POLI 6995 Report No. AUS0002107 Kyrgyz Republic Public Expenditure Review Better spending for building human capital March 2021 THE KYRGYZ REPUBLIC GOVERNMENT FISCAL YEAR January 1 – December 31 CURRENCY EQUIVALENTS (Exchange Rate Effective as of February 28, 2021) Currency Unit: Kyrgyzstani som (KGS) US$1.00 =84.80 KGS Regional Vice President: Anna Bjerde Country Director: Lilia Burunciuc Regional Director: Lalita Moorty Practice Manager: Sandeep Mahajan Lead Economist: Ivailo Izvorski Task Team Leaders: Appolenia Mbowe, Bakyt Dubashov, Ahya Ihsan ACRONYMS AND ABBREVIATIONS ADP Additional Drug Package MoH Ministry of Health CEATM Center for Educational Assessment and NBKR National Bank of the Kyrgyz Republic Testing COVID-19 Coronavirus Disease of 2019 NCD Non-Communicable Disease DB Defined Benefit NDC Notional Defined Contribution DRG Diagnosis-Related group NSBA National Sample Based Assessment ECA Europe and Central Asia NSC National Statistical Committee ECEC Early Childhood Education Care OECD Organisation for Economic Co-operation and ECDI Early Childhood Development Index Development EDI Early Development Instructions OOP Out-of-Pocket EDS Education Development Strategy PAYG Pay-as-You-Go EU European Union PEFA Public Expenditure and Financial Accountability EGRA Early Grade Reading Assessment EMIS Education Management Information PER Public Expenditure Review System FAP Feldsher-Obstetrical Ambulatory Point PHC Primary Health Care FGP Family Group Practice PISA Programme for International Student Assessment FMC Family Medicine Center PIT Personal Income Tax RT Republican Testing GDP Gross Domestic Product SGBP State Guaranteed Benefits Package GNI Gross National Income GPE Global Partnership for Education SOE State-Owned Enterprise IMF International Monetary Fund TA Technical Assistance KGS Kyrgyz Som UNICEF United Nations Children’s Fund MHIF Mandatory Health Insurance Fund VAT Value-Added Tax MICS Multiple Indicators Cluster Survey WDI World Development Indicators MoE Ministry of the Economy WHO World Health Organization MoF Ministry of Finance MES Ministry of Education and Science TABLE OF CONTENTS EXECUTIVE SUMMARY.................................................................................................................... 1 CHAPTER 1: EDUCATION ............................................................................................................... 9 A. INTRODUCTION ...................................................................................................................... 9 B. OVERVIEW OF THE EDUCATION SYSTEM ...................................................................... 10 C. PERFORMANCE OF THE EDUCATION SECTOR ............................................................. 13 D. GOVERNANCE AND PLANNING IN THE EDUCATION SECTOR ............................... 17 E. LEVEL, EFFICIENCY, AND EFFECTIVENESS OF EDUCATION EXPENDITURE .... 18 F. RECOMMENDATIONS ........................................................................................................... 31 CHAPTER 2: HEALTH .....................................................................................................................40 A. INTRODUCTION .....................................................................................................................40 B. OVERVIEW OF THE HEALTH SECTOR ............................................................................. 41 C. OVERVIEW OF THE HEALTH SYSTEM..............................................................................45 D. HEALTH INPUTS ....................................................................................................................48 E. UTILIZATION OF HEALTH SERVICES ...............................................................................52 F. HEALTH EXPENDITURE.......................................................................................................55 G. PAYMENT SYSTEMS ...............................................................................................................63 H. COVID-19 RESPONSE AND THE MACRO-FISCAL IMPACT OF COVID-19 ON HEALTH FINANCING ........................................................................................................................65 I. RECOMMENDATIONS ............................................................................................................69 CHAPTER 3: PENSIONS..................................................................................................................74 A. INTRODUCTION .....................................................................................................................74 B. OVERVIEW OF THE PENSION SYSTEM ............................................................................75 C. FISCAL FORECASTS OF THE EXISTING SYSTEM (STATUS QUO SCENARIO) ............83 D. ASSESSMENT OF THE EFFECT OF PARAMETRIC REFORMS .......................................88 E. RECOMMENDATIONS ...........................................................................................................92 ANNEXES ..........................................................................................................................................96 List of Figures Figure 1.1. Total student population in the country has increased by 40 percent from 1990 to 2019.................................................... 10 Figure 1.2. School-age population is expected to increase till 2045 and start declining after 2070 ......................................................... 10 Figure 1.3. Education system in the Kyrgyz Republic ..................................................................................................................................... 11 Figure 1.4. While 90 percent of 6-year-old children are covered by preschool or general education, around half of 5-year-old and 70 percent of 4-year-old children are currently not covered by the education system .................................................................................... 12 Figure 1.5. Weekly preschool hours, by type of guarantee, 2018/19 ............................................................................................................ 12 Figure 1.6. Net enrollment rates in preschool education have improved consistently but fall behind comparator countries ........... 14 Figure 1.7. Total net enrollment rates for primary education in the Kyrgyz Republic have increased to universal levels .................. 14 Figure 1.8. Total net enrollment rates for lower secondary have also increased to near universal levels ............................................... 14 Figure 1.9. Total net enrollment rates for upper secondary in the Kyrgyz Republic are much lower than universal levels ............... 14 Figure 1.10. Percentage of children ages 2–4 with whom a father or mother or any adult has engaged in four or more activities that promote learning and school readiness during the past three days ............................................................................................................... 15 Figure 1.11. Percent of children younger than 5 living in households with three or more children’s books ........................................ 15 Figure 1.12. About 72 percent of children younger than 5 are developmentally on track in health, learning, and psychosocial wellbeing in the Kyrgyz Republic, which is low compared to other countries. .......................................................................................... 16 Figure 1.13. The education budget is centrally controlled and executed. ..................................................................................................... 17 Figure 1.14. Education expenditure in nominal terms has continued to rise .............................................................................................. 19 Figure 1.15. Distribution of annual per-household expenditure on education by deciles of socio-economic status ........................... 20 Figure 1.16. Per-student public expenditures have increased over time ....................................................................................................... 22 Figure 1.17. Composition of public expenditure by education levels in 2017 ............................................................................................. 22 Figure 1.18. Except for Naryn Province, there is low variation in expenditures per capita on general education ............................... 22 Figure 1.19. Enrollment rates in general secondary education have increased for all students, inequities across regions persist ..... 22 Figure 1.20. Per-student public expenditure for preschool education across regions ................................................................................ 23 Figure 1.21. Composition of Kyrgyz Republic budget across personnel, capital and non-staff current expenses is comparable to other developing countries.................................................................................................................................................................................... 23 Figure 1.22. Expenditure by functional classification: Preschool .................................................................................................................. 24 Figure 1.23. Expenditure by functional classification: General secondary education (primary/lower secondary/upper secondary) 24 Figure 1.24. Expenditure by functional classification: Primary and secondary vocational education ..................................................... 24 Figure 1.25. Expenditure by functional classification: Post-secondary and tertiary education................................................................. 24 Figure 1.26. Despite major increases in education salaries, they are fall far below the national average ................................................ 25 Figure 1.27. Despite substantial increases in the number of teachers, student to teacher ratios in early grades remain stubbornly high .................................................................................................................................................................................................................................... 25 Figure 1.28. The system has an excess of teachers overall, but sub-optimal allocation of teachers across subjects and schools result in shortages of teachers for specific subjects, primarily language and math ................................................................................................ 27 Figure 1.29. The Kyrgyz Republic spends a considerable percentage of total government education budget on energy ................... 29 Figure 1.30. Energy performance under energy efficiency class B ................................................................................................................ 29 Figure 2.1. Life expectancy at birth ..................................................................................................................................................................... 42 Figure 2.2. Under-5 mortality rate per 1,000 live births .................................................................................................................................. 42 Figure 2.3. Maternal mortality ratio per 100,000 live births ............................................................................................................................ 42 Figure 2.4. Life expectancy at birth by sex, Kyrgyz Republic, 1990-2019 .................................................................................................... 43 Figure 2.5. Under-5 mortality rate per 1,000 live births and prevalence of stunting by wealth quintile, 2018 ...................................... 43 Figure 2.6. Probability of dying between the ages of 30 and 70 from one of the four main NCDs, 2016 ............................................ 44 Figure 2.7. Smoking prevalence, 2016 ................................................................................................................................................................ 44 Figure 2.8. Percentage of hypertensive persons taking prescribed blood pressure medication regularly, .............................................. 45 Figure 2.9. Share of insured population 2016-18 .............................................................................................................................................. 46 Figure 2.10. Hospital beds per 1,000 population .............................................................................................................................................. 49 Figure 2.11. Hospital beds per 1,000 population by oblast, 2010 –18 ........................................................................................................... 49 Figure 2.12. Physicians per 1,000 population, 1990–2017 .............................................................................................................................. 50 Figure 2.13. Physicians per 1,000 population, latest year available ................................................................................................................ 50 Figure 2.14. Nurses and physicians per 1,000, Kyrgyz Republic, 2004 –14 .................................................................................................. 51 Figure 2.15. Nurse-to-physician ratio, latest year available ............................................................................................................................. 51 Figure 2.16. Vacancy rates for physicians by facility, 2018 ............................................................................................................................. 52 Figure 2.17. Coverage of postnatal care, 2018 ................................................................................................................................................... 52 Figure 2.18. Utilization of health services, latest available year ...................................................................................................................... 53 Figure 2.19. Bed occupancy rate and average length of stay, latest available year ...................................................................................... 54 Figure 2.20. Cost per bed-day per hospital and occupancy rate ..................................................................................................................... 55 Figure 2.21. Current health expenditure per capita (US$) and as a share of GDP, 2017 (percent) ......................................................... 56 Figure 2.22. Current health spending per capita (in constant 2017 US$) by source, 2000-17 .................................................................. 56 Figure 2.23. Government health spending as a share of GDP, 2017............................................................................................................ 57 Figure 2.24. Government expenditures, 2014-18 ............................................................................................................................................. 57 Figure 2.25. Distribution of government health spending by program, 2018 ............................................................................................. 58 Figure 2.26. SGBP spending by level of care, 2018.......................................................................................................................................... 60 Figure 2.27. Salaries in health relative to national average and government salaries in other sectors, .................................................... 60 Figure 2.28. Program-based budget for MHIF, 2020–2022 ........................................................................................................................... 60 Figure 2.29. Inefficiencies of low public spending on health ......................................................................................................................... 61 Figure 2.30. Out-of-pocket spending as a share of total health spending vs. public spending on health as a share of GDP, 2017 (percent) ................................................................................................................................................................................................................... 62 Figure 2.31. Composition of out-of-pocket health spending in the Kyrgyz Republic, 2000-14 .............................................................. 62 Figure 2.32. Estimated cost vs. MHIF rate for select inpatient services ...................................................................................................... 64 Figure 2.33. Components of MHIF’s Results-Based Financing schemes .................................................................................................... 65 Figure 2.34. Breakdown of COVID-19 emergency response expenditures by government entity ......................................................... 66 Figure 2.35. Real GDP per capita growth actual (2000-19) and projections (2020-25), (percent) ........................................................... 67 Figure 2.36 Government revenues and expenditures in the Kyrgyz Republic, actual (2010-19) and projections (2020-25) .............. 67 Figure 2.37. Projections of government health spending per capita (2017 US$), 2020-25 ....................................................................... 68 Figure 3.1. Distribution of contributors by age ................................................................................................................................................ 78 Figure 3.2. Share of contributors among population, by age (percent)............................................................................................................ 78 Figure 3.3. Distribution of old-age pensioners by pension size at the end of 2019 (percent) .................................................................. 81 Figure 3.4. Projected number and age structure of population ...................................................................................................................... 85 Figure 3.5. Projected coverage of older people with old-age pensions ......................................................................................................... 