Report No. 43183-XK Kosovo Health Financing Reform Study May 6, 2008 Human Development Sector Unit Europe and Central Asia Document of the World Bank KOSOVO HEALTH FINANCING REFORM STUDY (ESW) Table of Contents ACKNOWLEDGEMENTS ..............................................................................................................i EXECUTIVE SUMMARY............................................................................................................. ii CHAPTER 1. INTRODUCTION..................................................................................................1 CHAPTER 2. HEALTH FINANCING IN CONTEXT ................................................................8 2.1 Macroeconomic And Fiscal Situation ................................................................... 8 2.2 Health Outcomes................................................................................................. 10 2.3 Delivery Of Care And Treatment Seeking Behavior .......................................... 10 CHAPTER 3. GOVERNANCE AND INSTITUTIONAL ARRANGEMENT FOR HEALTH INSURANCE.......................................................................................................15 3.1 Definition and Scope of Governance .................................................................. 15 3.2 Governance in Kosovo Today............................................................................. 16 3.3 Specific Governance Issues in Health Financing ................................................ 16 3.3.1 The importance of governance in health services........................................................ 16 3.3.2 The health financing system and governance.............................................................. 18 3.3.3 Health insurance models and governance.................................................................... 19 3.3.4 The search for a good governance structure for a health insurance system................. 19 3.3.5 Principles of good governance..................................................................................... 20 3.3.6 Main institutional elements of health insurance .......................................................... 21 3.3.7 Ownership and legal status .......................................................................................... 22 3.3.8 Mechanism of representation....................................................................................... 25 3.3.9 Fiduciary and human resource arrangement................................................................ 26 3.3.10 Forms and scope of governmental supervision.......................................................... 27 3.3.11 Information requirement............................................................................................ 29 3.3.12 External factors affecting the governance performance of insurance........................ 32 3.4 Lessons from the Kosovo Pension Pillar II Trust Fund ...................................... 32 3.4.1 Ownership and legal status .......................................................................................... 33 3.4.2 Mechanism of representation....................................................................................... 33 3.4.3 Fiduciary and other administrative arrangements........................................................ 33 3.4.4 Forms and scope of government supervision............................................................... 34 3.4.5 Information requirement.............................................................................................. 35 3.5 Conclusions and Recommendations.................................................................... 36 CHAPTER 4. SOURCES OF HEALTH FINANCING AND REVENUE COLLECTION.......39 4.1 Overview ............................................................................................................. 39 4.2 Kosovo: A Current Analysis ............................................................................... 41 4.2.1 Total health care expenditures ..................................................................................... 41 4.2.2 Out-of-pocket payments .............................................................................................. 44 4.2.3 General government financing..................................................................................... 46 4.2.4 Comparing Kosovo public expenditures with ECA countries..................................... 49 4.2.5 Lessons from the EU and OECD countries ................................................................. 49 4.3 Regional Lessons to Raise Public Revenues for Health in Kosovo .................... 52 4.3.1 Dedicated taxes............................................................................................................ 52 4.3.2 General tax revenues.................................................................................................... 53 4.3.3 Impacts to date: Challenges for enhancing revenues................................................... 54 4.3.4 Why did the use of the dedicated (payroll) tax fail?.................................................... 56 4.4 Implication for Kosovo ....................................................................................... 60 4.4.1 SHI: Is it a good idea for Kosovo? .............................................................................. 60 4.4.2 Economic analysis: Estimates of financial results under payroll tax........................... 61 4.4.3 Balancing revenues and expenditure scenarios under SHI.......................................... 62 4.5 Alternative Options for Increased Funding......................................................... 64 4.5.1 Option 1: Develop a health financing actuarial model ................................................ 64 4.5.2 Option 2: Consider dedicated taxes for health............................................................. 65 4.5.3 Option 3: Set medium-term expenditure targets to increase general revenues............ 66 4.5.4 Option 4: Encourage private VHI................................................................................ 68 CHAPTER 5. ORGANIZATIONAL FORM OF RISK POOLING ...........................................72 5.1 Overview and Definitions ................................................................................... 72 5.2 Risk-pooling Structure: Single or Multiple Risk Pools....................................... 72 5.2.1 Single risk pools .......................................................................................................... 72 5.2.2 Multiple risk pools with and without competition....................................................... 73 5.2.3 Multiple insurance and risk-equalization transfers...................................................... 74 5.2.4 Lessons from fragmented risk pooling in Bosnia and Herzegovina ............................ 75 5.2.5 Multiple risk pools under fiscal decentralization......................................................... 76 5.3 Current Form of Risk Pooling in Kosovo ........................................................... 77 5.4 Concerns About Imperfect Information in Risk Pooling .................................... 77 5.5 Concerns About Inequity in Access to Care........................................................ 79 5.6 The Financial Sustainability of Health Insurance ............................................... 80 5.7 Establishing an Insurance Fund........................................................................... 81 5.8 Implications for Risk Pooling in Kosovo............................................................ 81 CHAPTER 6. STRATEGIC RESOURCE ALLOCATION AND PURCHASING OF PRIORITY HEALTH SERVICES.......................................................................84 6.1 Overview and Definitions ................................................................................... 84 6.2 Purchaser-Provider Split..................................................................................... 85 6.3 Purchasing a Basic Benefit Package ................................................................... 86 6.4 Strategic Purchasing............................................................................................ 89 6.5 Health Facility Management Autonomy ............................................................. 90 6.6 Provider Payment System for Outpatient Care ................................................... 93 6.7 Performance-based Capitation Payment.............................................................. 95 6.8 Hospital Payment Mechanisms ........................................................................... 97 6.9 Implications for Purchasing in Kosovo............................................................... 99 CHAPTER 7. CONCLUSION AND RECOMMENDATIONS...............................................104 REFERENCES.............................................................................................................................111 MAP ­ IBRD 35406 Boxes Box 3.1 - Legal Status of Health Insurance in European Countries.............................................. 23 Box 3.2 - Slovakia: Health Insurance Companies and Surveillance Authoritya........................... 25 Box 3.3 - Estonia: Supervisory Board of the EHIF....................................................................... 28 Box 3.4 - EU Social Security Regulations .................................................................................... 29 Box 3.5 - Performance Scorecards, Health Fund Annual Report.................................................. 30 Box 3.6 - Illustrative Table from the EHIF Annual Report HIFs.................................................. 30 Box 4.1 - France and Financing of SHI......................................................................................... 51 Box 4.2 - Summary of the Tax Structure in Kosovo..................................................................... 65 Box 6.1 - A Possible Contracting Strategy for Kosovo................................................................. 89 Box 6.2 - PHC Provided by Private Sector ................................................................................... 91 Box 6.3 - Organizational Structures in Hospitals.......................................................................... 91 Box 6.4 - Accreditation of Health Care Providers......................................................................... 92 Box 6.5 - Examples of Performance Indicators of Access/Availability of Care........................... 96 Figures Figure 1.1 - Key Functions in the Health Financing System........................................................... 4 Figure 1.2 - Flow of Government Health Funds in Kosovo, 2007.................................................. 5 Figure 2.1 - Outpatient Visits per Person ...................................................................................... 12 Figure 2.2 - Most Frequent PHC Diagnoses ................................................................................. 12 Figure 2.3 - Hospital Beds per 100,000......................................................................................... 13 Figure 2.4 - Occupancy Rate, 2006............................................................................................... 13 Figure 2.5 - Hospital Admission Rate, 2006 ................................................................................. 13 Figure 2.6 - Average Length of Stay, 2006................................................................................... 13 Figure 3.1 - Analytical Framework for Good Governance of Health Insurance........................... 20 Figure 4.1 - Unpacking "Sources" and "Collection"..................................................................... 39 Figure 4.2 - Sources of Health Finances, 2000-2005 .................................................................... 43 Figure 4.3 - Public spending on health at different levels of per capita GDP ............................... 43 Figure 4.4 - Private Household Out-of-Pocket Payments in % of Total Health Expenditures...... 44 Figure 4.5 - Out-of-Pocket Spending as a Share of Total Income................................................. 45 Figure 4.6 - Out-of-Pocket Expenditures for Health as a Share of Household Consumption....... 45 Figure 4.7 - Health Grants to Municipalities, Per Capita per Year, 2005­06 ............................... 47 Figure 4.8 - General Government Expenditure on Health............................................................. 49 Figure 4.9 - Percentage of Total Health Funds Allocated through Public and Private Mechanisms, 2002............................................................................................................................................... 50 Figure 4.10 - Estonia's Health Expenditures as a Share of GDP .................................................. 56 Figure 4.11 - Mixed Funding Systems .......................................................................................... 56 Figure 4.12 - Bosnia-Herzegovina: Collection Rates by Canton, 2005 ........................................ 58 Figure 4.13 - Social Insurance Taxes as Share of Total Labor Cost ............................................. 59 Figure 4.14 - Payroll Tax Rates and Unemployment Rates: Comparison with EU15 Rates ........ 59 Figure 4.15 - Level of Informal Economy as Percentage of GDP................................................. 60 Figure 4.16 - General Government Revenue as Percentage of GDP (2004)................................. 67 Figure 5.1 - Revenue and Expenditures per Member per Year, across HIFs in Bosnia and Herzegovina (in convertible marka).............................................................................................. 76 Figure 6.1 - New Benefit Package: Kyrgyz Republic ................................................................... 87 Tables Table 1.1 - Key Labor Market Indicators, Western Balkans, 2004................................................. 2 Table 1.2 - Government Expenditures on Health, 2004-07............................................................ 4 Table 1.3 - Measurement of Governments' Capabilities, 2006....................................................... 6 Table 2.1 - Kosovo Main Economic Indicators 2004-08 ................................................................ 9 Table 2.2 - Reaching the MDGs in Kosovo .................................................................................. 10 Table 3.1 - Health Financing and Risks of Corruption.................................................................. 18 Table 3.2 - Health Insurance Models and Implications for Accountability .................................. 19 Table 4.1 - Advantages and Disadvantages of Different Methods of Revenue Enhancement...... 40 Table 4.2 - Kosovo: Health Financing, Main Indicators 2001­06 ................................................ 42 Table 4.3 - Poverty Impact of Health Expenditures ...................................................................... 46 Table 4.4 - Kosovo Government Expenditures on Health, 2004-07 ............................................ 47 Table 4.5 - Public Revenues: Central and Own Source Revenues: 2004-2007............................. 48 Table 4.6 - Municipality Revenue and Expenditures on Health, 2005, by Economic Category... 48 Table 4.7 - Financing of SHI, by Selected OECD Country .......................................................... 50 Table 4.8 - Equity in Health Financing ......................................................................................... 52 Table 4.9 - Contribution Base, 2005 estimates.............................................................................. 61 Table 4.10 - Scenario 1: Contribution Rate of 10%, and Health Expenditure equal to 2004........ 62 Table 4.11 - Scenario 2: Contribution Rate of 15%, and Health Expenditure 2004 plus 10% ..... 62 Table 4.12 - Scenarios 1-3: Estimated Health Spending by Source of Funding ........................... 63 Table 4.13 - Legal and Regulatory Issues for Implementing Private Health Insurance................ 70 Table 5.1 - Insurance Market Structure and Risk Adjusters in Multiple Insurance, 2006 ............ 75 Table 7.1 - Options for Different Health Financing Reforms ..................................................... 105 Table 7.2 - Proposed phasing for health financing reforms ........................................................ 108 ACKNOWLEDGEMENTS This report was prepared by a team led by Pia Schneider (main author), and comprising Jack Langenbrunner (health financing), Junghun Cho (governance and institutional arrangements), Flora Kelmendi (data collection and analysis) and Sreypov Tep and Viktoria Lebedeva (document processing). Famir Besimi contributed the section on the macro-economic and fiscal situation, and Linda Moll edited the document. The report was prepared under the overall supervision of Armin Fidler, Sector Manager (ECSHD) and sponsored by Jane Armitage Country Director (ECCU4) and the previous Country Director Orsalia Kalantzopoulos. Peer reviewers were Monika Huppi, Lead Economist (IEG), George Schieber, Advisor (EASHD), and Michael Borowitz, Senior Governance Advisor Access to Medicines Human Development Group, Policy Division, DFID, London. The team gratefully acknowledges the collaboration provided by the Ministry of Health, the Institute of Public Health, and the Ministry of Finance of the Government of Kosovo. i EXECUTIVE SUMMARY 1. Health financing reform is at the top of the policy agenda in Kosovo. The health sector in Kosovo is largely tax-funded, from general revenues taxes, while private sources contribute an estimated 40 percent. Total levels of health financing are comparable to other countries at similar levels of gross domestic product (GDP). Information about health outcomes, particularly about morbidity and mortality, is incomplete. Relatively high infant mortality and tuberculosis rates are a cause for concern. Utilization rates in public primary health care centers and in hospitals are low, suggesting that financial barriers exclude many people, especially those in vulnerable groups, from care, while those who can afford to pay high user fees seek care in the private sector, including outside of Kosovo (primarily in the Former Yugoslav Republic (FYR) of Macedonia, in Montenegro, and in Serbia). 2. Kosovo plans to introduce social health insurance (SHI). In April 2006, the Kosovo Parliament passed a health insurance law to introduce SHI that would be financed primarily through payroll taxes; after further analysis, the law was returned to the government. The objective of the SHI was to raise additional revenues for the public health sector and improve access to care through risk pooling. The further analysis concluded that the revenue potential of payroll-funded insurance was modest considering Kosovo's small formal sector, and that alternative revenue sources should be explored. Financial and economic implications of the benefits associated with the SHI scheme also raised concern, as did the administrative costs of introducing a payroll tax bearing in mind the limited availability of human, organizational, and institutional capacity to implement and manage an insurance fund. Thus, following the advice of the international donor community, the law was returned to the government. During meetings with the government and donor representatives, it was agreed that the World Bank would study health financing in Kosovo; this report is the result of that study. 3. The objective of this report is to present information on the different health financing reform options available to Kosovo, which should help the government to make informed policy decisions about financing reforms. The report focuses on the key insurance functions of revenue collection and management, risk pooling, and purchasing of health care, as well as the supportive regulatory and governance framework for health financing. These issues are discussed and options derived based on a review of the macroeconomic, fiscal, socioeconomic, labor, institutional, and health sector constraints in Kosovo. Lessons on health financing reforms are presented from elsewhere and implications for Kosovo discussed. The local experience with the Kosovo Pension Savings Trust (KPST) is presented to identify successful functions and activities that could be applied to a health insurance fund (HIF) as well as unsuccessful ones that should be avoided. Findings from this report may also help to inform future donor support to the Kosovo health sector. Under the UN Security Council Resolution 1244 (1999), Kosovo is administered by United Nations Interim Administration Mission in Kosovo (UNMIK). ii 4. Several challenges need to be addressed when developing and implementing health insurance in Kosovo. These challenges include (i) building a supportive governance framework and operating according to the rule of law, so that the population will trust the insurance programincluding adopting an insurance law, related by-laws, contracts, and fiduciary arrangements; (ii) identifying public and private revenue sources for health financing where the formal sector is small and the labor market cannot be burdened with additional payroll taxes; (iii) organizing a single risk-pooling arrangement to consolidate and streamline the current fragmented system; (iv) identifying purchasing contracts with public and possibly private sector providers for a benefit package that improves equity in access to health care and financial sustainability of the insurance fund; and (v) restructuring the health care delivery system by moving toward a purchaser-provider split that gives increased autonomy to health facilities and strengthens the stewardship role of the government. This report proposes ways to address these challenges in the context of Kosovo. 5. Before embarking on such major health financing reforms, the government will have to discuss and make some substantial initial decisions, and estimate the potential financial and organizational impact of those decisions. The government will need a clear vision about what kind of health insurance would be financially and organizationally most feasible in Kosovo and what the implications would be for the current health care system and related sectors such as labor markets and the private sector. Such a vision and the related strategy should be based on the overall health policy goals of the sector and take into account the macroeconomic, fiscal, labor, health financing, institutional, organizational, and human resource situation. For instance, in the current context of resource constraints and a small formal sector, a HIF financed with payroll taxes and private premiums will result in pooling among the better-off and rationing of care, which violates the policy goal of expanding equity of access to care. Hence, policymakers will have to make informed choices for each function including revenue composition, risk pooling, purchasing, and governance. The impact of each choice has advantages and disadvantages and the best trade-off should be sought given the socioeconomic and health context of Kosovo. 6. Any future insurance system must guarantee adherence to the rule of law, good governance, transparency, professional insurance management, and responsibility and accountability for results in the relationship between he insurer and contracted providers. This calls for substantial investment in standard insurance and health facility financial management systems, financial audits, quality, and utilization control in all health care facilities, and human capacity building. To gain the population's trust in the health system, insurance and facility managers need to be made responsible and accountable. Many countries provide examples of insurance funds that operate according to the rule of law, and with transparent and professional management: Austria, Estonia, France, Germany, and Switzerland. Closer to home, the KPST offers valuable lessons. 7. Based on the analysis presented in this report, the most feasible health insurance option in Kosovo is one comparable to the insurance funds of Estonia and the Kyrgyz Republic. Kosovo's labor market is too small and fragile to absorb a payroll iii tax increase. With only about 15 percent of the population working in the formal sector, while 100 percent of the population will try to use the insurance, a payroll tax-funded scheme is financially unsustainable. Therefore, insurance revenues will have to come mainly from the government budget, funded by direct and indirect taxation. Kosovo could explore increasing indirect taxes on tobacco and alcohol, as well as on luxury goods, to raise additional government revenues. Under a tax-funded system, the collection agent is the tax authority that transfers funds annually to the HIF. 8. A mixed funding system from private and public sources would enhance the likelihood of insurer sustainability. Based on a review of the current health financing and tax situation as well as neighboring countries' experience with payroll tax-funded insurance, it is recommended that Kosovo explore other potential revenue streams for the HIF. In addition to funding from general revenues as discussed in the preceding paragraph, Kosovo could use a mix of direct and indirect tax funding as do France and other countries, and direct contributions paid by individuals to the HIF as does Switzerland. The combination of different funding sources could be designed to achieve equity in financial access to health care through progressive health financing. Actuarial cost analysis of the benefit package will be conducted to adjust the package to the available resource envelope, estimate future funding needs, and propose a composition of funding from public and private sources that encourages equity and financial sustainability. Eventually, additional private funds could be raised to pay for benefits not covered by the basic package. To limit moral hazard among the insured, some level of cost-sharing should be introduced with adequate safeguards for the poor and other vulnerable groups, again, to ensure equity of access to care and progressivity in health financing. Mandatory enrollment for all residents would prevent risk selection and exclusion of high-risk individuals. 9. A single insurance system with mandatory enrollment for all residents is the preferred pooling solution. Kosovo's population is rather small for multiple risk pools. Kosovo does not have the institutional structure for risk-equalization transfers or risk- based premium definition that would be needed in a multiple insurance system. Single insurance would also contribute to solidarity among population groups (healthy and chronically ill, poor and rich, etc.), and prevent risk selection by the insurer. A single pool would also help to overcome the current fragmentation of risk pooling, with primary health care now being pooled at the municipal level and hospital care at the national level, and to build a continuum of care from the primary to the tertiary level. At the beginning, Kosovo may want to consider contracting out insurance management to an experienced international health insurance company that could also help build human capacity and information technology (IT) in Kosovo. 10. Substantial organizational and institutional reform of health care delivery is needed to allow providers to respond to the new financial incentives. Kosovo's health sector is still organized on the direct-provision model. To implement insurance, the organization and management of health facilities would have to be modernized to reflect a European standard. Reforms include separating purchaser and provider, increasing management autonomy for health facility directors, and passing laws to give health facility directors authority over production factors such as staff. The government iv will have to play a strong stewardship function through accreditation and quality control of all facilities. 11. Contracts between the insurer and providers and provider payment reforms will reward better quality and efficiency in the provision of care. A modern insurance company purchases care selectively, monitors and evaluates the performance of contracted providers, and excludes low-quality providers from contracting. In Kosovo, the insurer will need to select its payment method strategically. Payment reforms will take time andas in neighboring countriescould develop from the current line-item budgeting toward case-based payment for hospitals and a capitation formula that rewards better quality of care in outpatient settings. These reforms would require substantial investment in data collection infrastructure and connectivity between health facilities and the HIF, and training of staff in data collection and analysis of managers in use of analysis results in adjusting the care process in health facilities, and of HIF employees in contracting and purchasing. 12. If Kosovo chooses to implement health financing reforms, the roles and responsibilities of various health sector actors will change substantially in the sector and require concerted efforts to strengthen the accountability framework of the sector. The Ministry of Health would assume a policymaking and stewardship role, and implement reforms in quality assurance in the provision of care, organization of health facility management, and management of resources including staffing and pharmaceuticals. The Ministry of Economics and Finance would have to redirect the financial flow of health funds to the health insurer, which will assume full responsibility for risk pooling and purchasing of health care. The insurer will contract with accredited health care providers who offer a defined set of services to the population against an established price. Service providers will be given increased autonomy to produce these services most effectively and their performance in delivering the defined benefit package will be evaluated against an established set of performance standards. These changes will require investments in legislation, regulation, new governance and organizational structures, and information management so as to ensure proper accountability of all actors. The role of current stakeholders including municipalities will need to be revisited and redefined based on an institutional review to ensure more efficient flow of funds and resource management. 13. When constituting a HIF, Kosovo may wish to follow the successful performance of the Kosovo Pension Pillar II Trust Fund and apply its organizational and management form to the HIF. The Pillar II Fund is a public autonomous organization managed by an independent professional team (Gubbels 2007) that could serve as a model for a future HIF. Similar to the KPST, the future health insurer could be established as an independent autonomous organization with the purpose of risk pooling, purchasing health care from providers in Kosovo and abroad, and administering the health insurance system. The insurer would be staffed by Kosovar management and employees recruited from the international diaspora working in health insurance companies, as well as locally. The governing board could have a combination of local and foreign membership. The regulatory and oversight body of the health insurer could be the Banking and Payments Authority of Kosovo. Statutory reserves would have v to be securely invested, according to international best practices, and yield positive real net returns to the insurer, with comparatively low administrative fees. In planning to staff a future insurance organization, the governing board and its advisors should hire a small staff of highly motivated individuals who would be compensated on par with other financial sector organizations in Kosovo. The foundation for supervision of insurance will need to be strong, with effective transparency requirements, as in Estonia, for instance. An annual progress report should be conducted to evaluate whether the insurance system is comprehensively implemented, and to identify issues in functioning including the effectiveness of collection, record keeping, IT systems, and information reconciliation processes. 14. The impact of such major health financing reforms should be carefully monitored and evaluated to identify and correct for adverse effects at an early stage. It will take several years to develop and implement major health financing reforms such as the ones described above, and they will affect various stakeholders, whose support for the reforms must be gained. Some of the financial changes may lead to unexpected adverse effects that would need to be corrected, such as inequity in health financing or substandard quality of care. To monitor progress and detect problems, a monitoring and evaluation framework is needed. A baseline analysis should measure existing equity in utilization and financing of health care, patient satisfaction and care seeking behavior, quality of care, and cost and efficiency levels in health facilities. Findings from the baseline could help define the benefit package, levels of cost-sharing, and payment reforms and be used for comparison purposes with the results of later monitoring and evaluation of reforms, to identify progress and reform areas that need strengthening. vi CHAPTER 1. INTRODUCTION 1. Health financing reform is at the top of Kosovo's health policy agenda. The government aims to strengthen the health system through health financing reforms and modernization of health care delivery. Kosovo Health Law No. 2004/4 provides the legal base for health financing reform. The law proposes to establish a single health insurance fund (HIF) that will pool all government health monies, and pay for basic care provided in public health facilities.1 The Kosovo Health Strategy 2005-2015 is to guide reforms in financing and provision of care. A future health financing system in Kosovo is expected to raise additional health funds; create efficient risk pooling to avoid catastrophic health expenditures and ensure equity in access to care; and set incentives through the provider payment mechanism for improving the performance in the delivery of health care. A masterplan is being developed with the objective of describing a health care provider network for the population and their health needs of today, as well as in the medium and longer term given the available resources. In addition, a supportive institutional and governance framework will have to be created to sustain these reforms. 2. Kosovo aims to reform health financing by moving toward a health insurance system. To implement the proposed single insurer as described in the health law, a draft law for health insurance was discussed in Parliament in 2004. The draft law envisioned the revenue sources for the future health insurance as follows: (i) 60 percent of HIF revenue would be paid by the general budget, (ii) 13.5 percent would come from payroll contributions, (iii) 21 percent would come from direct payments by patients in the form of user fees, and (iv) 6.5 percent would come from patients' co-payments. Three years later, in April 2007, the Kosovo Parliament approved a health insurance law to introduce social health insurance (SHI) financed predominantly through payroll taxes, though it did not specify contribution levels.2 The revenue potential of such payroll- funded insurance was estimated to be modest considering Kosovo's relatively small formal sector and employment rate (Statistics Office of Kosovo 2005) (table 1.1). The law raised concerns about the financial and economic implications, as well as the additional benefits and costs of introducing a payroll system bearing in mind the limited availability of human, organizational, and institutional capacity to implement and manage a HIF. Thus, following the advice of the international donor community,3 the law was returned to the government. Under the UN Security Council Resolution 1244 (1999), Kosovo is administered by United Nations Interim Administration Mission in Kosovo (UNMIK). 1See Health Law Section 58 and 59, as well as Section 22 on definition of basic care. 2Assembly of Kosovo: Law on Health Insurance, April 2006. 3Letter from the Special Representative of the Secretary General, U.N. Mission for the Interim Administration of Kosovo to the President of the Assembly of Kosovo, June 15, 2007. 1 Table 1.1 - Key Labor Market Indicators, Western Balkans, 2004 Labor Employment Unemployment Long-term participation rate rate unemploymentb ratea Kosovo 52.3% 28.9% 41.4% 83.5% Albania 63.7 % 60.1% 5.6% 68.4% BiH 59% 46% 22% n/a Macedonia 51.2% 32% 37.2% 84.5% FYR Serbia 66.6% 53.5% 19.5% 71% Montenegro 65.1% 40.6% 23% 85% Sources: Statistics Office of Kosovo 2005; ILO (International Labour Organization) 2005 for FYR Macedonia. World Bank 2007 for other countries Note: BiH = Bosnia and Herzegovina a. Definitions of working age vary across countries. b. Share of unemployed with duration of 12 months or more. For Montenegro, based on registered unemployed. 3. The objective of this report is to inform the government of Kosovo of health financing reform options. The study examines different insurance options, focusing on revenue collection and management, risk-pooling arrangements, purchasing, and the supportive regulatory and governance framework for health financing. The analysis takes into account the macroeconomic, fiscal, and socioeconomic constraints as well as the institutional, organizational, and existing social insurance context in Kosovo. The report will draw from lessons learned from previous analytical work from Kosovo including on the sociodemographic, socioeconomic, labor, and health situation; present key issues on the health financing and supply side; and compare findings with the experience of financing reform in other countries in Europe and Central Asia (ECA). Findings from this report may also inform future donor support to the Kosovo health sector. The study will not address technical issues related to provider payment methods, performance incentives for primary health care (PHC), and the provision of care, as these topics are covered under the current Public Expenditure Management Technical Assistance Grant (PEMTAG). 4. The World Health Assembly put forward prepayment and risk pooling as a major option for improving the performance of health systems. In 2005, the World Health Assembly passed a policy resolution recommending that irrespective of the source of financing, health financing systems should include a method for prepayment, with a view to sharing risk among the population and avoiding catastrophic health expenditure and impoverishment of individuals as a result of seeking care (World Health Organization, 2005). In Kosovo, where health insurance is not available, and 15 percent of the population is estimated to live in extreme poverty (World Bank 2007), paying for health care has pushed some near-poor households into poverty, leading to a 3 percentage point increase in the poverty rate. 5. Introducing health insurance in Kosovo would be expected to better contribute to goals such as equity, efficiency, financial sustainability, and better 2 health outcomes than does the current financing and delivery system. Health financing reforms need to be accompanied by reforms on the delivery side. This will require substantial investment that is expected to pay off in the long run, and lead to sustainable health financing to pay for an affordable benefit package; improved access, equity in utilization and financing of health care, and protection against the poverty- producing impact of health spending; a reorganization of the delivery side that will lead to more efficient and better quality care; and the introduction of an active purchaser that uses modern contracting and payment methods to support reforms by setting adequate incentives. Government stewardship including an appropriate policy and institutional framework is fundamental for reforms to achieve their objectives. 