Report No. 22354 KZ Kyrgyz Republic Review of Social Policy and Expenditures (In Two Volumes) Volume 1: Overview of Health, Education and Social Protection Issues June, 2001 Human Development Sector Unit Europe and Central Asia Region Document of the World Bank Report No. 22354 KZ Kyrgyz Republic Review of Social Policy and Expenditures (In Two Volumes) Volume 1: Overview of Health, Education and Social Protection Issues June, 2001 Human Development Sector Unit Europe and Central Asia Region Document of the World Bank Kyrgyz Review of Social Expenditures June, 2001 CURRENCY EQUIVALENT (as of May 31, 2001) Currency Unit = Som US$1.00 = Som 49.1 FISCAL YEAR January 1 - December 31 ACRONYMS AND ABREVIATIONS ADB Asian Development Bank ALMP Active labor market program CDF Comprehensive Development Framework CHD City Health Department CIS Commonwealth of Independent States CPI Consumer price index CRH Central raion hospital CSAC Consolidated Structural Adjustment Credit DHS Demographic and health survey DRG Diagnosis related groups EF Employment Fund FAP Feldsher midwifery post FGP Family group practice FGPA Family Group Practice Association FSU Former Soviet Union GDP Gross domestic product HA Health association HIV Human immunodeficiency virus IDD Iodine deficiency disorders LAC Licensing and Accreditation Commission MESC Ministry of Education, Science and Culture MHIF Mandatory Health Insurance Fund MIF Medical Insurance Fund MLSP Ministry of Labor & Social Protection MOF Ministry of Finance MOH Ministry of Health NDC Notional defined contributions NHA National Health Accounts NIS Newly Independent States NSC National Statistical Committee OECD Organization for Economic Cooperation and Development OHD Oblast health department PAYG Pay as you go PF Pension Fund TB Tuberculosis SES Sanitary Epidemiological Surveillance SF Social Fund STDs Sexually transmitted diseases ii Kyrgyz Review of Social Expenditures June, 2001 SVAs Rural physician centers UMB Unified Monthly Benefit UNICEF United Nations Children's Fund VET Vocational education and training Vice President: Johannes Linn (ECAVP) Country Director: Kiyoshi Kodera (ECC08) Sector Director: Annette Dixon (ECSHD) Sector Manager: Maureen Lewis (ECSHD) Task Team Leader: Verdon Staines (ECSHD) iii Kyrgyz Review of Social Expenditures June, 2001 PREFACE In the first half of 1999, Kyrgyz authorities agreed with IDA that a review of Kyrgyz social expenditures, giving particular emphasis to health and education expenditures, would be valuable. This report presents findings from the resulting review, which an IDA team, working with Kyrgyz authorities and stakeholders, undertook between October 1999 and June 2000. The review team examined health and education expenditures particularly but also studied social protection expenditures. The review looked not only at the expenditures themselves but also at their efficiency and effectiveness as a means for achieving Kyrgyz objectives in the social policy arena. The review team first examined medium-term priorities for social policy under sharp resource constraints, then considered feasible strategies for achieving these priorities, and finally looked at their implications for policies, administration and public expenditures. The report is in two volumes. Volume 1 provides an overview of the findings on health, education, and social protection, emphasizing issues that policymakers in central ministries typically see as crucial for achieving effective patterns of social expenditure. Volume 2 offers more detailed critiques of expenditures and policies on health and education. The latter volume addresses a broader set of issues that interest sectoral policymakers. However, the two volumes are complementary. Volume 1, for example, discusses evidence on informal payments for social services and on access to education in greater depth than Volume 2. However, conclusions reported in Volume 1 often flow from analysis in Volume 2. iv Kyrgyz Review of Social Expenditures June, 2001 ACKNOWLEDGEMENTS This report has been prepared by an IDA team working with the extensive cooperation of a wide range of Kyrgyz officials and counterparts. The IDA task team consisted of Verdon Staines (task team leader), Maureen Lewis, Jan Bultman, Halil Dundar, and Dinara Djoldosheva. The team was strongly supported within IDA by Sue Berryman, Paulette Castel, Nancy Cooke, Annette Dixon, Vajeera Dobravila, Armin Fidler, Gabriel Francis, Philip S. Goldman, Kiyoshi Kodera, Joanna Maselko, Annie Milanzi, Mohinder Mudahar, Alexandra Posarac, Pedro Rodriguez, Michal Rutkowski, Kinnon Scott, and Jim Socknat. Peer reviewers were Fredrick Golladay, Jeni Klugman, and Michael Mills. The Review team would like to express its great gratitude to the large number of Kyrgyz officials and members of civil society who provided major assistance to the review team. The team would like particularly to recognize the deep support offered by Prof. Dr. Tilekbek Meimanaliev, who joined the mission team for its field visit to Naryn Oblast, only to return unexpectedly to be appointed Minister of Health. Written comments were received from the Ministries of Health and of Labor and Social Protection. Many others offered valuable contributions to the work of the review team. Joe Kutzin (health finance specialist) made major contributions to the analysis of health issues and data and was the primary author of key background papers on health. The team would like to acknowledge not only his role but also the generous support for this work from the UK Department for International Development (DFID) and Michael Borowitz. The USAID Zdrav Reform Team, led by Sheila O'Dougherty of ABT Associates, was generous in sharing its insight into the strategy and history of health reform within the Kyrgyz Republic, as well as related documentation and data. The Asian Development Bank (ADB) offered a high degree of cooperation to the work of the Review team. Among the ADB officials who met with members of the Review team were: Tom Crouch (Programs Manager); Yang Weimin (Director, Agriculture and Social Department); Hans-Juergen Springer (Deputy Director, Agriculture and Social Sectors Department); David Jay Green (Lead Economist); Manuel P. Perlas, (Senior Project Economist); J. Edgardo Campos (Senior Economist); Isabel Ortiz (Project Economist); Armin Bauer (Economist); Lan Wu, (Economist); and Cherian Alexander (Resident Representative in Bishkek) Harry Trines and Robert York from the International Monetary Fund provided helpful assistance with data and country perspectives Analysis of the Treasury health data would have been impossible without the work of Zhenya Samyshkin of the USAID-funded ZdravReform project, who converted the information provided by the Treasury into usable spreadsheets. The assistance of Zainagul Kataganova of the Kyrgyz Hospital Association also contributed greatly to the provision and understanding of the Treasury data. Work analyzing the National Health Accounts (NHA) study of health expenditures was undertaken by the Research and Analysis group (I.Z. Abdyldaev, J.T. Malabekov, A.-S.T. Meimanaliev) of the USAID- funded Zdrav Reform project. v Kyrgyz Review of Social Expenditures June, 2001 The analysis of poverty in 1998, which is based on the Kyrgyz Poverty Monitoring Survey of 1998, was the joint work of Ludovico Carraro, an external consultant, and the team of the National Statistical Committee of the Kyrgyz Republic, led by Shamsia Ibragimova, and including Alena Davydova, Rita Komandirova, Lada Nazarova and, in the first part of the project, Olga Sim. Consultants Natalia Pisareva and Lilia N. Ovtcharova undertook the analysis of privileges. vi Kyrgyz Review of Social Expenditures June, 2001 Review of Kyrgyz Social Expenditures Volume 1: Overview for Policy Makers TABLE OF CONTENTS Abbreviations and Acronyms ........................................................ ii Preface ........................................................ iv Acknowledgements ......................................................... v Executive Summary ........................................................ ix Introduction .........................................................1 Poverty in the Kyrgyz Republic ......................................................... 2 Medium-Term Priorities ......................................................... 4 Cross-Sectoral Issues ......................................................... 8 Sectoral Strategy for Health ......................................................... 15 Overview ......................................................... 15 Level, Source and Allocation of Health Spending ................................................. 16 Policy Priorities and Reform Strategies ......................................................... 21 Sectoral Strategy for Education ......................................................... 26 Overview ......................................................... 26 Fiscal Limitations ......................................................... 27 Access to Education during the Transition ......................................................... 28 Restoring the Quality of Education and Restructuring the Mix of Services .......... 36 Improving Efficiency and Ensuring Financial Sustainability ................................ 37 Sectoral Strategy for the Social Protection ......................................................... 41 Broader Context ......................................................... 41 Development of Programs and Policies ......................................................... 42 Current Structure of Programs and Policies ......................................................... 43 Action Agenda for Health ......................................................... 59 Action Agenda for Education ......................................................... 60 Action Agenda for Social Protection ......................................................... 60 References ......................................................... 62 Table 1. Economic Indicators in the Kyrgyz Republic, 1992-99 .............................. 1 Table 2. Dynamics of Main Poverty Indices (1996-1999) ........................................ 3 Table 3. Urban and Rural poverty 1996-1999 (head count) ..................................... 3 Table 4. Education and Poverty Status (1998 ......................................................... 4 Table 5. Public Spending on Health and Education, 1998 ........................................ 6 Table 6. Student/teacher Ratios for Selected Countries by Level of Education ...... 13 Table 7. Indicators of Trends in Nominal and Inflation-adjusted Expenditures on Health Services, 1995-1999 ......................................................... 17 Table 8. Recorded Health Expenditures, 1995-1999 ................................................ 18 Table 9. Distribution of Health Spending from the Government Budget, by Program (in percent) ......................................................... 20 vii Kyrgyz Review of Social Expenditures June, 2001 Table 10. Matrix of Responsibilities for the Delivery of Education services in the Kyrgyz Republic .......................................................... 26 Table 11. Public Education Expenditures, 1990- 99 ................................................. 28 Table 12. Trends in Enrollment Rates in the Kyrgyz Republic, 1990-97 ................. 29 Table 13. Reason for not Attending School Among those of School Age (ages 6 to 23 years), 1996 and 1997 .......................................................... 31 Table 14. Private Expenditures on Education in 1993 and 1996 for Children Aged 6 to 15 Years among those that Report Paying (in 1997 prices) ..... 33 Table 15. Total out of Pocket Expenditures on Education in 1997 for all Children ( soms per year) among those that Report Paying ...................... 34 Table 16. Private Expenditures on Education by Level (soms per year), 1996 ........ 34 Table 17. Average Annual Household Education Expenditures by Poverty Group (in Som and as percentage), 1997 .................................................. 34 Table 18. Basic Public Education Indicators by Level of Education, 1994-99 ....... 39 Table 19. Social Fund Income and Expenditures, 1994-2000 (% GDP) .................. 46 Table 20. Funds Allocated for Payment of Benefits, Awarded by the Ministry Labor and Social Protection and Arrears (1995-1999) (Som 'J00) ........... 53 Table 21. Number of Recipients of UMB and Average Amount of Benefit (1995-1999) .......................................................... 54 Table 22. Applications for UMB and Refusals (1998-1999) .................................... 54 Figure la. Estimated Frequency of Informal payments in Transition Countries ...... 11 Figure lb. Average Total Expenditures Per Capita for Selected ECA Countries (1995 US Dollars) .......................................................... 12 Figure 2. Distribution of Per Capita Health Spending by Oblast, 1998 .................... 21 Figure 3. Average Length of Acute Hospital Stays (days), 1996-97 ........................ 22 Figure 4. Education Attachment by Age Group, 1997 .............................................. 29 Figure 5. Enrollment Rates by Age, 1998 .......................................................... 32 Figure 6. Comparison of Current and Proposed Laws on Privileges ........................ 50 Box 1. Legal Status of Kyrgyz Informal Payments for Health Care ......................... 12 Box 2. Georgia Health Reform .......................................................... 25 Box 3. Impact of Poverty Status on Enrollment of Older Students .......................... 32 Box 4. Adapting Curricula and Teaching to the Skill Needs of a Market Economy 38 Box 5. What is a Notional Defined Contribution Pay-As-You-Go Scheme ............. 45 viii Kyrgyz Review of Social Expenditures June, 2001 REVIEW OF KYRGYZ SOCIAL EXPENDITURES EXECUTIVE SUMMARY I. Objectives. A review of Kyrgyz social expenditures on health, education, and social protection was undertaken between October 1999 and June 2000 by an IDA team working in consultation with Kyrgyz officials. Its main objectives were to: (i) Frame medium-term options for social policy that would facilitate World Bank and government initiatives over the next few years; (ii) Investigate the scale, pattern and effectiveness of public social spending and of ways to improve its efficiency, effectiveness, and impact on poor households; and (iii) Facilitate discussions about (a) policy priorities and (b) a medium-term strategy for social spending consistent with the resources likely to be available. II. Focus. Among the questions that the review addressed were: trends in social sector programs, expenditures, and outcomes; the appropriateness, effectiveness and sustainability of current programs; the efficiency and equity of resource use; and issues and links among social sectors and with other sectors. III. Medium-Term Priorities. The review identified three overarching medium-term priorities: to create conditions conducive to economic growth; to make much more effective use of resources allocated to health and education, which represented at least 8 percent of GDP and 37 percent of total public expenditures in 1998; and to maintain and extend reforms in social protection. IV. Cross-Sectoral Issues. The review also recognized five important issues that cut across the individual social sectors, and sometimes other sectors as well. (i) Education and economic growth strategies are interrelated. Realistic but far-sighted strategies for growth are therefore a prerequisite for appropriate policies on education. In particular; two questions are crucial. To what extent will the most rapidly expanding economic activities be knowledge- and skill-based products and services that require high levels of human capital? How can the country convert high educational levels into skilled services and products for which there is a market demand at home or abroad? (ii) Informal payments for health and education highlight problems with transparency and the probity of official conduct. Informal payments are large and growing, and might account for one-third to one-half of all spending on health. They mean that policy makers have lost control of ix Kyrgyz Review of Social Expenditures June, 2001 who gets what services. They also imply that poor families can lack access to these services. On a larger canvas, abuse of power destroys trust among citizens, donors and investors alike. (iii) Reforming health and education requires a start on civil service reforms. One need is for fewer staff who receive salaries that they can live on. (Addressing this will help reduce the informal payments problem.) Another need is to replace inflexible, input-oriented systems of resource management, expenditure control and performance assessment. This involves balancing financial flexibility with accountability in governance. Line-item budget controls should give way to performance indicators for outcomes achieved. (iv) Reforming health and education requires balancing equitable access with affordability and sustainability. Here, policy makers must address a number of questions. To work well for the poor, must health and education systems work well for everyone? What level of access for poor families is acceptable? What mechanisms will be used to support that access - e.g., differentiated user charges, scholarships, a limited basic health insurance benefit package, two-tier health systems? (v) Good policy must take account of linkages between health and education programs. The review team suggests that Ministry roles require realignment. In particular, the Ministry of Education, Science and Culture (MESC) now determines the number of new doctors and the number of medical faculties, but the Ministry of Health (MOH) must live with the financial consequences -higher health care costs. Although MECS did not agree with these views, the review team concluded that MOH needs to have the main role in setting the number of students per year trained as physicians, MOH needs veto power over new medical faculties, and no new medical faculties should be established in the next few years. Instead, the focus should be on improving the quality of medical education. A further linkage between these sectors could be exploited if MOH were to collaborate with other ministries and local governments on initiatives to improve health - e.g., through school-based programs. V. Main Findings on Health. Principal features of the health system include the following: * Real public sector spending on health has fallen in real terms, absorbs about 14 percent of government spending, and represents about 3 percent of GDP. Private spending on drugs and informal payments, however, might raise total health spending to 5-7 percent of GDP. * 70 percent of public health spending is for hospital care from four systems of hospitals that are substantially duplicative. Only 10 percent of public health spending is for primary health care. x Kyrgyz Review of Social Expenditures June, 2001 * Republican facilities and total health spending per capita are heavily concentrated in Bishkek. * Local governments provide two-thirds of the budgetary financing and the Republican government one-third, although the Republican share is rising. A 2-percent payroll tax supplements budgetary funding but has had a disproportionate impact on innovation within the health system. * Private payments - both formal and informal - have become significant. For poor households, they can represent a major baffier to care. Often, formal payments are in kind - food, laundry, bedding, and drugs. * The private-provider sector is small except for pharmaceuticals. * "Line-item" budgeting for inputs stifles providers' flexibility in resource allocation. VI. Stylized Facts of Health Sector. These findings can be supplemented with a more evaluative characterization of how the health sector functions in practice. The country has more health facilities than it can afford to operate. Health care is seriously skewed towards hospitals and away from primary health care. Government covers the fixed costs of running the system-mainly salaries, buildings and heat. Patients cover the marginal costs of care and most drugs from out-of-pocket payments. Poor households have less access to care, no risk-pooling and poorer health status. Rationalization of facilities could reduce fixed costs and release resources to help cover the marginal costs of care. A very sophisticated reform strategy has been prepared by MoH and its advisers. Health service delivery system and financing reforms are incorporated. Implementing the reforms might make the health system financially sustainable. However, reforms will require political support from the whole government. VII. Action Agenda for the Health Sector. The Kyrgyz health system now spends so much keeping buildings open, heated and staffed that it is starved of funds to finance actual care to patients. Staffing and utilities alone absorbed two-thirds of government health spending in 1998, whereas pharmaceuticals and supplies represented only 12 percent of the total. Accordingly, the chief priority for action is to reduce the excessive number of facilities through well-designed consolidation and closures and to use the financial resources released to fund the incremental costs of care for patients. VIII. Complementary priorities are to (i) reorient patterns of care away from overuse of hospitals and specialists and towards increased reliance on community-based primary care as well as modern approaches to specialized care; (ii) strengthen efforts to prevent disease and promote healthy lifestyles; (iii) reimpose a legal and equitable framework on the payments-both formal and informal-that patients make to providers; (iv) continue the introduction of more rational and flexible processes for allocating funds within the sector, on the national as well as on the oblast level; and (iv) continue implementing the planned reforms in the financing, regulation and delivery of health care that represent the route to a more effective and sustainable health system. xi Kyrgyz Review of Social Expenditures June, 2001 IX. Implied Health Policy Priorities. This action agenda implies that policy makers should establish spending levels, determine priorities, and move towards the required delivery system. Specifically, they should: (i) base the health budget on population size and availability of public revenues, not the number of MOH facilities or staff-for example, as a percentage of GDP or total public expenditures; (ii) align public spending with priority needs; (iii) extend Mandatory Health Insurance Fund (MHIF) payment methods to budget revenues; (iv) rationalize health infrastructure to deliver the services needed, not to support the facilities available; and (v) address future staffing levels with retraining and separation options. X. Implied Health Reform Strategies. The action agenda also implies reforming financing and regulatory systems to support the reformed delivery system. To do this, policymakers could: * Base strategy on three distinct but complementary roles of purchaser, provider and regulator. First, MiHIF could become the single health care purchaser which runs health financing (except revenue collection) for the whole population through universal health insurance based mainly on general tax revenue. It could propose policies on a benefit package, user fees, and exemptions from payment for care. It could have authority over streamlined pools of funds (one national and one per oblast). Second, providers could be allowed to be managerially more autonomous. Line-item budget controls could be relaxed. Health facility managers could be trained to manage. Third, MOH could establish the regulatory framework for health care and public policies that set the "rules of the game" for more autonomous providers. * Develop output-oriented accountability to replace input-oriented control. This would involve a shift from 'control' of sectoral inputs to monitoring of sectoral outputs. It would require the development of performance measures. MOH appears well-suited to be a 'pilot ministry' for this type of public sector reform. XI. Main Findings on Education. Despite various policy initiatives, the current system still largely resembles the inherited system, for which overall education indicators were good relative to income but varied by income and geographic area. The system is now highly decentralized and responsibilities for providing education are fragmented. The main current challenges are: * The sharp decline in the real budget - real spending per student fell 70 percent from 1990 to 1997, although public spending is stable at 4-6 percent of GDP and 20-22 percent of total government spending; * Inefficient resource use without overall sectoral rationalization - e.g., low student-teacher ratios, specialized institutions, wasted space, utility costs; * Deteriorating quality of education - e.g., poor teacher training, facility maintenance, outdated curricula and text books; xii Kyrgyz Review of Social Expenditures June, 2001 * Declining enrollment and attendance rates - especially at the preschool and upper secondary level; and * Poor linkage with the labor market - e.g., low demand for formal labor; slow response to the changing mix of required future skills; switch required from vocational programs to general education and wage differentials for educated workers are small but rising. XII. Responses to the Falling Real Budget for Education. These responses have included: (i) across-the-board cuts in education spending without establishing priorities; (ii) increasing local governments' financial responsibility for providing education; (iii) increasing local autonomy in decision-making; and (iv) diversifying revenue sources at all education levels via the introduction of formal and informal school fees and community contributions. XIII. Sectoral Restructuring is Needed. Despite