sufficient to cover catastrophic expendi- tures. These events require risk pooling across individuals, that is, health insurance. Medishield provides this insurance. Premiums are automatically deducted from Medisave, unless account holders re- quest otherwise. Reaching the threshold for catastrophic coverage usually entails a long hospital stay or one of several costly, ongoing outpatient treatments such as chemotherapy for cancer. Three dif- ferent Medishield programs offer different levels of coverage, based on deductible or excess per policy year, claim limits per policy year, and claim limits per lifetime. Medifund. Though established by government endowment, Medifund is also based on the saving concept. Hospitals receive grants to defray needy patients' hospital bills from the interest on sums of money that the government provides from budget surpluses. Sufficient funds have built up from the substantial capital endowment since the prin- cipal sum is untouched. Requests for assistance are considered on a case-by-case, basis, and "low-wage Medisave/Medishield contribu- tors and elderly persons whose accounts are not adequate to cover ex- penses" receive preference.6 Singapore's experience represents a successful experiment in the development of medical saving accounts. Despite continually rising costs and demand, unnecessary expenditures for inpatient care have been reduced without dramatic limits on physicians' incomes or on access to high-technology treatment. However, Singapore uses a fairly narrow definition of services eligible for Medisave expendi- Attacking Hospital Performance on Two Fronts * 457 tures (e.g., excluding most outpatient care), has a fixed fee schedule for medical services, and does not have universal insurance. How- ever, universal access is guaranteed through a system of targeted sub- sidies and subventions from tax-based sources as well as the last- resort Medifund endowment for the indigent. While the medical saving accounts are a critical component of the Singapore health sys- tem, it is not the sole mechanism for financing care, nor is it viewed as the entire solution to all health policy problems. In Singapore, the reform model was selected to avoid the prob- lems associated with welfare-state, tax-financed systems by distribut- ing the burden of financing among individuals, families, and em- ployers and by promoting involvement of the voluntary and private sectors. The strategy taken was to increase user costs by raising fees and to separate the provision of hospital services and financing from its usual identification with govermnent services. Hospital Reform The changes to the financing system enabled the hospital reforms Singapore launched in 1985. The primary goals of the reforms were to raise efficiency and service standards, improve productivity and cost control, and give management flexibility to respond quickly to changing needs. In keeping with the reform trends in the wider economy of privatization and market liberalization, Singapore opted to address the problems in its public hospital sector with marketiz- ing organizational reforms, in particular, the corporatization model. These reforms were applied to five acute care hospitals and six spe- cialist institutes. To increase management's financial discipline and accountability, commercial accounting systems were also introduced. Prices have been raised periodically to differentiate the quality of "hotel services" in the higher priced wards. In May 1984, the Ministry of Health announced a plan to reform the governance of public hospitals to give them greater autonomy in running their own affairs. It was hoped that the reforms would encourage efficiency, higher standards, and competition among hos- pitals, and hence improve services. Policymakers believed that man- 458 o Innovatons in Health Service Delivery agement autonomy would allow flexibility to innovate and to mo- tivate, and improved working conditions would inspire higher pro- ductivity and a more personalized service to patients. The pilot. Policymakers decided to first pilot the model in the new hospital at Kent Ridge.7 If the reform succeeded, the new gover- nance model would then be extended to the Singapore General Hospital and later to other government hospitals. The Kent Ridge Hospital was renamed and incorporated as University Hospital (Pte) Ltd, a subsidiary company of Temasek Holdings (a government- owned holding company), with a Board of Directors chaired by the permanent secretary/director of Medical Services of the Ministry of Health. The company was "to make the University Hospital the hospital of choice in Singapore" and "to provide excellent medical care, at the lowest possible cost to the patient." It would be run as an autonomous company "to give it maximum flexibility to introduce innovative and cost-effective management systems and to motivate and retain good staff." To fulfill its community responsibility of pro- viding medical care for the poor, it was allocated an annual subsidy from the Ministry of Finance to offset the operating deficit from subsidized care.8 The reform model was viewed as a success-as applied to the newly completed National University Hospital from June 1985. Health Corporation of Singapore. It was decided to roll out the reform to cover additional hospitals. However, rather than reforming indi- vidual hospitals, it was decided to apply the new governance model to them as a group or network. To this end, in 1987, a new govern- ment-owned structure, the Health Corporation of Singapore (HCS) was established, and the NUH (Pte) Ltd was placed under its con- trol. In December 1988, the National Skin Centre was integrated into the HCS, followed by the Singapore General Hospital in April 1989; Kandang Kerbau Hospital and Toa Payoh Hospital in April 1990; and Tan Tock Seng Hospital in April 1991. The HCS func- tioned as a vehicle to acquire and manage all reformed government Attacking Hospital Performance on Two Fronts * 459 hospitals. Legally, it was incorporated as a holding company with the hospitals and specialist institutions as its subsidiaries. The Health Corporation of Singapore became the largest health care provider in Singapore. It has 5,000 hospital beds (45 percent), more than 12,000 employees, and 200,000 admissions (50 percent) annually. Its listed business is to provide health care services through hospitals and specialist centers and to conduct education and re- search. As a holding company, it is governed by private law. How- ever, it is wholly owned by the government of Singapore and re- sponsible to the Ministry of Health. A Board of Directors, consisting of top government appointees and all the chairmen of its subsidiary companies, provides policy direction. Its mission statement is: "To own and manage an efficient network of health care institutions through which an excellent level of health and medical care is pro- vided to our patients in the most cost-effective manner." The reform model is described below, according to the framework elaborated in chapter 1. Decision rights. Reformed hospitals received extensive decision rights, including the right to recruit staff, set terms of remuneration, and decide on the deployment of labor and other resources. They also have the right to set some service prices. As noted above, the hospi- tals were established as private companies, which makes the transfer of decision rights irreversible without a complete reorganization. Residual claims. Reformed hospitals are formally the full residual claimants on their budgets. However, the linkage of the continued government subsidy to the level of operating deficits reduces the ex- tent of these claims. Over time, the subsidy has decreased, moving the hospitals closer to full residual claimant status. Market exposure. Increased prices and cost recovery were a central component of the reform model. These increases, combined with the decreasing government subsidy, heightened market exposure signif- icantly throughout the period of the reforms. Cost-recovery increases 460 0 Innovations in Health Service Delivery have generated a shift from about 15 percent before the reforms to more than 55 percent of hospital recurrent expenditure. The accu- mulation of mandatory savings under the Medisave scheme has en- abled these price increases and the related shift to private payment. Accountability. A Board of Directors was established to enable the government to indirectly hold the HCS accountable. The HCS management is accountable only to the Board of Directors for per- formance. The HCS mission is set by the Ministry of Health. The HCS continues to be subject to broad policy guidance by the gov- ernment through the Ministry of Health. Socialfunctions. Following the reforms, delivery of social functions by hospitals was ensured via a combination of mechanisms. The system of internal cross-subsidization from high-paying (Class A and B) pa- tients to low-paying patients (Class C) was continued. This system was altered in several ways, however. Service prices were increased, so even subsidized prices for poor (Class C) patients increased. The government established a ceiling, however, to constrain these in- creases. Hospitals continue to receive an annual subsidy from the government that is notionally linked to continuation of care for poor patients (though it is set relative to the operating deficit). In 1993, the establishment of Medifund, which makes payments to defray hospital bills for needy patients, made funding for social functions more explicit. While social functions have been more clearly delineated, there is no transfer tied to delivery of specific services. The funding system has effectively made funding for these services more explicit, in that they are increasingly paid for privately, with some support to needy patients. Political Economy There were serious concerns among the population and other stake- holders about the reforms. Because of widespread privatization in other sectors, there was much confusion and concern in Singapore Attacking Hospital Performance on Two Fronts * 461 about whether the corporatization model that was applied to the hospitals constituted privatization. Beneath the terminology, people clearly worried that the reforms reduced the scope of state responsi- bility. Besides consumers reacting to planned price increases, many health professionals and administrators believed the reforms would reduce their influence and control over health services, which they naturally opposed. Price regulation alleviated public fears, though policymakers viewed it as a stopgap policy, pending the development of a more stable and competitive market. Guarantees of appropriate opportunities and incentives in the new entity for an extended grace period reduced workers' resistance. These steps were taken in extending the reforms to the other gov- ernment hospitals after the National University Hospital pilot. In an- nouncing the reform of the Singapore General Hospital, the health minister attempted to buttress the case for the reforms by stating that the government hoped that "freeing the hospital from bureaucratic red tape would lead to better quality medical sources at economical rates."9 The government also expected the reforms to bring about more responsive services and to constrain further cost escalation. The minister also assured the public that SGH would continue to provide subsidized care to persons who could not pay full cost and waive fees for anyone who could not afford even the subsidized rates. Since the reforms obviously implied a great deal of upset for staff, the govern- ment also stressed that "there is likely to be job (re-)classifications and the signing of new contracts but the terms and conditions of service will be as good as what staff are now enjoying."'l To deal with these concerns, implementation was phased in over time. Staff members were given three options: Option A-accept the new terms and conditions to join the company; Option B-take up to a year to decide; or Option C-remain in the civil service. By Feb- ruary 1989, 80 percent of the 647 staff members in the Singapore General Hospital and the Ministry of Health's Pathology Depart- ment had opted for the new contract. Employment offers to the re- maining 2,000 or more staff, including the Department of Biomed- ical Engineering and School of Radiography were still being made. 11 The restructuring of the first six departments was started in January 462 o Innovations in Health Service Delivery 1989 to try out the staff-absorption program and to transfer em- ployees to the new working arrangements. This allowed the em- ployment shifts to be phased in over time and minimized transition problems. On April 1, 1990, management and ownership of the hospital were transferred from the Ministry of Health to the new government-owned company, Singapore General Hospital (Pte) Ltd. Trade unions communicated another concern about the reforms: emergence of differential access to high-quality services, either among wards or hospitals. Even before the reforms, there was a widespread perception that quality health care was associated with the higher class wards and was therefore less accessible to lower in- come individuals in the subsidized wards. The liberalized pricing and salary regime in the reformed hospitals seemed likely to exacerbate this segmentation. The emerging difference in remuneration seemed likely to attract doctors away from employment in (unreformed) public hospitals, and, within a hospital, away from lower class wards. Since the more lucrative settings cater to higher income groups, in- cluding rich foreigners, it seemed likely that access to high-quality health resources and expertise would increasingly depend more on ability to pay than on medical need. This issue was raised in November 1988 by the trade unions, even before the first public hospital was reformed. Unionists were con- cerned that their workers' benefits would be pegged to prices in (un- reformed) government hospitals, blocking their access to the higher quality services available in the reformed hospitals. The deputy sec- retary of the National Trades Union Congress (NTUC) proposed that hospitalization benefits be pegged to the new SGH price regime, rather than to that of the (unreformed) government hospi- tals (which most employers based their employees' benefits on at that time). "To the workers, the logical move would be to make the Sin- gapore General Hospital the standard, . . . [after the reform]," he said."2 After the SGH restructuring, the issue again flared up in July 1989 when the NTUC criticized some employers for classifying SGH as a private hospital and preventing workers from seeking treatment there. The NTUC also pointed out that many employers were shift- Attacking Hospital Performance on Two Fronts * 463 ing the increased costs onto their workers, by making them pay the extra costs associated with treatment in the reformed SGH.13 The government stated their opposition to these practices, by noting that SGH was still a government hospital and that unions should not hes- itate to take employers to the Industrial Arbitration Court if they held back on payment. The unions welcomed this clarification. Em- ployers for their part were concerned with the escalating costs asso- ciated with reforms. Choice and Competition The reforms envisioned that the hospitals would focus more on at- tracting consumers and that consumer choice would place pressure on the hospitals to improve. Signs emerged that hospitals were gen- erating excess demand for some services, and that this was leading to wasteful consumption and duplication.'4 Pressure increased for the government to exert price and quality controls. As the reforms pro- ceeded, it became clear that the increased choice and competition generated by the reforms necessitated additional regulation. Some of these problems had cropped up early in Singapore's hos- pital reforms. In the 1990 budget debate, the health minister provided reassurance that the government would closely monitor reformed hospitals for duplication of expensive services. He pointed out that the Ministry of Health continued to coordinate the development of services and purchase of costly equipment. Regarding the setting up of costly departments such as radiotherapy, the minister acknowl- edged the extremely high costs of duplication in such areas.",15 Notwithstanding the use of existing regulatory mechanisms, com- petition did lead to some wasteful duplication and cost inflation. In- vitro fertilization (IVF), for example, became available at seven dif- ferent hospitals in Singapore. In the 1990 health budget debate, the chairman of the Government Parliamentary Committee for Health asked, "How many people can actually benefit from the program.... [T] he procedures are expensive and the success rate is not fantastic.",16 Moreover, the prevalence of IVF raised another issue: cost shifting from private to government hospitals. TVF often increases the fre- 464 o Innovations in Health Service Delivery quency of multiple births and underweight babies who require ex- pensive neonatal intensive care. "Did the private hospitals have their own neonatal units or were they dumping babies with birth problems into the cheaper government hospitals?" asked the chairman. Costs and Pricing While employers and employees were equally concerned about the implications of the reforms on their own costs, the public at large was also concerned about the plight of the poor. Fears of price esca- lation and reduced access were rekindled when C class beds were phased out with the restructuring of SGH (Pte) Ltd. Members of parliament and political activists condemned this decision, and the first deputy prime minister admitted it was a mistake. The issue highlighted the explosively emotional nature of health care concerns, especially the humanitarian aspects of caring for the poor, the role of govermment, and the question of affordability. The present dif- ferential pricing for ward accommodation in public hospitals has much to commend it as an acceptable and workable system of cross- subsidization. The differential charges in hospital bills as a result of restructur- ing soon began to surface as political issues. Several members of Par- liament asked why the restructured Singapore General Hospital had raised its fees. The health minister had to justify the costs.17 The re- structured hospitals were autonomous and would charge differently, according to their own itemized pricing. C class beds would still be available, and anyone who could not afford to pay could seek a waiver or a reduction of fees. Whatever the charges, the policy was that nonsubsidized patients in the A and B wards pay the full cost of treat- ment. For the first time, comparative costs for hospital stays in dif- ferent hospitals and wards were made public (tables 12.1 and 12.2). Not surprisingly, the government hospitals had the lowest fees; pri- vate hospitals, the highest. The minister explained that SGH had raised fees to pay for upgrades in facilities, equipment, and staff. Sim- ilarly, N-UH's higher fees reflected improved patient services and its status as a teaching hospital, where doctors conducted more tests Attacking Hospital Performance on Two Fronts * 465 Table 12.1 Average Cost of One-Day Stay in Hospital, April 1989 (in Singaporean dollars) NATIONAL GOVERNMENT SINGAPORE UNIVERSITY PRIVATE WARD HOSPITALS GENERAL HOSPITAL HOSPITAL HOSPITALS Class A 255 3550 385b 480 715 Class B1 170 2300 240b 400 630 Class B2 55 70° Bob 97 380 Class C 40 - - - - Not applicable SGH Singapore General Hospital, NUH National University Hospital. a. Before April 1, 1989 b. After April 1, 1989 Source Straits Times, August 5, 1989. themselves. There was thus no need to standardize fees for all re- structured hospitals because an objective of restructuring was to allow the hospitals to run independendy and to compete for patients.18 The Ministry of Health later published in the newspapers an ex- planation and table on the comparative average costs of hospital stays in government-restructured and private hospitals.19 The article also sought to clarify comparative costs quoted in a new information pam- phlet on Medisave. The average costs per inpatient-day reported for government/reformed hospitals included doctors' fees, but those for private hospitals excluded both doctors' and surgeons' fees (table 12.3). Physician Payment Since the reforms allowed the hospitals to structure their own pay- ment system, disparities in physician incomes emerged.20 After the Table 12.2 Average Cost of "A" Class Hospitalization (in Singaporean dollars) SERVICE PUBLIC HOSPITALS SGH NUH PRIVATE HOSPITALS Medical case 1,500 1,680 2,430 3,100 Appendectomy 1,650 2,110 2,670 4,300 Removal of gall bladder 3,050 3,600 4,030 5,400 SGH Singapore General Hospital, NUH National University Hospital Source Straits Times, August 5, 1989 466 o Innovations in Health Service Delivery lrcioe 0 ]23 Average Cost of One-Day Stay in Hospital, December 1989 (in Singaporean dollars) AVERAGE PER INPATIENT-DAY AVERAGE BILL SIZE GOVERNMENT/ PRIVATE HOSPITALS GOVERNMENT/ PRIVATE HOSPITALS RESTRUCTURED (EXCLUDING RESTRUCTURED (EXCWDING HOSPITALS (INCWDING DOCTORS'/ HOSPITALS (INCLUDING DOCTORS'! WARD DOCTORS' FEES) SURGEONS' FEES) DOCTORS' FEES) SURGEONS' FEES) 1 bed 250-480 400-620 1,275-2,448 2,040-3,162 2-4 beds 170-420 200-520 867-2,142 1,020-2,652 6-10 beds 50-100 150-300 255-510 765-1,530 Open ward 40 n a 204 n a. n a. Not available. Source. Straits Times, May 2, 1990 reformed hospitals implemented schemes to pay senior and sought- after physicians higher salaries, it was necessary to extend this privi- lege to all government hospitals. Though intended to check the out- flow of specialists from the public to the private sector, the removal of ceilings on fees widened the income gap and raised questions about how much doctors should earn and their obligations to poor patients. It also rekindled fears about abuses in the competition to capture paying patients, declining interest in training and research, and spiraling cost inflation.21 Mimicking the liberalization of the SGH and NUH, the rest of the unreformed hospitals removed their fee ceilings to staunch their doctor drain. Of the 260 government doctors who left public service between 1986 and 1988, 140 went into private practice.22 But the as- sumption that public sector doctors could be motivated purely by monetary incentives to stay was not justified. At the National Uni- versity Hospital, 4 associate professors and 3 consultants resigned after the ceilings were removed, thus implying that reasons other than remuneration motivated their departure.23 There were also res- ignations from the Singapore General Hospital, and reports of sev- eral others planning to leave for the private sector. Most doctors who left discounted the importance of the lifting of the consultancy ceil- ing in keeping good staff, while citing nonmonetary and personal reasons. Attacking Hospital Performance on Two Fronts * 467 Among the host of problems created by the new incentives was the widening income gap between medical specialities, especially be- tween physicians and surgeons. A top surgeon could earn S$40,000 more a month than a top physician, since surgery commands a higher fee.24 The top money earners were reportedly specialists in heart surgery, eye surgery, and obstetrics and gynecology, who began to earn S$40,000 to S$60,000 a month, three to five times more than under ceilings. Top physicians earned S$15,000 to S$24,000, includ- ing consultancy fees, 50 to 100 percent more than previously. The new arrangements thus favored surgical and other costly treatments, using sophisticated facilities and investigative procedures. Since the hospital reform, public sector doctors appear to be rea- sonably well compensated. Physicians in government-owned facil- ities receive basic pay plus a "clinical faculty supplement" of 25 percent of base wage. Physicians with very heavy clinical loads, es- pecially in procedure-based specialities, may opt for an incentive based on their total billings instead of the fixed 25 percent supple- ment. A senior registrar (roughly equivalent to a postresidency fel- lowship in the United States) receives a salary equal to three times the country's average annual wage. A junior staff physician receives five times the average wage, and a senior physician earns about six times the average wage. These incomes are comparable to medical earnings in the United States, where five to six times the average wage is normal.25 Impact of Hospital Reforms Many positive results have been achieved through the reforms, in- cluding increased financial responsibility and commercial discipline, as well as improved standards of hospital services and responsiveness to patients' needs. From their new, quasi-independent status, hospi- tals have gained entrepreneurial flexibility and the ability to respond quickly to the changing supply and demand for hospital services. These hospitals also serve as benchmarks, competing against the pri- vate hospitals in terms of price and quality. 468 o Innovations in Health Service Delivery However, other, less favorable trends have emerged, and have had to be dealt with subsequently. Anecdotal evidence suggests that added quality may have come with administrative cost increases of 5 percent to 10 percent of recurrent expenditure. Vhile it is difficult to generalize from individual hospital performance, Singapore Gen- eral Hospital, the country's largest hospital and one of the two pub- lic tertiary centers where high-technology procedures are concen- trated, is the best example for studying the effects of the reforms. Labor's share of total hospital cost in Singapore is similar to that in the United States, but SGH is much leaner than its American counterparts. Of the nonphysician personnel at SGH (nurses and paramedical personnel, and others involved in ancillary services), only about one-sixth performs administrative and clerical functions. Even in benchmark American hospitals, the ratio of caregivers to support personnel is 2:1.25. The 5:1 ratio at SGH reflects increased efficiency as a result of the lightened bureaucratic and regulatory loads Singapore places on the delivery system.26 Although nurse-to- patient ratios in Singapore hospitals are in line with international standards, nurses are relatively less well paid, and demand for foreign nurses is increasing. Media reports have highlighted shortened waiting times and lengths of stay and the virtual disappearance of wait lists for many programs, including elective procedures. Innovative projects and the acquisition of new medical technologies and treatment regimes have also been noted. The governance and funding structure fueled a more competitive environment and generated a more market- oriented approach in the promotion of new products and services. However, whether any additional supply goes to fulfill previously unmet demand-or merely generates unnecessary consumption-is open to debate. 'rhe government has put in place revenue caps on reformed hos- pitals to prevent them from generating the "excess profits" associated with induced demand. The government has also established limits on average charges per patient-day and will adjust them annually. Hospitals that exceed the limits have their government subsidies cut by that amount; hospitals with a budget surplus now keep the addi- tional funds. Attacking Hospital Performance on Two Fronts * 469 Historically, the rate structure of public hospitals may have indi- rectly influenced private hospital rates, assuming public sensitivity to price differences. The government has also threatened to impose direct cost controls on private hospitals and doctors, especially for "balance billing," where providers charge much more than the lim- its set in various public financing programs. However, implementing direct cost controls is inherently difficult unless the payment system and fee schedule are standardized and adjusted for different risks and severity of disease conditions. The absence of a case-mix classifica- tion of patients related to diagnostic groups complicated this situa- tion at that time. A case-mix system that was modified from the Aus- tralian DRG model was subsequently introduced in 1999. Technical Efficiency Financial statements in the annual reports of the restructured hospi- tals under the Health Corporation of Singapore provide data on hos- pital expenditures and revenues, from which cost-recovery ratios could be calculated. Except for the National University Hospital, all hospitals have improved on their cost-recovery positions (figure 12.1). However, the NUH cost-recovery ratio is still higher than those for other restructured hospitals, except for the Singapore Na- tional Eye Centre, a specialist center. In this aspect, hospitals could have become more efficient as they recovered more of their costs. Not only are the hospitals improving upon their cost-recovery status, their cost-recovery indices also seem to be converging at between 0.4 and 0.6. This trend could be attributed to the govern- ment's policy of taxing a certain part of the hospitals' earnings. Hos- pitals may not have any incentives to improve their cost-recovery status if excess revenues will be taxed away. From the social perspec- tive, the rationale behind this policy is to prevent hospitals from overcharging patients in a quest for higher revenue. Different types of hospitals display different trends and cost- recovery ratios. Ang Mo Kio Community Hospital has the lowest cost-recovery ratio (0.4), which is consistent with its status as a com- munity hospital. Community hospitals provide services to the el- derly and the chronically ill, which are highly subsidized. Secondary 470 o Innovations in Health Service Delivery io$gure 12.11 Cost-Recovery Ratio, Restructured Hospitals, 1990-97 1.2= 0.8- 199 199 199 193 19-99 96 19 -o4 Year -O SGH {})- SNEC ---KKH X CGH ---NUH - AMKCH I TTSH HCS (Gp) Note- SGH Singpore Generol Hospit- l SNEC SHgapore Nationl Eye Center; KKH Kandang Kerbau Hospital; CGH Changi General Hospital, NUH National University Hospital; AMKCH Ang Mo Kio Community Hospital; TTSH Tan Tack Seng Hospital, HCS Health Corporation of Singapore Group Source Compiled from annual reports of various hospitals, Health Corporation of Singapore hospitals like Changi General Hospital and Tan Tock Seng Hospital are the next lowest in terms of cost recovery. Their charges are lower than the tertiary hospitals (SGH and N1E), which explains the higher cost-recovery ratios for tertiary hospitals. Kandang Kerbau Hospital for women and children, and Singapore National Eye Cen- tre, a specialist center, are not comparable with the rest. To provide a better insight into how hospitals' costs were recov- ered, figures 12.2 and 12.3 show revenue and expenditure trends. As seen in the figures, every hospital's revenues and expenditures rose. Since cost-recovery ratios have increased, revenues may have Attacking Hospital Performance on Two Fronts * 471 Figure 12.2 Hospital Revenues, 1990-97 (in Singaporean dollars) 600,000 500,000 LI 400,000 - 2 300,000- 0) 4D 200,000 _ _ _ _ _ 100,000 0 1990 1991 1992 1993 1994 1995 1996 1997 Year -- SGH -r- SNEC * KKH -h- CGH NUH |-+AMKCH -I--TSH -HCS (Gp) Note See figure 12 1 for full names of hospitals Source Compiled from annual reports of various hospitols, Health Corporation of Singapore Figure 12.3 Hospital Expenditures, 1 990-97 (in Singaporean dollars) 1,200,000 z1,000,000 ° 800,000 - , 600,000 x U. 400,000 /----- 200,000 1990 1991 1992 1993 1994 1995 1996 1997 Yea r SH -U- SNEC -*- KKH -- CGH --NUH --+ AMKCH -I- TTSH -HCS lop) Note See figure 12.1 for full names of hospitols Source Compiled from annual reports of various hospitals, Health Corporation of Singapore 472 o Innovations in Health Service Delivery Figure 112.4 Physician/Nonphysician Ratios, 1989-97 0.2000 0 1500 a 0.0500 - 0 0000 l I I l 1989 1990 1991 1992 1993 1994 1995 1996 1997 Year --| SGH --} KKH --NSC TPH Note SGH Singapore General Hospital; KKH Kandang Kerbau Hospital; NSC National Skin Centre, TPH Toa Payoh Hospital Source Singopore Official Statistics. increased more than expenditures. Higher admissions and bed- occupancy rates and increases in hospital charges could explain the increase in revenues. Purchases of more sophisticated and expensive medical equipment explain the inevitable increase in charges. PhysicianlNonphysician Ratios The hospitals' physician/nonphysician ratios have remained rela- tively constant or have fallen (figure 12.4). Hospitals may be recruit- ing more nonphysician staff for management and administrative duties, a more efficient arrangement than in the past, when doctors performed these functions. In addition, only a sixth of nurses, para- medical personnel, and others delivering ancillary services are in- volved in administrative and clerical functions. Even in benchmark American hospitals, the ratio of caregivers to support personnel is 2:1.25. The 5:1 ratio at SGH reflects increased efficiency as a result of Singapore's lighter bureaucratic and regulatory loads on the de- livery system. Attacking Hospital Performance on Two Fronts * 473 Figure 12.5 Admissions, Public and Private Hospitals, 1989-98 300,000 250,000 200,000. 0 150,000 iE 100,000 __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 50,000 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Year |+- Public -* - Privatel Source Department of Statistics, Annual Statistical Yearbook, various years Hospital Admissions Admissions to both public and private hospitals have increased (fig- ure 12.5). Until the regional financial crisis of 1998 reversed this trend, admissions had been growing faster at private hospitals than at public hospitals. Allocative Efficiency To ensure that the lower income groups are not deprived of health services because they cannot afford them, subsidies for hospital Class C wards are 80 percent. B2 wards cater to middle-income earners; B 1 and A wards are for high-income earners. From the equity per- spective, subsidizing the poor more than the rich makes sense, but this means sacrificing efficiency to equity. Based on data from Singapore General Hospital, the average length of stay in a C class ward is longer than in the other ward classes 474 o Innovations in Health Service Delivery Pigure 1 2.6 Admission Rates, Public and Private Hospitals, 1990-98 0.1I 0.05- 2 0 C 0 0 -005\ -01I -0 15 1990 1991 1992 1993 1994 1995 1996 1997 1998 Year | Public {} Pnvate Source: Department of Statistics, Annual Statistical Yearbook, various years (figure 12.7). However, there is no medical reason that a stay should be longer for similar cases and conditions. One explanation is that C class wards, being heavily subsidized, are overutilized. Patients, especially the elderly and chronically ill, though ready to be dis- charged, might choose to stay a few more days for lack of alternative facilities. The better-off may have domestics to take care of them after discharge. As a result of the increased length of stay, the bed oc- cupancy rates for C class wards are sometimes more than 100 per- cent. This is generally true for the bed occupancy rates of hospitals in the public sector (figures 12.8 and 12.9). Equity Subsidized care in Singapore accounts for about 20 percent of aUl health care expenditure, 0.7 percent of GDP. A key component of the government's policy for equity is a tiered structure of subsidies, based on the setting in which care is delivered and the amenities pro- vided with it. In the public hospitals, the different classes of wards re- ceive varying degrees of subsidy; private hospitals are unsubsidized. A major change after the national health policy review and hospital restructuring in 1991 was to ensure that subsidies were targeted by Attacking Hospital Performance on Two Fronts * 475 Figure 12.7 Average Length of Stay, Singapore General Hospital, 1990-97 9.- -z;- 8 -~- 2s 6 '-~ o 5- s - m4- -3- P) 2 60 - a0- ,, 40- 0 e 20 0 1990 1991 1992 1993 1994 1995 1996 1997 Year - A +B1 _B2 C Source Health Corporation of Singapore 476 o Innovations in Health Service Delivery IFigum. T 2.9 Bed Occupancy, by Ward Class, Public Sector Hospitals, Various Years 100 95 C o 7 90 198 19 94 19 96 19 C a 8 ------ 2 a 75- u 8 70- 65- 60- 1988 1993 1994 1995 1996 1997 - Class A 74.5 65 3 66.3 69.8 72 69 3 Class BI 80.7 81 1 80 80.9 82 3 80 4 Class B2 81 1 88 8 84 2 84.9 87.2 88 7 --- Class C 683 76 3 81.6 83.5 85 3 93.7 Source. Ministry of Health, quoted in H L Pang, "Healthcare Services in Singapore-A Demand Analysis," Singapore Nursing Journal 26(2) 21-24 (1999) appropriate channeling of patients into the appropriate ward classes. Their ability to pay is roughly determined by their Medisave account balances, a form of means test. Individuals are theoretically free to choose whatever ward accom- modation level they prefer. However, at the time of admission, pa- tients are advised of their responsibility to choose a ward class they can afford through a combination of Medisave, Medishield, or per- sonal funds. If necessary and with the owner's consent, patients can draw on their spouse's, children's, or parents' Medisave accounts. Class A wards have no subsidy and compete with private hospitals, offering private rooms with such amenities as air conditioning, tele- vision, and en suite bathrooms, in addition to the government's list of basic services. Care delivered in the remaining four wards-B 1, B2+, B2, and C-is supported by varying levels of government al- lowances. For example, the subsidy level in a Class B 1 ward is 20 per- Attacking Hospital Performance on Two Fronts * 477 cent of the total charges. Patients are responsible for the remaining 80 percent. The fraction paid directly by the government increases as amenities decrease, reaching 80 percent for C class. B 1 wards have four beds to a room; B2 patients do not have a choice of physicians; C wards are generally open wards. In addition, Medishield pays a greater part of the hospital charges for poorer patients and those who elect to receive their inpatient care in the subsidized hospital settings Quality Singapore has one of Asia's most sophisticated health care delivery systems. The admission rate for residents is 1.1 per 1,000 population, about the same as the more aggressively administered health main- tenance organizations in the United States.27 The average length of stay at an acute care hospital is five days, also comparable to the best U.S. managed-care programs and far less than that in other devel- oped countries. Occupancy rates are kept well above 80 percent, and in many reformed hospitals exceed the 90 percent mark. High-technology services are provided at what appear to be appro- priate levels. In 1993, of the 1,051 coronary artery bypass surgeries performed, 676 were provided to Singapore residents for a raw uti- lization rate of 24 per 100,000 population. Because Singapore's pop- ulation is relatively young, the age-adjusted utilization rate, though lower than in the United States, probably approaches that of Canada, Germany, and most West European nations.28 These data suggest that services are available at levels acceptable by most international standards and that high-technology medical services are generally available to the population. The reformed hospitals do not have the queues and long waiting times that characterize heavily regulated na- tional health service systems in many other developed countries. Conclusions The primary objectives of Singapore's hospital reforms was to con- trol government and total health care expenditures, to improve the quality of hospital services, and to improve the working conditions of government health personnel (box 12.2). 