6. A health system's financing functions consists of three interrelated componentsrevenue collection, risk pooling of resources, and purchasing of health care. These three functions are often carried out by the same organization, such as a health insurer. In voluntary health insurance (VHI), the pooling and collecting organizations are identical. In a mandatory insurance system, the SHI entity may pool funds and a government agency such as the tax administration collects them. The two other core functions of the health system are provision of care and stewardship (figure 1.1). These core functions are influenced by governments through their stewardship and governance function and by the population through their ability to co-finance and access the benefit package, which is affected by factors such as a population's health status and socioeconomic and sociodemographic background, and the availability of quality health care. 7. There are five alternative methods for collecting revenues for health care, resulting in different types of risk pooling. Health care can be financed through (i) the tax-funded government budget, (ii) a payroll tax earmarked and transferred to health insurance, (iii) a premium paid to mandatory or voluntary health insurance program, (iv) individual health savings accounts without risk sharing, and (v) patients' out-of-pocket payments. These methods result in different risk-pooling arrangements. Revenue collection through tax financing and payroll contribution aims at pooling the health risk of an entire population, and yields better results with respect to equity and risk pooling than VHI with limited risk pooling, or medical savings accounts and out-of-pocket payments with no risk pooling at all. 3 Figure 1.1 - Key Functions in the Health Financing System Health services Provision of services ngi dna ) Cost sharing/user fees noi nc naif onita at The Coverage Purchasing of services popu of p hi guler,e mrofni Choice? of Coverage lation Pooling of funds dsra nc Choice? we nar onis ovi St pr gove( Entitlement? Collection of funds Contributions Source: Kutzin, 2001 8. The Kosovo health system is predominantly tax funded. Government health spending is about 3 percent of gross domestic product (GDP) and 10 percent of general government expenditures (table 1.2). The Ministry of Economics and Finance (MEF) transfers health funds from the central budget to hospitals (51 percent), to municipalities in the form of an earmarked health grant for the provision of PHC services (26 percent), and to the Ministry of Health (MOH) for other services (22 percent) (MEF 2007). Figure 1.2 provides an overview of the flow of public health funds. The current level of government spending manages to cover about half of total health expenditures while patients co-finance care out-of-pocket at the point of service use. In the absence of recent household survey data and accounting systems in health facilities, the level of private spending is extrapolated based on estimates from the 2002 Living Standards Measurement Survey (LSMS) and more recent Household Budget Surveys (HBS). Table 1.2 - Government Expenditures on Health, 2004-07 Kosovo 2004 2005 2006 2007 (est) Government expenditures on health ( 64.6 71.4 68.3 70.1 million) Government expenditures on health (% 2.8% 3.2% 3.0% 3.1% of GDP) Government expenditures on health (% 10.4% 9.6% 9.8% 9.8% of general government expenditures) PHC grant (% of government 28% 25% 26% 27% expenditures on health) Source: MEF; Ministry of Local Government and Administration (MLGA) 4 Figure 1.2 - Flow of Government Health Funds in Kosovo, 2007 PIP MEF Capital Municipal Investment Revenues Primary Health Grants Municipal Health Budget Hospital MOH PHC Risk Pool Budget Budget Recurrent Wages & Wages & Budget Salaries Salaries Capital Primary Health Care Centers Hospitals Primary Health Care Staff Hospital Staff Source: Author's compilation based on information from MEF 2007, World Bank 2005, NHA MOH/WB 2007. Note: PIP=Public Investment Program 9. Substantial reforms in the health care delivery system are needed to implement effective health insurance. Kosovo is among the few health sectors that still have a Shemasko health care modelthe centrally planned, National Health Service (NHS) model of the former Soviet Union. Under this direct-provision model, the financing, risk pooling, and provision of health care is integrated and managed by the same organization and the budget is derived from the general state budget. Health financing reflects simple disbursement. There is no purchasing agent that sets financial incentives for providers to improve efficiency and quality of care. Public health facilities are owned by the state and all health care personnel are salaried, state employees governed by civil service law and budget rigidities. Implementing insurance would thus require a separation of the financing and provision functions in the health care delivery system and professional management in health facilities. 10. The effectiveness of insurance in revenue collection, risk pooling, and purchasing depends on Kosovo's governance capabilities such as effective and transparent government and the accountability of organizations. Good governance is measured by the quality of public and civil services, and the degree of government independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies. Governance of health insurance encompasses three functions (i) stewardship of the system, (ii) the structure for oversight, and (iii) administration of the insurance system. The necessary governance mechanisms include the definition of legal status, ownership, mechanisms for representing interests, forms and scope of governmental supervision, and information reporting requirements. The underlying principles are: transparency and the rule of law, consistency, accountability, inclusiveness and participation, and effectiveness and efficiency (Savedoff and Gottret 2007). 5 11. The health sector is limited by weak governance and institutions. The control of corruption measures the extent to which public power is exercised for private gain, including informal payments for health care, as well as "capture" of the state by elites and private interests (Kaufmann et al. 2007). In 2006, Kosovo scored -0.36 in government effectiveness and even lower, -0.63, in control of corruption, on a scale ranging from +2.5 to -2.5. With these results, the Kosovo government's effectiveness is comparable to Egypt and Tanzania; its control of corruption is similar to Albania and Iran but weaker than Tanzania and Egypt (table 1.3). Countries with a government effectiveness score of less than 1.0 tend to have difficulties in delivering effective and efficient health care independent of their financing and risk-pooling system (Hsiao 2007). Central European countries that reach an effectiveness score of 1 include Estonia and Slovenia. Any future health financing reform in Kosovo must therefore be built on principles of good governance, transparency, responsibility, and accountability for results, which would require investment in standard financial management systems in insurance and health facilities, financial audits, quality and utilization control in all health care facilities, and keeping managers and health staff responsible and accountable. Table 1.3 - Measurement of Governments' Capabilities, 2006 Government effectiveness Control of corruption Estonia +1.17 +0.87 Slovenia +1.11 +0.92 Tanzania -0.31 -0.37 Egypt -0.35 -0.42 Kosovo -0.36 -0.63 Albania -0.42 -0.59 Iran -0.80 -0.63 Source: Kaufmann et al. 2007 12. The remainder of this report is organized as follows. Chapter 2 presents macroeconomic and fiscal situation, and health outcome and delivery issues. Chapter 3 proposes the governance structure for a future HIF, focusing on the aforementioned three functions (i) stewardship, (ii) the structure for oversight, and (iii) administration of the insurance system. The chapter draws from insurance regulatory experience from other countries and from Kosovo's own experience with governance of pension insurance to propose an effective governance framework for health insurance in Kosovo. Responsibilities for the functions of governance are identified and governance mechanisms and principles described. Chapter 4 presents different options for health financing sources and collection function and proposes a most effective mix of financing sources for health based on a review of different countries' experience, and taking into account Kosovo's small formal sector. Chapter 5 reviews the experience with risk pooling and proposes an organizational form that leads to equity in access to care for the entire population. Chapter 6 is about purchasing health care. It presents the different payment methods for providers and how these can be used in Kosovo such that financial incentives will lead to better performance and improved health outcome for the population. Chapter 7 will conclude presenting different options of health financing reforms and practical recommendations addressed to the government of Kosovo. 6 Concrete actions are identified for developing and implementing these reforms in a phased approach over the next 10 years. 7 CHAPTER 2. HEALTH FINANCING IN CONTEXT This chapter describes the context in which health financing reform takes place in Kosovo. It starts with a description of the current macroeconomic and fiscal situation and outlook, and presents implementation constraints that limit the budget execution. It then identifies inefficiencies in the delivery of care and treatment seeking behavior. Current health outcomes are discussed, as well as limits in information about the health sector. 2.1 MACROECONOMIC AND FISCAL SITUATION 13. Kosovo's economy is growing moderately and the donor sector continues to exert a major influence on macroeconomic developments. In the first years following the conflict, a surge in donor support contributed to rapid economic rebound. However, as the subsequent scaling back of donor and UN presence counteracted the positive effects of increasing domestic investment, Kosovo's GDP growth stagnated. The sole exception was 2004, when a one-off surge in capital spending led to somewhat faster growth. Inflation has been low and stable, though in 2007 it reached 4.3 percent mainly as a result of the global increase of the food and oil prices. 14. Kosovo's external trade continues to be significantly out of balance and trending upward. While massive errors and omissions cast serious doubts on the exactness of the reported balance of payments data, Kosovo no doubt runs a trade deficit. Recorded exports amounted to only 6.3 percent of GDP in 2007, recorded imports 59.5 percent, resulting in a reported trade deficit was 53.2 percent of GDP. The chronic trade deficit drives the current account deficit (35.6 percent of GDP, before grants), which is largely financed by remittances (16 percent of GDP) and Official Development Aide (13.8 percent of GDP). The growth of the trade and current account deficits have been driven by private sector demand, the conservative fiscal position of the government over the past two years apparently having had little impact on Kosovo's imbalances. While these imbalances certainly pose risks to Kosovo's economic stability - and especially to its longer-term development prospects - the unusually high level of Kosovo's external financing makes it difficult to evaluate these risks with confidence. It is likely that Kosovo's reliance on official and private transfers is in relative terms a more stable pattern of financing than other alternatives. 15. A combination of expenditure limits, underspending on the capital budget, and rising revenues brought Kosovo's fiscal balance into surplus in 2006 and 2007 and increased the government's accumulated cash deposits to 18 percent of GDP by the end of 2007. From a low base, Kosovo has continued to increase general government revenues rapidly, from 26 percent of GDP in 2004 to 38.8 percent in 2007 (table 2.1), though they remain heavily reliant upon indirect taxes collected at the border. Spending, meanwhile, has been strictly curtailedfalling from 31.9 percent of GDP in 2004 to 26.5 percent of GDP in 2007. As a result, the Kosovo Consolidated Budget swung sharply from a deficit of more than 5.9 and 4.1 percent of GDP in 2004 and 2005, 8 respectively, to a surplus of 3.5 and 9.2 percent of GDP in 2006 and 2007, respectively. The decrease in government expenditure from 2004 to 2007 has been due in part to a conservative policy on recurrent spending, which has fallen over time, and to persistent underspending of the capital budget, which had been intended to benefit from the fiscal space created by limited recurrent spending. In addition, the budget for 2008 as approved by the Assembly of Kosovo does not reflect a sustainable fiscal policy since it incorporates a budget deficit of 7.6 percent of GDP, fueled by a massive planned increase in capital spending. This threatens to sustain the historic volatility of fiscal balance sheets. However, continued problems in spending capital budgets could again curtail actual deficits to more modest levels. Furthermore, an expected mid-term review of the budget building on a new Medium Term Expenditure Framework (MTEF) 2009-11, could provide an opportunity for corrective action. The MTEF forecasts that, given the limited fiscal space, the government will allocate about 10 percent of the general budget to health, thereby continuing the trend of the previous three years. Table 2.1 - Kosovo Main Economic Indicators 2004-08 2004 2005 2006 2007 2008 Est. Est. Est. Est. Proj. National accounts Real GDP growth (, %) 3.2 -1.0 2.6 2.8 5.1 Consumption (% of GDP) 111.8 118.2 119.0 121.5 119.5 Investment (% of GDP) 29.2 27.8 30.4 32.0 36.9 Price changes (%) Consumer price index -1.4 -1.4 1.5 4.2 4.0 General government budget (% of GDP) Revenues 26.0 27.5 30.4 38.8 30.8 Expenditures 31.9 31.6 26.9 26.5 31.7 Overall balance -5.9 -4.1 3.5 9.2 -0.9 External accounts (% of GDP) Trade balance -41.9 -45.2 -49.1 -53.2 -56.6 Current account -30.5 -32.5 -32.5 -35.6 -39.3 Foreign assistance 14.3 16.0 14.2 13.8 11.6 Remittances 9.3 12.0 15.0 16.0 16.6 Memorandum items GDP ( millions) 2,348 2,284 2,341 2,474 2,684 GDP per capita () 1,195 1,143 1,152 1,197 1,277 Population (000) 1,965 1,999 2,033 2,067 2,102 Source: International Monetary Fund staff estimates. Note: From local authority sources, the Kosovo Consolidated Budget 2007 (actual) and 2008 (as approved by the Assembly) are: (in percent of GDP) 2007 2008 Revenues 38.0 37.6 Expenditures 28.2 42.7 Overall balance 9.8 -7.6 9 2.2 HEALTH OUTCOMES 16. There is a lack of information on health outcomes, including morbidity and mortality statistics that could be used for international comparison. Kosovo does not collect data on common health indicators, including basic demographic indicators; lifestyle- and environment-related indicators; mortality, morbidity, and disability; and health care resources, comprehensive utilization, and expenditure. The existing data on the population's demographic characteristics and health status are highly contradictory, and highlight the need for investment in better monitoring and evaluation capacity at the MOH. This lack of data prevents analysis of trends and comparisons of international health statistics that could help to support the formulation and monitoring of health policy at the national level. 17. Kosovo needs to make considerable progress to improve basic health indicators. Child and infant mortality are considered to be the most complex issues impeding Kosovo's development. (United Nations Development Programme 2007). Major investment in the quality of basic health is needed to improve health outcomes for infant mortality, tuberculosis (TB), immunization, and reproductive health care (table 2.2). Life expectancy at birth is lowest in the South Eastern Europe region69 years in Kosovo, compared with 73 years in Serbia and 74 in Bosnia and Herzegovina, suggesting that the Millennium Development Goals (MDGs) are not the only health challenge in Kosovo and highlighting the need to establish comprehensive health statistics as reported by most ECA countries. Table 2.2 - Reaching the MDGs in Kosovo Kosovo Kosovo Indicators 2002/4 2006 MDG Goal 2015 Percent of population below 1.42 a day 44% (a) n/a 22% Perinatal mortality (per 1,000 live births)* 27 (b) 23(b) 9 Maternal deaths (per 100,000 live births) 22 (b) 7 (b) 5.5 Number of new TB cases per year 48 (c) 58 (d) 26 Number of new AIDS cases per year ** 4 (c) 2 (d) Reverse spread % of rural population with access to safe water 72% (e) 78% (e) >80% Sources: (a) HBS 2002; (b) UNDP 2007; (c) Kosovo HIS 2004; (d) Statistics Office of Kosovo 2006; (e) World Bank 2007; United Nations 2000. Note: *Perinatal mortality serves as a proxy for child mortality. No valid data on under 5 and infant mortality. **The low number of new HIV/AIDS cases could be due to underreporting and points to a need to strengthen the surveillance system. 2.3 DELIVERY OF CARE AND TREATMENT SEEKING BEHAVIOR 18. Kosovo has a health care system of direct provision, which is common in developing countries. Under direct provision, the financing and provision of health care is integrated and managed by the same organization, which in Kosovo is the government. The public financial management system executed by the MEF Treasury department is in 10 charge of the centralized budgets paid to hospitals that are owned by the MOH. For PHC, the MEF transfers the per capita health grant to municipalities, which are the owners and financial managers of PHC facilities. The system of direct provision relies on central planning. Studies estimate that inefficiency of the direct-provision model may average 30 percent of total health expenditures. As a result of these shortcomings, many countries that once used direct provision have moved to split the purchasing and provision function in health care (Hsiao 2007). 19. Health facility managers have little authority over spending, staff levels, staff selection and performance, and capital; they are unable to take steps that would improve efficiency of care, for example, through strategic staffing or pharmaceutical management. Hospitals are managed centrally by bureaucratic rules rather than on the basis of efficient operations. The municipal health directorate manages the PHC budget and is responsible for all resource management including hiring/firing of PHC staff; while the health facility director is nominally responsible for facility performance, he/she is not actually held accountable. Health facility managers are not given autonomy to perform the usual range of management functions such as planning, organizing, leading, and controlling resource utilization, and there is no award system for staff and managers, other than the negative possibility of reassignment. Physicians and health workers are civil servants with job guarantees, and staff is promoted based on civil service rules, independent of job performance and patient satisfaction. Some health workers pursue their own interests, including taking informal payments from patients as nontaxable salary increases and "moon-lighting" in private practice while on the public payroll (Gaumer 2007). This lack of management responsibility and accountability affects the delivery of health care (Boulton and James 2007). 20. PHC centers appear to be underutilized. In 2006, the visit rate for all outpatient care was 1.9 visits per capita, of which 0.6 visits are to Family Medicine practitioners. Of countries in the South Eastern Europe region, only Albania reports a lower visit rate (figure 2.1). PHC physicians presumably have capacity to see more patients; on average in 2006, there were 18.4 visits per physician per day, though in practice this may be less as not all of these visits are to a physician (for example, laboratory) (Statistics Office of Kosovo 2006).4 This low physician productivity combined with a growing private sector suggests that physicians use the public sector visits to refer patients to their private practice (Boulton and James 2007). The most frequent reasons for a PHC visit are locomotor system diseases by men age 50-64 years (34 percent), followed by respiratory infections (22 percent) (figure 2.2). 4Patient visits calculated assuming 250 working days per year per physician. Note, visit rates in Bosnia and Herzegovina and Serbia are about 20 visits per physician per day. 11 Figure 2.1 - Outpatient Visits per Person Figure 2.2 - Most Frequent PHC Diagnoses 7.0 6.0 6.0 Digestiv Other 4.6 5.0 system 22% Locomot 6% system 4.0 3.3 (men 50-64 3.0 Factors 1.9 age) 2.0 1.5 contact 34% 1.0 health services Respirat - Dermic 8% tissue system Albania vo ia ey 8% 22% so sn rk Ko Bo Tu cedonia Ma Source: Statistics Office of Kosovo 2006 Source: Statistics Office of Kosovo 2006; WHO Regional Office for Europe, Health for all database 21. Kosovo's hospitals are underutilized. Although Kosovo reports the lowest bed density in Europe (figure 2.3), the five regional hospitals report very low bed occupancy rates, ranging from a low of 41 percent in Vushtrri to an average of 65 percent in the other four hospitals (figure 2.4). Low occupancy rates are caused by low hospital admission rates and a short average length of stay (ALOS) (figures 2.5 and 2.6). Hospitals mainly treat secondary and primary care patients and provide only limited tertiary care, leading to a relatively low-severity case mix among patients and a short ALOS. The most frequent reasons for hospitalization in the five regional hospitals are: respiratory diseases (15 percent), pregnancy and childbirth (11 percent), and circulatory system diseases (10 percent). The Pristina Clinical Center reports the same patterns of diseases as regional hospitals (Statistics Office of Kosovo 2006). Low utilization rates in the public sector point to a significant demand for care by patients in the private sector and in neighboring countries including FYR Macedonia and Serbia, where they pay cash at the time of service use. 12 Figure 2.3 - Hospital Beds per 100,000 Figure 2.4 - Occupancy Rate, 2006 600 542.8 100% 86% 88% 462.7 90% 80% 500 75% 414.3 80% 69% 365.2 65% 66% 400 70% 58% 269.5296.9303.6 60% 300 50% 42% 40% 200 148.2 30% 100 20% 10% 0 0% o y ia l o ri i ja ov gr nia rbia hr ilan va ay Pe Ave Kos stina rw TurkeAlban snia Bo Portugaene edo Se Gj ako Vust Gj Prizren Pri ope IrelandNo Mont Mac Eur Source: WHO-Euro and Statistics Office of Source: WHO-Euro and Statistics Office of Kosovo Kosovo 2006 2006 Figure 2.5 - Hospital Admission Rate, Figure 2.6 - Average Length of Stay, 2006 2006 20 17.6 16.6 12 18 9.8 16 10 7.8 14 12.0 7.5 10.7 8 7.1 7.2 12 9.9 9.9 5.9 6.0 8.7 5.8 10 8.2 5.3 6 8 6.3 6 4 4 2 2 0 0 a vo y ni ro vo ar ro ia Koso snia rbia atiavenia rkey so eg sn Bo Alba Turkeydonianeg Se ng MaceMonte Cro Slo Tu oatiaen Bo Sl oveniaKo Hu Po rtugal doniaCr ce Ma Mont Source: WHO-Euro and Statistics Office of Source: WHO-Euro and Statistics Office of Kosovo Kosovo 2006 2006 22. Shortages of essential drugs and other supplies are reported to be widespread. Theoretically, essential drugs are provided to patients free of charge. But, public funding is not sufficient to provide access to essential drugs for the entire population. As a result, at least 80 percent of Kosovo's pharmaceutical market is financed by patients' out-of-pocket payments, contributing to the impoverishing effect of illness. In addition, the public pharmaceutical procurement system is struggling to overcome management and governance deficiencies, including the allocation of pharmaceuticals from the central level to health facilities. A major effort is needed to strengthen the pharmaceutical system, starting with a system to monitor the use of drugs in health facilities. Data on drug prices, availability, and utilization should be collected and then used to do pharmaceutical procurement, price-setting, and prevention of drug stock-outs in health facilities (Seiter 2007). 13 23. Patients who can afford paying higher prices seek care in the growing private sector, though in the absence of accreditation and licensing systems, concerns arise about quality of care. Shortages of drugs and supplies in public health facilities cause patients to seek care in the private sector and purchase drugs in pharmacies. As noted above, an increasing demand for private sector care also is driven to some extent by physicians working in dual-practice and referring patients to their private practice. To respond to consumer demand for private sector care, the number of private providers has been growing steadily in Kosovo, particularly in urban areas, and it is estimated that about 30 percent of PHC visits takes place in private facilities (Boulton and James 2007). There is no accreditation system for health care providers to ensure quality of care. 24. Getting better health outcomes will require better use of resources, and strong policy and institutional efforts. In the context of limited fiscal space, which precludes increasing public spending on health, it is essential to improve the efficiency of the existing level of resources. Strong policy and institutional reforms to strengthen public sector financial management, administrative capacity, and mechanisms for ensuring accountability are fundamental (Gottret and Schieber, 2006). Finally, substantial investment in health statistics and performance data collection and analysis is needed to allow for monitoring and evaluation and to inform decisions on health spending and resource use. 14 CHAPTER 3. GOVERNANCE AND INSTITUTIONAL ARRANGEMENT FOR HEALTH INSURANCE 25. The objective of this chapter is to present a governance framework for the performance of the HIF and introduce basic governance principles and the application of these principles to the main elements of an insurance fund. In addition, the chapter aims to suggest issues in governance framework that the Kosovo authorities may need to consider in introducing SHI. The chapter begins by defining the concept of governance and describing the current governance situation in Kosovo, which will influence decisions on how a HIF should be governed in Kosovo. The chapter then moves to specific governance issues in the health sector, such as its vulnerability to corruption and the types of governance risks posed by different financing methods. It then introduces the main governance elements and applies them to the main institutional elements of a HIF. Drawing lessons from the experience of the Kosovo Pension Pillar II Fund and of other countries, the chapter concludes by suggesting recommendations for the Kosovo authorities to consider in designing and implementing the governance structure. 3.1 DEFINITION AND SCOPE OF GOVERNANCE 26. Governance is generally defined as the traditions and institutions by which authority is exercised for the common good (World Bank 1997). This includes (i) the process by which those in authority are selected, monitored, and replaced; (ii) the capacity of the government to effectively manage its resources and implement sound policies; and (iii) the respect of citizens and the state for the institutions that govern the economic and social interactions among them. 27. For formal laws and rules to be effective, they must be fully reflected in the design of the main processes and workings of an institution and in the attitudes and behavior of the public officials involved. Good governance requires society's insistence that politicians act properly, and that society dishonors them for acting otherwise. It must be unthinkable, including among the elites of the society, for any politician to use violence against opponents, defy court authority, override administrative law, mislead parliament, or misuse coercive power - regardless of the circumstances. The elites, and particularly the senior public servants, must be aware that once core constitutional rules and conventions are broken, they cannot be restored without great political turmoil, a long lapse of time, and the risk of the collapse of the state. 28. Public trust is fundamental to governance because a democratic and fair society depends on widespread "consent to be governed" among its citizens. Improving governance is a way of bolstering the legitimacy of governmental institutions that will make the state more trustworthy - of seeing that states adopt constitutions, of the division of powers, and of other legal and institutional checks and balances. Governance arrangements have no reality until they are fleshed out by communities of politicians, 15 judges, and officials with the skills, motivation, and values to express those constitutional ideals in attitudes and actions. And even then, good governance is not embedded until the wider community of regions, clans, businesses, community groups, and the public at large feel that they have a stake in the wider nation, and trust the institutions of government to protect them and their legitimate interests even if their preferred party is not in power. 3.2 GOVERNANCE IN KOSOVO TODAY 29. There is no strong and sound domestic legacy of good governance in Kosovo and the nature of the stabilizing regime and the period of divided competencies between the United Nations Mission for the Interim Administration of Kosovo (UNMIK) and the Provisional Institutions of Self-Government (PISG) have further undermined the legitimacy of public institutions. While governance in Kosovo was much better under the Federation of Yugoslavia than under the Federation of Serbia, there is no strong domestic legacy of good governance on which Kosovo can draw in addressing the challenges of the future. A prolonged transition between the end of major conflict and the beginning of Kosovo's nationhood has been necessary for the establishment of security and stability. The UNMIK has sovereignty but is little trusted; as security concerns have subsided, it is seen as a foreign occupier, neither representative of nor accountable to the people). The PISG, while representative and somewhat accountable, is seen as lacking the authority essential to credible government. 30. As indicated in a number of surveys and reviews,5 the citizens of Kosovo greatly mistrust the institutions of state. The experience of internal conflict saw a breakdown of the authority and credibility of the state and the loss of trust in public institutions. Frustration seems highest with the court and justice system (especially concerning property rights), unfair and bad local public service, and poor energy supplies and health services. This situation is a key governance risk and presents a major problem to any post-Status administration. Mistrust undermines the very process the state would want to use to restore confidence and good governance: installation of those constitutional, political, legal, and organizational arrangements that are integral to free and democratic states. 3.3 SPECIFIC GOVERNANCE ISSUES IN HEALTH FINANCING 3.3.1 The importance of governance in health services 31. Corruption in the health sector is an especially critical problem in developing and post-conflict economies where public resources are already scarce. It discourages people from using and paying for health services and ultimately has a 5See "Kosovo Internal Security Sector Review" (KIPRED) 2006, Sixth Annual Report of the Ombudsperson, 2005- 6.July 2006, and USAID/UNDP "Early Warning Report" Jan.-June 2006. 16 corrosive impact on the population's level of health. Countries with high indices of systematic corruption also appear to have higher rates of infant mortality. 32. Health insurers in low-income countries are susceptible to corruption. In Kenya, compulsory SHI for hospital services suffered greatly from poor management and corruption, with only 22 percent of the SHI funds actually used to pay for benefits and a large portion of the reserves lost through corruption. This has impeded Kenya's plan to launch a new national SHI system that includes rules for constraining the board's behavior, including a 5 percent limit on administrative costs, and a requirement for board members to abstain from voting on investments or contracts if they have any financial links to them. In Armenia, health financing was taken outside of the realm of the treasury before the health financing agency had the capacity to actually ensure that the money would reach service providers. In the Philippines, the Office of the Actuary estimates that 10-20 percent of claims sent by providers to the insurer are fraudulent. Preventing abuse and reducing corruption are important steps for increasing the resources available for health services, for making more efficient use of existing resources, and, ultimately, for improving the general health of the population. 33. Existing research indicates that the health sector appears to be particularly vulnerable to corruption (CMI 2006). Experience shows that processes that have a high inherent risk of informal payments and kickbacks are the following: (i) the provision of services by medical personnel; (ii) human resource management; (iii) drug selection and use; (iv) procurement of drugs and medical equipment; (v) distribution and storage of drugs; and (vi) budgeting and pricing of the health financing system. In addition, the following characteristics of health sector operation increase the level of vulnerability in the health sector (Transparency International 2006). ˇ An imbalance of information in health systems: Health professionals have more information about illness than patients, and pharmaceutical and medical device companies know more about their products than public officials entrusted with spending decisions. Making information about prices transparent and widely available can reduce losses to corruption. A study from Argentina has shown that the variation across hospitals in prices paid for medical supplies dropped by 50 percent after the MOH began to disseminate information about how much hospitals were paying for their supplies. ˇ The uncertainty in health markets: Not knowing who will fall ill, when illness will occur, what kinds of illnesses people get, and how effective treatments are poses another challenge for policymakers, rendering it difficult to manage resources, including the selection, monitoring, measuring, and delivery of health care services and the design of health insurance plans. The risk of corruption is even higher in humanitarian emergency situations when medical care is needed urgently and oversight mechanisms are often bypassed. ˇ The complexity of health systems, particularly the large number of parties involved, exacerbates the difficulties of generating and analyzing information, promoting transparency, and detecting and preventing corruption. The 17 relationships among medical suppliers, health care providers, and policymakers are often opaque and can lead to distortions of policy that are bad for public health. 3.3.2 The health financing system and governance 34. Empirical research indicates that how and where corruption appears in the health sector depends to a large extent on the institutional arrangement for health financing (table 3.1). This is not surprising, as much of the corruption found in the health sector is a reflection of general problems of governance and public sector accountability. In fact, it is reported that institutional and organizational impediments have been the most serious obstacle to health financing reform in Eastern European and Central Asia in the past (Langenbrunner 2005). 35. A health financing system that directly finances the supply of services appears to be more vulnerable to corruption in procurement and to abuses that undermine the quality of services. On the other hand, a system that uses an insurance institution to finance health care is generally more vulnerable to diversion of funds. In addition, different types of insurance systems impose different types of accountability risks (table 3.2). Table 3.1 - Health Financing and Risks of Corruption Financing method Characteristics Corruption risks Taxes Normally associated with free or Large-scale diversion of public funds at almost free service delivery ministerial level If health costs increase, low-income High risk of informal and illegal payments countries may not be able to raise for treatment taxes sufficiently to pay for the increase Corruption in procurement The rich often receive a disproportionately high share of public subsidies SHI Compulsory, usually not every Most common abuses are excessive citizen eligible for coverage and medical treatment, fraud in billing, and benefits; premium and benefits diverting of funds described in social contracts (laws or regulations). Usually only for formal employees Private Insurance Buyers voluntarily purchase Same as for social insurance schemes insurance (can be done on individual or group basis) Out-of-pocket Patients pay providers directly for With weak regulatory capacity, high risk of payment goods and services. Costs are not overcharging and inappropriate prescribing reimbursable. of services. Also risk of employees pocketing official fees collected from patients. No guarantee that all health services are of value to those buying them Source: CMI 2006 18 3.3.3 Health insurance models and governance 36. Different models of health insurance have different types of governance risks. The insurance-related corruption risks (table 3.1) can be grouped by different health insurance models (table 3.2). When a single insurer provides care, the government may need to decide on its organizational structure, as there is no market mechanism by which to impose an effective and transparent model. For example, including government representatives on the governing board would be desirable. Oversight is needed to counterbalance the lack of incentive for a single provider to control costs and premiums; in contrast, provision by multiple providers could use competition to check undue cost inflation. Table 3.2 - Health Insurance Models and Implications for Accountability Number Selected implications for Model of Provider payment accountability mechanisms insurers - May have soft budget constraints Public - Lack of benchmarking information Direct administration Single - Risk of capture by providers provision and/or internal - Oversight requires political or contracting economic counterweight to the insurer - May have soft budget constraints Monopsonist - Lack of benchmarking information Single Single negotiating with - Risk of capture by providers payer multiple providers - Oversight requires political or economic counterweight to the insurer - Possible to rely on associations for Negotiation between overseeing certain aspects of Corporatist Multiple representative performance associations - Need to assure legitimate process for selecting representatives - Possible to elicit information about costs through comparative analyses Various forms of - Possible to rely on shareholders for contracting Regulated ensuring efficiency Multiple providers with market - Consumer protection procedures need different payment to be in place setting processes - Risk that insurers may "capture" regulator Source: Savedoff and Gottret 2007 3.3.4 The search for a good governance structure for a health insurance system 37. In most countries, the introduction of a HIF has suffered from a weak and ineffective governance system. Most countries have taken several years to implement SHI after introducing the concept. The reasons for this delay are closely related to governance structure and capacity, poor financial management, lack of guidelines for the 19 investment of insurance funds, weak banking systems, and incidents of fraud and the misappropriation of funds. 38. In addition, excessive levels of administrative bureaucracy have hampered performance, especially where the percentage of the enrolled population was low. The administrative costs of SHI funds varied throughout Central and South Eastern Europe. Estonia, Lithuania, and Poland, which operate single HIFs with regional branches and almost universal coverage, reported administrative costs of 1-2 percent of revenue.6 Albania, in contrast, spent 7 percent on administration of its health insurance scheme in 1998 while achieving a coverage rate of only 68 percent of its population (Langenbrunner 2005). 39. The search for the right insurance model includes the process of translating principles of good governance for public institutions into main institutional elements of health insurance. External factors also must be considered (figure 3.1). The next sections discuss the analytical framework for applying basic governance principles to the main institutional elements of health insurance models. Figure 3.1 - Analytical Framework for Good Governance of Health Insurance Institutional Elements of Governance principles Social Health Insurance External factors Rule of law Ownership and Fair legal status competition Transparency Mechanism of representation Public trust Consistency Mechanism of supervision Political Accountability process Management (fiduciary, human resources) Inclusiveness, Information Demand for participation requirement governance 3.3.5 Principles of good governance 40. The governance mechanism in insurance is often evaluated in terms of how effectively insurers fulfill their mandates to provide financial protection and access 6In Poland, administration costs of Sickness Funds in 1999 were 1.46 percent of total fund costs; in 2000, costs were 1.36 percent, and in 2001, 1.72 percent. 20 to health services for their members. This mandate can be divided into five general principles7 that are often used to define good governance for governments, corporations, and financial markets, and which also apply to SHI (see figure 3.1). ˇ Rule of law Rule of law requires that regulatory authority is legitimately exercised only in accordance with written, publicly disclosed laws adopted and enforced in accordance with established procedure. ˇ Transparency Transparency is a means of holding public decision makers accountable by making rules and decisions available for everyone to see, opening critical meetings to the public, and reviewing budgets and financial statements in an open manner. ˇ Consistency Consistency helps to avoid uncertainty surrounding rule-making and enforcement over time and through periods of political change. Stability buttressed by legitimate consistency is particularly important for health insurance because insurance necessarily entails commitment over time and must be financially sustainable over generations. ˇ Accountability Accountability ensures that decision makers and institutions are answerable and responsible for their actions. While transparency makes institutions' and individuals' performance visible, accountability involves consequences - rewards or sanctions. ˇ Inclusiveness, participation, and consensus orientation Inclusiveness and participation ask that stakeholder views be incorporated during the process of decision making. This stems from the belief that stakeholder views are integral to meaningful governance and that stakeholder participation adds value. 3.3.6 Main institutional elements of health insurance 41. There are many different ways to analyze institutional arrangements for a HIF. This section adopts a framework developed by Savedoff and Gottret (2007) around the five main institutional elements discussed below. Referencing the framework presented in figure 3.1, this section explains the importance of these five elements in determining governance effectiveness and risks and their implication for a Kosovo HIF. 7These governance principles are neither exclusive nor exhaustive. Rather they are the most commonly emphasized principles for governance of public institutions. 21 3.3.7 Ownership and legal status 42. Ownership of the SHI system creates a strong incentive for the owner to act in ways that will preserve or enhance their stake in the health insurance. Ownership of a mandatory health insurance system could be either in the hands of shareholders (as is the case with private firms) or of government (as is the case for publicly owned agencies). The most common case is more complex, even hybrid. In most Western European systems, ownership is shared among employers, employees, and the government, but the employer and employee shares are usually controlled by an association. However, the government may hold an implicit responsibility for keeping the insurer solvent, creating a situation of moral hazard and leading it to subsidize deficits rather than allow the inefficient and ineffective insurer(s) to fail. As a result, the deficits of social health insurers that are publicly owned agencies tend to be shifted to the ministry of finance; privately managed insurers are forced to actively manage their expenditures. 