significant achievements in education reform, inherited inefficiencies continue to undermine effective education reform at all levels. Without sectoral restructuring, a further deterioration of education indicators may be expected. XIV. Action Agenda For Education Sector. The implied action agenda for education has two main dimensions: * Policy makers should refocus primary and secondary education on general education (rather than specialized or vocational education) that: (i) restores high quality by using updated teaching materials and curricula; (ii) prepares people for long-term employment in a market-oriented economy (which will require a different mix of skills) by emphasizing the development of widely adaptable cognitive skills and concepts (rather than knowledge accumulation) and by training teachers accordingly; and (iii) reverses the decline in enrollment and attendance rates, especially at the upper secondary level. * To free resources for currently unmet needs within the education system, policy makers should pursue a variety of options for using resources more efficiently and diversifying sources of financing for education. XV. Main Findings on Social Protection. Kyrgyz social circumstances are bleak. The official poverty rate rose from 50 percent to over 60 percent in 1998. Available evidence indicates that poverty is positively correlated with lacking employment or being under- employed, rural location, large family size, living in a male-headed household, having a household head with less than higher education, Uzbek ethnicity, having limited livestock holdings, and engaging only in crop-growing rather than in diversified agricultural activities. It is clear that social protection programs cannot offset a sustained economic slump, which undercuts the revenue base for programs while swelling the need for assistance. Sustained economic growth is therefore central to any effective future social protection strategy. Reforms have begun, but not yet been completed, for most major social protection policies and programs. Overall, the policy agenda is to continue xiii Kyrgyz Review of Social Expenditures June, 2001 ongoing reforms and extend them where possible. The limited funds available should be allocated where they will have maximum impact. XVI. Action Agenda For Social Protection: Pensions. The review team recommends that policymakers should maintain the existing pension reform agenda. If accompanied by reasonable economic growth, that agenda will provide affordable and adequate pensions. It will also gradually but significantly lower the payroll contribution rate required to support the system. Projections undertaken when the pension reform package was adopted - prior to the Gove iment's decision in August 2000 to increase pensions by 20 percent - indicated that the reduction in payroll contribution rates could be from the prevailing rate of 40 percent today to 23.5 percent in 2010. Accordingly, policy makers should (i) postpone any significant increase in pensions until the financial performance of the Pension Fund clearly improves and shows that the additional expenditures can be sustained; (ii) refrain from introducing a voluntary, fully funded public pension pillar; (iii) act to improve collections; (iv) pass and implement the indexation law, while maintaining the present approach to benefit levels; (v) review rural pensions; (vi) initiate an organizational and administrative review of the Social Fund; (vii) adopt International Accounting Standards within the Social Fund; and (viii) monitor and evaluate the impact of the reform. XVII. Action Agenda For Social Protection: Targeted Assistance. Analyses show that targeting of assistance is far from perfect. Most extremely poor families receive no help and much help goes to non-poor families. Currently, there are several different systems that target assistance using different approaches and criteria. Those systems include privileges, energy-related price discounts with and without income-testing, social benefits, Unified Monthly Benefit (UMB), and other general price subsidies. The review team recommends that policymakers continue efforts to improve the targeting of social assistance, privileges, and energy-related assistance on those households in greatest need. To do this, they could (i) pursue the rationalization of privileges for which a draft law has been prepared; and (ii) initiate a broader review designed to develop a more coherent overall strategy for all targeted assistance and to rationalize the current programs within that framework. XVIH. Action Agenda For Social Protection: Employment Programs. The Government recently modified active labor market programs in response to a review of their effectiveness. The review team recommends that policymakers: Consider whether unemployment insurance is viable and appropriate in a context where (i) there has been a dramatic and sustained slump in economic output; (ii) it is extremely difficult, given the size of the informal sector, to verify employment status effectively and to determine whether an individual is involuntarily unemployed; (iii) a large number of people have frequently faced significant arrears in wages and benefit payments; and (iv) there are urgent potential alternative uses within the social sector for the revenues that the Unemployment Benefit program absorbs. xiv Kyrgyz Review of Social Expenditures June, 2001 REVIEW OF KYRGYZ SOCIAL EXPENDITURES VOLUME 1: POLICY OVERVIEW INTRODUCTION 1. Kyrgyz social circumstances are bleak, reflecting the country's economic difficulties. (Table 1). Output fell dramatically from 1990 to 1995 and has recovered only weakly since then. Over 60 percent of the population lived below the official poverty line in 1998, and 23 percent could not afford even the poverty-line food basket. In health, mortality from respiratory disease is the highest within the Commonwealth of Independent States (CIS); mortality from infectious and parasitic diseases climbed from 1990 to 1998; nutritional iodine-deficiency disorders affect 90 percent of population subgroups in some areas; and 45 percent of rural patients seeking inpatient hospital care sold produce or livestock to pay for it in 1996. Since 1990, educational enrollment rates have dropped from 37 to 33 percent for general secondary schooling and from 30 to 7 percent for kindergartens. It is likely that education outcomes have also dropped sharply, but no quantitative evidence to assess student learning achievements are available yet. Table 1 - Economic Indicators in the Kyrgyz Republic, 1992-2000 Indicators 1992 1993 1994 1995 1996 1997 1998 1999 2000 GDP Index (1990 = 100) 79.3 67.0 53.5 50.6 54.2 59.6 60.9 63.1 66.3 GDP real growth (%) -13.9 -15.5 -20.1 -5.4 7.1 9.9 2.1 3.7 5.0 Per capita dollar GDP (Atlas) 530 570 560 720 580 480 350 300 280 Inflation ( % dec-to-dec CPI) 2032.7 929.9 62.1 32.1 34.8 13 16.8 39.9 9.6 Poverty headcount (% of population) ... 40 ... 57 52 51 64 64 Estimated unemployment rate 0.3 1.7 4.1 5.7 7.8 5.7 5.9 7.4 7.9 Real wage index (1990 = 100) 59.4 49.6 42 43.5 44.5 49.1 55.6 50.6 51.8 Sources: Ministry of Finance, National Statistical Committee, IMF, World Bank staff calculations. 2. Not surprisingly, therefore, Kyrgyz spending on social protection, health and education is stretched to breaking point - stretched between the widely apparent human needs of a poor community and the limited economic base and tax base available to support government social programs and complementary private spending. Real public spending per student for education, for example, fell at least 70 percent between 1990 and 1997. 3. When public resources are so thinly stretched, a review of social expenditures must address two different questions. One is whether the social programs financed through these expenditures are efficient-that is, whether they achieve their intended objectives with the minimum expenditures necessary. The other, and logically prior, question is whether the programs financed are effective-that is, whether they seek to achieve the most appropriate objectives and do so using the most appropriate strategies. Although the review team met with many dedicated public employees, the review's Kyrgyz Review of Social Expenditures June, 2001 findings suggest that Kyrgyz social programs are often quite inefficient. Even more importantly, however, the findings suggest that many of the programs are ineffective and need to be significantly redesigned to improve their effectiveness. Accordingly, the focus of this report is primarily on what will be required to create far more effective social programs and secondarily on how to improve the efficiency of current programs. 4. The review of social expenditures begins with a snapshot of Kyrgyz poverty and its evolving nature. Against that background, it seeks first to propose medium-term priorities for social policy under sharp resource constraints. The review turns next to identifying important cross-sectoral issues that pursuing those priorities would raise. It then considers feasible strategies for addressing the priorities and explores the action agendas that they imply for policies, administration and public expenditures in each sub- sector. 5. Its report is in two volumes. Volume 1 provides an overview of the findings on health, education and social protection, emphasizing issues that policymakers in central ministries typically see as crucial for achieving effective, patterns of social expenditure. Volume 2 offers more detailed critiques of expenditures and policies on health and education. The latter volume addresses a broader set of issues that interest sectoral policymakers. However, the two volumcs are complementary. Volume 1, for example, discusses evidence on informal payments for social services and on access to education in greater depth than Volume 2. However, conclusions reported in Volume I often flow from analysis in Volume 2. POVERTY IN THE KYRGYZ REPUBLIC 6. The Kyrgyz Republic's population has experienced living standards that are among the lowest in the former Soviet Union. Access to much, although not all, basic social infrastructure is limited-only 45 percent of the population has access to running water. Housing and utilities represent a significant and growing proportion of household expenditure, and residential tariffs are slated to increase substantially to recoup a growing proportion of the costs of energy generation and distribution. Child malnutrition is on the rise as well. 7. In 1998, as growth in the Kyrgyz economy slumped once again, poverty increased substantially.' In 1997 about 50 percent of the population was living under the poverty line, but in 1998 more than 60 percent of the people in the country were poor (Table 2). Although poverty in 1998 was still primarily a rural phenomenon, the recession had affected production most in the sectors of industry, construction, and transport. Consequently, poverty had increased proportionally more in urban areas, reducing the gap between poverty patterns in urban and rural areas (Table 3). The changes more than wiped out growth-induced gains made in some urban areas in 1997. The poverty picture was worse in the urban areas of Naryn and Osh than in the corresponding rural areas. I Discussion of poverty in this section is based on the report, National Statistical Committee of the Kyrgyz Republic. Report on Poverty in tize Kyrgyz Republic - 1998, November 1999. The study analyzes poverty in 1998 and changes since 1996 and 1967. It is based on the Kyrgyz Poverty Monitoring Survey of 1998. 2 Kyrgyz Review of Social Expenditures June, 2001 Inequality had risen as well. Poverty appears to have changed little between 1998 and 1999. Table 2. Dynamics of Main Poverty Indices (1996-1999) 1999 1998 1997 1996 Food Poverty Food - Poverty Food Poverty Food Poverty poverty line poverty line poverty line poverty line line line line line Head-counta 23.3 64.1 22.98 63.61 14.8 51 20 51.9 Poverty gapb 0.07 0.25 5.99 24.65 4 18 6 19.1 Severity of - - 2.27 12.30 - - - - Poverty' I Source: Data for 1996, 1997 and 1998 from rounds of the Kyrgyz Poverty Monitoring Survey (KPMS). Data for 1999 from Kyrgyz Household Energy Survey (KHES) 1999. All surveys conducted by the National Statistical Committee (NSC) of the Kyrgyz Republic. Notes: (a) The head count is the proportion of families below the expenditure-based poverty line. (b) The poverty gap is the minimum additional income required on average to bring all poor families up to the poverty line, expressed as a percentage of the poverty line. (c) The severity of poverty is similar to the poverty gap, but gives greater weight to the poorest families. It is the average sq,uared proportional distance of families below the poverty line. 8. Poor households tended to be larger, with the difference in size between poor and non-poor households due mainly to the greater number of children. On average, however, poor households also had one more member who is a non-elderly, dependent adult. As in previous years, female-headed households were less likely to be poor. Ethnicity also influenced poverty. Table 3. Urban and Rural Poverty 1996-1999 (head count) Poor Extreme poor Urban Rural Total Urban Rural Total 1999 49.0 69.7 64.1 17.1 25.6 23.3 1998 50.7 71.3 63.6 18.3 25.8 23.0 1997 28.5 64.5 51.0 4.9 20.7 14.8 1996 37.1 58.9 51.9 10.3 23.3 19.1 Source: Data for 1996, 1997 and 1998 from rounds of the Kyrgyz Poverty Monitoring Survey (KPMS) carried out by the National Statistical Committee (NSC) of the Kyrgyz Republic. Data for 1999 are from NSC's Household Expenditure Survey (HES) for 1999. While focusing more on energy use, much of its data is comparable to the KMPS surveys. 9. Education was strongly linked to poverty (Table 4). Only households whose heads had higher education were under-represented among the poor. Households with 3 Kyrgyz Review of Social Expenditures June, 2001 heads who had completed secondary education, but not more, were the group most over- represented among the poor. They comprised 32 percent of the population but 47 percent of the extremely poor and 42 percent of the poor. This raises a serious question as to whether secondary education on its own is providing the combination of skills and attitudes needed to function productively in the current Kyrgyz economy. Has the additional education raised the job expectations of those completing secondary education more than their functional skills, or has it simply prepared them for a sector of the labor market and economy where, for now, the demand for labor is particularly depressed? Table 4. Education and Poverty Status (1998) Educational level Poverty status Incomplete No education Secondary Secondary High Education Extreme poor: % within group 9.1 20.4 46.7 23.8 Poverty gap 3.92 5.52 5.75 1.89 All Poor: % within group 10.4 16.6 42.4 30.6 Poverty gap 20.92 21.71 25.03 11.21 Non-poor (% within group) 7.2 14.0 21.9 56.9 Total population (% within group) 8.8 15.3 32.2 43.7 Source: National Statistical Committee of the Kyrgyz Republic. "Report on Poverty in the Kyrgyz Republic - 1998." Processed. Bishkek, November 1999. 10. Labor force participation rates fell-from 65 percent in 1997 to 60 percent in 1998. Poor members of the labor force had far higher rates of intermittent work during the year-particularly in rural areas, where much agricultural work is seasonal. Under- employment influenced the depth of poverty. 11. In rural areas, what differentiated poor and non-poor households most clearly was not primarily their ownership of land but the total value of the livestock they owned. Non-poor households, however, also owned more pasture land.2 A further differentiating factor was that poor households tend to grow only crops, whereas non-poor families appear more likely to have diversified among different agricultural activities. MEDIUM-TERM PRIORITIES 12. The review has identified three overarching priorities for social policy. These are: (i) to create conditions conducive to economic growth; (ii) to make much more effective use of resources allocated to health and education; and (iii) to maintain and extend reforms in social protection. 13. Priority One: Create Conditions Conducive to Economic Growth. The overriding social policy priority for government should be to create conditions conducive to 2 Possible additional reasons for the difference in livestock value might be (a) that families that become poor sell or consume their livestock or (b) that having more or better livestock increases families' incomes. 4 Kyrgyz Review of Social Expenditures June, 2001 economic growth. Social protection policies can help to manage and mitigate social risks. Ultimately, however, they cannot offer an alternative to successful macroeconomic performance when poverty is pervasive. In less impoverished countries, targeted social assistance measures transfer resources from relatively affluent groups to groups below the poverty line. In the Kyrgyz Republic, however, where most families are now below the poverty line or just above it, targeted social assistance measures face a much smaller group of affluent taxpayers and risk transferring resources from groups who themselves are barely above the poverty line to groups that are below it. In this setting, such programs can never substitute adequately for widespread productive employment and income generation. The action agenda implied by this priority, of course, is wide-ranging and extends well beyond the social sector itself.3 14. Priority Two: Make Much More Effective Use of Resources Allocated to Health and Education. The second most important priority should be to make much more effective use of the public resources allocated to health and education, which in 1998 together represented 8 percent of GDP and 36 percent of total government spending (Table 5). The acute scarcity of public funds in the Kyrgyz Republic argues for the allocation of public expenditures so as to achieve the best health and education outcomes possible. Many households are willing to bear part of the cost of supplying these services. The state should view its role as being to ensure that this private willingness to pay is combined effectively with public expenditures. Therefore, the state's roles in these circumstances are, first, to ensure that the needs of those who cannot afford to pay are met as fully as possible, and, second, to ensure that needs that cannot be addressed successfully by individuals acting alone are being addressed. The need for public sector intervention is especially great in areas of public health in which community intervention is required (control of communicable and vector borne diseases for example) or in which the costs of supplying a service in the short run outstrips the financial resources of most households (medical care for major illnesses and provision of education services to children who are members of large families). The appropriate test of the success of public spending is the improvement in the well-being of the collective community that the spending brings about. Applying this test requires comparing the outcome with and without government intervention rather than an assessment of the importance of a need for health or education services. The comparison of with and without situations requires the development of the "counterfactual": projections of what would have happened without government intervention. 3It involves seeking to produce a much larger flow of goods and services rather than just seeking to acquire a larger share of the existing flow of goods and services. The action agenda includes pursuing internal and external financial stability, a regional approach to reducing trade barriers, and a legal and regulatory environment that supports private sector development and overcomes existing disincentives to the creation of successful businesses. The development of robust banking and financial systems that use internationally recognized accounting and auditing standards will be essential. So will transparent regulatory and judicial processes, together with public policy-making and administration that are not distorted by conflicts between public and private interests. 5 Kyrgyz Review of Social Expenditures June, 2001 Table 5. Public Spending on Health and Education, 1998 Public Spending, 1998 As Percentage of Health Education Total Health and Education GDP 3.0 4.9 8.0 Total Public Expenditures 13.1 22.9 36.0 Note: "Health" and "Total Health and Education" public expenditures both include spending from the Mandatory Health Insurance Fund but exclude out-of-pocket payments. 15. Policies specifying which health and education services to fund publicly are still adjusting to the collapse in Kyrgyz economic output and its limited subsequent recovery. The greatly reduced revenue base implies that the public sector can no longer fund all health and education services that might be considered necessary. A corollary is that the private sector will need to fund a larger share of these services, if they are still to be provided. However, it is already clear that Kyrgyz private citizens show considerable willingness to pay for many health and education services if they would otherwise not be provided. In these circumstances, how can policy makers best structure public funding of health and education so as to take account of how private coilsjiuners wili respond to that structure? 16. Often, the structure of public funding will reflect circumstances (i) where private markets tend to fail (for example, because of information gaps and asymmetries, missing markets, "spillover" effects that one person's consumption or production choices can have for other people, and barriers to entering or leaving markets) or (ii) where poverty would lead to unacceptable inequities. For example, in the absence of efficient private financial markets that fund investments in healthy, skilled, and economically productive people- that is, in "human capital"- one could expect the costs of basic education and of crucial preventive and curative care for poor families to be largely publicly funded, with the share that is contributed privately growing over time as real incomes rise. Similarly, publicly funded scholarships for post-secondary education could assist highly talented students from poor families. Public subsidies might also encourage students who have the potential to become top performing artists or natural scientists-areas where society prizes excellence but where career openings are limited and, for each individual student, the probability of becoming a star is low. 17. Moreover, in health, policies on which services should be publicly funded for which Kyrgyz consumers, and to what extent, need to be established within the context of a broader set of health policies. All established market economies find health policy a perplexing challenge. Health care systems and their incentives must be carefully designed, managed and regulated, because markets for health care services differ from a competitive market structure. For example: * Modem health financing systems incorporate risk pooling because, in any given year, 70 percent of all expenditures are concentrated on about 10 percent of the population. 6 Kyrgyz Review of Social Expenditures June, 2001 * Providers-not consumers-often decide what care is consumed, because patients frequently delegate complex medical decisions to physicians. Once demand is not independent of supply, the incentives facing providers become crucial. * Both consumers and providers are insensitive to the cost of health services. Health insurance makes consumers insensitive, even though it is a rational policy response to high but uncertain medical costs. Medical training and professional standards make providers insensitive, by encouraging them to provide all care with positive expected medical or diagnostic benefits, regardless of its marginal cost. * Private insurance options require demanding regulatory systems if people who are poor or sick are not to be excluded from their scope. * Changes in medical technology fuel medical costs. 18. The country's plans for restructuring health policies accordingly are discussed below in Section 4. 19. Addressing this second priority will also require (i) a major rationalization and downsizing of the network of health facilities to release resources required to finance actual care for patients; (ii) a reorientation of the education system towards the new mix of skills required in a dynamic market-oriented economy, together with a rationalization of certain kinds of specialized institutions (vocational, specialized secondary, boarding, higher education, and research institutions); and (iii) addressing systemic sources of inefficient resource use in both health and education. 20. Systemic sources of inefficient resource use arise in both health and education primarily because: * As noted above, policies about which health and education services the public sector should pay for have not yet adjusted fully to the country's diminished revenue base; * Inflexible national processes of line-item budgeting and monitoring of performance, which focus on inputs rather than outputs or outcomes, limit the scope for managers of hospitals and schools to reallocate resources to where they are most needed; * Traditional staffing levels in health and education are unnecessarily high (and, in the case of health, involve the wrong mix of occupational specialties), while salaries are too small to live on; and * Policy makers have increasingly lost their influence over who gets access to what services and over the equity of access to these services, as formal and informal payments from the families of patients and students have become growing sources of income for both institutions and their staff. 