85 Figure 3.6. Average old age pension replacement rates ................................................................................................................................... 86 Figure 3.7. Projected government obligations ................................................................................................................................................... 87 Figure 3.8. Projected contributory revenues and total expenditures of the system .................................................................................... 87 Figure 3.9. Forecast results related to the establishment of qualifying period ............................................................................................ 89 Figure 3.10. Projected number of old-age pensioners if retirement age is changed ................................................................................... 90 Figure 3.11. Projected Level of NDC pensions for new old-age pensioners when using gender differentiation of actuarial coefficients .................................................................................................................................................................................................................................... 91 Figure 3.12. Projected level of NDC pensions for new old-age pensioners in case of indexation of accounts by 100 percent of wages growth ....................................................................................................................................................................................................................... 92 List of Tables Table 1.1. Education system in Kyrgyz Republic as of 2019 .......................................................................................................................... 11 Table 1.2. The Kyrgyz Republic ensures a legal right to preschool from a much earlier age as compared with the EU average ..... 12 Table 1.3. Usage of current school infrastructure............................................................................................................................................. 26 Table 2.1. Vaccination coverage, 2000-18 .......................................................................................................................................................... 43 Table 2.2. Number of health care facilities in the Kyrgyz Republic, 2000 –18 ............................................................................................ 48 Table 2.3. Avoidable hospitalizations in the Kyrgyz Republic, 2016-18 ...................................................................................................... 54 Table 2.4. Mandatory Health Insurance Fund revenues, 2018 and 2019 ..................................................................................................... 57 Table 2.5. MHIF revenues and expenditures for exempt categories of the population, 2019 ................................................................. 58 Table 2.6. MHIF’s reimbursement under the ADP compared to estimated need, 2017 ........................................................................... 59 Table 2.7. Public financing and the universal health coverage index of service coverage, 2017 .............................................................. 61 Table 2.8. Household health expenditures, 2000–19 ....................................................................................................................................... 63 Table 2.9. Summary of policy recommendations.............................................................................................................................................. 72 Table 3.1. Design of the pension system in the Kyrgyz Republic ................................................................................................................. 75 Table 3.2. Number of insurance contribution payers at year-end ................................................................................................................. 76 Table 3.3. Number of pensioners in the Kyrgyz Republic at year-end ......................................................................................................... 78 Table 3.4. Average size of the monthly pension in the Kyrgyz Republic at year-end (soms)................................................................... 79 Table 3.5. Budget of the PAYG system in the Kyrgyz Republic ................................................................................................................... 82 Table 3.6. Indicators of the mandatory funded pillar in the Kyrgyz Republic ............................................................................................ 82 Table 3.7. Actuarial coefficients used to calculate the NDC pension in 2019............................................................................................. 91 Table 3.8. Summary of policy recommendations.............................................................................................................................................. 93 Acknowledgements This report is part of the Kyrgyz Republic Programmatic Public Expenditure Review (PER) led by Appolenia Mbowe (Senior Economist, EECM1), Bakyt Dubashov (Senior Economist, EECM1), and Ahya Ihsan (Senior Economist, EEAM2). Zhypara Azhykanova (Team Assistant, ECCKG), and Sarah Babirye (Program Assistant, EECM1) provided technical and administrative supports. The preparation of sectoral chapters was led by: • Education: Syedah Aroob Iqbal (ET Consultant, HECED) with inputs from Gulmira Sultanova (Education Specialist, HECED), and Ayesha Y. Vawda (Lead Education Specialist, HECED); • Health: Iryna Postolovska (Senior Economist, HMNHN) with inputs from Christel Vermeersch (Senior Economist, HECHN); • Pension: Oleksiy A. Sluchynskyy (Senior Economist, HECHN) with inputs from Sergiy Biletsky (Senior Social Protection Specialist, HECSP), Lidiia Tkachenko (Pension Policy Consultant, HECSP). The team received overall guidance from Sandeep Mahajan (Practice Manager, EEMC1) and substantive technical inputs and guidance from Ivailo Izvorski (Lead Economist, EEMC1). The team appreciated the overall leadership and support from Lilia Buruncuic (Country Director, ECCCA) and Naveed Naqvi (Country Manager, ECCKG). The team appreciated the advice and contributions provided by the peer reviewers including, Lars M. Sondergaard (Program Leader, HECDR), Huihui Wang (Senior Economist, HHNGE), Igor Kheyfets (Senior Economist, HMNED), Volkan Cetinkaya (Senior Economist, HECHN), and Mitchell Wiener (Senior Social Protection Specialist, HECSP). The team also benefited additional inputs from Susanna Hayrapetyan (Program Leader, HECDR). The team is grateful to the authorities for the excellent collaboration with the Ministries, Departments, and Agencies during the preparation of this report, which included the Ministry of Finance, Ministry of Economy, Ministry of Health, Ministry of Education and Science, National Statistical Committee, and Social Fund. EXECUTIVE SUMMARY The Kyrgyz Republic economy has been hit hard by the COVID-19 pandemic, putting at risk the development progress achieved in recent years. The steady economic growth shifted to a large contraction in 2020 due to the pandemic. Lower external demand and weaker commodity prices, combined with domestic restrictions on economic mobility and activity, led to a contraction of real GDP growth by 8.6 percent in 2020. Higher food prices and the pass-through from a weaker currency pushed inflation to 9.7 percent in December 2020, above the National Bank of the Kyrgyz Republic (NBKR) inflation target of 5 –7 percent. With the closure of borders, especially the border with China, imports are estimated to have declined by about 30 percent in US dollar terms. At the same time, exports are projected to have increased by about 1 percent, owing to surging gold sales abroad. As a result, the current account is likely to be balanced in 2020 compared with a deficit of 12 percent of GDP a year before. Poverty is projected to increase as households face both external and domestic shocks. The COVID-19 pandemic has added substantial pressures on the country’s fiscal space for investing in human capital. The authorities’ response to the pandemic has been timely and appropriate, supported by development partners. The government enacted limits on mobility and economic activity in May 2020 to curb the spread of the virus. The government prepared two packages to combat the crisis, amounting to 5.2 percent of GDP. Weaker economic activity, deferred tax payments, and increased COVID-19-related spending widened the fiscal deficit to an estimated 4.2 percent of GDP in 2020. After years of decline, public debt is estimated to have risen to 68 percent of GDP 2020, from 52 percent of GDP in 2019. Once the pandemic is brought under control, the authorities should reduce the fiscal deficit to less than 3 percent of GDP to limit the vulnerabilities related to the size of government debt. Improving the quality of public finances and public finance management processes has become more critical for supporting the recovery and the continued investment in human capital. The pandemic has eroded the Kyrgyz Republic’s recent gain in human capital development. In 2020, the Kyrgyz Human Capital Index (HCI) was 0.6, suggesting a child born in the Kyrgyz Republic today will be 60 percent as productive as she or he could be under the benchmark of a complete education with good health. This level represents a slight improvement relative to 2018 (0.59) but is lower than the average for the Europe and Central Asia region (0.69). The pandemic and the containment measures including school closure are likely to negatively affect the accumulation of skills due to loss of learning which will affect today’s students and the skills proficiency of the future workforce and adults. Learning-adjusted years of schooling (LAYS) are estimated to decline due to school closure.1 On top of the direct health impact of COVID-19, the pandemic may also have an additional adverse impact on health outcomes, including in part due to a higher likelihood of individuals delaying seeking care and the disruption of health services for non-communicable diseases (NDC) during the COVID-19 pandemic. Education and health are critical factors for improving human capital and sustaining inclusive economic growth and development. Human capital consists of the knowledge, skills, and health that people invest in and accumulate throughout their lives, enabling them to realize their potential as productive members of society. Investing in early education is critical for human capital as learning from early childhood is essential for later learning and success. Investing in health systems that ensure all people have access to quality and affordable health services is also key for building strong human capital. Investing in early childhood development, resulting in improved infant and child health outcomes, is crucial for maximizing human capital. 1 “COVID-19 and Human Capital, Europe and Central Asia Econo mic Update� (Fall 2019), Washington, DC: World Bank. 1 The Kyrgyz Republic has continued prioritizing education and social protection as its development priority. On average, education and social protection consistently represent a high proportion of the public costs, not just the highest spending sectors (22.9 percent and 18.2 percent of total expenditure) between 2013 and 2018. Meanwhile health spending averaged only 10.2 percent of total expenditure during the same period. Between 2017 and 2018, the health function suffered a fall of 1.2 percentage points in its share of state expenditures, while public services gained 2.0 percentage points. Though efficiency gains are important, inadequate spending on sectors such as health is also a source of inefficiency. This report, phase 2 of the Programmatic Public Expenditure Review (PER2), aims to assist the government of the Kyrgyz Republic in identifying key constraints to efficient and effective public spending and policy options for improvements in three areas: education, health, and pensions, which have become more pressing post-COVID-19. Public spending on education and health is essential for improving human capital outcomes and creating a sustainable foundation for stronger and more inclusive growth. Ensuring effective and sustainable pension system may help to safeguard human capital. Hence, ensuring adequacy, effective, and efficient spending in education, health, and pensions are critical to close Kyrgyz’s human capital gaps. To this end, PER12 identified critical areas to increase fiscal space for development priorities including investing in human capital including curtailing the growth of public wages while reducing tax expenditures and the large energy subsidies and quasi-fiscal deficits. Every som in tax expenditures supporting inefficient enterprises or loss-making state-owned enterprises means one som fewer for improving the quality of the country’s human capital and fulfilling its growth potential. Education The Kyrgyz Republic has introduced several major reforms to improve the education system. In 2012, the government introduced the mandatory national 480-hour-long “Nariste� school preparation program to achieve nearly universal coverage for 6- to 7-year-old children. To reduce inequity in education expenditure, per capita financing methodology was introduced in 2013 and has been implemented in all general education and primary vocational schools and currently is being piloted in preschools and secondary vocational schools. The development of the new competency-based curriculum and implementation of sample-based assessments of education quality lay the foundation required to develop a high-quality education system. There have also been notable improvements in the areas identified in the 2014 PER (see Annex 1) including empowering the MES to guide education planning, expanding preschool coverage using alternative low-cost approaches, introducing a universal preparatory year for all 6-year-old children, and implementing per-capita financing. The reforms have been complemented with substantial education spending. Kyrgyz Republic is one of the world’s biggest spenders on education. Education spending by the government in the Kyrgyz Republic amounts to 7 percent of GDP, or 22.5 percent of the total government budget, substantially higher than the average for lower middle-income countries. This substantial level of education spending is needed to ensure human capital development. The education sector reforms coupled with a consistent substantial level of spending have expanded access to education and laid the foundation for improving education quality. Particular progress was 2 Phase 1 of the Public Expenditure Review (PER1) focused on fiscal sustainability and the role of fiscal policy as a counter-cyclical tool, expenditure performance and budget institutions, tax policy and tax administration for more efficient and higher revenue collection, performance and governance of energy sector state-owned enterprises (SOEs), fiscal decentralization and subnational public finances, and the management of the public wage bill (World Bank. 2020. Kyrgyz Republic Public Expenditure Review: Creating fiscal space for inclusive growth). 2 seen in access to preschool, primary, and lower-secondary education, and improved equity of education financing across regions. Enrollment rates for primary and lower-secondary levels significantly increased, reaching nearly universal levels, from 92 percent in 1996 to 100 percent in 2018 and from 89 percent in 2004 to 97 percent in 2018, respectively. The preschool net enrollment rates have also increased, from 6 percent to 40 percent over the past two decades). However, learning outcomes remain low and access to preschool and upper-secondary education remains a challenge. Education outcomes, according to Program for International Student Assessment (PISA) test scores, are low compared with the Central Asia and Caucasus region. The quality of education has only modestly improved over time, and there is currently a large gap (4.2 years3) between years of schooling and years of quality-adjusted schooling. Of particular concern, 64 percent of children in the country who should be finishing primary school today are not proficient in reading comprehension. The ongoing COVID- 19 crisis is eroding some of the learning gains attained over the past decades and risks shrinking the financing available for education expenditures. Despite improvements, enrollment rates in preschool and upper- secondary remain substantially lower than universal levels. Further, there is considerable variation in enrollment in preschool and upper-secondary education and significant disparities in per-student expenditure on preschool education across regions. Several factors have contributed to the low learning outcomes. First, efficiencies across the education system are low. Spending on salaries and social contributions accounts for almost 70 percent of total education spending and has been increasing over time. However, effectiveness and efficiency of salaries expenditure is limited by suboptimal allocation of human resources in the system. The school feeding program accounts for the largest spending within goods and services, but it is not based on nutrient standards and is not targeted toward economically disadvantaged students. Similarly, expenditure on energy is substantially high due to energy-inefficient infrastructure. Insufficient allocation for education inputs leads to many schools lacking necessary infrastructure like classroom spaces and laboratories and experiencing shortages of education inputs to facilitate better learning outcomes. Second, the governance of the education system needs to be strengthened. The Ministry of Education and Science (MES) should focus on providing strategic guidance education through enhancing of the analytical work. Governance of the education system is adversely impacted by the lack of data collection and analytical capacity within the MES. This limits the MES’s ability to lead education policy, resulting in policy fragmentation. The education system also lacks regular quality assessment to collect information on student outcomes and student, teacher, and school characteristics to guide education policy. Third, teacher professional development is lacking in the education system. In-service trainings are primarily conducted through centralized courses, sometimes delivered by international organizations, and teachers do not have the opportunity to choose courses on their own. Clear horizontal and vertical career trajectories with transparent performance-based criteria for promotions is also lacking. To build on the progresses in educational attainment and to achieve the sector priorities, the government needs to focus on a results-based approach and evaluate reforms for their impact on education quality. The Education Development Strategy (EDS) 2021–2040 lays out the main government education priorities which focus on increasing efficiency in the use of education public spending and improving the quality of education. To improve the overall education outcomes, the authorities can consider measures through enhancing the allocative efficiency of education spending, investing to improve education quality including quality assessment at all levels of education, and strengthening management and planning of the education system. 