43. A common feature in Central European countries is widespread and growing indebtedness among health insurers and hospitals, which has assumed alarming proportions in recent years. With generous benefit packages (of ineffective and non- essential services) and excessive infrastructures inherited from the pre-independence era, and increasing pharmaceutical costs and expectations of higher salaries in recent years, expenditures on health have consistently outstripped available resources, leaving huge unpaid bills for services already delivered. For example, in 2003, the deficit of the Hungarian health insurer amounted to 1.6 percent of GDP (Chawla 2007). These insurance deficits have posed financial obligations for ministries of finance, and have led countries, among them the Czech Republic, Hungary, and Slovakia, to change insurers' ownership status to private ownership (see box 3.1). 22 Box 3.1 - Legal Status of Health Insurance in European Countries Germany has statutory health funds and private health insurance companies that also act as purchasers of health care. Health funds are corporatist, nongovernmental organizations, operating on a not-for-profit basis. German law requires professional management of funds. France has a main health insurance scheme (régime général) with a network of 16 regional offices that are not-for-profit organizations with their own boards and a degree of managerial autonomy. They supervised by the national fund organization. Lithuania has a single statutory HIF that is a governmental budgetary institution largely financed by general taxation. The Estonia Health Insurance Fund (EHIF) is an autonomous public organization. Hungary's National Health Insurance Fund Administration (NHIFA) is a not-for-profit organization closely supervised by the MOH. In February 2008, the Parliament approved the creation of 22 insurance companies with 49 percent of stakes owned by private investors. Health insurers in the Netherlands are private not-for profit organizations. In Switzerland, insurers can be incorporated under public or private law and take various legal forms, including associations, not-for-profit stock company, and mutuality.a The Czech Republic has nine HIFs and one national fund (General Fund). Insurers are public, not-for-profit organizations that have a degree of autonomy from the government. Slovakia has two public health insurers and four private for-profit health insurance companies that until recently were allowed to pay dividends to their shareholders. Source: a. Schweizerisches Bundesgesetz fuer die Krankenversicherung 1994, Stand 2005: Art 11, 12. 44. International experience implies that if a country has a well-functioning public sector, direct public administration might be the best option. Where the public sector is less effective, autonomous public institutions could be considered. However, in both cases special attention would be needed to ensure accountability and transparency. While there is no simple approach, examples of well-functioning institutions within a country could provide a useful reference for the form of ownership and legal status that are best suited for a health insurance system. Box 3.1 enumerates selected examples in European countries. SHI systems consist of statutory sickness funds organized along corporatist, industry, or geographical lines (Germany, France, Austria)8 and are overseen and tightly regulated by the government. 45. Insurance and claims management can be contracted out to a private sector insurer. In the United States and the United Kingdom, for instance, it is common for large companies or state agencies to self-insure and contract out claims handling and 8In Austria, membership is assigned either according to occupation and/or region, whereas in Germany, Switzerland and the Netherlands, the population has a free choice of fund within a region. 23 provider management to an experienced private health insurance company. This involves a contract between both partners that clearly describes each party's responsibilities. A similar approach could be selected by a country with limited insurance management capacity and a need for training future insurance experts on the job. 46. The legal status of a health insurance system also determines rules for its operation. For example, its legal status will determine whether it manages its personnel according to civil service or labor codes. It also determines the terms under which it can be sued. Another important implication of legal status is the system's cultural and organizational behavior. If it is protected as a public agency and its employees enjoy civil service protection, the organization tends to inherit both the advantages and the disadvantages of the public sector, whereas a SHI system with private sector status is more conscious of financial solvency and operational efficiency. On the other hand, if the health insurance system is managed by the private sector, the law should contain provisions that allow competition and more autonomous management. It should also include provisions for surveillance in case of private health insurance company. (See box 3.2 for Slovakia experience). 47. For Kosovo, an independent HIF with the legal status of an autonomous quasi-public financial institution may be most appropriate. As a single insurer, the social insurance fund might well be governed as a quasi-public institution, with the public interest of the provision of quality health service as a main objective. Reflecting the very positive experience with the organizational and management form of the Kosovo Pension Savings Trust (KPST), a future health insurance pool could be organized similarly as an autonomous entity, managed by a professional team and based on international insurance standards. It would be important to prepare an enabling legislative environment for the HIF by spelling out the specific mandates and responsibilities of the organization and the supervisory body. Special attention must be paid to transparency and accountability of managers and staff. Alternatively, and maybe as a short-term solution, a claims management contract with a private international health insurance company could be considered, which would allow building local capacity in insurance management and train local staff on the job. This form of private-public collaboration could also reduce upfront costs for the government in insurance management information system investment. 24 Box 3.2 - Slovakia: Health Insurance Companies and Surveillance Authoritya Whereas the Act on Health Insurance defines a `product' to be provided by health insurance companies, the Acts on Health Insurance Companies and on the Health Care Surveillance Authority defines the `institutions' authorized to perform public health insurance (health insurance companies operating as joint stock companies) and supervision of public health insurance (the Health Care Surveillance Authority, or HCSA). The insurance company act transfers health insurance to the sphere of private companies authorized to perform health insurance activities. The State still has the possibility to provide insurance via a health insurance company established by the State with the State being the only or one of several shareholders. The act defines the operation of health insurance companies, including strict budgetary constraints, transparent financial relationships, financial and personal responsibility, transparent bookkeeping, and obligatory external audit. It lays down the legal provisions related to equality of opportunity for all health insurance companies and providers of health care. The law requires health insurance companies to define and publish the contracting criteria pertaining to staffing and equipment in health facilities; quality indicators; and a quality system certificate. The surveillance act stipulates the establishment of a special authority, HCSA, as a legal entity responsible for the continuous surveillance of adherence to professional standards of public health insurance and of the scope and quality of the health care provided.b HCSA's independence is ensured by not placing it in the system of state administration bodies, and by not subordinating it to any higher body. The act stipulates that HCSA is accountable to the Parliament and the government only for activities entrusted to it by this law and, being mostly financed from the contributions of the health insurance companies, shall not be dependent on the state budget. HCSA is authorized to supervise health care providers, that is, to check whether they provide health care within the scope and quality as required by law. Source: (Act 581/2004), described in World Bank 2006b a. These legal reforms have never been fully implemented in Slovakia, and to some extent been reversed under the new government. b. "Standards" mean meeting the specific conditions for establishment of a health insurance company and subsequent adherence to the conditions relating to the maintenance of permanent solvency when carrying out the obligations of the health insurance company towards contractual health care providers. 3.3.8 Mechanism of representation 48. The specific mechanism to represent owners as well as consumers, employers, medical care providers, and, at times, disinterested parties, has an important impact on the overall SHI governance system. Decision making regarding the mechanisms for selecting and maintaining a governing body should consider the impact of their choice on the degree of independence of the governing body and the incentives that its members face in managing the insurance institution. 49. It is common for a SHI system to be governed by a board of directors or governors, whose members are elected or appointed by shareholders, employers, or beneficiaries. The terms of appointment also need to be carefully designed (short or long 25 terms, ethical standards, and compensation), as they will influence the way in which the board is managed. For example, when their terms are short, board members may be more sensitive to the needs of those who elected them and may be focused more on short-term performance than on long-term strategy. 50. Countries often put representatives of employers, employees, and government on governing boards, which has advantages and disadvantages. This tripartite representation is useful in bringing the formal sector contributors to mandatory health insurance to the table. However, it is often criticized for being overly concerned with financial issues and for not truly representing the interests of patients. As an alternative, a range of civil society organizations could be invited onto the board, which could enhance the task of controlling corruption and conflicts of interest. 51. For Kosovo, the governing board whose membership would combine professional staff and independent government representatives could be considered. The appointment process and the terms of operation of the board and each member would need to be clearly spelled out in the law. Moreover, including a few international experts on the board might be considered until sufficient Kosovar professionals emerge. Alternatively, the representative of the International Civilian Operation could be a member during the transition period. 3.3.9 Fiduciary and human resource arrangement 52. Like all institutions, the SHI fund needs strong internal management processes. As a financial institution, the SHI fund needs to pay special attention to fiduciary arrangements. Recent surveys comparing budget allocations with actual spending at the facility level in a number of developing countries have confirmed that a significant amount of public resources have been leaked (Reinikka and Svensson 2007). This is largely due to lack of financial accountability and an ineffective budget process. 53. An improved resource control and accounting system is key to fiduciary strength. The SHI fund requires strong accounting and control standards and a system based on transparency, comprehensiveness, and timeliness. It is usually recommended that a SHI system should apply accrual-based accounting in order to capture changes in economic conditions when they occur rather than when a cash transaction is made. Also, an effective internal and external auditing system is crucial to ensuring an acceptable degree of integrity and efficiency. Countries often contract private auditing firms to conduct external audits, as health insurance accounting often differs from normal public sector accounting and public sector auditors are not familiar with private sector accounting standards. 54. Budget transparency and participation support fiduciary management. Opening up public policies, practices, and decision-making processes to the public is a powerful way to enhance accountability. When information is shared with the public, public officials are more mindful of the consequences of any misconduct. As a SHI system usually involves a high volume of participants, oversight should include representatives from various demographic groups. This participation requires that 26 management provide information in a way that is easily comprehended by non-experts in public finance. 55. In addition, a capable, motivated, and law-abiding staff plays a key role in ensuring efficient and accountable internal management. When the legal status of the SHI scheme is public sector or it is a publicly owned enterprise, low civil service salaries and rigid human resource management regulations often impede recruitment and retention of the caliber of staff required for SHI. In response, countries often allow higher salaries for SHI employees - usually equivalent or close to those of staff, working in private financial sector - but take away the lifetime civil service employment guarantee. In addition, more specialized staff, such as accountants and information technology (IT) specialists, need to be compensated on par with the same positions in the private sector. 56. It is important that Kosovo promptly establish an efficient and accountable financial management process, including accounting, reporting, and auditing. An international accounting standard - with a few modifications if necessary - should be applied. The Office of the Auditor General needs to develop an effective strategic plan and the capacity to provide a reliable auditing system for the HIF. This calls for immediate capacity building in financial management, so that it will be in place from the beginning of the implementation phase. 3.3.10 Forms and scope of governmental supervision 57. Government supervision is an important way of holding insurers accountable for their performance. The main objective of supervision is to ensure an adequate capital base, sufficient reserves, and investment in appropriate financial instruments as well as to manage political risk. The degree and nature of supervision are often affected by the type of health insurance institutions and market conditions in which they operate. If the market functions relatively well, government supervision could - and probably should - be restricted to checking that contracts are legitimately fulfilled and basic fiduciary responsibilities are followed. If the health insurance system is monopolized and subsidized by public resources, the government may need to be involved in a more comprehensive manner, including in standardizing contracts, defining the basic health plan, requiring the insurer to accept any applicant regardless of health status, and setting premiums. 58. In terms of practical steps for supervision, both financial and health care quality supervision is performed with both ex ante and ex post measures. As noted repeatedly here, a HIF is fundamentally a financial institution; therefore, ensuring that sound financial principles are applied is necessary in order to avoid insolvency. At the same time, a HIF, unlike other financial institutions, engages in contracts whose quality of service is not easily monitored or guaranteed. In other words, effective supervision of a HIF requires ensuring both financial solvency and the provision of a high quality of health care. Regarding the timing of supervision, if the internal control risk of health 27 insurance institutions is considered moderate, an ex ante supervision using auditing is more effective and does not excessively interrupt operations. If the risk is high, the government may need to be proactive to prevent serious damages. 59. Insurance supervisory bodies are called a board, a council, an assembly, or a commission (box 3.3). This body might be a single-tier organization, or it could have a larger group that meets infrequently and a smaller executive group (perhaps called a bureau) that meets regularly. The primary duty of a board of a social security institution is to ensure that the institution is acting in the interests of its constituents (contributors, beneficiaries, and future beneficiaries) and in accordance with its authorizing legislation, and to consider the institution's strategy and developments. The law will define the selection of board members, how they are appointed, and the role they are expected to play. More detailed regulations and guidance may come from the relevant government ministry or department. Nominated or elected board positions generally are not full-time posts. Board members are not employees of the institution, but rather holders of public offices. The institution's chief executive officer (CEO) and senior staff normally attend regular board meetings (ILO 2005). Box 3.3 - Estonia: Supervisory Board of the EHIF The Supervisory Board of the EHIF consists of 15 members. The Minister of Social Affairs, the Minister of Finance, and the Chairman of the Social Affairs Committee of the Parliament are ex officio members of the Board. Parliament designates one Member of Parliament, proposed by the Social Affairs Committee, to be a member of the Supervisory Board. The government designates one official of the Ministry of Social Affairs, proposed by the Minister of Social Affairs, as another member of the Supervisory Board. Five members of the board are nominated by organizations representing the interests of insured persons and another five members are nominated by employers' organizations. The Minister of Social Affairs is the ex officio chairman of the board. Board members elect a deputy chairman from among themselves. The deputy chairman assumes the chairman's functions when the chairman is absent. The Supervisory Board, among other things, approves the budget, structure, and administrative procedures of the EHIF. It also reviews and approves the reports submitted by the Management Board and appoints the auditor of the EHIF. The board also is responsible for immediately informing the government of any considerable deterioration of the EHIF's financial situation including any expenditure that causes the budgetary cash reserve of the EHIF to fall below 5 percent. Source: The EHIF Web site, http://www.haigekassa.ee/eng/ehif/sboard/ 60. Good governance provides clear rules of management. In Switzerland, the supervisory body for health insurers is the Federal Council (executive). The Swiss health insurance law requires health insurers to be managed like an organization under private or public law, which includes standardized accounting and financial management, annual reporting (Art. 60), and minimizing its administrative costs through cost-effective administration (Art. 22). The German health insurance law sets strict limits on insurers' 28 administrative costs.9 Social security regulations for the European Union (EU) are shown in Box 3.4. Box 3.4 - EU Social Security Regulations EU member countries, and candidates for membership must follow Directives and Regulations relating to the field of social security. The most important of these are: ˇ Regulation 1408/71, is intended to coordinate in detail the various social security systems of the European Economic Area; ˇ Regulation 1612/88, covers freedom of movement and equality of treatment for those defined as "workers", and the rights they have to social security benefits as a consequence; ˇ Regulation 883/2004, on the coordination of social security systems (this will in due course replace 1408/71, but for the moment the two are operating in parallel); and ˇ Directive 79/7/EEC, on the progressive implementation of the principle of equal treatment of men and women in matters of social security. Source: ILO 2005. 61. For Kosovo, the Ministry of Justice might head the HIF supervisory board to ensure independence. Board membership might include representatives of government, Parliament, insured persons, and employer organizations. As pointed out by the experience of the Kosovo Pension Pillar II Fund, the inclusion of international experts on the board should be considered as a means of sharing experience. 3.3.11 Information requirement 62. A health insurance institution is made more accountable by providing comprehensive and timely information to the public and to regulators. A rich governance literature with empirical evidence shows that greater disclosure of information significantly enhances institutional accountability. Standardizing the information reporting formats allows consumers and regulators to hold a health insurance institution accountable on a timely basis and is a good way to avoid an undue burden of information gathering and compilation on the part of consumers and regulators. The type of information to be disclosed is determined by need to reflect governance strength, the health insurance institution's capacity to monitor financial and nonfinancial performance and information, and the degree of demand from stakeholders and society in general. For example, the Commonwealth Health Fund in the United Kingdom produces an annual report containing a Treasury report, financial statements, performance scorecards (box 3.5), and the main program highlights during the fiscal year.10 In Switzerland, the annual report and detailed statistics for all health insurance companies is published and available on the government Web site.11 In Estonia, the EHIF publishes an annual report that provides detailed financial statements, an auditor's report, and information on the 9SGB 5: Paragraph 4: Krankenkassen. http://bundesrecht.juris.de/sgb_5/__4.html. 10http://www.commonwealthfund.org/usr_doc/site_docs/annualreports/2006/index.htm. 11Bundesamt für Gesundheit (BAG): http://www.bag.admin.ch/themen/krankenversicherung/01156/02446/index.html?lang=de 29 utilization of health insurance benefits. Box 3.6 provides an example of such annual reports. Box 3.5 - Performance Scorecards, Health Fund Annual Report Source: The Commonwealth Fund Annual Report 2006. Box 3.6 - Illustrative Table from the EHIF Annual Report HIFs Source: EHIF Annual Report 2006. 30 63. As a HIF normally deals with a high volume of participants and transactions, a high-capacity IT system is usually required. This is a challenge for many developing countries, including Kosovo, where IT infrastructure and professionals are not readily available. A careful, comprehensive strategy needs to be developed from the early stages of health insurance implementation. One important issue to be resolved is interaction with other revenue collection agencies. Usually, tax offices maintain an information system with a tax identification assigned to every tax-paying citizen. A national pension system may have its own IT system, using its own or a tax identification system. It is often recommended that a SHI scheme should rely on existing information systems and should build an IT system that will handle only additional information required specifically for health insurance management. 64. In an insurance company, the health management information system (HMIS) often has the following five distinctive components (Streveler 2006): (i) membership database with sociodemographic information for each insurance member, based on which membership cards can be issued to identify beneficiaries; (ii) provider management systems with information on contracting arrangements made with each provider; (iii) financial management systems including accounting, budgeting, human resources, and other financially based systems12; (iv) claims database with patient invoices sent by providers to document members' care seeking behavior and payment made by insurance; and (v) systems including a utilization and quality management component to detect abuse and low-quality performance.13 These systems are often "housed" in the insurer's department of actuary and contracting, which is also in charge of using the data for performance analysis and rate-setting. 65. It is important to carefully sequence and phase in HMIS implementation. As described above, the HMIS is a complex system that has a significant bearing on the way that health insurance operates. It also has multiple stakeholders whose incentives are both negative and positive for such a system. Therefore, HMIS implementation should start with developing a comprehensive strategy with wider consultation and, based on the strategy, prepare the required legislation. Following steps include analyzing and standardizing the business process and intensive training for capacity building at the early stage of implementation. 66. It is critical that the Kosovo HIF be equipped with a HMIS. Without a well- functioning IT system and HMIS, the HIF would not be able to function. In terms of information, the HMIS should produce financial information such as financial statements and audit reports, and the HIF should provide basic performance information on its main contracts with providers, as well as annual reports on the overall performance of the fund. Individual insurance members should receive regular information on changes in the 12A single Chart of Accounts and National Health Accounts is important. 13This an emerging field. Examples of parameters to track the use of scarce resources include drugs/prescription, visits/patient/year, occupancy rates of various bed types, and visits/provider/day, and using those statistics to actively arrange, plan for, and influence the growth of health services. Quality issues need to be monitored and flagged in important areas, such as concurrent tracking of nosocomial infection rates in hospitals, and medical error rates among physicians. This requires an active partnership with the MOH. 31 benefit package and health plans, and the fund's financial and management performance. Finally, it is important to recognize that the HMIS has an important bearing on the way in which the HIF operates and the type of information that it produces, as it is closely related to the business process. 3.3.12 External factors affecting the governance performance of insurance 67. The experience of many countries suggests that the search for an effective governance mechanism should pay close attention to how well the proposed governance mechanism "fits" the HIF structure and its context. In other words, the principles of good governance need to be translated into action in the main elements of health insurance institutions with careful consideration of context. As outlined in the framework presented in figure 3.1, there are four external factors that affect the performance of a HIF: ˇ The presence of fair competition in the market and an effective regulatory framework ˇ The degree of public trust in public institutions ˇ The effectiveness of political processes ˇ The presence of an organized civil society that demands accountability in society. 68. In many developing countries such as Kosovo, fair competitionthe market mechanism that sorts out and promotes the most effective suppliersoften does not exist. A lack of entrepreneurs willing to take risks, an uncertain and weak regulatory framework for guaranteeing fair competition, and information asymmetry between suppliers and consumers that undermines the market selection process are often the reasons for this situation. A low level of public trust in public institutions makes it difficult to convince the citizens that a SHI fund will manage the participants' contribution with due diligence. A political system that operates on patronage instead of performance is likely to appoint unqualified persons to governing and supervisory boards, which leads to mismanagement. Finally, the absence of an organized and active civil society that demands good governance from the SHI fund allows the fund to be less responsive to participants and less responsible for fund decisions and management. 3.4 LESSONS FROM THE KOSOVO PENSION PILLAR II TRUST FUND 69. It is useful to draw lessons from an insurance fund that functions well in a ? country. In this regard, the experience of the Kosovo Pension Pillar II Trust Fund is useful. This section summarizes the trust fund's experiences and lessons that several assessment reports have deemed relevant to a SHI fund (Gubbels et al. 2005 and 2007). 70. The Pension Pillar II Trust Fund is a mandatory, defined-contribution, pension savings program. The program requires all working residents of Kosovo who 32 are 18 years of age and older, and who were born in the year 1946 or later to contribute 5 percent of their gross salary, matched by a 5 percent employer contribution. 71. It is generally believed that the governance arrangement for the Pension Pillar II Trust Fund has been a success. Its savings trust, the KPST, is an independent body with strong governance and supervision, established solely for the purpose of administering the savings component of the pension system. The KPST tracks contributions, invests pension assets, and arranges for the payment of pensions. It creates an individual account for each plan participant, and credits the account accordingly throughout the participant's working life. 3.4.1 Ownership and legal status 72. As a reserved power of the UNMIK, the KPST is an independent financial institution. This means that the Special Representative of the Secretary General (SRSG) retains the right to appoint government board members, and to approve by-laws, annual budgets, and the annual administrative fee. This reflects the special status situation of Kosovo. 73. It is clear that its independent status has effectively insulated the KPST from fluctuating macroeconomic conditions in Kosovo. Since its establishment in 2002, KPST funds have been protected from any temptation on the part of the government to tap into its savings account. The changing political situation in Kosovo is likely to mean that the authority will eventually be transferred to the PISG as the status discussion reaches closure. 3.4.2 Mechanism of representation 74. The KPST is overseen by the Board of Governors, whose membership consists of four professional members, one employer representative, one employee representative, and one UNMIK representative. The legislation clearly defines qualification for membership and the appointment process. 75. One of the challenges was to fill the professional board member positions due to the lack of local capacity in the required areas of expertise. As of April 2007, all four professional members are foreign experts, who as a group hold decision-making power. There is increasing pressure from the Kosovo government to appoint Kosovar professional board members, but qualified individuals have not been identified. This is also likely to be the case for the SHI scheme. It would be worth considering combining domestic and foreign experts at the beginning and only gradually moving to full domestic representation. 3.4.3 Fiduciary and other administrative arrangements 76. As a financial institution entrusted to manage contributions on behalf of members, the KPST has well-established basic measures for internal fiduciary arrangements. For the administrative budget, clearly defined laws and by-laws regulate the preparation, approval, and execution of the KPST administrative budget. This 33 includes the authority and responsibility of the SRSG, the board, the managing director, and other staff, as well as the compensation policy for board members. 77. Transparent and effective accounting and auditing systems are important elements of the fiduciary arrangements, but accounting and auditing capacity is lacking. As a public pension agency with investment activity, the KPST has adopted a private sector accrual-based accounting standard, which has helped account for changes in economic conditions. However, moving from a cash to an accrual basis for recording revenue and expenses has posed challenges because of the lack of pension accounting experience and knowledge. The Board of Governors is required to provide for an independent audit of KPST finances and operations on an annual basis, the result of which is provided to the Banking and Payments Authority of Kosovo (BPK). Owing to the nascent status of the Office of the Auditor General, the audit has been conducted by an internal audit firm. 78. Another important element is accurate, timely reporting to participants. The KPST issues one account statement per year.14 To date, the KPST has issued five sets of accounts statements. The most recent issuance, in May 2006, prepared and mailed 193,000 statements directly to participant's homes. While issuing reports to all participants is costly financially and administratively, it is a necessary requirement, and building capacity for a timely and efficient reporting process deserves special attention. 79. In terms of staffing, the KPST recruited and maintains a small (initially 25) but highly motivated and qualified staff, who are compensated on a par with private sector financial organizations in Kosovo even though legally they are members of the civil service. A premium was paid for more specialized staff such as accountants and IT professionals. In this process, the board has approached the Kosovo diaspora through the international media and the Internet. 3.4.4 Forms and scope of government supervision 80. The authority to regulate the KPST rests with the Kosovo BPK, the overall regulator of financial services. The BPK's far-ranging powers allow it to oversee KPST operations to ensure that the best interests of participants are met. The KPST is required to provide various types of reports15 to the BPK on a regular basis. With the exception of the board member asset disclosure statements, all reports are also sent to the SRSG and are made available to the public. Other BPK responsibilities include regulating how the KPST calculates investment returns, and how it reports to participants on an annual basis. The BPK also has the authority to inspect the KPST. It is backed by enforcement authority, as detailed in the pension regulations. 14Initially, the KPST was required to report twice a year. However, an assessment of the level of effort and the cost of producing the statements led to a KPST/BPK agreement to reduce the requirement to once a year. 15Including: audited annual accounts statements; quarterly statements detailing the sale and acquisition of pension assets; annual statements of the value and allocation of pension assets; annual statements of fees paid to providers of financial services; asset disclosure information by members of the Board of Governors; and any other reports that the BPK finds necessary. 34 81. This hands-on supervision is considered appropriate and the role played by the BPK somewhat innovative. Given that the KPST is a sole operator in the market, and in view of the weak governance environment in which the KPST operates, this hands-on ex ante and ex post supervision has proved effective. Capacity building supported by the UNMIK and other international donors has allowed the BPK to play an effective supervisory role, in contrast to most countries with a similar system where the central bank has no supervisory power. This is considered an effective means of dealing with regulation in a territory in which regulatory capacity is limited. 3.4.5 Information requirement 82. The existence of a reliable HMIS for record and financial management has been a powerful tool for operational transparency, accountability, and efficiency. As the record keeper for more than 200,000 participant accounts, the KPST has a need for an advanced, secure, and stable HMIS, capable of processing a high volume of financial transactions (Gubbels et al. 2007). Early in the policy development stage, a decision was taken to pursue a strategy of unified collection, whereby the Kosovo Office of Tax Authority (TA) would assume responsibility for the collection and enforcement of pension contributions. This would take advantage of the TA's extensive network of regional offices and staff trained in collection, audit, assessment and enforcement. The approach allowed the KPST to develop as a centralized institution, with one office in Pristina and a relatively small staff, thereby maintaining a low administrative cost base. 83. While KPST staff members assist the TA with cleaning data, the KPST's primary role is to receive clean data for reconciled employer accounts from the TA. The KPST IT system has been designed to recheck the data provided from the TA for verification and reconciliation purposes. Once rechecked, the data can be attributed to the accounts of contributors. The KPST maintains a database of individual accounts. The participant retains the account throughout his/her working life, regardless of employment changes or periods of unemployment. For each quarter for which clean data are provided by the TA, each participant's account is credited for contributions as presented in the employer's quarterly report. In addition to contribution data, the KPST records data on date of birth, place of residence, mailing address, and beneficiary designation. Investment income and administrative costs are reflected in the KPST unit valuation. 84. Significant challenges were encountered in developing the KPST's IT/HMIS system, which could happen with the SHI fund. The process was long and challenging. The three most significant reasons for the delay were limited resources, changing KPST needs, and interaction with the TA's IT system. The development process stretched over two-and-a-half years and required considerable business process and programming revision to meet the needs of changing operational realities. The KPST now has an IT system that is functional but does not have the capacity and stability to support projects such as the implementation of multiple investment portfolios. The KPST experience is an important lesson for SHI-fund IT development, including 35 unwavering commitment, careful implementation sequencing and coordination, and capable IT staff. 3.5 CONCLUSIONS AND RECOMMENDATIONS 85. A sound governance structure is critical for health financing reform because of the high level of vulnerability in the health sector. No policy can be implemented as intended without a proper governance structure in place. Any successful reform, therefore, needs to take into account both policy and institutional aspects. A high degree of information imbalance, corruption in the health sector, and the highly decentralized and individualized health services that make it difficult to standardize and monitor service quality are the main reasons for the health sector's vulnerability to abuse and mismanagement. 86. Kosovo's weak overall governance makes creating a strong governance structure for a health financing arrangement challenging. Years of conflict and subsequent instability, the prolonged U.N. administration, and ethnic tension have all left Kosovar institutions highly personalized and lacking in proper checks and balance mechanisms. 87. Five principles of good governance could guide the Kosovo authorities in designing and implementing the appropriate governance structure for a SHI system. These principles are: the rule of law, transparency, consistency, accountability, and inclusiveness and participation. These principles are complementary and allow for a comprehensive and holistic approach to strengthen the governance framework. 88. In terms of ownership and legal status, Kosovo might create an independent HIF as an autonomous quasi-public financial institution. The Health Law No. 2004/4 provides the legal base to establish a single health fund that will pool all government health monies, and pay for basic care provided in public health facilities.16 According to the law, health care will be financed through a single fund and a central purchaser. The fund might well be governed as a quasi-public institution, with the public interest of the provision of quality health service as a main objective. This would reflect the positive experience with the organizational and management form of the KPST. A future insurer could be similarly organized as an autonomous entity, managed by a professional team and based on international insurance standards. Also needed is HIF enabling legislation that describes the specific mandates and responsibilities of the organization and the supervisory body. Special attention must be paid to transparency and accountability of managers and staff. 89. The Board of Governors membership would combine professional staff and government representatives. The law should specify the appointment process and terms of operation of the board and each member. Moreover, including international experts on the board might be considered until sufficient Kosovar professionals emerge. 16See Health Law Section 58 and 59, as well as Section 22 on definition of basic care. 36 Alternatively, the representative of the International Civilian Operation could be included as a member during the transition period. 