7 Kyrgyz Review of Social Expenditures June, 2001 21. Priority Three: Maintain and Extend Reforms in Social Protection. A third social policy priority is to continue and extend reforms in social protection policies, so that the limited funds available for these purposes are allocated where they would have maximum impact. A variety of recent reviews have signaled appropriate directions for the reforms needed next. 22. In particular, the country has already adopted a pension reform strategy that, if accompanied by reasonable economic growth, will provide affordable and adequate pensions while gradually but significantly lowering the rate of payroll contribution rate required to support the system from 40 percent today to 23.5 percent in 2010. There is no fiscal room, however, for any further policy initiatives that would raise pension expenditures above, or reduce collections below, those currently projected. The Government should postpone any significant increase in pensions until the financial performance of the Pension Fund clearly improves and shows that the additional expenditures can be sustained. It should refrain from introducing a voluntary, fully funded public pension pillar. It should act to improve collections. In addition, the Social Fund should initiate an organizational and administrative review and should adopt Internationa Accounting Standards. 23. Efforts should also continue to improve the targeftng of social assistance, privileges, and energy-related assistance on those households in greatest need. These could include a broader review designed to develop a more coherent overall strategy for targeted assistance and to rationalize the current programs within that framework. CROSS-SECTORAL ISSUES 24. Addressing these priorities will inevitably challenge policymakers in five further ways: (i) strategies for education and economic growth are interrelated; (ii) the issue of the transparency and probity of official conduct needs to be addressed; (iii) civil service and public sector management reforms are closely related to social expenditure policies; (iv) a balance needs to be established between equitable access to services with affordability and sustainability of the overall health and education systems; and (v) important linkages between health and education programs need to be recognized. 25. Strategies for Education and Economic Growth Are Interrelated. First, reorienting the education system will require policymakers to reconsider what levels of investment in education are affordable and sustainable. That will depend, in turn, on the extent to which education will function in practice as the economy's engine of future growth. Two key issues for policy towards education are therefore: * Whether the most rapidly expanding economic activities in the medium term will largely be knowledge- and skill-based products and services that depend crucially on maintaining and upgrading the high levels of human capital embodied in the current Kyrgyz workforce; and * How policy and regulatory distortions in the economy can be removed to allow emergence of a market - both domestic and international - for goods 8 Kyrgyz Review of Social Expenditures June, 2001 and services that can be produced using the skills and knowledge base of the current Kyrgyz work force. 26. Traditional Kyrgyz economic activities for which there remains a viable market include agriculture and mining. These, however, appear to offer opportunities only for a limited rate of growth that will not take full advantage of the country's educated population and that will take many years to restore living standards to their former peak levels. (Real GDP in 1998 stood at 58 percent of its 1990 level.) There is a theoretical possibility that the country might be able to establish a foothold within the world economy in niche markets that call for skill-intensive or knowledge-intensive production. Employment in such activities would generate higher earnings levels and higher rates of income growth. Those could permit a faster decline in poverty and would offer the country a more dynamic economic future. However, it is unclear how, or in what areas, employment for such niche markets could be established. Nevertheless, maintaining and strengthening education are critical to any growth strategy. Optimal strategies for education and for economic growth are therefore interconnected. 27. Transparency and Probity of Official Conduct. Second, addressing the widespread and growing phenomenon of informal payments to doctors, teachers, and their institutions will require policymakers to tackle the broader issue of the transparency and probity with which public officials conduct their official duties. The proportion of Kyrgyz patients making informal payments for inpatient care began at 11 percent in 1993, rose to 25 percent the following year and reached 75 percent in 1996 (see Figure la).4 In 1997, the average total per capita expenditure on informal payments for the Kyrgyz Republic was $US 68 (in 1995 dollars - see Figure lb).5 These informal private payments to providers -which are often illegal (see Box 1)-might have helped to keep services flowing in public systems characterized by inadequate revenues and by low wages that have often been unpaid or in arrears. Even if they have done so, however, they represent an inappropriate public policy response to that problem. 28. In some cases, informal payments have recently been replaced or supplemented by explicit, locally developed, user-charges policies specifying formal private contributions, to be made in cash or in kind to the institution concerned, for access to health and education services. Such policies, if they are integrated within national policies on user charges for health and education that take account of both equity and efficiency objectives, could represent a constructive development. Indeed, the government has recently approved a new policy for a health benefit package that incorporates a user-charges policy for health services. National policies on user charges need to differentiate these charges according to the income level of the user, the medical Abel-Smith, B. and Falkingham, J. (1996). "Financing Health Services in Krygyzstan: the Extent of Private Payments." Report for the British Government Know How Fund (KHF). London School of Economics. 5Lewis, Maureen A. (June, 2000). Governance and Health in Eastern Europe and Central Asia: Paying for Health Care. Washington: DC: World Bank. The data for the countries shown are either reported totals or aggregations of inpatient, outpatient, drugs and other categories. Fees for diagnostic tests, specialist consultations, direct physician contributions and consumables are only reported for some countries, which limits the comparability of the data, but nevertheless provides orders of magnitude. 9 Kyrgyz Review of Social Expenditures June, 2001 necessity of the services received (for health), and the public (as distinct from private) benefits of the services. Similarly, the Government has already introduced formal cost- recovery policies in public higher education institutions and, in 1999, it introduced school fees for general education (grade 1-11) in a number of pilot raions. 29. Informal payments for services have had a variety of undesirable effects on the health and education systems and on standards of governance more generally. Informal private payments to medical staff, for example, imply an informal market for health care within an ostensibly public system of health services using public facilities. The payments circumvent the country's normal financial controls, policy rubric and audits. They are usually unreported and constitute a form of corruption that exploits monopoly power, provider discretion and lack of accountability. They undermine government efforts to improve accountability and public sector management and they have adverse implications for equity and access, governance, government priorities, and incentives faced by both health providers and managers. Poor families are disadvantaged because they are unable to pay; the role of public policy in resource allocation decisions is diminished; and government objectives become marginalized. Exploitation of public positions for private gain, which anecdotal accounts suggest is common in many areas of government, has even broader ramifications. It undermines public trust in government and it discourages not only job-creating private investments but also potenti21 international partners through fear that official decision-making and regulatory action will not be transparent, impartial, consistent with stated policies, or fair. 30. Civil Service and Public Sector Management Reforms. Third, addressing the issues of (a) too many staff within the health and education sectors (see Table 6), most of whom receive inadequate wages, (b) inflexible, input-oriented systems of resource management, expenditure control, and performance assessment, and (c) unsatisfactory budget execution (that is, large differences between budget forecasts of spending and subsequent actual spending in numerous policy areas) will require policymakers to take initiatives that are central to reforms in the civil service and in public sector management. Addressing these issues will also help with the informal-payments issue just discussed because, as one official observed candidly, no-one in the Kyrgyz Republic can live on government wages and so, to live acceptably, all government workers must have other ways to supplement their incomes. 31. For the first issue of staffing levels, employing fewer staff but paying each of them adequately is a logical response, although severance pay, assistance in obtaining loans for new businesses, retraining (to address shortages of nurses and therapists), and other interventions may be needed to assist the process. 32. The second issue - input management - can also be interpreted as finding improved ways to balance financial flexibility with accountability in governance. At a general level, the implied action agenda for policymakers is to develop new approaches to accountability. These should allow institutions more flexibility in resource use, but simultaneously use budget practices that link budget allocations to past performance 6 See Lewis (June 2000) 10 Kyrgyz Review of Social Expenditures June, 2001 outcomes. In this way, institutions can be made accountable for using resources effectively to promote good health and educational outcomes. In practice, this will entail: * Relaxing restrictions that prevent oblasts, raions and managers of health and educaton facilities from shifting funds between line items in health and education budgets; * Developing suitable measures of good performance and outcomes for hospitals, schools, and other facilities; and * Developing an accountability system that uses these measures to evaluate performance by looking at performance and outcomes achieved, not inputs used. * To achieve the desired relaxation of restrictions imposed through the current policy of line-item budgeting, policymakers should consider increasing autonomy and giving managers and service providers full, or at least greater, control over the use of resources and the organization of service delivery. Figure la. Estimated Frequency of Informal Payments in Transition Countries Armenia (1999) 91 Vietnam (1992) 1181 Azerbaijan (1995) 78 Poland (1998) 178 Kyrgyz Republic (1996) 175 Russia (1997) 74 l percent Moldova (1999) 170 Tajikistan (1999) 66 Slovakia (1999) 160 Albania (1996) 22 Bulgaria (1997) 21 Source: Lewis, Maureen A. (June, 2000). Governance and Health in Eastern Europe and Central Asia: Paying for Health Care. Washington, DC: World Bank. 11 Kyrgyz Review of Social Expenditures June, 2001 Figure lb. Average Total Expenditures Per Capita for Selected ECA Countries (1995 US Dollars) Kyrgyz Republic (1997) 168.11 Georgia (1997) 51.21 Romania (1997) 20.18 Albania (1996) 18.68 Russia (1997) 17.84 Bulgaria (1997) 14.38. Source: Lewis, Maureen A. (June, 2000). Governance and Health in Eastern and Central Asia: Paying for Health Care. Washington, DC: World Bank. Box 1. LEGAL STATUS OF KYRGYZ INFORMAL PAYMENTS FOR HEALTH CARE According to the Constitution, all Kyrgyz citizens have "a right to have equal access to health services free of charge. Chargeable health services are provided to citizens according to the Kyrgyz legislation". The Government Decree, "On out- pocket payment in health sector," was adopted in 1993. It entitles some categories of citizens (for example, invalids, veterans of the Great Patriotic War, children under age 2, students, mothers with more than 10 children and citizens with specific diseases) to get health services free of charge. Other patients should make out-of-pocket payments for specific kind of health services. The Law, "On non- budgetary activities in health sector," was adopted in 1999. Because both the decree and the law are not precise, it can be hard to determine what payments are legal and which ones are illegal. Although some informal payments in the Kyrgyz Republic are legal, others-like bribes to doctors-are illegal. Accordingly, the MOH has developed a "benefit package" - a program of state guarantees on health services - which was recently approved by the government and will be effective shortly. This program clearly identifies what kinds of services are free of charge, what kind of services should be paid partially, and what kind of services patients should pay for fully. The intention is to make the payment system in the health sector transparent so that patients will know clearly for what kind of services they should and should not pay. 12 Kyrgyz Review of Social Expenditures June, 2001 33. These changes could help to energize bureaucratic decision-making and improve the outcomes, impact and quality of their programs. Parallel action to allow more scope for facilities to raise and retain revenue and to change compensation mechanisms for staff could also help restructure organizational incentives. As in OECD countries, policymakers should accompany increased autonomy with systems that ensure high levels of accountability. Table 6. Student/teacher Ratios for Selected Countries by Level of Education Country Primary Education Secondary Vocational Higher I_Education Education Education Kazakhstan(a) 12.5 9.7 9.9 Kyrgyz Republic (a) 15 15 7 Korea 31.7 23.7 Turkey 27.9 23.7 Ireland 23.6 16.2 _ _ 13.5 USA 16.7 16.5 __14.3 OECD Average(b) 16.3 14.4 16.7 14.4 Sources: OECD. (1997). Education at a Glance: OECD Indicators 1997; Rysalieva and Gibraeva, G. (1999). Educationalfinancing and budgeting in Kyrgyzstan, IIEP. Notes: (a) The data are for general education, which includes primary (grades 1-4), lower secondary (grades 5-9) and upper secondary (grade 10-11). (b) The data for OECD countries are for the 1995-96 academic year. 34. The third issue concerns improved budget execution. Formal budget processes mean relatively little when the amounts actually spent bear little relationship to the levels approved within the budget. Instead, mid-year reviews of spending or other arbitrary decision-making processes displace formal budgets as the method by which spending priorities and levels are actually set. This undermines the formal discipline that budget processes are intended to impose on government spending. In the Kyrgyz republic, actual public spending has often been well below budgeted numbers. For example, actual public spending on education in 1998 represented only two-thirds of planned (budgeted) public spending for the country as a whole. The shortfalls were much more pronounced for local government budgets than for the Republican budget. Actual spending represented only 62 percent of budgeted spending at the local government level, compared with 89 percent at the Republican level (see table 2 in Volume 2, ). Large discrepancies between budgeted and actual public spending make budgets a poor guide to policy priorities, make budget implementation an unsatisfactory process for those charged with implementing programs, and lead to budget cuts with significantly arbitrary elements. Usually, the discrepancies arise because revenue estimates have been unrealistically high. There are implied needs for: (a) better coordination of policies toward decentralization with patterns of revenue sharing and equalization grants to the oblasts and raions, (b) more realistic approaches to revenue forecasting, (c) a sharper setting of priorities by policy makers within the budget formulation process, (d) better monitoring of revenue and expenditure flows during the implementation period, and (e) more limited reductions in allocations during the year. 13 Kyrgyz Review of Social Expenditures June, 2001 35. Balancing Equitable Access to Services with Affordability and Sustainability. To restructure health delivery and financing systems and reorient the education system, policy makers will need to consider how best to balance equitable access to services with affordability and sustainability of the overall health and education systems. For example, can health and education systems be expected to reach people who are poor unless the systems also work well for everyone else? Within such systems, what level of access to services will be considered acceptable for poor Kyrgyz families? For education policy, what does this imply for existing cost-recovery policies in general and higher education and for the roles of merit-based student financial aid, student loan programs or other approaches in expanding access for poor students? For health policy, what approaches should be used to structure access of poor families to health services? In particular, what are the relative merits of systems of differentiated formal co-payments (user contributions) and alternative approaches to what insurance arrangements cover? 36. Linking Health and Education Programs. Important relationships between health and education deserve recognition. First, Ministry roles require realignment. The Ministry of Education, Science and Culture currently determines the number of university medical faculties and the number of students enrolled to become doctors, but the Ministry of Health has to live with the financial consequences. OECD countries have come to recognize that managing the supply of new doctors is a critical element in managing both health expenditures and the balance of skilled personnel within health care systems. The Review team considers it crucial that the Ministry of Health be given the principal role in determining whether new medical faculties will be established and, as well, the number of students per year to be admitted into medical faculties. The team also considers it crucial that no new medical faculties be established at this time. Rather, the critical issue in medical education at the moment is to improve its quality, not to expand its scale. 37. The Review team recognizes, however, that the Ministry of Education, Science and Culture disagrees with the views expressed in the previous paragraph. The Ministry considers that the planning of such higher educational programs is properly a function of the Mvinistry responsible for education, that the creation of new university medical faculties would not impose costs on the Kyrgyz national budget if the programs are financed from donor grant funds, and that expanding the number of well-trained health staff should reduce health care costs through competition and improved care, rather than increasing total health care costs. 38. Second, the Ministry of Health should collaborate with other ministries and local governments on specific health-improving interventions in non-health facilities. Schools provide a convenient potential location for health education, nutrition and other programs and the education system offers an excellent environment for promoting health. 14 Kyrgyz Review of Social Expenditures June, 2001 SECTORAL STRATEGY FOR HEALTH Overview 39. The Kyrgyz Republic inherited relatively good health indicators, relative to its average level of income. However, it also inherited an inefficient health system with too much infrastructure and with outmoded approaches both to clinical care and to the prevention of non-communicable disease. Accompanying the inheritance came a decline of about 50 percent in GDP between 1990 and 1995. Subsequently, with external support, the government prepared an overall strategy for health, known as the Manas program, and initiated important and interesting reforms designed to build up and restructure the primary health care delivery system. The latter involved the development of family group practices (FGPs) reimbursed through capitation payments to primary care providers. They reforms were accompanied by the gradual creation of new systems to make information on health care and health outcomes available to policy makers and new systems to monitor and assess the quality of care. The full impact of those reforms has yet to be evaluated, but they appear to be producing important improvements in the efficiency of services. Now, the challenges are to extend them to other parts of the country, to implement complementary reforms (for example, in the hospital sector), to address the problems of fragmentation in the systems for financing and delivering health care, and to find solutions for some of the generic public sector issues (such as salary levels, overstaffing, inflexible management and use of resources, and poor incentives). 40. The Government has continued to allocate about 12-13 percent of its substantially shrunken budget to health. The share of public health spending financed by local governments is around two-thirds but falling, while the Republican government's share is around one-third but rising. In addition, a payroll tax of 2 percent flows to the Mandatory Health Insurance Fund (MHIF) and provides a minor supplement to budgetary funding. MHIF was created in 1997 to help institutionalize a distinction between the purchaser of health care and the providers of that care. It is still small but is gaining in importance and has played a major role in innovations within the system. (That is partly because MHIF allocates to facilities funds that are additional to those they receive as fixed, line-item budgets.) As noted earlier, there is also a large and growing amount of private out-of- pocket spending for health care, of which much-and perhaps most-is undocumented and not transparent. 41. Nevertheless, despite the country's useful start on needed reforms, the combination of the inherited problems in the health system and the dramatically reduced budget for health has created a fundamental crisis for health policy makers. The Kyrgyz Republic has more health facilities than it can afford to operate. Its health care system is seriously skewed towards hospitals and away from primary health care and health promotion. The country has more hospitals per capita than all other Central Asian countries and most CIS countries; and it is the only CIS country with more hospitals per capita in 1998 than in 1990. 42. To a first approximation, public expenditures cover most of the "fixed" costs of running the health system-mainly salaries, building costs, and heating. The "fixed" costs of the system are also supported by informal payments to health workers, implicit 15 Kyrgyz Review of Social Expenditures June, 2001 subsidies from public utilities to the MOH, and a lack of investment in maintaining or upgrading buildings and equipment. Patients cover most of the marginal costs of care and most drugs from out-of-pocket payments. Relative to the costs of running the extensive infrastructure of the system, public revenues provide only a low level of funding. This, combined with the fragmentation of the public sector health finance and delivery system by level of government, results in very limited risk pooling for the population. The effects of this fall most heavily on poor households, which have less access to care, are at greater risk of impoverishment from out-of-pocket payments, and have poorer health status. Rationalizing facilities could reduce fixed costs within the system and release resources to help cover the marginal costs of care. 43. Fortunately, a very sophisticated reform strategy has been prepared by the Ministry of Health (MOH) and its advisers. The strategy incorporates both delivery- system and financing reforms. The chief priority for action is to reduce the excessive number of facilities through well-designed consolidation and closures and to use the financial resources released to fund the incremental costs of care for patients. Implementing the proposed reforms might improve the financial sustainability of the health system. However, reforms will require sustained political support from the whole government as they will entail reductions in both infrastructure and.personnel. 44. This section looks next at patterns in Kyrgyz health revenues and expenditures. That leads to consideration of related policy issues: how to establish spending levels and priorities; how to achieve a delivery system that reflects these; and how to reform financing and regulatory systems accordingly. Level, Sources and Allocation of Health Spending 45. Kyrgyz health expenditures are financed from multiple sources: (i) republican and local (i.e., oblast, raion, municipal) budgets; (ii) employer contributions to the MHIF (since 1997); (iii) out-of-expenditures for official and informal charges; and (iv) loans, grants and humanitarian assistance. Between 1995 and 1999, real government spending on health fell by 5 percent (Table 7).7 In the absence of growth in funding for health through MHIF and from official user fees ("special means"), however, this fall in real government spending would have been 20 percent. Moreover, because population was rising, real public spending on health per person fell by about 10 percent. Total well-recorded public expenditures on health [that is, spending from the budget plus the NIHEF plus official user fees (or "special means")] also fell as a share of GDP--- from 4.0 percent to 2.6 percent in 1999. Private spending on drugs and on formal and informal payments to providers is not well documented, but might raise total health spending to the range from 4.5 to 6.5 percent of GDP. Private and in-kind payments can represent a major barrier to care for poor households. 