478 o Innovations in Health Service Delivery Box 12.2 Assessing «ie Finance Reforns Resource mobilization A high priority on Singapore's health reform agenda is mo- bilizing enough resources to finance efficient and equitable provision of health care to meet increasing demand. Singa- pore has chosen to increase cost sharing by individuals and their families, supported by the introduction of mandatory medical savings accounts. Currently Medisave covers less than 10 percent of total health spending and about 30 percent of hospitalization ex- penditures. Over the longer term, the role of medical savings accounts is destined to grow considerably. Already, the assets accumulated in members' savings accounts are equivalent to four years' worth of Singapore's national health expenditure. This constitutes a substantial nest egg for the future aged population and could cover other contingencies such as eco- nomic recessions when welfare consumption rises. Efficiency Since Medisave balances are reserved to pay for infrequent but high-cost inpatient care, Singapore's financing system has succeeded in targeting actual insurable risks rather than predictable costs (such as outpatient care). However, be- cause they depend on intertemporal pooling over the indi- vidual's life cycle, it is not actuarially feasible for Medisave balances to insure against truly catastrophic contingencies. To solve this problem, Singapore introduced Medishield-a back-up health insurance program based on cross-sectional risk pooling, designed to finance the extreme catastrophic part of the risk distribution. Government policies on cost sharing to help control moral hazard and contain costs in social insurance programs vary greatly. Health financing policy in Singapore combines Attacking Hospital Performance on Two Fronts 0 479 appropriate deductible and coinsurance rates with explicit targeting of costly risks.a On average, the government sub- sidizes about 60 percent of hospitalization costs in public hospitals. The residual 40 percent charged to patients is cov- ered between Medisave and out-of-pocket payments. Thus, patients feel a double bite of individual responsibility-the 20 percent coinsurance paid out of their Medisave account and another 20 percent paid directly out of pocket. Claims for back-up Medishield coverage of catastrophic expenses are subject to 20 percent coinsurance on top of a high an- nual deductible. Equity Reforming pricing policy for publicly provided services is likely to be an important instrument to improve equity in the incidence of public spending. Users of publicly provided health services, especially inpatients, often face substantial out-of-pocket costs, which can bar access for the poor. Im- proving access may require selective price reductions, com- pensated by increased direct budget subsidies. Implementing this strategy calls for a pricing policy that consciously differ- entiates prices by users' ilcome. In this way, public subsidies can be targeted. Self-selection, by charging lower prices for services more likely to be used by the poor, is another tar- geting mechanism. Alternatively, subsidies can be targeted directly by mean-testing individual users. Singapore's health reforms demonstrate both approaches to promoting equity.b Budget subsidies, targeted to poorer users by self- selection through public sector pricing policy, continue to play a major back-up role in financing hospital inpatient care. Explicit price discrimination is built around the differ- ent classes of hospital wards in public sector hospitals. The subsidy ratios are highly differentiated, ranging from 84 (Box continues on the following page) 480 o Innovations in Health Service Delivery Box 12.2 (cnlinued) percent of hospital costs in the lowest Class C, to 71 percent in Class B2, 36 percent in Class B1, to 13 percent in Class A.c These differential subsidies are intended to help equal- ize the affordability of the class-specific prices relative to the incomes of patients who select them. As a last resort, patients who cannot pay their subsidized hospital bills can apply for a means-tested grant from their Hospital Medifund Committee. This safety net is targeted directly at households in the lower third of income distribu- tion. During its first three years of operation, Mediftnd paid the entire medical bill in 87 percent of requests for as- sistance. This now amounts to arolnd 5 percent of hospital admissions at the lower Class B2 and C levels, compared with 3.3 percent in 1993, 4.2 percent in 1994, and 5.8 per- cent in 1995 d a. L. M. Nichols, N. Prescott, and K. H. Phua, "Medical Savings Accounts for Developing Countries," Innovations in Health Care Financing: Proceedings of a World Bank Con- ference, March 10-11, 1997, World Bank Discussion Pa- per 365, Washington, D.C., 1997. b. Ibid. c. N. Prescott, "A Script-How to Manage Rising Health- care Costs in East Asia," Healthcare Asxa, Fourth Quarter (London: Economist Intelligence Unit, 1998). d. Ibid. Lack of data prevents a more comprehensive evaluation of the ef- fects of restructuring. This review of available indicators suggests that the reforms improved hospital efficiency in terms of administra- tion and cost recovery. Heightened competition has brought about a rise in admission rates for both public and private hospitals as well as Attacking Hospital Performance on Two Fronts * 481 an increase in service quality. There has also been a discernable shift from government expenditure to self-payment, with the expanded "insurance" coverage provided in the "Medi-schemes" described above. Thus far, service delivery reforms have concentrated on public hospitals. While the past system involved the government in exten- sive provision and financing of health services, the National Health Plan implemented a strong shift away from tax-based health care ex- penditure toward greater cost sharing. These reforms substantially altered the nature of demand, generating clearer signals about what services were needed, and what patients wanted. The application of the corporatization model to public hospitals was critical in creating a public hospital sector that, like the private sector, responded to such signals. By keeping ultimate control in public hands, Singapore appears to have found a structure that provides them with, in some sense, the best of both possible worlds. Acknowledgment: The author is greatly indebted to the late Nicholas Prescott who provided not only guidance and insights for this chapter but also wise advice and in- spiration for social protection and health care reform durmg the recent economuc turmoil in the East Asia and Pacific region. Notes 1. In 2000, the Health Corporation of Singapore was devolved further into two separate clusters of hospital groups that are inte- grated with their primary care network of polyclinics. 2. Other countries discussed in thds volume that operate these cross-subsidy schemes include Indonesia and Hong Kong. 3. Ministry of Health, Blue Paper on the National Health Plan, 1983. 4. Goh Chok Tong, Sunday Times, February 6, 1983. 5. K. H. Phua, "Saving for Health," World Health Forum 8 (1): 338- 41 (1987). 482 o Innovations in Health Service Delivery 6. K. H. Phua, "Comparative Health Care Financing Systems, with Special Reference to East Asian Countries," Research in Health- care Financial Management 5 (1): 113-33 (1999). 7. Joint press statement, Ministry of Health, National Univer- sity of Singapore and Temasek Holdings (Pte) Ltd, Singapore, Janu- ary 12, 1985. 8. Straits Times, January 13, 1985. 9. Straits Times, March 23, 1988. 10. Ibid. 11. Straits Times, February 2, 1990. 12. Lim Boon Heng, Straits Times, November 18, 1988. 13. Straits Times, July 8, 1989. 14. For example, competing hospitals might generate unnecessary services and pass on the costs to misinformed consumers, enabling them to procure unneeded high-technology equipment or to under- take other capital development programs. 15. Straits Times, March 28, 1990. 16. Dr. Aline Wong, Straits Times, March 27, 1990. 17. Straits Times, August 5, 1989. 18. Ibid. 19. Straits Times, May 2, 1990. 20. Straits Times, January 9, 1990, and January 18, 1990. 21. Straits Times, January 18, 1990. 22. Straits Times, September 19, 1989. 23. Straits Times, January 18, 1990. 24. Straits Times, January 9, 1990. Attacking Hospital Performance on Two Fronts * 483 25. T A. Massaro and Y. N. Wong. "Positive Expenence with Medical Savings Accounts in Singapore," Health Affairs 14 (2): 267 (1995). 26. Ibid. 27. Ibid. 28. Ibid. CHAPTER 13 Autonomous Structures with Incomplete Autonomy: Unusual Hospital Reform in Tunisia Hedi Achouri and Eva Jarawan Between 1992 and 1995, Tunisia undertook a multifaceted hospital reform aimed at its 22 teaching hospitals. The initiative included el- ements of technical, managerial, and organizational reforms. A large-scale investment program was undertaken to upgrade facilities, equipment, and management information systems. On the manage- rial development front, efforts were made to train existing managers and provide them with resources on the job to support better per- formance. In addition, significant efforts were made to recruit and retain new professionals possessing management skills in contrast to the administrative skills of the existing cadre.' The organizational changes were grounded in the conversion of the hospitals to legally independent status, the creation of a board structure, and initiation of multiyear performance contracts between the Ministry of Public Health (MPH) and the hospitals. The organizational changes were a relatively minor component of the reforms, with the endpoint arrangements closer to a budgetary than an autonomous entity. Somewhat perversely, in fact, despite the use of structures typically associated with enhanced autonomy, the 485 486 o Innovations in Health Service Delivery reforms ended up increasing the administrative influence of central authorities over the hospitals. Thus, the Tunisian reforms underscore the distinction between formal governance structures and actual governance practices. Vhile it is clearly important to put in place organizational arrangements that support the desired governance processes-it is not sufficient. In Tunisia, as in the United Kingdom, structures developed to support enhanced autonomy ended up serv- ing as a vehicle for centralization. The Tunisian reforms are also interesting in that they sought to make technical, managerial, and organizational changes simultane- ously. Further, the reform is viewed largely as a success in the country and has proven to be sustainable. Thus, its multidimen- sional design will be of interest to other countries faced with re- forming a hospital system with problems requiring efforts on all three fronts. It is not possible to draw conclusions about the impact of the or- ganizational changes separately, since changes on all these fronts took place simultaneously. The reader is encouraged to keep this in mind when reviewing the discussion of results. Coente and Background Between Tunisia's independence in 1956 and the 1980s, the govern- ment's main health care objective was to guarantee coverage to all via public health facilities. Little attention was paid to cost, and by the late 1970s, these facilities were under growing financial pressure. Other sources of funding had to be found. In 1981, the MPH began experimenting with pro forma invoicing of public hospital admis- sions in an effort to find out the cost of care and how much funding the system needed. In 1983, the contribution from the social security fund was increased, and user charges were introduced. This initiative was launched without arrangements for proper training or adequate financing and was subsequently abandoned. After several other false starts, political changes in November 1987 led to a study of health sector malfunctions. Autonomous Structures-with Incomplete Autonomy * 487 Two types of problems were diagnosed in public hospitals, partic- ularly university hospitals: inadequate funding mechanisms and in- ternal inefficiency. In funding, budget appropriations were insuf- ficient to meet emerging demand for services, social security contributions did not meet the cost of care for services rendered to their contributors, too many users received free care, input prices were rising, and information on the cost of care was scarce. Internal inefficiency was attributed to a range of interdependent factors, both general and technical in nature. These shortcomings included: managerial inability to evaluate performance, poor input monitoring through inventory control, waste of resources, igno- rance about the real cost of health services; rigid administrative and budgetary procedures; inflexible personnel management regu- lations; insufficient reliable management information for diagnos- tic assessment and well-targeted decisionmaking, and uncertainty about the distribution of powers and functions among the different administrative levels (individual hospital, regional authority, central government). Changing View of the Role of the State The public sector has always been predominant in Tunisia's health care system. In Tunisia, as in many other countries, current political thinking calls for a restructuring of the state's role in the economy as well as the private sector's roles to make room for individual initia- tive, healthy competition, and innovation. The broadly held vision is for a state that will concentrate on creating conditions conducive to social development and providing infrastructure and social services. The Reform Package The reforms were implemented in 22 hospitals; another 4 were elim- inated via merger. 488 o Innovanons in Healdt Service Delivery Objectives The Tunisian government's reform program is directed chiefly at raising the quality of hospital services and controlling public health expenditures, by ensuring appropriate billing and cost recovery for services in public facilities. The program was designed to: o Improve the internal efficiency of major teaching hospitals, while also containing costs and improving service quality o Enable adjustments to be made in financial burden sharing by for- mulating data to connect the use of hospital services with payment by insured and nonpoor uninsured patients. Organizational elements. To meet these objectives, teaching hospitals were converted into government-owned health corporations (EPSs). The EPS has legal status as a corporate entity under commercial law. It is financially independent, overseen by a Governing Board, and run by a chief executive officer (CEO) under MPH supervision. Multiyear performance contracts were initiated between the MPH and the CEOs. Technical elements. An infrastructure upgrade program was instituted for equipment to improve diagnostic and treatment capabilities, hos- pital maintenance, patient accommodations, and hospital hygiene. In addition, hospital administration and reception areas have been ren- ovated to accommodate the expansion of administrative business and new opportunities for treating patients. An integrated, computerized management information system (MIS) was introduced-although it covers only a part of hospital ac- tivities. This system is designed to enable the cost of health care to be determined and will require further expansion to cover the entire range of activities undertaken in the hospitals. Performance mea- surement criteria have been developed to allow comparison across hospitals (benchmarking) and to feed into the drawing up of the con- tracts. A Computing Center was established in the MPH, to enable the software designers to work closely with the hospital managers. Autonomous Structures-with Incomplete Autonomy * 489 An Information Technology Committee oversees the quality and consistency of these initiatives. Managerial elements. The management capabilities of hospital ad- ministrators were targeted for improvement. Management functions have been reorganized, following a standard organization chart. Tasks have been separated to eliminate incompatibilities and rein- force internal controls. A handbook of standardized management practices has been produced and put into computerized form. New managers were recruited to fill gaps, and personnel were trained to improve their managerial skills, especially in MIS use. Technical and financial support is being provided for a master's degree program and a diploma course in hospital administration at the School of Commerce. Complementary reforms. The Free Medical Care System has been reorganized and the criteria for entitlement to subsidized health care have been further elaborated. In addition, the prices at MPH-con- trolled public health facilities, and patient-assignment procedures have been refined. Subsequent to the reforms, billing for hospitalization of social se- curity contributors was introduced in January 1996, based on a flat admission charge. In January 1997, it was extended to outpatient consultations, using the same rate-setting system. However, it does not apply to all sources of funding. The Organizational Reform Decision Rights Day-to-day operation. Reduction of day-to-day intervention was planned, and performance contracts came to play a greater role in ensuring accountability. Therefore, the reforms provided for EPSs to negotiate performance contracts that set out operating and in- vestment budgets and funding plans, which were to be aligned with 490 o Innovations in Health Service Delivery the organization's goals and projected activities. The Governing Board was assigned the right to approve performance contracts on the hospital side. Personnel management. EPSs were given no significant decision rights in personnel management. All public health facilities are covered by the civil service regime and therefore have no autonomy in hu- man resource decisions (staff size, recruitment, salary levels, or pro- motions). The reforms did endow the EPS with the right to handle disciplinary procedures direcdy in the cases of certain personnel, excluding medical, pharmaceutical, and dentistry practitioners and certain classes of administrative personnel. The EPSs were also given certain rights to define how adminis- trative departments were organized, including the number and level of functional positions. Appointments of physicians, pharmacists, and dentists as heads of hospital departments are handled in the same manner for all public health facilities. o Property management. Apart from some small legal distinctions, there are no practical differences between the EPS and the tra- ditional hospitals with regard to ownership and management of property. o Organization of management. EPS Governing Boards formally re- ceived some of the MPH's rights to organize hospital manage- ment, within parameters set out by decree (though the proposals must receive MPH approval). This included the right to open and close departments. As noted above, EPSs did gain additional flex- ibility in determining the number and level of functional positions. o Procurement. The legal change to the EPSs gave them some deci- sion rights related to procurement. Public hospitals must follow competitive contracting procedures for contracts above D30,000. By contrast, EPSs can forgo these time-consuming tenders for contracts up to a value of D100,000, with permission from the Governing Board. Autonomous Structures-with Incomplete Autonomy * 491 * Financial management. The Governing Board formally received the right to draw up operating and investment budgets and asso- ciated funding plan. The EPSs received the formal right to take out loans. Residual claim status. The EPSs were given additional freedoms in managing their operating surpluses. All public hospitals can retain funds earned through their activities, whether in the form of gov- ernment subsidies, contributions from social security funds, or co- payment revenues from users. They may recycle operating surpluses (in excess of budget projections) into either their operating budgets or their capital budgets. However, for these allocations, public hos- pitals must receive approval from the Finance Ministry and MPH, while EPSs need only a decision from their board. Accountability Mechanisns Governing board. While public health facilities have only a consulta- tive body, a "health council," the reforms gave each EPS a governing board-a mechanism for exercising, at least formally, indirect guid- ance over the hospitals activities. A typical EPS Governing Board has 16 members including significant representation by hospital staff (9 chosen by MPH and 7 elected from the EPS medical and paramed- ical staff). The board chairman, chosen by the minister of public health, represents the MPH and sets board meeting agendas on the recommendation of the chief executive officer, who is not a board member. The board, by decree, "is invested with the widest powers to act in the name of the institution," including opening and closing departments, approving performance contracts, and drawing up op- erating and investment budgets and associated funding plans.2 Chief executive officer The reforms also provided for accountability to be exercised via ministerial control of the CEO's appointment. The CEO's responsibilities include: running the hospital's technical, ad- minustrative, and financial affairs; preparing the board's work and ensuring implementation of its decisions; representing the hospital in dealings with third parties; ensuring recovery of health care costs; 492 o Innovations in Health Service Delivery awarding procurement contracts; drawing up the hospital's operating and investment budgets and associated funding plans; and exercising authority over "all personnel."3 Performance contracts. A core part of the new accountability structure was grounded in performance contracts, which were intended to be established between the MPH and EPSs. Over time, as more infor- mation about hospital performance became available, the MPH budget allocation would be based on these indicators. A system of benchmarking hospitals would be developed, enabling MPH author- ities to compare a hospital's performance with that of other hospitals, and to its own the preceding year. Thus, even without direct compe- tition, pressure to improve performance would develop. Medical Committee. The reform also created a second, subordinate boardlike structure called the Medical Committee. The committee includes the CEO, department heads, and representatives of other groups of health professionals. Its prerogatives include: defining the hospital's annual medical research program; monitoring studies in progress; assessing departmental care, training, and research activi- ties; producing an annual report assessing the technical and eco- nomic care provided by the hospital; and answering all requests for its opinion from the minister of health or the Board of Directors. Market Exposure The reform design provided for little additional market exposure for the EPSs. Public hospitals were already earning a proportion of their revenue from out-of-pocket payment, so this relatively small portion of income was tied to the sale of services. At the time of reform, the social security institute was also supposed to pay for services ren- dered to contributors. However, this wasn't widespread until a sub- sequent, separate reform. Social Functions No reform provisions provided for social functions to be made ex- plicit, nor were steps taken to tie revenue explicitly to services ren- Autonomous Structures-with Incomplete Autonomy * 493 dered to the needy. By law, free health care is rendered to individu- als participating in scientific studies or disease prevention campaigns; victims of epidemics; citizens, spouses, and legally dependent chil- dren not affiliated with the National Social Security Fund (CNSS); and members of certain professional groups. External Environment In addition to changing the organizations (hospitals), the reforms also provided for some changes to their external environment, pri- marily in the area of the MPH funding processes. The three major sources of revenue for EPS hospitals were the MPH, the CNSS, and revenue they generate from the sale of ser- vices to paying customers. The MPH funding is segmented into op- erational and capital streams. The operational funding must cover 100 percent of payroll for as many employees as the MPH deems necessary. Though the base amounts were calculated according to various norms, prior to the reforms the annual allocations were cal- culated essentially as budget-balancing subsidies (excluding payroll and investment). Social security fund contributions were also deter- mined without reference to hospitals' real activities. The plan was that funding from the MPH would start to be tied explicitly to performance indicators (established in the performance contracts). This would reduce the uncertainty that came from the annual negotiating process-and reduce the perverse incentives as- sociated with adjusting the amount based on the hospitals' budget shortfalls. The reform package also supported efforts to tie social security contributions more directly to services rendered. This component was to include setting up a system for tracking services provided to contributors and sending out bills. Market Environment There was no attempt to put the EPSs in direct competition, either with one another or with private hospitals (other than the competi- tion that already existed for patients paying privately). However, the 494 o Innovations in Health Service Delivery added cost recovery has reduced their price advantage over private hospitals. EPSs do compete with other hospitals, both public and private, for medical staff. Low public salaries constrain the EPSs in attract- ing staff.4 The reforms provided additional flexibility to EPSs to help them recruit competent managers. EPSs exert no control over the demand for health care or rates, as the MPH determines their sources and levels of funding (table 13.1). Positions of Moaor Stakeh9olders Citizens and hospital users felt negatively about the reforms, accom- panied as they were with the shift to increased household payments. The population generally saw health care as a right and held public health facilities responsible for providing the best-quality care that was financially feasible. Health care professionals also held negative views about the re- forms, since they view services delivery as their "territory," and resented what they saw as an invasion by government. The positive results of the reforms are either less apparent or less important to staff members. Among providers of care, nurses in particular see the reforms as imposing extra work without any additional compensation. lrblls IS3. Operating Costs of Public Health Facilities, by Source of Financing 1985 1990 1995 1998 SOURCE AMOUNT° PERCENTP AMOUNT' PERCENTP AMOUNT0 PERCENTP AMOUNT0 PERCENTP Public sector 147.3 85 0 217.1 84.9 355 0 78.9 445.5 72.2 Social security funds 19.0 11.0 24.5 9 6 59.5 13.2 99.9 16.2 Own revenue 7.0 4.0 14.1 55 35.6 7.9 71 3 11.6 Total 173 3 255.7 450.1 616 7 a In millions of dinars at current prices b. Percentage of total operating costs Source Ministry of Public Health, Tunisia Autonomous Structures-with Incomplete Autonomy * 495 Implementation The personal interest taken by the president of the Republic pro- vided the driving force for the hospital reforms. The minister of pub- lic health personally holds regular evaluation meetings. A Hospital Reform Management Committee was formed at the beginning of the reform process, which brought together MPH officials, EPS CEOs, and officials from the Finance Ministry, the Economic De- velopment Ministry, and the Ministry for International Cooperation. Political backing was strong-as evidenced by the willingness to mobilize the required funding. It was also reflected in the many meetings of the Ministerial Select Committee, chaired by the presi- dent of the Republic, to monitor reform implementation. This po- litical and technical "pressure" has maintained momentum and has helped resolve difficulties that have arisen during implementation. Political and governmental stability, and the stability in office of the people responsible for putting the reforms into effect, contributed to sustaining the reform process. In addition to the Ministry of Public Health, other key minis- tries-Finance, Economic Development, and Social Affairs-have been extensively associated with the design of the reform process and have supported it. CEOs and chairmen of Governing Boards and Medical Commit- tees met periodically to debate every aspect of hospital reform. The EPSs and MPH headquarters coordinate closely, particularly on re- cruitment of managers to implement new organizational procedures and the MIS. Communication Strategy A consensus-building information strategy was conducted primarily toward medical personnel and only secondarily toward administra- tive personnel. No significant efforts were made to garner the sup- port of nurses at the start of the reform, but this choice had to be reversed as implementation proceeded. 496 0 Innovations in Health Service Delivery The general public was informed via the media. Debates in the Chamber of Deputies mirrored the course of the reform process, from analysis of objectives and implementation to results assessment and possible extension of reforms to other parts of the system. The government has also implemented a communication and in- formation strategy through seminars, day workshops, articles, and TV debates, all aimed at affecting Tunisians' relationship with their health care institutions. Public reaction was initially lukewarm, marked by fear that "the health sector was being privatized" and that access to health care would be more difficult, especially for poor or low-income groups. This mistaken assumption gained ground be- cause some EPSs had subcontracted out certain nonclinical activities and because rates and copayment levels had been raised. The general public also saw the increased seriousness about collecting from users as a sign of privatization. The university medical corps and its professional organizations were particular targets of the information campaign. Most often, re- forms were presented to them as a way of resolving their professional problems. The medical profession reacted with apprehension re- garding the new powers of EPS CEOs. At the same time, some doc- tors wanted to extend reform objectives to areas they were not de- signed for, such as the organization of training, university and hospital medical careers, and relations with the private sector. With time, understanding of the scope of the reform grew and the idea found more acceptance within the medical community. However, despite the strong political will and other favorable fac- tors, implementation was rocky. The many difficulties were evi- denced by the withdrawal of responsibility for management and staff salaries one year after it was given to the EPSs under the reforms. This event almost certainly served to undermine belief that the re- forms would "stay the course" in contentious areas. In the beginning, the hospital reform process had to be grafted onto the public sector's traditional working mechanisms. The MPH had apprehensions about the change from a vertical organization (structured around resources), to a horizontal organization (integrat- ing functional components), which led to implementation delays. Autonomous Structures-with Incomplete Autonomy * 497 This situation was especially difficult because the MAPH supervisory authorities had retained their existing organization, and the reforms were being put into effect by an external agent, the Hospital Reform project. Monitoring and Evaluation Reform monitoring was carried out by the Project Coordination Unit (PCU)5 and the MPH Computing Center. Individual facilities proposed management-procedure updates and submitted them to a steering committee (composed of EPS CEOs, computing center representatives, and the authors of the updates), which ensured stan- dardized procedures throughout the hospitals. The computing cen- ter then updated the system software to reflect the procedural changes, and the computer applications were tested and approved by users and the steering committee. In addition to its specific respon- sibilities for this project, the PCU also handled EPS board recom- mendations, monitored EPS performance (including billings), and drew up their operating budgets. Two evaluation programs were launched, one by the World Health Organization-Pan American Health Organization,6 the other, by a group of civil servants reporting to the prime minister7-in addition to periodic MPH and World Bank monitoring activities. Results Reform Design With regard to reform design, the changes that took place essentially brought about centralization under the guise of autonomy. The de- cision rights the hospitals were intended to receive were fairly mini- mal. However, the decision rights actually allocated to the hospitals and their boards were even less. Although the EPSs received the right to take out loans, this has not materialized in reality. As noted above, the reforms were not intended to delegate much labor man- agement decision rights to the hospitals, although a higher salary base was established for EPS CEOs. 498 o Innovations in Health Service Delivery The new governance framework, including lines of authority be- tween facilities and the MPH, was not elaborated. The EPSs and MPH headquarters have not been able to find a reasonable balance between centralization and decentralization, between autonomy and dependency. The powers of the EPS Governing Board turned out to be limited in practice. The medical committees did not take up their allocated tasks, mainly because of ambiguity regarding the way they should operate and their authority over staff. The performance con- tract mechanism was not established. MPH funding was never tied to performance. Efforts were made to tie revenues more directly to services rendered, and now services to CNSS contributors are billed separately. However, since the total payment is capped and the volume delivered always exceeds the ser- vices paid for, this funding stream also reduces to a block payment (untied to services). Process-Pros and Cons In addition to the reforms, other important changes were occurring. In particular, the proportion of the hospitals' income that came from the MPH was decreasing, and households were making greater con- tributions in the form of out-of-pocket payments. Increased copay- ments, and tighter eligibility for subsidized and free care, con- tributed to the shift toward household payments. The trend in shares of health care expenditures by the three major sources of funding are illustrated in table 13.2. The Ministry of Public Health resented the wide range of responsibility of the PCU in implementing the re- forms that encroached upon those of MPH departments. Some incentive-creating measures are beginning to produce im- provements in overall hospital performance. These include: adopt- ing cost-based billing and internal performance-based budgeting; giving computers to hospital departments that help to improve facil- ity performance; and increased consulting between hospital depart- ments on the formulation of hospital policies. The increased central influence reduced the independence of medical staff and was strongly resented by them. The professional Autonomous Structures-with Incomplete Autonomy * 499 Table 13.2 Health Care Spending (percent) SOURCE 1985 1990 1995 Public sector 51 38 34 Households 34 5 47 51 Social security 15 5 15 15 Source Ministry of Public Health, Tunisia, based on household's consumption surveys of the National Institute of Statistics medical organizations considered the CEO's official powers exces- sive. Medical staff also resented the ministry's prerogative of select- ing the chairman of the board, which they would prefer to see come from senior EPS staff. As implementation of reform gathered momentum, the MPH and the CEOs realized that