90. It is important to establish an efficient and accountable HIF financial management process, including accounting, reporting, and auditing. International accounting standards - with a few modifications if necessary - should be applied. An effective strategy and capacity building in financial management should be high priorities from the beginning of the implementation phase. The Office of Auditor General needs to develop a plan and the capacity to provide a reliable auditing system for the HIF. 91. It is equally important to recruit and maintain a motivated and qualified staff. As with the KPST, staff should be recruited internationally among the Kosovar diaspora working in health insurance companies in other countries. A competitive remuneration policy, with salaries on a par with the private sector, may be needed, as well as a merit-based promotion policy and practice. This implies that HIF staff might not have civil service status but rather that of private sector employees. The draft civil service law will need to consider this situation. 92. Regarding government supervision, the Ministry of Justice or Finance might head the board to ensure independence. Board membership might include representatives of government, Parliament, insured persons, and employer organizations. As seen with the KPSF, seating international experts on the board should be considered, as a means of sharing experience. 93. The insurance fund would need to invest in HMIS to be able to make relevant financial and performance information available to decision makers and members in a timely manner. Legislation needs to mandate the types and frequency of reports that the fund should produce. Starting with financial information such as financial statements and audit reports, the HIF should provide basic performance information on its main programs. In addition to annual reports on the overall performance, individual insurance members should receive regular information on changes in the benefit package and health plans, and the fund's financial and management performance. Customer satisfaction surveys should be conducted regularly by an independent firm and results published in the local newspapers as well as on the Web site of the insurer. 94. The HMIS needs to be implemented early and staff trained in order to process the high volume of transactions and produce reliable reports in a timely manner. Without a well-functioning HMIS, the HIF would be handicapped. Implementing an information system goes beyond technical issues. It has an important bearing on the way in which the HIF operates and the type of information that it produces as it is closely related to the business process, which should be settled before IT equipment is purchased and installed. The lessons from the delayed implementation of the KPST should be taken seriously. 95. In addition to intensive capacity building, HMIS implementation needs careful sequencing and phasing. There should first be a comprehensive strategy for 37 HMIS implementation. New legal provisions are probably needed as is a consultative business process review and revision process. The purchase of IT equipment needs to be protected from abuse and mismanagement. Based on international experience and existing conditions in Kosovo, the total process of developing and implementing a HMIS across Kosovo is estimated to take at least 3-4 years. 96. Improving governance requires committed leadership, an enabling legislative environment, competent management and staff, and society's continuing demand for good governance. Introducing a HIF requires fundamental changes in the governance structure for health care services, and it is impossible to achieve this without a committed leadership that will put the public's interest ahead of individuals' interests. It also requires competent managers and staff who will persist in the task until the necessary system is set up and running by itself. Finally, health financing reformers would find it highly effective if the demand for good governance from civil society manifests as tangible pressure to stay on the course of implementation. 38 CHAPTER 4. SOURCES OF HEALTH FINANCING AND REVENUE COLLECTION 4.1 OVERVIEW 97. This chapter attempts to describe, analyze, and derive lessons from reforms to change the sources of funds or revenue collection arrangements for the health system in Kosovo. Kosovo is currently debating SHI as one mechanism to raise more funding for health. The chapter draws on experiences of countries in the former Yugoslavia, Eastern and Central Europe transition economies, the Organisation for Economic Co-operation and Development (OECD) systems in Western Europe and the countries of the former Soviet Union. A number of scenarios are presented for SHI and a number of alternative options are developed as well. Conceptual differences between "collection" and "sources of funds" are often blurred and are summarized in figure 4.1 below. A collection agency such as an insurance fund can receive revenue sources in the form of different taxes, voluntary contributions, and grants. Figure 4.1 - Unpacking "Sources" and "Collection" Sources of Initial funding revenues/contribution Collecting sources mechanisms agencies Individuals/ 1) Direct taxes Central families/ government employees 2) Indirect taxes Local Employers/ government 3) Payroll taxes corporate Social Security entities 4) Other compulsory agency Foreign and contributions (mandates) Commercial domestic NGOs insurance fund and charities 5) Voluntary prepaid contributions Other Foreign insurance fund governments 6) Direct payment to and multilateral Employers providers at time of use agencies Earmarked 7) Grants savings fund Foreign and multinational Health care 8) Loans companies provider Source: Kutzin 2002 98. The main sources of revenue for health care are taxes, social insurance contributions through payroll taxes and other revenues, insurance premiums and user charges (formal and informal). The positive and negative features of each source are summarized in table 4.1. Most countries in the region rely on a mix of these sources. Taxes are compulsory for the whole population and are levied by government. Social insurance contributions are compulsory for all or some of the population; these are often 39 kept separate from other government revenues and are usually managed by a fund or funds independent of government. Table 4.1 - Advantages and Disadvantages of Different Methods of Revenue Enhancement Method of revenue Advantages Disadvantages collection Direct taxation Wide revenue base (all income) Compliance may be difficult Administratively simple Allocations subject to political negotiation Usually progressive and promotes solidarity Potential tax distortions Large risk pool Allows flexibility in spending across sectors Universal coverage Indirect taxation, for Visible source of revenue (all transactions) Potential tax distortions example, value-added Administratively simple Allocations rely on consumption levels Compliance easy Usually regressive taxes (VAT) SHI Earmarked for health Compliance difficult Separate from other government revenues Increases costs of labor and may reduce (May) Link contribution to benefit international competitiveness Low resistance to increases Revenue follows economic cycle Independent management of funds Strong regulatory framework May allow choice of insurer Narrow revenue base (only applies to earned income) VHI May allow choice of insurer Strong regulatory framework needed May relate payment to utilisation. Adverse selection (results in escalating premiums) Risk selection (leaves some uninsured) Access related to insurance cover Usually regressive User charges Relates payment to utilization May deter access to necessary services Access related to ability to pay Regressive Limited pooling of funds (no prepayment and no pooling among different risk groups) Source: Adapted from Dixon et al. 2002 99. All countries need to raise revenues for health in the context of social objectives, namely, to improve health outcomes and to better ensure financial protection of the citizenry. The "right mix" of revenues will vary, depending upon the social objectives and macroeconomic context. Any country can further benefit from clear criteria in identifying their optimal mix of revenues. These criteria include: ˇ Economic efficiency: do changes in revenue flows hurt production capacity or relative efficiency in production and allocation of goods and services? When governments tax the private sector, the economic cost in most cases exceeds the amounts of resources extracted. This efficiency cost results from the production and consumption inefficiencies associated with taxes that distort decisions made by firms and households. Excess burdens of taxes are directly related to the behavioral response of the taxed entity and the tax rates. The greater the responsiveness and rates, the greater is the excess burden. 40 ˇ Equity: is there an adverse impact on the lower-income groups, the poor, or vulnerable groups? ˇ Administrative simplicity: successful changes in tax policy can be achieved only through effective tax administration. Effective tax administration and policy means eliminating unproductive taxes and keeping differential tax rates and provisions to a minimum. ˇ Transparency: having clear tax provisions, and effectively communicating them. ˇ Revenue levels: potential for actual revenue generation. 100. Most insurers in ECA report some mix of financing sources. In the ECA region, the scope of reforms in the 1990s actually implemented in collection was rather narrow - principally the introduction of payroll taxation for new compulsory SHI. Many of the countries in the region come out a Soviet tradition in which they had a Shemasko- style NHS, a social insurance system that pools risks through general revenues for the entire population. In the early part of this decade, however, more reforms were initiated related to general revenues, and health systems relied more on a mix of payroll and general revenue financing, or an almost complete reliance on general revenues. While payroll taxes often depended upon levels of contributions and effectiveness in collection, the later reforms relied more on a diverse mix of other policy instruments and changes, including income-based premiums. 101. Effects of efforts to change the sources of funds should be assessed in terms of their impact on the overall level and mix of funding from all sources. Key criteria are listed above, but specific features found in Kosovo will further affect the feasibility of increasing the level of (especially public) funding for health including such factors as: ˇ Taxation capacity; ˇ Extent of formality of the economy and labor force; ˇ Capacity of the "health authorities" (usually the MOH, but also possibly one or more insurance funds) to lobby and influence the government for a larger share of public spending; and ˇ Extent of fiscal decentralization. 4.2 KOSOVO: A CURRENT ANALYSIS 102. Health financing reform is at the top of the health policy agenda in Kosovo. Health financing is a concern because the Kosovo health system produces the lowest health outcomes among countries in South Eastern Europe. The concern is that current levels of funding are inadequate. 4.2.1 Total health care expenditures 103. In terms of GDP, total health spending in 2006 is estimated at 6.7 percent. Public sector (or government) spending on health (71.4 million) was 3.2 percent of GDP (table 4.2). These levels of funding are thought to be low, and do not include key elements of health spending such as military health expenditures, health expenditures in 41 prisons, nonprofit organizations, expenditures of Kosovo residents on health treatment abroad, and health care services for Kosovo residents financed by the Serbian government. Informal payments for services are thought to be widespread. Historical levels of health financing may be higher and very volatile over time. 104. The current health system is mainly tax funded from general revenues. Government health funds are allocated from the central budget to hospitals. The MEF also transfers a health grant for PHC to municipalities; municipalities are responsible for co-financing PHC. Public health facilities also receive revenue from off-budget donor funds (direct project support). Patients contribute approximately 29 percent of total health finances in form of user fees, paid for drugs and health care services. Table 4.2 - Kosovo: Health Financing, Main Indicators 2001­06 2002 2003 2004 2005 2006 Total health expenditures ( million) 117.43 116.11 131.7 149.5 151 O Public spending 55.0 55.2 62.3 74.6 71.8 O Private out-of-pocket spending, estimates 46.2 47.9 62.3 69.3 72.8 O Donor spending (off-budget), estimates 16.2 13.0 7.1 5.6 5.6 Per capital health expenditures () 62.9 61.1 81.5 57.6 57.1 O Public 29.5 29.1 45.5 28.7 27.1 O Private out-of-pocket 24.8 25.2 32.3 26.7 27.5 O Donors 8.7 6.8 3.7 2.2 2.1 Total health expenditures (% of GDP) 4.7% 4.7% 6.2% 6.8% 6.7% O Public 2.2% 2.2% 3.5% 3.4% 3.2% O Private out-of-pocket 1.9% 1.9% 2.5% 3.2% 3.2% O Donors 0.7% 0.5% 0.3% 0.3% 0.2% Expenditure on health (% of government budget) 12.82 9.89 10.4 9.7 10 Sources: World Bank 2005, using reported sources of MEF, LSMS 2000, HBS 2002; MTEF 2008-2010; MLGA Note: Total public expenditures slipped in 2005 largely due to the change in accounting method, from accrual to cash basis. 105. The government is the main financer of health care. Numbers on household spending should be treated with caution as 2005 numbers use a different methodology, and reliable household survey data and information from accounting systems are not available.17 This highlights the importance of implementing standard accounting systems in all health facilities and collecting reliable household survey data based on population census, to document treatment seeking behavior. Donor project support has declined, contributing only 11 percent of total health finances in 2003 and less than 5 percent in 17While National Health Accounts (NHA) places out-of-pocket payments at 29 percent (based on HBS data), official statistics from the Ministry of Finance estimate out-of-pocket payments to constitute nearly half of the health financing envelope (46 percent in 2005, 48 percent in 2006), although this is based on the 2002 LSMS household survey and applies a simple 2 percent growth rate per annum. 42 2004 and 2005 (figure 4.2). In 2003, donor funds were used to supplement resources for the public health sector, and financed the MOH administrative costs at the central level (64 percent), the Institute of Public Health (26 percent), and the provision of care in public health facilities (10 percent). Figure 4.2 - Sources of Health Finances, 2000-2005 (Percent of Total) 100% 80% 60% 40% 20% 0% 2000 2001 2002 2003 2004 2005 Donors Household Government (Consolidated) Source: MEF, World Bank Staff estimates; NHA MOH/World Bank 2007 Note: Change in methodology from 2004 to 2005 106. As a share of GDP, total health expenditure is also at the lower end of the range reported by neighboring countries in 2004, but comparable to countries at similar GDP levels. While Kosovo spends considerably less on health than many of its SEE neighbors, relative levels of public expenditures is more or less in line with countries at similar levels of GDP (see figure 4.3). Figure 4.3 - Public spending on health at different levels of per capita GDP GOVERNMENT HEALTH SPENDING VS INCOME 15 P) D 10 G Serbia-Montenegro %( Czech Republic Turkey gn Croatia 5 Belarus Hungary Republic of Moldov aBosnia Lithuania and Herzegov inaPoland Ukraine Latvia pendi Turkmenistan Russian Federation Romania S Albania Uzbekistan ht Kosovo Ky rgyzstan Kazakhstan eal H Georgia Armenia ent Tajikistan nm Azerbaijan over G 10 100 250 1000 2500 10000 25000 GDP per capita Source: World Dev elopment Indicators, WHO 2007 Note: GDP per capita in current US$; Log scale 43 4.2.2 Out-of-pocket payments 107. Out-of-pocket payments for the region tend to track with levels of GDP per capita (figure 4.4). Out-of-pocket payments relative to public spending may fall over time as GDP grows though the specific relationship varies by country, as do the specific solutions. Figure 4.4 - Private Household Out-of-Pocket Payments in % of Total Health Expenditures 90 80 70 60 t 50 cen erP 40 30 20 10 0 Rep ia gary istanolandmaneration ia hstanulgarUkraineLa ia H p ia eniabaijajikistan n Czech Cro Est loven thua BelaSlovaHun rkem atia nia o nia rus kia P S Li sianFed Ro B tvia sovo Biekistan niayz ReGeorgArm Ko Albarg Azer Ta Tu Kazak Uzb Ky Rus Source: National Health Accounts, WHO 2007; MEF 2007 108. High out-of-pocket payments raise concerns about access to care for the vulnerable where poverty is widespread. Mendola (2007) using 2000 LSMS household data from Kosovo found that out-of-pocket payments were higher in urban than in rural areas and that spending on private sector services was 1.6 times that for public services on a per capita basis. Transportation costs were significant, at about 17 percent of total costs on average. The study found that economic status is not significant in shaping health care demand in Kosovo. Households pay more or less the same for health care across the income distribution, but the poor are hit hardest in terms of share of overall consumption. Expenses for health care represent 13 percent of total consumption of the lowest income quintile versus only 4 percent for the highest income quintile (figure 4.5). 44 Figure 4.5 - Out-of-Pocket Spending as a Share of Total Income richest quintile 3rd quintile poorest quintile 0 2 4 6 8 10 12 14 Source: World Bank Kosovo Poverty Report 2007 Note: Income was proxied by consumption 109. Health spending pushes some near-poor households into poverty. The HBS results of 2003/4 and again in 2005/6 show that for those who seek care, households classified in highest quintiles pay higher average amounts on health than those in lowest consumption groups, and health financing is regressive, with the better-off paying a smaller proportion of income on health than low-income groups (figure 4.6). The most recent HBS shows that the average household pays about 5 percent of its total expenditures to health care. This is very high and it is close to the catastrophic level of 7 percent out-of-pocket expenses in Albania, for instance. In 2005/06, the bottom quintile households increased the share that they spend on medical expenses by over 50 percent. The impact of health expenditures on the welfare status of households is high: simulations suggest that health expenditures increase the poverty rate by 3 and 4 percentage points for total and hospital expenses respectively (table 4.3). Figure 4.6 - Out-of-Pocket Expenditures for Health as a Share of Household Consumption 2 0 0 3 / 0 4 2 0 0 5 / 0 6 8 .9 es noi urti pt um pend nsoc 6. 1 40/ 5 .6 03 ex 60/ H htl 5 .1 5. 0 5 .0 ean 05 H 4 .8 M of 4. 4 4 .3 hea 3 .9 etavirP entcr pe as P o o r e s t 2 3 4 R ic he s t P o o r e s t 2 3 4 R ic h e s t G r a p h s b y w a v e Source: World Bank staff calculations, Kosovo Poverty Assessment, 2007 45 Table 4.3 - Poverty Impact of Health Expenditures Health Pre/post 2003/04 2005/06 Expenditure payment Poverty rate Total Pre 35.2 35.7 Post 37.1 38.9 Source: World Bank staff calculations from HBS data. Note: Simulations assume 100% propensity to consume out of that payment. Poverty rate calculated among households making the expense. 110. Out-of-pocket payments mainly go to drugs and private sector care. A major concern with out-of-pocket payments is the unnecessary prescription of drugs that are not included in the list of "free" drugs. In addition, essential drugs are often not available in health centers or hospital pharmacies and have to be purchased in private pharmacies. As a result, a high proportion of out-of-pocket payments is made for drugs that are supposed to be free. According to the HBS, the cost of drugs accounts for 65 percent of private expenditures. Population groups who can afford to pay higher fees seek care in the private sector. A growing private health sector also suggests that private providers respond to patients' demand for quality care. 111. The second most expensive out-of-pocket payment item is "gifts to providers," which are nontaxable income for physicians and health workers, highlighting the issue of corruption in health. Informal payments seem to be more prevalent in health facilities with Albanian workers. A survey conducted in 2003 found that 53 percent of the Albanian respondents but only 8 percent of the Serb respondents said that they had to make unofficial payments for health services (Early Warning Report No. 3, 2003), a situation that may be related to differences in provider attitudes, salary levels, funding of care and public accountability. Paying informal fees to providers is typical in the region as shown by findings from Albania (22 percent of patients pay informal fees), Armenia (91 percent), Slovakia (60 percent), and Poland (78 percent) (Lewis 2000). In the Kyrgyz Republic, the government lowered informal fees by introducing a formal co-payment policy, which includes defining and publishing official price lists for drugs and services and implementing sound financial management in all health facilities. 4.2.3 General government financing 112. Government funds for health care come from general revenues reflecting about 3 percent of GDP. Public health funds are used to finance health expenditures at the central level (76 percent). The remaining 24 percent is transferred in the form of a health grant from the center to municipalities (table 4.4). Municipalities spend a limited amount on capital investment, while major investment is actually financed through the central public investment program (PIP). The health grant allocation to each municipality is calculated annually based on a simple capitation formula that takes into account the total population of the municipality. 46 Table 4.4 - Kosovo Government Expenditures on Health, 2004-07 ( million) Total government of which: Total government Total PHC health Municipality health spending grant in % Year spending PHC grant in % of GDP of GDP 2004 64.6 17.8 2.8% 0.78% 2005 71.4 18.0 3.2% 0.82% 2006 68.3 18.3 3.0% 0.81% 2007 70.1 18.7 3.1% 0.82% Source: MEF 2007 113. The population-based capitation formula is simple to apply, but contains weaknesses resulting in different capitation amounts (figure 4.7). The total PHC grant allocated to municipalities is theoretically a fixed amount per inhabitant. However, in reality, the per capita allocation across municipalities ranges from 6 to 10, which is mainly a result of negotiations between municipalities and central levels and not based on cost or patient risk assessment. Also, capitation pays for recurrent costs only as capital costs are financed from the PIP.18 Figure 4.7 - Health Grants to Municipalities, Per Capita per Year, 2005­06 16 14 12 year 10 cap/ 8 per 6 rou 4 E 2 - Istog ushtrri jë iti e V tok Pe llogoc linë K jakovë rizaj jilan biliq G O htim Lipjan sovë açanik enicë eçan eçan D rizren V G ragash verage P G Zv rishtinë D kenderaj Fe itrovicë odujevë htërpcë M alishevëS ahovec R A K am uharekë Z.Po S P S M F.Ko K P S Leposaviç ovobërdë N Source: MLGA 2007 114. Municipality revenues from the health grant are complemented by their own source revenues (OSR). These revenues are relatively small and come from formal co- payments (table 4.5). The MEF uses a rounded 80:20 ratio to budget the amounts that municipalities should collect as OSR targets, that is, the health grant allocation represents 80 percent of municipality revenue, while the OSR comprises the remaining 20 percent. The OSR collection and allocation requirements have been in place only since 200619 and in practice there are inconsistencies in how much OSR the municipalities can collect and 18 More detail on this issue can be found in the report by World Bank and USAID 2007. 19Administrative Instruction 06/2006 47 how able (or willing) they are to meet the 80:20 ratio. In the case that a municipality cannot raise the target OSR for health through co-payments, municipal budget OSR from property taxes and other municipal fees and levies may be drawn on to close the funding gap. In reality, OSR allocations to health are at the municipality's own discretion. If OSR collected through co-payments exceed the targeted amounts, however, the surplus can be applied to fund other expenditure items in the PHC budget, although only after approval by the MEF, or carried forward to the next accounting year. But, at the same time, there is no adjustment for municipality's capacity to raise revenues. Co-payments may impact localities differently and have a negative impact on the population in poorer areas. Table 4.5 - Public Revenues: Central and Own Source Revenues: 2004-2007 ( millions) Budget Sources 2004 2005 2006 2007 Central health budget 61.0 70.0 67.8 70.2 Central OSR -- 2.7 2.5 4.0 Local OSR 1.3 1.9 2.3 3.4 Source: World Bank, as reported in MEF 2007 115. At the mayor's level, revenues for health are allocated into economic categories, to pay for salaries and wages of health workers hired by the municipalities based on MOH staffing norms, for goods and services, and for capital projects (table 4.6). MOH staffing norms limit local flexibility to make staffing changes and thereby reallocate revenues to more cost-effective inputs. The amount that municipalities report as wages and salaries is an underestimate, as it does not include informal payments made by patients to the staff and health facility, and wages paid to additional nonmedical staff hired by the municipality (Gaumer 2007). Table 4.6 - Municipality Revenue and Expenditures on Health, 2005, by Economic Category Total PHC Municipal health Health grant Health OSR expenditures % expenditures () () () distribution Wages & salaries 13,847,663 - 13,847,663 66% Goods & services 2,116,458 2,945,317 5,061,775 24% Municipal expenditures 873,512 - 873,512 4% Subsidies and transfers - - 0% Capital outlays 1,203,605 91,491 1,295,096 6% Total 18,041,238 3,036,808 21,078,046 100% Source: MEF 2007 48 4.2.4 Comparing Kosovo public expenditures with ECA countries 116. Health expenditure as a share of public budgets is low in Kosovo compared with neighboring countries (see figure 4.8). As a percentage of total budgetary expenditures, Kosovo spends relatively the same on health as Albania, but less than other former Yugoslav governments. The MTEF report (MEF 2007) found that government expenditures on health fell from 10.4 percent of total government expenditures in 2004 to 9.7 percent in 2005 and again in 2006, which is also low relative to the average of 11.8 percent for 28 countries across the ECA region. Figure 4.8 - General Government Expenditure on Health ) dgetu GOVERNMENT HEALTH SPENDING VS INCOME B 30 nte rnmevo GlatoT 20 The former Yugoslav Republic of Mace Turkmenistan Lithuania Turkey Croatia Czech Republic (% Republic of MoldovaKosovo Serbia-Montenegro Hungary ng Latvia Estonia Ukraine Bosnia and Herzegovina 10 Kyrgyzstan endip Albania Russian Federation Georgia S Uzbekistan Armenia ht Tajikistan ale Azerbaijan H nte 0 mnrevo 10 100 250 1000 10000 25000 GDP per capita Source: World Development Indicators, WHO 2007 G Note: GDP per capita in current US$; Log scale Source: National Health Accounts, WHO, 2007; MTEF 2008-2010 4.2.5 Lessons from the EU and OECD countries 117. The OECD picture of health financing is diverse. The health sectors in Canada, Spain, and the United Kingdom are almost exclusively financed out of general revenues (figure 4.9), though there are significant differences in health spending levels, with Spain and the United Kingdom reporting the lowest proportion of health spending in terms of GDP and Canada reporting much higher levels of spending. In fact, health spending in Canada in terms of percentage of GDP and per capita is comparable to health spending in Germany and the Netherlands, where payroll tax-funded SHI is the main health financing mechanism. 118. SHI, using some mix of payroll taxes and general revenues, is the main funding mechanism in other OECD countries. Primary examples are the Netherlands, Germany, France, and Austria, with relative health spending ranging from 7.7 percent of GDP in Austria to 11.2 percent in Switzerland. Per capita health spending is similar in 49 these countries, with the exception of Switzerland, where SHI features also differ. Premium contributions are levied by individual insurers as a per capita flat rate (community-rated) in Switzerland, independent of the individual member's income. Figure 4.9 - Percentage of Total Health Funds Allocated through Public and Private Mechanisms, 2002 100% Out-of-pocket 11 5.8 16 9.7 24.5 16.3 3.3 0 14.3 26.5 80% 13.2 32 12.7 8.4 4.1 0 1.4 3.4 60% 10 Private health insurance 50.5 40% 83.4 72.8 67.5 71.4 70 20% Social health 19.6 17 insurance 0% 4.3 7.8 5.6 in y NL ) General taxes Spa (2001) ada tria Can ance man Fr (2001 UK Ger Aus Switzl Source: Country information; Country Health Systems in Transition reports and OECD in Figueras et al. 2005 119. SHI is financed by contributions from employers and employees as well as tax-funded government subsidies. SHI funds in Germany and Austria receive most of their revenues from payroll contributions; in France, payroll contributions reflect only 54.5 percent of total revenues, the rest being paid by government revenues. Payroll contribution rates vary across countries and in France have also decreased over time (table 4.7). Government transfers are either paid directly to the Insurance Fund as in France (box 4.1), or indirectly by reimbursing premium-paying households based on their tax-statement as in Switzerland. Table 4.7 - Financing of SHI, by Selected OECD Country Characteristics Austria Germanya Franceb Switzerlandc Contribution in % of gross earnings 6.9­9.1 14.3 average 5.25 Flat per capita - Employer share, % of contribution 50 50 94 0 - Employee share, % of contribution 50 50 6 100 - SHI contributions, % of total 80.8 100 54.5 81.8 - Subsidies from tax revenue 0 45.5 18.2 - Other revenues 19.2 Source: Country specific HIT reports. Note: a) HIT Germany: Government transfer to SHI starts in 2004. Average contribution rates are 14.3%, but vary across funds. As of July 2005, SHI financing shares are 54% for employees and 46% for employers; b) HIT France: In 2000, annual income threshold to be exempt from premium payment is 6600; c) BAG 2004: Households with annual income threshold below specific level pay SHI premiums and are reimbursed by the government when income is assessed during the tax filing season. 50 Box 4.1 - France and Financing of SHI Reducing Payroll Taxes to Protect the Labor Market The French government decreased payroll taxes for SHI contributions for employees in 1998 from 6.8 percent to 0.75 percent of gross earnings, while employer contributions remained unchanged. This decline is compensated by `general social contribution' (CSG) revenue that is levied as a tax on total income. It results in 45 percent of SHI revenue to be tax-funded. According to the 2001 Social Security Funding Act, the CSG consists of direct income taxes (rates are 5.25 percent of earned income, capital and winnings from gambling and 3.95 percent on benefits (for example, pensions and allowances), state subsidies and earmarked taxes (on car usage and tobacco and alcohol), and taxes on advertising paid by the pharmaceutical industry. In addition, the 1999 Social Security Funding Act aimed to encourage the employment of people with low skills levels and to prevent SHI contributions paid by employers from having a negative effect on the labor market, particularly to prevent an increase in unemployment for low-wage earners. Thus, the act decreased employers' contributions for low-wage earners (about two-thirds of employees), which was compensated by a tax levied on the profits of companies with a turnover of more than FRF50 million and polluting activities. 120. Different methods of health financing result in inequity in health financing (regressivity). Wagstaff et al. (1999) measure inequity in health financing using household survey data from early 1990s. Table 4.8 presents results based on the Kakwani index for four financing sources and selected European countries. Findings suggest that indirect taxation and out-of-pocket payments are regressive in all countries. Health financing is progressive under direct taxation. SHI is regressive in Germany, as high-income groups are exempt from SHI and enroll in private insurance. SHI is progressive in France and Switzerland mainly because all income groups are insured and low-income groups are exempt from paying contributions. In France, private insurance is regressive due to a higher enrollment among high-income groups20 who purchase private insurance to insure against co-payments under SHI. Similarly, in Switzerland higher- income groups are more likely to purchase private insurance to cover care and services not included in the basic benefit package. 20Buchmueller, et al. (2004) finds that private insurance enrollment tends to be higher for managers and highly educated professionals than for semi-skilled and unskilled workers, and is lowest for the unemployed. 51 Table 4.8 - Equity in Health Financing Financing UK 1993 Spain France Germany Switzerland 1990 1989 1989 1992 Direct taxes Progressive Progressive Progressive Progressive Progressive Indirect Regressive Regressive Rregressive Rregressive Regressive taxes SHI Progressive Regressive Progressive Private Progressive Regressive Regressive Progressive Regressive insurance Out-of- Regressive Regressive Regressive Regressive Regressive pocket payments Source: Wagstaff et al. 1999 4.3 REGIONAL LESSONS TO RAISE PUBLIC REVENUES FOR HEALTH IN KOSOVO 4.3.1 Dedicated taxes 121. One of the fundamental choices in mobilizing public revenues for health is the balance between dedicated and general revenues. Dedicated taxes are taxes earmarked for health and consist of contributions/payroll taxes to a mandatory health insurance/health-related social security fund or income tax revenues earmarked for health. The earmarking for health makes such dedicated taxes in principle independent of the annual budget process. While this protects health revenue, the dedicated taxes often being levied on wages makes them particularly sensitive to changes in wages and unemployment as well as interaction with the informal sector. Alternatively, health care can be financed from general revenues, which include all funds generated by central and territorial governments allocated through the budget process by more explicit political choice. In general, earmarked taxes are less desirable because they limit the fiscal flexibility of government. In addition, in Egypt, for instance, non-earmarked taxes are progressive, with direct taxes substantially more progressive than sales taxes. In contrast, earmarked taxes that are used in part to finance the SHI scheme are broadly regressive (Rannan-Eliya et al. 1998). 122. The payroll tax, in particular, has emerged and remains a standard part of a diversified source of health care financing in the region. Of the ECA region's countries, 18 have introduced payroll taxes--in 12 countries as a predominant mechanism of financing and in six others as generally "mixed funding systems" using the payroll tax to complement general tax revenues and out-of-pocket payments. The contribution rate has tended to be higher in the EU's New Member States in Eastern and Central Europe, and in the former Yugoslav states, ranging from 9 percent in Poland to 17 percent in Bosnia and Herzegovina. In many former Soviet states, the tax rate was typically set at 2-4 percent of payroll; therefore, the major source of funding remained general revenues. Wagstaff et al. (2007) reviewed the impacts in all 18 countries following the post-Soviet tradition. They found only minor increases in funding using a 52 payroll tax, but also that systems did not necessarily benefit from the increases--indeed, health outcomes had actually deteriorated. 123. Few countries have earmarked taxes for health. In Latvia and Lithuania, the tax was set as a proportion of the personal income tax. While it was abolished in Latvia in 2004, Lithuania still relies on an earmarked share of income tax in addition to its payroll tax. In only two other countries, Turkey and Romania, were there other types of earmarked tax set up for health: Romania has a sin tax and in Turkey additional funding (5 percent) has been available since 1988 from earmarked taxes on fuel, new car sales, and cigarettes. In one country, Kazakhstan, the payroll tax was levied twice and cancelled twice, and currently the health sector relies solely on general revenues. 124. Changes in sources of funding were typically linked with organizational changes, but did not necessarily lead to actuarial rate setting. These funding and organizational changes have been perceived as catalysts for further system reforms such as changes in purchasing arrangements, provider payment policy changes, and organizational changes to improve efficiency. However, in almost all countries, the contribution rate was not synchronized with a detailed actuarial analysis of expected costs and revenues for the insured population (Ensor 1999). Indeed, there were few health actuaries in the region in the early 1990s, and many public insurance organizations such as in Croatia and Slovenia still did not have actuaries on staff even up to 1995. Today, the rate-setting process typically reflects a combination of optimistic "eye-balling" of desired revenues and guesses about the political acceptability of adding to the already heavy tax burden on employers and employees. This has resulted in frequent changes in the setting of the rates, or in the funds being depleted for periods of time. 4.3.2 General tax revenues 125. Despite the switch to payroll contributions, general tax revenues continue to play a significant role in health care funding in many countries. In several countries - Armenia, Belarus, Ukraine, and Uzbekistan, for example - general revenues have been the mainstay of funding. In Kazakhstan, there has been a return to general revenues following two periods of mixed payroll and general revenues. The Russian Federation has considered the switch in the context of tax policy reform to bring down excessive payroll taxes and spur small business growth. In two countries, Armenia and Latvia, a social insurance fund was created entirely from transfer of general revenues, though in the last case revenues reflected an earmarked income tax (which is still counted as general revenue). 126. At the central level, general revenues or dedicated funds such as employment funds (Slovakia) or social funds (Kyrgyz Republic) typically fund nonworking populations. In all Commonwealth of Independent States (CIS) countries with mixed models, contributions are centralized - either from the very beginning or in the course of health reforms over the last decade. In Georgia, the Kyrgyz Republic, and Moldova, they are made by the central government. In Russia, it was first done by regional and local governments, but in 2006 only by regional governments. This pattern is seen as 53 increasing financial sustainability and pooling, though some argue that decreasing responsibility of local governments lessens accountability. 127. The prevailing approach in CIS countries and Central/South Eastern Europe is that the government pays what it can afford in terms of nonworking population transfers. The legislation in most of the CIS countries that have a mixed model sets commitments to specific groups of nonworking population, but usually does not set specific rates. In Russia, a few regions pay a symbolic amount (the equivalent of $0.30- 1.00) for nonworking groups. In Georgia and the Kyrgyz Republic, these allocations are specified but also at the symbolic level. The most viable approach is in Moldova, where per capita rate of contributions for the nonworking population is linked to the per capita contribution for the employed, and to the cost of the SHI benefit package. This link has substantially improved revenue collection but caused problems of sustainability. In Serbia, despite contributions on behalf of vulnerable groups being specified based on the minimum wage, general revenue allocations are still based on historic levels and are ad hoc in practice. Croatia covered shortfalls in such subsidies retroactively, rather than budgeting for the entire cost (World Bank, 2005). In Slovakia, central government transfers to insurance companies for the nonworking population (pensioners, unemployed, social assistance recipients) are set at 4.5 percent of the average wage of two years ago. 128. General revenues can further fund public health-based programs and capital investment. Funding can go to TB, psychiatric care, HIV/AIDS, and addiction programs (for example, FYR Macedonia, Latvia, Poland, Russia, Slovenia), as well as high-priority programs such as diabetes (FYR Macedonia) and organ transplants (Poland). Capital investment funds flow from the central level (for example, Kazakhstan, Russia, Slovenia, Uzbekistan,) and regional level (for example, Estonia, Hungary, Poland, Slovenia) or both levels (Czech Republic). 129. Local and municipal general revenues funds are relatively small, though there are exceptions. In Russia, more than 80 percent of all public funding comes from the regional level and is allocated from the regional level. The majority of public funding is from general revenues though the insurance payroll tax, which typically makes up more than 25 percent of total revenues. In Bulgaria, more than 40 percent of public funding comes from local budgets and the remaining 60 percent from a dedicated payroll tax. 4.3.3 Impacts to date: Challenges for enhancing revenues 130. Countries can be grouped into three categories for primary sources of health financing. ˇ Dedicated taxes, either through payroll or earmarked share of income tax; ˇ General revenues; or, ˇ Mixed funding systems including direct and indirect taxes. 