3Human Capital Index (2020), World Bank. 3 Health The health outcomes in the Kyrgyz Republic have improved substantially over the past two decades, as the life expectancy increased with maternal and under-5 mortality decreasing. However, the evolving burden of non-communicable diseases (NCD) continues to exert significant pressure on healthcare provision. Although current health spending is slightly above average for the Kyrgyz Republic's income level, out-of-pocket (OOP) payments represent the majority of health expenditures in the country. There is a substantial gap between the government resources allocated to the health sector and the actual cost of delivering services outlined in the government health strategies and the law. That gap has increased in recent years. In 2017, the government spent 2.6 percent of GDP on health almost 1.8 percentage points lower than in 2012 and well below the 4.9 percent of GDP average in the Europe and Central Asia region. OOP spending is regressive, hurting households in the poorer quintiles that spend a higher percentage of the budget on health than those in wealthier quintiles. The health systems effectiveness needs to be improved, especially given the low and slowly declining healthcare utilization rate. The Kyrgyz Republic has the lowest outpatient care utilization rate in the region, with 3.5 outpatient contacts per person per year. Primary health care remains relatively weak and does not play a gatekeeping role. There are notable shortages of physicians and essential equipment and supplies, limiting the extent of services provided in primary care facilities. The COVID-19 pandemic has further exposed the country’s health system’s weaknesses and inadequate health infrastructure. Universal entitlement to the State Guarantee Benefits Package (SGBP) needs to be translated into an effective universal access to quality services. The SGBP accounts for 78 percent of government spending in the health sector, while the Additional Drug Package represents only 2 percent. To further improve the health system’s efficiency, there is a need to minimize hospitalization by strengthening primary care and revising payment methods. The cost-savings from avoided hospitalizations could be reinvested to implement cost-effective interventions in the health sector. In the fiscally constrained COVID-19 environment, a reprioritization of resources to the health sector is needed to achieve health outcomes gains while addressing inequities across socioeconomic groups to ensure universal health coverage. To further improve the health system’s efficiency, there is a need to avoid hospitalization using the cost savings to support cost-effective interventions to achieve the government’s broader development objectives. The prioritization of health in budget decisions in the Kyrgyz Republic is important given the low levels of domestic government financing, which means that there is currently a substantial gap between financing an essential package of quality services for the population and the available resources. Efficiency gains are essential, but a low level of public financing is also a source of inefficiency. Notwithstanding, the fiscal space can be obtained by increasing funding for health and/or by improving financing and expenditure efficiency to produce better outcomes without spending more (or, equivalently, to produce the same outcomes spending less). Increasing funding for health will have the desired impact on health outcomes if it is spent efficiently. Despite the low wages in the health sector (relative to other government sectors) and low number of physicians, personnel costs are the most significant expenditure item in public health expenditure, absorbing almost two-thirds of the total budget. However, the share of personnel expenditure cannot be reduced by freezing salaries, as these are still relatively low compared with other public sector wages in the 4 Kyrgyz Republic.4 Since 2017, average health sector salaries have stood at 63 percent of the average national salary. Meanwhile, public administration and education salaries have slightly increased relative to the average national salary during this period. Budget execution rates have been relatively high for the health sector (more than 94 percent), indicating that money allocated is effectively spent. However, this may be due to how the budget processes are implemented, including frequent amendments to revise the budget throughout the year that could ensure the achievement of certain budget execution rates. In 2018, the Kyrgyz Republic introduced program-based budgeting in the health sector. The 2020 program budget for health has eight programs (four programs in the Ministry of Health (MoH) and four programs in the Mandatory Health Insurance Fund (MHIF) budget). Since appropriations can only be made within ministries or government entities, the MoH and the MHIF have separate programs for their expenditure. The MHIF budget represents the bulk of government health spending and contains the following four programs: (1) planning, governance, and administration; (2) delivery of primary health care (PHC) services; (3) delivery of inpatient care; and (4) additional services outside the State Guaranteed Benefits Package (SGBP). The program-based budget fluctuates substantially over the three years, from 14.7 billion soms in 2021 to 17.2 billion soms in 2022. Program budgeting has yet to fully reflect the stated priorities of the government’s health program. The health sector priorities highlighted in the State Program for Health Development 2030 focusing on Public Health Protection and Health Care System Development for 2019–30 adopted the motto “Healthy Person—Prosperous Country� to emphasize the importance of investing in health to achieve economic development. In light of the health sector priorities, to further improve the overall health outcomes, the authorities can consider the following recommendations: (1) refine the scope and contents of the SGBP by ensuring that the benefits are well aligned with the country’s health needs and available financial resources; (2) award higher priority to the health sector to combat the rising trend in OOP spending by increasing government financing; (3) introduce new and increase existing health taxes (for example, taxes on tobacco and alcohol, sugar-sweetened beverages, and junk food) to reduce the burden of NCDs while increasing revenue; (4) improve the health system optimally by increasing priority to preventative and primary health services to avoid hospitalizations; and (5) develop the health system data to enhance the monitoring of quality access and for better use in decision-making. Pension Over the past five years, the pension system has remained financially stable and provided decent support for pensioners. However, parametric adjustments will be required to address growing demographic pressure and poorly calibrated parameters of the system. The pension spending is at around 8 percent of GDP, which is on the high end for countries with similar income per capita. However, the system is starting to come under pressure, as the number of contributors has not grown despite a substantial increase in the working-age population lately, largely because of growing informality. Due to the low participation rates of the current working-age population and prevailing informality, the coverage of the pension-age population is projected to decline from 85 to 63 percent over the next 30 years. This will result in considerably increasing 4 According to the PER1 (pages 141–42), the ratio of administrative management of staff in health organizations in Bishkek is about 4 percent, with a high percentage of about 96 percent of medical professionals than about 4 percent of administrative management personnel between 2016 and 2018. This could be referred to as “pro-service delivery staffing policy,� where support staff and administrative management personnel numbers are limited to only what is needed to assist frontline staff in delivering healthcare services. 5 expenditures on the programs alleviating poverty among the elderly to compensate for the falling efficiency of the pension program. Moreover, most of the contributions come from the public and organized formal sector employees, while those involved in the non-organized sector contribute very small amounts. The National Development Strategy (NDS) 2018-40 outlines the government objectives to improve the pension system. The government priorities include in particular: 1) reform the pension system to ensure its financial sustainability; 2) modernize the insurance component of the pension system to accommodate the needs of different age groups; and 3) streamline the early retirement provisions and increase the efficiency of the funded component of the pension system. In light of the pension reform priorities and this analysis’s attempts to further improve the overall pension outcomes, the authorities can consider the following recommendations: 1) consider further restrictions on early retirements and assess options for aligning the pension age for men and women, bringing it to 63 for both; (2) set a guaranteed minimum pension, for example, as a percent of the basic part of the pension; (3) simplify indexation of the basic part of the pension to make the system more transparent; (4) consider valorizing Notional Defined Contribution accounts by 100 percent of wage growth; (5) set the rules in the NDC system for compensation for the credit periods in the length of service (childcare time, mandatory military service, and so on); (6) assess options for indexing historic NDC balances retroactively; (7) relax the minimum account balance requirements for the regular periodic payments from the funded pillar to minimize lump-sum payouts; and (8) consider excluding farmers from the mandate of the NDC plan so their contributions can finance the basic part of pensions only (or have a separate DB plan just for that sector). The 2014 PER analysis pointed to several deficiencies in the national pension system and recommended certain parametric and structural reform options; some were reflected in the amendments introduced in the Law On the State Pension Social Insurance in 2017. Indeed, to address the issue of benefit adequacy, the 2014 analysis recommended: (1) introducing pension benefit indexation at the rate of 50 percent wages and 50 percent inflation; (2) valorizing NDC accounts, going forward, with the same rate of 50 percent wages and 50 percent inflation; and (3) retroactively valorizing all NDC accounts back to the date of the start of the reforms in 1996. The government adopted benefit indexation in (1), introduced valorization under (2) but with lower than recommended rate (inflation only), and did not address the legacy issue recommended in (3). Furthermore, the indexation of the basic part of pension remains complex and inadequate, which contributes to low benefits and ultimately to growing significant risks of poverty among the retirees. In this PER, the fiscal projections were updated using a new World Bank tool for pension policy microsimulations and the anonymous individual records, kindly provided by the Social Fund. Our simulations confirm a need for further parametric and structural changes in the benefit package to ensure that the system is socially sustainable. 6 Table ES.1. Summary of priority policy recommendations Priority Policy Recommendations Time Frame Expected Impact Education Improve the allocative efficiency of education spending Improve the allocative efficiency of human resources in the system by: Short term Improved student to teacher ratios • reducing the number of non-teaching and non-care workers in preschool and for preschool and primary (grades increasing the number of preschool teachers; 1-4) • shifting more teachers from secondary grades toward primary grades and towards teaching language and mathematics; and, • Ensuring that preservice training trains secondary school teachers in more than one subject, particularly in subjects facing teacher shortages, and sharing the current secondary subject teachers between schools where possible. Develop an effective program in collaboration with universities to train current secondary school teachers in more than one subject. Identify and stop reallocations from the education budget and increase spending on quality- Short term Adequate provision of learning enhancing education inputs including textbooks, methodological literature, teaching materials, materials for students laboratory equipment, library books, computers and internet access. Structure a comprehensive program for school and teacher-centered professional development with Short to medium term Improved professional regular classroom observations, peer-coaching and mentoring and robust career trajectory with vertical development for teachers and and horizontal opportunities for growth and allow teachers to use funds for professional development improved instruction in classrooms according to their training needs. Channel food expenditure through the Ministry of Health, develop nutritional standards for school Short to medium term Improved nutrition of students feeding programs and target public expenditure on school feeding programs towards students of from economically disadvantaged economically disadvantaged families. families Invest to improve education quality Prioritize universal coverage and shift from 10- and 24-hour preschools to 3- to 5-hour preschools; Short to medium term Improved access and equity in early prioritize public preschool funding for children ages 4-6 and from socioeconomically disadvantaged childhood education households, while enhancing service delivery through private provision. Ensure regular formative, summative and system - level quality assessments at all levels of education. Short to medium term Improved availability of data on education quality to orient education system towards quality Develop a forward-looking plan for infrastructure investments in education sector (with more efficient Medium to long term Increased time for instruction and and effective school designs) and reduce the percentage of schools with the three shifts to nil, halve the reduced class sizes and improved percentage of schools with three shifts to nil, and halve the percentage of schools with two shifts by efficiency of capital expenses adding classroom spaces to existing schools or constructing new schools. Strengthening management and planning of the education system Shift the role of the MES from the dominant education provider in the country to the provider of Medium to long term Evidence-based planning strategic guidance to local education departments and empower the MES with a strong analytical division of the education system that brings together data on school inputs, expenditures and results to inform policy. 7 Increase autonomy of local education departments and schools to reallocate funds across categories Medium to long term Increased capacity of local according to their unique needs and development priorities while holding them responsible for results. education departments and schools to invest in quality-enhancing inputs Health Increase existing health taxes (for example, tobacco and alcohol) and introduce new taxes (for example, Short term Positive effects on health and fiscal taxes on sugar sweetened beverages and junk food) to simultaneously reduce the burden of NCDs and savings increase revenue. Refine SGBP (State Guarantee Benefit Package) content and scope so that the benefits are aligned with Short to medium term Effective coverage of the SGBP the country’s health needs and available financial resources. Prioritize health within the government budget to maintain adequate levels of spending to respond to Medium term Stable increase in real per capita COVID-19, ensure continuity in service delivery, and combat the rising trend in OOP spending. financing for health Reduce the number of avoidable hospitalizations. Medium term Improved quality of primary healthcare service Invest in primary healthcare by prioritizing financing of PHC services. Medium term Enhanced efficiency of the health system Maintain the independence of the strategic purchaser; while improving its capacity and refining provider Medium term Improved efficiency of health payment methods. spending and quality care Develop data systems to monitor quality, access, and efficiency, and build capacity to use these data for Medium term Improved capacity to analyze the decision-making purposes. use of health system resources Pensions Revise the indexation of the basic part of the pension, so it reflects minimum income guarantee by the Short term Improved, predictable, and state, linked to macroeconomic indicators to make the system more predictable and transparent. transparent indexation of the basic part of pension Improve the NDC plan by avoiding gender differentiation in benefit calculation, setting the rules for Short term Improved and fair indexation of the compensation for the credit periods in the length of service (maternity leave, mandatory military service, insurance pension to preserve and so on) and assessing options of indexing historic NDC balances retroactively. benefit value over time Set a guaranteed minimum pension for people with an incomplete length of service, for Short term Improved adequacy of the pension example, as a percent of the basic part of the pension. benefit Consider an alternative proposal of switching to a modified DB formula, with benefit linked primarily Medium term Improved adequacy of the pension to service. benefit Consider further restrictions on early retirement and assess options for aligning the pensionable age for Medium term Reduced fiscal burden on the men and women. pension system Relax the minimum account balance requirements for the regular benefits from the funded pillar to avoid Medium term Improved adequacy of the pension unreasonable spending of the funds on one-off lump-sum payments. Assess options for better benefit coordination in payouts from the PAYG and funded pillars. 