more attention should be paid to nursing staff. Apart from a few workshops organized by the nurses' labor union and some professional organizations associated with the gov- erning party, no significant actions had targeted this group. Yet, be- cause of their natural place in the health system, nurses relate more frequendy and more closely to the general public than any other group of health care professionals. As a result, a number of actions have been initiated. The EPSs have poorly received some aspects of the reform process. These include: the practice of reducing, or merely main- taining, government subsidies for operating budgets; the withdrawal of responsibility for management and staff salaries one year after it was given to EPSs; continuing dependence on MPH headquarters for investment and manpower resources; delay in making temporary management staff permanent; and the allowance for university hos- pital physicians to work also in the private sector. Recruitment of competent, professional-rank management staff from outside the MPH remained problematic. Because of restric- tions on monetary compensation, some EPSs have begun to offer management staff benefits in kind. Despite the significant efforts made to develop capacity at the managerial level, supervision of the operational aspects of the hospital reform process has been weak. 500 o Innovations in Health Service Delivery Qualified personnel have been very hard to find and the supervi- sory authority turned out to be poorly organized. Supervision is pro- vided by a small number of experienced officials, assisted by younger staff who have no practical experience in the health care field. The operational capabilities of the executive and management teams in charge of implementing reforms in hospitals differ from one EPS to another and depend on numerous factors, such as the CEO's knowl- edge, aptitude, and methods, and the attitudes of the medical corps, especially its leaders and professional organizations. Important problems also arose in relation to CEO recruitment. The appointment criteria excluded some individuals who had di- rected university hospitals before their conversion into EPSs and, while CEOs' pay was increased, all other senior EPS management staff members remained at the same salary level as their civil service counterparts. Efforts to recruit competent managers have not been entirely successful, either, and some EPSs (about 11 percent) have hired individuals ill suited to their jobs. CEO performance still depends on the way the MPH perceives its changing role and functions and on the recruitment of EPS man- agerial staff, which has been slow. Reliance on temporary man- agement personnel has helped resolve immediate difficulties but, because temporary contracts can be precarious, many of these man- agers look for more stable and better paid jobs elsewhere. Essentially, the indirect accountability instruments were not devel- oped. Supervisory activities continued to be done in such a way that the true level of autonomy for hospital CEOs and boards is quite low. In this way, the reforms actually led to increased central control over hospitals. Overall Positive Outcomes Even though many of the organizational elements of the reform didn't work out, the hospital reform package and associated invest- ments have resulted in clear benefits. Included among these are improved patient and staff working conditions at EPSs due to Autonomous Structures-with Incomplete Autonomy * 501 building renovation and new equipment, expanded access to uni- versity hospital services, and more cost-conscious physician behav- ior. Physicians are making a transition from a department-oriented mentality to a corporate-hospital mentality, one more appropri- ate for an organization whose mission is to provide services. The government has been able to reduce the burden of covering defi- cits as health sector operating budgets have decreased, and social security funds have increased financing for their members' care. Clinical and administrative departments are also working together. Some observers believe that health care professionals no longer see an operating budget expansion as the outcome of a resource- mobilization campaign, but rather as the result of meeting facility goals. In addition, the number of EPSs running budget deficits de- creased between 1991 and 1997. In 1991, 18 out of 20 EPSs re- corded operating budget deficits of between 1 percent and 84 per- cent; only 12 EPSs ran deficits in 1997. Seven EPSs have almost eradicated their budget shortfalls (in the 3-to-84 percent range). Six EPSs have reduced their deficits since 1991, and 6 others have seen theirs worsen. Some budget deficits are explained by increases in production costs attributable to higher input costs, introduction of new technology, and persistent inefficient operation of certain health care and diagnostic facilities. However, the gap is gradually narrowing in most EPSs as all players become more cost conscious. Budgeting on a departmental basis has made staff members more accountable for the financial consequences of their activities. De- spite initial negative reactions, there is now more constructive crit- icism from all concerned and commitment to tracking costs in their departments. Although the steady rise in copayment levels is often perceived in negative terms, it appears that acceptance is growing among the populace that copayment for services is a legitimate financing means, and that subsidies to help the poor should be operated to benefit only the poor. 502 o Innovations in Health Service Delivery Wmpld Evaluation of the impact of reform is still hampered by a shortage of hard information, particularly on unit costs. Impact on the Healtb System The entire package of hospital reforms has increased the supply of services and productivity without any perceptible impact on equity. Equity and Accessibility Reductions in the length of stay have enhanced access to university hospitals by ensuring speedier turnover of hospital beds. As for out- patient care, introducing afternoon consultations and alternatives to hospitalization has also improved access, but there is still some work to be done in terms of organization. Low-income citizens who qualify for free or reduced-cost health care still use EPSs as much as ever. Under the Free Medical Care System, 130,000 households with limited incomes still receive care. For individuals subject to copayment (those entitled to reduced rates and social security insureds), all increases were imposed simultane- ously with income increases under the government's salary and wage policy, in an attempt to reduce any financial impact of the copay- ment increases. EPS activity grew between 1991 and 1997. Outpatient services rose by 32 percent, and hospital admissions, by 18 percent; however, they now seem to be leveling out. Access to university hospitals has improved: bed availability rose as the average length of stay fell and as alternatives to hospitalization improved access to ambulatory care. At the same time, productivity rose: the consultations-to-admissions ratio rose by 12.5 percent, average length of stay fell by one day, and bed turnover rose by 9 percent. Effectiveness EPS general performance indicators indicate an overall improve- ment in effectiveness. The increase in the number of hospitalizations Autonomous Structures-with Incomplete Autonomy * 503 helped reduce admission waiting lists. The decline in inpatient stays made 900 extra EPS beds available. Hospital services significantly improved as a result of building renovation, replacement of medical equipment, and the introduction of a maintenance strategy. Use of inpatient capacity also improved. Between 1992 and 1997, the bed-turnover rate increased from 34 to 3 7 patients per bed, while the bed-occupancy rate stayed at 75 percent. In addition, a new or- ganizational approach to the use of major equipment (e.g., MRI units, scanners, cardiac catheterization units) allows them to be shared by different hospitals. Recovery of receivables from third parties also improved, allowing additional resources to be mobilized and adding to the EPSs' own revenue. Productivity Resource flows grew by 3.3 percent between 1991 and 1997, en- abling several EPSs to wipe out or reduce their deficits. This growth rate, which was less than the increase in activity, nonetheless enabled EPSs to handle the increased activity while also improving patient care. Human resource mobilization by the EPSs in all categories rose by only 2.3 percent a year. Excluding management, the target of con- certed additional recruitment efforts, this rate drops to only 1.9 per- cent, whereas business developed much faster. This may denote an increase in employees' workload and an improvement in their productivity. Quality Because the data necessary to construct indicators to monitor EPS clinical quality are not yet available, changes in the quality of care cannot be fully assessed. Nonetheless, consumer quality has im- proved, especially in regard to patient admissions, board, and lodg- ing as a result of investments in these areas. Auxiliary services (e.g., beds and bedding, preparation and distribution of meals) have also improved since the contracting out of kitchen activities under the 504 a Innovations in Health Service Delivery technical supervision of EPS nutritionists and hygienists. In addi- tion, training and education initiatives have improved in-house man- agement of hospital waste and led to improvement in the hospital environment. Sustainability To last, reform has to be politically, economically, institutionally, and socially viable. There has been gradual expansion in the reforms' content and geographical scope, indicating growing commitment and acceptance of the reforms. The population is generally believed to view the changes positively, as these changes have improved both the quality of services and access to them and have not led to the elimination of services. There is evidence of a swing in the attitudes of EPS personnel to- ward acceptance of the reforms. This gradual change is a result of consistent political support, extensive sensitization and information programs, prompt and appropriate adjustments whenever necessary, and judicious resolution of internal conflicts. Economic and Financial Feasibility Financial sustainability is not a serious issue since the government retained control over budgets, investments, health care charges, and payment arrangements. Thus, the health sector operating budget (excluding payroll) has developed in step with available public re- sources (table 13.3). In fact, allowing for annual depreciation of 10 percent between 1991 and 1997, the budget has grown by an annual average of 3.3 percent. As for payroll increases, the government has kept a tight lid on both salaries and job numbers. At the hospital level, expenditures are now controlled through prescription monitoring, daily patient drug logs, and use of thera- peutic guidelines.8 Some EPS Medical Committees have formed therapeutic subcommittees to systematically review prescriptions identified as "unusual" by hospital pharmacies. Certain expensive di- Autonomous Structures-with Incomplete Autonomy * 505 Table 13.3 Changes in EPS Operating Budgets - TOTAL BUDGET PERCENTAGE PERCENTAGE AVERAGE ANNUAL jIN MILLIONS CHANGE, CHANGE, PERCENTAGE 1 YEAR OF DINARSO) YEAR TO YEAR 1997-91 CHANGE 1991 35 303 n a n a n a 1992 38 82 10 1993 44 480 15 1994 49 923 12 1995 56 279 13 1996 67 956 21 1997 80 733 19 129 15 * a Not applicable EPS government-owned health corporation a Current prices Source Ministry of Public Health, Tunisia agnostic procedures (e.g., scans, magnetic resonance imaging) are also monitored. Other aids to close the deficit include more successful cost recov- ery as a result of personnel training; more flexible management of loans; rate and copayment increases; and, since 1996, additional so- cial security funding. Budgeting procedures are also becoming in- creasingly sophisticated, enabling negotiations with EPSs to include elements of performance evaluation. It is now possible to precisely estimate each hospital's real self-financing capacities. The impact of hospital reforms on fast-rising health costs is diffi- cult to judge. Between 1991 and 1997, the overall nonpayroll oper- ating budget for the EPSs increased 15 percent a year (average, at current prices). This increase was covered by state allocations and the EPSs' own income, including billing for services to patients cov- ered by CNSS. Cost recovery accounted for 30 percent of the in- crease in the EPSs' own revenue. This increase in public expenditure remained in hne with overall budgets. In fact, the portion of the MPH budget, incorporating both operations (including payroll) and investments, fell from 7.16 percent of the national budget in 1992 to 6.95 percent in 1997, and rose from 2.49 percent to 2.64 percent of GDP over the same period. These 506 o Innovations in Health Service Delivery figures include all revenue of the public health facilities, 95 percent of it from private expenditure. Risks for the future. Risks include both systemic and implementation factors. The main recurrent risk is that the primary reform objec- tive-identifying the actual cost of services and pricing services to reflect cost-may not be achieved and instead may become an accounting exercise. Further raising of copayments might, in the medium term, limit access to health care. The increased emphasis on controlling costs without the ability to monitor quality may have an adverse impact on quality of care. Systemic factors presenting risks include the trend toward the centralization of decisionmaking, which may undermine hospital administrators' credibility, and the interdependence of hospital per- formance with other elements in the chain of health sector services. Implementation matters are also important: delays in reorganizing supervisory arrangements at MPH headquarters and in defining the new missions of the regional public health authorities may slow reform. The emphasis placed on the university hospitals was justified and timely. However, their performance is inextricably bound up with the performances of all the other elements in the health service chain. By continuing to direct resources toward these EPSs, Tunisia may dislocate the health system and undermine the credibility of other sector agencies, thereby producing the opposite effect of the one intended. Conclusions Reform has produced a number of positive results, which now need consolidation. It has also left areas needing improvement. Positive Results Positive results can be found in financial management, deployment of management staff, availability of reliable statistics, renewal of in- Autonomous Structures-with Incomplete Autonomy * 507 frastructure, revision of budget-financing mechanisms and entitle- ments, and governance. Financial management. The resource constraints of the university hos- pitals led to the introduction of ways of controlling expenditure and mobilizing additional resources, especially through the collection of receivables. The introduction of billing for social security produced excellent results in terms of control of procedures and the injection of supplementary financial resources into the hospital system. Availability of reliable statistical information. Since the reform process began, the MPH has kept up-to-date statistical information on EPS operations, although it is still fairly general in nature. All statistics on activities and resources communicated by the EPSs are used as the basis for budgetary negotiations with the EPSs themselves and with the social security funds. Infrastructure renewal. The package of reforms included a program to modernize the technical facilities of university hospitals and to reor- ganize maintenance. A previously thorny problem has thus been re- solved, improving the environment for patient care. Management organs. The development of the Governing Boards and Medical Committees in the hospitals, while problematic, did have positive effects. It encouraged members of the hospital community to work together and with the outside world to determine their hos- pitals' priorities. The boards and the Medical Committees began to jointly examine their hospitals' problem areas and to seek solutions to their difficulties. Areas Needing Improvement Despite the progress from these reforms, some shortcomings re- mained to be corrected. Revision of EPS payment arrangements. An important element of the incentives that the 1992 reforms did not address was the payment 508 o Innovations in Health Service Delivery arrangements. In 1996, changes were at last made in the way EPS operating budgets are funded. These changes were based on the principle that the EPSs are providers of services and must recoup their operating costs for treating patients from different sources of funding. The process began with social security fund contributors and will gradually be extended to the other population groups cov- ered by the state. Changes in socialfunctions. Subsequent to the 1992 reforrns, the au- thorities revised the conditions for entitlement to free health care. Processing and management have been transferred from the MPH to the Social Affairs Ministry, which maintains up-to-date records on needy population groups. CNSS has been given responsibility for processing applications for reduced-rate care to encourage the pub- lic to join social security programs. Failure to fully develop the concept of EPS autonomy. Transfer of power from MPH headquarters or the regional public health authorities to EPS management proved difficult, and, as noted, led to central- ization. Inefficiency and lack of credibility were cited as conditions requiring comanagement by the supervisory authorities. If man- agement is ever to be delegated to the hospitals, then hospital man- agement capacity will need to be increased. Under the current organizational structure, the MPH could easily delegate more autonomy to individual EPSs that demonstrated they could effec- tively manage their own planning, organization, assessment, and monitoring. Shortcomings of the MPH supervisory authority. Like the hospitals themselves, the central MPH in Tunis has accumulated its own man- agement experience and traditions, which are very hard to change, although everyone concerned recognizes that the current organi- zation is obsolete. Apprehension continues to feed sluggishness regarding the transition from a vertical organization, structured around resources, to a functional horizontal organization, integrat- ing all the operational components of the public health facilities. In Autonomous Structures-with Incomplete Autonomy * 509 order for the MPH to cease its direct intervention in hospital ad- ministration, there will have to be more development of the needed indirect mechanisms (such as performance contracts) to hold au- tonomous hospitals accountable. Lessons Learned Tunisia's experience with health care reform has resulted in the fol- lowing lessons: * Involve sector professionals and all sector constituencies in the re- form design process. * Define and implement a communication strategy from the start of the reform process that targets all sector constituencies. * Define the political, strategic, and operational goals of reform and the criteria to be used in determining whether the goals were accomplished. * Assign specific management responsibilities for implementation in advance of initiating reform measures. * Include feedback loops in reform management to allow necessary adjustments and flexibility in view of implementation experience and changes in the external environment. * MPH role and methods of interacting with and guiding the hos- pitals must be consistent with the envisioned level of autonomy- otherwise, they will undermine the reform. * Develop incentives and qualifications criteria to ensure recruit- ment of appropriate reform managers. * Ensure stability in the units involved in reform by avoiding staff turnover. * Promote information exchange between and among providers and the Central MPH. * Conduct continual staff training and institutional development. 510 ° Innovations in Health Service Delivery o Ensure that the focus of all action in health care institutions is on patients. Notes 1. See chapter 1 of this volume for a description of these types of reforms. 2. Government of Tunisia, Law No. 91-63, July 29, 1991. 3. Government of Tunisia, Decree No. 91-1844, December 2, 1991. 4. However, public hospitals do have an advantage in the compe- tition, namely, that physicians, pharmacists, and dentists interested in academic careers must practice in the public sector. In addition, public hospitals generally have access to new technology and major equipment, which makes them more attractive to physicians. 5. This was the PCU for the World Bank project that provided technical assistance to support the reform process. 6. World Health Organization-Pan American Health Organiza- tion, "Progress Report on Implementation of the Hospital Reform Project in Tunisia," Tunis, November 1993. 7. Ministry of Public Health (MPH)/National Institute of Public Health, 1956-1986: Trente ans au service de la sante: un engagement, une ethique (Tunis: MPH, May 1986). 8. All the therapeutic measures recommended for a given pathol- ogy. They are circulated in a review called Consensus, published by MPH since June 1997. CHAPTER 14 Autonomization in Indonesia: The Wrong Path to Reduce Hospital Expenditures Samuel S. Lieberman and Ali Alkatiri In the early 1990s, Indonesia initiated far-reaching organizational re- forms in 61 of its public hospitals, m response to a fiscal crisis and sharp cuts in the health budget. In these reforms, selected govern- ment-owned hospitals were permitted to operate as largely auton- omous institutions, according to specified rules and understandings. A core component of the reform was for these hospitals to increase their reliance on private funding. The government hoped to be able to shift budgetary resources toward lower level curative and preven- tative care. While some positive results were achieved, the primary objective of decreasing government spending on hospitals was not. Therefore, a decade later, policymakers, disenchanted with the swa- dana (autonomous) approach to hospital reforms, are looking for better options. Swadana Hospitals: Goals, Modalities, and Results The Public Hospital Sector At the end of 1989, Indonesia had 327 public hospitals, divided among four types or classes. At the lowest level, there were 182 Class 511 512 o Innovations in Health Service Delivery D hospitals. Class D hospitals provide general services, and often do not have specialist physicians or laboratory and x-ray services. The 120 Class C general hospitals are designed to provide specialist ser- vices in internal medicine, obstetrics and gynecology, and pediatrics. Both Class C and D hospitals are largely rural and are province- or district-owned. Most of their funding comes from higher levels of government, as well as external sources. The funding that comes to these hospitals is earmarked, and must be spent in its intended cate- gory. Funds for construction, equipment, and training costs are allo- cated as "development funds" by the Ministry of Health (MOH). Separate funding streams come from the MOH for building and equipment maintenance and operations, salaries (of centrally ap- pointed staff), and the costs of medicines and other consumables. While provincial and district governments supplement these funds with some allocations for salaries, other routine expenditures, and some investments, the centrally allocated revenues are typically much larger. Hence, unlike the situation in most countries, local governments here typically derive significant revenues from their public hospitals, creating some rather strong and perverse incentives vis-a-vis hospital governance. Class B hospitals usually operate 10 specialty departments, and are centrally owned and administered. Two Class A hospitals are highly specialized referral centers and are also centrally owned and adminis- tered. These are very large facilities, and combined have 2,900 beds. In all classes, any fee-based income the hospitals receive must be channeled directly to the central, provincial, or district bodies that own the facilities. Slightly older figures on other facilities indicate that in 1983/84, there were 115 facilities operated by other ministries and parastatal entities, as well as 175 private hospitals and 80 hospitals run by non- governmental organizations (NGOs). Private and NGO-owned fa- cilities accounted for roughly 40 percent of Indonesia's total hospital bed capacity. The five-year plan for 1984/89 (Repelita IV) was intended to focus on the hospitals needed to back up the fledgling network of largely rural health centers and affiliated service units. Improved referral ar- Autonornization in Indonesia * 513 rangements and capacity were seen as preconditions for increasing service use by the poor-policymakers had been aware for some time that government hospital services benefited middle- and upper in- come families disproportionately. During this period, the share of government health expenditure devoted to hospital investment was projected to rise sharply, while oudays on the health center network were to account for a much lower proportion of sector spending. Priority was to be given to improving the Class C and D hospitals, which would then provide technical support to nearby health cen- ters and handle an increased referral flow from them. The longer term strategy, beyond 1989, was to convert and upgrade the remain- ing Class D hospitals into Class C facilities so that they could pro- vide the general referral and support complement needed by rural health centers. Because of the oil shock-induced fiscal crisis that began in 1986, overall health development and public hospital spending fell in real terms by nearly two-thirds in 1986/87 and again by the same pro- portion in 1987/88, as the MOH focused on meeting the large and rising recurrent cost obligations associated with a decade of hospi- tal and health center expansion. Health development expenditures began to recover in 1988/89 and reached precrisis levels by the early 1990s.' Despite the focus on recurrent spending, a 1990 internal MOH study found that maintenance, staff performance, financial and hospital management, and quality assurance remained unsatis- factory in many hospitals.2 Confronted with increased fiscal constraints, the Indonesian gov- ernment limited construction of new public hospitals, and focused on priority recurrent spending and operations in existing Class C hospitals. At the same time, the MOH developed new policies to try to create accessible and affordable health services for the entire pop- ulation, including the poor. These initiatives included efforts to inte- grate health planning and budgeting at the district level, experiments with health cards as a means of targeting subsidized services to the poor, and a supportive stance toward expanding private health insur- ance and testing managed care and other prepayment arrangements to expand health spending by local communities. The government 514 o Innovations in Health Service Delivery also made regulatory changes to encourage domestic, and later for- eign, private investment in the hospital sector. Reform Design The policy package that emerged in the late 1980s also included or- ganizational reform of publicly owned hospitals. The reform was aimed primarily at securing additional private resources for health, thereby freeing up government funds for reallocation to promotive, preventive, and other public health services. Under the reforms, public hospitals meeting certain criteria were formally converted to autonomous status (Lembaga Swadana, LS). The government set specific criteria for transforming to swadana status, recognizing that not every hospital had the capacity to oper- ate successfully under the looser rules. Hospitals were eligible if the cost-recovery rate had increased during the previous three years and exceeded 40 percent, if their bed occupancy rate (BOR) was 70 per- cent or higher for centrally owned hospitals and at least 60 percent for province and district-owned facilities, if the average length of stay was 10 days or less, and if their surrounding communities were pros- perous enough to pay for medical services. Strong support from hos- pital administrators and central MOH or local government officials was also required. Swadana hospitals received some decision rights previously held by the MOH. They were authorized to increase or reduce certain fees, for example, and were allowed to retain resultant revenues. Their residual claim on these revenues was constrained, however, in that the hospitals were only allowed to use fee-based revenues for certain purposes. They could use fee-based revenue to improve ser- vices and facility utilization and in that way to augment future fee- based funding prospects. Thus, income from fees could be spent on drugs and medical consumables, but not on, for example, civil works and equipment. Fee-based revenue could also be spent on recruit- ment of new employees, staff incentives, and contracts with private service providers, which increased the hospitals' flexibility to manage human resources. These revenues had to be integrated, together Autonornization in Indonesia * 515 with conventional government grants and payments, into the annual funding proposals to central, provincial, or local authorities. Despite increased flexibility, personnel management is still highly constrained in swadana facilities. Hospital staff remained civil ser- vants, so most decision rights over staff, including hiring and firing, remained with the MOH. The swadana reforms expanded market exposure primarily in two areas. First, hospitals could contract with private providers, subject- ing their relations with at least some doctors to market forces. More significant, their increased reliance on fee-based revenue greatly en- hanced their exposure to pressure in the market for their services. The reforms did not provide for any new or reformulated ac- countability mechanisms, perhaps because of the fairly limited deci- sion rights that were transferred to the swadana hospitals. In addition to improving staff morale and motivation, it was also hoped that the reforms would make hospital services more accessible and affordable for the poor. And so, besides mobilizing additional private funds and reducing government subsidies to the hospital sector, the reformed hospitals were tasked with the important so- cial function of providing services to the poor. The plan was to ac- company the swadana initiative with an important complementary reform-to introduce more transparent means of subsidizing hospi- tal services for the poor. As a further protection, the MOH mandated that Class III wards, which provide the most basic services for the lowest fees (box 14.1), account for at least half the beds in each LS facility. Implementation, Outcomes, and Possible Determinants Implementation. An enabling presidential decree was enacted in 1991, and the first swadana hospitals were selected for pilot testing the same year. Implementation began in five hospitals in Java, one provincially owned and run and the rest under central MOH man- agement. Another 11 hospitals, owned and managed by provincial and district governments, were given swadana status in 1993. By 1997, 61 hospitals were operating on swadana principles. These in- 516 o Innovations in Health Service Delivery Lox I4.I Ward Classes and lniernal Coss-Subsidies in Indonesian Public HospiWls As in many East Asian public hospital systems, including Hong Kong and Singapore, Indonesia relies on a tiered price structure for hospital services according to the ward in which the patient is located. The pricing policy in Indone- sian public hospitals is to set Class I ward fees above actual costs, Class II ward fees equal to actual costs, and Class m fees below actual costs. Clinical services are intended to be the same across the wards. However, hotel services are dif- ferentiated, to encourage the better-off to opt for the higher price, Class I ward regime. This structure provides for sub- sidies to flow from the hospitals better-off patients to its poorer patients, and constitutes, in essence, a very decen- tralized mechanism for risk pooling. cluded 13 general hospitals and 2 specialized hospitals owned by the MOH. In addition, 46 hospitals, owned and run by provincial or local governments, had been converted-36 of these facilities were in Java. Of the rest, 6 were Class B facilities, including 4 in Sumatra. Outside Java, there were only the Class C swadana hospitals, and 2 of them were in Sumatra. Based on early results, some policymakers were predicting that the next five-year development plan would pro- vide for extensive replication of the swadana approach.3 Impact. The reforms have been evaluated through various case stud- ies. The studies include an initial assessment by A. Gani;4 follow-up reviews by T Bossert and colleagues,5 a review mission funded by the U.S. Agency for International Development (USAID)6 a team led by R. Malik;7 and a further look at facility-level outcomes conducted as part of the present study. Gani's evaluation was based on visits in 1995 to West Java's Tangerang and Sumedang hospitals, which were part of the second Autonomization in Indonesia * 517 cohort of facilities converted to swadana status in 1993. He con- cluded that the swadana approach had succeeded in increasing hos- pital revenue substantially and improving service quality, while sus- taining the use of facilities by the poor.8 Subsequent reviewers were somewhat less enthusiastic than Gani about the impact and usefulness of the swadana reform. The 1997 assessment by Bossert and coauthors looked at 5 facilities, 4 of them in the initial group of swadana hospitals. The sample also included 3 provincial or district hospitals that had not been run along swadana lines, and 2 private hospitals. The 15 hospitals evaluated by Malik and colleagues in 1997 included swadana and nonswadana facilities. Reference is also made to the 1998 USAID review of swadana per- formance in the Dr. Suradji Hospital in Klaten in Central Java.9 The sample for the present study consisted of six swadana units (two owned by the MOH, one province-owned, and three district- owned) and one traditional district hospital. These facilities were vis- ited in 1998. Results from these case studies are summarized below according to different performance dimensions. Fee Increases Managers of swadana hospitals can decide on fees, including daily charges for the use of beds in the (VIP) Class I and II wards. In our survey, we found that the fees set in different swadana hospitals vary substantially by service class within, as well as between, the facilities. User charges by ward class also rose at different rates in the period 1994/95 to 1997/98. For example, Fatmawati Hospital increased room charges in its VIP wings substantially in 1995/96, while little change was observed in the VIP or other wards of Hasan Sadikin Hospital (tables 14.1 and 14.2). Similar findings emerged for the sample reviewed by Bossert and colleagues.10 Funding, Expenditure, and Subsidy Trends Fee-based income increased in all of the hospitals, swadana and non- swadana, in this study due to increased bed charges and relatively inelastic consumer demand. The rise in income was largest in 008~0000 0 080808 0 o LO w) 0 'L OLt) ol o~0000000 0 Oin N L O w)t 00008800 o o so 0 0 ) 88888g8o L U o o o o 0 0 LI,~~~~c o W S-c o NCN o o o 88 8° O i >L m ^ N m , , N e 0 00 0 0 D 9 m CD O 0 ° O -D _ st ,,% - 7 tc '2 9 go 0008o D _ ~~N CN - nN _ 8 U E o ° > ° ° 22N >n I v) 0 O '6J_ ~~~00~~~~ -w o 0eco i >n L osi,Z 518 0 -u E _ a ~~~00 - -~~~~~~ U - >, o a .0- 0~~~~~~~ 518 V ~~~00 0 00 00 0 00 0 0 000 00 0 00 0 0 000 OO0O0O : O O Lo 0 0 0 0 Le) 10 n 'qOC) 0 N 00 0 00 00 00 0 00 0 00 00 0 00 "O - r 0 0 0 000'C N C'4 ltd 0 0 000 0 00 00C l 0o O >000 O 00 000 C\ O O O O 0 00 0 0 U iV) 1~~~ Y1 o 'O ''co 0'0 C4 1) C's 0 00 00 00 0 00 00 0 00 0 000 0 -'o 0 _^ 0 e O ) W) 0000 o ;v>' ~ ~ ' 0f< O C OO O q O _ co~ 0O " o-CN V o o T S sT COv N N O Nu q z, '. j-o T ~ o u- rN -1 Cn C N-O4 z N -NN 0 O O ' N co co CN - Z N n OL LO o 'W M2 CO 0 N CN Cn CN - CC 4 ~ ~ ~ ~ ~ c m CN | ig. 4 Il Iq 0~~~~~0 'nl CN 0 N O m cO 0 sttn O -. t _ 4 ~ _ C n Nq r- O Oq O N4 N0 .'