54 131. There is a trend in the OECD countries to reduce or eliminate payroll taxes. Western Europe and other parts of Europe have been decreasing their reliance on dedicated taxes. Germany and the Netherlands are the only countries that cover around 60 percent of total health expenditure through dedicated taxes levied on wages. Spain and Italy have moved from payroll to general revenue financing in the last few decades. France, as outlined in box 4.1, mainly relied on wage-related contributions, but shifted to a wider base for contributions in 1997. It did this to encourage small and medium enterprises to hire more employees and to better ensure that low-income and low-skill groups would be covered by insurance. In Switzerland, health insurance contributions are independent of payroll and paid directly as a monthly premium by each insurance member to their insurance company. 132. Payroll taxes are also being reduced in transition countries. Countries such as Albania, Croatia, Hungary, Kazakhstan, Armenia, and Georgia have each taken steps to reduce payroll taxes. In Hungary this process started in 1999 when employers' health insurance contribution decreased from 18 percent to 14 percent. At the same time the employers' lump sum health care tax has doubled to the level of 18 per month per staff. Tax exemption categories have narrowed substantially and the ceilings on employees' wage-based contributions were abolished in 2000. Ukraine and Uzbekistan have placed a freeze on new payroll taxes for health. Spurred by petrodollars and corruption associated with its insurance fund, Kazakhstan has dropped payroll taxes while Russia in 2001 dropped the rate of the payroll tax (for the entire social security) from 3.6 percent to 2.6 percent, and a taxation base scale was set. The use of a flat premium per capita is now being discussed. 133. The success of dedicated taxes depends on economic growth. Dedicated tax- based systems do best in terms of mobilizing public revenue for health as a percentage of GDP when compared with mixed models and general revenue systems, as long as the economy and with it the wage-base are growing. During economic recession and a decreasing wage base, social insurance revenue from payroll will go down. In Estonia, which relies almost entirely on a dedicated tax, the HIF used its reserves to compensate for the reduced real revenues caused by the economic crisis in the Russian Federation and the global economic downturn. At the same time, Estonia finds itself in a longer-term shift to an aging population, with a dwindling wage base. Public expenditures for health as a share of GDP (See figure 4.10) and as a share of total public expenditure have decreased since 1998 due to changes in the unemployment rate and demographic changes, as well as increased use of tax loopholes to evade paying higher rates (Couffinhal and Habicht 2005). 55 Figure 4.10 - Estonia's Health Expenditures as a Share of GDP Estonia 7 6 P D 5 Gfot 4 3 cen erP 2 1 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Year Dedicated taxes for health as % GDP General revenues for health as % GDP Source: Estonia, National Health Accounts, 2006 134. Despite the hope of increasing overall levels of funding, the diversified funding base approach has often failed to bring about expected revenues relative to pre-transition years. This was especially true in countries of mixed financing relying on payroll taxes and general revenues. Figure 4.11 shows results from other countries relying on mixed sources of revenues in the 1990s and early part of this decade. In Kazakhstan, available public expenditures actually dipped with the introduction of a payroll tax and the formation of a "mandatory" (or universal public) health insurance system. In other systems, the impact of introducing a mix of payroll tax funding with general revenues has been similarly weak in terms of revenues. Figure 4.11 - Mixed Funding Systems 7 6 P 5 D Gf 4 o 3 centre P2 1 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Albania Bulgaria Georgia Kyrgyzstan Latvia Lithuania Republic of Moldova Russian Federation Slovakia Source: World Bank 2006; WHO 2006 4.3.4 Why did the use of the dedicated (payroll) tax fail? 135. Several factors have influenced the success of dedicated (or payroll) taxes. Some countries, such as Estonia, Hungary, and the Czech Republic, have structural 56 characteristics that increase the likelihood of successful introduction of a payroll tax, including relatively higher per capita income, and a large percentage of the population living in urban areas and working in the formal sector (Ensor and Thompson 1997). Given the large number of employees in the government sector or number of large state enterprises in many countries such as in Kazakhstan, registration was initially made easier. But there have been major challenges in many countries including reduced public funding, low compliance rates, and negative impact on the labor markets. 136. Ministries of finance have reduced general revenue contributions to health. A new dedicated tax may be perceived by a ministry of finance as a substitute and not a complement to general revenue funds. In Kazakhstan, the state reduced general revenues in direct response to the new payroll tax, which in turn, could not compensate for this reduction. It was subsequently abolished in 1998. There were similar cuts in Russia in the 1990s. 137. Contribution evasion is a major impediment for payroll tax-funded SHI, and causes actual revenues to fall short of expectations. The SHI payroll contribution collection levels varied across countries but were typically low relative to initial expected or estimated flows. Compliance ranged from 9 percent to 52 percent of expected revenue in different oblasts in Kazakhstan in 1996, rising to only 40 percent in 1998. Several countries currently report differences in actual versus expected revenues, resulting in contribution compliance rates of 50 percent in Albania, 50­80 percent in Serbia, 80-90 percent in Romania, and 94 percent in Bulgaria. Other countries - Romania, Kazakhstan, the Kyrgyz Republic, Russia, and even Estonia - have reported similar problems of collection of payroll taxes for SHI due to evasion by labor and small businesses. 138. Bosnia and Herzegovina reports contribution evasion as high as 70 percent and could not improve collection compliance by moving the collection authority from the SHI to tax administration. One of the most thorough studies was in Bosnia and Herzegovina. It found that social insurance revenue collections from payroll taxes and collection rates (figure 4.12) varied from 30 percent to 84 percent due to such factors as levels of unemployment and informal employment. The study found that collections had moved from the HIF, where there was a direct link of efforts and revenues, to the general TA, where there was a lack of incentive to collect. The TA, furthermore, had little power for sanctions, and there was little coordination across ministries, such as the Ministry of Labor, for targeting firms and workers. 57 Figure 4.12 - Bosnia-Herzegovina: Collection Rates by Canton, 2005 90 80 70 60 Collection 50 Rates 40 30 20 10 0 a oj an snia nia zla vina na eiva inje a-S jevo stHerzego Tu -Dob Podr Um stBo ralBos t Posavi ica ian Sara We Cen Zen ovina-Ner zeg Bosn We Her Cantons Source: Sanigest Internacionale 2004 139. A variety of other issues have limited collection compliance, including the lack of a clear link between contribution and benefits. The historical legacy of the socialist era was that many countries had a constitutional right to health care for all, which was retained in the transition period in most countries. This no longer exists. Low levels of compliance may have been further exacerbated because there is often no link between contribution and benefit, as well as weak collection mechanisms, indebted enterprises, and large numbers of the population outside the formal workforce, particularly farmers and the unemployed. 140. The evidence of payroll taxes decreasing labor inputs in transition economies suggests that high payroll tax rates discourage firms and workers from coming into the formal economy (World Bank 2007). Payroll taxes can negatively impact job growth and capital formation, which can have a negative effect on future economic growth. The region as a whole has traditionally suffered from much higher social payroll taxes than other regions in the world (figure 4.13). A review of labor markets in OECD countries indicates that tax rates are a significant factor in explaining differences in the amount of market work undertaken by the working age population. A 10 percent increase in tax rates could decrease labor inputs by 1-3 percent for the working age population. Unemployment remains high in many countries with payroll taxes, and compares unfavorably with unemployment rates the EU15 countries (figure 4.14). 58 Figure 4.13 - Social Insurance Taxes as Share of Total Labor Cost Eastern Europe & Former Soviet Union (13) High-Income OECD (17) North Africa & the Middle East (13) Latin America & the Caribbean (19) East Asia & the Pacific (16) Sub-Saharan Africa (29) 0 5 10 15 20 25 30 35 Source: Palacios and Pallares-Miralles 2000 Note: Number of countries in region in brackets Figure 4.14 - Payroll Tax Rates and Unemployment Rates: Comparison with EU15 Rates 50 45 40 35 30 25 20 15 10 5 0 ina lic ia ia ia kia 15 ov oatia ngary on egro enia nia-Herzog mania CrechRepub Eston Poland Serb ova ov EU Hu ced nten Ro Sl Sl Ma Mo Cz Bos Payroll Tax Unemployment Rate Source: World Bank 2006 141. Weak tax collection systems and large informal economies limit revenue raising capacity. The payroll tax is less successful in countries with weak regulatory and organizational capacity to raise revenues, a large proportion of unemployed or self- employed workers, and weak tax collection systems, such as in Romania and Albania. Major challenges exist in contribution collection in countries with large informal economies such as in Albania, where the informal sector generates up to 50 percent of GDP (figure 4.15). 59 Figure 4.15 - Level of Informal Economy as Percentage of GDP 60 50 40 30 20 10 0 a ia ia a a lic bani don lgar enegr land oveni ovaki ngary ania m Al oldova ont Po M ace Bu Sl Sl Hu Repub Ro M M and ch a Cze Serbi Source: World Bank, as cited by Waters et al. May 27, 2006 4.4 IMPLICATION FOR KOSOVO 4.4.1 SHI: Is it a good idea for Kosovo? 142. Improved access to care through insurance may lead to higher health expenditures, and revenues from a payroll tax are expected to be insufficient to pay for this expenditure increase. The draft health insurance law in Art 13ff defines a basic benefit package under insurance that covers primary, secondary, tertiary care, and treatment abroad. Roughly, the draft law promises covering everything that is currently demanded by people living in Kosovo. Once SHI is available, as in other countries with SHI, utilization of health care services might increase among the insured and particularly among the better-off. Greater funding could be needed to meet this higher level of utilization and thus prevent a deficit. The recent patterns of increased utilization under SHI in Turkey, Saudi Arabia, and other countries provide ample evidence of this effect. 143. The revenue base under a payroll-funded insurance is limited by a small formal sector and possibly high contribution evasion rates. Kosovo has a very small formal sector. About 29 percent of the working age population is employed in the formal sector and estimates of unemployment vary from 23 percent to 40 percent depending upon the source (Statistics Office of Kosovo 2005; MEF 2007, World Bank 2007). The regional experience suggests that in Kosovo payroll contribution evasion could be at least 40 percent of expected total insurance contribution, yielding significantly less revenue for the health sector than expected. 144. In addition, a payroll-based insurance reform in Kosovo could bring about all the negative impacts noted above on the following: ˇ job formation especially in small and medium enterprises, and thus in overall economic growth; ˇ the overall revenue base for tax and fiscal polices as fewer jobs are created; and 60 ˇ formalization of the economy, as growth in payroll discourages employers and employees as it makes them less competitive. 145. The equal contribution rate for workers resulting from payroll taxes is not ideal. Rather than reflecting the carefully determined medical needs of the population, or benefits promised by law, or targeted provider reimbursement objectives, a defined payroll tax level reflects a politically acceptable additional increment to the current tax rate. No variation is allowed for predictable spending determinants like the number of dependents or a change in benefits package or service use. In addition, payroll taxes raise equity concerns: in raising funds for health care the government collects the same share of payroll from the rich and the poor, although when raising funds for itself the government progressively taxes higher incomes at higher rates. Payroll taxes also fail to address income from assets. 4.4.2 Economic analysis: Estimates of financial results under payroll tax 146. Two relatively simple scenarios are presented on the possible performance of payroll tax for health in Kosovo. The health insurance law of April 2007 suggested that SHI be financed through payroll contributions levied on wages of formal sector individuals; however, it did not define the level of contribution rate. The following scenarios estimate the relative magnitudes for potential revenues from payroll tax-funded SHI, future health expenditures, and the resulting financing gap. According to the Kosovo Labor Market Statistics from 2005, about 29 percent of the working age population is working. Most are employed in the private sector (69 percent), while 25 percent work in the public sector and 19 percent in agriculture. The annual wage sum of this population group constitutes the contribution base for a payroll tax to a HIF (table 4.9). Table 4.9 - Contribution Base, 2005 estimates Sectors Public Private Agriculture Total % working in sector 25% 69% 18.8% Employed, per sector, number individuals, estimate 90,313 249,985 67,915 408,213 Annual average salary, per sector 1,800 3,600 720 Total annual wage sum: Contribution base for 162,562,500 899,946,000 48,898,800 1,111,407,300 insurance, per sector Source: Kosovo Statistics Office 2005 147. Contribution evasion is expected to be high in Kosovo. The following three scenarios all assume the same contribution evasion rate, namely 0 percent in the public sector, 50 percent in the private sector, and 80 percent in agriculture. These rates correspond to estimated rates in Bosnia and Herzegovina. The following two revenue scenarios are based on different payroll tax levels that constitute the contribution rate. 61 148. Scenario 1 assumes a SHI contribution rate of 10 percent of payroll and agricultural revenues. This rate is lower than in FYR Macedonia, Serbia, and Turkey (all around 12 percent). Taking into account evasion, the expected SHI revenue is 62 million (table 4.10). Table 4.10 - Scenario 1: Contribution Rate of 10%, and Health Expenditure equal to 2004 Public Private Agriculture Total sector sector sector SHI contribution rate/payroll tax 10% 10% 10% Expected SHI revenue () 16,256,250 89,994,600 4,889,880 111,140,730 Contribution evasion rate 0% 50% 80% Annual SHI revenue () 16,256,250 44,997,300 977,976 62,231,526 149. Scenario 2 assumes a 15 percent contribution rate on payroll and agricultural revenues; and the same evasion conditions as in Scenario 1. A 15 percent contribution rate would be one of the highest in Europe, and results in expected SHI revenues of 93 million (table 4.11). Table 4.11 - Scenario 2: Contribution Rate of 15%, and Health Expenditure 2004 plus 10% Public Private Agriculture Total sector sector sector SHI contribution rate/payroll tax 15% 15% 15% Expected SHI revenue () 24,384,375 134,991,900 7,334,820 166,711,095 Contribution evasion rate 0% 50% 80% Annual SHI revenue () 24,384,375 67,495,950 1,466,964 93,347,289 4.4.3 Balancing revenues and expenditure scenarios under SHI 150. Payroll tax-funded SHI will most likely not result in additional revenues for the health sector. Table 4.12 shows total revenues for each of the above scenarios by source of funding, and compares this revenue side to different expenditure scenarios. The year 2004 reported in the Public Expenditure and Institutional Review (PEIR) serves as a baseline year for the expenditure side. Each scenario assumes that a cost-sharing policy is in place for insured patients who use health care. The resulting out-of-pocket payments in the form of co-payments are estimated at 20 percent of total SHI revenues, similar to Bosnia and Herzegovina. Donor funding is ignored as it reflected less than 10 percent in every year since 2004. All three scenarios show that in Kosovo, payroll tax- funded SHI will not result in additional revenue, but will shift financing for care from out-of-pocket to insurance contribution and to some extend to the central budget. 151. Scenario 1 assumes the same total amount will be spent on health as in the baseline year, though this is highly unlikely. In 2004, total health spending from the Kosovo government, donors, and private households amounted to 157 million. In 62 Scenario 1, SHI is expected to raise 62 million through a 10 percent payroll tax. Insurance coverage with a generous benefit package and limited cost-sharing will cause out-of-pocket spending to decrease substantially, to 12 million. To achieve the same total health spending level as in 2004, central budget spending can decrease by 5 million, still contributing 53 percent of total health financing. As evidence from all countries with insurance introduction shows, the utilization of health care will increase as will insurance expenditures. 152. Scenario 2 assumes that insurance will lead to increased service use resulting in a 10 percent increase in total health spending compared to the base year. The revenue scenario is based on a very high payroll tax of 15 percent resulting in higher SHI revenue of about 93 million. Private co-payments are estimated based on a co-payment rate of 20 percent of the total price per service, contributing about 11 percent of total insurance revenue, or 18.6 million. To finance the increasing demand for care that resulted in a 10 percent expenditure hike, the Ministry of Finance would still have to contribute 61 million or 35 percent of total health spending, although this is about 20 percentage points less than the base year. 153. Scenario 3 assumes on the revenue side a 10 percent payroll tax rate, and 10 percent increase in total health expenditures from 2004 due to higher service use. Under this scenario, expected revenues from payroll amount to 62 million while revenues from cost-sharing are estimated at 7 percent of SHI revenues. Increased health expenditures due to the insurance effect will result in a financing gap of 98 million, which will have to be financed through the central government revenues. This will necessitate an increase of 11 million from the central budget contribution to health care over 2004. Table 4.12 - Scenarios 1-3: Estimated Health Spending by Source of Funding Baseline year (PEIR 2004) Scenario 1 Scenario 2 Scenario 3 % % % % Payroll 0 0% 62,231,526 40% 93,347,289 54% 62,231,526 36% funds Cost- 62,300,000 40% 12,446,305 8% 18,669,458 11% 12,446,305 7% sharing Central 87,900,000 56% 82,752,169 53% 61,156,253 35% 98,495,169 57% budget Donors 7,200,000 5% - 0% - 0% - 0% Total health spending 157,430,000 100% 157,430,000 100% 173,173,000 100% 173,173,000 100% 154. An increase in the public share of total health expenditure may lead to reduced central budget contributions to health and could possibly hamper contribution compliance. Payroll tax revenue for health is counted as public revenue. Under Scenario 3 the sum of payroll taxes and central budget financing for health care is estimated at 160 million, which corresponds to about 7.3 percent of unadjusted 2007 63 GDP. While this ratio is comparable to Bosnia and Herzegovina, it is considerably higher than current public spending levels in Kosovo, which hover around 3 percent of GDP. This relatively high rate may become an argument for the MEF to decrease the general revenue budget contribution to health. This, as noted above, has occurred in other countries such as Kazakhstan and Russia. If this cut in financing is shifted to the consumer and only results in higher co-payments, it will question the effectiveness of insurance and could cause an increased unwillingness to contribute, leading to contribution evasion. 155. This scenario analysis, it should be noted, is static and does not reflect expected patterns over time. It is limited in that it lacks adjustments for changing labor markets, employment, and competitive effects over time. A growing economy with a growing level of formalization could generate greater revenues. At the same time, increases in payroll could force smaller (more competitive) firms underground and could inhibit capital and labor formation. Insurance sources generally also attract investors such as pharmaceutical and high-cost medical technology firms, which will try to sell higher-priced products to health care providers, often resulting in higher health care costs and growing financial liabilities. 4.5 ALTERNATIVE OPTIONS FOR INCREASED FUNDING 156. Following the experience of Latvia and France, Kosovo might want to consider alternative funding options for its SHI. These include the use of an actuarial health financing model to estimate revenue needs, direct and indirect taxation, and individual premiums, discussed below. 4.5.1 Option 1: Develop a health financing actuarial model 157. Analytic work would be initiated to develop a health sector-based actuarial model for estimating both revenues and expenditures. The model would be utilized for sector-wide analytic work. 158. On the revenue side, the model should develop the analytic capability to estimate future revenue flows. Estimates should be conducted considering different sources of funds (contribution mechanisms), and under alternative macroeconomic and demographic scenarios of growth or contraction. Possible sources of revenues can then be considered and should be built into the model using the government's current tax policies as well as estimating new tax structures and direct contributions by individuals as is currently the case in Switzerland and the Netherlands. An impact model might be built to look at changes in the tax structure over time and under various macroeconomic scenarios. The pension reform program relied on general revenues, and a pension is available to anyone older than 65 years. Perhaps it is a good lesson for the health sector that similar types of revenues need to be considered. Box 4.2 sets out the current tax structure in Kosovo. Any new tax must be considered in the overall framework of macroeconomic and fiscal policies as these exist now. 64 Box 4.2 - Summary of the Tax Structure in Kosovo ˇ Income tax: progressive (5-20 percent), with a maximum ceiling of 20 percent. ˇ Municipal and nontax revenues: mostly fees and charges collected primarily by municipalities and some central budget organizations ˇ VAT: 15 percent for companies with annual turnover of more than 50,000 ˇ VAT: 15 percent to almost all imported goods in Kosovo, except goods subject to regulation 2004/35 ˇ Customs duty: in most cases 10 percent on goods imported in Kosovo, except for goods originating from countries from which were signed free trade agreements (FTA) and goods subject to the regulation 2004/13 ˇ Excise duty: specific tax applied mostly to luxury goods (fuel, alcoholic beverages, tobacco, etc.) Source: MEF 2007 159. Whether on the demand or expenditure side, the work should begin with the benefit package. This package should reflect political and clinical consensus, be transparent and be costed out based on historic resource use. This will establish public health system costs. The benefit package can be done independent of the modeling of revenues. Once done, there is a mix of system efficiencies that could result in savings, but good estimates are necessary for choosing policy priorities. The model would build in age and gender utilization profiles and develop future or "outyear" estimates of utilization and overall costs to the sector. Once the supply and demand modules are developed, the model will be available for other key policy discussions such as the benefit package for one or more population groups, for instance pensioners. 4.5.2 Option 2: Consider dedicated taxes for health 160. These would include taxes designed to eliminate negative externalities, such as tobacco tax. Cigarettes and alcohol abuse increase risk factors, which increases health care demand and costs. Thus, a cigarette and alcohol tax could improve health status and raise revenues while reducing demand for health services, since fewer people would be smoking or drinking alcohol. 161. Cigarette tax increases are considered among the most cost-effective interventions to improve health and decrease demand for services. The OECD countries have utilized this tax extensively. France was discussed earlier in this Chapter (Box 4.1). In the United Kingdom, cigarette taxes have been raised repeatedly over the past three decades. Partly because of these increases, and partly because of the steady increase in awareness about the health consequences of smoking, cigarette consumption has declined sharply over the same period, with the number of cigarettes sold per year falling from 138 billion to 80 billion over three decades. Revenues, however, are still rising. For every tax increase of 1 percent in the United Kingdom, government revenues increase by between 0.6 and 0.9 percent. In the United Kingdom, this money is used to pay for health care, as the NHS is funded from general revenues. Some countries even have earmarked tobacco taxes, among them Australia, France, and Turkey. Middle- 65 income countries also rely on tobacco taxes. Brazil, Egypt, and Thailand finance their HIFs through earmarked tobacco and alcohol taxes. 162. Kosovo could consider increasing cigarette taxes and levying alcohol taxes to raise additional revenues for the health sector. In Kosovo, of the 105 million collected each year in excise taxes, at least 20 million comes from levies on cigarettes. Customs officials estimate that 8 million now eludes collection each year, an improvement over the 1999-2000 period when as much as 30 million may have escaped collection. According to the UNMIK, improved tax collection and a reduction in import taxes has increased the share of excise tax generated by cigarettes substantially (Bozicevic et al. 2004). Further such efforts could generate even more funds for health, but simultaneous efforts are needed to prevent adverse effects such as smuggling and informal sector vendors. In addition, Kosovo could also consider raising taxes on alcohol. 163. A false argument against raising tobacco taxes is that it has an unfair impact on lower-income groups. Tobacco taxes in other countries have been found to be regressive because the poor spend larger shares of their more limited household incomes on tobacco than the rich. However, as the price elasticities of the poor tend to be higher than the nonpoor, an increase in tobacco taxes will result in a larger reduction in consumption, and hence will be progressive in terms of the reduced tax burden for the poor. In other words, poor consumers will be more sensitive to price increases, cut their demand, and in the process pay less in taxes relative to the upper-income groups. While there are no measured price elasticities for cigarettes by income class in Kosovo itself, this result has been observed in South Africa, Morocco, Bulgaria, Ukraine, Sri Lanka, Thailand, and Indonesia. Tax increases might also be accompanied by other measures such as awareness campaigns (preferably targeted to groups that smoke most, that is, poor and lower-income groups) to inform people about the harm caused by smoking and - even more important - the benefits of quitting, and provide help programs for those who want to quit. 164. At the same time, the case for excise taxes may need to be balanced by the current levels of smuggling across countries such as Albania, Kosovo, Greece, and Italy. This could imply the concurrent need for greater enforcement and enforcement capacity in the years to come. 4.5.3 Option 3: Set medium-term expenditure targets to increase general revenues 165. Some governments have set medium-term targets for health revenues, often defined in terms of percentage in GDP or levels of government spending. This targeting is usually a policy decision within specific national conditions when health is seen as a priority sector, and can be utilized for pro-active planning. Still, most countries do not have middle-term targets for health expenditures, which can result in governments not maintaining allocations or relying too much on alternative sources such as payroll taxes or co-payments. In the end, the sustainability of health care systems depends largely on the ability to generate sufficient revenues. This is a key challenge given the number of contextual and structural problems. Part of the issue remains fiscal capacity to 66 raise additional revenues for the government. There is wide variation in ability to raise government general revenues across countries (figure 4.16). Kosovo's total government revenue stands at 28.7 percent of GDP, relatively low compared to other countries. The ability to raise general government revenues in Kosovo will be a key determinant for considering where funding will come for health. Figure 4.16 - General Government Revenue as Percentage of GDP (2004) 60.0 50.0 40.0 t cen 30.0 er P 20.0 10.0 0.0 eni stan stan a gia ia ia d ria ro blic nia ary a in eg l ArmTaj iki i Georgyz Rep lbanikh a stan baijanman ia istan raineldova akiadonia pu Se rbia oatia arusov A ov Es d Cr rkmen LatvUk Sl ace Re Slove Hung Kyr Kaza Azer Ro huanbek Lit Mo toniaerationPolanBulganten Fe Berzeg Uz M Mo Tu ssian Czech snia-He Ru Bo Source: WHO 2006 166. At the same time, the economy is growing moderately and conservative expenditure rule (on its capital budget) has resulted in a government surplus for 2006 and 2007 and swelled the government's accumulated cash deposits. This presents an opportunity to place health more prominently in the MTEF for 2008-2010, an argument that would have to be substantiated by improved efficiency in resource allocation. 167. Several countries are now developing new MTEFs that include expenditure targets for health. For example, in Belarus, health revenue collection did not fall in 1990s as severely as in other Former Soviet Union countries, and the share of government health expenditure stabilized at 4-5 percent of GDP. However, even this level was insufficient for an excessive network of medical facilities with a dominant hospital sector (for example, admission rate in 2001 was 28.7 per 1,000 population in some regions). The government then set targets of 6.5 percent of GDP for 2003 and 7.5 percent for 2007, based solely on presidential decree, but the actual implementation did not meet the targets - hospital expenditure was 4.3 percent of GDP in 2003, and there were similar shortfalls in subsequent years. Besides implementation failures, the target of additional funding levels may be discounted without active cost containment and service delivery restructuring in systems with bloated capacity or outdated protocols. In a sustainable funding model, medium-term targets need to be coupled with systemic changes to improve sector performance. 67 168. Similar attempts of medium-term targets were done in the region. In Armenia, the World Bank poverty reduction strategy program set targets for health that the government adopted in 2002. The share of health expenditure in the state budget was to rise to 8 percent in 2005 (from 6.5 percent in 2003), 9 percent in 2008, and 10 percent in the final stage of the program in 2015. Actual implementation is close to these targetsafter a long period of substantial decreases in public health funding, there are signs of a positive trend. In 2003-06, the health budget rose by an average of 25 percent annually, and its share in state expenditures was 7.9 percent in 2004 and 8.2 percent in 2005. Unlike in Belarus, the targets were much more realistic, and the service delivery patterns changed substantially in parallel. Medium-term targets have also been set for health in Albania, FYR Macedonia, and Montenegro. The Kosovo MTEF has outlined public spending levels for health. It proposes several focus areas including improving health status for different population groups and strengthening sector efficiency through increased spending for primary health care and hospital modernization based on a masterplan. 169. Different health care financing methods combined with different risk-pool arrangements impose different overall budget constraint on the health sector. A relatively high out-of-pocket share paid by patients limits the government stewardship role in managing health expenditures. In Kosovo, private spending is estimated at more than 40 percent of total health expenditures; hence almost half of health expenditures are outside the government control. Under tax financing and payroll funding, the total amount that would be spent for health care is organized and decided centrally. These single-payer financing methods can steer resource allocation and allow, for instance, more to be spent on prevention to prevent chronic illnesses from becoming acute problems requiring costly treatment. The single-payer tax-financing method of the United Kingdom and Scandinavian countries enables them to limit the diffusion of new technology and drugs that are not cost-effective and hence contain the rate of health spending inflation. Single-payer financing systems such as those in Canada and Taiwan reduce transaction costs and control what physicians and hospitals can charge to patients beyond the amount insurance pays (Hsiao 2007). In countries with multiple payers (that is, insurers and patients), government policies such as budget constraints are less effective as costs can be shifted across payers such as patients. 170. A further complication in Kosovo is private provision. Even with single- channel financing, rules and regulatory capacity will be needed to ensure coordination across public and private providers and payers. This will be important for patient utilization and use of high technology such as CT scans and MRIs. Even in countries with coordinated financing, unfettered private provision can result in an oversupply of equipment and overutilization of services, as evidenced in many countries including China and Iran (World Bank 2007). 4.5.4 Option 4: Encourage private VHI 171. The development of a private insurance sector depends on an advanced policy environment in health. The initiative, if designed correctly, could provide better access to care and possibly raise additional resources, which might have a positive impact 68 for all through improved quality and availability of services. Private insurance could be utilized to gain access to highly qualified providers, high-end diagnostic tests in the private sector, and amenities such as treatment abroad and hotel-like rooms in hospitals not covered under the basic benefit package. It could protect against out-of-pocket costs and costs for catastrophic events or chronic illnesses as well as possibly reduce reliance on informal payments. 172. Different types of private insurance exist: complementary, supplementary, and duplicative insurance. Private health insurance complements financing from public programs in some OECD countries by covering cost-sharing under those arrangements. This type of coverage predominates in France, where complementary insurance reaches more than 90 percent of the population. In the United States, individuals eligible for Medicare can buy policies covering co-payments or other service gaps in the public program. In many OECD countries, private health insurance supplements public systems by financing goods and services that are excluded from the basic benefit package. In Switzerland, for example, about 70 percent of the population has supplementary insurance. In Australia, Ireland, New Zealand and the United Kingdom, where privately funded providers operate in parallel to the public delivery system, private health insurance duplicates existing public universal coverage, offering a private alternative to public insurance. Nearly half of the Australian and Irish populations purchase a private health insurance policy, making these the largest duplicate markets in the OECD (OECD 2004). 173. The development of private insurance would require some initial baseline review and synthesis of information, and substantial public sector involvement. Information reviews will include updating of out-of-pocket expenditures including informal payments; a marketing review of consumer demands and current level of trust in insurers and insurance market; a review of incentives and disincentives for consumers and employers in the purchase of insurance; a review of legislation and regulation regarding allowance of Kosovo firms and EU-market firms to sell health insurance in Kosovo; and a review of current regulatory structure and options for strengthening capacity. The new private insurance market will need to create clear legislation and a transparent benefit package that is supported by public funds. Only when this occurs will a private market be able to emerge. A number of legal and regulatory mechanisms are needed to foster a well-functioning market of private insurance, including definitions on standards for market entry, rules for reporting and market exit, and consumer protection mechanisms (table 4.13). 69 Table 4.13 - Legal and Regulatory Issues for Implementing Private Health Insurance Financial and nonfinancial Rules for reporting and Consumer protections standards for market entry exit of health insurance and mechanisms to and operation plans improve fairness ˇ Capital and surplus ˇ Reporting of financial ˇ Language and requirements and market information marketing of ˇ Common accounting and ˇ Use of accounting and contracts actuarial practices actuarial professions to ˇ Provider-plan ˇ Reinsurance requirements conduct onsite relations ˇ Approved business plan examinations ˇ Guaranteed ˇ Citizen/residency of owners ˇ Notice to policyholders issue/renewal ˇ Lawful organization forms and financial plan for ˇ Premium rating ˇ Prohibited products paying incurred but not ˇ Rate review/approval reported expenses ˇ Mandated/standard ˇ Guaranty funds benefits Source: World Bank 2008 Regional Experience with VHI 174. VHI has been a modest source of revenue in many countries in the region. VHI was conceived in many countries, such as Switzerland, Germany, Poland, and Kazakhstan, as a supplement to SHI, covering those services excluded from the benefits of the SHI scheme. In some countries, such as Hungary, Poland, and Russia, there have been more recent discussions and debates about introducing a model where private insurers compete for the primary benefit package, or consumers can opt-out of the public system. Failed attempts to introduce private competitive health sector insurance occurred in the Czech Republic and Kazakhstan. The revenue raising potential of VHI is modest, contributing at most 1 percent of GDP to the health sector in Switzerland and France, where VHI is most developed. 175. Complementary insurance to cover co-payments led to financial sustainability problems and inequity in access, for which it is prohibited by law in Switzerland. Some countries have introduced VHI to insure co-payments charged under the SHI, among them Croatia, France, and Slovenia. Complementary insurance did not have the expected effect on equity and financial sustainability, which is one of the reasons why it is forbidden by law in Switzerland.21 Evidence shows that insurance to cover co-payments is generally purchased by those who use a high amount of health care (Tapay 2001). The resulting adverse selection endangers the financial sustainability of insurance. In Slovenia, in 2002, 95 percent of the population had entered a VHI scheme to cover co-payments charged under SHI. In Croatia and Slovenia, insurers ran into 21See Art 64 in the Swiss health insurance law (KVG): Schweizerisches Gesetz fuer Krankenversicherungen. 70 financial sustainability problems caused by adverse selection of higher risk-groups (Skledar 2004; Mastilica and Kusec 2005). In France, complementary insurance negatively affects equity in health, as higher-income groups are more likely to be insured; and when insured, they use care more often than the lower-income groups (Buchmueller and Couffinhal 2004). 176. The market in private VHI remains small in most countries. The market share of private insurance has been limited by several factors: (i) most transition countries have an insufficient number of high-quality provider networks to contract with; (ii) the SHI benefit package is not clearly defined; (iii) there is a tradition of informal payments at the point of service; (iv) countries have limited capacity for regulation, management, and actuarial studies; and (iv) insurance markets are immature, focused only on the provision of indemnity insurance, often bundled with life and other insurance products to stay profitable, such as in Russia. 177. These lessons suggest that private VHI is not a strategic choice for Kosovo. Private VHI may not yield significant new revenues for Kosovo, and the lack of institutional and regulatory capacity in many of its neighboring countries may be an additional signal of caution for Kosovo, at least in the short term. At a minimum, the government of Kosovo would need to do a comprehensive regulatory analysis of current regulatory capacity and assess costs, a reasonable medium-term timeline, and needed inputs to ensure an adequate regulatory status to oversee the private insurance market. 