8 1. CHAPTER 1: EDUCATION The Kyrgyz Republic is one of the world’s biggest spenders on education and is leading several reforms to improve the education system. Education spending in the Kyrgyz Republic amounts to 7 percent of GDP, or 22.5 percent of the total government budget, substantially higher than the average for lower middle-income countries. Significant level of spending in the education sector has been accompanied by several reforms implemented in recent years. These have resulted in considerable improvement in access to education, particularly for pre-school, primary, and lower secondary education, and improved equity of education financing across regions. Additionally, government measures over the past decade, such as the development of the new competency-based curriculum and implementation of sample-based assessments of education quality, lay the foundation required to develop a high-quality education system. Despite the significant level of funding and several reform efforts, learning outcomes remain low. The quality of education has only modestly improved over time and there is currently a large gap (4.2 years) between years of schooling and years of quality- adjusted schooling. Learning poverty is also the highest in the Europe and Central Asia (ECA) region, with 64 percent of children in the country who should be finishing primary school today not proficient in reading comprehension (World Bank, 2019). The National Development Strategy (NDS) 2018-2040 rightly emphasizes “access to quality education� as the central focus for national development. To ensure quality education, the NDS 2018–40 and the Education Development Strategy (EDS) 2021–2040 lay out the main government priorities as: (1) increasing efficiency of the use of funds allocated to education while maintaining education as one of the priorities for public sector investment; (2) improving quality of education; and (3) strengthening management and planning of the education system. To achieve these priorities, the government requires a results-based approach and needs to evaluate all reforms for their impact on education quality. The need for the efficient use of current resources is exacerbated by the COVID-19 crisis, which is eroding the modest learning gains attained over recent decades and risks shrinking the financing available for education expenditures. In light of the sector priorities, this policy note reviews the Kyrgyz Republic’s public expenditure in education. The government has successfully implemented several recommendations highlighted in the 2014 Public Expenditure Review (PER), including empowering the Ministry of Education and Science (MES) to guide education spending, expanding preschool coverage using alternative low-cost approaches, and introducing a universal preparatory year for all 6-year-old children and implementing per capita financing. This PER provides direction for further improvements in the efficiency of education spending, quality of education and overall management of the sector. A. INTRODUCTION 1.1. This chapter assesses the efficiency and effectiveness of education spending in the Kyrgyz Republic to help the authorities improve access to and the equity and quality of education in the country. The chapter is organized as follows: Section B provides a brief overview of the education system. Section C focuses on the performance of the education sector followed by Section D, on governance and planning. Section E discusses the level, efficiency and effectiveness of the education expenditure. Finally, Section F presents recommendations and policy options that the authorities can consider improving education outcomes that will lead to enhanced human capital in the Kyrgyz Republic. 9 B. OVERVIEW OF THE EDUCATION SYSTEM 1.2. In 2019, the education system in the Kyrgyz Republic educated 1.82 million students from preschool to higher education. The number of students in the system increased by 40 percent from 1990 to 2019 (Figure 1.1Error! Reference source not found.). Additionally, the school-age population is expected to continue growing till 2045 and only start declining after 2070 (Figure 1.2). The education system therefore must prepare to cater to an increasing school-age population. Figure 1.1. Total student population in the country Figure 1.2. School-age population is expected to has increased by 40 percent from 1990 to 2019 increase till 2045 and start declining after 2070 Higher educational institutions 0-4 5-9 10-14 15-19 20-24 25-29 Vocational schools General education schools 5 000 2000 Preschool institutions 4 500 Number of students (Thousands) 4 000 Population (Thousands) 1500 3 500 3 000 1000 2 500 2 000 1 500 500 1 000 500 0 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2080 2085 2090 2095 2100 Source: National Statistical Committee (NSC). Source: World Population Prospects 2019. United Nations, Department of Economic and Social Affairs, Population Division. 1.3. Article 45 of the Constitution of the Kyrgyz Republic establishes the right to education for all citizens. Preschool education is provided through state and community kindergartens for children of ages 6 months to 7 years and through a mandatory 480-hour school preparation program for 6- and 7-year- old children. Primary (grades 1-4 starting at ages 6-7 years), basic general (grades 5-9), and secondary general education (grades 10-11) are compulsory, and the government provides free of charge education in state and municipal educational institutions. After the successful completion of basic general education, students receive a certificate issued by the state and continue their education in organizations of secondary general education or primary vocational education (compulsory grade 11). Graduates of secondary general education can continue to secondary vocational or higher professional education (Figure 1.3). State grants are provided to support primary and secondary vocational education and higher education in state educational institutions. Entry to higher education and the award of government grants are based on the results of the Republican Testing (RT)5 conducted by the independent Center for Educational Assessment and Testing (CEATM). 5 ORT consists of two parts – the main test of a projective nature, which determines the possibility of training an applicant at a university, and subject tests, built on the material of the school curriculum, which determine the level of knowledge of applicants. Currently graduates can choose from six subject tests: chemistry, biology, a foreign language, mathematics, history and physics. ORT is carried out once a year (May). To obtain the right to study at a university, in 2015 it was necessary to score 110 points in the main test and 60 points in the subject test. ORT certificates are issued simultaneously with obtaining a certificate of secondary education. 10 Figure 1.3. Education system in the Kyrgyz Republic Secondary Higher Education Vocational (Bachelors, Masters, Education PhD) Upper Secondary or Primary Out of School Organizations Secondary General Vocational (Grades 10 & 11) Education Adult Education Lower- secondary or Basic General (Grades 5 - 9) Primary or Initial General (Grades 1-4) Preschool Education 6 months to 7 6/7 to 10 11 to 15 years 16 to 17 years 18 & + years years 1.4. The Law on Education provides all children of preschool age a legal entitlement institution to preschool education; however, the state is facing challenges in meeting this obligation. The state guarantees financial and material support for preschool education and ensures the availability of places in preschool educational organizations for all segments of the population (Law on Education). Despite this legal obligation, not all children of preschool age attend preschool (Figure 1.4) and there is a considerable demand for the available preschool places. For example, in 2017, there were 126 children per 100 places in preschool institutions. Table 1.1. Education system in Kyrgyz Republic as of 2019 Level of Education Number of Institutes Number of students Number of educators Preschool education 1617 207884 8312 General education 2283 1311478 78,552 Primary vocational 99 29,684 3,403 education Secondary vocational 144 91,531 7,181 education Higher professional 55 183,778 12,160 education Source: NSC. 1.5. The number of hours of preschool education provided in the Kyrgyz Republic is higher than most European countries. Preschool education is provided in educational institutions offering stay for 3 hours, 10 hours, and 24 hours, with state preschools providing stay for 10 or 24 hours. This is considerably higher than in most European countries, which guarantee between 20 and 20 ECEC (Early Childhood Education and Care) hours a week (Eurydice, 2019; Table 1.2Error! Reference source not found.). 11 Figure 1.4. While 90 percent of 6-year-old children Table 1.2. The Kyrgyz Republic ensures a legal right are covered by preschool or general education, to preschool from a much earlier age as compared around half of 5-year-old and 70 percent of 4-year- with the EU average old children are currently not covered by the education system Country Kyrgyz Estonia Russia EU Preschool General education Republic average 100% Starting age Percentage of children enrolled 90% (in years) of 0.5 3 5.5 3.2 80% guaranteed 70% ECEC place 60% 50% 40% 30% 20% 10% 0% <3 3 4 5 6 Age Source: NSC (2017) Source: Eurydice 2019. 1.6. In general education, 72 percent of schools operate in two shifts, while another 3 percent operate in three shifts. The high percentage of schools operating in multiple shifts limits time and space available for instruction above the compulsory instruction times specified in the curriculum (Figure 1.5). The years of compulsory education (11 years) and the compulsory instruction time for general education in the Kyrgyz Republic is comparable to that in most Organization for Economic Co-operation and Development (OECD) countries. However, in most OECD countries, classroom instruction time accounts for only part of the time students spend learning. Students spend considerable time in after-school study activities and remedial “catch-up� programs (OECD, 2015). Figure 1.5. Weekly preschool hours, by type of guarantee, 2018/19 ⓧ No top-level regulations Δ No restrictions (full day) Note: * “Legal entitlement� to ECEC refers to a statutory duty of ECEC providers to secure publicly subsidized ECEC provision for all children living in a catchment area whose parents, regardless of their employment, socio-economic or family status, require a place for their child. Source: Eurydice (2019). 1.7. While the government remains the primary provider of education services, especially for preschool and general secondary education, the role of the private sector is gradually expanding. Private preschools have increased more than twofold from 2013 to 2017, though comprising about 10 12 percent of the total preschools in the Kyrgyz Republic and accounting for only 4.5 percent of the preschool students in the country. Similarly, private general secondary schools have increased by 75 percent during the same period, but still account for only 2 percent of the total general secondary education students. While there are no private primary vocational schools, the private sector is a bigger player in secondary vocational education and higher education. The private sector makes up a quarter of secondary vocational schools and a third of higher education institutes, accounting for about 14 percent of the students in secondary vocational education and higher education in the Kyrgyz Republic. C. PERFORMANCE OF THE EDUCATION SECTOR 1.8. Access to education has improved over time, but substantial challenges remain for preschool and upper secondary education. Enrollment rates for primary and lower-secondary levels significantly increased, reaching nearly universal levels from 92 percent in 1996 to 100 percent in 2018 and from 89 percent in 2004 to 97 percent in 2018, respectively (Figure 1.7 and Figure 1.8). The preschool net enrollment rates have also increased, from 6 percent to 40 percent over the past two decades). However, despite improvements, preschool enrollment rates remain substantially lower than universal levels. Across age groups, 11 percent of 6- year-olds and 49 percent of 5-year-olds were out of school in 2017. Preschool enrollment rates increase with age, with 27 percent of 3-year-olds and 31 percent of 4-year-olds enrolled in preschool in 2017. Further, enrollment rates in upper-secondary classes are also well below universal levels and have declined over time from 79 in 2004 to 72 percent in 2018, even though secondary education (or primary vocational education) is provided free of cost. 1.9. There is significant regional variation in preschool and upper-secondary attendance6 in the Kyrgyz Republic. The 2018 Multiple Indicators Cluster Survey (MICS) attendance rates data show that there is considerable variation in attendance in preschool and upper-secondary education. For example, while 63 percent of children ages 3–5 years attend early childhood education in Naryn, it is only 30 percent in Chui. Similarly, while 96 percent of upper-secondary age children in Naryn attend upper-secondary, it is only 78 percent in Osh city. 6 Multiple Indicators Cluster Survey (MICS) ask about each household member whether ‘at any time during the current school year did (household member’s name) attend school or any pre -primary education programme?’ The attendance rates reported are based on this and other attendance questions in the MICS survey. 13 Figure 1.6. Net enrollment rates in preschool Figure 1.7. Total net enrollment rates for primary education have improved consistently but fall education in the Kyrgyz Republic have increased behind comparator countries7 to universal levels Kyrgyzstan Georgia Kazakhstan Kyrgyzstan Georgia Kazakhstan Tajikistan Armenia Tajikistan Armenia 70 105 Net Enrollment Rate - Pre-primary Total Net Enrollment Rate - Primary 60 100 50 40 95 30 90 20 85 10 0 80 1995 2000 2005 2010 2015 2020 1990 2000 2010 2020 Figure 1.8. Total net enrollment rates for lower Figure 1.9. Total net enrollment rates for upper secondary have also increased to near universal secondary in the Kyrgyz Republic are much lower levels than universal levels Kyrgyzstan Georgia Kazakhstan Kyrgyzstan Georgia Kazakhstan Tajikistan Armenia Tajikistan Armenia 105 Total Net Enrollment Rate - Lower 100 120 Total Net Enrollment Rate - Upper 95 100 90 80 Secondary 85 Secondary 80 60 75 40 70 20 65 60 0 1990 2000 2010 2020 1990 1995 2000 2005 2010 2015 2020 Source: UNESCO Institute of Statistics (UIS). 1.10. Access to preschool and upper-secondary education is highly dependent on families’ socioeconomic status. Mothers’ education and family wealth are strongly related to attendance in preschool. While 62 percent of children of mothers with higher education attend preschool, only 31 percent of children of mothers with only basic secondary education attend preschool. Similarly, for secondary school, 97 percent of children of mothers with higher education attend upper-secondary school and only 59 percent of children of mothers with only basic secondary education do so. Attendance is also correlated to household’s income. Children from the poorest quintile are half as likely as children from the wealthiest quintile to attend early childhood education. Similar trends hold for upper-secondary education, where only 78 percent of the children in the bottom wealth quintile attend school compared with 96 percent of children from the top wealth quintile. Access to education for preschool, primary, and lower-secondary does not significantly vary across ethnicities; however, net attendance rates for upper-secondary are much lower for Uzbek children (60 percent) and other ethnicities (63 percent) as compared with children of Kyrgyz (93 7 Total net enrollment rate for preschool education not available from UNESCO Institute of Statistics. 14 percent) and Russian (89 percent) descent (MICS, 2018). 1.11. Despite its importance in promoting learning success, parental engagement in early childhood development is low. In the Kyrgyz Republic, the percentage of children ages 2–4 whose mothers engaged in four or more early childhood developmental activities is low (only 44.5 percent compared with 90 percent for Serbia and 84 percent for Montenegro). Only 21 percent of children 36–59 months in the Kyrgyz Republic live in households with three or more children’s books compared with Kazakhstan (51 percent), Montenegro (72 percent), and Serbia (57 percent). Figure 1.10. Percentage of children ages 2–4 with Figure 1.11. Percent of children younger than 5 living whom a father or mother or any adult has engaged in households with three or more children’s books in four or more activities that promote learning and school readiness during the past three days Adult Father Mother 80 70 100 90 60 80 50 70 60 40 50 30 40 30 20 20 10 10 0 0 Kyrgyz Kazakhstan Montenegro Serbia Kyrgyz Kazakhstan Montenegro Serbia Republic Republic Source: Multiple Indicators Cluster Survey (MICS) 2018. 1.12. Quality of early childhood education and care is low in the country, but attendance in preschool is linked with improved outcomes. Only 72 percent of children (36–59 months) are developmentally on track in health, learning, and psychosocial wellbeing (Figure 1.12).8 Also, only 14 percent are on track in literacy-numeracy.9 Attendance in early childhood education helps, with children attending early childhood education twice as likely to be developmentally on track in literacy-numeracy as children who are not attending. 