~~~ ~~ O. 0 0 N 0cN * ol 0s * O CN C)(O 0) c X C, 8 - o o o4 o4 N N N n n 0 10 OCN4 0 0 LO C ) L Lo D M C N N -0 CN (0 C' ' . O O Ei O E 0 0) 0) >i > 0 U N N N u O O : _ _ _ O O ___ IJ Q , >_ = DC 519 520 o Innovations in Health Service Delivery lrcaWe 14.3 Increase in Sample Hospital Revenues, by Year (percent) PERCENTAGE INCREASE BY YEAR NO. HOSPITAL 1995/96 1996/97 1997/98 1 Hasan Sadikin 60 18 13 2 Fatmawati 22 18 20 3 Syaiful Anwar 41 25 29 4 Serang 35 52 7 5 Banyumas 70 41 33 6 Karawang 38 30 42 7 Cianjur 27 22 25 Note Shoding denotes nonswadona Source Data collected from sample hospitals Banyumas Hospital (table 14.3). Revenues also rose sharply in Cianjur Hospital, a nonswadana facility. What was surprising was that rising fee-based revenues in MOH-owned facilities were accom- panied by increased outlays, supported by subsidies from the Indo- nesian government, directed mainly at construction of new wards and equipment upgrades (tables 14.4 to 14.6). A similar result was seen in the province- and district-owned hospitals, including the nonswadana Cianjur facility but excluding the swadana Banyumas facility. Thus, increased subsidy flows to reformed as well as nonre- formed hospitals was seen in most of the cases. Bossert and his coau- thors1I noted the same results regarding subsidy trends. Moreover, the USAID review team12 found that the long-running swadana pilot in the Dr. Suradji Hospital in Klaten also benefited from con- tinuing substantial budgetary support from the MOH. Physical Outcomes and Efficiency Trends Swadana hospitals appeared to improve efficiency. Bed occupancy rates and other efficiency indicators rose in the swadana hospitals we examined (table 14.7). Bossert and coauthors13 and Malik and his team14 reported similar findings. Service Quality Standards of service and patient satisfaction are notoriously difficult to measure. All facilities covered in this study established total Cl) - 0 N ONL LO MM VN >- 0 0 0 cs Xv 0 0 : )o c0 o- C O O 0 10 v O' O 'C; O 00MM L O V - N LO O 0 K ": -C' oC "I co f - D L C N CNN a0 00 0' N U D O0''0C4 C (N4 - 0'~~0 Q 0-00C"J0c'0 - ' 0 0N- J0 o oCl 0cn01 C- C'N'a O 0i 10 4 CO e K_ -o ,, 0 - C C O -C A 5 Z Q t 0 4 0 1 ) t co e l CO S C 0' 0 -o 0 C) - , 00L 0QN- O C) > 0 o, N O-ON a) Cg Iq : U N0' CM ' CN t E I o .2- ~0 U, ~ ~ ~~ 00) o oo<( oo U 0) D _ I_ O E- ° S ° o 0 521 522 o Innovations m Health Service Delivery T7caWe 14.5 Beds in Sample Hospitals, by Year NUMBER OF BEDS HOSPITAL 1994/95 1995/96 1996/97 1997/98 Hasan Sadikin 882 958 925 951 Fatmawati 564 569 582 598 Syaiful Anwar 760 760 744 753 Serang 272 272 272 272 Banyumas 180 200 220 235 Karawang 167 167 167 207 Cioniur 150 150 150 152 Source Data collected from sample hospitals quality-management programs with wide staff involvement, backed by quality assurance and medical committees. Moreover, this sample of swadana facilities developed standard operating procedures cover- ing medical interventions as well as financial and reporting activities. Karawang Hospital initiated the use of patient surveys and other feedback mechanisms, a sign that at least some hospitals were start- ing to focus more on consumer quality. Personnel Developments and Incentive Payments There were some improvements in staff behavior, probably as a result of management ability to structure the incentive payments, and to contract with private providers. We compared staff remuner- ation in two facilities, one run on swadana principles (Karawang Hospital) and one on traditional (Cianjur Hospital). The swadana approach appears to have had little impact on staff pay packages overall (table 14.8). However, interviews revealed that staff incentive packages seem to have been better designed in the swadana facility. Hospital and personnel directors in the swadana hospital felt that staff discipline and motivation had improved, while absenteeism had fallen. Similarly, Bossert and colleagues15 found reduced absenteeism as well in swadana hospitals. But despite increased flexibility, person- nel management remains constrained in swadana facilities. This is because the hiring and firing of hospital staff members who are civil servants remain a responsibility of the MOH. _ 0 N. CN N ol Wn o m i C _ cs 0 C. ) c ) O C . . -0' NX 0' 1'- m > 0 Ol 0s c0 u) N r 0' C- 00 0 as W ) CN 10 aa 01 0 E 10 0t .0 04 , L 0 C -O 0 00 OD-'0 0 CN -- cN 0 O l O w ur _ 0 ')~~ Civ' 'C-0~ ) _ n 0 (N0'J - -Co O LO u-N CO W 0, r .0 C E ) o--- CL ._ _ s~~0 n cn N c" m N I ) O O 0 0 ~~~~~O (> N V 'O CD L)O s ~~~~1 , 0 - cot CN v eN S n ~~CN CO C x . a) . i. I . c, o o _ a) N ( -o X C' m 0 0 N (N a:- X . 0 LO l q In LO N E I~ ~~~~o C, C- ffi o = C-4 F- "0)~~~~~~ e o -00< O C 523 524 o Innovations in Health Service Delivery al3e 14.Y Bed Occupancy Rates in Sample Hospitals BED OCCUPANCY RATES NO HOSPITAL 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1 Hason Sadikin 62.3 63 1 63.1 63.4 63.9 62.3 63.0 2 Fatmawati 67 6 69.0 69.5 74.9 75.3 77.7 72.2 3 Syaiful Anwar 62 9 63 9 61 2 61.7 59 1 57.7 58.0 4 Serang 65 4 70.3 74 3 54.5 70 3 78.2 75.3 5 Banyumas 81 2 80.6 85 6 82.0 86.0 90.0 92 5 6 Karawang 65 7 58.2 57.5 57.4 64.7 78.2 84.3 7 Cianjur 58.3 52 4 57.5 62 6 63 4 63.0 69 2 Note Shading denotes period of reformed status. Source. Data collected from sample hospitals Utilization and Equity As noted above, the Class I wards are used by wealthier patients, who effectively subsidize the third-class beds that are assumed to be used by the poor. Prices for services to patients in Class Em wards must be approved by the MOH for centrally owned hospitals and by local leg- islative bodies for facilities owned by local governments. Thus, stat- ed fees in Class m wards showed negligible increases in the hospitals studied, swadana and nonswadana (table 14.1), and the intended TaiDe 1I4.8 Trends in Average Incentive Package, by Staff Category (in rupiahs) STAFF INCENTIVES NUMBER CATEGORY OF STAFF OF STAFF 1994/95 1995/96 1996/97 1997/98 KARAWANG HOSPITAL Medical staff 42 7,611,480 7,919,640 9,981,610 13,916,260 Nurses and paramedics 288 461,890 383,850 463,490 606,860 Administrative staff 171 434,730 511,920 643,730 888,040 Total 501 1,106,780 1,077,300 1,331,600 1,810,590 CIANJUR HOSPITAL Medical staff 24 5,998,100 7,167,470 9,912,050 9,039,410 Nurses and paramedics 180 73,480 174,930 801,210 872,660 Administrative staff 128 216,790 209,640 356,850 360,830 Total 332 567,180 778,880 1,233,970 1,265,690 Source. Data collected from sample hospitals Autonomization in Indonesia * 525 Table 14.9 Occupancy Rates in MOH-Owned Hospitals 1994/95 1995/96 1996/97 1997/98 WARD CLASS BEDS BOR BEDS BOR BEDS BOR BEDS BOR HASAN SADIKI VIP 22 42 4 26 74 8 17 56 6 34 45 4 Class 1 79 55 2 76 58 1 76 44 8 99 54 4 Class 11 148 58 8 154 65.0 138 65 0 173 60 1 Class III 544 72 2 606 70 5 598 70 9 596 66 0 Special class 89 27 6 96 244 96 300 49 70 7 Total/average 882 63 4 958 63 9 925 62 3 951 63 0 FATMAWATI Super VIP 0 - 9 16 7 7 38 1 7 26.0 VIP 48 70 1 50 760 59 78 1 22 673 ClassI 17 708 17 718 17 776 44 65 1 Class l 118 640 118 663 121 704 132 612 Class III 371 79 3 365 80 3 368 81 4 329 78 7 Special class 10 64 7 10 64,0 10 74 5 9 95 3 Total/averoge 564 74 8 569 75.3 582 77 7 598 72 2 BOR Bed occupancy rate Source. Data collected from sample hospitals cross-subsidization seems to have been achieved (table 14.2). The MOH requires that Class Ell beds not fall below 50 percent of the total-placing a floor on how much hospitals may reduce their money-losing beds and patients. However, the number (and share) of Class m beds did appear to fall in some swadana hospitals (table 14.9). For example, the share of Class 111 beds in total bed capacity in Fatmawati fell from 65 percent in 1994/95 to 55 percent in 1997/98. Accordingly, the net impact on the poor remains unclear. An Overview of the Swadona Initiative In short, several assessments of the performance of swadana hospitals have arrived at similar findings, namely, that the swadana approach has not made major advances toward goals related to hospital financ- ing, access for the poor, personnel management, service quality, and patient satisfaction. 526 e Innovations in Health Service Delivery Yet this does not mean that individual swadana hospitals have "failed." On the contrary, facility managers have shown they are capable of using their discretionary powers to alter the range of ser- vices offered to their customers and to raise fees and increase rev- enues from patients. There are also indications that steps have been taken to improve staff benefits and morale and to increase service quality in swadana units. However, these measures have not been used to underpin the broad reforms of the public hospital sector that were envisaged in the late 1980s. In particular, the swadana initiative has not reduced the subsidies provided to many government-owned hospitals-public hospitals have continued, even during a major fi- nancial crisis, to absorb a fifth or more of the health budget. This support has continued despite continuing staffing misallocations and imbalances, facility underutilization by all income groups but espe- cially the poor, and other indicators of inefficiency. Accounting for these seemingly disappointing results is not diffi- cult. The swadana initiative began to extract public hospitals from the maze of regulations, procedures, understandings, and expecta- tions associated with being integrated units of the MOH. As dis- cussed, measures were adopted to alter the governance (i.e., the way the MOH and other owners interacted with their hospitals) of a small but not insignificant number of public hospitals. These reallo- cations of decision rights gave facility managers wider discretion in crucial spheres such as setting user fees, which resulted in enhanced exposure to various market pressures, including the demand for ser- vices from publicly owned health insurance companies. These fac- tors also allowed management to make further adaptations, which changed the incentives faced by staff and patients, as well as other stakeholders of swadana hospitals. All these steps were not strong enough, however, to substantially relieve the problems with the public hospital sector. For instance, while certain gains in efficiency and service quality were apparent in the swadana hospitals, the government was unable to parlay this development into making reduced subsidies available to participating hospitals. And the impact of the reforms on services to poor patients is not clear. WVhile swadana facilities had to maintain half their beds Autonomization in Indonesia * 527 for the use of subsidized (i.e., money-losing) patients, it's not clear how many services these patients actually used. Looking back then, Indonesia's swadana measures lacked the coherence and critical mass needed to engender desired changes in facility-level behavior and performance. Reformed hospitals were granted only limited decision rights, were only weakly exposed to market pressures, and were expected to fund sigmficant social costs. With such limited and inconsistent changes in the external and inter- nal incentives environments, it is perhaps surprising that the swadana approach yielded any results at all. The outcomes associated with the swadana initiative will not come as a surprise to anyone who followed health sector trends and poli- cies in Indonesia in the 1990s. Until the 1997 economic crisis, pub- lic resources allocated to health increased at a satisfactory rate, loos- ening some of the budget constraints experienced in the late 1980s.16 And with funding less restrictive (until 1997), implementation of the swadana approach could afford to advance in a gradual and selective fashion, while the MOH proceeded to exercise significant constraints on the discretionary authority of swadana facility administrators. For example, crucial decisions with respect to personnel, such as the number and reimbursement of permanent staff, were not devolved to the hospital level, while facility managers were obligated to serve many poor patients who generated far fewer revenues than they cost. In addition, it seems that MOH authorities used increased flows of support, at least to centrally run hospitals, as a means of making swadana more palatable to hospital staff. Moreover, MOH leaders did not lobby aggressively for swadana principles within the ministry itself and in policy and budget discussions with the Ministries of Finance and Planning. Officials in both ministries were never con- vinced that swadana was an indispensable element within a coherent and realistic program of health financing and service reforms. Similarly, decisionmakers within the Ministry of Home Affairs and in its attached hierarchy of provincial and district governments remained unreceptive to the swadana approach, which was seen as a direct threat to the handsome revenues typically derived from public hospitals. Nor did MOH polhcymakers convince their counterparts 528 o Innovations in Health Service Delivery in Finance and Planning to eliminate the salary, drug, and other sub- sidies that allowed local governments to obtain large hospital rev- enues while putting relatively limited funds into these facilities. Further, the important complementary reform to explicitly fund services provided to poor patients never went beyond the pilot phase. This significant setback forced the hospitals to continue to deal with these responsibilities internally or implicitly. Other efforts to rationalize the channeling of funds to provincial and district governments and to give local authorities greater discre- tion in health planning have proceeded very slowly, as have experi- ments with different forms of prepaid financing of hospital and other health services. What Next for Hospital Reform in lndonesia? This chapter should not be read as an obituary for the swadana ap- proach. Indeed, Indonesian experience indicates there is nothing in- trinsically wanting in the framework of marketizing organizational reforms. Hospital managers have demonstrated their readiness and capacity to play a more active role in managing their operations, and to make decisions to lead their hospitals to a sustainable financial footing. Moreover, the goal of increasing revenues to provide additional support for public health interventions and services targeted at the poor is even more pertinent during the current crisis. But rolling out the swadana form to additional public hospitals must be accompa- nied by complementary reforms, especially with regard to funding services for the poor. Swadana status, in its current form, is certainly not sufficient to move the hospitals into effective and stable operation in the present setting-it may not even be a necessary element of a reform package. This is because constraints on government health spending promise to be even more restrictive and longer lasting than what prevailed in the late 1980s. The swadana reforms as noted above were not able to Autonomization in Indonesia * 529 contribute to a reduction in public hospital funding, which must be a core part of any envisioned reform in the current situation. Other important differences between the present context and the one a decade ago include the increased role of private hospitals and the insurance sector and the growing strength of provincial and dis- trict governments. Private hospital capacity has continued to expand and now comprises more than a third of general hospital beds, up from a quarter in the mid-1980s. And private hospitals have begun to exercise policymaking clout. At the same time, availability and affordability of high-quality hospital services cause concern among private health insurance companies. Increasingly, the MOH is being reminded of its important but largely unfulfilled regulatory functions as regards public and private hospitals. Finally, the role and place of swadana principles are sure to reemerge as provincial and district governments are handed increased responsibilities and reduced cen- tral transfers for health services. Local authorities and their con- stituents will have to decide whether retaining ownership of hospi- tals, even if managed on a swadana basis, is in the public interest. So the future of a swadana approach depends on strategies and decisions at the central level, within the MOH and the key funding ministries, and likely most of all at the provincial and district levels. At the central level, the issue is no longer how fast to replicate swadana in the remaining MOH-run hospitals. The key question is rather why the MOH needs to continue owning and managing any hospitals, whether run on a swadana basis or not. A plausible option would be for the MOH to transfer all of its hospitals to provincial and local governments, including the few general facilities it still operates and the 14 mental and other specialized hospitals it owns and manages. This transfer of facilities would proceed with some handing over of budgetary support (i.e., Operational and Mainte- nance Budget for Hospital [OPRS] funds as well as resources com- ing through other channels). This support is scheduled to decline sharply over a three-year period, to be replaced by a funding chan- nel that would support hospitals according to the poverty, remote- ness, and other salient characteristics of their service areas. Mean- 530 o Innovations in Health Service Delivery while, Indonesian government support for such hospital functions as research on key health issues and training medical students could be secured through direct, competitive grants to local government- owned facilities and also private hospitals. Finally, the MOH would need to quickly strengthen its regulatory capacities vis-a-vis local government- and privately owned hospitals, and regarding its tech- nical assistance and health advocacy roles. The agenda at the local (provincial and district) government level is even more challenging. Faced with possibly declining real health transfers from the Indonesian government (except in more impover- ished or remote areas), local authorities will have to decide whether it makes sense for them to own and operate hospitals, including the 60 general and specialty hospitals handed over by the MOH. The swadana approach has not appealed thus far to local officials who feel that autonomy would reduce flows of hospital-generated revenues. The likely decline in centrally channeled hospital subsidies will force a reassessment of the need to retain control of local hospitals and will increase the attractiveness of the swadana approach. However, direct support to private and NGO-run hospitals may emerge as a more tractable and attractive option. In this scenario, province- and district-owned hospitals, whether run on swadana principles or not, would be contracted out, fully privatized, or merged with existing privately owned hospitals. Ntloes 1. F. Saadah, S. Lieberman, and M. Juwono, Indonesian Health Sec- tor Expenditures during the Crisis: Have They Been Protected?, Watch- ing Brief 5 Jakarta: World Bank, East Asia and Pacific Region, Au- gust 1999). 2. S. S. Rukmono, "Hospital Diagnosis Study," Health Sector Fi- nancing Project, Monograph Series 5, Jakarta, 1990. 3. The plan, now postponed, would have been Repelita VII, 1999/ 2000-2004/2005. Autonomization in Indonesia * 5 31 4. A. Gani, "Improving Quality in Public Sector Hospitals in In- donesia," International_Journal of Health Planning and Management 11: 275-96 (1996). 5. T Bossert, S. Kosen, B. Harsono, A. Gani, HospitalAutonomy in Indonesia: Data for Decision-Making Project (Boston, Mass.: Harvard School of Public Health, April 1997). 6. C. N. Johnson, F. A. Duncan, Jr., D. R. Hotchkiss, S. Jacobalis, R. Noor, and C. Rice, "Assessment and Lessons Learned from the Klaten Integrated Health Care Reformn Field Trial" (report prepared for the U.S. Agency for International Development, Jakarta, In- donesia, September 1998); processed. 7. R. Malik, M. Nadjib, B. Susetyo, D. S. Argadiredja, Evaluation of Health Financing Reforms in Indonesia (akarta: World Health Or- ganization, 1998). 8. Gani, "Improving Quality in Public Sector Hospitals in In- donesia." 9. Johnson et al., "Assessment and Lessons Learned from the Klaten Integrated Health Care Reform Field Trial." 10. Bossert et al., Hospital Autonomy in Indonesia. 1 1. Ibid. 12. Johnson et al., "Assessment and Lessons Learned from the Klaten Integrated Health Care Reform Field Trial." 13. Bossert et al., Hospital Autonomy in Indonesia. 14. Malik, Evaluation of Health Financing Reforms in Indonesia. 15. Bossert et al., Hospital Autonomy in Indonesia. 16. Saadah, Lieberman, andJuwono, "Health Sector Expenditures during the Crisis." CHAPTER 15 From Plans to Actions: Hospital Reform in Ecuador Patricio V Marquez, Fernando Sacoto, and Marta Molares-Halberg Despite political and social turmoil, Ecuador has made steady progress toward reforming its health system since the mid-1990s. Nevertheless, its health indicators are still below those of other countries at a similar level of development. The infant mortality rate of 29 per 1,000 live births is twice as high as in Chile, and maternal mortality, at 150 per 100,000 live births, is three times higher than in Costa Rica. Higher exposure to risk factors and limited access to effective medical care puts an especially heavy burden on the poor. Ecuador is trying to direct its limited resources, over the short term, to the spots where they will do the most to reduce mortal- ity and morbidity. The Ministry of Public Health (MPH) has tar- geted a set of key public health interventions directed at specific maternal, childhood, and adult problems. This approach is in- tended to avoid the pitfalls of classical universality, whose prom- ise of "everything for everyone" has proven unsustainable-even in the richest countries. Over the longer haul, to help create a more equitable, efficient, and integrated system, Ecuador is trying to separate financing from delivery of health care in the public sec- tor and is also seeking changes in institutional configuration and management. 534 0 Innovations in Health Service Delivery The leaKi Reform Framework With support from the World Bank-financed FASBASE and MOD- ERSA projects as well as from other international agencies, broad policy and institutional reform proposals have been designed and adopted.1 These proposals, which enjoy a high degree of consensus among sectoral stakeholders, build upon laws enacted for reforming the Ecuadorian state as a whole. Legal Framework for Health Reform A major change in the government view of the health sector was stated in the 1998 Constitution, adopted on August 10, 1998. The state explicitly guarantees health promotion and protection through food safety, water supply, and basic sanitation. It also guarantees un- interrupted access to health services in accordance with principles of equity, universality, solidarity, quality, and efficiency. For people who cannot pay, health services will be provided in public facilities, and no one will be denied emergency services. The Constitution also stipulates that the National Health System, consisting of public, au- tonomous, private, and community-based entities, will be deconcen- trated, decentralized, and participatory. The centralized functions of the MPH cover policymaking, norm setting and regulation of health care, and coordination of public health measures. Within the institutional framework set forth in the Health Code, the Municipalities Law empowers the municipalities, within their jurisdictions, to regulate and deliver water for human consumption and sanitation and, in coordination with MPH, to reg- ulate hygiene and related health matters, including the clean opera- tion of businesses dealing with food and public buildings. Two complementary laws were enacted during the 1990s to pro- vide an enabling framework for decentralizing and implementing organizational reforms in the health system. They are the 1993 Law for Modernization of the State, Privatization, and Delivery of Public Services by the Private Sector and its Regulations; and the 1997 Decentralization and Social Participation Law. Under these From Plans to Actions: Hospital Reform in Ecuador * 535 legal instruments, the central government transferred its powers and functions under the decentralization and deconcentration regimes: * Decentralization is "a transfer of functions to a functionally or ge- ographically decentralized entity, and ... includes also the power to create new entities to discharge functions which were originally centrally assigned."2 It consists of the delegation of political, eco- nomic, administrative, and financial management powers and du- ties from the central government to subnational governments (provincial or municipal). Executive decrees are required for such decentralization. * Deconcentration consists of the delegation of administrative and fi- nancial functions from the central government to its own depen- dencies. In this framework, the ministries are required to delegate their powers and responsibilities within economic or geographic regions and through ministerial resolutions. The only specific re- quirements they must meet are that the legal framework specify the geographic scope (boundaries) where the delegate will act and that the Ministry of Finance and Public Credit give prior approval of the necessary budgetary transfers. Recent measures. To operationalize these mandates, presidential de- crees and ministerial resolutions were issued in the first quarter of 1999, allowing a transfer of authority in decisionmaking from the MPH to the provinces and from the provincial health authorities to the hospitals and health areas.3 These key legal measures will affect the success of the reform. A presidential decree in June 2001 made the decentralization of the health sector to the municipalities opera- tional. Municipalities are responsible for requesting the health ser- vices to be decentralized. Since each municipality can ask for the transfer of a different subset of health service provision to its juris- diction, the process can lead to greater entropy. The MODERSA project is working closely with the MPH and the national Modern- ization Council (CONAM) in helping municipalities fully under- 536 0 Innovations in Health Service Delivery stand the costs and benefits of the resource and responsibility trans- fers they request. Health care is the first sector to receive this dele- gation of authority. Political and Administrative Decentralization A major step in the decentralization process has been the integration of public and private providers into the Municipal Integrated Health Care Networks (MHCN), which involve substantial commu- nity participation. These networks are managed by a Consejo or Junta de Salud (Health Board) that coordinates services and infor- mation among health providers in order to guarantee the delivery of essential health care interventions. The Health Boards are responsi- ble for developing a local health plan (including organization and financial arrangements); classifying and registering health care us- ers; developing new health care financing mechanisms (e.g., a local health fund); and supporting the modernization of management structures and practices (information, quality-control mechanisms, human resources). Several instruments have been essential for public and private in- tegration of service delivery. First, the Health Boards determined the scope of the basic health package and the number of people covered, and specified its components according to the region's epidemio- logical characteristics. The package was then priced following cost manuals generated in the MODERSA project. For delivery of basic package interventions, private providers are also recruited via public tenders and brought in under contract, especially in areas where public providers are few. The contracts and procedures used (also provided by MODERSA) specify the mode of payment, incentives, and quality-control mechanisms. The Decentralized Health System in Tena Canton is a good ex- ample of decentralization with strong community participation (box 15.1). Performance agreements (compromisos de gesti6n) are the main instruments that MODERSA is supporting as incentives to public providers to improve service quality and efficiency. Empha- sis on public health targets is also expected to improve equity. From Plans to Actions Hospital Reform in Ecuador 0 537 Box 15.1 Decentralized Health System in Tena Canton The Junta de Salud was established on November 7, 1997. Its structure includes a Health Assembly with political con- trol, a Board of Directors with managerial power, and a Technical Team, responsible for delivery. Decentralization requwred legal instruments such as the ministry-level agree- ment with the Municipality of Tena, and an ordinance to delegate power to the Junta de Salud. A first plan was constructed, based on citizen participa- tion through canton-level assemblies, rural workshops, technical workshops, topic-specific agreement boards, con- sultations, and opinion polls. This junta has four components: Healthy Lifestyles and Environment (involving promotion, prevention, and com- munity participation); a Services Network (organization and enrollment of users, entitlement of services, hospital modernization, and consolidation of the fluvial system); a Support System (management, information, and human re- sources); and the creation of the Fondo Local de Salud. The Fondo de Salud classifies and registers people, de- fines the basic package and its cost, establishes a subsidy and copayment system, manages the fondo's operations, and contracts with registered providers. The subsidy and copayment system came up with an in- teresting innovation. It allows for six potential discount cases, based on socioeconomic evaluation: 0 percent, 25 percent, 50 percent, 80 percent, 90 percent, and 100 per- cent. For the Ecuadorian system as a whole, only four cases are possible: a 0 percent, 25 percent, 50 percent, or 100 per- cent discount. This shows that, if given enough flexibility, local juntas may be able to work out more precise classifi- cations and appropriate discounts, which almost guarantee both cost recovery an-d efficiency gains. (Box tontinues on the following page) 538 o Innovations in Health Service Delivery Box 15.1 (continued) The fondo's management consists of a manager, a med- ical supervisor, a financial-administrative head, and a sec- retary. The seed fund is composed of MODERSA (60 percent), local government (10 percent), copayment (20 percent), and central government and other institutions (10 percent), for a total of about US$2.2 million. Source: Ministry of Public Health, "Tena: En la Ruta de la Salud y el Bienestar-Propuesta de Creaci6n del Sistema Descentralizado de Salud del Canton Tena," Quito, 1998. On the basis of extensive discussions with the key players in the system (unta de Salud, provider network, users), draft performance agreements have been prepared in Cuenca and Tena. These are ex- pected to be signed in May 2002. These agreements specify the ser- vices that the provider networks are expected to deliver to a defined population over a period of one year. Specific targets are set for every six months, when an evaluation committee of the Junta de Salud assesses compliance. Providers meeting the agreed targets are re- warded with payment incentives. Economic Decentralization and Hospital Reform Recent decentralization and state modernization laws have opened opportunities for reforming public hospitals. Although many public hospitals need substantial improvements, the MPH, with support from MODERSA, has also set up some hospital autonomy demon- stration models. To move the public hospital sector toward more au- tonomous decisionmaking, the pilot autonomy initiative has been expanded to eight hospitals, and modernization agreements have been signed with them. The two original hospitals were the Enrique Garces Hospital in the south of Quito and the Vicente Corral Mos- From Plans to Actions. Hospital Reform in Ecuador * 539 coso hospital in Cuenca. Two additional pilot hospitals are Guaya- quil hospital and the Eugenio Espejo Hospital in Quito, a teaching hospital affiliated with the Central University Medical School. The arrangement with Eugenio Espejo adds an interesting dimension to the autonomy initiative, as the medical school has a large degree of autonomy, but its teaching hospital has not. Four additional provin- cial hospitals are being selected. The government considered six options regarding hospital auton- omy, ranging from keeping the status quo to enacting a law man- dating autonomy for public hospitals. An intermediate option was chosen, whereby a few pilot hospitals would enter into performance agreements with the MPH (at the provincial or national level). Per- formance agreements between the pilot hospitals and the provincial health authorities are the first step toward linking financing with production. These hospitals operate under a Board of Directors and a general director appointed by the board. The MPH Task Team will closely monitor the implementation of these far-reaching measures, delegating to the four pilot hospitals decision-making authority on organizational, financial, personnel, and procurement matters now concentrated in the hands of the ministers of health and of finance. Deconcentration instruments. The deconcentration of management and administrative functions from the MPH to hospital directors, leading to the hospitals' gradual autonomy, is done through ministe- rial resolutions, issued as each hospital qualifies to participate under the MODERSA project. The most important powers transferred are: contracting for goods and services, as authorized by the domes- tic procurement law; approving payment of budgeted expenditures under specific contracts; and planning and monitoring implementa- tion of investments for hospital improvements. Hospital modernization and selection. Decentralization included the modernization of public hospitals-their organization and gover- nance structure, management practices, management information systems-and improvements in health care quality over the medium term. Hospital modernization was also to include: implementing au- 540 o Innovations in Health Service Delivery tonomous managerial arrangements, including user participation; strengthening planning and decisionmaking; developing human re- sources through selective training and continuing education, per- formance incentives, and career development; implementing finan- cial administration systems and management information systems; revamping cost-recovery policies and systems and developing new resource allocation mechanisms for hospital care; and financing civil works, equipment, maintenance, and supplies for refurbishing and upgrading existing facilities. Modernizing institutions with strongly entrenched traditions and weaknesses requires experimentation and the careful selection of candidates for change. The MPH selected the initial demonstration hospitals using the hospital management assessment instrument. This evaluation assessed seven key areas: senior management capa- bilities; strategic planning; information systems; accounting and fi- nancial management; human resources and labor relations; facilities and equipment; and relationships with the community and with other providers. Other factors for selecting hospitals for the project were: having powers transferred from the MPH to hospital directors via an MPH resolution with deconcentration of authority provision; classifying hospitals as provincial or national hospitals; choosing a hospital director and staff willing to accept changes in organization and financial management and improvements in quality; using a hos- pital with strong influence in its region and credibility to serve as a model; providing a signed agreement between the hospital director and the MPH, with endorsement of respective provincial/municipal authorities, to undertake modernization changes; and ensumng that a hospital was willing to be part of a "learning network" of other hos- pitals included in similar change processes. Accounting, reporting, and cost recovery. Cost recovery in public hospi- tals is an important policy option that benefits both the health sys- tem and the public that uses it. With demonstrated improvements in quality, Ecuador's MPH could eventually recover a significant part of recurrent costs. To this end, support is provided the demonstration hospitals to develop a revised system, including legal safeguards, for From Plans to Actions: Hospital Reform in Ecuador * 541 Table 15.1 Fee Schedule for Selected Procedures (thousands of sucres) IPROCEDURE COST TOTAL REFERENCE FEE Cholecystitis (surgical) 270 110 Appendicitis (surgicall 700 280 Gastrointestinol endoscopy 260 104 Normal birth 650 260 Elective caesarean birth 1 ,000 400 Tonsil extraction 444 200 Prostate biopsy 485 194 Source Ministry of Public Health, Reglamento al Decreto Elecutivo 502, Funcionamiento de Comit6s de Participaci6n Social y Control de la Gesti6n en Servicios de Solud, Anexo 1, Torifario Referencial, Quito, 1999 charging and collecting fees and managing revenues from user charges. The MPH has developed a fee schedule that includes both total estimated costs and reference fees (tarifas referenciales) for each intervention (table 15.1).4 Because the blind use of such fees might have a regressive effect, the MPH allows hospitals to proportionately adjust the fee schedule according to regional circumstances. Moreover, a socioeconomic categorization of user procedures was designed to apply discounts to the user fees, according to family characteristics (box 15.2). Promoting allocative efficiency in hospitals. The current budgeting sys- tem in MPH hospitals does not promote allocative or technical effi- ciency.5 Budgets are based on historical spending patterns, not on the hospital's output. Changing the method of paying hospitals is one possible way of improving both their technical and allocative effi- ciency. To this end, the MODERSA project supports development of a payment methodology; implementation of "shadow prices" to ac- quaint hospitals with reimbursement based upon output; and gradual implementation of new hospital financing methodology. For exam- ple, initially the payment rate would be set equal to the total hospi- tal budget divided by the total number of discharges. All participat- ing hospitals would receive the same payment rate. The uniform payment rate would be used until it could be refined to include fac- tors such as case mix, cost of living, and perhaps hospital-specific 542 o Innovations in Health Service Delivery Blo 15.2 Sodciconomic Cl sEcagion § H= Cuire Utra in Ecuador The Ecuadorian government adopted criteria to classify users of public health facilities in May 1999. This classifi- cation required information about residence (urban/rural; transportation; owned dwelling; access to water, sewerage, electricity, and phone); education (household head, number of children attending school or university, type of school); household demographics (female-headed, number of de- pendents, disability, chronic diseases); and occupation and income (job location, household income, durable goods, fi- nancial assets). Questions on each item were weighted from 0 to 210 