178. Overall, Kosovo is actively exploring ways to improve public revenues. The case for higher public sector spending on health cannot, however, be made in a vacuum. The government will need to take a closer look at how much fiscal space there will be in the medium term to expand public sector spending on health and what inter-sectoral trade-offs such increased health sector spending entails. Somewhat linked to this is also the question of desirability of earmarked funding through either payroll or mechanisms such as sin taxes. While this is obviously appealing from a health sector perspective, the question is how desirable earmarking of taxes for health is from an overall macroeconomic and public finance viewpoint. 71 CHAPTER 5. ORGANIZATIONAL FORM OF RISK POOLING 5.1 OVERVIEW AND DEFINITIONS 179. The objective of this chapter is to present an overview on how risk pooling is organized in different countries. The focus will be on different risk-pooling structures; organizational and legal forms; and concerns about equity and financial sustainability that arise in the context of risk pooling. Based on this review, implications for the Kosovo health sector are presented. 180. The risk-pooling function of a health insurer pools and spreads different and independent financial risks related to illness across individuals and time. Through risk pooling, insurance lowers the out-of-pocket price for medical care at the time of consumption, and it reduces the individual risk caused by the uncertainty about future health status by cross-subsidizing health care from "low-service users" to "high-service users" (Hurley 2000). A lack of risk pooling and insurance has led to health-related impoverishment in many countries.22 181. The effectiveness of a risk-pool organization depends on several factors. These include its structural configuration, organizational form, and incentives regimes (Preker et al. 2007), the inclusiveness and thus financial sustainability of the benefit package, and institutional capabilities, including effective government and the accountability of organizations. A risk-pool organization can be a political entity such as a government ministry, an independent social insurance fund, many private insurers operating in a competitive market, or a combination thereof. When risk pools are identical with purchasers, then they have the financial decision-making power to decide about contracting and payment methods with providers. 5.2 RISK-POOLING STRUCTURE: SINGLE OR MULTIPLE RISK POOLS 5.2.1 Single risk pools 182. Many countries have single risk pools for the entire country that can be financed out of general government revenues or payroll taxes. Single pooling reflects the equity objective of securing universal coverage, often referred to as the solidarity principle (Hsiao 2007). In countries with a single risk pool, revenue (whether generated by general taxation, payroll contributions, or individual premiums) is placed in a single central pool that seeks to cover a chosen package of health care services. Payments are then made to providers for services provided to patients, based on the terms agreed upon in the contract between the risk-pooling organization and providers. Among the single- 22For an overview, see O'Donnell et al. 2007. 72 pool countries are all the countries of the former Yugoslavia except Bosnia and Herzegovina (see below), and most Central and Western European countries. 183. Some "single" risk-pooling arrangements are exclusive to defined population groups, which may lead to unequal access to health care. The organizational form of risk pools is influenced by health policy objectives, institutional and market contexts, and the population's socioeconomic and cultural backgrounds. In high- and middle-income countries with strong formal sector employment, compulsory, state-organized health insurance systems, and private VHI are predominant. These insurers contract with or provide medical care in public and private health facilities. Few developing countries rely on risk pooling through insurance, mainly because their small formal sectors would result in small risk pools for the more affluent segments of the society, while the poor and informal sector groups tend to be excluded from these risk-sharing arrangements. Such limited risk pooling and exclusion of the informal sector population groups contributes to inequity in service use in a population. 184. Depending on the population size, unified pooling may be more efficient as it spreads risks over a bigger population. Smaller risk pools tend to report larger variations between actual and predicted spending than larger pools. Studies found that for an insured population of 10,000, there is a 33 percent probability of a more than 10 percent variation of actual spending compared with the predicted spending; this probability falls to less than 1 percent for a population of 100,000 and becomes negligible for populations of more than 400,000 individuals. International experience suggests that, due to limits in data, risk pooling should be initiated for populations of at least 500,000 or more. Therefore, countries place limits on risk exposure on small pools, and have central health funds pay for health expenditures (Ensor 2005). 5.2.2 Multiple risk pools with and without competition 185. Some countries have opted for multiple risk pools. The United States has multiple risk pools, being based on factors such as age (Medicare), poverty (Medicaid), or employment (employer-sponsored enrollment in private insurance). European countries with multiple risk pools in the form of insurance include Bosnia and Herzegovina, Czech Republic, Slovakia, Germany, Austria, Switzerland, Belgium and the Netherlands. In these countries, pooling is based on region; employer and professional affiliation; and, to some extent, consumer choice. 186. Some countries with multiple risk pools expect competition between insurers to have an impact on health expenditures. Competition between multiple insurance funds, mainly for lower-risk members, has led to a consolidation of the insurance market through mergers and market exits of insurers. In Germany, the number of insurers has declined from 1,209 in 1991 to 238 in July 2007 (dfg-Ranking 9.8.2007). In Switzerland, the number of insurers dropped from 145 in 1996 to 87 in 2006.23 So far, competition has not had a moderating impact on health expenditures, and in fact, the two most expensive 23BAG: Statistik der obligatorischen Krankenversicherung, 2006. http://www.bag.admin.ch/themen/krankenversicherung/01156/index.html?lang=de 73 health systems in the world - the U.S. and Swiss - also happen to have multiple insurance systems. 187. In several middle- and lower-income countries, multiple insurance systems have not performed well. Colombia, a lower-middle-income country, established SHI in 1993 with a competitive model to reform its health care delivery. Under this model, competing private insurance plans purchase health care to achieve greater efficiency and better quality, though neither has materialized. Risk selection by insurers has been a problem, along with political obstacles to corporatizing public hospitals and forcing them to compete on a level playing field (Hsiao 2007). 188. The Kyrgyz Republic, as well as Estonia, has abandoned fragmented risk pooling and moved to a single insurance system, which substantially improved access to care. From 1997 until 2000, the Kyrgyz Republic introduced a mandatory health insurance fund (MHIF) as a national pool that coordinated with local government (oblast and rayon) pools under a "joint systems" approach. In 2001, the implementation of a "single payer" reform began with pooling of rayon and oblast revenues in the oblast department of the MHIF. Nationwide implementation was completed by 2005, and resulted in one pool for the entire population of each oblast and one contributory national pool for the insured providing a complementary benefit package; both pools are managed by the MHIF. In 2005, following legislative changes related to fiscal decentralization, the Kyrgyz government decided to centralize the fragmented health budgets at a republic level. In 2006, reforms led to a shift from oblast-level pooling to merged and centralized pooling of general budget funds in a single risk pool on the national level, the MHIF. Preliminary findings suggest that creating a single risk pool has contributed to key policy objectives such as improved efficiency in service delivery and administration, transparency, and equity of access and the distribution of health spending (Kutzin et al. 2007). 5.2.3 Multiple insurance and risk-equalization transfers 189. Risk-equalization transfers across multiple insurers depend on sophisticated data collection and analysis, and sophisticated governance structures. Countries with multiple risk pools and community rating of premiums have risk equalization schemes across insurers to equalize the risk distribution of individuals across different pools. Risk-equalization schemes use risk adjusters to predict the total health care expenditure of individuals enrolled in an insurance pool. Based on risk-distribution results, there is a redistribution of insurance finances from pools with "lower risk" to pools with "higher risk" enrollment. This redistribution aims to set a disincentive to insurers for risk selection (of low risks). Table 5.1 provides an overview on the different risk adjustment variables being used in selected countries. Redistribution of funds across insurers has been criticized as incomplete mainly because of the low quality of risk adjusters and incomplete information about individuals' health status. Simple risk adjusters based on age, gender, and region such as in Switzerland predict only about 5-7 percent of total expenditures, whereas more complex adjusters as currently used in the Netherlands are able to predict about 25 percent of total expenditures of a risk pool (Gress 2006). Risk-adjustment and equalization transfers across multiple pools needs 74 substantial investment in HMIS to collect data on individuals' health and demographic background and a transparent financial and auditing system to ensure the validity and reliability of data based on which funds are transferred. Table 5.1 - Insurance Market Structure and Risk Adjusters in Multiple Insurance, 2006 Market Germany Switzerland Netherlands Czech Slovakia Rep. Number of 275 87 33 9 7 insurers Risk adjuster Age, gender, Age, gender, Age, gender, Age, Age, in risk disability, region region, disability, gender gender equalization entitlement sick working status, scheme leave, income, pharmacy, disease diagnosis cost management groups program 5.2.4 Lessons from fragmented risk pooling in Bosnia and Herzegovina 190. Fragmented risk pooling results in ineffective insurance coverage. In Bosnia and Herzegovina, political decentralization resulted in fragmented risk pooling, with 13 HIFs for a population of 3.9 million people, including the central HIF in the Republic Srbska (RS), the insurance fund in District Brcko, 10 cantonal HIFs and the Federal Solidarity Fund in the Federation of Bosnia and Herzegovina. Insurance membership is defined by place of residence. The number of members in the 13 HIFs ranges from 35,000 in the smallest cantonal pool (Gorazde) to more than 402,000 members in Sarajevo Canton HIF, and 1.1 million members in the RS HIF, the largest pool in 2004. Indeed, four of the cantonal HIFs have fewer than 100,000 members. This stands in contrast to the single health insurer (MHIF) for the Kyrgyz population of 5.3 million and the single risk pool for Moldova's 4.2 million people. Because of the administrative limitations and small formal sector economies in some Bosnian cantons, the population insurance coverage rates vary widely, ranging from 61 percent to 95 percent across the 10 cantonal HIFs in the Federation. The funds are financed through payroll taxes from formal sector employees, individual premiums paid by those active in the informal sector and agriculture, and general budget transfers for low-income groups. Fragmented pooling combined with high contribution evasion has resulted in a financially unsustainable situation. As a result, co-insurance rates paid by patients at the time of service use are up to 80 percent of the service or drug price, which questions the effectiveness of insurance in Bosnia and Herzegovina against paying out-of-pocket fees. 191. Fragmented risk pooling without risk-equalization transfer leads to high out- of-pocket payments and unequal access to care in Bosnia and Herzegovina. The inability to redistribute funds across pools means the relative size of each pool reflects the contribution capacity of the territory (cantonal or entity level) it serves, rather than the 75 underlying health care needs of the covered population. Some insurers in richer cantons are "richer" than others and can afford paying for a larger benefit package. Not surprisingly, "poorer" cantons charge higher co-payments to patients to raise additional funds for health, thereby increasing the financial barriers to access to care. The consequences of such fragmented pooling are geographical (and probably related socioeconomic) inequities in access and financing of care, and inequity in resource allocation (figure 5.1) that in turn contributes to unequal benefit packages for the insured. Figure 5.1 - Revenue and Expenditures per Member per Year, across HIFs in Bosnia and Herzegovina (in convertible marka) 600 500 Avg insurance revenue per insured Avg spending per insured 400 300 200 100 0 Una-SanHercegovina a Zenica-DoSrednja boj Bosna TuzlaPosavinarceg-Bosnsko na Podrinje Mostar Sarajevo Zapadna He Bosa Source: Federal Solidarity Fund Bosnia and Herzegovina) 2003 192. Fragmented risk pooling has led to ineffective insurance coverage and inequity in health financing in Bosnia and Herzegovina. A 2001 household survey indicated that there was no significant difference in the out-of-pocket amounts paid by insured and uninsured patients who had to pay for outpatient care, suggesting that insurance does not improve access to care. In addition, households classified in the poorest quintile spent about 10 percent of their total consumption on health care; this share was 5 percent for the richest quintile of households, pointing to inequity in health financing (World Bank 2003). 5.2.5 Multiple risk pools under fiscal decentralization 193. Under fiscal decentralization tax-funded health systems generally have risk- equalization transfers across regions. Risk pooling through fiscal decentralization generally includes some form of equalization payments, in which less-well-off regions or municipalities receive additional funds. Some countries have alternative equalizing mechanisms, such as special funds (Italy, Spain). In Spain and Italy, these health equalization mechanisms have become enmeshed in broader regional-central funding disputes. The Cohesion Fund created in 2003 in Spain, intended to ensure adequate standards of health care across the 17 regions, will likely be changed or abandoned. In Italy, the National Solidarity Fund, created as part of a broader national-regional funding agreement to provide transfers from wealthier Northern to poorer Southern regions, has never been funded. In Scandinavia equalization schemes are based upon the tax base and potential revenues based on the average tax rate, rather than actual revenues (Saltman et al. 2007). 76 5.3 CURRENT FORM OF RISK POOLING IN KOSOVO 194. Kosovo still has a direct-provision model that does not split between risk- pools, purchasers and providers. All risk pools (municipalities and central level) are public entities that are also the owners of public facilities, responsible for the provision of care in Kosovo. This direct provision results in a situation where the financing, pooling, and provision of health care is integrated and managed by the same organization. This model of direct provision and financing is common in most developing countries. 195. In Kosovo, fragmented risk pooling for different levels of care without risk equalizations contributes to unequal service provision. Kosovo has fiscal decentralization with basic health care delegated to the municipal level. This has resulted in 32 risk pools; all of them are political entities that pool risk based on regional criteria. The hospital risk is pooled nationally at the MEF, while the PHC risk is pooled by the 31 municipalities that are also the owners of PHC centers. The MEF transfers an annual PHC grant to municipalities based on each municipality's total population (MLGA 2007). In the absence of demographic and health data on the municipality population, the financial situation of the PHC risk pool is influenced by political negotiations and historical allocations. The resulting risk pooling without risk-adjusted transfers does not account for providers' different costs due to different case mixes, and may cause drug and other shortages, in turn causing patients to purchase drugs in private pharmacies. To prevent such inequalities, Kosovo could possibly learn from the experience of the Kyrgyz Republic and pool all health risks in one national-level fund that provides coverage for all levels of care (PHC, hospital, and outpatient specialist). 196. Fragmented pooling in small risk pools increases financial risks and limits access to care. In Kosovo, only two municipalities have more than 200,000 inhabitants and 21 municipalities have fewer than 100,000 inhabitants, resulting in risk pools that are too small. Small risk pools at the municipality level, unequal revenue-raising capacities for OSR, and a lack of local capacity for strategic purchasing across municipalities results in inefficient risk distributions, and limits people's access to care as shown by the low visit rate of 1.3 visits per capita per year to PHC facilities (Kosovo Statistics Office 2005). Findings from the HBS show that out-of-pocket health expenditures leads to regressive health financing with the poor paying a higher share of income than the better- off. 5.4 CONCERNS ABOUT IMPERFECT INFORMATION IN RISK POOLING 197. Insurers have specific mechanisms to limit inefficient service use caused by imperfect information. While health insurance leads to better access to care, it also sets behavioral incentives to providers and the insured, which may lead to inefficiency and endanger the financial sustainability of insurance (Cutler and Zeckhauser 2000). These incentives include adverse selection, cream-skimming, moral hazard, and supplier- 77 induced demand. They arise because of asymmetric information, that is, the inability of insurers to monitor the actions of the insured and the provider. The resulting overuse of care can endanger an insurer's financial sustainability. To prevent this, insurers try to influence their members' behavior by setting co-payments. They also try to shift some of the financial risk related to insurance to providers through the provider payment mechanism. 198. Adverse selection by individuals into different risk pools is a problem under VHI enrollment. Adverse selection means that primarily high-risk individuals insure, whereas low-risks remain uninsured. It results in high-risk pools. To prevent adverse selection, in most European countries, risk pooling is mandatory, the enrollment period is at least one year, and rich or healthy citizens cannot opt out of contributing. However, where enrollment is voluntary, insurers' financial solvency is threatened if mainly the ill insure, and premiums are not risk adjusted such that higher-risk individuals would be paying a higher rate. 199. Adverse selection jeopardized the financial viability of the voluntary complementary insurance in Slovenia. In Slovenia, in 2003, Vzajemna, the largest VHI company, had more than 1.1 million members with complementary health insurance and more than 80 percent of the market share. In the first five months of 2004, Vzajemna reported a loss of SIT 1.4 billion (5.8 million), whichaccording to Vzajemnawas primarily the consequence of the high share of pensioners among its insured members (adverse selection). On August 1, 2004, Vzajemna was due to raise the insurance premiums for complementary health insurance by 13.5 percent for 383,000 insured people over the age of 60, but this was met with political opposition (Skledar 2004). 200. Cream-skimming or risk selection by insurers is a common strategy in a multiple insurance system when premiums are defined independent of individuals health risk. If insurers set premiums independent of individuals' health risk (community rating), they will have an incentive to select the "good risk," that is, healthier individuals with potentially lower service use. Individuals in bad health will find it difficult to insure. Most countries with multiple insurance systems and community rating have risk- equalization schemes across insurance funds to set a disincentive for cream-skimming to insurers. However, risk-equalization schemes are of low quality. They are based on incomplete information on individuals' health status and the related financial risk. Risk selection is thus the predominant strategy in multiple insurance systems (Van de Ven et al. 2007). 201. Insurers try to influence moral hazard and supplier-induced demand by increasing co-payments paid by patients and through prospective provider payment. Moral hazard occurs when insured individuals overuse medical services because they co-pay a reduced price for care at the time of use. Overuse implies that the benefits are less than the risks and costs, and it can result in unnecessary cost increases. Supplier-induced demand happens when providers oversupply care as a result of being reimbursed based on the number of services provided under fee-for-service payment. Both may lead to variations in the package received and inefficient provision of care. Insurers can respond to moral hazard and supplier-induced demand by shifting part of the 78 insurance financial risk to the insured and providers, who as a result will bear more risk than they would like - patients will pay a higher co-payment and providers will be paid through prospective payment methods such as capitation, which increases the financial risk born by the provider. 202. Demand-side policies such as cost sharing are often introduced based on the moral hazard argument. Cost sharing refers to out-of-pocket payments made by patients to the providers at the time of service use. For patients with health insurance, there are three main forms of cost-sharing: ˇ deductible: an amount that must be paid out-of-pocket before benefits of the insurance become active; ˇ co-payment: a flat amount that the insured patient must pay for each service used; ˇ co-insurance: a percentage of the total charges for a service that must be paid by the beneficiary. Other policies that are frequently associated with cost-sharing mechanisms include benefit maximums, out-of-pocket maximums, extra billings, pharmaceutical reference pricing, and coverage exclusions. All these options aim to influence patients' care seeking behavior. 203. Risk-shifting to providers may result in substandard care and requires monitoring and evaluation of provider performance. Providers who carry a larger financial riskas, for example, under capitation paymenthave an incentive to reduce costs. Cost-reduction strategies may cause adverse effects such as skimping on patient care. The government and purchasers must therefore strengthen their capacity for monitoring and evaluation to identify substandard quality of care and to ensure that all providers are offering uniform levels of care, in line with the chosen package. 5.5 CONCERNS ABOUT INEQUITY IN ACCESS TO CARE 204. Inequity in access to care occurs when low-income groups are more likely to be excluded from insurance. In low- and middle-income countries where universal coverage is not achieved, individuals in skilled or technical professions who are salaried are more likely to be covered under insurance and to have more comprehensive coverage than the unemployed, and agricultural and informal sector workers. In Bosnia and Herzegovina, about 30 percent of the population is not insured mainly because they work in agriculture or in the informal sector. The experience with VHI in France (Buchmueller and Couffinhal 2004) shows that complementary insurance negatively affects equity in health. People in lower-income categories are considerably more likely (54.5 percent) to be excluded from VHI than the better-off. Equity is of particular concern in countries with high out-of-pocket payments and where the lack of single risk pools allows different income groups to pool among themselves in fragmented risk pools and private insurance (France, Germany, and the Netherlands). Exclusion of the poor from insurance will lead 79 to inequity in access, which highlights the importance of universal and mandatory enrollment. 205. Insurance may create inequity in access to care because the rich insured are more likely to use specialist care than lower-income groups. Based on 1996 household survey data, and after adjusting for insured individuals' need of care, van Doorslaer and Masseria (2004) find evidence of pro-rich patterns of specialist care use among the insured, with higher-income groups using more specialist care than what would be expected on the basis of their need for care. Similarly, household survey findings from Switzerland indicate that service use for general practitioners (GPs) and hospitals is distributed equally across income groups, whereas the distribution of specialist visits is significantly pro-rich (Leu and Schellhorn 2004). Different treatment- seeking behavior across income groups could be addressed by charging higher co- payments for specialist care with clear exemption policies for vulnerable groups. 5.6 THE FINANCIAL SUSTAINABILITY OF HEALTH INSURANCE 206. The financial viability of risk pools is endangered if expenditures exceed revenues. Many payroll tax-funded social health insurers in the Central and Eastern European region have a mandate to cover the whole population but struggle with revenue collection and expenditure management, resulting in regular deficits. The most common origin of this problem is incomplete financing and pooling arrangements to cover the inactive part of the population such as the elderly, children, and population groups active in the agricultural and informal sector. In addition, insurance laws propose a comprehensive benefit package, and people often consider the benefit package as strict entitlements, making it politically very difficult to adjust benefits in line with financial and economic pressure. The resulting insurance deficit tends to be financed by the ministry of finance, and by the population through increasing contribution levels and co- payments paid by patients. Alternatively, instead of being a passive disbursement agent, insurers can play an active role and contract strategically with more efficient providers. Accurate estimates of the benefit package and of costs based on actuarial methods will determine the financial sustainability and survival of risk pools. 207. The performance of health insurance organizations has suffered from poor financial management and high administrative costs. Poor financial management, lack of guidelines for the investment of insurance funds, weak banking systems, and incidents of fraud and misappropriation of funds aggravate performance of the insurance systems. Excessive levels of administrative bureaucracy also hamper performance especially where the number of enrolled population is low. Administrative costs of SHI funds varied throughout the Central and Eastern European region. Estonia, Lithuania, and Poland, which operate single HIFs with regional branches and almost universal coverage, reported administrative costs of between 1 and 2 percent of revenue.24 Albania, in 24In Poland, administration costs of Sickness Funds in 1999 were 1.46 percent of total fund costs, in 2000 1.36 percent, and in 2001 1.72 percent. 80 contrast, spent 7 percent on administration of its health insurance scheme in 1998 while only achieving a coverage rate of 68 percent of its population. 5.7 ESTABLISHING AN INSURANCE FUND 25 208. Developing and implementing health insurance organizations takes time and needs constant adjustments and enhancement. In Estonia, it took two years from presentation of the health insurance legislation to Parliament in 1990 until its entry into force on January 1, 1992. The establishment of SHI in Poland was carried out with speed, but with consequences. Following initial discussions of a draft law in 1996/7, the health insurance act was passed at the end of 1998, and health insurance started to operate as of January 1999. In retrospect, it became apparent that speed took its toll - important aspects of administration and organization were not planned for. In the Kyrgyz Republic, it took almost five years to implement health insurance after the governing legal framework was approved in 1992. The decree to establish a HIF was approved in January 1996. Premium collection started in January 1997, and funding for service provision began in July of that year. In Romania, initial discussions of a health insurance law started in 1994, and it was passed by one chamber of Parliament in the same year. Final approval by Parliament, however, was delayed by another three years until 1997, and the collection of funds and expenditure started in 1998, then administered by the Ministries of Health and Finance. The HIF as a separate administrative entity was established in 1999. Over time, insurance needs constant refinement to address such issues as cost escalation (Croatia, Hungary), and develop new capacity for legislation and regulatory and administration. 5.8 IMPLICATIONS FOR RISK POOLING IN KOSOVO 209. Kosovo's socioeconomic, demographic, and historical situation will influence the nature of health risk pooling. Kosovo has an estimated population of about 2.5 million, of which roughly 361,000 (15 percent of the population) are active in a very small formal sector. The rest of the working-age population is either unemployed, or active in the informal economy or in agriculture. Kosovo has the lowest level of per capita GDP in Europe and a relatively modest GDP growth rate. If only 15 percent of the population contribute to a risk pool but 100 percent benefit, then major additional funding is needed to ensure the pools' financial sustainability. 210. Kosovo lacks the institutional capacity to manage multiple risk pools. To prevent financial insolvency caused by adverse selection and cream-skimming into smaller pools, a system with multiple risk pools would require either (i) risk-rated premium setting or (ii) community-rated contributions with risk-equalization transfers across risk pools, as in Switzerland. Multiple risk pools with risk-adjusted or community-rated premiums may not be feasible to implement in Kosovo given the small 25This section draws on Langenbrunner2005. 81 population size; the lack of institutional, fiduciary, and financial management capacity; and the complete absence of census and individual health status data. Given the small population, a single risk pool with mandatory enrollment may be easier to implement and will prevent problems related to risk selection. 211. A future risk pool for health care in Kosovo needs to overcome current inefficiencies and inequities in the health system. These are caused by a missing purchaser-provider split, fragmented pooling at municipality and central levels, high out- of-pocket payments, low productivity in the health sector, and insufficient resources and quality of care in health facilities. In addition, Kosovo lacks institutional and technical capacity in financial and insurance management. Insurance can contribute to a more efficient health sector, if providers are rewarded for better-quality care, patients are charged a defined co-payment, and financial funds are managed professionally following standard accounting principles. To limit adverse effects such as moral hazard behavior leading to overuse of care, affordable co-payment levels need to be charged. Equity in risk pooling will require universal enrollment, and legally exempting vulnerable population groups from paying co-payments. 212. A sustainable single risk pool that integrates all levels of care may lead to improved access and continuity across the different levels of care. If Kosovo opts for implementing a HIF, this would imply that all public health funds go through a single health fund that is the purchaser of care. Recurrent costs of PHC provision will then be paid through the central fund instead of the block grant. The financial viability of a risk pool requires actuarial management of the benefit package. Additional revenues generated by patient co-payments must stay within the system, retained and registered by health facilities following international accounting practices. Providers must be equipped with accounting and financial management and planning systems and human capacity to ensure they reinvest their revenues from patients and insurance into improved quality of care, and the delivery of the contracted benefit package. 213. Several reform steps are needed to combine funds in a single pool, including, first, exclude health from the block grant paid to municipalities and, second, redefine the role to be played by providers and municipalities. As a result and in line with other European countries, the role of municipalities in a system with provider autonomy would be limited to ownership of facilities, and being a supervisory and complaint body. Third, devolve PHC management to the provider level, following the subsidiarity principle, and make health facility directors responsible and accountable for the management and provision of care. Fourth, create a single risk pool in charge of purchasing all levels of care including PHC and hospital care in the public sector and possibly also in the private sector. Fifth, hire actuarial capacity in the risk pool to calculate the cost of the benefit package covered by insurance and resulting revenue needs; set the public expenditure target and identify based on the actuarial analysis a basic benefit package that can be financed with public funds. Sixth, define the flow of funds from sources to users of health monies. Seventh, based on actuarial analysis, develop contracts to be negotiated with providers. Eighth, monitor and evaluate the flow of fund and impact on the effectiveness of PHC delivery. 82 214. The successful KPST organization could serve as a model for the organizational form of a health risk pool. Following the KPST example, the Kosovo Health Fund could be established as an independent autonomous organization solely for the purpose of risk pooling and managing the funds of a future health insurer, and purchasing care from providers. Legally this organization could be a not-for-profit body organized under private or public law. It should be staffed by Kosovar management and employees recruited locally as well as internationally from the Kosovar diaspora working in insurance in other countries. The governing board should include a combination of local and foreign insurance experts. The BPK could act as the regulatory and oversight body to ensure that the funds managed in the KHF are professionally managed and reserves securely invested, according to international best insurance practices, and with comparatively low administrative fees. 83 CHAPTER 6. STRATEGIC RESOURCE ALLOCATION AND PURCHASING OF PRIORITY HEALTH SERVICES 6.1 OVERVIEW AND DEFINITIONS 215. The objective of this chapter is to present an overview on purchasing and provider payment in European countries, including purchaser-provider split, strategic purchasing, health facility autonomy, provider payment methods for PHC and hospitals, and key issues in contracts for a basic benefit package. Options are discussed for provider payment reforms in Kosovo. 216. Purchasing is the transfer of pooled funds from the purchaser to health care facilities to buy a benefit package defined in a contract. The benefit package can include health services such as ambulatory visits, diagnostic tests, surgery, and hospitalization. Purchasers are insurers or agents who act on behalf of the government in the absence of insurance. Under strategic forms of purchasing, purchasers take proactive decisions about which health care services should be purchased from providers, at what quantity and price, how and from whom (Figueras et al. 2005). Purchasers use different provider payment methods to pay for care received from providers on behalf of a population. 217. PHC providers are increasingly paid a risk-adjusted capitation budget. Provider payment methods for PHC include (i) fee-for-service, (ii) capitation, (iii) an input-based budget, or (iv) a combination thereof. Input-based line-item budgets allow payers to control PHC costs directly, but budgets set an incentive for increasing the number of factors based on which the budget is defined (for example, beds, staff) as well as underprovision of services and excessive referrals to specialists. Fee-for-service sets an incentive to providers to expand the volume of services and negotiate for higher prices. Under capitation, providers are paid a fixed amount for each individual registered, usually adjusted for factors such as age and gender. While capitation sets an incentive for reducing costs, it also may lead to adverse effects, such as skimping on patient care and excluding high-risk patients from registration with the PHC provider (Cashin et al. 2005). In South Eastern Europe, most countries are in the process of or have already moved from input-based payments to capitation payments for outpatient care. 218. Hospitals are increasingly being paid by case-based payment. Hospital payments include different options: (i) line-item budgets; (ii) per diem (per bed day); (iii) fee-for-service; (iv) case-based payment such as diagnosis-related groups (DRGs) for each case hospitalized; (v) global budget, (vi) per episode of illness; and (vii) capitation per insured member. Per diems, DRGs, and fee-for-service all have an incentive to increase the number of days, cases, or services; and they pose less financial risk to the hospital than simple capitation. Most countries in the South Eastern Europe region have started the process of hospital payment reform, with most of them moving from input- 84 based to some kind of case-based payment with the long-term objective of implementing DRGs. 219. Pay-for-performance (P4P) is more common in countries with strong institutions and well established monitoring and evaluation systems in health facilities. Output-based reimbursement methods, such as fee-for-service, DRGs, and capitation, provide little financial reward for improvements in quality of care. Some countries have therefore added a performance-based financial incentive to the provider payment method, to reward better-quality provider performance and to prevent negative consequences such as inappropriate referrals to hospitals and specialists and substandard quality of care. P4P pays different amounts to providers based on their performance. Performance measurement can include five dimensions: (i) patient satisfaction, (ii) clinical process, (iii) outcome, (iv) IT, and (v) efficiency indicators to identify areas of inappropriate utilization (Rosenthal et al. 2006). 220. Strategic purchasing requires several factors including active purchasers, providers and consumers, additional information, and possibly some legal changes. Insurers in South Eastern Europe tend to be passive purchasers and operate as simple disbursement agents. Several conditions would need to be in place to implement strategic purchasing. First, a purchaser needs adequate information to assess provider performance and use results in rate-setting and contracting as well as in evaluating utilization of care and implications for the purchaser's financial situation. Second, providers need some degree of managerial and financial autonomies to be able to react to the incentives set by the provider payment method. This includes decisions such as adjusting the number of beds and staff. Third, consumers must have a choice of providers so that the money follows the patient. When implemented together with increased management autonomy and population choice, reforms such as capitation PHC payment and case-based hospital payment create financial incentives for providers to make more cost-effective decisions about internal resource allocations (for example, staffing). Legal changes are often needed on health financing, contracting, provider payment, the basic benefit package, and the health workforce in order to move from passive to strategic purchasing. 6.2 PURCHASER-PROVIDER SPLIT 221. Middle- and high-income countries have adopted an indirect-provision model, also called "purchaser-provider split." Indirect provision separates the purchasing organization from the organization that provides health care, resulting in two organizationally independent entities: the third-party payers acting on behalf of patients, and the health service providers. The payer or purchasing agency ideally would rely on market competition and select the highest-quality providers at the lowest price and contract with them for health care on behalf of the insured. 222. The purchaser-provider split exists in two forms. First, a government agency (such as a SHI fund) acts as the purchaser and buys the services from public or private 85 health care providers who are financially autonomous. This model is common in Europe and other countries with SHI. Eastern European countries that have moved in this direction include Armenia, the Czech Republic, Estonia, Georgia, Hungary, Latvia, Lithuania, Poland, Romania, the Russian Federation, Slovakia, and Slovenia. In tax- funded systems (United Kingdom, Sweden, Italy, Portugal), purchasers are agencies or private fundholders acting on behalf of the government. Purchasers are responsible to the budgetary authorities for cost control and to patients for the quality and accessibility of care through contracts with providers (Docteur et al. 2003). The government delegates the purchasing function to private intermediaries (such as GP fundholders in the United Kingdom and community health boards in Tanzania) that manage the health funds on a community level. These intermediaries act as agents of the government and can purchase health care from public as well as private providers for the people (Hsiao 2007). 