8 MICS 2018 9 Children are identified as being developmentally on track based on whether they can identify/name at least 10 letters of the alphabet; whether they can read at least four simple, popular words; and whether they know the names of and recognize the symbols of all numbers from 1 to 10. If at least two of these are true, then a child is considered developmentally on track. 15 Figure 1.12. About 72 percent of children younger than 5 are developmentally on track in health, learning, and psychosocial wellbeing in the Kyrgyz Republic, which is low compared to other countries. (in percent) 100 ECDI Literacy-Numeracy age who are developmentally on track in Proportion of children under 5 years of health, learning and psychosocial well- 80 60 being (%) 40 20 0 Kyrgyzstan Mongolia Kosovo 2013 Republic of Kazakhstan Georgia 2018 Montenegro Serbia 2014 2018 2018 Moldova 2015 2018 2012 Source: MICS. Note: The current Early Childhood Development Index (ECDI) is composed of 10 questions, covering four domains of early childhood development: language/cognitive, physical, social-emotional, and approaches to learning. 1.13. Similarly, quality of education is low for students in primary and secondary education. Assessments in the early years of education in the Kyrgyz Republic give an idea of the education quality in the country. The country has participated in Early Grade Reading Assessments (EGRAs) each year from 2014 to 2017. In 2017, the first nationally representative EGRA was conducted and around 10 percent of the students in grade 2 and grade 4 were unable to respond to any of five comprehension questions asked after reading a short grade-level passage. Similarly, the 2018 MICS data show that 42 percent of children ages 7–14 are unable to demonstrate foundational skills in reading, and about half do not have foundational skills in numeracy.10 The Kyrgyz Republic also conducted National Sample Based Assessments (NSBAs) for grade 4 and grade 8 students in 2007, 2009, 2014, and 2017. According to 2017 NSBA, about 60 percent of the students in grade 4 perform below basic proficiency in both reading and mathematics. This is, however, an improvement as compared with 2014 NSBA, when 65 percent and 64 percent of students performed below basic proficiency in reading and mathematics, respectively. The quality of education is even more dismal in Uzbek primary schools, where 79 percent of the students performed below basic proficiency in their native language and 72 percent below basic proficiency in mathematics. 1.14. Low quality of education in the early years affects learning acquisition in later years, as students lack the foundational knowledge needed to develop more complex concepts. An adult skills survey implemented in 2019, adapted from the Programme for the International Assessment of Adult Competencies (PIAAC), shows that about 90 percent of adults surveyed performed below the minimum proficiency level (level 3 in PIAAC) compared with the OECD countries’ average of 54 percent. 10 United Nations Children’s Fund, Guidelines for Adapting the Foundational Learning Module to Non-Multiple Indicator Cluster Household Surveys, UNICEF, New York, August 2019. 16 D. GOVERNANCE AND PLANNING IN THE EDUCATION SECTOR 1.15. The education budget is centrally controlled and executed, except for the city of Bishkek. Before 2013, local governments controlled and executed a substantial portion of the education budget. This was reversed by Decree No.302, of May 30, 2013, which transferred funding for the costs of the educational organizations from the local budgets (except for Bishkek Institutions) to centralized bookkeeping offices at district and city departments of the MES. As a result, in 2017, 81 percent of the education budget was controlled and executed by the central authorities (Figure 1.13). The local governments are responsible for utilities and maintenance of buildings and premises. For preschool, out-of-school, and other educational organizations, local governments have somewhat greater responsibility and are responsible for financing expenses, except wages and deductions from the Social Fund of the Kyrgyz Republic. Local governments control a slightly larger percentage (31 percent) for preschool education. Figure 1.13. The education budget is centrally controlled and executed. 40 5 Local -- All local budgets 4 Local -- Rural community (aiyl okmotu) budget 3 Local -- City budget 30 Billion Soms 2 Local -- District (rayon) budget 1 Central 20 10 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: World Bank BOOST data. 1.16. Education planning is formalized and guided by the NDS 2008-2040, which places “access to quality education� at the center of the country’s development. Following the guidance of the NDS, the Education Development Strategy (EDS) 2021-40 lays out the sector goals till 2040, with clear indicators and targets. Three-year action plans starting from Action Plan 2021-2023 will lead the implementation of these goals. 1.17. Over the past decade, the education system in the Kyrgyz Republic has undergone several major reforms that have improved access to preschool and laid foundations for improving education quality. In 2012, the government introduced the mandatory national 480-hour-long “Nariste� school preparation program to achieve nearly universal coverage for 6- to 7-year-old children. To reduce inequity in education expenditure, per capita financing methodology was introduced in 2013 and has been implemented in all general education and primary vocational schools and currently is being piloted in preschools and secondary vocational schools. Additionally, a new competency-based curriculum has been approved for general education and assessments of education quality were implemented in the country with the support of international development organizations. 1.18. Governance of the education system is adversely impacted by limitations in human resources for data collection and analysis within the MES. According to the EDS 2021–40, the 17 Education Management Information System (EMIS) has been in operation for almost a decade but is only 60 percent effective as of 2020. In contrast, the National Statistical Committee (NSC) collects information directly from schools through forms filled out each year by educational organizations. However, this information is not available from the MES. The EMIS does not have data on school finance and student performance either though the ministry collects data on school finance and information on student performance available with CEATM. Additionally, the analytical division in the MES is not well-staffed and can benefit from expanding the human resources to undertake the analysis required to inform policy making. 1.19. The MES should focus on strengthening human resource capacity by using robust analytical tools to enhance education quality and outcomes. For example, three different tools assessing quality at the preschool level have been piloted through support of development partners (CLASS by the World Bank, Early Development Instrument (EDI) by the Global Partnership for Education (GPE), assessment tools piloted by UNICEF for program Zolotoi Kluchik). However, currently, a national standardized mechanism to assess education quality in preschool is not yet developed. Similarly, there is a need to regularize assessments of education quality at the school level and ensure budgetary funding for these assessments. The government will be allocating funds for the NSBA from the budget, which is a positive step and should be continued. 1.20. MES reforming efforts need to be paired with rigorous evaluations of reform impact on education outcomes to inform policy. Per capita financing has been implemented in all general education schools and is currently in the pilot stage in preschool and secondary vocational education institutes. However, the pilots are not accompanied with rigorous impact evaluations to inform policy. Similarly, a stimulating salary fund for the staff of educational organizations was adopted at the rate of 10 percent of the salary fund; however, the impact of these stimulating funds on teacher effort and student outcomes is not evaluated. 1.21. Regular quality assessments and increased school autonomy to direct funds to improve quality can spur quality improvements. Results from quality assessments provided to districts and schools with data on inputs and some autonomy over expenses have a huge potential to inform school development plans and improve the teaching and learning process. 1.22. Limited autonomy of schools in deciding financing priorities limit schools’ ability to more efficiently use available funds. The transition of education funds managing function to the MES in 2013 provided the MES greater clarity of financial flows of education to different regions, and, thereby, allowed the ministry to shape spending priorities for the overall development of the sector. However, regulating expenditure, through developing budgetary standards for expenditure on different categories, limits schools’ ability to reallocate expenditure across categories based on their unique needs. This also limits schools’ ability to plan and fund school development plans targeted to improve education quality. E. LEVEL, EFFICIENCY, AND EFFECTIVENESS OF EDUCATION EXPENDITURE11 11 Effectiveness and efficiency of public spending are common concepts used in evaluating the quality of public spending. Effectiveness relates to how input translates into output and outcomes (for example, the final objectives). It shows the success of the resources used in achieving the objectives set. Outcome is often influenced by multiple factors, including the quality of regulatory and institutional framework, economic development, and political choices. For example, how does education spending (input) affect learning outcomes or quality of labor force? Efficiency relates to the link between inputs and outputs. In other words, how much and what is the optimal mix of input to produce a certain amount of output? For example, how does education spending (input) affect access to education (number of teachers, classroom, or textbook)? (Mandl, Ulrike, and others 2008, “The Effectiveness and Efficiency of Public Spending.� European Commission). 18 Level of Education Spending 1.23. The Kyrgyz Republic is one of the biggest education spenders in the world, both measured as a share of total government expenditure and as a share of GDP. In 2017, the Kyrgyz Republic spent about 7 percent of GDP on education. As a share of total government expenditure, education spending accounted for 22.5 percent. This education expenditure is substantially higher than the average for lower middle-income countries, which on average spend 4–5 percent of GDP on education and 16 percent of total government expenditure. Overtime, real spending on education has also increased continuously. From 2013 to 2017, while the consumer price index decreased from 106.6 to 103.2, nominal education expenditure increased by 44 percent and remained stable as a share of overall government expenditure and as a share of GDP (Figure 1.14). 1.24. The Kyrgyz Republic spends Figure 1.14. Education expenditure in nominal terms more on preschool education and less on has continued to rise general and higher education on a per Total State Budget expenditure student basis than the average of Total State Budget expenditure as % of GDP countries in the lower-middle-income group. In 2017, the Kyrgyz Republic spent Total State Budget expenditure as % of total state budget expenditures on average 28,000 soms per student on 40000 30% preschool, 16,000 soms on general and 25% vocational education, and 35,000 soms on 30000 Million Soms 20% higher education. In terms of purchasing power parity (PPP), these expenditures 20000 15% amounted to $1,582, PPP $904, and PPP 10% 10000 $1,978 respectively. The per-student 5% expenditure on preschool education is 0 0% substantially higher than the average of the 11 2013 2014 2015 2016 2017 lower-middle-income countries in 2017 (PPP Source: NSC “Education and Science in the Kyrgyz Republic 2013 - $1,150).12 However, the per-student 2017,� Tables 5.1 and 5.2. expenditure on general school education and tertiary education is substantially lower than the average of lower-middle-income countries in 2017 (PPP $1,337 and PPP $2,112, respectively).13 1.25. The household OOP spending on education is small because general education is free for all citizens in state and municipal educational institutions and the government provides grants to students for vocational and higher education. Households spend primarily on tuition for preschool, vocational and higher education (Figure 1.15). On average, household outlays on education account for only 1.5 percent of total annual education spending in the Kyrgyz Republic (Kyrgyzstan Integrated Household Survey [KIHS] 2018). The country's average annual household expenditure in 2018 was 230,000 soms, while the average annual expenditure on education was only 4,200 soms. As 80 percent of the Kyrgyz Republic population live below twice the national poverty line, free provision of general education and 12 The eleven lower-middle income countries for which data are available from UNESCO Institute for Statistics (UIS) for 2017 are: Armenia, Cabo Verde, Côte d’Ivoire, Egypt, El Salvador, Guatemala, Jordan, Republic of Moldova, Mongolia, Myanmar, and Ukraine. 13 For per-student expenditure on primary and secondary school education, data are available for 13 lower-middle-income countries for 2017 from UIS: Armenia, Cabo Verde, Côte d’Ivoire, Egypt, El Salvador, Guatemala, Jordan, Republic of Moldova, Mongolia, Myanmar, Sri Lanka, Ukraine, and Zambia. For per-student expenditure on tertiary education, data are available for nine lower- middle-income countries for 2017 from UIS: Cabo Verde, El Salvador, Georgia, Jordan, Republic of Moldova, Mongolia, Pakistan, Sri Lanka, and Ukraine. 19 budget grants for vocational and higher education has been necessary to ensure access to education. 1.26. Though proportionately low, household OOP spending totals 6.56 billion soms (KIHS 2018), amounting to 17.5 percent of the overall public expenditure on education. Household OOP spending is therefore an important part of the overall education expenditure in the country. However, households lack information on the quality of different educational institutes needed to make informed spending decisions. 1.27. International donors provide Figure 1.15. Distribution of annual per-household substantial financing, primarily for expenditure on education by deciles of socio-economic preschool education and system level status improvements. In fact, the entirety of the education investment plan in NDS 2018–2040 relies on funds from international development organizations including the European Union, the Asian Development Bank, and the World Bank. The three organizations are providing in total $114.4 million for four projects over the life of the projects (four to six years) which make around $19 million per year, that is around 4 percent of the annual budget in the year 2017 (US $478 million). This reliance on funding from international organizations for system level improvements in the sector limits the MES’s ability to lead system-level reform. In fact, the Source: Author’s calculations based on Kyrgyz Integrated Household Strategies for Development of Education Survey (KIHS) 2018. (SDE) 2012–2020 includes as one of the indicators the percentage of financial support for educational development programs from the state budget versus donor participation and aims at improving the ratio from 20:80 percent in 2011 (baseline) to 70:30 percent in 2020. Efficiency of Public Education Spending 1.28. The increase in real education expenditure for preschools over time has not translated into a proportionate increase in enrollment rates and the per-student expenditures have increased. While both the budget and the number of students in preschool have increased over time, the preschool budget has increased by a much higher percentage. The nominal budget for preschool education doubled between 2013 and 2017, translating into a real increase of 77 percent (the price increased by 19 percent between 2013 and 2017). Over the same period, the number of preschool students in state institutions increased by 39 percent, resulting in an increase in real per-student expenditure in preschool education of 21 percent from 2013 to 2017 (Figure 1.16). 1.29. In fact, preschool expenditure in 2017 was enough to provide three hours of preschool education to all children ages 3 to 6, yet only covered 27.4 percent. According to the per capita financing framework for preschools, average annual per capita costs of three hours of preschool education is 8,032 soms for rural students and 6,433 soms for urban students. At an average rate of 7,490 soms per child (weighted by the urban-to-rural ratio), the current expenditures in preschool education in 2017 of 4,792.2 million soms (after accounting for the fact that per capita norms for preschool do not incorporate utility costs) could cover 607,827 students. In 2017, the number of 3- to 6-year-old children was 593,289 20 (NSC). 1.30. Real per-student expenditures in general secondary and vocational education have only marginally increased over time. For general secondary and vocational education, nominal budget increased by 41 percent between 2013 and 2017, translating into a 22 percent increase in constant prices. Over the same time period, number of students studying in general secondary and vocational education in state institutes increased by 17 percent, leading to a gradual increase in nominal per-student expenditure (Figure 1.16) and only a slight increase in real per-student expenditure. Given that the current per-student expenditure on school education is still low compared with the average for the countries in the lower- middle income group, this increase will be positive and should lead to quality improvements Error! Reference source not found.. 