points. Public subsidies were applied as follows. CATEGORY POINTS PAID BY USER ( PERCENT) A < than 50 Free service B 50-100 25 percent of reference fee C 101-125 50 percent of reference fee D 126-180 100 percent of reference fee E > 180 100 percent of total cost Individuals holding a Bono Solidario are entitled to Cate- gory A. Individuals enrolled in public or private health in- surance correspond to Category E. Source: Ministry of Public Health, Reglamento al Decreto Ejecutivo 502, Funcionamiento de Comites de Participacion Social y Control de la Gesti6n en Servicios de Salud, Anexo 1, Tarifario Referencial, Quito, 1999. factors. This initial payment rate would supposedly accustom hospi- tals to working under a per-case administration payment system. It would not be used to pay hospitals, only to simulate what a hospital would receive under a per-discharge system. It would also allow time to refine the payment method. From Plans to Actions. Hospital Reform in Ecuador * 543 Improving hospital governance. Key stakeholders agree that important hospital decisions should be made closer to the population served to improve flexibility and responsiveness to the specific needs of Ecuador's diverse geographical areas and population groups. To im- prove decisionmaking in hospitals, the following measures have been implemented: * A decisionmaking framework has been developed to support greater autonomy of selected public hospitals by redefining the re- lationship between the MPH and the provincial and municipal Health Boards or Councils, autonomous health entities, and indi- vidual hospitals, with respect to key hospital governance decisions. * One or more governance models have been implemented to allow regional authorities and local communities to share greater re- sponsibility for governing and managing demonstration hospitals. * Improved management processes have been designed and im- plemented, supported by management information systems, in- cluding total quality management initiatives, that would support strengthening cost accounting, and improving performance, mea- surement, and reporting systems in the demonstration hospitals. * Hospital management educational programs have been designed and established to train individuals for positions as senior hospital and health system leaders, managers of clinical and administrative departments, and technical support staff. * The educational program infrastructure in hospital and health system management have been strengthened. Human resource management. Overall, the MPH concentrates 23.2 percent of its labor force in health services; the Ecuadorian Social Security Institute (IESS), 11.7 percent; and other private, for-profit institutions, 35 percent. Out of 16,000 physicians in Ecuador, the MPH and IESS account for about 6,600 physicians in the system compared with 6,000 in private health provider businesses.6 The main human resource-management problems are the lack of labor 544 - Innovations m Health Service Delivery incentives to improve the quality of health service; the inability of hospital directors to choose an appropriate labor mix between ad- ministrative and medical personnel; and the lack of a career profile in the public sector.7 These problems are determining factors in the high turnover of qualified personnel, the percentage share of public and private jobs, and other unresolved labor arrangement problems. With the support of the MODERSA project, decentralization is encouraging output-based incentives-which still require legal in- struments-that delegate power to hospitals for hiring, managing layoffs, making decisions on training, and developing information systems to organize these decisions. Since the project involves other public and private institutions, contracts with providers specify ap- propriate incentives and accountabilities. The process is intended to move the incentive system from one based on fixed salaries and budgets to a system where providers and users share risks, and sal- aries reflect those risks (partial- and full-capitation). The experience of Colinas del Norte suggests some avenues for strengthening reform in human resources management (box 15.3). Conclusions The success of health care reforms in Ecuador, as in other countries around the world, depends on consensus on the major directions and principles of reform. The government that took office in Ecuador in August 1998, building on the previous government's efforts, began to attack the most critical bottlenecks in the system, but delays in im- plementing the reforms occurred with the change of administration in January 2000. Sets of interventions have been adopted to establish new organizational models for public health care delivery while cul- tivating support for a broader reform agenda among the different sectoral actors, particularly for health insurance reform. Experience in other countries shows that this strategy offers the greatest proba- bility of success in the reform endeavor, particularly in country situ- ations characterized by high political instability and a lack of policy and institutional continuity. From Plans to Actions Hospital Reformi in Ecuador a 545 Box 15.3 A Public HMO: Community System for Integral Healih in Colinas del Norte The Community System for Integral Health (Sistema Co- munitario de Salud Integral, SICSI) in Colinas del Norte builds on the Municipal Health Council model and seeks to improve the incentive environment in which health work- ers and service recipients interact. It targets one neighbor- hood with about 2,000 families north of Quito. The SICSI operates like a small, public health maintenance organiza- tion (HMO) in which families pay small monthly fees to become members of a clinic. Physicians are paid their us- ual monthly base salary plus bonuses for each new mem- ber farnily and for each individual attended and service rendered. A physician in a project clinic receives a basic monthly salary independent of productivity. For each new paying af- filiated member family, the physician receives a monthly bonus of 0.1 percent. This encourages physicians to help sell the program in the communities they serve and, quite likely, to treat members better than nonmembers. In addi- tion, physicians are paid for each service rendered, within certain limits. They receive s/1,250 per consultation up to a limit of s/500,000 (400 consultations) a month. Finally, no more than 40 percent of the additional pay for consulta- tions may relate to morbidity, leaving 60 percent for "pub- lic health" consultations. For US$2.50 a month, affiliated families receive a com- prehensive array of health services, including preventive care, pediatric care, and pre- and perinatal care. The local clinic is the main point of contact but has referral rights to the local hospital. By contrast, unaffiliated families are served in the traditional manner-in a system with copay- ments for most services. In the first four months of the (Box conti7ues on thefollowing page) 546 o Innovations in Health Service Delivery Box 15.3 (continued) project, 700 families became affiliated with the clinic, and about 400 of them pay the monthly fee regularly. Families eligible for the bono can affiliate without paying the fee (just as the unaffiliated do not have to make most copay- ments). In addition to encouraging HMO-style efficiency incen- tives, another promising development is the cooperation on related service provision (such as pre- and perinatal care). The MODERSA project has helped forge this cooperation between the local clinics in the Area de Salud and the hos- pital geographically located within it. Source: A. Gershberg, "Education and Health Decentrali- zation in Ecuador: Practice, Plans and Prospects" (back- ground paper prepared for the World Bank, Washington, D.C., October 1999). Constitutional change has opened a real window of opportunity for pushing the health care reform agenda in Ecuador. These changes strengthen the policymaking and regulatory role of the MPH within the health system while reaffirming public and private participation in the delivery of health services within a decentralized context. However, major medium- and long-term institutional build- ing efforts will clearly be required to make the proposed policy and institutional reforms fully operational in the Ecuadorian health sys- tem. Evaluations to measure the impact of reform will have to be part and parcel of these efforts. Notes 1. Other international agencies supporting health care reform in Ecuador include the U.S. Agency for International Development, From Plans to Actions: Hospital Reform in Ecuador * 547 the Pan American Health Organization, the World Health Organi- zation, and the United Nations Development Programme. 2. Government of Ecuador, Regulations 30 and 31 of the Law of the Modernization of the State, Privatization, and Delivery of Pub- lic Services by the Private Sector, Law no. 50 published in the Offi- cial Gazette, no. 349, December 31, 1993. Executive Decree no. 2328, published in the Official Gazette, no. 581 (Supplement), De- cember 2, 1994. 3. Presidential Decree 502,January 22, 1999. Ministerial Resolu- tion from the Ministry of Finance and Public Credit 018; and AIPH Ministerial Resolution 1292. 4. Reference fees include only operational costs, excluding med- ical personnel expenses. 5. R. Lucio and E. Lasprilla, "Costos y eficiencia de los Servicios de Salud en el Ecuador a 1999" (background paper prepared for the World Bank, July 1999). 6. Organizaci6n Panamericana de la Salud/Organuzaci6n Mun- dial de la Salud, Ecuador: Perfil del Sistema de Servicios de Salud, Pro- grama de Organizaci6n y Gesti6n de Sistemas y Servicios de Salud (Quito: OPS/OMS, 1999). 7. Lucio and Lasprilla, "Costos y eficiencia de los Servicios de Salud en el Ecuador a 1999." Glossary of Terms Accountability Responsibility for achieving objectives and reporting on activities. Accountability, funding Responsibility for achieving objectives that is generated or payment by the structure of compensation for services. Accountability, Responsibility for achievmg objectives that is created ownership through the governance relations between the government- owner and the organization (i.e., corporate plan, perfor- mance contract). Agency theory Highlights the mechanisms used to reconcile divergent interests among mdividuals under conditions of wide- spread uncertainty and uneven access to information. An agency relationship is one in which one person (the agent) acts on behalf of another (the principal). For example, an employee is an agent on the employer's behalf; a doctor is an agent on the patient's behalf. Allocative efficiency A situation In which resources cannot be reallocated to achieve more of one objective without accepting less of another. Autonomization A reform of a public organization that shifts day-to- day decisions from supervisory agents in the hierarchy to the management of the organization ("letting managers manage"). Bilateral monopoly Bilateral monopoly is the combination of a monopoly market (single seller) on the selling side and a monopsony market (single buyer) on the buying side. A market domn- nated by a profit-maximizing monopoly tends to charge a higher pnce. A market dominated by a profit-maximizing 549 550 o Innovatons in Health Service Delivery monopsony tends to pay a lower price. When combined into a bilateral monopoly, the buyer and seller are forced to negotiate a price. The resulting price could end up any- where between the higher monopoly's pnce and the lower monopsony's pnce. Where the price ends ups depends on the relative negotiaung power of each side. Board Group of individuals, formally consttuted in a super- (or "company board") visory role over an orgamzation. Board members repre- sent individuals or entities outside an orgarnzation that have interests in the organization's performance-either as dispersed owners (shareholders) or other stakeholders. The board functions to concentrate the more intense su- pervisory responsibilities among a small number of repre- sentatves to reduce the cost of overseeing the firm, and to enable some specialization by board members. In pubhc organizations, the boards may also be used as a means to "regularize" supervision of the organization, enhance trans- parency, and depoliticize decisionmaking. Business case A tool that supports planning and decisionmaking, includ- ing decisions about whether to buy, which vendor to choose, and when to implement. Business cases are gen- erally designed to answer the question: What will be the financial consequences if we choose X or do Y) A good business case shows expected cash flow consequences of decisions, over ume, and it includes the rationale for quantifying benefits and costs. Critical success factors and significant nsks will be discussed, if relevant. The case also describes the overall impact of a proposed initiative in terms of net cash flow, discounted cash flow, payback pe- riod, and internal rate of return. Capex Funds used by a company to acquire or upgrade physical assets such as property, plant, or equipment become fixed assets on your balance sheet. This can include everything from repainng a roof to building a fire escape. Capitation Payment system whereby managed-care plans pay health care providers a fixed amount per insured person to pro- vide care to a group of patients for a defined period. Under this system, providers are not paid for services that exceed the allotted amount of costs. In the United King- dom, for example, covered persons register with a general practitioner, selected from a list of registered and partici- pating GPs, who receive a fixed monthly fee from the Na- tional Health Service for each person registered with him or her. See also provider paymnent arrangements. Glossary of Terms * 551 Contestability The existence of competition "for the market," whereby potential suppliers are compelled to bid competitively for the right to be the sole supplier of a service for a defined length of time. While holding this nght, the supplier will be a monopoly provider, yet it faces competition for that nght at regular intervals-hence the market is "contest- able" but not competitive. Contracting reforms Reforms that shift public resource allocation in health from automatic funding transfers to organizations, based on historical resource use for contracts that create com- mitments to deliver specific services and other outputs. Control group Comparison of behavior and performance among reform- approach ing hospitals with behavior among hospitals presumed to be unaffected by the reform, allowing attribution of ob- served differences to the reform. Corporatization A reform of a public organuzation that reproduces some elements of private sector business structure in an attempt to engender the efficiency of private corporations while assuring contmued emphasis on social objectives through public ownership. Counterfactual What would have happened without the reform. Compar- ing the counterfactual with the actual will provide insights about a reform's impact. Cream skimming Selection of healthy patients who need less treatment than other patients and therefore cost less. Decentralization Form of organizational reform that shufts decision-making control, and often revenue rights and responsibilities, from central to lower level government agencies. Decision rights The managerial prerogative to make decisions about aspects of an organization's actuvites-including pro- curement, use of inputs, scope of activities, financial man- agement, clinical management (hospitals), nonclinical administration (hospitals), strategic management, market strategy, sales, and the production process. Diagnostic-related A patient classification scheme that provides a clinically group (DRG) meaningful way of relating the number and types of pa- tients treated in a hospital to the resources required by the hospital. DRGs constitute a payment system that reim- burses health care providers a fixed amount for all care in connection with a standard diagnostic category. The sys- tem is a form of case-rate payment. 552 0 Innovations in Health Service Delivery Dial settngs Shorthand for the aggregate status of the five key compo- nents of the incentive regime for a reformed hospital: magmtude of a hospital's decision rights; the extent of a hospital's residual claims to net revenue; balance between indirect and direct (hierarchlcal) accountability mecha- nisms; magnitude of a hospital's market exposure; and de- gree of explicitness of arrangements for social functions. Economics Academic discipline that applies a variety of economic of organization perspectives-especially principal agent theory, transac- tion cost economics, and property nghts and public choice theory-to examine the effectiveness of different owner- ship and governance arrangements for structuring activi- ties. The analytical focus is on the factors influencing the choice and design of organizational arrangements govern- ing commercial transactions. External financing The amount of finance that an autonomous public entity hmit may raise durng the financial year through government grant or borrowmg. Externality The impact on another entity or individual that a decision- maker does not take into account when making a decision. Governance The sets of rules or authonty structures, either sponta- neously evolved or externally imnposed, that determine the manner in which organizations are managed and the na- ture of accountability of the managers to the owners. Governance structure A set of rules and institutions for administenng an eco- nomic exchange relationship. The free market/common law is an example, as is common ownership with hierarchy. Influence activities Self-interested activities designed to shape others' deci- sions. Within organizations (public as well as private), these are often aimed at redistributing rents and take the form of political activity or misrepresentation and distor- tion of information. Influence costs The costs incurred m attempts to influence others' deci- sions in a self-mterested fashion, in attempts to counter such influence activities by others, and m the degradation of the quality of decisions because of influence. Institutions Rules (formal and informal) and customs of a game or acnvity-the humanly devised and socially shared con- stramts that shape human interaction and the mechamnsms by which these rules are enforced. Examples include mar- nage, licensing, burial customs, the contract, wage labor, Glossary of Terms * 553 a currency, the handshake, weights and measures, and the budgeting process. Examples from the health sector in- clude formal and informal norms regarding payment of hospitals and physicians, formal and informal norms re- garding staffing, care givmg, care seeking, prescription and use of medication. Encompasses both rules and or- ganizations that shape and enforce these rules. Institutional capacity The degree to which the institutional structure supports desired behaviors of relevant individuals and organiza- tions. In the context of a reform program, institutional ca- pacity is the degree to whuch the existing institutional str-ucture encourages the actions that further the reform. Institutional Institutionalized mechanisms (such as competitive pay, mechanisms prestige, contracting arrangements, or training proce- dures) influence the overall level of capability of an organ- ization, for example, by influencing the supply of qualified labor and relevant expertise. Institutional Changes made in the institutional framework to make it development more conducive to economic development; alternatvely, m the context of a reform program, changes made in the in- stituuonal framework to proimote the desired policy actons and related behaviors of individuals and organizations. Institutional The "rules of the game" that determine the rewards as- framework sociated with different kinds of knowledge, skills, and behaviors. Integration, backward When a firm establishes formal linkages or merges with another firm that is m the previous stage of the production- distribution process to increase efficiency and/or market power. Backward mtegration in health care often takes the fonm of an acute care facility linking to doctors or groups of doctors, to increase their referrals, and to dilute pur- chasers' influence. Integration, forward When a firm establishes formal linkages or merges with another firm that is in the next stage of the production- distribution process to increase efficiency and/or market power. In health care, an example is when a provider inte- grates with the enuties in the markets to which they refer patients, in order to capture part of the revenue from these referrals (e.g., diagnostic labs, pharmacies). Integration, vertical Integration with a buyer or supplier-as compared with horizontal integration, which is integration with a firm at the same stage of production. 554 o Innovations in Health Service Delivery Interventions The identified actions taken (e.g., clinical interventions (m evaluation) public health interventons, and policy mterventon). Managerial capacity The way that management skills and orgamzation combine with accountability arrangements to determine the capa- bility of managerial staff to conduct key managerial tasks (e.g., directing, adminustering, and making decisions). Market exposure Extent to which an organization relies on earning its rev- enue In a market, where purchasers have choice, instead of relying on automatic financial transfers. Marketizing reform A set of planned, structural changes to public service or- ganizations and their operating environment, that seek to improve incentives through creating or enhancing market pressures, especially those associated with the exercise of choice by the service user or payer. Moral hazard The form of postcontractual opporturmsm that anses when actions required or desired under the contract are not freely observable. Examples from health care include overuse of health services in response to third-party pay- ment. In insurance, it refers to the insured's tendency to relax attention to preventing events or losses covered by insurance. Natural monopoly A monopoly based on an incumbent firm's overwhelming cost advantage stemming from access to some unique nat- ural resources or past capital installations that a competi- tor would have to duplicate (e.g., a power grid). Neoclassical The mainstream approach to economics, which has in- economics volved primarily a single set of idealized rules goverming market exchange. It has provided valuable insight mto the fundamental nature of exchange and resource allocation in decentralized markets. However, the model has over- looked two areas of inquiry cntical to health sector policy: how do alternative sets of rules and organizations affect performance, and why does the form of economic organ- ization differ among activities? New public Application of private sector management principles to management the public sector. Thls approach often includes elements such as: shifting away from process accountability to ac- countability for results; devolution of management con- trol combined with enhanced reportng and monitoring mechanisms; disaggregation of large bureaucratic struc- Glossary of Terms * 555 tures into quasi-autonomous agencies; emphasis on mon- etary rather than nonmonetary incentives, and a stress on cost cutting and efficiency. Nondistribution A critical component of the governance framework for constraint nonprofit orgarnzations that supports the nonprofit es- sence and motivation of the organizations and their em- ployees through blocking distribution of residuals to any individuals. Organuzations The way individuals are structured to engage in an activ- ity. Organuzations are groups of individuals (or subordi- nate organizations) bound by some common purpose to achieve objectives. They have well-defined boundaries be- tween what is inside and what is outside the organization. Examples include political parties, firms, churches, schools, and hospitals. Organizational The set of formal and informal administrative rules and behavior procedures for selecting, deploying, and supervising re- sources and undertaking activities in the most efficient way to achieve institutional objectives Organizational reform Changes to the constitution or structure of an organiza- tion, and hence the interaction of mdividuals inside the organization Such changes are also likely to change the way the organization interfaces with other individuals and organzations-hence, organizational reform can change the institutional framework-by altering formal and in- formal norms of behavior between the organization and others. Examples of organizational reform include changing the organizational form of an entity (i.e., changing a firm from a proprietary or family-run business to a limuted lia- bihty company; converting a state-owned company to pri- vate ownership; decentralization within public sector, and delegation of authority and accountability). Organizational reform Efforts to enhance the performance of public organi- modalities, marketizing zations by reducing direct government control and in- creasing exposure to market or marketlike mcentives (au- tonomization, corporatization, or privatization). Oversight, government External government regulatory structure and activities to monitor and oversee subordinate or regulated entities. Performance The process of funding in exchange for delivering certain budgeting services or products-where the amount of funding is 556 o Innovations in Health Service Delivery tied to explicit performance results and quality indicators (e.g., utilization, average length of stay, staffing ratios, in- fection rates). Privatization Transfer of a public hospital to nongovernmental own- ership, either as a for-profit or nonprofit organization. The most extreme version of marketizing organizational reforms. Property rights theory This discipline examines property rights as an economic concept. It examines the range of orgatization and con- tract types as resulting from the attempts of individuals to maximize the value of their property rights. Provider payment The organization and structure of the system for fund- arrangements ing or paying providers; may also include formula for price setting. Pubhc choice theory Academic discipline focused on the behavior of public sec- tor bureaucrats. While bureaucrats are supposed to work In the public interest, putting into practice the policies of government as efficiendy and effectively as possible, pub- lic choice theorists see bureaucrats as self-interested util- ity maximizers, motivated by such factors as salary, pre- requisites, public reputation, power, patronage, and the ease of performing duties. This discipline focuses on the importance of the political process in shaping and imple- menting policy. Public corporation Public corporations are government-chartered but gov- erned by private rather than public law. They occupy a muddle ground between purely prnvate corporations and government agencies. They often operate with some sort of government backing or mandate, and enjoy explicit or implicit government financial guarantees. They may be exempt from national or local taxes. Such organizations are often established to run a nationalized industry or state-owned enterprise. A government mimster usually appoints the chairman and board members. Reflexive comparison Making the implicit assumption that a hospital's behavior and performance prior to an intervention would have con- tinued unchanged, so that any differences are due to the reform. Reform, funding Addresses system- or provider-performance problems by and payment altering the structure of funding or payments to providers, usually by tightening the link between resource allocation and user or payer selections. Glossary of Terms * 557 Reform, management Efforts to strengthen managenal performance within the preexisting structural parameters through changes in re- cruitment, training opportunities, improvements in infor- mation systems to facilitate effective decisionmaking, and new management practices. Reform, organizational Efforts to improve performance by altering the structure of hospitals and relations with other parts of the health care system instead of by working within the system's parameters. Rents A return received in an activity that is in excess of the mm- imum needed to attract the resources to that activity. Reputational risk The potential that negative publicity regarding an institu- tion's practices or activities, whether true or not, will cause a decline in the customer base, costly litigation, or revenue reductions. Residual revenue rights Entitlement to whatever revenue remains after all funds have been collected and all debts, expenses, and other con- tractual obligations have been paid. Residual nghts Entitlement to make any decisions regarding an asset's use of control not explicitly contracted by law or assigned to another by contract Sector neutrality Environment in which government policies are not dis- ("level playing field") criminatory and all companies in a given market must follow the same rules and are given an equal abibty to compete. Separation of provider Health sector reforms that seek to create a degree of com- payment reforms petition among public providers through establishmg dis- tinct government organizations to undertake the funding/ purchasing function and the provision function. Service-profit chain Conceptual approach to labor management in the service sector, under which employee productivity is perceived as deriving most directly from employee loyalty. Social functions Services or products priced and delivered below costs- either for distributional purposes or because of externai- ties in consumption. Soft budget constraint A situaton under which public or pnvate entities are (or "soft budgets") operating, where the government is explicitly or implicitly hable for losses or budget overruns. Statistical approach Relies on the assumption that measures of the most in- fluential refornm charactenstics and of the exogenous 558 o Innovauons in Health Service Delivery environment are mcluded in the available data or can be gathered. Strategic management The proactive approach to management under which an organization determines its long-run direction and per- formance by formulating institutional objectives, under- taking ongoing internal appraisal, and analyzing manage- ment and functions. Strategic management also mvolves careful appraisal of the external environment, hence it in- cludes evaluation of changing market conditions. Technological capacity The ability of an agency, or group of agencies, to gather, assess, and share information. Related to its organizational structure, collective knowledge, and skill of staff, as well as adequacy of related technical resources, including infor- mation technology (hardware and software). Total factor The relationship between output and the aggregate of all productivity inputs. In health care, usually measured as the number of patients served, or services rendered, per index unit of inputs. Transaction cost Investigation of alternative governance structures, partic- economics ularly the differences between the natures of markets (in- terorgamzational relationships) and firms (intraorganu- zational relationships), to determine which governance structure, among a hst of those available, is most efficient (production and transaction cost minimizing) for govern- ing a given transaction or set of activities. Bibliographies All notes appear with chapters. Chapter 1 Bennett, S. 1992 "Promoting the Private Sector: A Review of Developing Country Trends." Health Polity and Planning 7 (2): 97-110. Berryman, S. E., N. Boyle, F. Golladay, M. Holmes, P. Keefer, and K. Sigrist. 1997. "Assessing Institutional Capabilities." Processed. Boubakri, N., andJ-C. Cosset. 1998. "The Financial and Operating Performance of Newly Privatized Firms Evidence from Developing Countries." 7ournal of Finance 53: 1081-1110 June). http://www.cob.ohlo-state.edu/-fin/journal/ archive-papers/issjun98/ms5897.pdf. Casagrande, J 1999. "Social Sector Pnvatization in Mongolia: Concepts Learned through Expenence." World Bank, Private Sector Development Department, Washington, D.C. Processed. Donaldson, D. J. 1995. "Pnvatization: Principles and Practice." Lessons of Experience Series 1. Washington, D.C.: International Finance Corporation. Duncan, I., and A. Bollard. 1992. Corporatization and Pnvatization Lessons from New Zealand. Auckland, N.Z. Oxford University Press. Fama, E. F. 1980. "Agency Problems and the Theory of the Firm." Journal of Polit- ical Economy 88 (2): 288-307. Galal, A., L. Jones, P. Tandoon, and I. Vogelsang. 1995. Welfare Consequences of Sell- ing Public Enterprises: An EmpincalAnalysis. New York: Oxford University Press. Ginshankar, N. 1999. "Reforming Institutions for Service Delivery: A Framework for Development Assistance with an Application to the HTNP Portfolio." Pol- icy Research Working Paper 2039. World Bank, Policy Research Department, Washington, D.C. 559 560 o Innovations in Health Service Delivery Gregory, R. 1995. "Accountability, Responsibility and Corruption: Managing the Pubhc Production Process." In J. Boston, ed., The State under Contract. Well- mgton, Mass: Bndget Williams Books. Gwatkin, D. 1997. "Poverty and Equity and Health in the Developmg World: An Overview." World Bank, Washington, D.C. Processed. Harding, A., and A. S. Preker, eds. 2003. Pnvate Participatzon in Health Services. Washington, D.C.: World Bank. Jensen, M. C., and W. H Mecklng. 1976. "Theory of the Firm: Managerial Be- havior, Agency Costs and Ownership Structure." Journal of Financial Economics 3. 305-60. Lane, J-E., ed. 1997. Public Sector Reform: Rationale, Trends, Problens. London: Sage Publications. Lewis, M. A., G. M. La Forgia, and M. B. Sulvetta. 1996. "Measuring Public Hos- pital Costs Empirical Evidence from the Dominican Republic." Social Science Medicine 43 (2): 221-34. Manmng, N. 1998. "Unbundling the State: Autonomous Agencies and Service De- hvery." Washmgton, D.C.: World Bank, Public Sector Group. Processed. Manning, N., and D. Parker. 1994. "Nationalization, Privauzation and Agency Sta- tus within Government: Testmg for the Importance of Ownershlp." In P Jackson and C. Price, eds., Privatisation and Regulation. London. Longman Group. Mathur, S., and G. Torres. 1996. The Third Wave of Pnvatization: Pnvatization of So- cial Sectors in Developing Countries. World Bank, Private Sector Development Department, Washington, D.C. Mllgrom, P., and J. Roberts. 1990. "Bargaining Costs, Influence Costs, and the Or- ganmzation of Economic Activity." In J. Alt and K. Shepsle, eds., Perspectives on Positive Political Economy. New York: Cambridge Umversity Press. - . 1992. Economics of Organization and Management, chap. 9. Englewood Cliffs, NJ.: Prentice-Hall. MNllstein, I. 1998. "The Corporation as an Ownershup and Management Structure." Paper written for the World Bank Private Sector Development Forum, Balti- more, Md. Olson, M. 1982. The Rise and Decline of Nations: Economnic Growth, Stagflation and So- cial Rigidities. New Haven, Conn.: Yale University Press. OECD (Organisation for Economic Co-operation and Development). 1992. Regu- latoiy Reform, Prvatisation and Competition Policy. Paris: OECD. Bibliographies * 561 Osborne, D, and T Gaebler. 1993. Reinventing Government. New York Plume. Pauly, M. 1996. "Health Systems Ownership: Can Regulation Preserve Community Benefits?" Frontiers of Health Services Management 12 (3). 3-34 (spring) Preker, A. S., A. Harding, and P. Travis. May 2000. "Make or Buy Decisions in the Production of Health Care Goods and Services: New Insights from Institu- tional Economics and Organizational Theory." Bulletin of the World Health Or- ganization 78 (6): 779-90. Robinson, J. 1997. "Physician-Hospital Integration and Economic Theory of the Firm." Medical Care Research and Review 54 (1) 3-24; discussion 25-31 (March) Saltmian, R., and J. Figueras. 1997. European Health Care Reforni. Analysis of Current Strategies Copenhagen: World Health Organization. Sappmgton, D. E. 1991. "Incentives in Principal Agent Relationships." Journal of Economic Perspectives 5 (2). 45-66. Schick, A. 1998. "No Leapfrogging." Presentation made at the World Bank. "Why Most Developing Countries Should Not Try New Zealand's Reforms." World Bank Research Observer 13: 123-31 (February). Schieber, G., ed. 1997. Innovations in Health Care Financing Proceedings of a World Bank Conference, March 10-1 1, 1997. World Bank Discussion Paper 365. Wash- ington, D.C. World Bank. Williamson, 0. 1991. "Comparative Economic Organization The Analysis of Dis- crete Structural Alternatives." Adminstrative Saence Quarterly 36: 269-96 (une). - 1989. "Transaction Cost Economucs." In R. Schunalensee and R. Willig, eds., Handbook of Industrial Organization New York: North-Holland. 1 989. Bureaucracy- What Government Agencies Do and Why They Do It. New York: Basic Books. . 1985. The Economic Institutions of Capitalism: Firnis, Markets and Relational Contracting New York: Free Press. Wilson, J. Q. 1991. Bureaucracy What Government Agencies Do and Why They Do It. New York: HarperCollins. World Bank. 1997. World Development Report 1997: The State in a Changing World New York- Oxford University Press. . 1995. Bureaucrats in Business The Economics and Politics of Government Owner- ship. World Bank Policy Research Report. New York: Oxford University Press. World Health Organization. 2000. World Health Report 2000. Peiforn7ance of Health Systems Geneva: WHO. 562 o Innovations in Health Service Delivery Chlapler 2 Alford, R. 1975. Healtb Care Politics. Chicago: University of Chlcago Press. Arnell, A. 1996. "The Monopolistic Integrated Model and Health Care Reform: The Swedish Experience." Health Policy 37: 19-33. Barr, N., ed. 1994. Labor Markets and Social Poliey in Central and Eastern Europe. Ox- ford. World Bank/Oxford University Press. Bennett, S., S. Russell, and A. Mills. 