223. The absence of any purchasing activity reduces the role of the MEF and municipalities to simple disbursement agents. The public financial management system executed by the MEF Treasury manages the centralized budgets paid to hospitals, which is defined based on economic categories. The resulting incentives to providers are to disregard patients' preferences and needs, leading to widespread shortages of essential drugs and other supplies, and major difficulties in the supply chain of pharmaceuticals to the primary care level, affecting availability of and access to even basic drugs and medicines from the essential drug list (Seiter 2007). 6.3 PURCHASING A BASIC BENEFIT PACKAGE 224. Kosovo's current benefit package is comprehensive, and financially not affordable from public funds, resulting in regular stock-outs of basic material and drugs in health facilities. The current benefit package is supposed to provide coverage in all public health facilities and a limited treatment package abroad. However, in reality more than 80 percent of drugs purchased by patients in Kosovo are financed out of pocket, mainly because the small, government-run program for essential drugs is underfunded and suffers from logistical problems. 225. A basic benefit package should only include health care that is financially affordable to the payer and improves health status in a cost-effective manner. The definition of the benefit package has proven a politically difficult process in most transition economies, and some countries have chosen to introduce a negative list of services, which are explicitly excluded from coverage. Other countries have defined a list based on analysis of high-frequency hospital admissions for conditions that could be treated cost effectively at the PHC level or based on interventions that permit the highest health outcome gains among the poor and vulnerable groups.26 Priority should be given to financing preventive care, and affording health protection to the poor, which can yield 26Several Latin American countries, including Mexico and Colombia, have defined positive lists for vulnerable groups based on the highest expected health gains. The Kyrgyz Republic defined its reimbursable drugs list based on an analysis of main diagnoses for hospital admissions that can be cost-effectively treated at the PHC level with drug therapy. 86 substantial improvements in health outcomes in a cost-effective manner. These might be services (for example, immunization) that would be provided free-of-charge or at a lower co-payment, while others would be provided against a co-payment (for example, pharmaceuticals) or full payment (for example, infertility treatment). Georgia had a basic benefit package that covered mostly PHC and some secondary care. Armenia has developed a similar package of outpatient services only, with secondary care only for the poor. 226. Defining the benefit package requires taking decisions on who will benefit from publicly financed services (breadth of coverage) and the services to be financed (depth of coverage). The benefit package also defines the levels of out-of-pocket contributions that beneficiaries will need to pay and where, and under what conditions services can be accessed. The Kyrgyz Republic has developed an innovative basic benefit package that has shifted drug benefits to outpatients (figure 6.1). The package is divided into three components that are paid by different funding sources. Thus the breath and depth of coverage is defined by the level of funding available, which ensures the financial sustainability of the system. Early results of the new benefit package design show a reduction in informal payments and increased revenues through formal channels. Figure 6.1 - New Benefit Package: Kyrgyz Republic Funding Source Benefits Private U ncovered Services (non-contracted) Private, out-of-pocket Co-paym ent tp pt M Services Service Social Fund payroll tax; mex emx E E Supplemental B enefits: reduced Co Co R epublican B udget yll "uninsured" copayment,H outpatient drugs ntracted vera transfers for "insured" yll Fu tiaraP ge (De I B udget ­ local by Basic Benefit Package: free primary care from enrolled governm ents buy pth) FG P, referral care w ith co-payment, from w hich defined M universal coverage for H F their populations population groups are fully or partially exem pt IF 0% 100% Population C overage (Breadth) Source: Kutzin et al., 2002 Note: FGP=Family Group Practice 227. Actuarial costing of the benefit package is essential to ensure a financially sustainable risk pool. Accurate estimates of the benefit package and of costs determine the financial sustainability and survival of risk pools. Actuarial methods require information such as household utilization data broken down by age, gender, employment, and income. Furthermore, utilization and unit costs will have to be projected to reflect 87 moral hazard, supplier-induced demand, and price increases associated with the introduction of health insurance. For example, health services of higher quality than in the past, without financial barriers to their use, are likely to lead to a large increase in service use. Thus, the level of finances transferred to the risk pool needs to be determined based on accurate actuarial costs of the membership pool. This process requires technical skills and data and involves the following steps: (i) defining the benefit package, (ii) conducting a cost analysis of the benefit package, (iii) organizing political consultations to discuss the benefit package and its costs, and health revenues available to fund the package, and (iv) adjusting the benefit package to ensure financial sustainability, and conducting new actuarial cost analyses to confirm this. That is, if total revenue needs based on analysis exceed the total available resource envelope, then the benefit package will have to be reduced to ensure the financial sustainability of the system. 228. The challenge for Kosovo will be to introduce a benefit package with financially realistic but socially acceptable limits, while ensuring equity in utilization and financing of health care. A limited basic benefit package (for example, covering public health and preventive care services) could be made available to all and fully financed by current general tax revenues. The cost of a more extensive benefit package (financed primarily by general revenues) could be supplemented by different levels of co- payments, while certain services would be excluded and on the negative list. To ensure equity in access to care, some low-income groups would enjoy access at conceivably lower co-payments or be exempt from cost-sharing. For example, in Hungary, population groups and services exempt from fees include (i) children and adolescents under 18 years of age; (ii) homeless people and Roma; (iii) emergency services as defined in a separate law (emergency status involves the patient's potential death or permanent damage to health); (iv) preventive services, public health services, pregnancy, and obstetric and neonatal care services. Guidelines about medical necessity need to be determined so that the insurer is able to deny coverage for any service not medically necessary. For example, the Swiss health insurance law (Article 32) dictates that health care included in the basic benefit package must meet the criteria of effectiveness, efficiency, and efficacy. 229. The benefit package can offer some coverage for essential drugs for the entire population, though a related regulatory system for this would need to be established. Including pharmaceuticals in insurance coverage would require a contract between the insurance organization and providers in the public and private sector (pharmacies and health facilities). Insurance members could then obtain their medicines from all outlets that have contracts with the insurance fund, either for free or at a defined co-payment. The necessary regulatory and contractual framework for such a solution includes (Seiter 2007): ˇ a strengthened regulatory authority that ensures quality of drugs on the market and high professional standards in manufacturing and distribution; ˇ a defined, transparent process to select drugs for the reimbursement list, based on public health priorities and cost-effectiveness; ˇ a purchasing process that creates competition between qualified providers and makes pricing transparent for the public; 88 ˇ an insurer whose contract with providers specifies quality standards and prices at different levels of the benefit package; ˇ monitoring of utilization to avoid system abuse and overprescribing; linked with contractual incentives for rational prescribing and use of medicines; ˇ transparent cost-sharing rules for patients (for example, 10 percent of drug price is paid by patient). 230. Contracts are increasingly used as a legal base for collaboration between purchasers and providers. Contracts define the price a purchaser will pay for a specific benefit package purchased from a provider at specified conditions. To be an effective tool for managing health care costs, the purchaser must have discretion and flexibility in contracting with all types of health care providers. Thus, purchasers often have a legislative mandate to contract with both public and private providers. However, the low payment rates often discourage private providers from seeking contracts. This has prevented competition among providers and thereby not fully utilized possible market mechanisms to increase efficiency. Russia, for example, enacted legislation in 1993, but its insurance purchasers have never contracted with nongovernmental providers. Hence, a purchasing organization should have a specific vision about contracts and contracting processes (box 6.1). Box 6.1 - A Possible Contracting Strategy for Kosovo A contracting strategy would include the following decisions: Establish criteria for selecting providers for contracting Specify clear contracting objectives such as fair provider compensation, access to services, budget limitations Establish a priority list of providers Give priority to quality, efficient providers instead of to providers that charge less but overprovide services Evaluate provider payment alternatives to line-item budgeting, and set some pay for performance conditions in the contract Allow special terms Achieve budgetary goals through contract management and not spending limits only Source: Adapted from: Korjenek 2006 6.4 STRATEGIC PURCHASING 231. Most South Eastern European insurers use passive purchasing. Under passive purchasing, contracts between purchasers and providers are defined based on providers' working plan for the future year and the related input factors (for example, number of staff and beds). Insurers are disbursement agents who follow a predetermined budget or simply reimburse bills based on a fee-for-service basis. To operate within the available budget, where hospitals have little opportunity to adjust their resources and input mix, providers modify their work plans resulting in volume ceilings for treatments and waiting lists. Under passive purchasing, provider performance or outcome results 89 (for example, patient satisfaction, infection rates, ALOS, or bed occupancy rate) have little relevance in contracting or provider payment. 232. Purchasing tends to be limited by accurate information and analysis on provider performance. In South Eastern Europe, detailed and valid data are rarely available, as many hospitals and PHC centers as well as insurers lack the necessary IT. Also, information connectivity across facilities and payers is incomplete, and patient information tends to "get lost" once patients change providers. Data collection is mostly on paper, causing the purchaser and Public Health Institute (PHI) to receive incomplete or low-quality data. 233. Strategic purchasing would require an active purchaser and investment in legal, organizational, human, and information capacity. Insurers will need to decide with whom to contract for which health care services, at what quantity and price, and how to pay for it. To make these decisions, a purchaser will need a unit in charge of provider performance analysis that receives adequate information to assess provider performance, conduct analysis, and forward results to the contracting department, to the actuaries for rate-setting, and to the quality management department for utilization and disease management. Such data collection and analysis would need investment in an HMIS in all health facilities and the insurance company. Insurers would have to publish provider performance results so that consumers were better informed and could select better providers. Few insurers have used their purchasing power strategically, mainly because they face strictly defined entitlements by providers and government price regulation, resulting in limited space for price negotiation with providers. 234. Implementing strategic purchasing in Kosovo would require constituting a purchasing agent, and investment in data collection, performance analysis and public reporting. Kosovo's HMIS is inaccurate, and there is no patient registration system in PHC facilities.27 The purchaser organization would need a provider performance department in charge of monitoring and evaluation of service provision and comparing performance across health facilities. Health facilities would be profiled against each other, and flagged for quality supervision when reporting low performance. The insurer's contracting department could then use results in defining contracts with providers. Legal and procedural revisions might be needed to allow flexible price-setting in contract negotiations between the purchaser and providers. A report about providers' key quality performance should be made widely accessible to consumers, posted on the insurer's Web site and published in the local newspaper. 6.5 HEALTH FACILITY MANAGEMENT AUTONOMY 235. In OECD countries, health facility managers have management autonomy, and are kept responsible for the facility performance, as most of PHC is privatized. Management autonomy goes along with organizational changes in health facilities resulting in 27In Albania, the patient registration system for PHC and the population census differ by 1 million individuals. See World Bank 2006. 90 privatized PHC in European countries and increasing private participation in the hospital sector (see box 6.2). Management responsibility for key areas include hiring and firing of staff; determining the number of staff and its skill mix; doing financial management; determining the level and scope of activities; and deciding on capital developments, the number of beds, and technology mix. Facility managers are responsible and are held accountable for cost and facility performance, and they are fired if performance targets (financial results, patient satisfaction, staff satisfaction) are not met. Provider autonomy must be accompanied by increased transparency, reporting, and annual financial and performance audits to prevent management failure. Box 6.3 describes the main organizational structures for hospitals that currently exist in central Europe. Box 6.2 - PHC Provided by Private Sector In Bulgaria, PHC physicians have the legal status of independent contractors rather than civil servants. All PHC physicians who wish to contract with the National Health Insurance Fund (NHIF) must register as a single or group practice and contract with the NHIF. Some group practices employ a practice manager to handle financial affairs, logistics, and interactions with the NHIF. A doctor who does not contract with the NHIF may provide services to private patients on a private pay basis. Most public polyclinics have been transformed into diagnostic and consultation centers or medical centers and registered as trade companies. These new organizational forms are housed in the buildings of former polyclinics, owned by the municipalities. Single and group practices have the right to acquire ownership of premises and medical equipment or to pay low rents to the municipality for consulting rooms in the former polyclinics. Alternatively, privately owned premises or rented privately owned offices may be used. Source: Adapted from European Observatory on Health Systems 2003 Box 6.3 - Organizational Structures in Hospitals Country Organizational status Director appointed by: Croatia Separate legal entity Mayor Estonia Three types of non-tertiary hospitals (a) (a) Mayor Slovakia municipal not-for-profit (b) joint-stock company (b) company board law (c) trust form (c) trust board Hungary (a) Separate legal entity that can enter into Municipality contractual arrangements with HIF and private (assembly or mayor) sector, (b) budget unit subject to public finance law Lithuania Not-for-profit institutions under Law on Health Owning local Care Institutions from June 1996 government Poland Budgetary units; legislation passed on Owning local `independent units' government Romania Extra-budgetary units governed by the Public District health Finance Law from 1998 authority based on MOH criteria Source: Based on McKee and Healy 2002 91 236. Insurers contract and purchase health care from licensed and accredited providers. With a licensing system, the government grants permission to an individual practitioner or a health facility to operate based on minimal standards. The development of a strong licensing system, based on transparent licensing standards to ensure a minimum level of hospital care, generally precedes the development of an accreditation system that aims at continuous quality improvement (see box 6.4). In OECD countries, health insurers generally contract with licensed and accredited providers, to ensure quality care for their members. Box 6.4 - Accreditation of Health Care Providers Accreditation is a means to systematically promote the continuous improvement of health services quality. Accreditation combines internal improvement and assessment with external assessment mechanisms, and uses a set of standards, specifically designed or adapted for a country's need. Hospitals may request accreditation of an international accreditation agency, but most countries prefer to develop their own system by adapting existing models and standards for services. Various types of health care providers can be subject to accreditation: hospitals, PHC centers, GP offices, etc.; specific activities or departments can also be accredited (laboratories, etc.). In most accreditation systems, providers do their own internal assessments, working continuously to reach the required level (standard) of care, while an accreditation body provides periodic external assessment and grants accreditation status to the surveyed provider. The status may be with or without conditions, for the standard period or a shorter period. (In the latter case, needed improvements would be specified.) Various models exist - mandatory or voluntary, public or privately organized - and countries may have more than one accreditation body (for example, for PHC, for hospitals). Many accreditation bodies also provide training for quality management staff at the provider level as well as for external surveyors. Standards of service delivery are periodically upgraded to match the development of the health sector in general and medical and nursing science and practice in particular. Standards can cover all aspects of health care provision, that is, physical and human resources infrastructure, management and administration, organizational components, various processes (clinical and administrative), as well as the desired outcomes of the care process. If clinical practice guidelines are developed and endorsed, they should be referred to in the accreditation standards. International experience indicates that financial and institutional questions need to be dealt with up front to ensure that investments in the development of an accreditation system are not in vain. Some countries have come to realize that the operating costs for an accreditation system should be mobilized only after making the initial investments. Successful introduction of quality improvement mechanisms in general and accreditation systems in particular also call for: (i) awareness-raising about the importance of continuous quality improvement among the hospital staff and patients; (ii) active involvement of hospital management in the quality improvement process, (iii) appointment or designation of special staff in the hospitals to deal with quality issues on a daily basis and to assist in the preparation for the internal assessment as well as for the external survey; (iv) training of a sufficient number of external surveyors (eventually with a train-the-trainers program); (v) a legal framework; and (vi) involvement of the HIF or other purchasing agency. Source: Based on World Bank 2006 92 6.6 PROVIDER PAYMENT SYSTEM FOR OUTPATIENT CARE 237. The provision of PHC in Kosovo is financed by an earmarked health grant, transferred from central general revenues to municipalities. The grant is calculated annually based on a simple capitation formula taking into account the total population in a municipality. Allocationi = PerCap × POPi where PerCap is the total health grant divided by the total population and POP is the population of each municipality as reported by municipalities. Capitation payment requires a population census and a system of patient registration in PHC, neither of which exist in Kosovo. Capitation based on an unknown target population results in incorrect payments. This simple capitation system creates incentives for public providers to report a higher population number registered and reduce the workload by referring patients to secondary providers or their own private practice. 238. PHC in most European countries is increasingly paid by capitation. While in Finland, Greece, and Iceland, GPs are still government employees and paid a salary that is negotiated centrally (between physicians' associations and the government), other European countries mainly pay for outpatient care via fee-for-service or capitation. Fee- for-service is widely used for specialists working in ambulatory care. Fee levels are either negotiated centrally (as in Japan, Canada, Germany, and in France) or set by the individual practitioners (Docteur et al. 2003). Cost escalation under fee-for-service caused many countries to shift to capitation payment, mostly case-mix adjusted to account for differences in the severity of illness in the registered population. In the 1990s, with the adoption of SHI systems, Croatia, the Czech Republic, Slovakia, and Ukraine moved from input-based payment to fee-for-service reimbursement, which quickly led to increased activity levels and put financial pressures on purchasers. Insurers reacted and put ceilings on the total amount, or negotiated volume contracts within a capped budget, or prospective global budgets with activity caps. The Czech Republic moved from salaries to fee-for-service and then to capitation. Ireland shifted from fee-for-service to capitation, leading to an estimated decline in doctor visits of 20 percent. In Slovakia, the cost increase caused by fee-for-service led to a quick move to capitation in 1994, and in 1998 to a 60:40 capitation/fee-for-service mix (Langenbrunner et al. 2005). Thailand's universal SHI uses a capitation payment method to force providers to vertically integrate their services and improve the continuity of care (Hsiao 2007). 239. Most South Eastern European countries apply simple capitation, adjusted based on age and gender, and geographical differences. In Albania, the HIF pays PHC practitioners based on a modified capitation basis: base salary plus capitation supplement depending on location and number of registered patients. PHC personnel, operations, and maintenance costs are paid from a different source, which gives the physicians limited control over the performance of their entire operation (World Bank 2006). Montenegro is implementing a mixed capitation and fee-for-service formula, to prevent PHC providers from skimping on services. To limit the providers' financial risk, 93 salaries tend to be excluded from the capitation amount and funded separately as a budget line item. 240. Simple capitation sets financial incentives to providers that may lead to underprovision of care and exclusion of high-risk patients (Ellis and McGuire 1996). To prevent negative effects, capitation often includes some sort of case-mix adjustment and output-based incentives for delivering specific services including preventive care measures. Most capitation mechanisms are based on the following need and cost formula for funding to an area or a provider i: Allocationi = PerCap × POPi × (1+ ai )× (1+ ni )× (1+c ) where (Ensor 2005): ˇ PerCap is the per capita budget (total allocation divided by population registered); ˇ POP is the population registered with each PHC provider; ˇ a is an adjustor for needs by age and sex (providers with more elderly or women of reproductive age than the national average would receive a higher allocation); ˇ n is an adjustor for health need (a provider with specially high morbidity might receive more funding); ˇ c is an adjustor for cost (provision of care in a sparsely populated area might cost more). Once information and analytical capacity has been established, the current simple capitation formula in Kosovo can be expanded to account for PHC providers' different case mixes and reward better quality care. 241. Monitoring and evaluation of provider performance will detect and prevent negative effects of capitation. Capitation sets an incentive to produce efficiently by adjusting the treatment intensity within a medically acceptable quality range. However, providers also have an incentive to reduce their costs by encouraging healthier individuals to register, and exclude individuals with costlier health problems; or to overreport the number of patients registered with the providers, based on which the capitation budget is defined. In Albania, the population registration system is not properly implemented and the number of people registered with PHC providers is about 1 million more than Albania's total population. In Kosovo, the provision of care and the impact of the current capitation payment on quality and efficiency are not monitored. This highlights the importance for the future purchaser to monitor and evaluate provider financial and utilization performance and incorporate findings in contracting and payment. 242. Consumers need information about providers' performance to choose a good PHC provider. Consumer choice means that individuals are informed about provider performance and can freely choose their provider. While it is argued that consumer choice over PHC doctors, coupled with the principle of "money following the patient" may moderate the negative effects of capitation, using consumers as control agents would require several conditions to be in place. These include a PHC market large enough to 94 choose from, patient mobility, and consumers having the necessary information and ability to judge providers' quality of care. In reality it may be difficult for patients to identify substandard technical quality. Also, when providers are government-owned, effective choice is limited. Choice is particularly limited in remote geographic areas with only one provider (Cashin et al. 2005). Thus, to prevent negative effects under capitation, performance targets should be established and provider performance monitored and evaluated. Public reporting of performance results will allow consumers to be informed and register with better-performing PHC practitioners. 243. Kosovo can strengthen the implementation of capitation payment by investing in the delivery and health financing side. On the delivery side, management autonomy would need to be given to PHC facility managers and management responsibility centralized at the PHC director level. This change would allow facility managers to make decisions about input factors such as staffing, pharmaceuticals, and service provision, which could result in improved productivity and quality of care. Regular performance and financial audits in health facilities could help detect possible management failure. Investment in data collection in PHC centers would be needed, including patient registration, utilization, finances, and quality of care, and the data need to be forwarded to the purchaser department in charge of provider performance analysis, and to the MOH. On the financing side, results from analysis on expenditure, utilization, and case-mix patterns can be used to adjust the per capita budgets paid to providers for differences in costs across age/gender groups in different PHC facilities. 6.7 PERFORMANCE-BASED CAPITATION PAYMENT 244. Some countries have started to pay providers based on performance to improve quality of care. Output-based reimbursement methods such as fee-for-service or per-case payments provide little financial reward for better quality of care. Traditional strategies that stimulate quality improvement include regulation, measurement of performance and subsequent feedback, and marketplace competition. The motivation to pay providers based on performance comes from a response to rising medical cost trends, the growth in health care utilization, and demands by purchasers and patients for improvements in the quality of care (Dudley et al. 2004). P4P provides a financial reward for improved performance outcomes and pays different amounts to providers based on their performance differences. P4P is designed to reward better-quality performance and prevent negative consequences such as inappropriate referrals to specialists and hospitals and substandard quality of care, which could occur under simple capitation. 245. Few South Eastern European countries have performance-based payment. In FYR Macedonia, the capitation amount for PHC providers is adjusted by age, gender, and region, with higher amounts paid to providers residing in mountainous areas. To limit skimping on care, performance is measured for a series of indicators related to preventive care, immunization, diabetes, cardio-vascular diseases, cancer prevention, prescription medicines, referrals, issuing of sick-leave certificates, etc. Of the total monthly capitation amount, the 70 percent base payment is paid monthly while 30 percent is withheld to be paid at the end of each quarter, based on quarterly performance 95 evaluation of the agreed benchmarks. Capitation covers all recurrent expenditures including salaries of privatized physicians and other input factors such as materials to treat patients. A performance payment requires information on quality compliance, and cost-relevant factors of the registered population such as severity of illness. 246. Capitation payment for PHC has been adjusted to pay for better performance in the United Kingdom and the United States. Most performance- adjusted capitation formulae include performance variables, such as: (i) patient satisfaction, (ii) clinical process, (iii) outcome, (iv) IT, and (v) risk-adjusted efficiency indicators. In the United States, P4P focuses on improving PHC measures including the Health Plan Employer Data and Information Set28 (HEDIS) scores (box 6.5), patient satisfaction, physician access or electronic claims submission (Baker 2003). Results are still largely anecdotal, and P4P participation among providers is largely voluntary. Performance-based capitation has encountered several difficulties, including substantial underestimation of total payments, defining measures across different reporting initiatives, risk adjustment for outcome measures, and administrative burden on smaller health facilities (Rosenthal et al. 2005). Box 6.5 - Examples of Performance Indicators of Access/Availability of Care ˇ Adults' access to preventive and ambulatory health services ˇ Children's and adolescents' access to primary care practitioners ˇ Prenatal and postpartum care visits ˇ Annual dental visit ˇ Initiation and engagement of alcohol and other drug dependence treatment ˇ Call answer timeliness Source: HEDIS, USA. www.ncqa.org 247. P4P can result in considerably higher costs for the purchaser. In the United Kingdom, the NHS introduced a P4P contract with family practitioners in 2004, and committed $3.2 billion in additional funding over a period of three years for the program. Results from the first year show that financial incentives affect physicians' behavior. Providers attained a median of 96.7 percent of the available points for clinical indicators, which greatly exceeded the 75 percent predicted. Consequently the cost to the payer was considerably more than expected. The P4P program increased the gross annual income of the average family practitioner by US$40,200. From this amount, the family practitioner paid for any additional nursing and administrative costs of meeting the targets. 248. Public reporting of provider performance results may have a similar effect as P4P and is less expensive. In the United States, the evidence base linking P4P programs to better quality of care is thin (Epstein 2007). P4P is often introduced in combination 28HEDIS is a tool created by the National Committee for Quality Assurance (NCQA) to collect data about the quality of care and services provided by the health plans in the United States. HEDIS consists of a set of performance measures that compare how well health plans perform in key areas: quality of care, access to care, and member satisfaction with the health plan and doctors. NCQA requires health plans to collect this information in the same manner so that results can be fairly compared with one another. Health plans can arrange to have their HEDIS results verified by an independent auditor. 96 with public reporting of performance results. Research suggests that the incremental effect of P4P over public reporting is small, resulting in around 3 percent performance improvement over two years, with the largest improvements observed among hospitals that were poorest performers. The costs of administering P4P programs are likely to be higher than those for public reporting of provider performance programs. 249. The MOH in Kosovo is currently experimenting with some form of P4P for PHC and could draw on lessons learned with P4P in the United Kingdom and the United States when adjusting its capitation formula at some later stage to reward better- performing providers: ˇ P4P programs are costly and require substantial additional investment in IT systems to monitor performance; ˇ a baseline analysis and careful monitoring and evaluation of progress are needed to avoid paying for improvements that have already occurred and to prevent abuse. ˇ information about provider performance needs to be transparent and made available to consumers through public reporting; and ˇ health facility managers need to have the management power to respond to the financial incentives set by P4P leading to active quality and cost management. These reforms would have to go along with management autonomy in PHC centers to allow directors to respond to the new financial incentives. 6.8 HOSPITAL PAYMENT MECHANISMS 250. Hospital payment methods include: (i) line-item budgets; (ii) per diem (per bed day); (iii) fee-for-service; (iv) case-based payment such as DRGs for each case hospitalized; (v) global budget, (vi) per episode of illness; and (vii) capitation per insured member. Per diems, DRGs, and fee-for-service all set an incentive to increase the number of days, cases, or services, and they pose less financial risk to the hospital than simple capitation. 251. Line-item budgets create an incentive to underprovide services, and increase the numbers of staff and beds on which the budget is based. Line-item budgets are determined prospectively, at the beginning of the budget year, and are based on projected input use, including the number and type of staff employed in the hospital and controls on nonsalary expenditures. This budget formulation creates relatively low administration costs, and there is limited need for information systems. It is also a central planning and budgeting tool for rigid control of government expenditures and often results in rationing and cost-shifting to patients who pay out of pocket, raising concerns about equity in access to care. This method of payment sets only weak incentives for hospitals to adopt innovative management and improving efficiency. In Albania, for instance, input-based hospital payment caused a lack of provider accountability for low-quality performance and a high level of informal payments in hospitals (World Bank 2006). 97 252. Kosovo's hospitals are still paid a line-item budget to cover operating costs, while several countries have started payment reforms. Most countries in the South Eastern Europe region have started the process for hospital payment reforms, with most of them moving from input-based to some kind of case-based payment with the long-term objective of implementing DRGs. Implementing these reforms requires increased managerial autonomy in hospitals to ensure that directors can react to the financial incentives set through the payment system. 253. The trend in hospital payment reform is toward paying hospitals based on DRGs with budgetary caps to control hospital expenditures. To contain growth of hospital expenditures and improve productivity, hospital payment in Western Europe has gradually moved from fee-for-service and per diems to global budgeting and case-based payments, such as DRGs (Docteur et al. 2003). Global budgets have been developed in response to volume problems under per diem and per case payment systems. A global budget at the hospital level is a payment fixed in advance to cover the aggregate expenditures of that hospital over a given period to provide a set of services that have been broadly agreed upon. Budgetary caps are widely used for setting volume limits and controlling hospital expenditure, and are often complemented by spending caps on subsectors, including ambulatory care and pharmaceuticals. Case-based payment mechanisms such as DRGs29 reflect the average cost of producing a "case" in an average hospital, which may be adjusted to account for regional economic conditions. The U.S. Medicare system, Australia, and several countries in Europe started experimenting with DRGs in the early 1980s. A number of middle-income countries have introduced case- based payment systems, including Korea, Taiwan, Thailand, and Hungary (Langenbrunner et al. 2005). Serbia is in the process of moving from line-item budgets to case-based hospital payment, which should help reducing high bed numbers and relatively long ALOS in hospitals (WHO 2006b). 254. To develop DRGs, hospitals have to document and report all diagnosed cases based on diagnoses and procedure codes. The clinical data necessary to develop DRGs include age and sex of the patient, the International Classification of Diseases (ICD-9 or ICD-10) code for the primary diagnosis, the length of stay, and other details, such as whether there was a surgery and whether the patient spent time in intensive care. Actual cost data need to be collected on the patient at the department level to represent the relative costliness of producing a DRG. In a DRG payment system, the hospital revenue is the total sum of DRG points multiplied by the base rate, which reflects the aggregated average cost per hospital case across all or a representative group of hospitals. DRG payments increase the awareness of resource utilization and set the incentive to increase the number of discharges and productivity resulting in shorter hospital stays. In the United States, moving to case-based payments led to a reduction in the ALOS by 15 percent in the first three years after the DRG payment was implemented (Cashin et al. 2005).30 29Cashin et al. (2005) provide detailed technical information in a user manual on the development and implementation of DRGs. 30Note: The manual written by Cashin et al. gives detailed explanations and guidance about how to implement DRG payments in hospitals. 98 255. Hospitals will have to invest in administrative systems (information and billing) to report their cases and be reimbursed by the purchaser. DRG payment requires investment in hospital HMIS and data collection. The purchaser will need an information system that computerizes the recording of cases by hospitals and the grouping of cases into payment categories. Additional staff is needed and they must be trained in the collection of activity and cost data and clinical coding. An agency needs to be made responsible for receiving the relevant data, applying the case-mix grouper, monitoring quality standards, and benchmarking (Wiley 2007). 256. Managing hospitals under DRG payment requires modernizing disease management beyond the hospital, focusing on discharge management. To respond to financial incentives, hospitals will have to improve productivity, and managers will need to select the input mix that allows producing a case at the lowest cost while maintaining quality. DRG involves collaboration with outpatient and community care providers who may need to provide a greater degree of follow-up to discharged patients, as their hospitalizations will become shorter. 257. Implementing hospital payment reforms is a long-term process and needs to be accompanied by increased managerial autonomy in hospitals. The U.S. Medicare system began reimbursing hospitals with a case-based payment using DRGs in 1983, and it is now in the process of adding performance-based payment criteria to reward better- quality care. Hospital payment reform needs inter alia constant refinement of rates based on cost and financial analysis, and legal and procedural changes to increase managerial and financial autonomy of hospitals and allow hospitals to react to the financial incentives set by the payment system, and substantial investment in human capital, organizational performance, data collection and management, and financial and disease management. 