7 Given the absence of special means from in the recorded expenditure figures for 1995, the true degree of this decline was slightly greater. 16 Kyrgyz Review of Social Expenditures June, 2001 Table 7. Indicators of Trends in Nominal and Inflation-adjusted Expenditures on Health Services, 1995-1999 Broad Economic Indicators 1995 1996 1997 1998 1999 Current GDP (million soms)a 16,145 23,399 30,686 34,181 48,321 Population ('QOOs)' 4,596 4,661 4,732 4,806 4,823 Revenues as a percent of GDPa 17.0 16.8 16.6 17.9 16.2 Expenditures as a percent of GDPa 28.6 22.2 21.8 21.0 18.7 Deficit as a percent of GDPa 11.5 5.4 5.2 3.0 2.5 "Public Health" component of CPI (1995 = 100)' 100.00 120.57 145.97 165.01 207.02 Percent increase in "Public Health" component of CPI 38.8 20.6 21.1 13.0 25.5 Exchange rate: soms to one US dollar (annual ave.)a 10.82 12.84 17.37 20.77 39.03 Ave. annual depreciation of som to the dollar -0.2 18.7 35.3 19.6 87.9 Current Health Expenditures ('000 soms) Total recorded spending (budget + fees + MHIF) 639,743 744,567 904,140 1,047,436 1,255,056 Budget as a percent of total recorded spendingb 100.0 96.7 92.9 87.3 85.3 Special means as percent of total recorded spendingb 3.3 6.3 8.4 6.9 MHIF as a percent of total recorded spendingc 0.0 0.0 0.7 4.3 7.9 Current Health Expenditure Indicators Per capita total recorded expenditures (dollars) $12.9 $12.4 $11.0 $10.5 $6.7 Total recorded health spending as percent of GDP 4.0 3.2 2.9 3.1 2.6 Total budget + MHIF as a percent of GDP 4.0 3.1 2.8 2.8 2.4 Gov't health (incl. fees) as % of total public spending 12.3 13.2 11.7 11.9 12.8 Real Health Expenditure Indicators (1995 soms) Health spending from government budget 639,743 597,450 575,717 554,169 516,916 Percent increase in real budget expenditures -6.6 -3.6 -3.7 -6.7 Expenditures by MHIF 0 0 4,556 27,332 47,739 Total budget + MHIF spending 639,743 597,450 580,272 581,501 564,656 Percent increase in real budget + MHIF spending -6.6 -2.9 0.2 -2.9 Total recorded spending (incl. fees) 639,743 617,553 619,408 634,777 606,255 Percent increase in real total health expenditures -3.5 0.3 2.5 -4.5 Per capita budget spending 139.2 128.2 121.7 115.3 107.2 Per capita total recorded spending (incl. fees & MHIF) 139.2. 132.5 130.9 132.1 125.7 Percent increase in real per capita expenditures -4.8 -1.2 0.9 -4.8 Per capital total recorded expenditures (dollars) $12.9 $12.2 $12.1 $12.2 $11.6 Sources: aNational Statistical Committee; bKyrgyz Government Treasury Data; cMandatory Health Insurance Fund. No data are available on expenditures from special means (official user fees) for 1995. 17 Kyrgyz Review of Social Expenditures June, 2001 Table 8. Recorded Health Expenditures, 1995-1999 Health Expenditure Indicators 1995 1996 1997 1998 1999 Total recorded health spending 639,743 744,567 904,140 1,047,43 1,255,056 (Budget + fees + MHIF) 6 Comprising: Total government budget health spending, 639,743 720,330 840,364 914,426 1,070,109 excluding official user fees ('000 soms) Local government health spending 487,295 516,782 555,712 623,337 725,052 ('000 soms) I Republican government health spending 152,448 203,548 284,652 291,089 345,058 ('000 soms) Local Budget Share (% of total) 76 72 66 68 68 Republican Budget Share (% of total) 24 28 34 32 32 Official user fee collections ("special means," in '000 soms): Total --- 24,237 57,126 87,910 86,118* Local 13,904 34,647 36,039 46,654 Republican --- _ 10,333 22,479 51,871 39,464* Fees as % of total government health 3.3 6.4 8.8 7.4* Spending Fees as % of local health spending 2.6 5.9 5.5 6.0 Fees as % of Republican health 4.8 7.3 15.1 10.3* Spending MHIF expenditures ('000 soms) 0 0 6,600 45,100 98,800 MHIF as % of total recorded spending 0.0 0.0 0.7 4.3 7.9 Sources: Data from Kyrgyz Government Treasury and Mandatory Health Insurance Fund. No data are available on expenditures from special means for 1995. The figures for 1997 exclude 95.7 million soms of counterpart funding for World Bank-supported health project, due to the absence of similar data for 1998. 46. Most government health spending comes through local budgets, although the Republican share increased from less than one-quarter to almost one-third during this short time period (Table 8). Health is a major activity at the level of local administration, commanding more than 25 percent of total local budget spending. The contribution of revenues from official user fees, while small, grew in importance between 1996 and 1998. (That trend might well have continued in 1999, but the Republican data for 1999 are not comparable with those for the two previous years. This is because two Republican institutions-the Diagnostic Center and the Medical Academy, which accounted for 44 percent of reported Republican fee income in 1998- were reclassified as "self-financing entities" and stopped reporting user-fee data to the Treasury in 1999.) Although the MHIF made only a small contribution to total expenditures in its first two years, its role grew as it contracted with an increasing number of providers. 47. The Kyrgyz health sector also benefits from development loans, grants and humanitarian aid. Preliminary findings from a National Health Accounts (NHA) study of health expenditures suggest that grants and humanitarian assistance may play a very 18 Kyrgyz Review of Social Expenditures June, 2001 important role in supporting the Kyrgyz health sector. For example, in 1998, three large projects contributed about $12 million in loan funds8. Although data are not available on a consistent basis over time, if expenditures from this source are added to the total recorded health spending figures from Table 8, they would represent about 4 percent of health spending in 1997 and about 9 percent in 1998. 48. Allocation of Health Spending. Of all public health spending, 70 percent is for hospital care. This is provided by four systems of substantially duplicative hospitals (Republican, oblast, raion, and city systems). Only about 15 percent of public spending on health is for primary health care and the public health system (Table 8). Republican facilities and total health spending per capita are heavily concentrated in the City of Bishkek (Figure 2). 49. Resource Allocation. The accumulation ("pooling") of public funds for health care and their allocation to service providers is organized according to the level of government which 'owns' the health facilities. Hence, through the end of 1999, the Republican, Oblast, raion, and municipal pools of funds overlapped and so did the catchment populations of the facilities that they fund.. Funds are allocated to providers through strict line-item budgets on the basis of norms driven primarily by the size of facilities (number of beds) and their staffing levels. These norms create incentives for each level of government to expand the physical capacity of the health system which it 'owns'. In practice, MOH, Oblast Health Departments and raions are passive intermediaries in the resource allocation process and do not exercise their legal power to change this allocation. (An exception involved reforms in 1998 to establish capitation payment of primary care providers in pilot family group practices.) 50. Facilities lack discretion to alter how line-item funds are used so as to improve quality of care or save costs. Although they may in principle consolidate partially used wards or close unneeded wards to save heat and other overheads, they cannot move money from line-item budgets around to do so. A similar constraint applies if they wish to substitute medications or outpatient care for additional days of inpatient care, to vary the mix of medical staff so as to expand skilled nursing or paramedical care, or to change from obsolete to more appropriate clinical protocols for care. Sometimes, however, the critical inhibiting constraint can be outmoded soviet clinical protocols that require hospitalizations for conditions that could probably be treated in lower-cost settings. Nevertheless, it remains the case that inflexibility frequently raises costs. 51. In contrast to the treatment of budget funds, the MIHIF operates a single pool of funds at the national level for primary and inpatient care from contracted providers. The MHIF is an active intermediary, linking its health care "purchasing" decisions to information on the population and the performance of health providers. While called an insurance fund, the MHIF has never had full financial responsibility for the costs of care for insured persons. Its main role has been to introduce new incentives and monitoring 8The World Bank-supported Health Project accounted for about half of this. The other main contributors were the Asian Development Bank's Social Sector Project and the Islamic Development Bank's support to the National Cardiology Center. 19 Kyrgyz Review of Social Expenditures June, 2001 arrangements into the health sector, while providing contracted health facilities with modest but important additional revenues that are not tied to strict govemment line items. 52. Redirecting Health Spending to Match Health Needs. If health spending is not directed towards the most important health needs, scarce resources are used ineffectively. In fact, there is a serious mismatch between Kyrgyz health needs and the way the country uses resources for health. Priority health needs center on (i) improving the availability and quality of primary curative care, (ii) addressing maternal and neonatal mortality through pre-natal, ante-natal and post-natal care, (iii) preventive interventions to promote healthy lifestyles, and (iv) targeted interventions to improve infant and child health and nutrition in poor households. Health spending, however, focuses primarily on hospital care (Table 9). Table 9. Distribution of Health Spending from the Government Budget, by Program (in percent) Program 1995 1996 11997 1998 1999 Wide profile (child and adult general) hospitals 53.3 52.3 48.6 49.1 47.4 Speciality hospitals 14.5 15.6 ' 8.2 S 19.7 20.7 Maternity hospitals 3.8 3.1 3.0 3.4 3.2 Rehabilitation hospitals 0.0 0.0 0.1 0.0 0.0 Other hospitals 0:2 0.3 0.2 0.2 0.2 Hospital Sub-total 71.7 71.3 70.2 72.5 71.5 General polyclinics & OPD physicians 7.9 8.0 7.5 7.3 8.0 Speciality polyclinics & speciality physicians 0.2 0.2 0.2 0.2 0.3 Dental polyclinics 1.2 1.3 1.1 1.1 1.1 Ambulance stations 1.0 1.1 0.9 1.0 1.1 Ambulatory Care Sub-total 10.3 10.6 9.7 9.5 10.5 Public health (SES, etc.) 7.1 6.8 6.5 5.7 6.1 Health research institutes 0.7 0.6 0.6 0.7 0.8 Administration and accounting 0.7 0.6 0.7 0.8 0.9 Central maintenance services 0.9 1.3 0.0 0.1 0.1 Capital investments 4.4 5.2 3.5 1.7 1.6 Education of health professionals 1.9 1.5 1.8 1.5 1.4 Other services not included in other categoriesa 2.4 2.1 7.0 7.5 7.1 Consolidated Budgetary Health Spending 100 100 100 100 100 Source: Kyrgyz Government Treasury data. a Includes Department of Drugs and Supplies (beginning 1997), Republican Immuno-Prophlaxis Center, Medical Information Center, and a variety of other centralized units of the MOH. 20 Kyrgyz Review of Social Expenditures June, 2001 Figure 2. Distribution of Per Capita Health Spending by Oblast, 1998 Per Capita Health Expenditure from Well-Recorded Sources, 1998 700 607 600 500 400 _ o MHIF E . *speca tnds 0 0 anpRfrmSan budget 53.______________________ Chng theBaisfo_Dtemiin theSz ndAlcton~ ofteHatbudgets 300 -_ _ __ _ _ 200 -1 190 2 128 121 l j 100- 0 DlaIa-Abad, Issyk-Kul Naryn Osh Chul Talas Blshkek Kyrgyzstan Source: Kyrgyz Treasuiry 'anid MHIF expenditure data; population data from National Statistical Committee. Policy Priorities and Reform Strategies Establishing Spending Levels, Priorities, and the Required Delivery System 53. Change the Basis for Determiining the Size and Allocation of the Health Budget. To create a more disciplined fiscal framework for health and to eliminate undesirable incentives that thwart rationalization, the size of the overall budget for the health sector should be unrelated to the number of MOH facilities or staff. So a formula should be negotiated between the MOH and the MOF to determine the total level of the health sector's budget, as is done in many OECD countries. In principle, it should be related to the size of the population. In practice, it will be related to the availability of public revenues. Therefore, the level might be fixed as a percentage of projected GDP or of total public expenditures. Within this total sectoral budget, the mix of publicly raised funds (that is, taxes and social insurance contributions) and out-of-pocket payments by users would be determined through explicit policy decisions concerning the role that is envisaged for the government within the health sector (for example, the policy on covered services and associated user charges). 54. Align Public Spending with Priority Needs. The share of health funding for primary curative care and health promotion should increase, while the share for hospitals, especially specialist facilities, should decrease. [Primary care should continue to be restructured into family group practices (FGPs), building on nationwide pilots.9] In 9 FGPs usually consist of at least 3 physicians-a "therapist" (adult internal medicine), a pediatrician, and an obstetrician/ gynecologist-together with several nurses. In rural areas, feldsher-midwifery posts should continue to be incorporated into the FGP model. To complement growth in FGPs, outpatient 21 Kyrgyz Review of Social Expenditures June, 2001 addition, M4HHIF methods for paying providers should gradually be extended to budget revenues, given their greater size and impact.'0 Because the case-based payment system employed by the MEIHIF contains an incentive for hospitals to increase the volume of their admissions, however, it would be wise for the MHIF to use methods (for example, negotiating volume limits or a budget cap for each hospital) in addition to the case-based system to ensure financial control while maintaining productivity incentives. 55. Streamline Health Service Delivery Infrastructure. There is considerable scope to reduce the fixed costs of the health system without harming population coverage or physical access to needed services. For example, one aspect of the high cost structure of Kyrgyz health care is the average length of stay in Kyrgyz hospitals, which is more than 50 percent higher than in Hungary and almost three times the levels in Sweden and the United Kingdom (Figure 3). Rationalizing hospital care and the facilities that provide it would greatly improve the effectiveness of health spending. The principle is to organize the health system to deliver the services needed by the population rather than to support the facilities owned by different levels of government. Figure 3. Avcrage Length of Acute Hospital Stays (days), 1996-97 Figure 8. Average length of acute care hospital stay (days), 1996/97 Belgium 8 Italy 8 Austria 7 Denmark 6 France 6 Sweden 5 United Kingdom 5 Conatia 12.6 Latvia 12.5 Bulgari. 12.5 Lithuania 1.8 Slnunkia 11.3 Poland 10.4 Estonia 10.3 Romania 10 Hungary 9.1 Czech Republic 9 Moldova 17.6 Azerbaijan 17.5 Ruossi 16.3 Kazakbttan 15.9 Ukraine 15.7 Kyrgyztatn 14.7 Belarns 14.6 Tujikisatn 14 0 2 4 6 8 10 12 14 16 18 20 Source: OECD health data; WHO (2000). departments in central raion hospitals should be developed as providers of referral services so as to remove an overlap with FGP specialists and to reduce the number of buildings that the system must fund. 10 MHIF pays FGPs on a capitation basis that reflects patients' enrolment choices. It reimburses hospitals according to the number and complexity of the cases they treat and the quality of care they provide. Reimbursement per inpatient is based on diagnostic and treatment data that are part of clinical records. 22 Kyrgyz Review of Social Expenditures June, 2001 56. In Bishkek, the rationalization process will be helped by putting Republican, CHD, and urban Chui Oblast hospitals and polyclinics on a territorial basis to facilitate population-based planning of services. Progress on the implementation of existing rationalization plans might begin through a merger of the management structures of several facilities that are located in close proximity to each other. However, the MOH will soon need to implement proposals for a more significant consolidation of capacity involving the closure of some facilities and the merger of others into two or three multi- specialty facilities that primarily serve the populations of Bishkek and urban Chui. A careful review is needed to identify those specialty services that are available only in Republican facilities and that should remain available to the entire population. However, this does not imply the maintenance of entire health facilities for this purpose. 57. It has been suggested that donor grant financing might be available for a new hospital in Bishkek. That could provide an important opportunity to upgrade care and de- emphasize facilities by creating a new, relatively small, teaching hospital. That investment should only proceed, however, under the conditions that other facilities are closed, total hospital running costs in Bishkek are decreased (which would facilitate the redistribution of funds), and a more multi-disciplinary setting is created (contrary to the existing situation of single-issue institutes like those for oncology, surgery, and cardiology). Such an investment could provide an incentive for the chief doctors and professionals to agree on a rationalization process and could provide a new perspective on care. 58. At the oblast level, it is necessary to close specialty hospitals and merge their services into general hospitals as departments. Where the physical layout of facilities does not allow for an immediate merger, a general hospital can be created through a managerial change. This is done in the recent MOH decree that gives specific guidance for mergers of oblast hospitals into a single "combined oblast hospital" in each of the oblasts. Each will be a single legal entity managed by a combined oblast hospital director. If this management unit is given greater authority to allocate its financial resources (see below), it would have an incentive to restructure the combined hospital's services to reduce costs-especially fixed costs-while maintaining an adequate level of care for the population it serves. 59. Address Future Staffing Levels. While some savings for reallocation within the health sector will be gained by closing facilities and redistributing revenues formerly spent on fuel and electricity, the gains will be limited if the health system is required to re-employ all of the personnel who were working in the closed facility. Thus, the government needs to be prepared to allow the health workforce to be rationalized as well. This is a very sensitive issue with the potential to do considerable political damage to the sector if it is not handled effectively. 60. A reasonable and humane approach to this would be to develop, in concert with the labor union representing health workers, a process offering counseling and retraining opportunities for those who can no longer be supported by the publicly funded health system. In addition, the government needs to be prepared to meet the costs of the termination benefits associated with the separation of former civil servants. Because of the importance of this change to the creation of a sustainable health system, such 23 Kyrgyz Review of Social Expenditures June, 2001 "severance payments" can be a critical, one-time investment to facilitate the process of change. As such, they are an especially good candidate for donor funding (or government co-financing of health projects). Reform Financing and Regulatory Systems to Support a Reformed Delivery System 61. Earlier discussion outlined most elements of the proposed Kyrgyz strategy for restructuring health financing, regulatory systems and resource allocation methods so as to support reforms in service delivery and clinical training. The functions of two elements of this strategy merit fuller comment. 62. The government should consider seriously the MOH proposal to have the oblast affiliates of MHIF manage budget money for personal health care services. MHIF would then be responsible for implementing health care financing functions (other than revenue collection) for the entire population, creating a universal health insurance system funded largely from general tax revenues. Pooling funds (from the Budget and the Health Insurance Fund) on the national level would also allow funds saved in one geographic area to be reinvested elsewhere in the health system. That is precisely what is needed to allow a gradual re.istribution of funds from Bishkek to the other Oblasts as health care facilities in Bishkek are dowNnsized and rationalized. This would implicitly involve, as well, a progressive reallocation of resources from tertiary and secondary care facilities to primary care. 63. Accompanying this should be development of coherent policies on user fees and exemptions from payment for care that define the services for which the insurance system will and will not pay. These policies on covered services and user charges could evolve over time as average real incomes within the country grow. They would serve as a principal mechanism for distinguishing operationally between what the health sector should provide and what the public sector should pay for. Estimating the Costs and Benefits of Sectoral Reform Programs 64. The discussions above present a broad strategy for sectoral reform. Although reformed sectors can deliver better services at lower costs, reform programs involve costs as well as benefits and the process of change entails higher costs in the short term.. Old buildings must be sold, or remodelled and reequipped. New buildings could be required. Some staff will no longer be needed and there will be associated retrenchment costs. Other staff must be retrained, or moved to new work locations. New management systems must be developed, and so on. Quantifying the costs and benefits of a sectoral reform program, so that these can be analysed and compared, is a major and complex task. To do so also requires that there be a detailed, specific proposal for reform. Accordingly, a benefit-cost analysis for reform of the Kyrgyz health sector has not been undertaken as part of this study. Such an analysis would be an important accompaniment to the preparation of a detailed reform plan for the Kyrgyz health sector. 65. However, to provide some sense of what is involved in undertaking a detailed analysis of health sector restructuring programs and of illustrative numbers from an 24 Kyrgyz Review of Social Expenditures June, 2001 attempt elsewhere, Box 2 presents an overview of Georgia's experience in planning such reforms. BoX 2 GEORGIA HEALTH REFORM Georgia began planning health reforms in 1991. Its health care delivery system faced serious problems: very low efficiency (for example, an average occupancy rate of 28 percent, with 1.5 physicians per patient), overspecialization, poorly maintained facilities and medical equipment, and poor coordination between different levels of care. Reform efforts focused on increasing the efficiency and quality of care, while reducing redundancies in the hospital system. To optimize the hospital network, the main goal was to decrease the hospital system's bed capacity by 45-50 percent and its staffing levels by 40-45 percent. Concurrent reforms would: implement mechanisms to monitor costs and performance; alter the ratios of different kinds of personnel; and shift responsibilities. These changes were expected to increase occupancy rates to an acceptable 75- 80 percent and to yield a decrease in the cost of hospital care by 35-40 percent. The hospitals that were to be closed would either be sold as real estate or, potentially, privatized. The first phase of the reforms involved restructuring 51 hospitals (with an initial capacity of 10,000 beds) in Tbilisi. Decisions about which hospitals to sell, and estimates of associated selling or remodeling costs and savings, were based on their potential real estate value (given their age, size, location, and potential uses), the quality of care they provided, and their efficiency, together with their current cost structure, rate of capacity utilization, and relevance to current health-service needs. However, in practice, projected sales revenues were predicted quite inaccurately. For example, one group of 5 facilities was originally projected to yield revenue of USD 10.2 million. Later estimates showed that one of the facilities could not be sold, while the others would sell for a combined value of only USD 900,000. The expected revenues from liquidating facilities were then combined with the expected costs of upgrading the remaining facilities for 5,000 targeted beds. An initial estimate was that to make improvements to Georgian standards, without providing new equipment, would cost USD 51 million. Additionally, personnel costs would need to be added for providing retraining, relocating staff who would lose their jobs as a result of a facility shutdown, and any unemployment benefits that might temporarily accrue. During this process, the government found it difficult to obtain much of the information needed for accurate predictions of the costs and savings of the restructuring project. This is partly because, historically, much of the data has been collected only sporadically, or it has been aggregated at an inappropriate level for the given task. Moreover, many of the estimates depend on future conditions, such as economic growth, that influence real estate values. Source: Kaiser Permanente International. (1998). Georgia Restructuring Master Plan for Tbilisi. Unpublished Consultancy Report for the Ministry of Health, Republic of Georgia. 