1.31. Higher education in the country went through a major reshuffle between 2013 and 2017; the government undertook a process of optimization of higher education institutes that closed five institutes and led to a steep decline in the number of higher education students. Between 2013 and 2017, the number of students in state higher education institutes dropped by 29 percent (from 196,232 to 138,695). Over the same time period, the budget for higher education increased by 55 percent (from 3.1 billion soms to 4.8 billion soms). The per-student expenditure on higher education therefore more than doubled (from 15,686 soms to 34,628 soms) between 2013 and 2017 (Figure 1.16). Between 2017 and 2019, the trend was reversing, and the number of students has risen by 13.9 percent but is still lower than the level in 2013. 1.32. A large percentage of education expenditure in 2017 (10 percent) was classified as “Education―not classified elsewhere� (Figure 1.17). Eighty-six percent of these unclassified funds go toward fixed assets, 6 percent toward salary, and 6 percent toward social security contributions/payments. The Public Expenditure and Financial Accountability (PEFA) Assessment for the Kyrgyz Republic in 2015 noted that though the classification systems for the budget as presented to Parliament and the Charter of Accounts (CoA) is fully aligned with the UN Classification of Functions of Governments (COFOG), the integrity of the classification system is compromised by the practice of attributing large expenditures to “other� sub-functions rather than to the specific sub-functions. 21 Figure 1.16. Per-student public expenditures have Figure 1.17. Composition of public expenditure by increased over time education levels in 2017 7096 Subsidiary services to 7097 Research and development -- Higher education Number of students - state education Education Secondary general and vocational education Students - State 4% 0% Preschool education Number of students - state 7098 7095 Education Education not Preschool education Per-student expenditure not definable by classified General and vocational education Per-student expenditure 7091 Pre- levels elsewhere school Higher education Per-student expenditure 1% 10% education Per-student Expenditure (Thousand Soms) 40 1800 14% Number of Students Thousands 35 1600 1400 7094 Tertiary 30 education 25 1200 13% 1000 20 800 15 600 7092 10 7093 Post- 400 Secondary secondary 5 education 200 education 56% 0 0 2% 2013 2014 2015 2016 2017 Source: NSC “Education and Science in the Kyrgyz Republic Source: World Bank BOOST data. 2013–2017.� 1.33. The introduction of per-capita financing for general education has reduced inequities in expenditure across regions although inequities in access to education persist. In 2017, the per-student expenditure in general education varied across regions from a high of around 16,700 soms in Batken Province to a low of around 14,000 soms in Bishkek compared with a national average of about 15,500 soms, excluding Naryn Province which has a much higher per-student expenditure of 25,752 soms per student due to a greater percentage of mountainous areas in the region (Figure 1.18). The expenditure is also progressive with the poorest region, Batken, receiving the highest expenditure per student (excluding Naryn). Though enrollment rates in general education increased for all regions in the country from 2013 to 2017, inequities in access persist. For example, enrollment rates in Osh oblast improved slightly, from 83 percent in 2013 to 86 percent in 2017 but remain well below the national average of 99 percent (Figure 1.19). Similarly, learning outcomes have modestly improved in all regions but wide regional gaps persist. Figure 1.18. Except for Naryn Province, there is Figure 1.19. Enrollment rates in general secondary low variation in expenditures per capita on education have increased for all students, general education inequities across regions persist Source: NSC, “Education and Science in the Kyrgyz Republic 2013–2017.� and World Bank BOOST data. 22 1.34. There are significant disparities in Figure 1.20. Per-student public expenditure for per-student expenditure on preschool preschool education across regions education across regions. Per-student expenditures in preschool have increased across all Bishkek oblast Chuy oblast Issyk Kul oblast Naryn oblast municipalities and range widely from a high of Talas oblast Osh oblast about 39,000 soms per student in Naryn to a low 40 Jalal-Abad oblast Batken oblast of about 22,000 soms in Batken oblast in 2017 (Figure 1.20). The expenditure is also not progressive with the poorest oblast, Batken, 30 Thousand Soms allocating the lowest funds for preschool per student. As of 2020, per-capita financing is only being piloted in 40 early childcare centers in the 20 country, but the EDS intends to implement per- capita financing for all early childcare centers by 2030. However, per-capita financing should be 10 2013 2014 2015 2016 2017 aimed not only at closing the differences in per- capita expenses across regions but also addressing the inequity in education outcomes of access and Source: NSC and World Bank BOOST data. quality across regions. 1.35. Salaries and social security contributions account for the largest share of the education budget. Their contributions constitute up to about 69 percent of the overall education budget. Capital expenditure and current expenditure, other than staff compensation, accounted for 11 and 20 percent of the public education budget, respectively. This composition is comparable to other developing countries, but developed countries allocate a higher percentage to non-staff current expenditures. The comparatively higher rate of capital expenditure in the Kyrgyz Republic aligns with the need for increased infrastructure investments in the country. Figure 1.21. Composition of Kyrgyz Republic budget across personnel, capital and non-staff current expenses is comparable to other developing countries 100 Percentage of total government 90 80 70 60 expenditure 50 40 30 20 10 0 All-staff compensation Capital Current expenditure other than staff-compensation Source: Edstats Data (2018 or latest year available), World Bank BOOST data, Kyrgyz Republic, 2017. 1.36. Staff compensation also accounts for the largest share across all education levels: 60 percent for preschool, 85 percent for general secondary education (primary/lower secondary/upper-secondary), 74 23 percent for primary and secondary vocational education, and 60 percent for tertiary education. Increases in salaries and social security contributions, due to increases in nominal salaries and stock of teacher, made up for almost the entire increase in education expenditure from 2011 to 2017 (Figure 1.22 - Figure 1.25). Figure 1.22. Expenditure by functional Figure 1.23. Expenditure by functional classification: Preschool classification: General secondary education (primary/lower secondary/upper secondary) 314 Land and other non- 6 20 314 Land and other non- produced assets 312 Inventory produced assets 18 312 Inventory 5 16 311 Fixed assets 311 Fixed assets Billion Soms 14 Billion Soms 4 282 Miscellaneous other 12 282 Miscellaneous other expenses 272 Social assistance expenses 10 272 Social assistance 3 benefits to population benefits to population 223 Utilities 8 271 Social security 6 benefits 2 222 Purchases of goods 223 Utilities and services - other 4 221 Use of goods and 222 Purchases of goods 1 services 2 and services - other 212 Social security 0 221 Use of goods and contributions/payments 2011 2012 2013 2014 2015 2016 2017 services 0 211 Salary 212 Social security 2011 2012 2013 2014 2015 2016 2017 contributions/payments Figure 1.24. Expenditure by functional Figure 1.25. Expenditure by functional classification: Primary and secondary vocational classification: Post-secondary and tertiary education education 1.4 311 Fixed assets 311 Fixed assets 6 1.2 282 Miscellaneous other expenses 5 282 Miscellaneous 272 Social assistance other expenses 1 Billion Soms benefits to population Billion Soms 4 272 Social assistance 271 Social security 0.8 benefits to population benefits 223 Utilities 3 271 Social security 0.6 benefits 222 Purchases of goods 2 0.4 and services - other 223 Utilities 221 Use of goods and 1 0.2 services 212 Social security 222 Purchases of 0 2… 2… 2… 2… 2… 2… 2… contributions/payments goods and services - 0 211 Salary other 2011 2012 2013 2014 2015 2016 2017 Source: World Bank BOOST data developed from Treasury data obtained from Ministry of Finance. 1.37. Despite increases over time, teachers’ salaries remain below the national average. Teachers’ salaries were increased three times over the past decade (2011, 2015, and 2019). From 1993 to 2011, the United Tariff Scale (UTS) was used in the Kyrgyz Republic for teachers’ salaries. The actual amount of payment depended on education, qualification category, and work experience. However, the starting salary for teachers was so low that even at subsequent qualification levels, salaries remained lower than the national survival wages (Ivanov 2017). In 2011, the UTS was abandoned, and a new payment system based on the hourly workload and education level of the teacher was introduced. Through this new system, the state was able to increase teachers’ salaries again in 2015 and 2019. Teachers’ salaries have therefore, on average, almost quadrupled, from 3,500 soms to 13,800 soms (EDS 2021 –40). However, salaries in the education sector are still well below the national average and other lucrative professions, for example, financial services and ICT (Figure 1.26). 24 1.38. While the number of teachers has risen overtime, there are imbalances in the allocation of teaching staff across education levels and student to teacher ratios are substantially higher in early grades of education. The number of teachers has risen across all education levels, reducing student- to-teacher ratios across all levels except for primary education (grades 1 to 4) (Figure 1.27). Despite increases over time, student-to-teacher ratios in preschool remain high (27:1; Figure 1.27), do not meet the legislative standards in the country,14 and are well above international norms.15 Similarly, student-to-teacher ratios for elementary school (grades 1-4) are high with an average of 28 students per teacher in 2017 compared with Kazakhstan (17), Armenia (15) and Georgia (9). According to the 2017 NSBA survey covering 203 schools, more than half of the schools faced a shortage of elementary school teachers. This is despite the fact that in 2019, 74 percent had fewer than 15 students per teacher. On the other hand, student- to-teacher ratios are low for secondary education and comparable with student to teacher ratios in developed countries (grades 5-11). Figure 1.26. Despite major increases in education Figure 1.27. Despite substantial increases in the salaries, they are fall far below the national average number of teachers, student to teacher ratios in early grades remain stubbornly high 40 80 40 35 Student - Teacher ratio Monthly Salary (Thousands) 30 Teachers (Thousands) 60 30 25 20 40 20 15 20 10 10 5 0 0 0 2013 2014 2015 2016 2017 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Secondary general education Teachers (5-11 (12) grades) - Total excluding school managers Information and communication Secondary general education Teachers (1-4 grades) - excluding school managers Financial and insurance activities Preschool education Number of educators - state Professional, scientific, and technical activities Preschool education Student - Teacher Ratio - state Public administration Secondary general education Student - Teacher ratio (grades 1 to 4) Education Human health and social work activities Secondary general education Student - Teacher ratio (grades 5 - 11) Source: NSC Source: NSC “Education and Science in the Kyrgyz Republic 2013–2017.� 1.39. Sub-optimal allocation and utilization of teaching staff results in severe shortages of language and math teachers. The education system has an overall excess of teachers with around 13,000 more teachers than required (using the legal norms of 25 students per class and a weekly load of 20 hours 14 Government Decree September 30, 1995 No. 404: for nursery groups up to 1-year old students, the maximum group size is 20; for nursery groups from 1 to 3 years: 25 students; for preschool groups from 3 years and older: 30 students. There needs to be 1.5 educators per nursery group and 1.5 educator + 1 assistant educator per preschool group. This implies that there should be one educator per 17 students in a nursery group and at least one educator per 20 students in a preschool group. 15 In the US, though states have their own requirements for pre-K (which is similar to preschool) group size and child-to-staff ratios, class sizes of no more than 20 children, and ratios of no more than 10:1 students per teacher are recommended by the American Public Health Association (APHA), the American Academy of Paediatrics (AAP), and the National Association for the Education of Young Children (NAEYC) for 3 and 4-year-olds (National Institute for Early Education Research 2004). In countries, such as Finland, Ireland, and the UK, the maximum number is below 10 children per staff member until the year before entering primary education. 25 of teaching). Despite this general excess, there is a severe shortage of language and mathematics teachers, while the excess is of subject teachers for secondary education (Figure 1.27). In 2017 (NSBA), 71.2 percent of schools had vacancies for one to two teachers, especially for primary grades and for specific core subjects such as mathematics, language and literature. Despite an overall excess of physics, chemistry, and geography teachers, individual schools faced a shortage of subject teachers. This indicates the need for better teacher resource management and to train teachers to teach more than one secondary school subject. 1.40. A large number of small schools providing comprehensive education (grades 1 to 9 or 11) limit the efficient use of the available teaching resources. In 2019, 138 schools, that is, 6 percent of total general education schools, catered to primary education (grades 1–4). All other schools covered grades 1 to grade 9 (274 schools, 12 percent) or till grade 11 (1,803 schools, 79 percent) (Table 1.5). This pattern of utilization of school infrastructure places high demand on teaching resources. According to the instruction times specified by the basic education curriculum, recruitment of a subject teacher becomes cost efficient only if a school operates two sections for secondary grades. For example, the total instruction time for biology for one set of secondary grades from grade 5 to 11 is 9 hours per week. Given that a teacher is expected to have a load of 20 hours of teaching a week, recruitment of a biology teacher is only efficient if the school operates two sections of each grade in secondary education. However, 34 percent of basic and comprehensive schools have up to 25 students per grade, that is, only enough students to operate one section. This further emphasizes the need to ensure that pre-service training institutions train secondary teachers in more than one subject. In the short run, this highlights the need to share subject teachers across secondary schools and design effective training program in collaboration with universities to train current secondary school teachers in more than one subject. Another alternative is to restructure utilization of current infrastructure to have a higher percentage of primary schools feeding into a smaller percentage of secondary education schools of larger sizes. Table 1.3. Usage of current school infrastructure Number of 1803 schools 274 138 Primary (grades Lower- Secondary Upper- Secondary 1- 4) (grades 5–9) (grades 10–11) 1.41. In addition, there is substantial share of workers (other than teachers, managers, methodologists, medical workers, psychologists, and cooks) with unspecified job descriptions, especially in preschool education. In 2019, other workers comprised 26 percent of the workforce for preschools. The nature of the recruitment and jobs for these “other workers� is not certain and needs deeper scrutiny, as they form the largest share of total workers in preschool education, second only to educators (27 percent). Similarly, nonteaching jobs make a substantial share of the total workforce in general and vocational education. Clear data on nonteaching staff in general and vocational education and their job descriptions are not available. However, from the number of total employees and teachers in general and vocational education, it is estimated that nonteaching staff made up around 25.9 percent of the total workforce in general and vocational education in 2019. 1.42. Expenditure on the use and purchase of goods and services constitutes the second largest portion (20 percent) of the total education budget. Goods and services make up 35 percent for preschool, 13 percent for general secondary education, 21.6 percent for primary and vocational education, and 34 percent for post-secondary and tertiary education. Food purchase and energy costs are the two 26 biggest expenditures under goods and services, comprising 32 percent and 25 percent, respectively, of the total expenditure on goods and services in the education sectors in 2017. 1.43. Within goods and services, the school feeding program accounts for the largest expenditure but the program is not based on nutrient standards to ensure adequate nutritional health for all children. For preschool, 26 percent of the total expenditure goes toward the food program (funded by local governments, with 59 percent of the expenditure financed using special funds). Early-life health and nutritional interventions, especially in the first 1,000 days of a child’s life, have significant impacts on schooling, earnings, and productivity over the lifecycle of an individual (World Bank 2018). The Kyrgyz Republic has shown a strong commitment to ensuring meals provision in all early childcare centers by including it as one of the obligations for preschool educational organizations in the Law on Preschool. However, improvements are required in the current system to ensure that the aims of the school feeding program are met. High-income countries, including Chile, Mexico, the UK, and the US, have introduced nutrient-based standards in school feeding programs to enhance the contribution of school meals to recommended dietary intake. Similarly, for vocational school, while 10 percent of expenditure goes toward the Kyrgyz Republic’s food program, there is no regulation that sets the standard for the quality of food provided. 