1999. "The Changing Role of Government in Adjusting Economies." Paper 3: Health Sector, summary paper presented at a workshop at the University of Birmigham, March 28-30. Ferhe, E., L. Ashbumer, L. Fitzgerald, and A. Pettigrew. The New Public Manage- ment in Action. Oxford: Oxford University Press, 1996. Ham, C. 1997. Health Care Reform: Learningfrom International Experience. Bucking- ham, U.K.: Open University Press. - 1994. "Where Now for the NHS Reforms?" BMI] 309: 351-52. Ham, C., and M. Brommels. 1994. "Health Care Reform in the Netherlands, Swe- den and the United Kingdom." Health Affairs 14: 106-19. Heclo, H. 1974. Modern Social Politics in Britain and Sweden. New Haven, Conn.: Yale University Press. Hogwood, B., and L. Gunn. 1984. Politcy Analysis for the Real World Oxford: Oxford University Press. Hood, C. 1991. "A Public Management for All Seasons." Public Administration 6: 3-19. Immergut, E. Health Politics- Interests and Institutions in Western Europe. Cambridge, U.K: Cambndge University Press, 1992. Kingdon, J. W 1995. Agenda, Alternatives and Public Policies. 2d ed. New York: HarperCollins. Klein, R. 1995. The New Politics of the NHS. 3d ed. London: Longman. - 1995. "Big Bang Health Care Reform-Does It Work? The Case of Britain's 1991 National Health Service Reforms." Milbank Quarterly 73: 299-337. Lipsky, M. 1980. Street-Level Bureaucraiy. New York: Russell Sage. Machiavelli, N. 1950. The Prince and the Discourses. New York: Random House. March, J., and J. Olsen. 1989. Rediscovering Institutions. The Organizational Basis of Poltiacs. New York: Free Press. Bibliographies * 563 Marmor, T R. 1994. Understanding Health Care Reform. New Haven, Conn.: Yale Umversity Press. Mintzberg, H. 1987 "Crafting Strategy." Harvard Business Review 65 (4). 66-75. Moore, M. 1995. Creating Public Value. Cambridge, Mass.: Harvard University Press. OECD (Organisation for Economic Co-operation and Development). 1994. The Reform of Health Care Systems: A Review of Seventeen OECD Countnes Pans: OECD. . 1992. The Reform of Health Care A Comparative Analysis of Seven OECD Countries. Paris: OECD. Preker, A. S., and R. G. A. Feachem. 1996. Market Mechanisms and the Health Sector in Central and Eastern Europe. World Bank Technical Paper 293. Washington, D.C. Pierson, P. 1994. Dismantling the Welfare State? Cambridge, U.K.: Cambridge Uni- versity Press. Pressman, J., and A. Wildavsky. 1973. Implementation. Berkeley: University of Cah- fornia Press. Reich, M. 1996. "Applied Political Analysis for Health Policy Reform." Current Is- sues in Public Health 2: 186-91. . 1994. "The Politcal Economy of Health Transitions in the Third World." In L. C. Chen and N. Ware, eds., Health and Social Change in International Per- spective Boston, Mass.. Harvard School of Public Health. Saltman, R., and J. Figueras. 1997. European Health Care Reform: Analysts of Current Issues. Copenhagen: World Health Organization. Saltinan, R., J. Figueras, and C. Sakellarides, eds. 1998. Cntical Challengesfor Health Care Reform in Europe. Buckingham, U.K.. Open University Press. Tuohy, C. H. 1999. "Dynamics of a Changing Health Sphere: The United States, Britain and Canada." Health Affairs 18 (3): 114-34. van der Gnnten, T. E. D. 1996. Scope for Policy Essence, Operation and Reforni of the Policy System of Dutch Health Care. Rotterdam: Erasmus Umversity. Walt, G. 1998. "Implementing Health Care Reform: A Framework for Discussion." In R. Saltman,J. Figueras, and C. Satellarides, eds., Cntical Challengesfor Health Care Reform in Europe. Buckingham, U.K.: Open University Press. Walt, G., and L. Gilson. 1994. "Reforming the Health Sector in Developing Coun- tries: The Central Role of Policy Analysis." Health Policy 9: 353-70. Wildavsky, A. 1979. Speaking Truth to Power: The Art and Craft of Policy Analysis. Boston, Mass. Little, Brown. 564 o Innovations in Health Service Delivery Chalper 3 Abedian, I., B. Strachan, and T. Ajam. 1998. Transformation in Action: Budgeting for Health Service Delivery. Cape Town: University of Cape Town Press. Aday, L. A. 1975. "Economic and Noneconomic Barriers to the Use of Needed Medical Services," Medical Care 8 (6): 447-56. Aday, L. A., and R. Andersen. 1981. "Equity of Access to Medical Care: A Concep- tual and Empincal Overview." Medical Care 14 (12), supplement- 4-27. Alderman, H., and V Levy. 1996. "Household Responses to Public Health Service: Cost and Quality Tradeoffs."World Bank Research Observer (International) 11: 3-22 (February). Andersen, R. 1978. "Health Status Indices and Access to Medical Care." American Journal of Public Health 68 (5) 458-63. Barnum, H., and J. Kutzin. 1993. Public Hospitals in Developing Countries: Resource Use, Cost, Financing. Baltimore, Md.: The Johns Hopkins University Press. Bennett, S., S. Russell, and A. Mills. 1996. Institutional and Economic Perspectives on Government Capacity to Assume New Roles in the Health Sector: A Review of Evpe- nence. PHP Departmental Publication 22, Health, Economics, and Financing Programme, Health Policy Unit, London School of Hygiene and Tropical Medicine, London. Bossert, T 1998. "Analyzing the Decentralization of Health Systems in Developing Countries: Decision Space, Innovation and Performance." Social Science and Medicine 47 (10). 1513-27. Bossert, T, S. Kosen, B. Harsono, and A. Gani. 1997. Hospital Autonomy in Indone- sia. Data for Decision Making Project, Department of Population and Interna- tonal Health. Boston, Mass.: Harvard School of Public Health. Bradford, C. K., and J. R Tiscormia. 1987. Monitoring the Hospital's Financial Health: A Guide for Trustees. Chicago: Amencan Hospital Association. Chawla, M., R. Govmdaraj, P. Berman, and J. Needleman. 1996. Improving Hospital Performance through Policies to Increase Hospital Autonozmy: Methodological Guide- lines. Data for Decision Making Project, Department of Populaton and Inter- national Health. Boston, Mass.: Harvard School of Public Health. Collins, D., J. Quick, S. Musau, and D. Kraushaar. 1996. Health Financing Reform in Kenya: The Fall and Rise of Cost Sharing, 1989-1994, Stubbs Monograph Series 1. Boston: Management Sciences for Health. Cullis, J., and P. West. 1979. The Economics of Health: An Introduction. Oxford, U.K.: Martin Robertson. Bibliographies * 565 De Geyndt, W 1995. Managing the Quality of Health Care in Developing Countries. World Bank Technical Paper 258, Washington, D.C. Duncan, W J., P. M. Ginter, and L. E. Swayne 1998. Handbook of Health Care Man- agement Malden, Mass.: Blackwell. Dunlop, D., and J. Martins, eds. 1995. An International Assessment of Health Care Fi- nancing: Lessons for Developing Countries. EDI Semmnar Series, World Bank, Washington, D.C. Evans, R. 1984. Strained Mercy: The Econiomics of Canadian Health Care Toronto: Butterworths. Filmer, D., and L Pritchett. 1997. Child Mortality and Public Spending on Health: How Much Does Money Matter? Policy Research Paper 1864. World Bank, Policy Re- search Department, Washington, D C. Filmer, D., J. Hammer, and L. Pritchett. 1998. Health Policy in Poor Countries Weak Links in the Chain. Policy Research Working Paper 1874. World Bank, Policy Research Department, Development Research Group, Poverty and Human Resources, Washington, D.C. Gerder, P., E. Rose, and P. Glewwe. 2000. "Health Module." In M. Grosh and P. Glewwe, eds., Designing Household Survey Questionnaires for Developing Coun- tries Lessons fi-om Ten Years of Living Standard Measurement Erperience New York Oxford University Press. Girishankar, N. 1999. "Reforming Institutions for Service Delivery." Policy Re- search Working Paper 2039 World Bank, Policy Research Department, Wash- mgton, D.C. Gordon, J. R. 1987. A DiagnosticApproach to Organizational Behavior Newton, Mass Allyn and Bacon. Govindaraj, R., and M. Chawla, 1996. Recent Experiences with Hospital Autonomy in Developing Countries-What Can We Learn? Data for Decision Making Project, Department of Population and International Health. Boston, Mass.: Harvard School of Public Health. Harris,J. 1977. "The Internal Orgarnzation of Hospitals: Some Economic Implica- tions." Bellyournal of Economics 8: 467-82. Hee, Song Dae. 1998. "New Public Enterprise Policy and Efficiency Improve- ment." Korea Development Institute Working Paper 8811. As cited in L.Jones, "Comments," in R. Picciotto and E. Wiesner, eds., Evaluation and De- velopment The Institutional Dimension, New Brunswick, NJ.: Transaction, 1998. Hersey, P., and K. H. Blanchard. Management of Organizational Behavior: Utilizing Human Resources. Englewood Cliffs, NJ. Prentice-Hall, 1993. 566 o Innovatons in Health Service Delivery Heskett, J. L., W E. Sasser, Jr., and L. A. Schlesinger. 1997. The Service Profit Chain: How Leading Companies Link Profit and Growth to Loyalty, Satisfaction, and Value. New York: Free Press. Holland, W, ed. 1983 Evaluation of Health Care. Oxford: Oxford University Press. Hunter, D. J., and G. Fairfield. 1997. "Disease Management (Managed Care, Part 3)." Bntish MedicalJournal 314 (7099): 50-53. Jimenez, E., and Y. Sawada. 1999. "Do Community-Managed Schools Work? An Evaluation of El Salvador's EDUCO Program." World Bank Economic Review 13(3): 415-41. Joshi, J. M. 1979. Theory of Value, Distribution and Welfare Economics. New Delhi- Vikas Publhshing House. Kaplan, R., and D. Norton. 1992. "The Balanced Scorecard: Measures That Drive Performance." Harvard Business Review, Harvard Business Review on Measunng Corporate Performance. Boston. Harvard Business School Press. Kmg, E. M., and B. Ozler. 1998. "What's Decentralhzation Got to Do with Learn- ing? The Case of Nicaragua's Autonomy Reform," Working Paper 9 on Impact Evaluation of Education Reforms. Washington, D.C.: World Bank, Develop- ment Economics Research Group, April. Kutzin, J., and B. McPake. 1997. Methods for Evaluating Effects of Health Reforms. World Health Orgamzation ARA Paper 13, Geneva. Lewis, M. A., M. B. Sulvetta, and G. M. La Forgia. 1990. Estimating Public Hospital Costs by Measunng Resource Use: A Dominican Case. Washington, D.C.. Urban Institute. Luft, H. 1987. Health Maintenance Organizations: Dimensions of Performance. New Brunswick, NJ.: Transaction Books. Mackay, K., ed. 1998. "Public Sector Performance-The Critcal Role of Evalua- tion." Selected proceedings from seminar on the experience in developing na- nonal evaluation systems, World Bank Operatons Evaluation Department, Washington, D.C., April. Melum, M. M., and M. K. Sinioris. 1992. Total Quality Management: The Health Care Pioneers. Chicago: American Hospital Publishing, American Hospital Associa- tion Company. Mills, A. 1997. "Improving the Efficiency of Public Sector Health Services in De- veloping Countries: Bureaucratic versus Market Approaches." In C. Colclough, ed., Marketizing Education and Health in Developing Countries: Miracle or Mirage? Oxford: Clarendon Press. Millstem, I. 1998. "The Corporaton as an Ownership and Management Structure." Paper written for the World Bank Private Sector Development Forum, Sep- tember 16-18, Baltimore, Md. Bibliographies * 567 Millstem, I., M. Albert, A. Cadbury, R. Denham, D. Feddersen, and N. Tateisi. 1998. Corporate Governance. Improving Competitiveness and Access to Capital in Global Markets: A Report to the OECD by the Business Sector Advisoiy Group on Corporate Governance. Pans: OECD (Organization for Economic Cooperation and Development). Mwabu, G., M. Amsworth, and A. Nyamete. 1993. Qzzalty of Medical Care and Choice of Medical Treatment in Kenya: An Empincal Analysis. Techmcal Working Paper 9. World Bank, Africa Technical Department, Human Resources and Poverty Division, Washington, D.C. Numerof, R. 1982. The Practice of Management for Health Care Professionals. New York: AMACOM [Div. American Management Associations]. ODA (Overseas Development Administration). 1996. "A Guide to Appraisal, De- sign, Monitoring, Management and Impact Assessment of Health and Popula- tion Projects." Draft 3. ODA, Health and Population Division, London. Ovretveit, J. 1993. Health Service Quality: An Introduction to Methods for Health Ser- vzces. London: Blackwell Science. Partnerships for Health Reform. 1997. Measunng Results of Health Sector Reform. A Handbook of Indicators. Bethesda, Md. Abt Associates. Pauly, M. 1996. "Health Systems Ownership: Can Regulation Preserve Community Benefits)" Frontiers of Health Services Management. 12 (3). 3-34 (spring). Picciotto, R., and E. Wiesner, eds. 1998. Evaluation and Development: The Institu- tional Dimension. New Brunswick, NJ.: Transaction. Reeder, L. 1999. "Anatomy of Disease Management Program." Nursing Manage- ment 30 (4): 41-45. Salvatore, D. 1986. Microeconomics: Theory and Applications. New York. Macmillan. Scott, W R., and A. B. Flood. 1987. "Costs and the Quality of Care in Hospitals: A Review of the Literature " In A. B. Flood and W R. Scott, eds., Hospital Struc- ture and Performance. Baltimore, Md.: The Johns Hopkins Umversity Press. 1987. "Conceptual and Methodological Issues in Measurmg the Quality of Care in Hospitals." In A. B. Flood and W R. Scott, eds., Hospital Structure and Performance. Baltimore, Md.: The Johns Hopkins University Press. Shirley, M. 1998. "Why Performance Contracts for State-Owned Enterprises Haven't Worked." World Bank, Finance, Pnvate Sector, and Infrastructure Network, Viezpoint, no. 150, August. . 1998. "Performance Contracts: A Tool For Improving Public Services?" In R. Picciotto and E. Wiesner, eds., Evaluation and Development: The Institutional Dimension, New Brunswick, NJ.: Transaction. Stout, S., A. Evans, J. Nassim, and L. Raney. 1997 Evaluating Health Projects: Lessons from the Literature. World Bank Discussion Paper 356. Washington, D.C. 568 o Innovations in Health Service Delivery Tarlov, A. R., and R. V Colley. 1992. "Clinical Outcome Management." In M. M. Melum and M. K. Simoris, Total Quality Management: The Health Care Pioneers Chicago: American Hospital Publishing Inc., American Hospital Association. Thomason, J., and K. Edwards. 1991. "Using Indicators to Assess Quality of Hos- pital Services in Papua New Gunea." International Journal of Health Planning and Management 6: 309-24. Trisolini, M. G., S. S. Russell, G. Gretchen, and D. Zschock. "Methods for Cost Analysis, Cost Recovery, and Cost Control for a Public Hospital in a Develop- ing Country: Victoria Hospital, St. Lucia." international Journal of Health Plan- ning and Management 7: 103-32. Wagstaff, A., and N. Watanabe. 2000. "The Poverty Impact of Out-of-Pocket Pay- ments for Health Services: China and Egypt Compared." Note in WHO, World Health Report 2000, Geneva: WHO, 2000; based on A. Wagstaff, N. Watanabe, E. van Doorlaer, and K Xu, "Poverty and Equity in the Financing of Health Care in Selected Low and Middle-Income Countries," World Bank. Processed. World Bank. 1993. World Development Report: Investing in Health. New York: Oxford Umversity Press. - 1995. Bureaucrats in Business: The Economics and Politics of Government Own- ership. A World Bank Policy Research Report. New York: Oxford University Press, 1995. Young, M., J. van der Gaag, and D. de Ferranti. 1998. "Learning from Early Child- care and Education Programs." In R. Picciotto and E. Wiesner, eds., Evaluation and Development: The Institutional Dimension, New Brunswick, NJ.: Transaction. Zander, R. N. 1992. "Critical Pathways." In M. M. Melum and M. K. Sinions. 1992. Total Quality Management: The Health Care Pioneers. Chicago: American Hospi- tal Publishung, American Hospital Association Company. Chapter 4 Batley, R. 1999. "An Overview of Findings." Paper 2 presented at the Workshop on the Changing Role of Governments in Adjusting Economies, at the University of Birmingham, March 28-30. Collins, D., G. Njeru, J. Meme, and W Newbrander. 1999. "Hospital Autonomy: The Expenence of Kenyatta National Hospital." International Journal of Health Planning and Management 14:129-153. Mills, A., S. Bennett, and S. Russell. 1999. "Health Sector-Summary Paper." Paper 3 presented at the Workshop on the Changing Role of Governments in Ad- justing Economies, University of BLrmingham, U.K, March 28-30. Bibliographies * 569 OECD (Orgamnsation for Economic Co-operation and Development). 1994. The Reform of Health Care Systems- A Review of Seventeen OECD Countries. Paris OECD. Walt, G. 1998. "Implementing Health Care Reforn- A Framework for Discussion." In R. Saltman, J. Figueras, and C. Sakellarides, eds., Critical Challenges for Health Care Reforwt in Europe. Buckingham, U.K.: Open Umversity Press. Walt, G., and L. Gilson. 1994. "Reforming the Health Sector m Developing Coun- tries: The Central Role of Policy Analysis." Health Policy 9: 353-70. Chapter 5 Adeyi, O., S. Radulescu, S. Huffman, C. Vladu, and R. Florescu. 1998-99. "Health Sector Reform in Romania: Balancing Needs, Resources, and Values." Euro- health 4 (6) Special Issue: 29-32 (winter). Cuyler, A. 1989. "The Normative Economics of Health Care Finance and Provi- sion." Oxford Review of Economic Policy 5: 34-58. Eisenhardt, K. 1989. "Agency Theory An Assessment and Review." Ameencan Man- agentent Review 14 (1): 75-83. Ensor, T 1993 "Health System Reform in Former Socialist Countnes of Europe" InternationalJournal of Health Planning and Management 8. 169-87. Goldstein, E., A. S. Preker, 0. Adeyi, and G Chellaraj. 1996. "Trends in Health Sta- tus, Service and Finance: The Transition m Central and Eastern Europe." World Bank Technical Paper 341 Social Challenges of Transition Senes. Washington, D C. Ho, T 2000. "Hospital Restructunng in Central Eastemn Europe and the Former Soviet Union." In M McKee, ed., An Appropriate Role for the Hospital Buck- ingham, U.K.: Open University Press Hsiao, W 1997. "Risk and Incentive- The Tale of Modern Provider Payment Sys- tems." Paper presented at the International Conference on Health System Re- form, Taipei, Taiwan, China. Jakab M., A. S. Preker, and A. Harding. 2002 "Linking Organizational Structure to the External Environment Experiences from Hospital Reform in Transition Economies." In M. McKee and J. Healy, eds., Hospitals in a Changing Europe Buckingham, U.K.: Open University Press. Klugman, J., and G. Schieber. 1997. "A Survey of Health Refonn in Central Asia." In R. Feachem, Z. M. Henscher, and L. Rose, eds., "Implementing Health Sec- tor Reform in Central Asia." EDILearning Resources Series, World Bank, Wash- ington, D.C. 570 o Innovations m Health Service Delivery Langenbrunner,J., and M. Wiley. 2000. "Paymg the Hospital: Payment Policies and Reformns in Eastern Europe and the Former Soviet Union." In M. McKee, ed., An Appropriate Role for the Hospital. Buckingham, U.K.: Open University Press. Lewis, M. 2002. "Informal Health Payments in Eastern Europe and Central Asia." In E. Mossahos and A. Maresso, eds., Funding Health Care: Options in Europe. Buckingham, U.K: Free Press. Maynard, A., and K. Bloor. 1999. "Payment, Contracting, and Regulation of Providers." World Bank, Flagship Course Matenal on Health Sector Reform and Sustainable Financing, Washington, D.C. Milgrom, P., and J. Roberts. 1992. Economics, Organization, and Management. Engle- wood Clhffs, NJ.: Prentice Hall. Observatory (European Observatory on Health Care Systems). 1999. Health Care Systems in Transition: Albania. http.//www.who.dk/observatory/hits/hits.htm. - . 1999. Health Care Systems in Transition: Croatia. http://www.who.dkl observatory/hits/hits.htmn. - 1999. Health Care Systems in Transition: Hungary. http://www.who.dk- observatory/hits/hits.htm. - 1999. Health Care Systems in Transition: Kazakhstan. http://www.who.dk- observatory/hits/hits.htm. - . 1999. Health Care Systems in Transition: Poland. http://www.who.dkl observatory/hits/hits.htm. - 1996. Health Care Systems in Transition: Czech Republic http://www.who.dkl observatory/hits/hits.htm. - 1996. Health Care Systems in Transition: Estonia. http://www.who.dkl/ observatory/hits/hits.htm. - 1996. Health Care Systems in Transition: Latvia. http://www.who.dkW observatory/hlts/hits.htm. . 1996. Health Care Systems in Transition: Lithuania http-//www.who.dk/ observatory/hits/hits.htm. - . 1996. Health Care Systems in Transition: Romania. http://www.who.dkW observatory/hits/hits.htm. Preker, A., and R. Feachem. 1995. "Market Mechanisms and the Health Sector in Central and Eastern Europe." World Bank Technical Paper 293. Washing- ton, D.C. Preker, A., M. Jakab, and M. Schneider. 2000. "Health Financing Reform in Tran- sition Economies." In E. Mossalios and A. Maresso, eds., Funding Health Care: Options in Europe. Buckingham, U.K. Free Press. Bibliographies * 571 Saltman, R. B., and J. Figueras. 1997. European Health Care Reform: Analysus of Cur- rent Strategies Copenhagen: WHO Regional Office for Europe. Schiavo-Campo, S., ed. 1994. Institutional Change and the Public Sector in Transitional Economies. World Bank Discussion Paper 241. Washington, D.C. Staines, V 1999. "A Health Sector Strategy for the Europe and Central Asia Re- gion." Health, Nutrition and Population Series World Bank, Human Develop- ment Network. Washington, D.C. Wiley, M. M. 1995. "Hospital Financing and Payment Systems: A Review of Se- lected OECD Countries." Report for the Organization for Economic Cooper- auon and Development, Pans. . 1992. "Hospital Financing Reform and Case-Mix Measurement: An Inter- national Review." Health Care Financing Reviezv 13 (4): 119-33 (summer). World Bank. 1999. "Health Policy Note: Croatia." World Bank, Europe and Cen- tral Asia Department, Human Development Unit. Washington, D.C. Chapter 6 Alvarez, M., and G. Prieto 1999 "Descentralizaci6n en salud en La Republica de Argentma." ISALUD, Buenos Aires. Processed. - 1999. "Descentralizaci6n en salud en la Republica de Argentina." World Bank, Washlngton, D.C. Processed. Bennett, S., and A. Mills. 1998. "Government Capacity to Contract. Health Sector Experiences and Lessons." PublicAdministration and Development 18 307-326. Government of Argentina. Decree 9/93. Decree 578/93. Law 17.102/66. Government of Chile, Ministerio de Salud. 1998. Desarollo e impacto de la imple- mentacian de compromisos de gestudn en el sistema puiblico de salud de Chile. Santiago, November. , Ministeno de Salud, Subsecretaria de Salud, Unidad de Gesti6n de Servi- cios 1998. Compromisos de gesti6n del SNSS, 1998. Santiago, October. Government of Uruguay, Ministeno de Salud Publica, Administraci6n de Servicios de Salud del Estado. 1998. "Compromiso de gesti6n, antecedentes, anexos, marco normativo." Montevideo, October. Processed. 572 o Innovations in Health Service Delivery Greer, A. L. 1977. "Advances in the Study of Diffusion and Innovation in Health Care Organizations." Milbank Memronal Fund Quarterly, Health and Society 55: 505-32. Ham, C. 1998. "Mix and Match: A Fashionable Love Affair with Market Forces Starts to Fizzle Out." Guardian, February 18. Harding, A., and G. Prieto. 2001. "The Public Hospital System in Buenos Aires Province: Current Situation." Health Systems Development Thematic Group Case Study. World Bank. Washington, D.C. Harding, A., and A. S. Preker. September 2000. "Understandmg Organizational Re- form. The Corporatizauon of Public Hospitals." HNP Discussion Paper. World Bank, Washington, D.C. Kaluzny, A., and J. Veney. 1977. "Types of Change and Hospital Planning Strate- gies." Amencan Journal of Health Planning 1 (3, January). 13-19. Kaluzny, A., and S. R. Hernandez. 1983. Organizational Change and Innovation in Health Care Management: A Text in Organizational Theory and Behavior. New York. Wiley. Le Grand, J., N. Mays, and J-A. Mulligan. 1997. Learning from the NHS Internal Market: A Review of the Evidence. London: Kmg's Fund. Lemus, J. 1999. "Transformaci6n del modelo de atenci6n de la salud en el marco de la reforma sectonal de la Republica de Argentina." Report presented at the First Europe-Latin Amenca and Caribbean Health Reform Forum, May 10-12, Paris. Processed. Mills, A. 1998. "Government Capacity to Contract: Health Sector Experience and Lessons." Publc Administration and Development (U.K) 18 (October): 307-26. Osborn, D., and P. Plastrick. 1997. Banishing Bureaucracy. New York: Addison. Schick, A. 1996. The Spint of Reform: Managing New Zealand State Sector in a Time of Change. Wellington: State Services Commission. Shortell, S. 1977. "The Role of the Environment in a Configurational Theory of Orgamizations. Human Relations 30 (3): 275-302. Shortell, S., and A. Kaluzny, eds. 1983. "Organizational Theory and Health Care Management." In Health Care Management: A Text in Organization Theory and Behavior New York: Wiley. Sojo, A. 1996. "Reformas de gesti6n en la salud p6blica en Chile." CEPAL Review 3 (August): 127-45. Bibliographies - 573 Walt, G., and L. Gilson. 1994. Reforming the Health Sector in Developing Coun- tries: The Central Role of Policy Analysis." Health Poliy 9: 353-70. Will, B, and L. Harrison. 1983. "Quasi-Markets and the National Health Service Reforms." In J. Le Grand and W Bartlett, Quasi-Markets and Social Policy. Lon- don: Macmillan. Chapter 7 Appleby, J., P. Smith, W Ranad, V Little, and R. Robinson. 1994. "Monitonng Managed Competition." In R. Robinson and J. Le Grand, eds., Evaluating the NHS Reforms. London: King's Fund Institute. Barrick, A., and P. Balcombe. 1998. Anglian Harbours-The First NHS Trtst Dtssolu- tion. N.p.: NHS Executive. Bartlett, W, and J. Le Grand. 1994. "The Performance of Trusts." In R. Robinson and J. Le Grand, eds., Evaluating the NHS Reforms, London: King's Fund Institute. Bloor, K., and A. Maynard. 1994. "An Outsider's View of the NHS Reforms " BMJ 309: 352-53. Ferlie, E., L. Ashburner, L. Fitzgerald, and A. Pettigrew. 1996. The New Public Man- agement in Action. Oxford: Oxford University Press. Flynn, R., G. Williams, and S. Pickard. 1996. Markets And Netwvorks. Contracting in Community Health Services. Buckingham, U.K.: Open University Press. Foster, C D., and E J. Plowden. 1996. The State under Stress. Buckingham, U.K.: Open University Press. Goodwin, N. 1998. "GP Fundholding." In J. Le Grand, N. Mays, and J. Mulligan. 1998. Learning from the NHS Internal Market. London: King's Fund Institute. Hall, S., and M. Jacques, eds. 1989. New Times. London: Lawrence and Wishart. Ham, C. 1994. "Where Now for the NHS Reforms?" BM] 309 351-52. 1996. Public, Pnvate or Community? What Next for the NHS? London: DEAMOS. - 1997. Management and Competition in the NHS. Oxford: Radcliffe Medical Press. Hamblin, R. 1998. "Trusts." In J. Le Grand, N. Mays, and J. Mulligan, eds., Learn- ingfrom the NHS Internal Market, London: King's Fund Institute. 574 o Innovations m Health Service Delivery Hood, C. 1991. "A Public Management for All Seasons." Public Administration 69: 3-19 (spring). Klein, R. 1995. "Big Bang Health Care Reform-Does It Work? The Case of Britam's 1991 National Health Service Reforms." Milbank Quarterly 73: 299-337. Le Grand, J., N. Mays, and J. Mulligan, eds. 1998. Learning from the NHS Internal Market. London: King's Fund Institute. Locock, L., and D. Welsh. 1994. Pulling the Right Levers. Oxford: NHS Executive Anglia and Oxford. Maynard, A., and K. Bloor. 1996. "Introducing a Market to the United Kingdom's National Health Service." New England yournal of Medicine 334: 604-608. McHale, J., D. Hughes, and L. Griffiths. 1997. "Conceptualising Contractual Dis- putes in the National Health Service Internal Market." In S. Deakmn and J. Michie, eds., Contracts, Co-operation and Competition, Oxford: University Press. Newchurch. 1993. The Third Newchurch Guide to NHS Trusts London: Newchurch. Newdick, C. 1996. Who Should We Treat? Oxford: Clarendon Press. OECD (Organisation for Economic Co-operaton and Development). 1994. OECD Economic Surveys: United Kingdom 1994. Paris: OECD. Osborne, D., and T Gaebler. 1992. Reinventing Government. Reading, Mass.: Addi- son-Wesley. Polhtt, C., J. Birchall, and K. Putman. 1998. Decentralising Public Service Manage- ment. London: Macmillan. Propper, C. 1995a. "Agency and Incentives in the NHS Internal Market." Social Sci- ence and Medicine 40 (12). 1683-90. - . 1995b. "Regulatory Reform of the NHS Internal Market." Health Econom- ics 4: 77-83. - . 1996. "Market Structure and Prices: The Response of Hospitals in the U.K. National Health Service to Competition." Journal of Public Economics 61: 307- 35 (1996). Propper, C., and N. Soderlund. 1998. "Competition in the NHS Internal Market: An Overview of its Effects on Hospital Prices and Costs." Health Economics 7: 187-97. Raftery, J., R. Robmson, and J. Mulligan. 1996. "Contracting In the NHS Quasi- Market." Health Economics 5: 353-62. Ranson, S., and J. Stewart. 1994. Management for the Public Domain: Enabling the Learning Society. London Macmillan. Bibliographies * 575 Robmson, R., and J. Le Grand, eds. 1994. Evaluating the NHS Reforms. London: King's Fund Institute. Smee, C. 1995. "Self-Govermnng Trusts and GP Fundholders: The British Experi- ence." In R. Saltman and C. von Otter, eds. Implementing Planned Markets in Health Care. Buckingham, U.K: Open University Press. Soderlund, N., I. Csaba, A. Gray, R. Mdne, and J. Raftery 1997. "Impact of the NHS Reforms on Enghsh Hospital Productivity: An Analysis of the First Three Years." BMJ 315: 1126-29 Walsh, K. 1995. Puhlhc Services and Market Mechanis7s. London: Macmillan. Government of the United Kingdom Comptroller and Auditor General. 1998. NHS (England). Summarized Accounts 1996-97. London: Her Majesty's Stationery Office. Department of Health. 1989. Self-Governing Hospitals. Working Paper 1. London: Her Majesty's Stationery Office. 1990. NHS Trusts: A Working Guide. London: Her Majesty's Stationery Office. - 1994. The Operation of the NHS Internal Market. London: Her Majesty's Sta- tionery Office. Department of Health and Social Secunty. 1983. NHS Management Inquiry ["Grif- fiths Report]. London: DHSS. National Health Service Executive. 1996. Seeing the Wood, Spanng the Trees Leeds: NHS Executive. National Health Service and Community Care Act 1990. 1990. London: Her Majesty's Stationery Office. Secretary of State for Health. 1989. Working for Patients. London: Her Majesty's Stationery Office. Chapter 8 Government of New Zealand Area Health Board Act 1983. CCMAU [Crown Company Monutoring and Advisory Unit]. 1996. "Crown Health Enterprises: Briefing to the Incoming Minister." Wellington. Controller and Auditor-General. 1998. "Second Report for 1998." Wellington. 576 o Innovations in Health Service Delivery Department of Statstics. 1999. Publhc Health Financial Statistics: Annual Operating Statements june 1998, 1999 Quarters. Wellington. - 2000. "Quarterly Results for December 1999." Wellington. Health and Disability Commnssioner. 1998. "Canterbury Health Ltd: A Report by the Health and Disability Commissioner." Wellington. Health Disability Services Act 1993 and amendments. Health Reforms (Transitional Provisions) Act 1993. Hospital and Related Services Taskforce. 1987. "Unshackling the Hospitals." Welling- ton. Minister of Finance, Mmiister for Crown Health Enterprises. 1996. "Statement of Shareholders' Expectations of Crown Health Enterprises." Wellngton, April. Minister of Health. 1988. "Health: A Prescription for Change." Wellington. 1991. "Your Health and the Public Health." Wellington. 1996. Letter to Chairman, Central RHA, regarding a response to concerns raised on the proposed rate of return requirement on CHEs, May 9. - 1996. Letter to CHE Chairs and CEOs, regarding business plans, Decem- ber 20. - 1998. "Hospital Services Plan." Wellington. Minister of Health and Associate Mimster of Health. 1997 Letter to RHA and CHE Chairs and CEOs. "Operationalising the Coalition Agreement," March 20. Minuster of Health and Minister of Finance. 1997. Letter to CHE CEOs and CFOs. "1997-98 CHE business plannig process," February 13. Ministry of Health. 1986. Health Benefits Review. Choices for Health Care. Wellngton. . 1993. New Zealand Hospital Sector Performance 1983-1992. Wellington. 1994. Letter and attachments to CEO, Central RHA. Letter entitled "Meet- ings with Mnistry 11-13 January 1994: Recent Government Decisions on RHA Funding and Other Policy; Comment on RHA Strategic Purchase Plans" and attached documents entitled "Impact of Changed Planning Assumptions on the RHA Funding Path." "RHA Funding Assumptions." "Addressing Adjust- ment Pressures Facing RHAs and CHEs Over the Next Three to Five Years" and "Addressing Health Expenditure/Funding Imbalances," January 11. - 1994. Facsimile to RHA and CHE CEOs, "Data Used in Bench Marking to Estimate Aggregate Efficiency Gains in CHE Services," February 15. Bibliographies * 577 . 1995. Facsimile to RHA CEOs, tided "Draft Owners" Letter to Crown Health Enterprises," December 22. . 1997. Health Expenditure Trends in New Zealand 1980-1997. Wellington. . 1998. Purchasingfor Your Health 1996-97. Wellington. . 1998. "1998 Survey, Health Funding Authority," Wellington. . 1999. Health Expenditure Trends in iVew Zealand 1980-1998. Wellington. National Interim Provider Board. 1992. Providing Better Health Care for Nevw Zea- landers Report to the Government and to the Public. Wellington. Steering Group. 1997. Iniplementing the Coalition Agreemient on Health. Wellington Other Sources Arthur Andersen & Co. 1987. Publhc Hospital Performance Assessment. Report for De- partment of Health. Wellngton. . 1994. Letter to Crown Health Enterprises and Regional Health Authonty Chairs and Chief Executive Officers, regarding an outline of the process that arnved at the new funding scenario, January 26. Chambers, K. 1996. "Understanding Consumer Needs, Wants and Perceptions." Report for Central Regional Health Authority. Wellington. Clayton, C. 1996. "Review of Managed Care." Report for Central RHA. Wellington. Coopers & Lybrand. 1997. "Central Health: Secondary Services Study." Report prepared for Central Health. Wellington. - 1996. "Health Centres Project, A Framework for Small Health Centres." Report prepared for the Central Regional Health Authority. Wellington. CS First Boston. 1992. "Structural, Organisatnonal and Legislative Issues Relating to the Refonn of Public Health Care Providers." Report for Department of the Prime Mmruster and Cabinet. Wellington. Deloitte Touche Tohmatsu 1996. "Trends in Area Health Board/CHE Perfor- mance 1989/90-1997/98." Report for CCMAU. Wellington. Hawkins, L. 1997. "CHE Shareholder and Banker Roles and Linkages to the Role of the Funder and of Contracts." Report prepared for the Central Regional Health Authonty. Wellington. Vaithianathan, R. 1997. "Institutional Issues in the CHE Sector." Unpublished document. - 1999. "The Failure of Corporatization: Public Hospitals m New Zealand." Agenda 6 (4):325-38. 578 0 Innovations in Health Service Delivery Chopter 9 Adeney, E. 1997. "The Invisible Reconstruction of the Victorian Metropolitan Hos- pital System." Journal of Law and Mfedicine 4 (3): 267-75 (February). Cameron, P., M. Kennedy, and J. McNeil. 1999. "The Effects of Bonus Payments on Emergency Service Performance m Victoria." Medical Journal of Australha 171 (6): 243-46 (September). Duckett, S. 1998. "Economics of Hospital Care." In G. Mooney and R. Scotton, eds., Economics and Australian Health Pohiy. Sydney: Allen and Unwin, pp. 93-104. HIC (Health Issues Centre). 1995. Networking Hospitals. Consumers and Communities. Melbourne. - 1999. "Health Services Policy Review Response." Melbourne, July. Processed. Healy, W, and H. Varney. 1995. "Mainstreamnng and Integrating Psychiatric Ser- vices: The Victorian Experience." Australian Joournal of Social Issues 30 (2): 179- 94 (May). Lin, V, and S. Duckett. 1997. "Structural Interests and Organizational Dimensions of Health System Reform." In H. Gardner, ed., Health Poltcy in Austraha. Mel- bourne: Oxford University Press. St. Vincent's Hospital. 1997. Annual Report 1997. Melbourne. - . 1998. Annual Report 1998 Melbourne. Swenssen, H., and S. Duckett. 1997. "Health Policy and Financing." In H. Gardner, ed., Health Policy in Australia. Melbourne: Oxford University Press, pp. 13-45. Victorian AIDS Council/Gay Men's Health Centre. 1999. "Response to Health Ser- vice Policy Review Discussion Paper." Melbourne. Processed. Walker, C. 1994. "The Emergence of the Hospital System m Melbourne: 1846 to 1975." Ph.D. diss., La Trobe University, Melbourne. Government of the Commonwealth of Australia AIHIW (Australian Institute of Health and Welfare). 1999. Australian Hospital Sta- tistics 1997-98. Canberra: AIHW. SCRCSSP (Steering Committee for the Review of Commonwealth/State Service Provision). 1997. Reforms in Government Service Provision. Canberra. - 1997. Report on Government Service Provision 1997. Canberra. Bibliographies * 579 1999. Report on Government Services 1999 Canberra: www.pc.gov.au/ service/gspindex.html (accessed 2 August 1999). TQAHC (Taskforce on Quality m Australian Health Care). 1996. The Final Report of the Taskforce on Quality in Australian Health Care Australian Health Mnisters' Advisory Council. Canberra, June. Government of the State of Victoria DH&CS (Department of Health and Community Services). 1995. Hospital Services Report: December 1995. Melboumne. DHS (Department of Human Services Victoria). 1995. Hospital Services Report: De- cember 1995. Melbourne. . 1996. A Healthier Future: A Plan for Metropolitan Health Care Services. Melbourne. . 1996. DPC (Department of Premier and Cabinet). Competitive Neutrality: A Statement of Victorian Government Policy. Melbourne. 1997. A Healthier Future- The Metropolitan Hospitals Projects. Melbourne. 1999. Public Hospitals Policy and Funding Guidelines 1999-2000. Melbourne. 1999. Financial Management Act, Annual Reporting Requirements, Publhc Hos- pitals 1998/9 Melbourne. . 1999. Hospital Services Report: December 1999. Melbourne. DTF (Department of Treasury and Finance). 1998. The Capital Assets Charge in 1998-99. Melbourne Duckett, S., L. Hunter, and A. Rassaby. 1999. Health Services Policy Reviezw Discus- sion Paper. Melboumne: Department of Human Services, March. Economic and Budget Review Committee. 