25 Kyrgyz Review of Social Expenditures June, 2001 SECTORAL STRATEGY FOR EDUCATION Overview 66. At independence, the Kyrgyz Republic inherited relatively good education indicators compared to countries with similar GDP per capita. It had a relatively highly educated population for its level of per capita income, with particularly strong science and mathematics education. Enrollment rates were high at all levels of education, regardless of gender, income level or geographic location. The delivery of education services was also satisfactory, as measured by some crude input indicators (such as student-teacher ratios, an extensive network of education institutions, and the availability of basic learning materials), although it was inefficient and poorly managed. The country's network of education institutions was extensive at all levels, with a large number of highly trained teaching and non-teaching staff. 67. The Kyrgyz education system is now highly decentralized with responsibilities divided among the national, oblast, raion, and municipal levels of government. Responsibility for the provision of education is highly fragmented because of the devolution of education management to numerous ministries and state agencies (Table 7). The Ministry of 'Education; Science and Culture (MESC) has overall responsibility for policy formulation and regulation of the sector (including curriculum and textbook development, teacher training), administers national level institutions-(e.g., vocational and specialized secondary education and higher education), defines the educational standards to be offered at all education institutions, and determines the overall level of financing for education that will be provided by local budgets through centrally established norms (Table 10). Table 10. Matrix of Responsibilities for the Delivery of Education Services in the Kyrgyz Republic Type of Responsibility Education National Oblast Rayon Village Primary and Planning, Regulation, Oversight Delivery and Maintenance Secondary and Salaries maintenance and materials Vocational and Planning, financing, Delivery None None Technical regulation, and delivery I I Higher Planning, financing, Delivery None None Education regulation, and delivery Source: Adapted from ADB Social Services Delivery and Finance Project. 68. In addition, the Republican Ministries of Labor and Social Protection, Health, Agriculture, Transport, the Economy, and the Interior as well as other state committees are responsible for the remaining vocational and technical schools and higher education institutions, largely in the fields of agriculture, health or other specialties. 26 Kyrgyz Review of Social Expenditures June, 2001 69. Each oblast and raion has a department of education responsible for the overall management of preschool and general education." Local governments are responsible for the delivery of non-tertiary education services within their jurisdiction. In recent years, local governments have had greater autonomy in raising and spending additional revenues for non-salary items of the education budget. 70. Main Issues: Coverage, Quality. Efficiency. The main issue appears to be deterioration of both the coverage and quality of education at all levels because of fiscal retrenchment during the transition. In addition, the education sector has not been responsive to the new skills needs of an emerging market economy. Because education is particularly important for both poverty reduction and economic development, the government should continue to pursue policy reforms identified in the Bilim Program and supported by the ADB-financed Education Sector Reform program. Among these are streamlining the organization of education management at the oblast and raion levels by eliminating raion education departments, and introducing transparent formal school fees. These could improve efficiency in the use of resources available for education and prevent further deterioration in the coverage and quality of education services, particularly in basic education. The government has already addressed some of the problems at the level of primary and secondary education by making it the priority area for a significant share of the sector's available budget. Several essential issues and challenges need to be resolved to help improve the quality of education services and access to them. 71. Current Challenges. Like other CIS transition economies, the Kyrgyz Republic has experienced rapid deterioration of its social services (including education) during the transition from a centrally planned economy to a market economy. Because of decreasing resources and an increasing competition for limited funds, the country has been struggling to sustain previously achieved education outcomes, while reorienting its education system to the needs of a market economy and a democratic society. It now faces five major challenges in the education sector: (i) sharp declines in real budgetary resources, which in turn have had quite adverse effects on the coverage and quality of education; (ii) declines in enrollment and attendance rates according to anecdotal reports, especially in poor oblasts and raions; (iii) deterioration in the quality of education at all levels; (iv) a mix of outputs that is inappropriate for an emerging market economy; and (v) inefficiency in the use of available resources and therefore uncertainty about financial sustainability. Fiscal Limitations 72. The budget for public educatioh has fallen during the last decade, both in absolute terms and as a share of GDP, although its share of total government expenditures has returned roughly to its level in 1990 (Table 11). 73. Despite the Government's significant efforts to address the complex set of problems in the education sector over the past decade, there remains a substantially I The Bishkek City rayon education departments were consolidated in 1999 as part of the education management reform. 27 Kyrgyz Review of Social Expenditures June, 2001 incomplete reform agenda. Even under the best conditions, national income will not be restored to its pre-independence level over the medium term. This suggests continuing fiscal limitations in the education sector. Accordingly, given current economic difficulties and the future budget outlook, the government is not likely to be able to increase the current education budget and might have difficulty maintaining it. Assuming an increasingly tight public budget for education, what options and strategies could be addressed by 2005, given the current problems facing the education sector? What are the implied policy changes, administrative requirements, and spending priorities for these options and strategies? Table 11. Public Education Expenditures, 1990-99 Total educational expenditure Current education Year expenditure As % of As % of total In constant As % of total GDP government prices education expenditure (1990=100) expenditure 1991 6.1 23.0 91.1 1992 5.0 16.1 - 90.3 1993 4.2 18.5 91.1 1994 6.1 26.0 - 95.4 1995 6.6 23.1 - 96.8 1996 6.2 23.5 - 96.0 1997 5.0 22.6 - 94.8 1998 4.8 22.7 1999 Source: Statistical Yearbook of Kyrgyzstan, 1997 and 1998. Access To Education During The Transition 74. At independence, enrollment and attainment rates at all levels of education were relatively high, although levels of access were far lower than in other parts of the former Soviet Union. Enrollment and completion rates in basic education were close to universal (97 percent) for all children ages 7-15. Adult literacy was also nearly universal (96 percent in 1995). Of the population aged 15 years and older, over 80 percent had attended at least eleven years of education, and over 16 percent had completed higher education. 75. During the transition, however, the Kyrgyz Republic suffered sharp declines in enrollment and attendance rates at all levels of education except higher education. It is likely that a combination of supply-side and demand-side factors underlie these patterns (Table 12) summarizezrecent trends in enrollment rates by level of education. 28 Kyrgyz Review of Social Expenditures June, 2001 Figure 4. Education Attachment by Age Group, 1997 Highest Education Lewl Completed by Each Age Grouj, Kyrgyz 1997 65+ 55-64 45-54 35-44 25-34 15-24 0% 20% 40% 60% 80% 100% O None *Basic O Secondary OUniversity Table 12. Trends in Enrollment Rates in the Kyrgyz Republic, 1990-97 Level of Education 1990 1991 1992 1993 199 1995 1996 1997 Kindergarten (ages 1-6) 30.3 26.7 23.3 13.4 8. 7.7 8.2 7. Basic education (gross, ages 7-15) 91.8 90.6 90.3 89.7 89. 89.1 89. 89.2 General secondary (gross, ages 15-18) 36.6 36.7 36.0 32.5 28. 27.3 28.8 32.4 Vocational education (gross, ages 15-18) 12. 12. 11.7 11 9. 8., 6._ __ Technical secondary (gross, ages 15-18) 12.5 12. 11. 10.S 9.2 8. 7.d 7.' Tertiary education (gross, ages 18-22) 10.E 10. 9. 9.7 10. 11.8 12. 15. Source: UNICEF-TransMONEE Database, 1999. * Preschool education. The preschool level has experienced the sharpest declines in enrollment rates, dropping from 30.3 percent in 1990 to 7 percent in 1997. The number of preschools declined from 1,696 to 405 over the period 1990-98. It appears that the public system of pre-schooling has virtually collapsed. This is especially true in rural areas because of the massive closures of preschool establishments linked to the network of state-owned enterprises. These had previously provided preschool education for their employees' children free of charge. * Basic and secondary education. There was also a slight deterioration in the enrollment rate of basic education (grades 1-9), which is compulsory and meant to be free of charge. The gross enrollment rate for basic education dropped from 92 percent in 1990 to 89 percent in 1997. The gross enrollment rate for upper secondary education also dropped from 36.7 percent to 32.4 percent during the same period. A household survey in 1997 also showed that enrollment rates for children in compulsory grades (ages 6-15) were almost universal, regardless of poverty status and geographic area. However, 29 Kyrgyz Review of Social Expenditures June, 2001 enrollment rates drop sharply at age 16 when children complete 9 years of compulsory basic education. At this age, poor children are less likely to attend school than non-poor children. According to the survey results, only 60 percent of children aged 16 and 17 enrolled in school. The rate is about 68 percent for the non-poor students but only 53 percent for poor children. * Despite high official enrollment rates in basic education, preliminary data on attendance rates show that the number of children who are not attending school is much higher than official statistics indicate. More importantly, student attendance rates have been worsening in recent years. In analyzing non-attendance, Eversmann (1999) noted that official data do not acknowledge the phenomenon of non-attendance at school. Rather, non- attendance figures are usually underreported because of financial and prestige pressures on school administrators and teachers to maintain high official attendance figures. Based on his qualitative survey, Eversmann (1999) reported that: (1) about 2-3 percent of children of school-going age (20,000- 30,000 pupils) are not attending school, seven to ten times more than official figures suggest; (2) the daily attendance rate is around 85-90 percent; (3) about 5-10 percent of children miss 40 or more days of school per year; and (4) about a quarter of students miss 25 o01 moie days of school per year. * Vocational and Technical Education. The country has kept the extensive network of vocational education institutions it inherited without restructuring their course offerings much to reflect the demands of a new economy. While the number of schools remained about the same (i.e., 105), the number of students has declined sharply from 50,400 in 1991 to 26,800 in 1999. The share of the age cohort attending vocational education has also fallen considerably from around 12 percent in 1990 to well below 6 percent in 1996. The enrollment rate in technical education also dropped from 12.6 percent in 1990 to 7.7 percent in 1996. * Higher Education. There has been significant growth in enrollment rates of college-age students attending higher education institutions, from 10.8 percent in 1990 to about 15.2 percent in 1997. That growth offset a decline in enrollment at public institutions (from 34 thousand in 1994 to 25 thousand in 1998) with an even larger increase in enrollment at private institutions. The latter, in tum, reflects growth in the number of private institutions during the decade. Although the total number of higher education institutions increased from 12 in 1991 to 43 in 1998, 13 of these were private. Thirteen private higher education institutions enrolled about 12 percent of total higher education students. Many of the private institutions offer programs for which student demand appears to be quite high, such as business administration, economics, law, and foreign languages. While it is not clear what factors affected social demand for public and private higher education, it is likely that private institutions were able to offer programs that attracted students in areas of the labor market where there is greater demand. 30 Kyrgyz Review of Social Expenditures June, 2001 76. Increased Private Costs of Education to Households. On the demand side, one major factor contributing to changing enrollment and attendance patterns is the clear trend toward increased private costs of education for households. This cost increase took place as poverty and income inequality increased and parents had progressively fewer resources available for education and everything else. For example, the 1997 household survey found that the financial constraint was the main reason for school non-attendance (Table 13). Poverty assessments also show that education enrollment rates are closely related to the levels of family income and education - children from poorer and less well- educated families are more apt to leave school before completing basic education or at the time of its completion (see Box 3). Eversmann observes that increased poverty rates represent "the greatest obstacle to children attending school as children are more likely to be working, without clothing and school materials, neglected or ill."12 Table 13. Reason for not Attending School among those of School Age (ages 6 to 23 years), 1996 and 1997 Reason 1996 1997 Graduated 44.79 48.99 Cost 34.57 33.46 Distance 0.70 0.54 Health 2.20 0.74 Do not like school 10.62 9.86 Never in school 1.20 0.00 Other 5.91 6.41 Total 100 100 Number 998 1674 Source: Dorabawila, V. 2000. Trend and Status of Education and Private Expenditure, Kyrgyz 1993-1997 77. One can observe a clearly increasing trend in private expenditures on education, regardless of level, during the period from 1993 to 1997 (Table 14). In addition to the size of fees, the number of individuals reporting school fees also increased substantially. Out-of-pocket expenditure on school fees rose from only 4.70 soms per student to 6.60 soms per student at the primary level during the period from 1993 to 1996. Among individuals from the age group enrolled in each level, the proportion who report paying has also increased from 15 to 95 percent of those at the primary level. 12 Eversmann, 1999, p.4 31 Kyrgyz Review of Social Expenditures June, 2001 Box 3 IMPACT OF POVERTY STATUS ON ENROLLMENT OF OLDER STUDENTS Data for 1998 show that the gap in educational attainment between poor and non- poor children begins at age 16 (Figure 5). Although the overall national enrollment rate drops suddenly from over 95 percent among 11-15 year olds to only 63 percent among 16-17 year olds, the drop is much sharper among poor students. Among 16- 17 year olds, three-quarters of the non-poor were still in school, compared with half of the extremely poor. The enrollment discrepancy grows with age: whereas 37 percent of non-poor 18-22 year-olds were still in school, only 10 percent of the poor and 5 percent of the extremely poor were still studying. Figure 5. Enrollment Rates by Age, 1998 100% 900/0 - I .. ,... 800/o- || C C 1A ; . 700/ - __- i iLK-1 600/oL- ll-5 I 50%* 7 L rL i 01617 200% -]II 10o0 I I I - E)drerme Poor Poor Non Poor Total Source: Data from the Kyrgyz Poverty Monitoring Survey (KPMS) carried out by the National Statistical Committee (NSC) of the Kyrgyz Republic, plus calculations by NSC and World Bank staff. 32 Kyrgyz Review of Social Expenditures June, 2001 Table 14. Private Expenditures on Education in 1993 and 1996 for Children Aged 6 to 15 Years among those that Report Paying (in 1997 prices) 1993 1996 Fees Total' Number Fees2 Tota13 Number (%) (%) Paying Paying Primary 4.70 137.32 143 (13.9) 6.60 291.71 837 (97.6) Lower Secondary - 15.37 171.09 102 (15.4) 38.52 407.31 782 (96.5) Upper Secondary 152.00 4 (17.4) 18.84 566.45 7 (81.9) Total 9.00 151.39 249 (14.5) 22.00 348.49 1626 94.8) Note: 'Includes only fees and books. Fees include both tuition and fees in 1996, while in 1993 the question asks only for fees. 3 Includes tuition and fees, books and supplies, uniforms, sports clothes, tutors and extra classes, transport to and from school, meals for school, clubs, sports, literature, newspapers, school library fees, and expenditure through parents committees. This is net of allowances and scholarships received. Expenditures have all been adjusted to 1997 prices. Adjustment was made using the consumer price index available in the 1998 Country Assistance Strategy. These numbers have been restricted to children aged 6 to 15 years who were currently enrolled in the primary or secondary levels. There were also children in this age group that were currently not enrolled. These children have been excluded to match the available data from 1993 when calculating the averages, but their expenditures are given in some of the tables below. 78. Out-of-pocket expenditures on education continued to rise during the period from 1996 to 1997 (Table 15). The sharpest rise-more than doubling-can be observed among those that report not currently being enrolled. Among those enrolled the sharpest rise in expenditure both in percentage and in absolute terms is at the tertiary level (Table 15). The average annual expenditure per tertiary student has increased by about 1500 soms, which is a 66 percent rise. 79. As a fraction of per capita consumption, a large share of expenditure is allocated to education. This is most pronounced at the tertiary level where total out of pocket expenditure per student is close to 60 percent of household per capita consumption. Even primary and lower secondary education, however, absorb an average of 8-9 percent, which would represent a substantial proportion of available disposable income for a poor family. 80. Out-of-pocket costs of families for education peaked at the post-secondary level, as did receipts from government for scholarships (Table 16). Receipts from government, however, represented only 7 percent of out-of-pocket costs on average. Poorer families spend larger shares, but smaller absolute amounts, on education than less poor families (Table 17). 33 Kyrgyz Review of Social Expenditures June, 2001 Table 15. Total out of Pocket Expenditures on Education in 1997 for all Children (soms per year) among those that Report Paying Education Level 1996 1997 % Rise As % of PC Consumption 1997 Not enrolled currently 393.55 806.85 105.0 14.82 Primary 292.73 352.87 20.5 7.96 Lower Secondary 400.48 415.03 3.6 9.67 Upper Secondary 659.70 868.94 31.7 16.83 University 2431.58 4041.18 66.2 57.14 Missing /Not available 796.30 0 Total 537.84 813.57 51.3 14.78 Note: Includes tuition and fees, books and supplies, uniforms, sports clothes, tutors and extra classes, transport to and from school, meals for school, clubs, sports, literature, newspapers, school library fees, and expenditure through parents committees. This is net of allowances and scholarships received. All 1997 expenditures have been adjusted for regional variation in prices. Table 16. Private Expenditures on Education by Level (soms per year), 1996 Level 'f I School Parent Total out- Receipts Net.t-out-of 1 Education expenses association of-pocket from pocket I________ expenses expenses Government expenditures Not enrolled 309 20 350 0 350 Primary 218 39 256 12 244 Lower Secondary 275 46 321 9 313 Upper Secondary 387 65 452 28 423 University 739 75 814 190 624 Other/missing 305 45 351 103 247 Total 297 48 345 24 320 Note: Total out-of-pocket expenses includes tuition and fees, books and supplies, uniforms, sports clothes, tutors and extra classes, transportation, meals, and expenditures through parents committee. Net-out-of pocket expenditures are net of any allowances and scholarships received from the Government. Table 17. Average Annual Household Education Expenditures by Poverty Group (in Som and as percentage), 1997 Category of Household Type of Expenditure Extremely Poor Non- National Poor Poor Tuition 21 53 375 204 Books, uniforms, fees, tutors, 193 228 356 288 School repair, classroom supplies 36 43 64 53 Meals, transport, other 43 82 350 207 Total 292 406 1145 752 Education as share of annual consumption (%) 16.9 14.1 13.1 13.6 Source: World Bank. (1999) Update on Poverty in the Kyrgyz Republic. Report No: 19425. Washington, DC: World Bank 34 Kyrgyz Review of Social Expenditures June, 2001 81. Other Demand-Side Influences on Enrollment and Attendance. Private costs of education were not the only demand-side influence on enrollment and attendance rates. For instance, although the country has achieved overall gender balance in education, gender may have begun to affect attendance rates in recent years. The 2001 Poverty Update indicates no disparities in access or attendance by gender.'3 Consultations with the Poor reported, that "boys are being chosen, instead of girls, to attend school" and that gender disparities may be the result of the increased direct costs of education and poverty in the country.'4 Similarly, children who have special needs are affected most negatively during the transition. For children who have mental and physical disabilities, who are internal migrants or refugees, or who live in small villages, access to education is becoming harder. 82. Ethnic origin is also important in explaining enrollment in education. About 60 percent of all people living in households with ethnically Kyrgyz heads are poor, compared with about 17 percent for households with ethnically Russian heads. The School Mapping Survey shows that the Kyrgyz-speaking population is under-represented in preschool enrollment. Although approximately 65 percent of general education students are Kyrgyz-speaking, only 34 percent of preschool enrollments are Kyrgyz- speaking. The main reason for under-representation of the Kyrgyz among preschool- aged children is the lack of preschool education facilities in rural areas where the majority of Kyrgyz-speaking children live. 83. Supply-Side Influences. On the supply side, the availability, quality and relevance of educational institutions appear to play significant roles in determining enrollment rates. First, many school facilities (especially preschools and vocational schools) formerly financed and controlled by state enterprises and collective farms have been closed or transferred to local education authorities without additional financial support. Many vocational schools, for example, reduced their numbers of students as graduates were no longer hired by state enterprises. Second, the quality and relevance of these schools have deteriorated considerably. Similarly, vocational schools failed to invest to improve their quality and relevance to the needs of a market economy. The reduced quality of schooling is indicated by lack of education inputs (e.g., shortages of textbooks and other teaching materials and deteriorating facilities), ineffective processes (e.g., the poor quality of student-teacher interaction in the learning process), and deteriorating and increasingly inadequate outputs and outcomes (e.g., low student achievements). This has also reduced the demand for schools. As a result, students were less inclined to select vocational and technical education, because they have suffered considerably from an old-fashioned and increasingly irrelevant curriculum, lack of basic education materials and supplies, worn-out equipment, and insecure labor market possibilities. 84. Surprisingly, as noted above, there has been significant enrollment growth in higher education due largely to the growth of private institutions during the transition. 3 World Bank, 2001 4 World Bank (2000b). Consultation with the Poor: Uzbekistan: Washington, DC: World Bank, p.21. 35 Kyrgyz Review of Social Expenditures June, 2001 The enrollment declines in public higher education institutions can be attributed to the falling quality of public higher education as well as the institutions' failure to adapt to the changing expected skill requirements of the labor market. Private institutions, in contrast, have been able to increase their enrollment during the transition-due largely, it appears, to their focus on teaching such skills as economics, law, business administration, and foreign languages that are demanded in the labor market. 