6.9 IMPLICATIONS FOR PURCHASING IN KOSOVO 258. Kosovo should follow international experience and introduce a purchaser­ provider split. The split between purchaser and providers reduces administrative rigidities generated by hierarchically structured command-and-control models. Hospital and PHC facilities could be corporatized as nonprofit autonomous organizations, competing for patients with private for-profit providers. On the municipality level, a community oversight body should be formed to supervise the functioning and performance of the PHC center or hospital and ensure transparency in major financial and operational decisions. 259. Management decisions should be centralized at the facility director level. Health facility directors will need greater autonomy and flexibility to respond to financial incentives and improve productivity. Following the subsidiarity principle and the Kosovo Health Law,31 the responsibility and accountability for management and financial decisions rests with the hospital and PHC directors. Management autonomy involves expanding management responsibility for key areas including hiring and firing; 31Sections 18 and 28. 99 determining the number of staff and its skill mix; doing financial management, determining the level and scope of activities, and deciding on capital developments, the number of beds, and technology mix. This autonomy needs to be accompanied by professional financial management in health facilities with transparency, reporting, and annual financial and performance audits to prevent management failure. Managers must be equipped with the necessary resources and tools to make decisions, and kept accountable for performance results. The following steps lead to management autonomy in health facilities (Widmer 2007). (i) Facility manager: Director takes the role of a CEO, who is responsible, will be held accountable for cost and facility performance, and will be fired if performance targets (financial results, patient satisfaction, staff satisfaction, occupancy rate, and ALOS) are not met. The hospital director reports to the hospital board and the minister of health. This is a full-time position for an experienced manager. If the director is a physician, then he/she should not be allowed to see patients. (ii) Staff management: Decisions about staff hiring and firing, remuneration, and fringe benefits should devolve to hospital and PHC facility directors who select a staffing mix that increases revenues, productivity, quality of care, and patient satisfaction. Directors may release redundant staff, and introduce part- time employment and P4P-based salaries to all staff based on patient satisfaction. Existing staff allocation and efficiency should be analyzed. Legal changes to the civil service and labor law may be needed to allow flexibility in employment of health staff. (iii) Financial management: Financial managers and health staff are trained on how to improve productivity. Facilities devise a business and investment plan to manage finances and prevent financial loss, and allocate funds efficiently to pay for recurrent costs through provider payment. They conduct an annual independent financial audit and publish results. (iv) Other input factors (for example, drugs): Facilities identify ways to improve efficiency in provision of care by changing the quantity and type of drugs (generics), supplies, and other input factors, and by working in networks with other providers (Seiter 2007). (v) Physical assets: The facility director under supervision of the board and within the provisions of the masterplan has decision power about disposing of existing capital stock, including buildings and equipment, or acquiring new capital, merging departments, entering into private-public partnerships, etc. Hospitals and PHC facilities pay a capital fee (similar to rent) based on return of investment to owner. (vi) Organizational structure: The hospital manager implements the most efficient management structure, organization of departments, and ancillary services based on the masterplan, and is allowed to contract out services (laundry, kitchen, cleaning, laboratory, and others) to the private sector.. (vii) Output mix: The hospital manager makes decisions about composition of inpatient care versus outpatient and day-surgery, which are less resource- intensive and less costly. The objective is to shorten the ALOS and increase the bed occupancy rate. 100 (viii) Data and analysis: Facilities collect and analyze finance and utilization data for unit cost analysis and process evaluation; do a patient satisfaction survey with each patient, use results for salary definition, and publish results on the Web site or in the local newspaper; conduct quality and performance analysis on infection rate, re-admission rate, bed turnover, productivity, etc. and use results to change production of care and input mix (including staffing) and in decisions to contract out specific services to private sector. There is timely transfer of claims data from health facilities to purchaser. (ix) Use of surplus revenues: Facilities have a business plan that describes how to use profits generated from efficiency gains for investment in quality and facility improvement, staff training, and innovation to make facility more attractive. 260. A series of policy and financing changes would have to be implemented on the financing side to allow a future health insurer to become an active purchaser: (i) Make institutional and legal changes to allow the purchaser's contracting and analysis units to monitor and evaluate provider performance, negotiate prices and volumes with providers, set incentives to contain growth of expenditures, and contract selectively with accredited providers. (ii) Create a central risk pool and a purchaser that contracts with providers for price, quality, and volume of services. The purchaser has a performance department that conducts provider performance analysis to define which services should be purchased based on the legally defined basic benefit package, from which health facilities and how providers are paid. (iii) Invest in IT and data collection capacity at the purchaser/health insurer and in health facilities. Collect provider and patient data on the provision of care, demographics and health status, and finances to evaluate provider performance and use results in purchasing decisions to contract selectively with providers. 261. The capitation amount paid to PHC providers should be revisited and be based on factors such as (i) the number of individuals registered with the PHC facility; (ii) the age/gender distribution of registered population; and (iii) possible performance indicators, to measure performance of providers and public reporting of results (for example, preventive care compliance). Kosovo would have to implement the following steps to support capitation: (i) Do a population census with sociodemographic and economic data to calculate the overall capitation budget and adjustment payments for higher- risk patients. (ii) Do population registration with PHC providers that is compatible with census data. Ensure that vulnerable groups and minorities are included in PHC registration. (iii) Collect data on PHC and use information for population profiling and case- mix adjustment. 101 (iv) Build capacity in health facilities and purchasing agency for establishing claims database. Claims data are collected and transferred to purchaser to evaluate provider performance and detect underprovision or substandard quality of care. (v) Publish provider performance results regarding patient satisfaction, cleanliness, infection rates, etc. in a local newspaper and on the Web site of the purchaser and the MOH. (vi) Monitor and evaluate the impact of capitation payment on provider costs, and quality and utilization of care. Based on the findings, the capitation formula should be further refined to ensure the resulting financial incentives contribute to overall health policy goals. 262. If Kosovo plans to reform its hospital payment system by moving from input- based to some case based payment, investment in a number of factors are needed32: (i) Create enabling legal and institutional settings that support the effectiveness of hospital payment reforms. For example, labor laws and regulations may interfere with policies to grant hospitals autonomy over hiring and firing staff or setting salary levels. (ii) Define contracts between purchasers and providers, including private providers. Develop contracts that specify which services providers agree to deliver and what prices the purchaser agrees to pay, which party has the authority to make which decisions, and what recourse is available to each party if the terms of the contract are not met. (iii) Develop the analytical and management capacity of the purchaser and provider to manage the new payment system, including capacity to develop and implement purchasing contracts, manage information systems and quality assurance systems, and monitor and evaluate purchasing policies. (iv) Build information and financial management capacity among providers to manage their internal resources, including accounting, billing, and information system. (v) Invest in two main components of a basic information system to support the development and implementation of a case-based hospital payment system, both of which are established at both the provider and the purchaser level: (a) hospital case database, including basic discharge information about each case at each hospital included in the payment system; and (b) financial database, including cost accounting and expenditure information. (vi) Build necessary care management capacity at the provider level in the wider health system to support implementation of DRGs; include setting up continuum of care paths, home care agencies, and improved referral practice. 263. The purchaser must have the competence to select qualified and accredited providers (public and private alike), set performance standards, bargain with the providers on payment methods and rates, and monitor their performance. The staff performing these functions must have a strong public interest and be willing to confront 32Cashin et al. (2005) provide a thorough description about the enabling factors for implementing DRGs. 102 and negotiate with public and private providers. The government needs to be willing to let poorly performing public hospitals and clinics close, or change their management. Whether purchasing is successful in Kosovo will depend on the institutional competence of a government agency and governmental accountability (Hsiao 2007). 103 CHAPTER 7. CONCLUSION AND RECOMMENDATIONS 264. Kosovo choosing health financing reforms that lead to the implementation of health insurance will entail substantial changes in the roles and responsibilities of various actors in the health sector and require concerted efforts to strengthen the accountability framework of the sector. The MOH would assume a policy-making and stewardship role, and implement reforms in quality assurance in the provision of care, organization of health facility management, and management of resources including staffing and pharmaceuticals. The MEF would have to redirect the financial flow to the health insurer, which will assume full responsibility for risk pooling and purchasing of health care. The insurer will contract with accredited health care providers who offer a defined set of services to the population against an established price. Service providers would be given increased autonomy to effectively produce these services and the defined benefit package and their performance evaluated against an established set of performance standards. Actuarial cost analysis of the benefit package should be conducted to estimate the future funding needs, adjust the benefit package to an affordable level, and propose a composition of funding from public and private sources that leads to equity and financially sustainable health financing. These changes will require investments in legislation, regulations, new governance and organizational structures, and information management so as to ensure proper accountability of all actors. Implementing these levels of reforms will have cross-sectoral effects and require close collaboration. The role of current stakeholders including municipalities will need to be revisited and redefined based on an institutional review to ensure more efficient flow of funds and resource management. 265. Before embarking on such major health financing reforms, the government will have to discuss and make some substantial initial decisions and estimate their financial and organizational impact. As outlined in this report, the government will need a clear vision about what kind of health insurance would be financially and organizationally most feasible in Kosovo and what the implications would be for the current financing functions. Such a vision and the related strategy should be based on the overall health policy goals of the sector and taking into account the financial, institutional, organizational, and human resource situation. In a context with resource constraints and a small formal sector, for instance, an insurance fund financed with payroll taxes and private premium could result in pooling among the better-off and rationing of care, which would be against policy goals to improve equity in access to care. Table 7.1 provides an overview of the different choices policymakers will have to make for each financing function. Each choice has advantages and disadvantages and the best trade-off should be sought taking into account the macro- and socioeconomic, fiscal, labor, and health context of Kosovo. 104 Table 7.1 - Options for Different Health Financing Reforms Financing Current system Potential future health insurance system function Option 1 Option 2 Option 3 Collection TA TA Health insurer (or Health insurer agent social security Private insurers collect organization) Out-of-pocket for PHI or government payments collected if publicly run Out-of-pocket and kept by payments collected providers Out-of-pocket and kept by payments collected providers and kept by providers Funding General revenue General taxation and Mix of public Risk-rated or method out-of-pocket (payroll and general community-rated payments taxation) and private premiums paid by Eventually premiums premium, out-of- individual members for VHI pocket payments Enrollment Not applicable Mandatory all Mandatory for Voluntary residents formal sector Risk pool MEF for Single risk pool on Single risk pool for Multiple risk pools hospitals national level for paying members for enrolled Municipalities entire population only members for PHC Purchaser None; simple Not-for-profit Private not-for- Private for-profit disbursement independent single profit insurers insurance companies from MEF health insurance Provider Line-item budget Case-based in Global budget Fee-for-service and payment to hospitals hospitals per diem payments Line-item Capitation with in hospitals and budgets to PHC performance outpatient component in PHC Contracting None Selective contracting All providers in At insurer's with public and private Kosovo discretion providers Benefits Universal in Financially sustainable Basic benefit At insurer's public sector basic benefit package package to cover discretion, for PHC and hospital prevention only depending on care targeted to needs VHI to cover contracting of low-income groups additional services arrangements and to cover cost- at full cost paid by effective treatment private premium Eligibility Residents Residents Paying members Paying members Cost-sharing Formal and Co-payment Deductible and co- Co-insurance informal payment Regulatory / MOH Health Care Financing BPK International oversight MEF Supervisory Board Insurance body with stakeholders and Commission insurance professionals 266. Based on the analysis presented in the previous chapters, the most feasible health insurance option in Kosovo is comparable to the insurance funds of Estonia 105 and the Kyrgyz Republic. Kosovo's labor market is too small and fragile to absorb a payroll tax increase. Therefore, insurance revenues should mainly consist of general government budget funds from direct and indirect taxation. Kosovo could explore increasing indirect taxes on tobacco and alcohol, as well as on luxury goods, to raise additional government revenues. Under a tax-funded system, the collection agent is the TA that transfers funds annually to the HIF for risk-pooling. 267. A single risk pool would be enough for Kosovo's population, especially since Kosovo does not have the necessary institutional and risk-equalization capacity in place that would be needed for multiple insurance systems. This single risk pool should cover all levels of care (outpatient and inpatient) to ensure a continuum of treatment. To ensure solidarity among the healthy and the sick and limit adverse selection, insurance enrollment and eligibility for service use should be mandatory for all people living in Kosovo. To ensure political independence, the legal form of the insurer could be a public autonomous not-for-profit organization similarly as the KPST. At the beginning, Kosovo may want to consider contracting out of insurance management to an experienced international health insurance company that could also help build human capacity and IT in Kosovo. 268. When constituting an insurance fund, Kosovo may wish to follow the positive experience with the Kosovo Pension Pillar II Trust Fund and apply its organizational and management form to a future HIF. The Pillar II Trust Fund/KPST is a public autonomous organization managed by an independent professional team (Gubbels 2007). A future health insurer could be established as an independent autonomous organization similar to the KPST, to do risk pooling, purchase health care from providers in Kosovo and abroad, and administer the health insurance system. The insurer would be staffed by Kosovar and international management and employees recruited from the international diaspora working in health insurance companies, among international insurance experts, as well as locally. The Board of Governors could comprise a combination of local and foreign members. The regulatory and oversight body of the health insurer could be the BPK. Statutory reserves would have to be securely invested, according to international best practices, and yield positive real net returns to the insurer, with comparatively low administrative fees. In planning to staff a future insurance organization, the Board of Governors and its advisors should hire a small staff of highly motivated individuals, who would be compensated on par with other financial sector organizations in Kosovo. The foundation for supervision of insurance will need to be strong, with effective transparency requirements, as is the case in Estonia. An annual progress report could be conducted to evaluate whether the insurance system is comprehensively implemented, and to identify issues in insurance functioning including the effectiveness of collection, record keeping, IT systems, and information reconciliation processes. 269. The risk-pooling agent or HIF is also the purchaser of health care. Once reliable information systems and analysis have been set up, Kosovo could consider moving toward capitation payments for outpatient care adjusted by age and gender, and case-based payments for inpatient care. Only accredited private and public providers should be considered for contracting. Based on performance analysis results of providers, the insurer should contract selectively and exclude low-quality and inefficient 106 providers from contracts to protect patients' safety. This information should be made widely available. The basic benefit package of insured services should undergo an actuarial cost analysis, and a list of benefits defined based on the total insurance revenues available, to prevent the accumulation of insurance debts. Modest co-payments should be introduced at all levels of care with clearly defined and transparent exemption policies for vulnerable groups. Standard accounting systems should be installed in all health facilities so that the use of patient revenues is accounted for and reinvested in the provision of care. 270. Timing and sequencing is critical to allow the health system to prepare for and absorb the necessary changes and not overwhelm implementation capacities. A three-phase gradual approach is therefore recommended, with the Phase 1 focusing on preparation, and Phase 2 and 3 each extending over four to five years to implement, adjust, and complete reforms. Phase 1 is crucial and may take one or two years; it serves mainly to set up the legislative, regulatory, governance, and strategic directions to implement health financing reforms. Phase 2 focuses on implementing and adjusting reforms. Phase 3 is the completion phase where the new financing and delivery system should prove its sustainability. The timing and sequencing of key steps are presented in table 7.2. 271. The impact of health financing reforms should be carefully monitored and evaluated to identify and correct for possible negative effects at an early stage. Some of the financial and organizational changes may lead to unexpected adverse effects that would have to be corrected, such as inequity in health financing, or substandard quality of care. To monitor progress and detect negative effects, a monitoring and evaluation framework should be adopted that assesses baseline results before the introduction of reforms and conducts regular follow-up surveys to measure progress. A baseline could establish results based on several surveys on equity in utilization and financing of health care, patient satisfaction and care seeking behavior, quality of care and cost and efficiency levels in health facilities. Findings from the baseline could help to define the benefit package, levels of cost-sharing, and payment reforms. Baseline results can be compared against any future follow-up survey to identify progress as well as areas that would need strengthening, and to ensure that reforms contribute to policy goals such as equity, financial sustainability, and better quality of care. Any valid household survey data collection and analysis in Kosovo will require a population census, as the last census is from before the early 1990s. 107 Table 7.2 - Proposed phasing for health financing reforms Objective Phase 1: Phase 2: Phase 3: Year 1­Year 2 Year 2-5 Year 6­Year 10 Start-up phase Implementation phase Completion phase Establish Establish health insurance Identify and establish the Review and revise governance law following international governance structure for a regulations and structure for best practice, including a future health insurer, legislation on the basis health financing negative list and explicit supervisory board, and of experience gathered reforms, and criteria for benefit package hospitals in first few years support MOH definition stewardship Decide on licensing and function Develop strategy and accreditation system for all capacity-building plan for providers, requiring ensuring accountability and regular relicensing, and transparency for health exclude nonlicensed and financing (internal control, non-accredited providers accounting, reporting and from insurance contracts auditing) Define co-payment policy in health facilities Define a human resource development strategy for the health sector Purchaser­ Establish legal base for Implement reforms provider split purchaser­provider split following laws and related decrees Hospital Finalize masterplan for Implement masterplan Continue autonomy and health sector implementation of reorganization Develop strategy for masterplan Establish legal base for partnering with private hospital autonomy, and sector providers Refine standards for provide responsibility and accountability and accountability to health Decide on transition plan governance based on facility managers with benchmarks for experience from implementing previous years Conduct an institutional organizational reforms in review of health sector health sector Continue authorities to define roles implementation of and make adjustments in Grant autonomy to human resource and context of insurance hospitals in a phased management manner and provide on the restructuring Decide on governance, job guidance to hospital organizational, and managers Continue evaluating management structure of impact of co-payment hospitals and outpatient Agree on performance policy and make facilities indicators and benchmarks refinements of good governance and Hire financial management accountability in staff in each health facility autonomous health and establish standard facilities accounting systems 108 Objective Phase 1: Phase 2: Phase 3: Year 1­Year 2 Year 2-5 Year 6­Year 10 Start-up phase Implementation phase Completion phase Following human resource strategy, establish optimal management and staff mix with job description and action plan for each health facility Implement co-payment policy with directions for use of funds in provision of care Introduce single- Decide on revenue sources Continue hiring insurance Set up regional branch payer health of insurer staff following business offices responsible for finance system plan for different complaints resolution Decide legal form and departments with providers and vision of health insurance members and develop business plan Train staff in revenue and for years 1-3, with job expenditure management, Continue conducting descriptions for each claims management, annual actuarial, management position contracting and rate financial, and burden of setting, quality disease analysis Create insurance improvement and disease organization, nominate management, and provider Meet regularly with governing board, and hire performance analysis providers to present and insurance director, actuary, discuss results from and heads of contracting, performance analysis benefits, and HMIS departments Create provider performance unit within insurance fund that does actuarial, financial, and burden of disease analysis Establish basic Create benefits unit in Finalize scope of covered Develop and project cost benefit package insurance fund and hire services and compute unit for a private insurance and target delivery director costs for rate setting package that is fully to vulnerable premium financed and groups Develop basic benefit Manage high-cost diseases covers services excluded package and conduct from basic benefit actuarial cost analysis to package (such as adjust benefit package to "luxury services" in available resources Switzerland) Develop and implement transparent cost-sharing policy and prohibit informal payments Establish standard treatment protocols for high-cost 109 Objective Phase 1: Phase 2: Phase 3: Year 1­Year 2 Year 2-5 Year 6­Year 10 Start-up phase Implementation phase Completion phase diseases included in benefit package Implement Decide on outpatient and Conduct a cost and Monitor impact and provider payment hospital payment reforms efficiency analysis in all refine payment reforms reforms hospitals and PHC in hospitals and PHC Legislate reforms facilities and use results in centers rate setting and input Develop strategy and action factor adjustment plan for implementing payment reforms during Continue payment reform Phase 2 and 3 in hospitals and PHC centers Develop contracts and contract selectively with most efficient providers Implement HMIS Establish legal basis for data Install data collection in all Continuously improve and collect data collection and protection in health facilities and data collection and for performance health sector, and connectivity with insurer reporting analysis responsibility for transfer and storage Install claims, provider Train staff and member database in Decide on HMIS strategy to insurance fund, and ensure data collection and provider connectivity transfer among contracting partners (providers and Implement patient insurer); mobilize necessary registration systems in all resources for health facilities implementation Hire, train, and keep staff Install in all health facilities responsible for valid and and pharmacies a timely data collection and pharmaceutical prescription transfer system with an online real- time feedback loop Monitor and Develop and adopt M&E Follow-up surveys and Update M&E evaluate impact of framework for 5 years comparative analysis framework reforms Collect and evaluate Use results to propose Follow-up surveys and baseline data, present results corrective measures to comparative analysis and use in policy advice reform Use results to propose corrective measures to reform 110 REFERENCES Assembly of Kosovo. 2006. Law on Health Insurance, April 2006. BAG. 2006. Statistik der obligatorischen Krankenversicherung, 2006. http://www.bag.admin.ch/themen/krankenversicherung/01156/index.html?lang=de Boulton, G. and John James. 2007. Primary Health Care Restructuring. Final Inception Report. Bozicevic, Ivana, Anna Gilmore, and Stipe Oreskovic. 2004. The Tobacco Epidemic in South East Europe. HNP Policy Note. Washington, DC: The World Bank. March. Bredenkamp, C. and M. Gragnolati. 2007. Sustainability of Healthcare Financing in the Western Balkans: An Overview of Progress and Challenges. World Bank Policy Research Working Paper 4374. Buchmueller, T., and A. Couffinhal. 2004. Private Health Insurance in France. OECD Health Working Paper No. 12. Busse R. et al. 2004. "Organization and financing of social health insurance systems: current status and recent policy development." In Social health insurance systems in Western Europe, ed. R. Saltman, et al. European Observatory on Health Systems and Policy Series. Cashin, C., Samyshkin, Y, O'Dougherty, S. et al (2005): Case-Based Hospital Payment Systems: A Step-by-Step Guide for Design and Implementation in Low- and Middle-Income Countries. USAID Abt Associates Inc. Zdrav Reform Project. Chawla, Mukesh. 2007. Health care spending in the new EU member states. Controlling costs and improving quality. World Bank Working Paper Nr. 113. CMI (Chr. Michelsen Institute). 2006. Corruption in the Health Sector. U4ISSUE. Cutler, David M. and Richard J. Zeckhauser. 2000. The anatomy of health insurance. In Handbook of Health Economics 1(1): Chapter 11, 563-643, ed., A. J. Culyer and J. P. Newhouse. Elsevier. Docteur E, Oxley H. (2003): Health-Care Systems: Lessons from the Reform Experience. OECD Health Working Papers No. 9. Dudley RA, Frolich A, Robinowitz DL, Talavera JA, Broadhead P, Luft HS. Strategies to support quality-based purchasing: a review of the evidence. Rockville, MD: Agency for Healthcare Research and Quality, July 2004. (AHRQ publication no. 04-0057.) Ellis R.P., and T. McGuire. 1996. "Hospital response to prospective payment: moral hazard, selection, and practice-style effects." Journal of Health Economics 15: 257­77. Ensor, T. 2005. Formula Funding for Health Care: International Review. June 22, 2005. (Manuscript). 111 European Observatory on Health Systems. 2003. Health Care Systems in Transition: Bulgaria. Federal Solidarity Fund, Bosnia and Herzegovina. 2003. Annual Report, 2003. Figueras, J., R. Robinson, and E. Jakubowski (eds) (2005): Purchasing to improve health systems performance. European Observatory on Health Systems and Policies Series. Gaumer, G. 2007. Report on Management Accountability in Kosovo's Health Sector. The World Bank, Europe Central Asia Human Development Department. Gottret, P., and G. Scheiber G. 2006. Health Financing Revisited: A practitioner's guide. Washington, DC: World Bank. Gress, S. 2006. "Regulated competition in social health insurance: A three-country comparison." International Social Security Review 59(3/2006). Gubbels, John, David Snelbecker, and Lena Zezulin. 2007. The Kosovo Pension Reform: Achievements and Lessons. SP Discussion Paper No. 0707. Washington, DC: World Bank. April. Hsiao, W. 2007. Why is a systematic view of health financing necessary? Health Affairs 26(4): 950-961. Hurley, J. 2000. An Overview of the Normative Economics of the Health Sector. In Handbook of Health Economics, Volume 1, ed. T. Culyer and J.P. Newhouse. Amsterdam: Elsevier Science: pp. 56-118. ILO (International Labour Organization). 2006. Social Security Governance: A Practical Guide for Board Members of Social Security Institutions in Central and Eastern Europe. Budapest. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2007. Governance Matters VI: Governance Indicators for 1996-2006. World Bank Policy Research Working Paper No. 4280. July. SSRN: http://ssrn.com/abstract=999979 Korjenek, Patrice. 2006. Poland's National Health Insurance Fund: Strategy and Operations Assessment. Washington, DC: World Bank. KIPRED (Kosovo Internal Security Sector Review). 2006. Kosovo Ministry of Health and the World Bank. 2007. Kosovo National Health Accounts. Kutzin, J. 2001. "A descriptive framework for country-level analysis for health care financing arrangements." Health Policy 56 (3; June):171-204. Kutzin, J., Shishkin, S., Antosová, L., Schneider, P., and Pavel Hrobo: Reforms in the pooling of funds. European Observatory for Health Systems and Policies Series. Forthcoming. Langenbrunner, J., Orosz, E., Kutzin, J., Wiley, M. 2005: Purchasing and paying providers. In: Figueras, J. et al. eds.: Purchasing to improve health systems performance. European Observatory for Health Systems and Policies Series. 112 Langenbrunner, John C. 2005. Health Care Financing and Purchasing in ECA: An Overview of Issues and Reforms. Washington, DC: World Bank. Letter from the SRSG UNMIK to the President of the Assembly of Kosovo, June 15, 2007. Leu, R, and M. Schellhorn. 2004. The evolution of income-related inequalities in health care utilization in Switzerland over time. Diskussionsschriften. Universitaet Bern. www.vwi.unibe.ch Mastilica, Miroslav and Sanja Kusec. 2005. "Croatian health care system in transition from a user point of view." In: BMJ 331 (23 July 2005). Matheson, Alex. 2006. An Issue Paper on Governance in Kosovo. Washington, DC: The World Bank. McKee, Martin and Judith Healy (eds.). 2002. "Hospitals in a Changing Europe." European Observatory on Health Care Systems, Open University Press, Buckingham: 177-186. In Albania Health Sector Note. World Bank., 2006. MEF (Ministry of Economy and Finance). 2007. Medium-Term Expenditure Framework 2008- 2010, Provisional Institutions of Self-Government, Sector Spending Strategies. Vol. II. Sept 12, 2007. MLGA (Ministry of Local Government and Administration) Department for Supervision. 2007. "Effectiveness and Reform of Municipalities (DSERM)." Annual Municipal Statistics Report, 2005-2006. Pristina. O'Donnell, O. et al. 2007. Analyzing health equity using household survey data. The World Bank Institute. Chapter 18. OECD (Organisation for Economic Co-operation and Development). 2004. Private health insurance in OECD countries. Policy brief. http://www.oecd.org/document/10/0,3343,en_2649_37407_33913226_1_1_1_37407,00.html Palacios, R. and M. Pallares-Miralles. 2000. "International Patterns of Pension Provision." Social Protection Discussion Paper No. 0009. Washington, DC: World Bank. Preker A. et al. 2007. "The organizational structure of purchasers." In: Public Ends, Private Means: Strategic purchasing of health services. Ed. A. Preker, X. Liu, E. Velenyi, and E. Barris. World Bank. Rannan-Eliya, R. et al. 1998. Egypt National Health Accounts 1994-95. Special Report No. 3, Bethesda, MD: Partnerships for Health Reform Project, Abt Associates Inc. for the United States Agency for International Development. Reinikka Ritva, and Jakob Svensson, 2003. "Survey Techniques to Measures and Explain Corruption." World Bank. Republic of Serbia. 2006. National Health Accounts. Ministry of Health Serbia. 113 Rosenthal M.B., R.G. Frank, A.Z. Li, and A.M. Epstein. 2005. Early experience with pay-for- performance: from concept to practice." Journal of the American Medical Association (JAMA) 294:1788-93 Saltman, R. et al. 2007. Decentralization in Health Care: Strategies and Outcomes. European Observatory on Health Systems and Policies. Savedoff, William D, and Pablo Gottret (eds.). "Governing Mandatory Health Insurance: Learning from Experience." Version of March 26, 2007. Washington, DC: World Bank. Schneider, Pia. 2007. Provider Payment Reforms: Lessons from Europe and America for South Eastern Europe. Health Policy Note. Washington, DC: The World Bank. October. Schweizerisches Bundesgesetz fuer die Krankenversicherung (KVG) 1994, Stand 2005. Seiter, A. 2007. Analysis of Pharmaceutical Governance in Kosovo 2007. Ministry of Health and the World Bank. SGB 5: Paragraph 4: Krankenkassen. http://bundesrecht.juris.de/sgb_5/__4.html. Skledar, S. 2004: Institute of Macroeconomic Analysis and Development, European Industrial Observations Observatory. http://www.eiro.eurofound.eu.int/2004/12/feature/si0412302f.html Statistics Office of Kosovo. 2005. Kosovo Labor Market Statistics, 2005. http://www.ks- gov.net/ESK/esk/pdf/english/social/helth_statistics_2006.pdf Statistics Office of Kosovo. 2006. Health Statistics 2006. http://www.ks- gov.net/esk/esk/pdf/english/social/labour_mark_stat_05.pdf Streveler, Dennis. 2006. The strategic use of health management information systems at the National Health Fund. Warsaw. November. The World Bank. Tapay, N. 2001. "Private insurance, public health." OECD Observer December. http://www.oecdobserver.org/news/printpage.php/aid/566/Private_insurance,_public_health.html Transparency International. 2006. Executive Summary, Global Corruption Report, 2006. United Nations Declaration. 2000. The Millennium Development Goals in Europe and Central Asia. September. UNDP (United Nations Development Programme). 2007. Second Millennium Development Goals Report for Kosovo. USAID/UNDP "Early Warning Report" Jan.-June 2006 Van de Ven, Wynand, et al. 2007. "Risk adjustment and risk selection in Europe." Health Policy 83(2-3, October 2007): 162-179. Van Doorslaer E, Masseria C. 2004. Income-Related Inequality in the Use of Medical Care in 21 OECD Countries. OECD Health Working Papers No. 14. 114 Wagstaff, A, et al. 1999. "Equity in the finance of health care: some further international comparisons." Journal of Health Economics (18): 263-290 Wagstaff A., R. Moreno-Serra. 2007. Europe and Central Asia's great post-communist social health insurance experiment: impacts on health sector and labor market outcomes. Policy Research Working Paper #4371. Washington DC: World Bank. Widmer, W. 2007. "Moving from input-based to output-based payment: Requirements for hospital autonomy and impact on management." Presentation given at the World Bank workshop on Health Financing and Strategic Purchasing. Ljubljana, May 21, 2007. Wiley, Miriam (2007): Factors for consideration when developing a case mix based hospital payment system. The Economic and Social Research Institute. Dublin, Ireland www.esri.ie. Presentation given at World Bank GDLN session. April 11, 2007. World Bank. 1997. "The State in a Changing World." World Development Report. Washington, DC. World Bank. 2004. The Tobacco Epidemic in South-East Europe Consequences and Policy Responses. HNP Discussion Paper. Economics of Tobacco Control Paper No. 18. Washington, DC. World Bank. 2005. Kosovo Public Expenditure and Institutional Review, Volume I. Washington, DC. World Bank. 2006a. Albania Health Sector Note. Washington, DC. World Bank. 2006b. Slovak Republic: Reforming the health sector for greater fiscal sustainability. Washington, DC. World Bank. 2007. Kosovo Poverty Assessment. Report no. 39737-XK. Washington, DC. World Bank: 2008. Slovakia Public Expenditure Review. Washington, DC. World Bank and USAID. 2007. Kosovo: Selected Issues in Fiscal Decentralization. Washington, DC. WHO (World Health Organization). 2005a. Sustainable Health Financing, Universal Coverage, and Social Health Insurance, Agenda Item 13.16, Fifty-eighth World Health Assembly. Geneva. WHO (World Health Organization). 2005b. World Development Indicators. Geneva. WHO (World Health Organization). 2006a. National Health Accounts: Estonia. Geneva. WHO (World Health Organization). 2006a. National Health Accounts: Serbia. Geneva. 115 IBRD 35406 20°00' E 20°30' E To Kraljevo 21°00' E 21°30' E This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations To Kraljevo and any other information shown on this map do not imply, on the part of SERBIA The World Bank Group, any judgment on the legal status of any territory, or KOSOVO ^ any endorsement or acceptance of To Nis such boundaries. Ibar Toplica Leposavic´ ^ (Leposaviq) To Nis S 43°00' N Gazivodsko E 43°00' N Jezero ^ R B I A To Leskovac M Zvecan MONTENEGRO NTENEGRO Zubin Potok (Zvecan) Mitrovica Podujevo (Mitrovicë) ^ Batlavsko Jablanca Jezero To Ivangrad Vucitrn Istok (Vushtrri) (Istog) Srbica Lab Beli ^ To Djurakovac Ivangrad Drim Klina Obilic´ PRISTINA (Obiliq) (PRISHTINË) Pec´ Gracanicko (Peje) Glogovac Kosovo Polje Jezero ^ Klina (Kline) (Gilogoc) (Fushë Kosovë) Novo Brdo Kosovska Decani (Novo Berde) Kamenica To Leskovac (Decan) ^ Radonjicko Sitnica Lipljan (Lipjan) 42°30' N Jezero ^ 42°30' N Malisevo Ereni Gnjilane k Orahovac (Gjilani) (Rahovec) Stimlje (Shtime) ^ Morava Djakovica Binacka (Gjakove) Suva Reka Urosevac (Suhareke) (Ferizaj) Vitina ^ (Viti) ^ Gabrica Kacanik A L B A N I A Strpce (Kacaniku) Prizren (Shterpce) BeliDrim Liqeni i Lepenac Fierzës F Y R To Durrës M A C E D O N I A Gora (Dragashi) SKOPJE To Sofia MAIN ROADS 42°00' N 21°30' E RAILROADS 30° SELECTED CITIES* AUSTRIA 20° To Durrës H U N G A R Y MOLDOV NATIONAL CAPITAL SLOVENIA R O M A N I A A CROATIA PROVINCE BOUNDARY To Bitola BOSNIA AND INTERNATIONAL BOUNDARIES Adriatic HERZEGOVINA SERBIA Black B U L G A R I A Sea KOSOVO To Bitola ITALY MONTENEGRO FYR *When two city names are shown, Serbian spellings Sea MACEDONIA are shown first, followed by Albanian spellings in parentheses. 0 10 ALBANIA 20 30 Kilometers 40° When only one name is shown, the spelling is the same in 40° both languages. Tyrrhenian Aegean 0 10 20 Miles GREECE TURKEY Sea Sea 20°00' E 20°30' E 21°00' E 20° APRIL 2007