1992. Hospital Services in Victoria. Effi- ciency and Effectiveness of Health Service Agreements Parliament of Victoria, Mel- bourne, August. Fahrer, J. 1995 Hospital Costs in Melbourne An Economet-ic Analysis. Melbourne Metropolitan Hospitals Planning Board. HOI (Health Outcomes International). 1999. Evaluation of the Relocated HiVIAIDS and Infectious Diseases Services. Melbourne. Department of Human Services. IEHCN (Inner and Eastern Health Care Network). 1999. Progress to Date East Melboumne, August. - 1999. Response to the Health Services Policy Review Discussion Paper. East Melbourne. 580 o Innovations in Health Service Delivery MHPB (Metropolitan Hospitals Planning Board). 1995. Developing Melbourne' sHos- pital Network: Intenm Pbase 1 Report. Melbourne, April. - . 1995. Developing Melbourne's Hospital Network: Phase 1 Report. Melbourne, June. - . 1995. Taking Melbourne's Health Care Networks into the 21st Centuiy: Phase 2 Report. Melbourne, December. Nicholls, D. 1992. Independent Review of Victoria's Public Sector Finances. Victorian Government, Melbourne. NWH (North Western Health). 1998. North Western Health Annual Report 1997-98. Melbourne. SHCN (Southern Health Care Network). 1997. Annual Report 1997. Melbourne. . 1998. Annual Report 1998. Melbourne. Victornan Auditor General's Office. 1998. Acute Health Services Under Case Mix. A Case of Mtxed Priorities. Melbourne, May. . 1999. Report of Ministerial Portfolios. Melbourne, May. Victonan Comnmssion of Audit. 1993. Report of the Victornan Commission ofAudit: Vol- ume Two. Melbourne, May. WHCN (Western Health Care Network). 1997. Western Health Care Network An- nual Report 1996-97. Melbourne. Chapr 1O Cheung, A. B. L. 1995. "The Politics of Administrative Reforms in Hong Kong: Corporatization of Public Services during the 1980s." Ph.D. diss., London School of Econornics. Gauld, R., and D. Gould. 2000. The Hong Kong Healthcare System. Otago, New Zealand: University of Otago Press. Hong Kong Annual Digest of Statistics. 1985 ed. Census and Statistics Department, Hong Kong. Hospital Authority. 1992. Annual Report. Hong Kong. - 1993. SMC-P80. Initial Plan to Sustain the Orgamzatnon Transformation Following the Current Phase of the NMI Effort. Hong Kong. - 1994. Human Resources Policies Manual. Hong Kong, July 26. Bibliographies * 581 . 1996. Statistical Report 1995/96. Statistical Information Section, Hospital Authority. Hong Kong. - 1997. HA-P62 1. Progress Report of the Puiblic Complaints Committee. Hong Kong. . 1997. Procurement and Materials Management Manual Hong Kong. - 1998. Patient Satisfaction Survey-Analysts of Annual Survey on Image of HA and Tracking Polls, April 17 Hong Kong. 1998. HAB-P3. Report of the Public Complaints Committee. Hong Kong. 1998. HA-P647. Budget Allocation 1998/99 Hong Kong. 1998. PC-P53. Overview of Government Funding of HospitalAuthority Services. Hutcheon, R. 1999. Bedside Manner: Hospital and Health Care in Hong Kong. Hong Kong: Chinese University Press. Scott, W D. 1985. The Delivery of Medical Services in Hospitals: A Report for the Hong Kong Government. Hong Kong. Yeoh, E. K 1993. "The Corporatisation of Our Public Hospitals." HKAccountant (September-October). Yip, W, R. Rannan-Eliya, and W C. Hsiao 1999. Health Care Expenditure Estimates for Hong Kong. The Domestic Health Accounts. Special Report I to the Hong Kong Government, January. Yip, W, and W Hsiao 1999. "Organizational Reform in the Public Hospital Sector- The Case of Hong Kong." Harvard University, Cambridge, Mass. Processed. Chapter 1 1 Economist Intelligence Unit. 1998. "Profile-Malaysian Healthcare: The Slow Road to Prvatization." Healthcare Asia (Thurd Quarter): 67-76 Hussein, R. H., and S. Nyunt-U. 1998. "Estimating Out-of-Pocket Health Care Expenditure in Malaysia." Paper presented at the Eighth Annual Public Health Forum, Reforming Health Sectors, London. InstitutJantung Negara Sdn. Bhd (NHI). 1997. Annual Reports, 1992-97. Suleiman, A. B. 1995. "The Approach of the Mnistry of Health in Malaysia in Fur- ther Upgrading Quality in Health Services." Paper delivered at EDI/ UNFIPA/ICOMP/World Bank Seminar on Quality of Reproductive Health Care as the Way Forward. Kuching, Malaysia, November 14-17. 582 o Innovations in Health Service Delhvery - . 1997a. "Health Care Issues in Malaysia' A Review of Government Policy Priornties." Paper presented at the Economist Conferences, Health Care Sys- tems in Asia-Challenges and Opportunities, Singapore, January 21-22. - 1997b. "Health Challenges Facing Malaysia." Fourteenth Global Scientific Meetng, INCLEN, Penang, February 18-24. . 1998. "A Script-Looking at Malaysia's Impending Healthcare Crisis." Economist Intelligence Unit. Healthcare Asia (First Quarter): 1-12. World Bank. 1998. World Development Indicators, 1998. Washington, D.C.: World Bank. Government of Malaysia Second Outline Perspective Plan (OPP2), 1991-2000: New Development Policy (NDP). 1991. Kuala Lumpur: National Press. Seventh Malaysia Plan 1996-2000. 1996. Kuala Lumpur: National Press. Ministry of Health. 1998. Annual Report 1996. Kuala Lumpur: National Press. Chapteir 12 Government of Singapore Ministry of Health. 1983. Blue Paper on the National Health Plan. Singapore. - 1985. "Joint Press Statement" with the National University of Singapore and Temasek Holdings (Pte) Ltd. Singapore, January 12. - . 1992. Toward Better Healthcare: The Main Report of the Review Committee on National Health Policies. Smgapore. -1 1993. Affordable Healthcare: A White Paper. Singapore. Ministry of Trade and Industry. 1993. Report of the Cost Review Committee 1993. Singapore. - 1996. Report of the Cost Review Committee 1996. Singapore. Other Sources Brntish Medical Journal. 1982. "Singapore Restructured." British Medical journal 285: 914-15. DBS Bank. 1990. The Medical Services Industry in Singapore. Report 15. Economic Research Department, June. Bibliographies * 583 Lancet 1990. "Singapore A Restructured Hospital Service." Lancet 335: 1208-1209. Lim, Meng-Kn. 1998a. "An Overview of Health Care Systems in Singapore." 7our- nal of Publhc Health Medicine 20 (1) 16-22. . 1998b. "Restructuring of Public Hospitals." Singapore General Hospital, Ninth Annual Scientific Meeting, Apnl 23-26, Singapore. SGH Proceedings 7 (4): 73-77. Massaro, T A., and Y N. Wong. 1995. "Positive Experience with Medical Savings Accounts in Smgapore." Health Affairs 14 (2): 267. Nichols, L. M., N. Prescott, and K. H. Phua. "Medical Savings Accounts for De- veloping Countries." Innovations in Health Care Financing. Proceedings of a World Bank Conference, March 10-11, 1997. World Bank Discussion Paper 365. Wash- ington, D.C. Pang, H. L. 1999. "Healthcare Services in Singapore-A Demand Analysis." Singa- pore Nursing_Journal 26 (2). 21-24. Phua, K. H. 1987. "Saving for Health." World Health Forum 8 (1). 38-41 - 1991. Pnvatization and Restructunng of Health Services in Singapore. IPS Oc- casional Paper 5, Institute of Policy Studies. Singapore: Times Academic Press. - 1997. "Medical Savings Accounts and Health Care Financing in Singa- pore." Innovations in Health Care Financing: Proceedings of World Bank Conference, March 10-11. World Bank Discussion Paper 365. Washington, D.C. - . 1998. "Singapore Case Study." Paper prepared for World Bank Experts' Seminar on Innovations in Organization of Health Care Delivery, November 12-13, Washington, D.C. Processed. . 1999. "Comparative Healthcare Fmancing Systems with Special Reference to East Asian Countries." Research in Healthcare Financial Management 5 (1): 113-33. Prescott, N 1997. "A Script-How to Manage Rising Healthcare Costs in East Asia." Chapter 3 in HealthCare Asia, Fourth Quarter. London: Economist In- telligence Unit. Schieber, G. J., ed. 1997. Innovations in Health Care Financing Proceedings of a World Bank Conference March 10-11 World Bank Discussion Paper 365. Washington, D.C. Thynne, I., and M. Ariff, eds. 1989. Pnvatization: Smngapore' Experience in Perspec- tive. Singapore: Longman. 584 o Innovations in Health Service Delivery C[haper 13 Government of Tunisia Decree No. 81-1130, September 1, 1981. Decree No. 91-1844, December 2, 1991. Decree No. 91-1845, December 2, 1991. Decree No. 92-884, May 11, 1992. Decree No. 94-1704, August 8, 1994. Decree No. 95-1634, September 4, 1995. Decree No. 98-409, February 18, 1998. Law No. 91-63, July 29, 1991. Ministry of Public Health (MPH). 1997. Manuel deproceddures de gestlon desaffazresdes EPS. Tunis: MPH, March. Ministry of Public Health (MPH)/National Institute of Public Health. 1986. 1956-1986: Trente ans au service de la sante- un engagement, une 6thique. Tunis: MPH, May. Order of the Ministry of Public Health, March 1, 1995. Other Sources Achouri, H. 1995. Paper prepared for seminar "Pour le XX[ si6cle, des echanges m6diterran6ens," Lyons, France, October 19-20. Processed. 1995. "Le chemin vers l'autonomie de l'h6pital public tunisien." Paper pre- pared for World Health Organization Forum, "Renforcer les capacit6s des pays en economie de la sante," Geneva, December 6-8. Processed. Bach, S. 1993. "Health Care Reforms in the French Hospital System." International Journal of Health Planning and Management 8: 189-200. Balique, H. 1994. "Les programmes de sante publique en Afnque." Le Coumer 147: 56-60 (September-October). Barnum, H., and J. Kutzm. 1993. "Public Hospitals in Developing Countries: Re- source Use, Cost, Financing, The World Bank." Baltimore, Md.: Johns Hop- kins University Press. Bui Dang, and Ha Doan. 1993. "La Sante au prix coitant. Cahiers de sociologie et de d&nographbemidzcales23 (2): 153-68. Bibliographies * 585 Creinadez, M. 1987. "Les cles de 1'evolution du monde hospitalier." Les cahiers de Gestions hospital:eres, no. 59 and Gestions hospitaliNres, no. 266 (May). Crozier, M. 1987. "L'analyse strategique appliqu6e en milieu hospitalier: pertinence et m6thodologie. Gestions hospitaliires no 261 (December 1986-January). Crozier, M., and E. Fnedberg. 1977. "Organisation et pouvoir." L'acteur et le syste?ne. Paris: Seuil Delorme, J. 1994. "Un nouveau statut pour l'h6pital." Le concours medtcal: 166-211 (March 19) De Tovar, A. R., and G. Vallet. 1998. "CHU. Pour une r6flexion strat6gique." Ges- tions hospitalh?res 293 (February). Frenck, J. 1994. "Dimensions of Health System Reform." Health Poltcy 27 19-34 Goldstein, E, A. S. Preker, 0. Adeyi, and G. Chellaraj. 1996. Trends in Health Sta- tus, Services, and Finance The Transition in Central and Eastern Europe World Bank Technical Paper 341, vol. 1. Washington, D.C. Hermel, P. 1990 "Management et demarche strat6glque." Gestions hospitalitres 298 (August-September). Huard, P. 1977. "L'h6pital comme organisation formnelle " Techniques Hospitaliires 377 (February). Lefebvre,J. 1992. "Directeur d'h6pital quelle fonction?" Techniques HospitalNres 321 (December). Leonard, J. L. 1990. "Grer la complexite a l'h6pital." Gestions hospitalieres 298 (August-September). Michel, M. 1990. "Strategie de l'h&pltal ou strategies dans l'h6pital." Gestions hojpi- talieres 298 (August-September). Nadeau, J. A. 1996. "La reforme du systeme de sante au Quebec." World Hospitals and Health Services 32 (2). Newbrander, W, H. Barnum, and J. Kutzin. 1992. "Economue hospitalh6re et fi- nancement des h6pitaux dans les pays en developpement." Geneva- World Health Organization. Quebec Hospitals Association. 1994. "Demarche de la recherche de la performance AHQ, direction des resources humaines." Paper prepared for training session on "La recherche de la performance." Montreal: QHA, January. . 1994. "Le tableau de bord des etablissements de sante: D6scription des in- dicateurs de gestion." Montreal: QHA. Riley, C. 1987. "Le management ou la gestion? Un point de vue." Les cahtersfrancais 233 (October-December). 586 o Innovations in Health Service Delivery Saltman, R. 1994. "A Conceptual Overview of Recent Health Care Reforms." Euro- pean Journal of Publzc Health 4: 287-93. Schleber, G. 1995. "Preconditions for Health Reform: Experiences from the OECD Countnes." Health Policy 32: 279-93. Serdar, D. B., and I. Sheiman. 1996. "Contractng Models and Provider Competi- tion in Europe." WHO Book on European Health Care Refrrms. Copenhagen: World Health Organization. Serieyx, H. 1992. "Les strategies de changement." Revue Politiques et Management Public 10 (1): (March). Van Doorslaer, E., and A. Wagstaff. 1992. "Equity m the Delivery of Health Care: Some Internatnonal Compansons." Journal of Health Economics 11: 389-411. World Bank. 1993. "Investing in Health." World Development Report 1993. New York: Oxford Umversity Press. - . 1997. "Health, Nutrtion and Population: Sector Strategy." World Bank, Human Development Network, Washington, D.C. World Health Organization (WHO). 1993. "Evaluation des recentes reformes operees dans le financement des services de sante." Sinre des rapports techniques 829. Geneva. World Health Organization (WHO), Regional Office of Europe. 1996. "European Health Care Reforms: Analysis of Current Strategies." Copenhagen. Chapter 14 Government of Indonesia Laws 22 and 25, 1999. Presidential Decree No. 38, 1991. Otber Sources Bossert, T., S. Kosen, B. Harsono, and A. Gani. 1997. Hospital Autonomy in Indone- sia: Data for Decision-Making Project. Boston, Mass.: Harvard School of Public Health, April. Gani, A. 1996. "Improving Quality in Public Sector Hospitals in Indonesia." Inter- national Journal of Health Planning and Management 11: 275-96. Johnson, C. N., E A. Duncan, Jr., D. R. Hotchliss, S. Jacobalis, R. Noor, and C. Rice. 1998. "Assessment and Lessons Learned from the Klaten Integrated Bibliographies * 587 Health Care Reform Field Trial." Report prepared for the U.S. Agency for In- ternatonal Development, Jakarta, Indonesia, September. Processed. Malik, R., M. Nad)ib, B. Susetyo, and D. S. Argadiredja. 1998. Evaluation of Health Financing Reforms in Indonesia. Jakarta: World Health Organization. Rukmono, S. S. 1990. "Hospital Diagnosis Study." Health Sector Financing Project, Monograph Series 5. Jakarta. Saadah, F., S. Lieberman, and M. Juwono. 1999. Indonesian Health Sector Expendi- tures during the Crisis: Have They Been Protected Watching Brief 5. Jakarta: World Bank, East Asia and Pacific Region, August. Chapter 15 Government of Ecuador Ministry of Public Finance and Public Credit. 1999. MInisterial Resolution 018, March 18. Ministry of Public Health. 1998. "Tena: En la Ruta de la Salud y el Bienestar- Propuesta de Creaci6n del Sistema Descentralizado de Salud del Canton Tena." Quito. - . 1999. Reglamento al Decreto Ejecutivo 502, Funcionamiento de Comit6s de Participacian Socialy Control de la Gestidn en Servicios de Salud, Anexo 1, Tarifario Referencial Quito. . 1999. Ministerial Resolution 1292, April. Presidential Decree 502, January 22, 1999. Regulations 30 and 31 of the Law of the Modernization of the State, Privatization, and Delivery of Public Services by the Pnvate Sector, Law no. 50 published in the Official Gazette, no. 349, December 31, 1993. Executive Decree no. 2328, published in the Official Gazette, no. 581 (Supplement), December 2, 1994. Otber Sources California Consortium (Lewin-VHI, Inc., The Healthcare Redesign Group, Inc., Prof. Alam Enthoven of Stanford University), with assistance of Patricio Mar- quez and Juan Luis Londono of the World Bank. 1996 "Options for Health Reform in Ecuador." Prepared for the Ministry of Public Health and the Na- tional Health Council of Ecuador. San Francisco, Calif. Gershberg, A. 1999. "Education and Health Decentralization in Ecuador Practice, Plans and Prospects." Background paper prepared for the World Bank, Wash- ington, D.C., October. 588 ° Innovations in Health Service Delivery Johns Hopklns University. 1996. "A Plan to Modernize Hospital Services and Re- form the Health Sector in Ecuador." Prepared for the Mnuustry of Health and the National Health Council of Ecuador. Baltimore, Md. Lucio, R., and E. Lasprilla. 1999. "Costos y eficiencia de los Servicios de Salud en el Ecuador a 1999." Background paper prepared for the World Bank, Wash- ington, D.C., July. Marquez, P., and E. Murrugarra. 1999. "Health m Ecuador." Ecuador: Crsis, Poverty and Social Services. 2 vols. Report 19920-EC. World Bank, Washington, D.C., December 17. Organizaci6n Panamericana de la Salud/Organizaci6n Mundial de la Salud. 1999. Ecuador: Perfil del Sistema de Servczos de Salud. Programa de Organizaci6n y Gesti6n de Sistemas y Servicios de Salud. Quito: OPS/OMS. World Bank. 1992. "Staff Appraisal Report for the Ecuador Second Social Devel- opment Project-Health and Nutrition" (FASBASE), Washington, D.C. - 1998. "Project Appraisal Document for the Ecuador Health Services Mod- erruzation Project" (MODERSA), Washington, D.C. About the Authors The Editors Alexander S. Preker (Chief Economist for Health, Nutrition, and Population and editor of the HNP Publication Series) is responsible for coordinating the World Bank's Health Systems Development team and overseeing its analytical work on health financing and ser- vice delivery. He coordinated the team that prepared the Bank's 1997 Sector Strategy on health care in developing countries, was one of the authors of the World Health Report 2000 on health systems, and was a member of Working Group 3 of the WI-HO Commission on Macroeconomics and Health, which reported in 2001. Mr. Preker has published extensively on topics related to health systems devel- opment and is a frequent speaker at major international conferences. He is on the editorial committee of several journals and on the sci- entific committee of the International Health Economics Associa- tion Conference in San Francisco, 2003. Mr. Preker has an appointment as associate professor at the George Washington University, is on the External Advisory Board for the London School of Economics Health Group, and is a mem- ber of the teaching faculty for the Harvard/World Bank Institute Flagship Course on Health Sector Reform and Sustainable Financ- ing. His training includes a Ph.D. in economics from the London School of Economics and Political Science, a fellowship in medicine from University College London, a diploma in medical law and ethics 589 590 o Innovations in Health Service Delivery from King's College London, and an M.D. from the University of British Columbia/McGill. April Harding is a senior economist in the Health, Nutrition, and Population Department of the World Bank. Since arriving at the World Bank, she has provided technical assistance to governments in more than 13 countries, primarily on issues related to private sector development and privatization. Ms. Harding currently manages the analytical work, operational support, and training related to private participation in the health sector. She speaks and publishes in nu- merous forums on public-private partnerships, private participation, privatization, and reform of health services. She recently edited a vol- ume titled Private Participation in Health Services. Prior to joining the World Bank, Ms. Harding was a research fel- low in economic studies at the Brookings Institution. From 1987 to 1993, Ms. Harding was a fellow of the Russian and East European Center at the University of Pennsylvania, where she also received a Ph.D. in economics, with a concentration in comparative economic systems and public finance. The Conkributors Alexandre Abrantes is sector manager for human development in Africa at the World Bank. Born in Portugal, he trained as a medical doctor in Lisbon. He also received an M.Sc. in community medicine from the London School of Hygiene and Tropical Medicine and a Ph.D. in health policy and administration from the University of North Carolina at Chapel Hill. Dr. Abrantes started his career as a general practitioner in a rural county in Portugal but progressively moved into a career in public health and health services administra- tion. He also pursued an academic career, becoming associate pro- fessor of preventive medicine and public health at the Lisbon Med- ical School and the National School of Public Health. Before joining the Bank, he worked at the Commission of the European Union as About the Authors * 591 scientific officer in the advanced informatics in medicine and science programs. Hedi Achouri is director of the Hospitals Department at Tunisia's Ministry of Public Health. He implemented the hospital reform sup- port project, targeting teaching hospitals, and currently contributes to activities targeting regional hospitals. His professional career began in 1980 with a short period of medical activity in a maternal and infantile protection center, later as head of primary health care, then regional director of public health. During this period, he par- ticipated in the conception and the implementation of the primary health care policy and decentralization of the health system. Syed AJ-Junid is professor of health economics and head of the De- partment of Community Health, Faculty of Medicine, Universiti Kebangsaan Malaysia (the National University of Malaysia). After completing his M.D. training at the university in 1985, he continued his study and obtained an M.Sc. in public health from the National University of Singapore and a Ph.D. from London School of Hy- giene and Tropical Medicine. Currently, he is the chairman of the health economics group of the Academy of Medicine of Malaysia and president of the Public Health Specialist Association of Malaysia. Dr. Al-Junid is actively involved in a number of health economics and health financing research projects in Malaysia funded by local and international agencies. He serves in a number of committees in the Malaysia Ministry of Health, consulting on issues related to health economics and financing. In 2000-01, he served as a consult- ant for the WHO (WPRO) on situational analysis for a health pol- icy project involving a number of Pacific Island countries, including Fiji, Western Samoa, Kiribati, and Vanuatu. Ali Alkatiri is director of Vamed Engineering Asia. Previously he worked in a number of hospital-related positions in the Indonesian Ministry of Health, which he joined in 1983 on completing medical school at the University of Indonesia. Dr. Alkatiri earned an M.Sc. in public health from the National University of Singapore. In the In- 592 o Innovations in Health Service Delivery donesian Ministry of Health, he worked primarily in the Directorate General for Medical Services, ultimately becoming head of the Plan- ning and Evaluation Division, and then head of the Sub-Directorate of Health Maintenance. Dr. Alkatiri has published extensively on hospital-related and other topics regarding Indonesia. Yahaya Baba is director of the Health Education Division at the Ministry of Health in Malaysia. Simon Corden was the assistant commissioner at the Australian Gov- ernment's Productivity Commission when he wrote this chapter. He has an M.Sc. in economics from the London School of Economics; a Bachelor of Economics (Honors) from Monash University, Aus- tralia; and a Bachelor of Applied Science from the Royal Melbourne Institute of Technology. Willy De Geyndt is president of the InterHealth Institute, adjunct pro- fessor of health services management and policy at the George Wash- ington University, Washington, D.C., and senior vice president of MASHURA Health Care Consulting, Riyadh, Saudi Arabia. Previ- ously, he was principal public health specialist at the World Bank. Chris Ham is professor of health policy and management at the Uni- versity of Birmingham, England. Since March 2000, he has been on secondment to the Department of Health in London where he is currently director of the strategy unit. Originally trained as a politi- cal scientist, he has worked in the field of health policy and manage- ment for almost 30 years, both in the United Kingdom and interna- tionally. Mr. Ham's work focuses on the middle ground between theory and practice. He is the author of 18 books and numerous ar- ticles, papers, and reports. Loraine Hawkins is a senior health specialist in the Eastern Europe and Central Asia Human Development Unit of the World Bank. Currently she is on secondment to the Strategy Unit of the Depart- ment of Health in England. She has previously worked as a health economist and policy adviser in the New Zealand Treasury and De- partment of Health, and in the Treasury in the United Kingdom. About the Authors * 593 William C. Hsiao is the K. T. Li Professor of Economics at Harvard School of Public Health, and also directs the university's program in health care financing. Rozita Halina Tun Hussein is a member of the Health Systems Re- search Division, Public Health Institute, at the Ministry of Health in Malaysia. Melitta Jakab was a researcher in the Human Development Net- work at the World Bank while working on the Social Re project. Cur- rently she is completing a Ph.D. degree in health economics at Harvard University. Prior to her assignment to the Human Devel- opment Network, she had worked in the Bank's Resident Mission in Hungary as a research and operations assistant. She has an M.A. in health policy and management from the Harvard School of Public Health and a Bachelor's degree in economics and political science from McGill University. Eva 7arawan is a lead health specialist in the Human Development Group of the Africa Region at the World Bank. Since joining the World Bank 11 years ago, she has provided operational and analytical support to countries in the Middle East, Maghreb, and Sub-Saharan Africa in the areas of health sector development, hospital manage- ment, health financing, population and HIV/AIDS control, and pub- lic health. Before joining the Bank, Ms. Jarawan was an assistant pro- fessor of health management and policy at the American University of Beirut and at the University of Michigan. Her academic credits in- clude a Ph.D in Health Administration, a Master of Public Health, and a Master of Business Administration. Samuel S. Lieberman, currently lead specialist in the World Bank's East Asia and Pacific Region, Human Development Unit, recently returned from a four-year assignment in Jakarta as country sector co- ordinator for human development. Following graduate work in eco- nomics at Cambridge (M.A.) and Harvard (Ph.D.), Mr. Lieberman worked at Harvard on water resources in India, and at the Popula- tion Council as resource associate; he was eventually assigned to field 594 o Innovations in Health Service Delivery stints in India and Zambia. At the World Bank, Mr. Lieberman worked on poverty, employment, and social sector policies in India before joining the East Asia Region. Patricio V Marquez is a lead health specialist in the Latin America and Caribbean Office of the World Bank. Currently, Mr. Marquez is responsible for managing health sector projects in Mexico, Vene- zuela, Dominican Republic, and the Multi-Country HIV/AIDS Pre- vention and Control Adaptable Program Lending for the Caribbean Region, including projects in the Dominican Republic, Barbados, Grenada, Trinidad and Tobago, St. Kitts and Nevis, and Dominica. Previously, he managed World Bank-financed projects in Argentina, Colombia, Ecuador, and Paraguay. Prior to joining the World Bank in 1988, he worked for the Pan American Health Organizaion/World Health Organization, and the United Nations Fund for Population Activities. Mr. Marquez did his undergraduate work at George Wash- ington University and his graduate work at the Johns Hopkins Uni- versity School of Public Health, with additional course work done at the London School of Tropical Medicine and Harvard University. Lynne McKenzie is a director of Southern Cross International Ltd. and works as an international public sector consultant. She has held senior management positions in an area health board, a regional health authority, and the Health Funding Authority of New Zealand. Marta Molares-Halberg is a senior counsel in the World Bank's Legal Department. Soe Nyunt- U is a lecturer in the Departmnent of Social and Preven- tive Medicine at the University of Malaya. Mead Over is lead economist in the World Bank's Development Re- search Group. He joined the Bank in 1986 as a health economist after teaching health and development economics and econometrics at Williams College and Boston University. He was assigned by the About the Authors * 595 Bank, as part of a 1987 agreement with the WHO's Global Pro- gramme on AIDS, to do research on the economics of the AIDS epi- demic. MIr. Over has written several articles on the economic impact of AIDS and on the economics of prevention programs and has spo- ken on these topics at conferences and symposia around the world. He is coauthor, with Martha Ainsworth, formerly of the Bank's Development Research Group, of Confronting AIDS: Public Priorities in a Global Epidemic (first published in 1997 and revised in 1999). While maintaining his interest in the economics of the AIDS epi- demic, Mr. Over increasingly focuses on the evaluation of health sector policy intervention and reform in developing countries. Mr. Over received his Ph.D. in economics from the University of Wis- consin, Madison, in 1978. Kai Hong Phua is associate professor and head of the Health Ser- vices Research Unit, Department of Community, Occupational and Family Medicine, at the National University of Singapore; he is also senior adjunct fellow at the Institute of Policy Studies, Singapore. He worked previously as medical care admmnistrator in the New York State Department of Health, Office of Health Systems Management, and in Singapore's Ministry of Health. He is currently serving as chairman of the Technical Advisory Group for Health Sector De- velopment at the World Health Organization, Western Pacific Re- gional Office; he is also a member of the Health Commission of the International Federation of Red Cross and Red Crescent Societies and serves on the board of directors or councils of several health care organizations in Singapore. Dr. Phua has researched and published extensively on health sector reforms, health systems, health care financing, and population aging, with a focus on the newly industri- alizing countries of Asia. After premedical studies at Harvard Uni- versity, he did graduate work in health services administration and population studies at the Harvard School of Public Health and ob- tained a doctorate at the London School of Economics. Fernando Sacoto is a former undersecretary of the Ministry of Health in Ecuador. 596 0 Innovations m Health Service Delivery Graham Scott is chairman of Celtic Pacific Limited, Southern Cross International Limited, and Graham Scott (NZ) Limited. He is an af- fliliate of the Law & Economics Consulting Group in Emeryville, New Zealand, and works as an international public sector consultant. He served as chairman of a regional health authority (1995-97), chairman of the Health Funding Authority (1997-2000), and secre- tary to the New Zealand Treasury. Naoko Watanabe was a consultant in the Human Development Net- work at the World Bank in Washington, specializing in research of inequalities in health. She currently works in the World Bank's De- velopment Data Group. James Webster is manager of market management at Providian Na- tional Bank (United Kingdom). He previously consulted in the New Zealand health sector. Winnie C. Yip is associate professor of international health econom- ics and policy at the School of Public Health, Harvard University. She holds a Ph.D. degree in economics from the Massachusetts In- stitute of Technology. She was a coprinciple investigator of a study to evaluate the performance of Hong Kong's health care system and to design reform options. Her current major research interests are: fi- nancial incentives on provider behavior; international health system assessment and designs; economic development and health and well- being in developing countries; and dynamics and interactions be- tween public and private health care sectors in transition economies. Index Abedian, I., 133-134 Indonesia, 515 Abrantes, Alexandre, 239-261 managers and, 119, 121 access to services, 69-70, 91-92 network boards, 354-355, 356 Hong Kong, 413-414 New Zealand, 338 Indonesia, 515 NHI, 435 Malaysia, 445-446 NHS trusts, 275, 286, 294, 295 New Zealand, 332 organizational structure and, 45, Singapore, 464, 475-477 46-47 Tunisia, 496, 502, 503 Singapore, 460 Victoria, 380-381 tenn use, 117 Accident Rehabilitation to the Ministry of Health, 227 Compensation Insurance (ARCI), to patients, 227 306, 341n13 to stakeholders, 154 accident-related services, 306 transition econornies, 224-227, accountability, 7, 11, 13, 14, 95, 97, 229, 231 236n16 Tunisia, 491-492, 500 AHBs, 307, 308 Uruguay, 250, 252 alternative mechanisms for, accounting systems, 320 190-191 accreditation, 380 Argentina, 242, 244, 250 Achouri, H6di, 485-5 10 autonomization and, 53-54, acute care services, 345-346 76n33 access indicators, 359-360 CHEs, 308, 313-315 merging of providers of, Chile, 248-249, 250 363-366, 367 competition and, 51 Singapore, 452 corporatization and, 54-55 Victoria, 357-358, 363-366, HFA, 323 367, 378-379 Hong Kong, 14-15, 400-401, Administration of Public Health 404, 413, 420, 421 Services, Uruguay, 252 qo7 598 o Index administrators, 75n22 autonomy, 215-221, 228, 244, 250 admission rates autonomy index, 245 NHI, 439 readmussions, 380 Baba, Yahaya, 425-449 Singapore, 472-473, 477, backward integration, 56-57 480-481 Baltics, 223, 224 Tunisia, 503 Barnum, H., 126 advocacy groups, 371, 372 Bartlett, W., 283-284 The Age, 350, 377 beds agency theory, 30-31 Hong Kong, 392-393, 398, AHBs. See area health boards (AHBs) 413-414 AIDS. See HIV/AIDS Singapore, 459 Al-Jumd, Syed, 425-449 Swadana facilines, 522 Alford, R., 89 Tunisia, 503 Alkatiri, Ali, 511-531 See also occupancy rates allocative efficiency, 74n20, 130-132, benchmarking, 56-57, 76n35, 488 134, 160, 202 Bennett, S., 83, 88-89 Ecuador, 541-542 bilateral monopolies, 312, 316, Latn America, 253 341n17 New Zealand, 329-330 Bloor, K., 290 policy trade-offs and, 188 boards, 55, 272, 361, 372 Singapore, 474-475 accountability and, 119, ambulatory care, 429 225-226 Anglian Harbours Trust, dissolution AHBs, 307, 308, 311 of, 287 Argentina, 242, 243 area health boards (AHBs), 300-306 capital investments and, 355, changes to RHAs and CHEs, 384nl7 310-311 CHEs, 308, 313-314 overview of, 307 Ecuador, 536 powers pre- and post-1993 functions, 352-353 reform, 308-309 governance arrangements, technical efficiency, 324-329 349-356 trends m costs and deficits, Hong Kong, 397, 404-405, 325 409 Argentina, 11, 240-245, 255 legislative constraints applied to, assets, decision rights over, 218, 231 354 Australian Council on Health membership of, 55, 57, 353, Standards, 380 370, 377, 491 autonomization, 2-3, 11, 24, 53 New Zealand, 331, 338 Indonesia, 16-17 NHI, 434, 435, 449n8 management capacity and, 200 NHS trusts, 280-281, 294 New Zealand, 13 Singapore, 460 results outside health sector, Tunisia, 491, 498 64-65 Victoria, 347, 350-353, 361, Umted Kingdom, 12-13 383n3 autonomized organizations, 53-54 Bossert, T., 516, 517, 520, 522 Index * 599 Bradford, C. K., 123 criteria for ranking countries vs. budgets, 193, 211, 501 objectives, 172-173 Chile, 246, 247 design diversity, 171-172 demand-dnven, 245, 246 evaluation of reforms, 201-202 Ecuador, 541 implementation lessons, fee-for-service, 318 194-201 Hong Kong, 405 key features in implementation, Malaysia, 428 177-183 NSH trusts, 287 lessons learned from, 187-194 as residual claimant, 223 mutigatng strategies, 189-190, shadow, 247-248 192, 193-194 transition economies, 217 momtoring of reforms, 202-203 Tunisia, 504 problems in implementanon, Uruguay, 251 183-187 Victoria, 374-375 success in implementation, See also hard budgets; residual 172-177 claims; soft budgets trade-offs in health policy, bureaucracies and bureaucrats, 62-63, 187-189 80, 86-87, 393 See also names of individual Bureaucrats in Business, 107-108 countries business management strategy, 124, case-based payments, 213 126-127, 157 case-mix funding, 357, 358, 364, business plans, 193, 314 375-377, 379-380, 381 bypass fees, 131 catastrophic cases, 110, 146-147n12, 456, 478 Canada, 92 Central Eastern Europe. See transition capacity building, 96 economies capital investnents, 191, 309 Central Provident Fund (CPF), costs, 37 Singapore, 453, 455 decision rights over, 218-219 Chawla, M., 134, 136 Hong Kong, 399, 403, 420, checklists, 142 423nl5, 424n28 CHEs. See Crown Health Enterprises Indonesia, 520 (CHEs), New Zealand maintenance and development, chief executive officers (CEOs), 320, 343n30 Tunisia, 491-492, 499, 500 New Zealand, 311 Chile, 11, 109, 245-249, 255 NHI, 442-445 China, 108-109 NHS trusts vs. directly managed chronic illnesses, 426, 448n2 units, 276, 278, 279 civil service Victoria, 355, 384n17 Hong Kong, 403, 409, 410, 440 capitation, 311 Indonesia, 522 cardiac care, 432-443, 446, 447 Malaysia, 428, 435-436, care, continuity of, 188-189, 366 437-438 case studies transfers of personnel, 437-438 complexity effects, 190-192 Tunisia, 490 credibility issues, 192-193 clinical outcomes, 330-331 600 o Index clinical pathways, 127, 149n32 social functions and, 109-110, Clinton plan, 93, 94 146nn11-12 CNSS. See National Social Security for statewide and cross-network Fund (CNSS) services, 368 coalitions, 90, 93, 181 Tunisia, 493-494, 510n4 Chile, 255-256 within networks, 369 New Zealand, 338, 343n3 1, complementary reforms, 405-407 343n37 consumer price index, 326, 342n21 Uruguay, 255-256 consumer satisfaction, 125, 134, 135, commercial-in-confidence culture, 162 287, 294 complaint process, 416 communication exit surveys, 134, 135 Latn America, 256-258 Hong Kong, 414-416 New Zealand, 321, 322 Malaysia, 446 Tunisia, 495-497 New Zealand, 33 1-332 Community Care Act of 1990, U.K., Victoria, 371, 372 272, 273, 280 contestability, 292-293 community input, Victoria, 370-373 continuity of care, 188-189, 366 Commumty System for Integral contracting, 153, 212, 226-227, Health, Ecuador, 545-546 228-230 Companies Act, New Zealand, block contracts, 287 312-313 CHEs, 317-318 competition, 51, 65, 66, 67, 95, 269, competition and, 190 285 New Zealand, 