85. In summary, these developments suggest that vocational, technical and public higher education institutions have failed to respond to the needs of an emerging labor market and require substantial reforms to improve the quality and relevance of their programs. At primary and secondary education levels, there is also a concem about whether the present high levels of enrollment at lower age-levels can be maintained, given the sharp increases in the direct costs of education during the transition. Because of the close relationship between poverty and education, an increase in the existing schooling gap between poor and non-poor families could also hamper the country's poverty alleviation efforts. A challenging question, therefore, is how to prevent the development of significant inequities in access to education at a time when the Government's constrained fiscal revenues are already stretched by its efforts merely to provide existing education services. Restoring the Quality of Education and Restructuring the Mix of Services 86. Preserve and Improve Equity in Access to Education. To reverse the trend since 1990 toward declining enrollment rates, particularly for children from poor households, it is fundamental that poor children have guaranteed access to quality education, especially at the basic educational level in rural areas. If necessary to achieve that objective, the public system of basic education should be maintained fully and should be protected from cuts in education spending. This education level has the highest social rates of return to education expenditures and deterioration at this level affects the incidence of poverty very strongly. Moreover, any improvement in primary and secondary education could be achieved through reallocation of public resources from vocational education, boarding schools, specialized secondary, and higher education institutions to primary and secondary education (especially lower secondary). Nevertheless, the discussion in paragraphs 14-16 of the role of the public sector also applies to basic education. Private willingness to pay for basic education, as well as subsequent levels of education, is clearly evident among many families. In a country where public revenue is scarce, it is important to consider what structure of formal user fees (possibly linked to family income or local living standards) and what roles for private-sector provision might contribute to maximizing overall educational outcomes in terms of access, quality and efficiency of resource use. 87. Improve the Quality and Relevance of General and Vocational Technical Education. Priority should be given to the following options: (i) Continue to improve curricula and teaching to meet the new skill needs of a market economy and a democratic country (see Box 4). (ii) Introduce a new system of student monitoring and evaluation. Establishing a new system to assess and monitor students is highly important, given the 36 Kyrgyz Review of Social Expenditures June, 2001 dramatic decentralization of education management and finance over the past decade. The central government should continue to play a role in assessing and monitoring education outcomes and providing support to disadvantaged school districts. (iii) Provide basic education materials. New curricula and teaching methods must be accompanied by appropriate textbooks and instructional materials. This will require the development of textbooks and learning materials to meet the new standards. Sufficient resources should also be allocated for primary and secondary education to address the shortage of basic instructional materials and supplies. (iv) Continue to provide personnel training. Under the ADB-financed Education Sector Development Project, the country is now training existing teaching and administrative staff. It should continue to provide training programs to improve the effectiveness of teachers and education administrators in introducing new standards and methods. (v) Continue to strengthen management and policy development capacity, including management information systems. Weak management and policy development capacity contribute to poor service delivery and inefficient use of existing resources. Management roles and responsibilities at all levels of government need to be clearly defined now that the provision and management of education have been decentralized and additional burdens and expectations have been placed on local governments. (vi) Efforts are also needed to improve the management and policy development capacity of the Ministry of Education, Science and Culture (MESC), local education departments, and schools. Improving Efficiency and Ensuring Financial Sustainability 88. To improve quality, reduce inequity and realign the sector in ways just discussed will require substantial resources. As education already receives about 22 percent of all public expenditures, additional public funding for education is most unlikely unless significant growth occurs. To improve education and its efficiency in using resources, policy and spending priorities should be to maintain and improve the delivery of essential education services. These are likely to have high social rates of return and to contribute to poverty reduction efforts. Adjusting to reduced public funding would then require efficiency gains in resource use, mobilization of additional resources, and possibly an expanded role for non-government providers of some services used within the sector. (See Table 18 for indicators of current sectoral inputs and student flows.) These broad policy directions allow significant scope to improve educational coverage and quality. 37 Kyrgyz Review of Social Expenditures June, 2001 Box4 Adapting Curricula and Teaching to the Skill Needs of a Market Economy Outdated school curricula and teaching methods, which were inherited from the Soviet system, still pose concerns for educational quality. The Government has not been able to replace old textbooks and curricula, which have little relevance to the needs of a market economy. With the assistance of the ADB and several other international donors, numerous titles for lower grades have been printed, though school curricula have not been revised yet. According to the official data, only 46 of the 124 textbooks in general schools meet the requirements. The number of new teaching materials published also declined from over 80 in 1991 to less than 15 in 1997. Teaching methods are not responsive to the new skill demands of a market economy. They usually do not reflect modern teaching practices, such as student-based learning. Rather, they are rigid and authoritarian, with insufficient attention paid to the needs of individual children. They focus on learning facts, rather than acquiring skills that allow knowledge to be applied in various situations, and respond to tests with minimum critical thinking, problem solving, and decision-making skills. This shortcoming was especially important in the move away from a planned economy to a market economy that equires greater individual responsibility and flexible skills with an emphasis on I general training. Student assessment is geared primarily towards ascertaining what facts students can remember. The education system also lacks a modern student assessment and monitoring system. Attempts to reform school curricula and teaching methods need to continue to reorient the country's education system to meet the new skills needs of a market economy. In addition, the overall thrust of these reforms should be to reorient the country's monitoring and evaluation process from one concerned mainly with inputs towards a system which monitors inputs, teaching and learning processes (e.g. teaching and learning methods and their flexible application) and outcomes (what knowledge and skills are acquired by students and how these are relevant to the labor market, the requirements of the higher education system, and societal goals). The establishment of a new system of student assessment and monitoring is especially important at the present time, given the dramatic decentralization of education management and finance over the past decade. The central government should continue to play a role in assessing and monitoring education outcomes and providing support to disadvantaged school districts. 38 Kyrgyz Review of Social Expenditures June, 2001 Table 18. Basic Public Education Indicators by Level of Education, 1994-98 Indicators by Level of Education 1994 1995 1996 1997 1998 Preschool Schools 639 448 449 405 405 Students 59,000 46,000 47,000 47,000 46,000 Students per school 122 111 111 112 113 Number of teaching staff . 5,260 4,522 4,490 4,49( Number of non-teaching staff 5,895 5,432 5,275 5,26( Student/teacher ratios 8 E 9 9 1( Primary and General Secondary Schools 1,81( 1,819 1,830 1,839 1,84 Students 943,90 946,555 972,489 1,012,773 1,042,716 Classes 40,087, 39,834 39,156 39,901 40,262 Students per school 520 520 531 551 567 Students per class 24 24 25 25 26 Number of teaching staff . 70,904 69,863 70,517 71,062 Number of non-teaching staff . 35,065 32,719 32,348 32,302 Student/teacher ratios . 13 14 14 15 Vocational and Technical Schools 105 105 105 105 105 Students 36,125 30,109 21,959 22,447 21,644 Classes 1,308 1,070 954 930 721 Students per school 344 287 209 214 206 Students per class 28 28 23 24 30 Number of teaching staff 2,347 2,121 1,780 1,303 1,472 Number of non-teaching staff 6,641 4,372 3,921 3,581 3,592 Student/teacher ratios 15 14 12 17 15 Specialized Secondary Schools 46 46 46 44 44 Students 21,889 18,343 16,823 16,854 15,60 Classes 783 755 721 660 595 Students per school 476 399 366 383 355 Students per class 28 24 23 26 26 Number of teaching staff 3,835 3,716 3,188 3,085 2,876 Number of non-teaching staff 3,667 3,263 3,176 2,374 2,374 Student/teacher ratios 6 5 5 6 5 Higher Education Institutions 1 8 19 21 21 21 Students 34,430 31,919 29,514 27,695 25,269 Students per school 1,913 1,680 1,405 1,319 1,203 Number of teaching staff 4,832 4,657 3,859 3,673 3,397 Number of non-teaching staff 6,861 6,682 5,972 5,934 5,580 Student/teacher ratios 7 7 8 8E 7 Source. Ministry of Education, Science and Culture (1999); Wicks (1999). 39 Kyrgyz Review of Social Expenditures June, 2001 89. Scope for More Efficient Resource Use. Despite recent progress, there are still numerous ways to use resources more efficiently in the education sector. Examples include: * Reducing the number of teaching and support staff. Overall, student- teacher and support staff ratios appear to be much lower than the average at OECD countries and in many developing countries. This suggests that there is potential to reduce teaching and support staff at all levels of education even if levels as low as the OECD average are set as a norm.; * Rationalizing existing physical excess capacity through consolidation or merging of schools and programs with low student demand and high costs, such as vocational schools, boarding schools, and higher education institutions that overlap in their functions or are under-developed or have declining enrollments; * Improving budgeting and the resource allocation process through the introduction of a more flexible and decentralized school finance policy as well as a system of accountability for the use of resources; * Better targeting of the resources available for education (such as student stipends); and * Reviewing research centers affiliated with universities or various state ministries and agencies to consider the scope for reducing the number of ineffective institutions or programs. These are costly and should be reduced to free resources for pressing priorities. 90. Reducing Excessive Reliance on State Funding. Alternative funding sources are potentially available, with some already in place. Options include: * Improving cost-recovery efforts through user charges, including (i) introducing formal tuition fees for all students in secondary specialized and higher education; (ii) charging students for boarding, meals, educational materials, and student activities; and (iii) replacing the current student stipend program with an income-tested scholarship program in the short-term and, perhaps, a student loan program in the medium-term after a sound financial and banking sector is established and after auditing and financial management processes are overhauled in both the public and private sectors. Such programs could be available for students coming from both public and private institutions; * Charging students for boarding, meals, educational materials, and student activities; * Encouraging income generation from goods and services, including (i) teaching services to non-students (e.g., extension or continuing courses for working adult students); (ii) sale of education and research related 40 Kyrgyz Review of Social Expenditures June, 2001 products (e.g., books, inventions, farm products); (iii) sale of consultancy services; and (iv) privatizing various non-education related services (e.g., dormitories, cafeterias, health care, etc.); * Increasing formal contributions of parental and community resources; and * Providing a more supportive an enabling environment for the development of the private sector in education at all levels. 91. The policy options outlined could potentially achieve significant cost savings, improvement in the quality of education, and reduced inequities in access to education. Some of these options, however, will require both significant time and substantial initial costs before yielding any significant benefits. Submitting specific reform proposals to the discipline of a systematic analysis of benefits and costs would therefore be valuable. SECTORAL STRATEGY FOR THE SOCIAL PROTECTION Broader Context 92. Families face a variety of risks to their economic well-being. They include temporary downturns in macroeconomic performance as well as family-specific risks such as the death, disablement, loss of employment, illness, or retirement of a family member. Families also face a more general risk of experiencing inadequate income. 93. Social protection programs of three kinds can help families and individuals to reduce, mitigate, or cope with these risks. Labor market policies and programs seek to improve labor market performance and expand job opportunities. Provided they are financially sustainable, insurance arrangements (such as pension programs) enable people to transfer income from high-income to low-income periods of their lives; and they can be managed by either the public or private sectors. Social assistance programs transfer income from better-off to worse-off households. They can be thought of as programs for compulsory sharing that tax higher income groups and transfer these resources to lower income groups that meet defined eligibility conditions. 94. Social assistance programs, however, are not satisfactory programs for dealing with severe and sustained economic contractions, particularly in countries that had relatively low per capita incomes before the downturn. That is primarily because the strategy of transferring income between groups of people does not work well unless (a) a substantial proportion of the population has high, or at least adequate, incomes and (b) a much smaller proportion of the population is below the poverty line and receiving benefits. In countries like the Kyrgyz Republic, where over 60 percent of the population are officially poor and many of the remainder have incomes close to the poverty line, programs of this kind cannot provide adequate income support to all who are poor. 95. For this reason, economic growth must be central to any future Kyrgyz strategy for social protection. There is currently a limited supply of goods and services available to Kyrgyz citizens and growth is the only way to increase that supply. While individuals 41 Kyrgyz Review of Social Expenditures June, 2001 might succeed in improving their position by obtaining a larger share of the existing flow of goods and services, that strategy cannot work as a national policy for the country as a whole. It is a zero-sum game in which one person's gain is another person's loss. For a national policy, the only solution is to increase the flow of goods and services. The country requires a larger economic pie, not a new approach to slicing the existing economic pie. Development of Programs and Policies 96. Like expenditures on health and education, social protection expenditures have been forced to contract sharply since independence by the large reductions both in economic activity and in the country's tax revenues. In 1991, social expenditures on pensions, allowances for families with children, education and health comprised about 55 percent of on-budget government expenditures at all levels of government combined, and close to 65 percent if the budget of the Pension Fund is consolidated with those of the republican and local governments. Those expenditures represented 25 percent of GDP"5. However, substantial restructuring of these programs later became inevitable. 97. The pension system that was inherited from the former Soviet Union remained largely unchanged through 1996. The initial efforts of the Government focused on organizational and administrative reforms of the system rather than fundamental shifts in policy approaches. After independence, the Government created the Social Fund (SF) as a benefit collection and payment agency under the general supervision of the Ministry of Labor and Social Protection (MoLSP). The Social Fund now administers the Pension Fund (PF), the Social Insurance Fund (SIF), the Employment Fund (EF) and the Medical Insurance Fund (MIF). Policy reforms followed later. 98. Allowances for families with children have also given way to new social assistance policies. The inherited system was both expensive and very poorly targeted. In 1995, a new targeted social assistance benefit, the 'Monthly Common Subsidy,' was introduced to compensate the poor for the phasing out of certain consumer subsidies. In May 1998, a "Law on State Benefits" established a new Unified Monthly Benefit (UMB) as the basis for a simplified and unified cash social assistance policy. UMB is designed to guarantee a minimum income under means testing, and is to be adjusted annually in line with an assessment of needs and available fiscal resources. As discussed below, however, serious administrative problems are experienced in targeting social assistance. 99. Kyrgyz employment policies-as these are understood today-are a creation of the decade since independence. After the collapse of the underlying central planning system, people began to face uncertainty about income and job tenure for the first time. The country was badly hit by unemployment, especially disguised unemployment, which rose sharply and contributed to the substantial growth in poverty. Initial responses to weakness in the labor market followed OECD models, which might have had limited applicability to Kyrgyz circumstances. The employment service was reformed, unemployment benefits were introduced, and the vocational training system was 5 The World Bank (1993). Kyrgyzstan: The Transition to a Market Economy. Washington D.C: The World Bank, p.103. 42 Kyrgyz Review of Social Expenditures June, 2001 upgraded to help it anticipate demand for new skills. When the country became increasingly unable to maintain these new systems as the economy slipped further, the government introduced further active labor market programs, especially training and the development of small businesses.'6 Current Structure of Programs and Policies Social Insurance Pensions 100. The Kyrgyz pension system has been very broad in scope, with benefits that were over 60 percent of the average wage in 1998 and with a variety of categorical privileges for early retirement or earnings calculations inherited from the prior system. The system's financial problems have been compounded by a drop in formal sector employment which, combined with the low retirement age, led to a drop in the ratio of contributing workers to pensioners. Payroll contributions to the Pension Fund are 32 percent of gross wages (24.7 percent of labor costs), which depresses employment levels and strengthens the incentive to evade contributions. The above factors have contributed to high expenditure levels, particularly notable for a country with the Kyrgyz Republic's young demographic profile. 101. In 1996, the Kyrgyz Republic took the first step toward pension reform when it implemented record-keeping for individual contributions and eliminated many early retirement provisions. The "Law on State Social Insurance," which became effective in June 1997, introduced a fundamental reform of the state pension system by introducing a notional defined contribution scheme (see Box 5). Pensions are progressively linked to workers' complete working history and retirement age. The reform was accompanied by a generous recalculation of all existing pensions and no increase in the retirement age. The system, however, was not sustainable in the short-term. So the retirement age is finally increasing gradually to 58 for women and 63 for men in 2007. There are no plans for a funded system. 102. Acquired rights in the old system were redefined and scaled down. Each year of service accrued before 1996 gives the right to a replacement rate of 1 percent of worker's average wage. The latter is calculated on workers' five best consecutive years before 1996, although at 1997 prices. Workers contributions in the new system earn an interest rate equivalent to 75 percent of the growth rate of the average wage. The new system incorporates a minimum pension per year of service that is added to each workers' earning-related benefits. The overall benefit should converge to an average of 35 to 40 percent of individkj earnings.. 103. The minimum pension per year of service should be an important tool to prevent pensioners-especially very low-income earners like farmers-from falling below the 16 Abrahart, Alan (1997). Kyrgyz Republic Employment Policies: Facing East. Paper prepared for the Ministry of Labor and Social Protection Seminar, "A Strategy for Social Policy of the Kyrgyz Republic as a Framework for the "SOSAC' Project." Bishkek, September 24-26, 1997. 43 Kyrgyz Review of Social Expenditures June, 2001 poverty line without excessively distorting the system's incentives to contribute. A minimum of five years of service is required to qualify for this addition. (Pension contributions for farmers are based on a tax on land rather than on earnings; and it is likely that most farmers will qualify only for the minimum pension.) 104. While the reform is a step forward, there are still a number of outstanding issues. Indexation of current pensions is on an ad hoc basis, and a draft law is before Parliament that would put indexation on a more transparent and sustainable footing. Disability pensions grew rapidly through 1998 and, although they stabilized in 1999, they still represent a worrisome trend in Pension Fund expenditure. Early retirement privileges have been reduced, but there are still important provisions for certain groups. 105. Fiscal position. The Social Fund deficit, financed by the state budget, widened substantially from 0.8 percent of GDP in 1994 to 1.3 percent of GDP in 1996 (Table 19). This was mostly due to a policy that protected pension benefits against inflation and provided excessively generous increases in the minimum pension. Moreover, there was an increase in. the number of early retirements (in turn generated by the downsizing of the public sector and the closing of loss-making enterprises). This increase offset the expected decrease in the stock of pensioners resulting from the smaller size of the retiring cohorts born during WWII. During 1998, the Government designed a package of measures to address the Social Fund deficit, which has reduced the overall deficit of the Social Fund to 0.4 percent of GDP in 1999 and to 0.3 percent of GDP in 2000. The overall sustainability of the Social Fund and the possibility of future surpluses depend crucially on the strength of economic recovery and the capacity of the authorities to encourage the formalization of new businesses so that these contribute payroll taxes. 106. Future trends. Projections made when the pension reform package was adopted indicated that the new system would produce a surplus in a few years unless collection further deteriorated. The stock of pensioners will decrease until 2007 as a result of the retirement age increases. In 2010, each pensioner will be supported by approximately 4.3 workers instead of 3.2 in 1997. The reduction in the average replacement rate will also help to keep Pension Fund expenditures under control, reducing in turn the tax burden of the system. The projections showed that, in 2010, a contribution rate of 23.5 percent - instead of 40 percent as then applied-should be enough to balance the system. However, those projections were undertaken before the Government increased pensions by 20 percent from August 1, 2000. Accordingly, the projections need to be updated. 107. System Organization and Administration. Models of social insurance organization vary throughout the world. The aim is often to have the ministry set policy, with the social insurance agency administering the policy in the most cost-effective way. The current Kyrgyz system places the Social Fund under a Supervisory Council (Nabliudatel'nyi Sovet). 