1.44. Additionally, the school Figure 1.28. The system has an excess of teachers overall, feeding program is currently not but sub-optimal allocation of teachers across subjects and targeted toward economically schools result in shortages of teachers for specific disadvantaged students. General subjects, primarily language and math educational institutes provide food to all primary students (grades 1-4) and vocational educational institutes provide food to all students. According to per-capita financing norms in the country, the monetary standard for food per student per day for primary students in grades 1 to 4 is only 7.0 soms. This amount is too low to provide the recommended daily allowance of energy and key micronutrients (the World Health Organization [WHO] recommends that school feeding programs contribute 30-45 percent of the recommended daily allowance of energy and nutrients for half- day schools, and 60-75 percent for full-day Note: Author’s calculations based on school level data on students and teachers from NSC. Teacher excess is estimated using the legal schools [WHO 1998]). These funds can be norms in the country - average class sizes of 25 students per class and more effectively used by directing school that all teachers teach 20 hours a week. feeding programs toward students of economically disadvantaged families rather than covering all students. While schools ensure provision of food for all (to avoid any segregation of students based on socioeconomic status), systems can be set up to ensure that food costs are shared by parents when parents are socioeconomically advantaged and can afford the costs. In fact, such systems are in place in the Kyrgyz Republic for pre-school education where parents do contribute towards food costs. In vocational schools too, the entire expenditure on food purchases (10 percent of total expenditure on vocational schools) is funded through transfers from the central government. Vocational education organizations should be allowed to decide if the money from the central transfer for meals can be targeted more progressively (by making the priority providing for the socioeconomically disadvantaged) and with greater emphasis on results. 27 1.45. Similarly, within goods and services, expenditure on energy is substantially higher, amounting to 5 percent of the total government expenditure on education in 2017. This is substantially higher than some countries, like Poland and Bulgaria, which spend 3 percent of total government budget on energy, but lower than others, like Ukraine (Figure 1.29). The energy costs are high due to the fact that on average, schools in the Kyrgyz Republic are operating with suboptimal heating and lighting, and the normative energy need for schools is about 20 percent higher than the actual energy consumption.16 1.46. Energy costs for educational institutes are financed from local budgets. The per capita financing models for general education and preschools therefore do not include energy costs. However, the per capita financing methodology for primary vocational institutes includes per-student costs for energy by type of heating. The per-student cost for energy for primary vocational institutes was 3,233 soms in 2018, making 8–11 percent of the per-student expenditure for the year (depending on location and area of specialty). 1.47. Energy retrofitting of school buildings can lead to a 70 percent reduction in energy consumption and utility bills. The average specific heat energy consumption of schools averages 120– 140 kWh/m2 per year, which is roughly the equivalent of the energy efficiency classes “C�–“D� for energy used for heating purposes.17The implementation of energy efficiency measures will reduce energy consumption of buildings by an average of 70 percent compared with nominal energy consumption, according to regulatory benchmarks. Upon energy efficiency retrofit, school buildings are expected to comply with the level of energy performance at energy efficiency class B. Figure 1.30 provides the resulting maximum energy performance in different climatic regions. The total number of buildings eligible for optimal energy renovation is 1,531 (according to CAIAG database, http://schooldb.caiag.kg). The total costs required for proposed energy retrofitting of schools nationwide is approximately $279 million ($279,342,207).18 16 World Bank, 2019. 17 Energy class in accordance with energy efficiency scale of buildings’ energy certification, establis hed by Kyrgyz Government Regulation “On the rules of energy certification of buildings� from 2/08/2011, #531 18 World Bank, 2019. 28 Figure 1.29. The Kyrgyz Republic spends a Figure 1.30. Energy performance under energy considerable percentage of total government efficiency class B education budget on energy 600 G; 566 6 Percentage of total government 500 expenditure on education 5 E; 403 400 E; 366 G; 263 F; 295 F; 298 300 F; 244 D; 230 D; 221 C; 216 C; 195 3 3 200 D; 162 B; 103 B; 92 B; 84 B; 77 B; 73 100 B; 51 0 I II III IV V VI Specific total real consumption Specific total real consumption NORM) Bulgaria Poland Kyrgyzstan Ukraine Specific total energy demand after EE Source: BOOST databases, Bulgaria (2018), Poland (2014), Source: World Bank, 2019. Kyrgyz Republic (2017) and Ukraine (2018). 1.48. After employee compensation and goods and services, capital expenses account for the third largest expenditure category for education, constituting 11 percent of the total government expenditure on education in 2017. These capital expenses are predominantly financed from the central budget. A large majority of these capital expenses (66.6 percent) are incurred on acquisition and construction of public structures. Another 20 percent is spent on capital repair of structures. Though rules on energy certification (Government Regulation No. 531, dated February 8, 2011) exist, these rules are not binding under the building norms for public buildings and facilities. The new constructions and repairs therefore fail to ensure improved and energy-efficient designs. 1.49. There is a need to ensure sufficient budgetary funding towards a national infrastructure investment plan and to guide the efficiency and effectiveness of capital expense. Given that the student population is expected to increase until 2045, that 75 percent of general education schools currently operate more than one shift, and that the current infrastructure design is inefficient in energy utilization and educational instruction, the MES needs a forward-looking national infrastructure investment plan. This plan should allow the MES to identify needs for infrastructure investments, provide a transparent criterion for needs prioritization, and ensure that the capital expenses are used for efficient and effective infrastructure designs. 1.50. Schools face a shortage of necessary infrastructure like classroom spaces to deliver quality education. Though overall school sizes are moderate, with 64.4 percent of schools in the country educating fewer than 500 students in 2019, the schools lack classroom spaces to divide students into small groups. According to a survey of school directors and teachers in a nationally representative school sample (NSBA 2017), 70.4 percent of schools in a grade 4 sample had an average class size of more than 25 students per class. The large class sizes indicate the need for more class infrastructure within schools to allow smaller groups of students in a class. 1.51. In addition, schools experience shortages of education inputs to facilitate better learning outcomes. Expenditures on textbooks, manuals, and books decreased between 2013 and 2017, both in value and as a percentage of the overall government expenditure on education (from 125 million soms in 2013 to 105.5 million soms in 2017, and from 0.5 percent of overall government expenditure on education 29 in 2013 to 0.3 percent in 2017). According to the average cost of a set of textbooks by grade as specified in the methodology for determining the standards of budgetary financing of educational organizations,19 a total of 1.37 billion soms (in 2017 prices) was required to provide textbooks to all students of general education in 2017 as compared with 105.5 million soms spent on textbooks in 2017. The NSBA 2017 survey therefore observed severe shortages in the provision of textbooks and methodological material to schools. Around 38 percent of grade 8 teachers reported that textbooks were shared by two or more students (21 percent reported that textbooks were shared by three or more). Similarly, 57 percent of grade 4 teachers reported difficulty in finding the right methodological literature for their subject, and 19 percent of grade 8 teachers reported that their school did not have methodological literature on their subject (NSBA 2017). 1.52. Similarly, schools experience shortages in critical educational inputs like laboratories, computers, and the internet. In NSBA 2017, about half of the schools in the grade 8 sample reported that they did not have equipped chemistry and physics laboratories. Similarly, while the availability of computer labs and access to the internet has increased, 12.4 percent of schools did not have any computers connected to the internet and 54 percent of schools had only one computer with internet access in 2019. Effectiveness of Education Spending 1.53. Despite increases in education expenditure, learning outcomes improved only slightly. According to the National Sample Based Assessment (NSBA) 2017, there was a slight improvement in learning outcomes for grade 4, performing below basic reading and mathematics proficiency, which accounted for only 60 percent compared with 64 percent in literacy and 62 percent in numeracy in 2007. According to the EGRA, NSBA, and Adult Skills assessments, education quality remains very low. Results of the Adult Skills survey show that higher skilled groups of people in the Kyrgyz Republic earn higher wages, emphasizing the high returns on human capital development. However, the low quality of education from early childhood through tertiary education can lead to missed opportunities in lifelong incomes for the Kyrgyz population. The Kyrgyz Republic is not participating in international learning assessments currently, but it is aiming to participate in the Programme for International Student Assessment (PISA) in 2024. 1.54. The increased salaries in education have attracted individuals with higher qualifications to the system. The typical qualification profile (highest education attained) of teachers has improved from 2014 to 2019 for all education levels in the country. For general education, the percentage of teachers with higher professional education has increased from 86 percent in 2014 to 94 percent in 2019. Similar trends are seen at other levels, with the percentage of teachers with higher professional education increasing from 62 percent to 66 percent and from 96 percent to 98 percent from 2014 to 2019 in kindergartens and vocational education, respectively. 1.55. However, pedagogy is still one of the lowest rated specialties for higher education in the Kyrgyz Republic. In 2019, pedagogy and methodology for primary education was the fifth lowest rated specialty (out of about 70 specialties) for higher education positions funded by government grants and contracts with employers.20 The average score for applicants enrolled in the specialty was 128.4, and 124.2 for grants and contract funding, respectively, where 110 was the threshold score determined by the MES 19 Government Resolution No. 302, dated May 30, 2013. 20 Each year, students in the Kyrgyz Republic study on positions funded by government budget grants or by employers based on a tripartite agreement between a student, an educational organization and an employer. These positions are allocated based on students’ performance on ORT. 30 for all forms of study (grant, contract, full-time, part-time, and distance learning). This is, however, an improvement over 2014, when the average RT score for applicants enrolled in pedagogy was 115, and 108 for grant and contract places, respectively (the threshold was 110). 1.56. Currently, there is a lack of clarity of a career development path for young specialists entering the education sector. In-service trainings are conducted on an ad-hoc basis through courses delivered by international organizations, and about a third of teachers of elementary education and 40 percent of grade 8 did not have the opportunity to choose courses on their own (NSBA 2017). There is a need to professionalize the career, starting with a move from hourly remunerations to annual salary specifications, with a clear map of horizontal and vertical career development paths in the sector, whereby teachers can become mentors, subject coaches, methodologists, and directors. 1.57. Systematic information at the system level on the in-class teaching and learning process quality is not available. Student and teacher interactions within the classroom determine the quality of the teaching and learning process and thereby the quality of learning. However, there is no formal or informal standardized process of classroom observations in the country. According to a research study (LoCasale-Crouch, Arango, Sweeney, and Kerr, 2017) based on 250 preschool classroom observations using the Classroom Assessment Scoring System (CLASS: Pianta, LaParo, and Hamre 2008),21 however, teachers fared poorly in instructional support although they performed in the medium range in providing emotional support to students and in classroom organization. This weak performance in instructional support is likely due to large class sizes and consequently a large amount of time taken up in classroom organization. A better understanding of the current instructional practices used by teachers and challenges faced by them will allow the teacher training institute to develop targeted training content and support teachers better. F. RECOMMENDATIONS 1.58. The government of the Kyrgyz Republic has successfully implemented a number of recommendations highlighted in the 2014 PER, including empowering the MES to guide education spending, expanding preschool coverage through the use of alternative low-cost approaches, introducing a universal preparatory year for all 6-year-olds, and adapting per-capita financing (see Annex 1).22 This PER builds on these successes and provides direction for further improvements. 1.59. This review agrees with the NDS 2018–2040 that the government should focus on the efficient use of the funds allocated to education, while maintaining education as one of the priorities for public sector investment. The PER further identifies scope for efficiency improvements within the existing system and provides recommendations to ensure that investments in education improve education quality. The 21 The CLASS (Pianta and others 2008) assesses three broad domains of teacher-child interactions: Emotional Support, Classroom Organization, and Instructional Support. These domain scores are generated from observer ratings on 10 dimensions of practice (for example, Teacher Sensitivity, Behaviors Management, Quality of Feedback). For each 15-minute observational cycle, data collectors assign scores on a scale from 1 to 7 on each of these dimensions. Low-range scores (assigned a numerical value from 1-2) indicate that very few of the desired behaviors are observed; mid-range scores (3-5) reflect some of the behaviors observed, and high-range scores (6-7) indicate that behaviors are observed consistently among the majority of students. The dimension scores are then averaged to create domain-level scores reflecting teachers’ emotional support, classroom organization, and instructional support. 22 Resolution No. 302, dated May 30, 2013, introduced per capita financing in school education and all general education institutes receive financing using per capita methodology; Resolution No. 209-r, dated May 16, 2016 introduced pilot testing of normative financing for preschools; Resolution No. 545 dated Nov 22, 2018 introduced normative funding in state educational institutions of primary vocational education; Resolution No. 66-r, dated March 5, 2020, introduced pilot testing of the model of normative financing in the system of secondary vocational education. 31 key policy measures recommended by this report are summarized in the following table. Improve the allocative efficiency of education spending 1.60. Improve the allocative efficiency of human resources by reducing the number of non- teaching workers with unspecified job descriptions in preschool to nil, allocating more teachers toward primary grades, and teaching language, mathematics and training teachers in more than one subject, particularly those facing teacher shortages. Spending on salaries and social contributions accounts for almost 70 percent of total education spending and has been increasing overtime. However, effectiveness and efficiency of salaries expenditure is limited by suboptimal allocation of human resources in the system. In 2019, other workers (workers other than teachers, managers, methodologists, medical workers, psychologists, and cooks) comprised 26 percent of the workforce for preschools. The nature of job of these “other workers� is not clear and the number of workers with unspecified job descriptions needs to be reduced to nil. Similarly, while student-to-teacher ratios for primary education (grades 1–4) are high, student-to-teacher ratios in secondary and higher levels are quite low. Secondary school teachers need to be reallocated toward primary grades, trained in more than one subject and shared between schools where possible to effectively utilize the current human resources in the system. More information on workload of secondary school teachers will be needed for a successful reallocation. 1.61. Structure a comprehensive program for schools and teacher-centered professional development and allow teachers to use funds for professional development according to their training needs. MES has made significant investments in teacher training and teachers can choose courses according to their professional development needs. The impact of these centralized trainings can be significantly enhanced by structuring a system of school and teacher-centered professional development with regular classroom observations, peer feedback, and coaching and mentoring. This will ensure ongoing support to teachers tailored to improve classroom instruction. There is also a need to institute clear horizontal and vertical career trajectories with transparent performance-based criteria for promotions. According to the monetary norms for general education, 6990 soms (in 2013 prices) are allocated for trainings per teacher over a course of five years. The effectiveness of this expenditure can be increased by allowing teachers to use the money to pay for courses according to their development needs. 