323 admission rates and, 480-481 NSH trusts, 280, 282, 291 anticompettive practices, 68, 71 reforms of, 41 CHEs, 315-318, 342n18 Tunisia, 490 contracting arrangements and, control group approach, 137, 138, 140 190 copayments, 76n32, 221, 222 forces determnining level of, Chile, 247, 248 51 Ecuador, 537, 545 governance and, 59-60 Tumsia, 501, 502 Hong Kong, 398 Corden, Simon, 345-389 internal markets and, 50 Core Service Conunittee, New market environment and, 56-57 Zealand, 310 for market share, 367-368, corporatization, 2, 24, 54-55, 57, 83, 387n40 107-109 NHI, 440, 441 Argentina, 243, 245 NSH trusts and, 281, 287, challenges for, 438 294-295 collaborative networks and, 189 organizational reforms and, 61 funding of health care, 84 performance and, 107-108 governance in Hong Kong as reform design component, before and after, 406-407 171-172 impact evaluation, 9 replacing regulation, 73n8 incentives and, 95 Singapore, 463-464 Malaysia, 15, 430-432 Index * 601 management capacity and, 200 Crown Company Monitoring and mitigating strategies for Advisory Unit, New Zealand, 311, designing of, 193-194 315 National Heart Institute, Crown Health Enterprises (CHEs), 432-443 New Zealand, 13 New Zealand, 312 accountability, 313-315 results outside health sector, barriers to achieving reform 65-66 gains, 328-329 term use, 106 changes from AHBs to, 310-311 costs, 26, 231, 324-326, 357 decision rights, 313 as basis for pricing, 279-280 external environment, 315-318 containment of, 187-188 legal structure of, 312-313 cost-recovery ratios, 469-471 name use, 13, 306, 340nn3-4 Ecuador, 540-541 powers pre- and post-1993 escalation, 70-71 reforms, 308-309 Hong Kong, 403, 416-420, purchasing arrangements, 423nl6, 424nn26-27 314-315, 317-318 Malaysia, 428 restructuring of, 338-339 networks, 377-379 technical efficiency, 324-329 nonlabor component, 126 Csaha, T., 285 NSH trusts, 287 Czech Republc, 218, 220, 224, 226 one-day stay, 466 227 228 recovery of, 410-411 related to outputs, 326-327, data collection 343n37 Singapore, 464, 464-465, health care utilization, 134 466 patient activity in NSH trusts, Swadana facilities, 519 285 trusts, 283-285 productive efficiency and, 416 Tunisia, 494 prospective measurements, counterfactual, 137, 138 130-141, 143 CPF. See Central Provident Fund reliability of information, 507 (CPF), Singapore retrospective data and, 139-140, cream skimming, 57 143 critical care, 446 unit costs of in Victoria, 378 Croatia, 218, 219, 220, 222, 223 wages, 132 cross-section analysis, 138, 139 Davidson, K, 378 cross-subsidization, 47, 76n3 1, De Geyndt, Willy, 425-449 121-122, 148n23 debt, 155, 223, 224, 278, 319-320, from fee-paying patients, 436, 342n23 437 debt-equity ratios, 319 Indonesia, 516, 525 decentralization, 1-2, 10, 13-14, 40, internal, 49, 61, 69-70 230 Singapore, 460 character of, 212, 213-215 tiered pricing, 452 Ecuador, 535, 536-544 trusts and, 280 extent of, 91-92 602 0 Index decision rights, 46, 120, 152-153, 191, DHS. See Department of Human 231 Services (DHS), Victoria AHBs, 309 diagnostic-related group (DRG), 14, Argentina, 241, 244, 250 213, 357 assets and capital, 218-219 discharges, 327, 343-344n38 CHEs, 309, 313-315 disclosure of information, 337-338 Chile, 247, 250 disease management, 127, 149n32 Hong Kong, 399, 401, 421 district health boards, 338-339, human resources issues, 344n45 215-217, 437-438, 490 District Health Insurance House, Indonesia, 514 Romania, 220 inputs, 218-220 DRG. See diagnostic-related group internal incentive environment (DRG) for, 229 Duckett, S., 356 management, 220-221, 230n16, 245 Economic Planning Unit, Malaysia, NHII, 434-435, 437 443-445 NHS trusts, 275-277, 294, economics 295 agency theory and, 30-31 organizanonal structure and, 44, design of networks and, 374 45, 231 Ecuador, 538-544 pricing to purchasers, 220 Indonesia, 513 Singapore, 459 macroeconomic environment Tunisia, 489-490, 497 for reform, 83-84 Uruguay, 250, 251 neoclassical, 28-29, 147 decisionmaking, 95 of orgamzations, 29-30 AHBs, 311 role of state in, 26-28, 487 bundling of, 189 TCE, 31-32, 34 control of, 45 economies of scale, 56 Ecuador, 535-536, 543 Ecuador, 17 managers, 108 overview of health services, 533 politicization of, 61, 66 reform framework, 534-544, deconcentration, Ecuador, 535, 539 546-547nl delivery systems, 26 education programs, 256-258 demographics, Asian countries, efficiency trends, 253 426-427 Indonesia, 520 Department of Human Services Malaysia, 443-445 (DHS), Victoria, 347 United Kingdom, 267, 268 governance arrangements and, emergency services, 357, 359-360, 452 353-356 Emergency Services Enhancement performance monitoring, Program, Victoria, 357 358-360 employment, 215-217 staff conditions, 355-356 inequities of conditons, denvatve strategies, 423n23 409-410 developing countries, prospects for NHS trusts vs. directly managed reform in, 203-204 units, 275 Index * 603 See also civil service, human federal system vs. urntary system, resources 90-91 endowment, 200-201 fee schedules, 517, 541, 542 Enthoven, Alain, 269 fee-for-service arrangements, 318, EPSs. See government-owned health 410-411, 434-435, 436-437, 440 corporations (EPSs) Ferlie, E., 97-98, 266, 291 equity of access. See access to services finance systems, 123, 124, 152, 155 equity of services, 69-70, 91-92, AHBs, 311 134-136, 163-164, 502, 524 CHEs, 311, 318-321, 342nl9, Estonia, 218, 220 342nn21-23 evaluation of reforms, 8-9, 106, 113, external financing limits, 114 277-279 Ecuador, 544-546 Hong Kong, 403, 412-413 Hong Kong, 411-419, 420-422 Malaysia, 428-429 lessons learned from case networks, 353-354 studies, 201-203 NHI revenues and, 440-441 Malaysia, 443-448 NHS trusts vs. directly managed New Zealand, 324-332 units, 276 NSH trusts, 293-296 options for, 115-116 recommendations and private finance initiatives, 279, conclusions, 144-145 289-290 Singapore, 477-481 Singapore, 478-480 successful implementation, sustainability of, 254 172-177 trusts, 283-284 Tunisia, 497, 506-509 Tunisia, 491, 504-506, 507 Victoria, 377-383 Fondo de Salud, Ecuador, 537-538 evidence-based medicine, 291 forward integration, 56 excess capacity, 282 Foster, C. D., 266 exit surveys, 134, 135 fraud, 226 expenditures Free Medical Care System, Tunisia, Indonesia, 513, 517-520, 523 489, 502 NIHI, 441-442 fundholders, 270, 287, 291, 296 ratio to revenues, 443-445 funding, 44, 184, 269, 325, 334 Singapore, 470-471 accountability arrangements, Tunisia, 498, 499, 504, 505-506 76n33 extended care providers, 366, 367 case-mix, 357, 358, 364, external environment, 210-215 375-377, 379-380 CHEs, 315-318 DHS, 356-358 hospital behavior and, 47-5 1, external environment and, 113-117, 230 48-49 NSH trusts, 293 governance and, 67-68 payment systems and, 48-49 Hong Kong, 396, 397-398, 403, Tunisia, 493 422-423nll, 424n28 external financing limits (EFLs), Impact of monutoring on, 359 277-279 Indonesia, 512, 517-520 external incentives, 209-210, 230-231 market exposure and, 50-51 604 o Index funding (continued) budget deficits, 501 NIf, 439-440 decision rights and, 497 NHS trusts vs. directly managed management staff, 500 units, 275 market environment, 493-494 output-based, 345-347 nonsupport of reforms, 499 reforms, 39-40 payment systems, 507-508 Tunisia, 487, 493, 494 personnel management and, See also finance systems 490 quality of care, 503-504 Gaebler, T, 266 residual claim status and, 491 Gani, A., 516-517 Govindaraj, R., 134, 136 general practitioner (GP) fundholders, GP. See general practitioner (GP) 270, 287, 291, 296 fundholders Georgia, 220 gratuty payments, 221, 222 Ghana, 136 Griffiths Report, 267-269 Gilson, L., 177 Grosh, Margaret, 15 1n44 Glewwe, Paul, 151n44 Gyogyinfok, Hungary, 220 governance, 4, 11, 116, 230-231, 362 autonomization and, 53 HA. See Hospital Authority (HA), CHEs, 327 Hong Kong definition, 25, 43, 75n28 Ham, Chris, 79-105, 169-207, Ecuador, 543 265-297 external environment and, 48 Hamblin, R., 283, 285-286 funding and, 67-68 hard budgets, 54, 107, 109, 120, 153 Hong Kong, 393, 395, 406-407 lack of constraints, 440-442 hospital accountability and, procurement and, 126 225-226 residual claimant status and, 224 networks, 345 Harding, April, 23-78, 207-237 ownership and, 30 Harris, J., 132 of privatized nonprofits, 58 Hawkins, Loraine, 79-105, 169-207 public hospitals, 59-63 Health and Disabilities Service Act of sectoral, 75n28 1993, New Zealand, 312, 341n17 Singapore, 458-460 Health and Medicines Act of 1988, transition economies, 210, 211, 289 212 Health and Welfare Branch (HWB), Tunisia, 16, 486 Hong Kong, 393, 395, 402, 403, United Kingdom, 12-13 408, 423nl6 vertical integration and, 32 health care professionals, 87-88, Victoria, 349-356 180-181, 197 governance structure, 30-31, 116 See also human resources; Government Medical Store, Malaysia, personnel; physicians 430 Health Corporation of Singapore, government-owned health corpora- 458-459, 469, 481nl tions (EPSs), 488, 489-490, 492 Health Funding Authority (HFA), access to care in, 502 New Zealand, 306, 323, 325-326, autonomy of, 508 330, 332, 343nn30-31 Index * 605 Health Insurance Fund, Hungary, 219, hospital behavior 225 accountability and, 225-226 Health Issues Centre, Victorna, 370, determinants of, 209-210 372 external environment and, Health Service Agreements, Victoria, 47-51, 113-117 356-361 implications for, 230-232 Health Services Act of 1988, Victoria, informal payments and, 223 350, 352 measurnng response to, health-technology assessment, 291 122-127 Heart Transplant Fund, Malaysia, organizational structure and, 447-448 41-47 Heskett,J. L., 125 hospital capacity, 210-211, 226, 394 HFA. See Health Funding Authority hospital governing committees, 397 (HFA), New Zealand Hospital Medifund Committee, high-technology services, 477 Singapore, 480 HIV/AIDS, 368, 371, 426, 448n3 Hospital Reform Management holding companies, 182, 193, 458, 459 Committee, Tunisia, 495 Holland, W., 133 Hospitals and Charitable Institutions Hong Kong, 14-15 Act, Victoria, 351 assessment of impact of reforms, Hospitals and Charities Act of 1958, 411-419 Victoria, 351-352 criteria for ranking reform Hsiao, William C., 391-424 success against objectives, human resources, 124, 125-126, 153, 173-175, 177 156-157 external environment, 396-407 decision rights over, 215-217 hospital sector issues in the Ecuador, 543-544, 545-546 1980s, 394-395 excess, 231 hospital sector prior to reforms, feature of successful reform 392-393 design, 178 implementation of reforms, 177, impact of reform on, 412 178-183, 407-411 NHI, 437-438 overview of reform, 391-392 opponents to reform, 409 reform design, 171-172, 205n2, role in reform implementation, 396 181 Hong Kong Special Admmistrative Tunisia, 498-499, 503 Region, 402 See also civil service; Hospital Authority (HA), Hong Kong, employment; personnel 14, 15, 402 Hungary, 217, 219-220, 221, 223, 224, creation of, 391, 396, 408 225, 227 funding, 397-398 Hussein, Rozita Halina Tun, incentive regime, 398-401 425-449 market environment and, 398 HWB. See Health and Welfare Branch objectives of, 397 (HVWB), Hong Kong quality of care issues, 414-416 relations with individual iatrogenic diseases, 133, 162 hospitals, 398-401 Immergut, E., 93-94 606 ° Index implementation of reforms, 86-90, incentves, 54, 67, 347, 420, 421 197, 321, 336 autonomizaton and, 65 Argentina, 243-245, 256-257 corporatization and, 95 authonzing environment, Ecuador, 17 196-197 external, 209-210, 230-231 Chile, 243-245, 247-249, Hong Kong, 398-401 256-258 Indonesia, 522, 524 complexities of, 79-81, 190-192 internal environment and, cross-cutting analysis overview, 229 9-11 overview of, 5-6 decentralization and, 91-92 for reform implementation in effectiveness of, 98-100 U.K., 87 evaluation of reasons for success residual claims and, 35-36 in, 172-177 response to, 69-72 Hong Kong, 406-407, 407-411 service delivery and, 42-43 Indonesia, 177, 178-183, Tunisia, 498 183-186 vertical integration and, 32 issues arising in corporatization India, 136 of NHI, 437-443 Indonesia, 16-17 lessons from case studies, design of reform, 171-172, 194-201 205n2 Malaysia, 15 future of reform in, 528-530 managing of, 96-98 key features of implementation, market exposure of NHS trusts, 177, 178-183 281-283 overview of Swadana Initiative, mitigatmg strategies, 198, 199, 525-528 201 problems with implementation, New Zealand, 13-14, 312-322 183-184, 185-186 overview, 6-8 Swadana hospitals, 511-525 pace of change, 93-94, 97 infectious diseases, 146nll, 426 policy design link to, 336 influence activities, 32, 33 politcal context of NHS trusts, influence costs, 33 274-281 information asymmetries, 118, problems encountered, 131-132, 153 183-187 informanon strategy, Tumsia, 495-497 reform approaches and, 92-93 information systems, 3 7-38, 141-142, success in, 172-176, 177-183 200, 317 Swadana facilities, 515-517 infrastructure, 27, 28, 507 Tumsia, 495497 inpatient services, Hong Kong, United Kngdom, 12-13, 177, 416-420 178-183, 194 inputs Uruguay, 251-252, 258 allocative efficiency and, 160 values and, 86 Argentina, 241 Victona, 373-377 decision rights over, 218-220 See also policy making equity of services and, 163 in vitro fertilization (IVF), 463-464 Hong Kong, 402-403, 416-420 Index * 607 performance mdicators in Jarawan, Eva, 485-510 management contracts, Junta de Salud, Ecuador, 537-538 253-254 surveys and, 143 Kaplan, R., 123, 126 technical efficiency and, Kennett Government, Victoria, 345, 158-159 347-349, 373-375 institutional capacity, 25, 96, 183, Kenya, 136 199-200 Klein, R., 290 institutional innovations, 27, 65, 73n8 Korea, 109 institutions, 116 Kutzin,J., 126 definution, 25 Ecuador, 539-540 labor and labor markets, 6, 65, 119, environment for reform, 68 217 political, 93-94 Labour-Alliance coalition, New insurance, 257, 305 Zealand, 338, 343n31, 343n37 Argentina, 246 Langenbrunner,J., 212 Chile, 248 Latin America, 255-256 income-protection, 306 Le Grand, J., 283-284 insured patients, 241, 248, leadership, 180 260n3 legal issues, 214, 534-536 Malaysia, 428 legislation, 255, 257 Medishield, 456 board member issues, 354 NIS trusts vs. directly managed Ecuador, 534-536 units, 276 health services payments to hospitals, 218 recommendations, 375 for public patients, 241, 260n3 Malaysia, 428-429 uninsured, 240, 260nl New Zealand, 307, 312-313, insurance market, 110, 146-147n12 314, 341n17 integration, backward, 56-57 United Kingdom, 272, 273, 280, integration, forward, 56 289, 297-301 interest groups, 89-90, 106, 197 Victoria, 351-352, 370 Intenrm Phase I Repon, 350, 366, 375 length of hospital stay internal allocation, 132 average cost, 466 internal efficiency, 253 Hong Kong, 418, 420, 424n27 interventions, 112-113 Indonesia, 514 characterization of, 117-122 New Zealand, 327 definition, 25 Singapore, 474 external environment and, Lewis, M., 222 113-117 Lieberman, Samuel S., 511-531 interviews, with health system life expectancy, Malaysia, 426, 427 workers, 307, 340n5 line-item allocations, 211 lVF. See in vitro fertilization (IVF) Lithuania, 220 Living Standards Measurement Survey Jakab, Melitta, 207-237 program, 142-143 Jamaican Survey of Living Conditions, local governments, 213, 218, 225, 143, 151n44 230-231 608 ° Index Machiavelli, N., 85 implementanon arrangements McKenzie, Lynne, 305-344 and, 96-98 majority governments vs. coalitions, influence on policy 90 implementation, 87-88 Malaysia, 15 political pressure and, 217 context and background of privatizanon and, 58 health services, 426-432 See also management corporatization of the NHI, market environment, 6, 51, 68, 115, 432-443 436 criteria for ranking reform Hong Kong, 396-398 success against objectives, structure and competition, 173-175, 177 56-57 design of reform, 171 Tunisia, 493-494 key features of reform Victoria, 362-369 implementation, 177, market exposure, 118-119, 120, 153, 178-183 231, 308 overview of reform, 425-426 Argentina, 241-242, 244, 250 policy trade-offs, 188 case studies and, 174 Malaysia Incorporated, 431 Chile, 247-248, 250 Malik, R., 516, 517, 520 competinon between networks managed markets, 286 for, 367-368, 387n40 management, 6, 126, 248, 257, 307, external environment and, 366 50-51 Argentina, 240-241, 243 Hong Kong, 400, 404 autonomy of, 2-3 Indonesia, 515 best practce techniques, 38-39 NHI, 435-436 CHEs, 313, 327 NHS trusts, 281-283, 294, 296 decision rights over, 220-221, organizational structure and, 43, 230n16 44,45-46 deconcentration of, 535, 539 Singapore, 459-460 Hong Kong, 405-407, 421 transition economies, 221-223, new public, 2, 41, 83, 266 228, 229 NHS trusts, 267, 275, 292 Tunisia, 492 reforms, 38-39, 40, 75n22 Uruguay, 247-248, 250, 251 relationship with governments, market failures, 28-29 30 market structure, 68, 315-316 selected indicators, 155-158 marketing, 123-125, 156 support for reforms, 7 marketizing reforms, 2 trusts, 271-272, 286 overview, 9-11 Tunisia, 489, 490, 499-500, privatization and, 57-58, 63-64 507 Singapore, 15-16 management capacity, 88, 199-201 Marmor, T. R., 80-81 management information system Marquez, Patricio V., 533-547 (MIS), 37-38, 488-489 Maynard, A., 290 managers, 54, 108, 120 medical committees, Tunisia, 492 accountability of, 117, 119, 121 medical equipment, 37-38 Index * 609 Medical and Health Department lessons learned from case (MHD), Hong Kong, 393, 395, studies, 201-203 408 Tunisia, 497 medical management strategy, 124, Victona, 358-360 127, 157 monopolies, 27-28, 73n8, 316-317 medical profession, 92, 94, 496 bilateral, 312, 316, 341n17 medical savings accounts, Singapore, cardiac care, 436 16, 455-457 geographic, 56, 315, 341n16 Medifund, 455, 456-457, 460 recognition of problems with, Medisave, 453-456, 465, 475, 476, 336-337 478-479 Moore, M., 88, 96-97 Medishield, 455, 456, 476, 477, moral hazard, 71, 131 478-479 MPH. See Ministry of Public Health mergers, 282 (MPH) acute care providers, 363-364 Municipal Integrated Health Care of acute and nonacute service Networks, Ecuador, 536 providers, 364-366, 367 CHEs, 315, 341n]15 National Business Review, 331-332 Victoria, 363-364, 375, 376, National Health Fund, Chile, 247 379 National Health Plan (NHP), MHD. See Medical and Health Singapore, 453-455 Department (MHD), Hong Kong National Health Service and Millstein, Ira, 107, 109, 110 Community Care Act of 1990, Ministry of Health (MOH), Chile, 297-301 247-249 National Health Service (NHS), Ministry of Health (MOH), Indonesia, United Kmgdom, 171, 267-269 512, 516-517, 529 National Health System, Ecuador, 534 Mlnistry of Health (MOH), Malaysia, National Heart Institute (NHI), 426-427, 428 Malaysia, 425, 432-443 Ministry of Public Health (MPH), National Social Security Fund Ecuador, 533, 534, 540 (CNSS), 493 Ministry of Public Health (MPH), National Trades Union Congress, Tunisia, 485, 491, 493, 508-509 Singapore, 462-463 Mintzberg, H., 97 neoclassical economics, 28-29, 147n18 MIS. See management information Netherlands, 90, 93 system (MIS) networks, 182 Model Facility Survey, 143 community input in, 370-373 Modernization Council (CONAM), competition for market share, Ecuador, 535 367-368, 387n40 MODERSA project, 534, 535, creation of, 375-376 536-544 Ecuador, 536, 538 Molares-Halberg, Marta, 533-547 governance arrangements, monitorng mechanisms, 106 350-359 AHBs, 308 hospital closings and, 363-364 CHEs, 308, 311 management of, 366 Hong Kong, 400 objectives of, 377 610 o Index networks (continued) power of, 272-273 role of economncs in design of, private finance initiatives, 279, 374-375 289-290 statewide and cross-network structure of, 271 services, 368 nondistribution constraints, 77n46 Victoria, 345-347, 383n2 nongovernmental organizations New Economic Policy, Malaysia, 431 (NGOs), 306, 512 New Management Initiative, Hong nonprofit hospitals, 393, 394-395, 409 Kong, 405-407 nonprofit orgamzations, 58, 71-72, new pubhc management, 2, 41, 83, 76n36, 77n46 266 Norton, D., 123, 126 New Zealand, 13-14, 92, 109, Nyunt-U, Soe, 425-449 387-388n52 criteria for reform success vs. occupancy rates objectives, 176-177 fee rates and, 517, 518 design of reform, 171 Hong Kong, 413-414 health care spending, 305-306 Indonesia, 514, 520, 524-525 implementation of reforms, NHI, 439 13-14, 195, 312-322 Singapore, 452, 475, 476 key features of reform Tunisia, 503 implementation, 177, See also beds 178-183 offshore business developments, 311, lessons drawn from reform in, 341n9 332-338 Organisation for Economic policy trade-offs, 188 Co-operation and Development problems with implementation, (OECD), 290 184, 185, 186 organizational behavior, 122-123 transition costs, 197 organizational reforms, 2, 3-4, 24, 40, NHI. See National Heart Institute 61, 106 (NHI), Malaysia conceptual framework of, 5-6 NHP. See National Health Plan desired effects of, 8-9 (NHP), Singapore dimensions of, 117-122 NIHS trusts, 12 evaluation overview, 113, 114 assessment of, 293-296 influence activities and, 33 compared to directly managed Latin America, 255-256 units, 275-276 New Zealand, 324-332 design of reform, 269-274 political choice theory and, dissolution of, 287-288 36-37 dynamics of, 286-290 poliucal decisions and, 191 EFLs, 277-279 progress of, 107-108 legislation, 297-301 residual rights and, 35-36 lessons learned from reform selected indicators of, 152-154 experience, 290-293 social functions and, 110-111 performance of, 283-286 Tunisia, 489-494, 514 political context of organizational structure, 209-210, implementation, 274-281 211, 212, 215-230, 231 Index * 611 orgamzations private, 28-29 definition, 25 public, 28-29, 266 economics of, 29-30 residual rights and, 34-36 factors influencing behavior of, transfer of, 213, 218-219 49 governance and, 60-62 pareto optimal performance, 117, internal, 31 147n18 private, 38-39 Parliamentary Opposition, Victoria, vertically integrated, 32 354, 376 Osborne, D., 266 patient records, 133 out-of-pocket payments, 221, 223, Pauly, M., 76n31 446, 498 payment systems, 6, 130-131, 213, Outline Perspective plans, Malaysia, 224, 226 430, 431 copayments, 76n32, 221, 222, outpatient services, 416-420, 439 247, 248, 501, 502, 537, 545 output-based systems, 224, 345-347 external environment and, outputs 48-49 allocative efficiency and, fee-paying patients, 318, 130-132 436-437,440 Argentina, 241 formal, 221, 222, 223 capital investments and, 355 Hong Kong, 397-3 98 decision rights over, 219-220 informal, 222, 223, 228 Health Service Agreements, market exposure and, 221-223 356-361 options for, 115-116 HFA, 323 out-of-pocket, 221, 223, 446, Hong Kong, 402, 416-420, 498 424nn26-27 performance-based, 212, 213 performance indicators in provider payment, 14, 41, 179, management contracts, 213,218,309,356-361 253-254 reforms of, 39-40 related to costs, 326-327, revenues and, 212-213 343n37 Tunisia, 507-508 surveys and, 143 performance, 113, 308 technical efficiency and, agreements, 536 128-130 funding and, 375 Over, Mead, 105-165 hospital capacity and, 210-2 11 oversight, 116-117, 293, 350-353, implications for, 230-232 435 improvement of, 106-107 ownership, 25, 214, 230 measuring the impact of accountability arrangements, reforms on, 127-135, 76n33, 225-226 149n33, 150n34 Argentina, 260n2 NSH trusts, 283-286, 294 CHEs, 311, 341nlO rewards and sanctions linked to, NHI, 433-434, 436-437, 443 360-361 NHS trusts vs. directly managed social functions and, 111 units, 276 performance budgeting, 50 612 o Index performance contracts, 64, 108-109, social functions and, 228 489-490, 492 wages, 545 performance indicators Pierson, P., 85, 89 Argentna, 253-254 Plowden, F.J., 266 Chile, 253-254 Poland, 217, 220 Hong Kong, 405 policy design, 194, 336 NHI, 444 policymaking, 92, 180 Tunisia, 502-503 content of, 99 Uruguay, 253-254 framework for policy analysis, personnel, 410, 441, 489 81, 82 accountability of, 117, 119, 121 institutional capacity to benefits link to pricing, 462-463 implement change, 96 concerns of dunng interest groups and, 89-90 implementation, 461-462 internal consistency and Hong Kong, 399, 402-403, coherence of, 95-96 408-410, 423nl6 pohticians and bureaucrats and, Indonesia, 522 86-87 Malaysia, 428 politics of, 81 NHS trusts vs. directly managed public sector health workers units, 275, 280 and, 89 payment for health care services, role of managers and health care 436 professionals in, 87-88 physician/nonphysician ratios, role of public and patients, 472 88-89 physicians' pay, 465-467 specificity of objectives, 94 retention of, 433, 447 trade-offs in, 187-189 Singapore, 468 political choice theory, 36-37 training of, 447 political economy, 6-8, 81-82, 95-96, Tunisia, 490, 492, 498-499 460-463 Victoria, 355-356 political systems, 90, 93-94 views of reform, 494 politics, 84, 191, 217, 336, 495 See also civil service; human case-mix funding and, 376 resources decision making and, 61, 66 PFIs. See pnvate finance initiatives differential charges and, (PFIs) 464-465, 466 pharmaceuticals, 132-133, 134, 428 implementation and, 86-87, Phua, Kai Hong, 451-483 186, 196-197,198-199 physical capacity, 231 Latin America, 255-256 physicians management processes and, 221 decision rights on employment, market exposure of NHS trusts, 215-218 281 nonphysician ratios, 472 network boards and, 354 number of, 26, 27n2 New Zealand, 321-322, 343n31 payments, 465-467 NHS trusts, 274-281 resistance to change, 328 privatization costs and, 105-106 Index * 613 public sector health workers results outside health sector, and, 89 66-67 Victona, 273-374 of utilities, 266 politics of change, 336 procurement, 124, 126, 152, 157 poor patients decision rights over, 219 Argentina, 240 Hong Kong, 403 equity of care and, 134-135, Tunisia, 490 151n42 productive efficiency, Hong Kong, Indonesia, 513 416-420, 424nn26-27 Malaysia, 436 productivity, 35, 108 pricing policies, 479-480 competition and, 285 Singapore, 460 improvement of, 39 subsidies and, 447 Malaysia, 443-445 population-based information, 144 technical efficiency and, Preker, Alexander S., 23-78, 207-237 128-130, 149n33, 150n34 pricing total factor, 111 based on costs, 279-380 Tunisia, 503 decision rights over, 220 U.K. hospitals, 270 fee-for-service, 318 profit maximizers, 28, 29 Hong Kong, 397 profits Indonesia, 516 cream skimming and, 57 link to benefits, 462-463 as motivating force, 70-71 NES trusts vs. directly managed privatization and, 58, 63-64 units, 276, 279-280 property management, 490 policies, 130-131, 135 property rights theory, 34, 36 regulation of, 461 prospective measurements, 130-141, regulation of, 461 143 Shnadowe, 4521 404645, provider payment mechanisms, 41, sapo, 45460 179, 213, 218 466, 479-480 AHBs and CHEs, 309 under HFA, 323 performance-based, 14 private finance initiatives (PFIs), 279, prctoria, 356-361 289-290 See also payment systems private good, 122, 146nl11pbcdlltain 8 Private Healthcare Facilities and public administration, 180 public choice theory, 36-37 Services Bill, Malaysia, 428-429 Public Finance Act, New Zealand, 314 private hospitals, 393, 529 public goods, 110, 119 private patients, NHI and, 439-440 public health sector workers, 89 private sector organizations, 91-92, public hospitals 428-429 best practice management privatization, 2, 24, 41, 57-58, 63-64, techniques in, 38-39 75n26 governance of, 59-63 Argentina, 255 Indonesia, 511-514 Malaysia, 430-432 modernization of, 539 political costs of, 105-106 Victoria, 345, 369, 383nl 614 o Index Public Hospitals Policy and Funding reform, 7, 68, 84, 96 Guidelines, 1999-2000, Victoria, alternative measurement 354-355 instruments, 131-144 public opinion, Tunisia, 496-497 approaches to, 8, 18-19, 91, public sector, 179-180 92-95, 321, 376 influence on policy capacity to implement change, implementation, 88-89 233-234 organizations, 430-431 complementary, 69-72, reforms and, 32, 34 405-407, 489 public utlities, 27 consistent with previously Pugner, Klaus, 4 introduced reforms, 232-233 purchaser funding, 318, 342n19 dominant values and, 85-86 purchasing arrangements, 116, 294 external consistency of, 67-68 CHEs, 314-315, 317-318 future of, 528-530 competition for service delivery governance and, 60-62 and, 315-316, 342nl8 health care professionals' role in internal markets, 50 in, 87-88 New Zealand, 310-311, 340n8 history of and commitment to, NT-IS trusts, 270, 287, 289 82-83 RHAs, 307, 310, 314-315, institutional environment and, 317-318 68, 85 transition economies, 209, 210, interests groups' role in, 89-90 211, 212 internal consistency of, 67 under NSH trusts, 270 macroeconomic environment, 83-84 quality improvement strategies, managers and health care 365 professionals' role in, 87-88 quality of services, 130, 149n33, measuring impact of, 127-13 5, 161-162, 254 149n33, 150n34 cost escalation and, 70-71 need for, 24, 26 Hong Kong, 414-416 objectves and timelines of, hospital performance and, 347-349 132-134, 150n34 parallel reforms, 18-19 Indonesia, 520, 522 politicians and bureaucrats role Malaysia, 446 in, 86-87 NSH trusts, 285-286 prospects for in developing and policy trade-offs and, 188 transition countries, 203-204 Singapore, 477 public and patients' role in, Tunisia, 488-489, 503-504 88-89 Victoria, 360, 366, 379-380 public sector health workers' role in, 89 randomized trials, 130-141 Singapore, 457-460 Ranson, S., 266 sustainability of, 504-506 Reagan, Ronald, 85 Tunisia, 488-489 recessions, 83-84 underlying pressures for, reflexive comparison, 137, 138 429-430 Index * 615 views of stakeholders in, 494 organizational structure and, 44, See also evaluation of reforms; 45 implementation of reforms; Singapore, 459 names of specific countries; Uruguay, 248, 250, 252 specific reform issues, i.e , residual rights, 34-36 governance; market exposure resource allocation, 74n20, 132, 291 reform design Hong Kong, 402 experimental, 137-141 influence costs and, 33 Hong Kong, 396 New Zealand, 310-311, 330 Indonesia, 514-515 Victoria, 363, 384-385nl7 lessons learned from case retrospective measurements, 138, studies, 187-194 139-140, 143 NHS trusts, 269-274 revenue rights, 218 Tunisia, 497-498 revenues, 53, 223, 224 regional health admunistration (RHA), AHBs, 324-325 247-249, 307 Argentina, 243, 245 changes from AHBs to, 310-311 CHEs, 314, 317, 325-327, purchasing arrangements, 341nl3 314-318 Chile, 248 technical efficiency of, 325-329 Indonesia, 514 regular information systems, 141-142 market exposure and, 221 Reich, M., 81, 99 NHI, 440-441 research payment systems and, 212-213 conceptual framework of, ratio to expenditures, 443-445 112-113, 114, 170 RHA, 319 cross-cutting analysis of, 9-11 Singapore, 470-471 future investigation areas, 4-5 Swadana facilities, 520, 521 scope of, 111-112 Tunisia, 493 residual claimant status, 119, 153, 178, See also residual claims 193 RHA. See regional health AHBs, 308 admunistration (RHA) Argentina, 242, 244 risks CHEs, 308 finance systems and, 115-116 HIong Kong, 399-400 market exposure and, 221 imposition of, 120-121 shifting of, 50, 70 NHI, 435-436, 442-443 Tumsia, 506 transition economies, 223-224, Romania, 217, 219-220, 221, 228 228, 229 rules, definition, 25 Tunisia, 491 rural health services, Malaysia, residual claims, 35, 46, 54, 58 426-427, 429 Argentna, 250 budgetary units and, 52, 76n32 Sacoto, Fernando, 533-547 Chile, 248, 250 savings and loan industry, 35 Hong Kong, 404 Schedule of Procedures, Czech Indonesia, 514 Republic, 220 NHS trusts, 295 Scott, Graham, 305-344 616 o Index Scott Report, 396, 408 Singapore, 460 self-managed public hospitals transition economies, 227-228, (SMPHs), 241-244, 257 229 service agreements, 354, 384nl5 trusts and, 287, 296 service-profit cham, 125 Tumsia, 492-493, 508 shadow budget system, 247-248 Uruguay, 250, 252 shadow prices, 541 social insurance, 83, 212, 213, 215, side payments, 135, 162 230, 240, 478-479 Singapore, 15-16 social services, 28, 61-62 context and background of sociocconomics health services, 452 Asian countries, 425-427 critena for ranking reform Ecuador, 542 success against objectives, equity of care and, 134-135 173-175, 177 health care and, 436-437 design of reform, 171 Soderlund, N., 284-285 evaluation of reforms, 477-481 SOEs. See state-owned enterprises impact of reforms, 467-477 (SOEs) implementation successes and soft budgets, 6 failures, 195 Soviet Union, former. See transition institutional reforms and, economues 453-467 specialty care, 381, 415, 452, 467 key features of reform state implementation, 177, debate on role in reform, 84 178-183 political choice theory, 36-37 overview of reform, 451 role in economy, 26-28, 487 timelne of reform, 454 role of in Sweden, 91 Smee, C., 285, 286 state-owned enterpnses (SOEs), SMPHS. See self-managed public 107-110, 193, 310 hospitals (SMPHs) state-owned industries, 266 social functions, 52, 53, 95, 119, 154 statistical approach, 138 AHBs and CHEs, 309 stewardship, transiton economies, allocative efficiency and, 209, 210, 211, 212 130-132 Stewart, J., 266 Argentina, 242-243, 243, 244, strategic alliances, service delivery and, 250 362-363 Chile, 249, 250 strategic management, 88, 152, 353 corporatization and, 55 capital investment decisions and, definig and protection of, 355, 384n17 109-111,121-122, HongKong, 401-402, 423nl4 146nnl1-12, 148n22 subsidies, 55, 447 external environment and, 49 Ecuador, 537 Hong Kong, 405 Indonesia, 517-520 NIB, 436-437 Singapore, 461, 474, 475, 479 organizational structure and, 45, Tunisia, 498, 520 47 supervision mechanisms, 191, 193 Index * 617 surgery, 316 prospects for reform in, Ecuador, 541 203-204 fees for, 467 residual claimant status, Malaysia, 438-439, 440 223-224, 228, 229 New Zealand, 327, 332, 342n28 social functions, 227 Victoria, 359-360, 381 term use, 234nl surpluses, 276 transparency, 219, 221, 377 surveys, 134, 135, 142-143 trusts. See NIHS trusts Swadana hospitals, 511-528 Tunisia, 16 Sweden, 91, 93, 109 changing role of the state, 487 Swerissen, H., 356 context and background of health care services, 486-487 technical capacity, 182 criteria for ranking reform technical efficiency, 128-130, 149n33, success against objectives, 150n34, 202 173-175, 177 New Zealand, 324-329 design of reform, 171-172, policy trade-offs and, 188 205n5, 497-498 Singapore, 469-471 evaluation of refonm, 506-509 Tunisia, 488-489 impact of reforms, 502-504 Victoria, 377-378 implementation of reforms, 177, teclnological reforms, 37-38 178-183, 495-497 Tena Canton, Ecuador, 536-538 organizational reform, 489-490 tertiary acute care facilities, 346 overview of reform, 485-486 Thatcher, Margaret, 85, 265, 267, 268 position of stakeholders in third-party payments, 116, 130-131 reform, 494 Thong, K. L., 395 reform package, 487-489 Tiscomnia, J. F, 123 sustainability of reforms, Tomlinson, Bernard, 282 504-506 total factor productivity, 111 Tuohy, C. H., 92, 94 total purchasing projects, 270 trade unions, 280, 462-463 unitary system, 90-91 transaction cost economics, 31-32, 34 United Kingdom, 12-13, 109 transition costs, 197 cnteria for reform success vs. transition economies, 10 objectives, 175-176, 177 capacity to implement change, design of reform, 171, 194 233-234 dynamics of health care reform, changes in hospitals, 207-209 92 decision rights on labor input, financial incentives offered for 215-217 reforms, 87 external environment of, implementation and, 97-98, 210-215 177, 178-183, 185, 186 internal incenuve environment lessons learned from reform during, 229 experience, 290-293 market exposure, 221-223, 228, pace of change, 93 229 policy trade-offs, 188 618 o Index United Kingdom (continued) policy changes, 347-349 power of majority governments, policy trade-offs, 188, 189 90 provider payment system, transition costs, 197 356-361 See also NHS trusts Victorian Commission of Audit, United States, 90-91, 92, 93, 94 373-374 urban health services, Malaysia, 429, 431 wages, 132, 179, 465-467, 545 Uruguay, 249-252, 255, 258 waiting time, 415 U.S. Agency for International Devel- Walsh, K, 267 opment (USAID), 516, 517, 520 Walt, Gill, 10, 81, 177, 195 user fees, 410-411 Watanabe, Naoko, 105-165 utilization patterns, 438-439 Webster, James, 305-344 welfare, 110-111 values, 85-86, 197 welfare reform, 85, 89 vertically integrated organizations, 32 Wellmgton, Clive, 361 veto power, 93-94 Western Massachusetts Health Victoria, Australia, 13, 14 Planning Council, 147n17 assessment of reforms in, Wiley, M., 212 377-381 Wilson, James Q., 45 criteria for ranking reform Working for Patients, 268, 269, success against objectives, 270-272, 273, 274-281 173-175, 177 World Bank, Hospital Reform design of reform, 171 Projects, 165 governance arrangements, World Health Orgamzation (WHO), 349-356 428 implementauon of reform, 177, 178-183, 373-377 Yip Wmnie C. 391-424 market environment, 362-369 i W overview of health care system, 345-347 Zimbabwe, 136 Selected titles from the World Bank's Health, Nutrition, and Population Series: Better Health Systents for India's Poor: Findings, Analysis, and Options. 2002 (D. lHl. Peters, A. S. Yazbeck, R. R. Sharma, G. N. V. Ramana, L. H. Pritchett, and A. Wagstaff The Burden of Disease aniong the Global Poor: Czutrrent Situation, Futl/r-c Trends, and Impipcationzsfor Strategy. 1999 (D. R. Gwatkmi and M. G iillot) Health Expenditures, Sei-vices, and Ourconies in Afi-ica: Basic Data and Cross- National Co7mparisons, 1990-1996. 1999 (D. H-l. Peters, K. Kandola, A. E. Elmendorf, and CJ. Chellaraj) Health, Nutrition, and Population Indicators A Statistical Handbook. 1998 (E. Bos, V. I-Hon, A. M4aeda, G. Chellaraj, and A. S. Preker) Improving Wonien's H-eIalth in Pakistan. 1998 (A. G. Tinker) Mea curing Countiy lerfornian7ce on Heilth: Selected Indicatorsfor 115 Countit-1es. 1999 (T. Wang, D. T. Jamison, E. Bos, A. S. Preker, and J. Peabody) Private Par-ticipation in Healtlh Services. 2003 (A. Harding and A. S. Preker, editors) Prospectsfor Imnproving Nutrition in Eastern Europe and Central Asia. 2002 (C. Rokx, R. Galloway, and L. Brown) Reproductive Health in the Middle East an7d North Anierica. Wl"ell-Belizgfor All. 2001 (A. Aoyama) Social Re-Insurance. A Ne7v Approach to Szustainable Coniniunity lealth Financing. 2002 (D. Dror and A.S. Preker, editors) Toward a Virtuous Circle: A Nutrition Review of the Middle East and North Afi-ica. 1999 (A. Aoyama) - s iv@soo!e14494 ~~~9 7881 341 ! E ~~~~~~~~~~~~~~