44 Kyrgyz Review of Social Expenditures June, 2001 Box 5 What is a Notional Derined-Contribution Pay-As-You-Go Scheme? Pension schemes, which provide income for people if they retire or become disabled, can be financed and structured in various ways. * "Pay-as-you-go" (PAYG) pension systems, which can be established quickly, exploit the fact that the state continues to exist as a legal entity, even though individual people eventually die. These pension systems use contributions from the current generation of contributing workers (or their employers) to finance pension payments to the current generation of eligible retirees. When the current workers retire, their pensions will be financed from contributions of the next generation of workers (or their employers). * "Funded" pension systems work differently. Instead, they invest the contributions of each worker (or the worker's employer) in income-earning securities that are backed by income-generating assets and managed through a large investment fund. Each covered worker has a personal account that grows with investment earnings over the person's working life. When people become eligible for a pension, their share of the investment fund's assets is used to finance their prornised pension benefits. > In "defined-benefit" funded pension systems, the employer guarantees to pre- fund sufficient assets to finance a specified stream of pension payments. > In "defined-contribution" funded pension systems, the size of pension payments is set at whatever levei can be funded from the accumulated value of contributions to the worker's account by the time of retirement. * A "notional defined contribution" (NDC) pension system structures benefits just like a defined-contribution funded pension system, basing them on the accumulated value of the worker's individual account. However, it funds those benefits from PAYG contributions. Thus, an NDC scheme records each covered worker's contributions in a notional account that returns interest until retirement and it mimics a defined- contribution funded scheme by converting the accumulated value in the worker's account into a stream of "annuity" payments for the person's remaining lifetime. The new Kyrgyz state pension system phases out the old earnings-related pension, replacing it with an NDC system of personal accounts. The notional defined contributions made on behalf of workers after 1996 are registered in the individual account record, and are used to finance current system liabilities (in line with the PAYG approach). Once the new system is fully phased in, the individual's total contributions, which at present are compounded by average wage growth, will be multiplied by a coefficient based on the total expected post- retirement life span to calculate the monthly pension benefit. The benefit is structured like an annuity, but is financed by current PAYG revenues rather than by assets such as securities. The NDC scheme has three main differences from a fully funded scheme. First, workers' contributions are not invested, but only recorded, as they are used to pay current pensioner benefits. Second, the return on contributions is not directly linked to the results on the financial market but is predetermined. The interest rate is more often related to the growth of the average wage of the economy or the contributors. Finally, after retirement, a worker's notional capital does not earn interest. Consequently, benefits are calculated by dividing the amount accumulated in the notional account by the average number of months that workers (with no gender differentiation) are expected to live at each specific age of retirement. NDC pension systems, like defined-contribution funded systems, are intended to enhance the incentives for workers to participate in pension programs. 45 Kyrgyz Review of Social Expenditures June, 2001 Table 19. Social Fund Income and Expenditures, 1994-2000 (% GDP) 1994 1995 1996 1997 ^ 1998, 1999 -2000 (budgel) TOTAL INCOMEE 5.3 7.3 6.7 6.2 6.3 5.5 5.0 Pension Fund 4.3 6.0 5.5 5.2 5.2 4.4 4.3 Social Insurance Fund 0.7 1.0 0.9 0.6 0.5 0.4 0.3 Employment Fund 0.3 0.3 0.3 0.3 0.3 0.3 0.2 Health Insurance Fund 0.0 0.0 0.0 0.1 0.3 0.4 0.2 T OTAL EX]PEN"DITJURES 6.1 8.2 7.9 7.4 7.8 5.9 5.3 Pension Fund 5.1 7.3 7.2 6.6 7.0 5.2 4.8 Social Insurance Fund 0.9 0.7 0.5 0.4 0.4 0.3 0.2 Employment Fund 0.1 0.2 0.3 0.3 0.2 0.2 0.2 Health Insurance Fund 0.0 0.0 0.0 0.1 0.1 0. i 0.2 DEFICrr -0.8 -0.9 -1.3 -1.3 -1.5 -0.4 -0.3 State Budget Transfers 0.6 0.9 1.3 1.5 1.0 0.6 0.2 ATrrears in benefits payments 0.2 0.1 0.3 0.2 0.6 0.3 0.2 Memorandum: State Budget Transfers (Som mln) 75.5 139.7 308.1 446.2 344.0 300.0 150.0 Source: World Bank estimates based on Social Fund data. 108. Even as it manages roughly six percent of GDP, the Social Fund administration is extremely weak. Processing of pension applications is still handled manually and can take several months in many cases. Accounting procedures are being reviewed but are currently non-transparent and do not conform to International Accounting Standards. Roughly ten percent of expenditures and revenues are in-kind, compounding the inefficiencies and lack of transparency of the system. Policy monitoring and evaluation are also weak. On the more positive side, the Social Fund and the Tax Inspectorate have begun implementation of a project to coordinate collection efforts, including the introduction of a common identification number for workers and employers, joint audits, and exchanges of collection related information. To improve efficiency and accountability of the Social Fund, an overall review of the organization is necessary, with any recommended changes taking place step by step. 109. Recommendations for Pension System. The Review reached nine recommendations concerning pensions. * Pension adjustment policy. Balancing the Social Fund budget should be the first priority of the Government so the system does not drain public resources that are very much in need in other sectors. Accordingly, the Government and the Social Fund should carefully analyze the consequences of any increase in pension benefits to ensure that such an adjustment would be compatible with maintaining continuing balance in the financial position of the Social Fund (which is currently fragile) and that it does not postpone the improvement in this position that is both necessary and expected. Moreover, the Government 46 Kyrgyz Review of Social Expenditures June, 2001 should itself refrain from shifting expenditure responsibilities from the Social Fund to the State budget unless a decrease in State budget transfers to the Social Fund or a decrease in the payroll tax would offset such a transfer, so that the overall tax burden of the system remained unchanged. * Pass and implement the indexation law. The Govemment should pass this law, while maintaining the present approach to benefit levels. * Review rural pensions. Rural pensions are a problem for many countries because rural incomes are hard to measure, rural poverty rates are high, collecting contributions is an ongoing problem, and benefit entitlements for rural contributors can often be low. The proposed review would consider possible responses to these difficulties. * Poverty alleviation. The Government should postpone any significant increase in pensions until the financial performance of the Pension Fund clearly improves and shows that the additional expenditures can be sustained. Indeed, although many pensions are very low, any excessive increase in pension benefit will lead the Social Fund to accumulate new arrears in benefit payment. Arrears have a strong regional dimension, as the ability of the system to pay benefits depends on the regional level of contributions. Arrears have been more common in rural areas, where contributions follow the agricultural cycle, and local pension offices collect most of their revenue in the summer. Any excessive increase in pension benefit will result in the payment of higher pensions in Bishkek and higher arrears in other parts of the country. That would possibly raise poverty rates. * Increasing coverage. Social Fund revenues from contributions have been decreasing from 7.3 in 1995 to 5.5 percent of GDP in 1999. The Government should direct its efforts to holding and reversing this trend. Otherwise, the system will not be financially sustainable and successful in preventing poverty among the elderly. The Government should keep developing innovative approaches to maintaining and increasing coverage among farmers and the self-employed, such as the fees collected in the market place (bazaars). To maintain collections, the Government should refrain from introducing a voluntary, fully funded, public pillar. Besides overall macroeconomic and financial considerations, the preconditions for creating such a scheme in the Kyrgyz Republic are lacking. First, the contribution rate is too high and there is no room for any additional increase. Second, the Pension Fund continually faces financing problems, which result in pension payment arrears. Under these circumstances, it would be difficult to build public confidence in the Social Fund investment policy, if at the same time the pensions are not paid on time. * Disability Certification. The disability certification commissions need additional support to strengthen the improvement realized over the past years. Indeed, since the disability certification commissions have been transferred to the Social Fund, significant efforts have been made to improve efficiency and 47 Kyrgyz Review of Social Expenditures June, 2001 effectiveness of the disability certification process. In 1998, new criteria were introduced which emphasized medical reasons for disability eligibility while excluding social consideration. New medical experts were hired and some local commissions were reorganized, particularly in Osh and Jalal-Abad. The disability certification commissions, however, face a severe lack of equipment such as computers and means of transportation for outreach services they provide, as well as for verifying eligibility. * Social Fund Monitoring and Accounting Capacity. The Social Fund should pursue its effort in developing a modem accounting environment by introducing International Accounting Standards. The information such a system would provide is crucial for the Social Fund to be able to monitor the reform and the increase in the level of pensions, thereby fighting poverty among the elderly. The Government should support the creation of solid teams in the financial and accounting department if the implementation and use of such a tool is to be successful. In that respect, the repetitive changes in the management and the teams of this department has probably been counter- productive. c Monitor and evaluate the impact of the reform. The refonrm is based. on a variety of assumptions about future rates of economic growth and wage growth, labor market performance, and so on. There is uncertainty concerning all of these variables. Monitoring and evaluation are needed partly to assess whether the assumptions prove to be realistic and partly to assess the evolution and impact of the reforms themselves. Adjustments to the reform program could be considered if there are unforeseen developments. * Initiate a review of the organization and administration of the Social Fund. As noted in paragraph 109, the Social Fund faces a variety of organizational and administrative challenges that the proposed review could address. Price Discounts and Privileges 110. General Price Privileges. Kyrgyz legislation and decrees at the national level authorize numerous categories of individuals to be granted a wide array of pricing privileges for which the Republican budget is supposed to reimburse the providers of those goods and services. In practice, reimbursements are often delayed, partial or absent. Similar arrangements also apply at lower levels of government. 111. The estimated cost of these privileges in 2000 was Som 452 million. To finance these costs, about Som 343 million were provided in government budgets (300 million in the Republican budget and 43 million in local budgets). This left a funding gap of about Som 110 million for legislated privileges. 112. In a constrained fiscal environment, many of these privileges represent low- priority expenditures. Restructuring these privileges could therefore form part of a broader strategy for re-targeting assistance towards the poorest and most needy households.. Accordingly, the Government submitted to the Parliament, within the 48 Kyrgyz Review of Social Expenditures June, 2001 context of discussions on the proposed Consolidated Structural Adjustment Credit (CSAC), a draft law that would restructure privileges. The draft law is designed to be new controlling legislation on privileges. It would supersede existing privilege legislation and would streamline current legislative privileges by focusing principally on people in greatest need and by reducing substantially the aggregate cost of privileges. The Government is committed to financing fully all privileges and benefits authorized through the proposed law. 113. The draft privileges law would reduce the total estimated cost of privileges, the number of qualifying individuals and their family members who are eligible to benefit from them, and the number of different privileges. * According to MLSP calculations in 2000, the cost of financing all privileges in existing legislation would be about Som 439 million. The state budget actually included Som 343 million as funding for these privileges. Privileges in the draft law were estimated to cost Som 276 million, a reduction of 39 percent from the full cost of current privileges and a reduction of 19 percent from the 2000 budget allocation. * The number of individuals qualifying for privileges in 2000 would have fallen by 12 percent from 373,000 to 328,000, while the number of people benefiting (that is, qualifying individuals and their eligible family members) would have fallen by 31 percent from 475,000 to 330,000 * The number of different privileges authorized through existing legislation is essentially unknown, but the law on Chernobyl victims included about 60 types of privileges and the law on World War H participants included more than 40. The draft law incorporates 6 kinds of privileges: discounts for medicine, transportation, utilities, a place in a sanatorium, telephone charges, and extraordinary grants. 114. Estimates suggest that, under existing legislation, 38 percent of expenditures on privileges assist people or families who are poor (or in categories likely to be poor) or who have limited ability to work17. This proportion could rise to 69 percent under the draft law. Figure 6 presents a comparative analysis of the current and proposed new laws. 17 This includes as poor people in high-mountain areas, invalids, people eligible for utility subsidies under decrees 281 and 578, and people with per capita income less than 200 soms. 49 Kyrgyz Review of Social Expenditures June, 2001 Figure 6. Comparison of Current and Proposed Laws on Privileges Existing Laws Households with pc incoms below 200 somstnmth 3%/ Poor households 21% War veterans * t X ~~~~~~~~~46% Highlanders 5%- Invalids People wAth State awards 0 _ People with pensions for spedalactivities Sevicemer ' . 2% 3% Pol - - 3% Chemobyl vicdms Rehabilitated people 5% 3% Draft Law Households with pc income below 200 soms/mth 5% Poor households 281% War veterans 27%/ Highlanders 6% - Chemobyl victims Rehabilitat4dapeople 1% Invalids 29% 50 Kyrgyz Review of Social Expenditures June, 2001 115. Energy-related Price Discounts. Reforms to establish a viable economic basis for the energy and infrastructure sectors will require that tariffs be raised substantially to cost-recovery levels. These tariff increases will reduce significantly the already low purchasing power of the whole population. Accordingly, the Government has decided to structure energy subsidies that would help protect the poorest consumers against anticipated energy price increases through a 3-part strategy. (i) It will retain existing (largely non-means-tested) energy pricing privileges for five categories of households: recipients of Unified Monthly Benefit; pensioners with pensions below 400 soms; families with disabled children under 16; retirees living alone and not working; and low-income families in which children have lost at least one parent. The estimated cost in 2000 is Som 94 million. (ii) For gas, hot water and district heating, the Government has created a new set of "socially protected energy prices," for all parts of the country except Bishkek City. The new prices would rise more slowly than regular prices, for households with per capita incomes below 200 soms per month. (These will initially reflect non-privileged tariffs of the fourth quarter of 1999.) The measure will give poor households additional time to adjust their consumption of energy and other items as normal tariffs rapidly increase. In effect, it introduces a new income-tested category into the privilege structure. The incremental estimated cost [over and above (i), as the categories overlap] is Som 9.3 million in 2000 and Som 21 million in 2001. (iii) For electricity, where consumption is universally metered, a "lifeline" tariff structure will entitle households to receive the first 90 kilowatt hours per month at a heavily discounted rate. 116. Bishkek City will continue to operate and fund a city-wide "pilot" housing allowance program for poor apartment residents. For eligible households, the allowance covers the difference between expected utility costs and 30 percent of the household's income. For private houses, the City proposed to introduce, on a pilot basis, a program of income-tested compensation for energy tariff increases that would have effects similar to those of socially protected prices. Social Assistanct,Paid Primarily in Cash: State Benefits 117. Social assistance programs are designed to target assistance to poor or otherwise needy families and individuals. Kyrgyz State benefits, which constitute the main Republican system of cash social assistance for low-income families and individuals, include (a) social benefits, which are given to categorical groups regardless of income, and (b) a Unified Monthly Benefit (UMB), which is paid to eligible families on a means- tested basis. Expenditures on these benefits in 1999 were Som 350 million. 51 Kyrgyz Review of Social Expenditures June, 2001 118. Social Benefits. Social benefits complement the pension insurance system and protect people unable to work - invalids and those who are not entitled to the employment-based insurance pension. Social benefits are the analogs of social and disability pensions. The recipients of social benefits are invalid children, orphaned children without an economic provider, invalids from childhood, and "mother heroines," as well as people of pensionable age without a right to pension. 119. The basic rate of social benefit is an income standard, which is currently set at a level identical to the per capita income threshold used to determine a family's eligibility for the UMB. Through March 2001, this income standard was Som 100 per month. The Parliament's Legislative Assembly (lower chamber) decided on April 10, 2001 to increase the income standard, which is also known as the guaranteed minimum level of consumption, from Som 100 to Som 120, with effect from April 1, 2001. 120. Rates of social benefit are set as percentages of this income standard that vary with the category of recipient. For example, the social benefit for full orphans is 150 percent of the incomc standard.8 Social beneficiaries are also categorically eligible for privileges on housing and utilities payment. (The income standard for social benefits, and the UMB threshold with which it now coincides, are each substantially lower than both the "general" and "severe" official poverty lines, which were Soms 412 and Soms 216 per person per month in 1998. These are based, respectively, on an overall consumption basket and a food consumption basket. . 121. Unified Monthly Benefit. A new "Law on State Benefits" established a new Unified Monthly Benefit (UMB) in May, 1998, to provide a basis for a simplified and unified cash social assistance policy. It aims to guarantee a minimum income under means-testing, and is to be adjusted annually in line with an assessment of needs and available fiscal resources. 122. For each eligible family member, the UMB payment is (usually) the amount by which per capita family income is less than the minimum consumption guarantee-from April 1, 2001, Som 120 per month.'9 The decision of the Parliament's Legislative Assembly to increase the guaranteed minimum level of consumption will increase the cost of UMB in 2001 by SOM 130 million. 123. The formula that is used to calculate the minimum consumption guarantee yields that maximum level for the consumption guarantee that is affordable within a budget ceiling equal to 1.5 percent of GDP, given the estimated number of UMB benefit recipients, the estimated average income of UTMB recipients, and the estimated number of social benefit recipients (Tables 20 and 21). In recent years, however, until the decision in April, 2001, thi minimum consumption guarantee under the UMB was not revised upwards, despite iNTation, for fiscal reasons. Consequently, its real value had dropped and fewer families than otherwise met the increasingly stringent standard. With recipient 18 The Parliament also decided to increase the rate of social allowances for invalid children with effect from January 1, 2001, at a cost of Som 36.4 million. 9 For some categories of dependents, benefits are a multiple of the consumption guarantee itself. 52 Kyrgyz Review of Social Expenditures June, 2001 numbers and the real level of benefits both held down in this way, UMB program expenditures have fallen below the threshold of 1.5 percent of GDP. 124. An evaluation of the new administrative arrangements in September, 1998, highlighted a number of achievements, including the implementation of clear eligibility criteria, improved client friendliness such as ease of application and protection of confidentiality, the thorough investigation of applications including home visits, and better staff competency. It also pointed, however, to significant remaining issues including the high level of benefits paid in kind (40 to 60 percent in some rural areas), and the availability and quality of program administrative data to be used for program monitoring (Table 22).2o The administrative data indicate that caseloads require each benefits staff member to process applications for an average of 3 - 4 per families per day. If each application is assumed to require 2-3 office visits, a home visit and some follow- up to check documentation provided by the applicant, then a caseload of 2-3 families per staff member per day would seem manageable. MfLSP reports that, of its 2115 total staff, only 331 are involved in assessing and paying state benefits. (Other staff work on privileges and social services, for example.) That suggests an average caseload of 688 applications per year. MILSP estimates that caseloads for staff specializing in payment of state benefits are actually higher than this-762 applications in towns and 846 applications in districts. Table 20. Funds Allocated for Payment of Benefits, Awarded by the Ministry of Labor and Social Protection and Arrears (1995 - 1999) (Som '000) Approved in Actual Credit actually Financed, Arrears from Actual arrears the budget requirement of given by the including the approved including (including the Ministry of Ministry of delivered budget of the delivered arrears) Labor and Finance products Ministry of products at Social Finance Ministry of Labor Protection and Social Protection 1995 428 774.8 496 629.4 369 232.0 377 317.5 59 542.8 119 311.9 1996 478 569.6 298 661.6 405 397.3 411 546.7 73 172. 3 6 426.8 1997 391719.4 305 754.0 305 419.2 312 180.8 86 300. 2 0.0 1998 341 246.0 410 256.0 242 902.1 306 524.7 98343.9 103731.3 1999 478 343.9 349 288.2 418 996.6 453019.5 59 344.3 0.0 Source: Statistics provided by the MLSP. 20 Systematic differences in administrative data from different Ministries often arise. For example, MLSP records UMB assistance as paid to recipients when they receive their benefit in cash or kind at a post office. MOF records the assistance as paid when a commission authorizes an offset within official accounts for delivered food products. MLSP notes that there are sometimes significant delays between these events. 53 Kyrgyz Review of Social Expenditures June, 2001 Table 21. Number of Recipients of UM1B and Average Amount of Benefit (1995- 1999) 1995 1996 1997 1998 1999 Average number of Total UMB (people) 1002731 672944 643726 583423 487517 Average amount of 39.54 33.65 35.56 51.30 47.30 benefit (soms) Source: Statistics provided by the MLSP, Natstatkom Report. Table 22. Applications for UMB and Refusals (1998-1999) Total Appl- Number Refused Number % of Refused % of Approv- staff ications of appl- applicat of controls after refused ed Year for ications -ions controls control after applicat benefits per staff control -ions member 1998 252782 9168 44118 18.1 17465 39.6 226149 1999 331 227564 687 55429 89615 52.1 11134 12.4 161001 Source: Data provided by the MLSP. 125. A recent report on the effectiveness of social assistance in the Kyrgyz Republic concluded that, although the government had taken "concrete steps towards a means tested targeted system of social assistance," there is still substantial scope for improvement. Fully 50 percent of state benefits were paid in kind rather than cash. UMB was too low. Targeting was quite weak: 73 percent of recipients were not extremely poor, and many resources were given to non-poor. families; while many of the extremely poor were still left out from the system. Problems implementing the actual criteria were "remarkable.'2 The review incorporated several technical recommendations for improving U3MB and state benefits and the efficiency of their administration. MLSP agreed with the report's conclusions that the level of IvMB needed to be revised upward (a step subsequently taken), methods for means-testing require improvement, and administrative reforms are needed within MLSP's Division of Social Protection. MLSP disagrees with numerous other findings, however. It notes, for example, that regional imbalances in the distribution of MLSP local staff arise partly because local governments finance their salaries. In respect of the effectiveness of targeting, it notes that MLSP administers UMB using reported income (with adjustments) whereas household surveys measure poverty from consumption and expenditure data. 21 National Statistical Committee (2000). Effectiveness of Social Assistance in the Kyrgyz Republic. Bishkek: National Statistical Committee, pp. 28-29. 54 Kyrgyz Review of Social Expenditures June, 2001 126. Social Assistance Issues for Policy Makers. For Kyrgyz social policy makers, one of the most frustrating dilemmas is that the national budget cannot finance the level of assistance they would like to offer to poor Kyrgyz families. Behind this dilemma lie three inherently conflicting objectives of cash transfer programs. These are high levels of assistance to the poorest groups, strong incentives for them to work, and low overall expenditures on social assistance. They conflict because they involve three policy variables-the maximum benefit, the effective marginal tax rate, and the ceiling income level for benefit eligibility-but only two of these can be set independently. Unfortunately, raising the basic benefit amount for someone without other income will also raise total program spending, unless higher spending is offset by faster withdrawal of benefits as private income increases. But that would worsen work incentives! More generally, policy changes to achieve any of the three goals more completely would worsen attainment of at least one of the other two. So balancing benefit adequacy, work incentives, and program affordability unavoidably involves tradeoffs among competing objectives. 127. A second issue for social assistance policy makers is how to target assistance more effectively. (The targeting issue is made more difficult by high levels of informal employment and self-employment among the families potentially eligible.) This issue can be viewed in a broader way, however. The existing national policy framework for targeted assistance incorporates a number of discrete benefit systems that were designed largely independently of one another and without a fully coordinated overall strategy for coherent, equitable, cost-effective and administratively efficient targeted assistance. These benefit systems include privileges, energy-related assistance (including non- income tested categorical privileges and income-tested socially protected prices), social benefits, and UMB. Within Bishkek, there are also housing allowances. To these one might in principle add any remaining general price subsidies. The overall effectiveness of the Kyrgyz social protection system could be enhanced by a review that sought to develop such a strategy and incorporate its findings within a further rationalization of these benefit systems. 128. There is no rigorous theory of the best way to target social assistance in the complex circumstances found in the Kyrgyz Republic, but the following principles are relevant. * Equity is usually thought of as having both vertical and horizontal dimensions. The vertical dimension ranks individuals or households according to their resources and implies that related assistance should be targeted towards those with low incomes or means. The horizontal dimension involves differences in "needs" between households at any given income level and implies that related assistance should be targeted towards those individuals or households with special needs (e.g., because of family size, illness or disability, or remote or harsh geographic circumstances). * In a budget-constrained environment, one possible social judgment might be to give vertical aspects of equity higher priority than horizontal 55 Kyrgyz Review of Social Expenditures June, 2001 dimensions. Alternatively, horizontal equity aspects (special needs) might attract extra public assistance only for those who also have low incomes. * "Categorical" assistance involves providing specified benefits to individuals or family members who are members of a category that is defined by criteria other than income or wealth-for example, people above retirement age, families with children, or war veterans. * Pricing privileges for groups designated as eligible constitute one form of categorical assistance. Pricing privileges therefore have the advantages and disadvantages shared by all forms of categorical assistance. Categorical assistance is often juxtaposed with means-tested or income- tested assistance, although the two approaches are not mutually exclusive. * Income- and means-testing are administratively complicated, relatively expensive, and often inaccurate. They are a relatively expensive way to target assistance that has a low average value per recipient. When eligibility for benefits cuts out suddenly at a particular income level, inequities can be created along with incentives to maintain incomes below the threshold level. * Categorical assistance is usually administratively simpler and less expensive. It can be an effective way to reach groups with special needs. But it is often inefficient as a method for targeting assistance toward the poorest families. When eligibility for categorical benefits is linked to (that is, defined by) eligibility for another social benefit, administrative costs can be particularly low. However, categorical assistance also creates potentially undesirable incentives for people to self-select themselves into the favored categories (e.g., one-parent families or unemployed people). * Even where categorical approach to assistance seems appropriate, there are further questions about whether that assistance should be provided in cash, through vouchers, through pricing discounts, or in other ways. * Finding the most desirable balance among pricing privileges, other categorical approaches, and income- or means-tested approaches therefore requires judgements about the best way to balance administrative simplicity and economy against efficiency in limiting assistance to the desired target group and the avoidance of undesirable incentives. Typically, countries include both categorical benefits and means-tested or income-tested benefits within their systems of social assistance. 129. In short, the most effective and equitable system of social assistance in the Kyrgyz Republic (a) should probably target assistance not just to those who are poor and (b) should provide assistance to those who are poor through a combination of income- tested, means-tested and categorical programs which could include pricing discounts. In other words, a doctrinaire prescription to use only a single approach would probably be misguided. 56 Kyrgyz Review of Social Expenditures June, 2001 130. Accordingly, the best approach to Kyrgyz social assistance targeting issues might be to conduct a review that: (i) establishes priorities about groups meriting public assistance in some form; (ii) documents the assistance that these and other groups already receive; (iii) establishes guidelines about when each alternative approach to providing assistance should be used; (iv) applies these guidelines to propose a set of (categorical and/or income-tested or means/tested) beneficiary categories, associated benefits, appropriate delivery mechanisms, and improved administrative arrangements, and (v) on this basis recommends ways to rationalize the structure of assistance across the current set of benefit systems so as to achieve a more coherent, equitable, cost-effective and administratively efficient system of targeted assistance. Employment Programs 131. The Employment Fund finances two kinds of employment programs: passive and active. Unemployment Benefit is the only passive program, but there was a range of active labor market programs (ALMPs) in 1999.22 132. Passive Programs: Unemployment Benefit. The Review draws attention to three pertinent questions that policy makers might wish to address further concerning Unemployment Benefit. First, can an unemployment insurance program be administered satisfactorily in an environment where determining whether an individual is involuntarily unemployed is extremely difficult? Second, is unemployment insurance viable and appropriate in the context of a dramatic and sustained slump in economic output and when a large number of people have frequently faced significant arrears in wages and benefit payments? Third, is unemployment insurance the best use of social protection funds, given the size of the informal sector, the inability to verify employment status effectively, and the broader economic context? 133. Active Programs: Active Labor Market Programs Including Micro-credit. Two comprehensive reviews of active labor market programs in the Kyrgyz Republic have reported since 1999. The earlier one found that: the demands placed on the Employment Services were requiring them "to be all things to all people;" consistent criteria for eligibility for programs were not applied in practice; and job search and vocational training were core Employment Service activities, although international experience with 22 Of these, the main ones were: (i) Job Placement Activities, including job fairs, youth brokerages that linked youth with jobs, job clubs to assist long-term unemployed job-seekers, and a "job check" program that allowed unemployed youth to arrange two-year job contracts that entitled the employer to a wage subsidy; (ii) Training and Retraining for registered unemployed people by contracted training centers (mainly vocational schools) for a period of up to 12 months; (iii) Public Works that created a public benefit and were managed by a contracted implementing organization, for subgroups of the registered unemployed and for a maximum period normally of 12 months; (iv) Self-Employment Program which provided the lump-sum equivalent of 12 months UB to registered unemployed eligible for UB who have a viable business plan for self-employment; (v) Micro-Credit Program, which provides a loan of up to 30 times UB, at an interest rate of 2 percent per month, for registered unemployed people initiating local community projects outside main urban areas through a contract with legally independent Centers of Micro-Credit; and (vi) Employment Promotion Companies, which were NGOs that implemented public works and other similar ALMP activities approved by State Department of Labor that had public infrastructure or other benefits, primarily in rural areas, and that had cofinancing local sponsors. 57 Kyrgyz Review of Social Expenditures June, 2001 the effectiveness of vocational training suggests that it has only a marginal impact on unemployment.23 134. The later report found mixed performance among the various ALMP programs24. It classified the cost and the quality of each program as either high, moderate, or low and then arrayed programs in a matrix on this basis. For example, several job placement programs (youth brokerages, job clubs and job check) scored well, with moderate or good quality and low costs. Conversely, others (self-employment and micro-credit programs) scored poorly with low or moderate quality and high costs. In addition to numerous program-specific recommendations, the report also recommended: * Grouping programs into broader categories for budgeting purposes while giving local offices greater discretion in allocating these funds; * Reallocating funds from poorly performing programs (self-employment and micro-credit) lo well performing program categories (primarily job search activities and, secondarily, training/retraining); * Allowing the NGOs running micro-credit programs to assume future responsibility for micro-credit activities; * Improving (a) monitoring and evaluation systems for the programs, (b) management information flows from the Employment Fund to the State Department, and (c) budgeting and financial management processes within the Employment Fund; and * Considering a gradual reduction and restructuring of contributions to the Employment Fund. 135. After considering these reports on the efficiency of Kyrgyz labor market employment programs, the Supervisory Council on State Social Insurance recently decided to cease funding the Employment Fund's self-employment and micro-credit programs for unemployed people from the second quarter of 2000. It reallocated those funds to be used for vocational education, public works and other labor market programs for which the Employment Fund is responsible. The share of the Employment Fund's expenditures on active labor market programs would be limited to no more than 35 percent of its total expenditures. Children's Services 136. There is an absence of community services for children and families. Although cash assistance can alleviate the stress on a family suddenly facing unemployment or a severe drop in income, it cannot help a family to grapple with more complex and longer 23 Hess, Jolant (January, 1999). The Kyrgyz Republic: Assessment of Labor Market Institutional Reforms. (Processed). 24 Wolfgang Schwegler-Rohmeis (2000). Review of Active Labor Market Policy in Kyrgyzstan: A Report to the World Bank. (Processed). 58 Kyrgyz Review of Social Expenditures June, 2001 term problems, such as the severe disability of a child or substance abuse by a parent. Prior to independence, Kyrgyz families could rely on a comprehensive range of social supports. However, when families faced severe difficulties, such as a child's disability, alcohol or drug abuse, sudden unemployment, child abandonment or spousal battering, the primary service available was placement in a residential institution (such as orphanages, homes for physically and mentally disabled, and detention centers). These institutions tend to be large, deteriorating and demoralizing facilities. Little community based help tailored to the needs of the individual was or still is available. Such services include counseling and social work, specialized educational services, day care, and shelters for battered women. The situation has deteriorated severely with the transition. 137. Residential care for children in any country is considered by most social work professionals to be a measure of last resort, and is associated with delayed emotional and behavioral development. It is also extremely expensive when compared to the cost of alternatives that keep families at home - projects that are successful in reducing institutionalization are likely to generate rapid economic returns.25 As importantly, community based services, when implemented effectively, can help children and families to lead more productive and satisfying lives through programs tailored to the needs of the individual. ACTION AGENDAS Action Agenda for the Health Sector 138. The Kyrgyz health system now spends so much keeping buildings open, heated and staffed that it is starved of funds to finance actual care to patients. (Staffing and utilities alone absorbed two-thirds of government health spending in 1998, whereas pharmaceuticals and supplies represented only 12 percent of the total.) So the chief priority for action in health is to reduce the excessive number of facilities through well- designed consolidation and closures and to use the financial resources released to fund the incremental costs of care for patients. Complementary priorities are: * To reorient patterns of care away from overuse of hospitals and specialists and towards increased reliance on community-based primary care as well as modern approaches to specialized care; * To strengthen efforts to prevent disease and promote healthy lifestyles; * To reimpose a legal and equitable framework on the payments-both formal and informal-that patients make to providers; * To continue the introduction of more rational and flexible processes for allocating funds within the sector, on the national as well as on the oblast level; 25 A recent economic analysis on the impact of community multi-service centers in Lithuania projected annual savings of US$1,800 for each client receiving services that prevented institutionalization. 59 Kyrgyz Review of Social Expenditures June, 2001 * To continue implementing the planned reforms in the financing, regulation and delivery of health care that represent the route to a more effective and sustainable health system; and * To coordinate the activities and realign the roles of MoH and MESC so as to foster health-improving interventions in schools and other non-health facilities and to give MoH the primary role in determining whether new medical faculties will be established and the number of students per year that medical faculties should train. (MESC disagrees with this realignment in roles.) Action Agenda for the Education Sector 139. In education, in addition to addressing fundamental strategic questions about the role of education in the future Kyrgyz economy, policy makers should refocus primary and secondary education on general education (rather than specialized or vocational education) that: * Restores high quality by using updated textbooks, teaching materials and curricula; * Prepares people for long-term employment in a market-oriented economy (which will require a different mix of skills) by (i) emphasizing the development of widely adaptable cognitive skills and concepts (rather than knowledge accumulation) and (ii) training teachers accordingly; and * Reverses the decline and maintains universal enrollment and attendance rates, especially at the upper secondary level. 140. In addition, to free resources for currently unmet needs within the education system, policy makers should pursue a variety of options for using resources more efficiently and diversifying sources of financing for education. Action Agenda for the Social Protection Sector 141. Continue and extend reforms in social protection policies, so that the limited funds available for these purposes are allocated where they would have maximum impact. Maintain the existing pension reform agenda that, if accompanied by reasonable economic growth, will provide affordable and adequate pensions while gradually but significantly lowering the rate of payroll contribution rate required to support the system from 40 percent today to 23.5 percent in 2010. Accordingly, (i) pass and implement the indexation law, while maintaining the present approach to benefit levels; (ii) review rural pensions; (iii) postpone any significant increase in pensions until the financial performance of the Pension Fund clearly improves and shows that the additional expenditures can 60 Kyrgyz Review of Social Expenditures June, 2001 be sustained; (iv) refrain from introducing a voluntary, fully funded public pension pillar; (v) act to improve collections; (vi) initiate an organizational and administrative review of the Social Fund; (vii) adopt International Accounting Standards within the Social Fund; and (viii) monitor and evaluate the impact of the reform. * Continue efforts to improve the targeting of social assistance, privileges, and energy-related assistance on those households in greatest need. In addition, (i) pursue the rationalization of privileges for which a draft law has been prepared; and (ii) initiate a broader review designed to develop a more coherent overall strategy for all targeted assistance and to rationalize the current programs within that framework. * Consider whether unemployment insurance is viable and appropriate in a context where (i) there has been a dramnatic and sustained slump in economic output; (ii) it is extremely difficult, given the size of the informal sector, to verify employment status effectively and to determine whether an individual is involuntarily unemployed; (iii) a large number of people have frequently faced significant arrears in wages and benefit payments; and (iv) there are urgent potential alternative uses within the social sector for the revenues that the Unemployment Benefit program absorbs. 61 Kyrgyz Review of Social Expenditures June, 2001 REFERENCES Abrahart, Alan. (1997). "Kyrgyz Republic Employment Policies. Facing East." Paper prepared for Ministry of Labor and Social Protection Seminar, "A Strategy for Social Policy of the Kyrgyz Republic as a Frameworkfor "SOSAC' Project. " Bishkek, September 24-26, 1997. Abel-Smith, B. and Falkingham, J., (1996). "Financing Health Services in Krygyzstan: the Extent of Private Payments." Report for the British Government Know How Fund.. London School of Economics. Dorabawila, V. (2000). "Trend and Status of Education and Private Expenditure, Kyrgyz 1993- 1997." Washington, DC: World Bank. (Processed). Asian Development Bank. (1996). Kyrgyz Republic: Education and Training Master Plan. Report prepared by ABU Consultant. Manila: Asian Development Bank. Asian Development Bank (1997). Education Sector Development Program. RRP:KGZ 28441. Asian Development Bank. (1998) Social Services Delivery and Finance Project, 1998, RRP:KGZ 29660. Bauer, A., Green. D.,and Kulnast, K. (1997). The Kyrgyz Republic in Transition: Women and Gender Relations. Manila: Asian Development Bank. Brooks, D.H. and Myo, T. (1998). Social Sector Issues in Transitional Economies of Asia. New York: Oxford University Press. Cusan, A. and Motivans, A. (1998). "Fiscal Adjustment, Education Equity and Access in Transition", Florence: International Child Development Centre Mimeo. Eversmann, E. (1999). "School attendance in the Kyrgyz Republic". Unpublished report for UNICEF Bishkek Office. Gwatkin, D.G., S. Rutstein, S., Johnson, K., Pande, R., and Wagstaff, A. (2000) . "Socio- economic differences in health, nutrition, and population in the Kyrgyz Republic." Discussion Draft. Washington, DC: The World Bank: HNP/Poverty Thematic Group. Gedik, G., J. Kutzin, A. Fawcett-Henesy, and F. Apfel (2000). "Report on the implementation of health care reforms in The Kyrgyz Republic for the period December 1998 - May 1999." Copenhagen: World Health Organization, Regional Office for Europe. Government of the Kyrgyz Republic (2000). "Resolution #10 on the Reform of the Local Governance Structures. " Bishkek: Government of the Kyrgyz Republic. Hackett, R. (1999). "Evaluating Educational Expenditures." Unpublished draft report prepared for the Government of Kyrgyzstan. USAID/Barents Group. Harding, A. and Preker, A. (1998). "Innovations in Health Care Delivery: Reform with the Public Sector. Background and Conceptual Framework." Draft paper presented at: Organizational Reforms in Health Services: Lessons Learned from the Corporatization of Hospitals. World Bank seminar. November 12-13, 1998, Washington, DC. 62 Kyrgyz Review of Social Expenditures June, 2001 Hess, J. (1999, January). "The Kyrgyz Republic: Assessment of Labor Market Institutional Reforms." (Processed). Howse, G., Wicks, J., and Stackpole, E. (1999). Inception Report for Support for Financing Reform in the Education Sector. Education Sector Development Project. Bishkek, Krygyz Republic: Ministry of Education, Science, and Culture. Sargaldakova, A. and J. Healy. (2000) Health Care Systems in Transition (HiT): The Kyrgyz Republic. Copenhagen: European Observatory on Health Care Systems. International Monetary Fund (1995). Unproductive Public Expenditures: A Pragmatic Approach to Policy Analysis. Washington, DC: International Monetary Fund. Kaiser Permanente International. (1998). Georgia Restructuring Master Plan for Tbilisi. Unpublished Consultancy Report for the Ministry of Health, Republic of Georgia. Klugman, J. (1999). "Financing and Governance of Education in Central Asia". In Central Asia 2010: Prospects for Human Development. New York: Regional Bureau for Europe and the CIS, UNDP. Kyrgyz National Statistical Committee (2000, March). Effectiveness of Social Assistance in the Kyrgyz Republic. Bishkek: Kyrgyz National Statistical Committee. Lewis, M. S. (2000). Informal Health Payment in Eastern Europe and Central Asia. Issues, Trends and Policy Implications. Washington, DC: World Bank. Lewis, M. S. (2000). Governance and Health in Eastern Europe and Central Asia: Paying for Health Care. Washington, DC: World Bank. Ministry of Health, The Kyrgyz Republic (1996). Manas National Programme on Health Care Reforms (1996-2006). Bishkek: Ministry of Health, The Kyrgyz Republic. Ministry of Health., (January 18, 2000). "Decree #8 On Reorganization of Oblast Health Care Institutions." Bishkek: Ministry of Health, the Krygyz Republic. Ministry of Education, Science and Culture. (1999). Country Report on Education for All. Bishkek: Ministry of Education, Science and Culture, the Kyrgyz Republic. National Statistical Committee. (1999). Women and Men in the Kyrgyz Republic. Bishkek: National Statistical Comnmittee. National Statistical Committee. (1997). Statistical Yearbook of Kyrgyzstan 1997. Bishkek: National Statistical Committee. National Statistical Committee. (1998). Statistical Yearbook of Kyrgyzstan 1998. Bishkek: National Statistical Committee. National Statistical Committee the Kyrgyz Republic. (1999). Report on Poverty in the Kyrgyz Republic - 1998. Bishkek: National Statistical Committee of the Kyrgyz Republic. 63 Kyrgyz Review of Social Expenditures June, 2001 Newell, A. & Reilly, B. (1999). Rates of Return to Educational Qualifications in the Transitional Economies. Education Economics, Vol. 7, No. 1,67-84 O'Dougherty, S and Wickham, C. (1999). "Separation of Health Purchaser and Providers and the Development of a Single Health Purchaser: Next Steps in Funds Flow and Ministry of Health Organization." Almaty: ZdravReform Project, Abt Associates, Inc. OECD (1997). Education at a Glance 1997. Paris: OECD. Ormonbekov, T.O. (2000, May). "On the Results of the Kyrgyz Higher Education System for 1999 and the Main Objectives of the Kyrgyz Ministry of Education, Science and Culture for 2000." (Processed). Psacharopolous, G. (1994). "Returns to Investment in Education: A Global Update". World Development Report, Vol. 22, pp. 1325-1343. Rohmeism, W. S. (2000, February). Review of Active Labor Market Policy in Kyrgyzstan. A Report to the World Bank. (Processed). Ryslaveia, and Gibraeva, G. (1999). Educational Financing and Budgeting in Kyrgyzstan. Paris: Institute for Internationa! Education Planning. Sargaldakova, A. and J. Healy (2000). Health Care Systems in Transition (HiT): The Kyrgyz Republic. Copenhagen: European Observatory on Health Care Systems. Schieber, G. and A. Maeda. (1997). "A Curmudgeon's Guide to Financing Health Care in Developing Countries." In Schieber, G., (Editor), Innovations in Health Care Financing: Proceedings of a World Bank Conference, March 10-11, 1997. World Bank Discussion Paper. Washington, DC: World Bank. WHO/EURO (2000). Health for All Database, European Region. January Update. Copenhagen: World Health Organization, Regional Office for Europe. Wicks, J. (1999). Early Childhood Consultancy. Quarterly Report. Bishek: Ministry of Education, Science and Culture. ADB Loan 1555-KGZ. World Bank (1993). Kyrgyzstan: The Transition to a Market Economy. Washington, D.C.: World Bank. World Bank (1999). The Consultation with the Poor: The Kyrgyz Republic. Washington, DC: World Bank. World Bank (1999). Kyrgyz Republic: Update on Poverty in the Kyrgyz Republic. Report No. 19425-KG. Washington, DC: World Bank. World Bank (2000). Education and the Poor in ECA Countries. In ECA Poverty 2000. Washington, DC: World Bank. World Bank (2000 in preparation). Kyrgyz Republic Growth Study. 64 IMAGING Report No.: 22354 KZ Type: SR