1.62. Identify and stop reallocations from the education budget and direct funding towards quality-enhancing education inputs especially textbooks, methodological literature, laboratory equipment, library books, computers and internet access, to ensure adequate provision of learning materials. In 2017, a large percentage of education expenditures (10 percent) was classified as “Education– not classified elsewhere� and a large portion (72 percent) of these unclassified expenditures was directed to the Poverty Reduction Fund. In 2017, reallocation from the education budget to the fund amounted to 2.6 billion soms. These reallocations need to be restrained and funds directed toward meeting shortfalls for critical educational inputs. While textbooks required a total of 1.37 billion soms in 2017 for general education, only 105.5 million soms were spent on the acquisition of textbooks, manuals and books the same year. Similarly, 214 million soms were required in 2017 to cover the cost of educational supplies (teaching materials, methodological literature, equipment for science laboratories, and so on), however, this requirement was also largely unmet. 1.63. Channel food expenditure through the Ministry of Health, develop nutritional standards for school feeding programs and target public expenditure on the feeding programs toward students of economically disadvantaged families. Channeling food expenditure through the MoH will allow school feeding programs to support improved nutrition. Though food costs account for 4.8 percent of the total education budget, there are no nutritional standards specifying energy and micronutrient requirements for school feeding programs. Additionally, the daily allowance of 7 soms per day for students 32 of primary grades is too low to ensure adequate nutrition. The effectiveness and efficiency of expenditure on food programs can be substantially increased by setting up a system whereby socio-economically advantaged families can contribute to food costs. This will allow public funds to be targeted to provision of food for students from economically disadvantaged backgrounds. Food costs in preschool are partially covered by parents and a similar model of differentiated payment based on economic background of families can be adapted to ensure that public money is directed to ensure adequate nutrition for those in need. Invest to improve quality 1.64. Prioritize universal coverage expansion of cost effective 3- to 5-hour preschools; prioritize public preschool funding for children ages 5 and 6 and from socioeconomically disadvantaged households while enhancing service delivery through private provision. Kyrgyz Republic has made impressive improvements in coverage of preschool education over the past decade. However, some gaps remain. Around 11 percent of 6-year-old children and 49 percent of 5-year-old children were not covered by preschool or school education in 2017 and the system is very overloaded with an average of 126 children per 100 seats in preschools and average group occupancies of 35 students. This is because the state preschools generally provide 10 and 24 hours of stay which exponentially increases the costs. This duration of stay is also much higher than any country in Europe. As a result, preschool education covered only 27 percent of 3 to 6-year-old children in 2017, though the preschool expenditure was large enough to provide 3 hours of preschool education to all 3 to 6-year-old children. Moving from 10 to 24 stay preschools to 3 to 5 hours of preschools will allow the country to reach universal coverage and focus on reducing group sizes, student to teacher ratios, and improving the quality of preschool education. Additionally, the government has instituted a system of electronic queuing for preschool places of children’s age and socioeconomic background should be included as criteria for enrollment in public preschools. 1.65. Ensure regular formative, summative and system - level quality assessments at all levels of education. The EDS 2040 rightly highlights regular quality assessments as a key priority, however, quality assessments in the Kyrgyz Republic are few and are not conducted regularly. The regular system level assessment at grade 4 and grade 8 through the NSBA is recommended through ensuring allocation of money through republican or donor funds toward the NSBA. The assessment is nationally owned and costs 17 million soms (around 0.02 percent of the overall education budget when conducted once in two years), and provides a wealth of information on student outcomes, student, teacher, and school characteristics to guide education policy. Similarly, the MES needs to standardize school leaving examinations of IGA-9 and IGA-11 to measure student competencies and inform education policies. Additionally, it is recommended to consolidate learnings from the assessment tools implemented in the country for preschool education and develop a national standardized tool for a quality assessment of all preschools. 1.66. Develop a forward-looking plan for infrastructure investments (with more efficient and effective school designs) in the education sector to reduce the percentage of schools with three shifts and two shifts by adding classroom spaces to existing schools or constructing new schools. The high rate of schools operating in double and triple shifts in the country limits the instruction time and space available. Additionally, high class sizes adversely impact the quality of education. Furthermore, the current stock of infrastructure is deteriorating and inefficient in energy utilization and quality instruction. It is recommended to conduct a detailed study on infrastructure requirements to develop a long term and forward-looking infrastructure investment plan. This plan should allow the MES to identify needs for infrastructure investments, provide a transparent criterion for needs prioritization and ensure that the capital expenses are used for efficient and effective infrastructure designs. Given the scale of this exercise, 33 the MES should prioritize donor funding towards infrastructure needs. Strengthen central and local governments’ capacity for evidence-based planning 1.67. Shift the role of the MES from the dominant education provider in the country to provider of strategic guidance for education development through analytical functions. The National Testing Center currently houses the EMIS and can be upgraded through the addition of technical staff members and affiliation with researchers at Kyrgyz universities to provide analysis and inform policy making at the MES. The EMIS needs to be completed, includes data on education quality, and EMIS data brought together with data on expenditure, school inputs and teachers from the NSC. This will ensure that education policies are guided by rigorous analysis of available data on the Kyrgyz education system. 1.68. Increase the autonomy of schools to reallocate funds across categories according to their unique needs and development priorities while holding them responsible for results. It is recommended that the MES continues to ensure that data collected by the analytical agencies are provided to local education departments and schools and can be used to guide spending decisions at the local level. Increasing autonomy of local education departments and schools on spending decisions while increasing their responsibility for student results will encourage schools to spend on quality-enhancing inputs according to their unique needs and development priorities. Table xx. Summary of priority policy recommendations Priority Policy Recommendations Time Frame Expected Impact Improve the allocative efficiency of education spending Improve the allocative efficiency of human Short term Improved student to teacher ratios for resources in the system by: preschool and primary (grades 1-4) • reducing the number of non-teaching staff in preschool and increasing the number of preschool teachers; • allocating more teachers from secondary grades toward primary grades and towards teaching language and mathematics; and • Ensuring that preservice training trains secondary school teachers in more than one subject, particularly in subjects facing teacher shortages, and sharing the current secondary subject teachers between schools where possible. Develop an effective program in collaboration with universities to train current secondary school teachers in more than one subject. Identify and stop reallocations from the education budget Short term Adequate provision of learning materials for and increase spending on quality-enhancing education students inputs including textbooks, methodological literature, teaching materials, laboratory equipment, library books, computers and internet access. Structure a comprehensive program for school and teacher- Short to medium Improved professional development for centered professional development with regular classroom term teachers and improved instruction in observations, peer-coaching and mentoring and robust classrooms career trajectory with vertical and horizontal opportunities for growth and allow teachers to use funds for professional 34 development according to their training needs. Channel food expenditure through the Ministry of Health, Short to medium Improved nutrition of students from develop nutritional standards for school feeding programs term economically disadvantaged families and target public expenditure on school feeding programs towards students of economically disadvantaged families. Invest to improve education quality Prioritize universal coverage through expansion of cost Short to medium Improved access and equity in early effective 3- to 5-hour preschools prioritize public preschool term childhood education funding for children ages 4-6 and from socioeconomically disadvantaged households while enhancing service delivery through private provision. Ensure regular formative, summative and system - level Short to medium Improved availability of data on education quality assessments at all levels of education. term quality to orient education system towards quality Develop a forward-looking plan for infrastructure Medium to long Increased time for instruction and reduced investments (with more efficient and effective school term class sizes and improved efficiency of capital designs) in education sector and reduce the percentage of expenses schools with three and shifts by adding classroom spaces to existing schools or constructing new schools. Strengthening management and planning of the education system Shift the role of the MES from the dominant education Medium to long Evidence-based planning of the provider in the country to the provider of strategic guidance term education system with enhanced analytical function that brings together data on school inputs, expenditures and results to inform policy. Increase autonomy of local education departments and Medium to long Increased capacity of local education schools to reallocate funds across categories according to term departments and schools to invest in their unique needs and development priorities while quality-enhancing inputs holding them responsible for results. 35 Annex 1. Progress of policy recommendations for education sector proposed in 2014 Public Expenditure Review (PER) Policy Options Progress Areas for Improvement Policy Area Short-term measures (1- Medium-term measures 3 years) (3-5 years) Strengthen system Train MES staff to gather Empower the MES to be The MES houses the Budgetary Policy, Financial The Budgetary Policy, Financial Analysis, management and and analyze relevant data the main management Analysis, and Investments Department, and and Investments Department is constrained oversight to monitor school authority tasked with collects information on school expenditures. The by lack of human resources and lack of spending practices and setting financing policies in World Bank Project “Sector Support for analytical expertise as well as the recommending steps to the education sector, Education Reform Project� supported school Monitoring, Strategic Planning and improve equity and including through the use report cards (SRCs). Out of 95 percent of schools Information Support Office. There is no efficiency in resource use of evidence and data on operating under the per-capita financing modality, department within the MES for analysis of which to base policy 78.8 percent post their annual school budget and education quality. As a result, budgetary decision. expenditure reports, as part of the digitized SRCs. policy and financial analysis is not adequately supported by all data and analytics to target policies and funds towards improving education outcomes of attainment and quality. Expand coverage of Expand pre-school Introduce a universal In 2012, the government introduced mandatory There are still gaps in coverage with 11 pre-primary education. coverage in underserved preparatory year for all 6- 480-hour long “Nariste� school preparation percent of 6-year old and 50 percent of 50 areas, including through year old, followed by program to achieve nearly universal coverage for percent of 5-year old out of school (nor in the use of alternative low- expanded preschool 6 to 7-year-old children. The government is also preschool or school). Additionally, wide cost approaches such as coverage for 5-year old in expanding pre-school coverage in underserved socioeconomic gaps exist in preschool half-day programs. underserved areas. areas through half day, community-based coverage. kindergartens, primarily through GPE and WB projects. As a result of these initiatives, preschool enrollment has continued to increase over time. Invest in quality Adopt modern teaching Reform teacher Limited progress has been made on this front. There is considerable room for enhancing inputs. and learning practices and recruitment and pay Teacher salaries were increased thrice over the last improvement. The new system of Teacher ensure adequate provision practices with the aim of decade and the percentage of teachers with Attestation maintains the earlier three levels of learning materials in all attracting top graduates professional higher education has increased over of qualifications rather than providing schools. into the teaching time. attractive vertical and horizontal career profession. trajectories. The new system also assesses Additionally, in 2016, with the support of the teachers only once every five years rather Asian Development Bank, the working team set than incorporating yearly assessments and up by the MES developed a regulation on regular meetings with mentors and “Teacher Attestation,� which was approved by the managers. Additionally, the associated government in June 2017. The new system was incentives are not structured as a part of piloted in 2016 in 30 innovative schools; however, salary but as a separate allowance. Similarly, the pilot was halted until 2020 and has yet to be despite improvement, pedagogy is still one 36 implemented. of the lowest rated specialties for higher education. Investment on learning materials has not increased. Reduce spending on Reduce outlays on food Reduce energy costs by Detailed study on energy efficiency of the The school feeding program is still managed non-education items. provision from the improving relevant school current school infrastructure has been by the Ministry of Education and Science. It education budget by infrastructure and conducted under the World Bank’s accounts for the largest expenditure within converting it into a social modernizing school “Enhancing Resilience in the Kyrgyzstan good and services, but it is not based on assistance program for heating practices. Project�. nutrient standards and not targeted to needy students. economically disadvantaged students. Spending on energy remain significant, amounting to 5 percent of the total education public spending (2017) due to school infrastructure that is energy inefficient. Streamline school level Set and monitor Implement a streamlined Some progress has been made on this front. Utility costs are not included in the per procurement practices. compliance with national public procurement Monetary standards for school feeding programs capita financing framework. guidelines for unit costs framework in the education exist and per capita food costs are included in per (such as price caps for sector. capital financing formula in the country. Law on food and energy) to be public procurement has been developed and followed by schools and approved by the government in 2015. local authorities. Adopt modern Implement PCF Continuously monitor and Progress has been made on this front. Per capita Currently, per capita financing formula does education financing nationwide in general improve the PCF financing has been implemented for general not account for the unique school and mechanism. secondary education; mechanism, eventually school education in the country and is being student needs for example, infrastructure or abolish outdated staffing expanding it to pre-primary implemented for other levels of education input improvements required for schools or norms, especially for non- and vocational education including preschool and vocational education. schools serving students from teaching staff. tiers. socioeconomically disadvantaged backgrounds. Additionally, there is no evaluation of the impact of per-capita financing on education outcomes of attainment and quality. 37 References Learning Poverty: Historical data and sub-components | Data Catalog. (2019). Retrieved January 8, 2021, from https://datacatalog.worldbank.org/dataset/learning-poverty. Education at a Glance 2015: OECD Indicators. (n.d.). [Text]. Retrieved January 26, 2021, from https://www.oecd- ilibrary.org/education/education-at-a-glance-2015_eag-2015-en. United Nations, Department of Economic and Social Affairs, Population Division (2019). World Population Prospects 2019, custom data acquired via website. Law of the Kyrgyz Republic “On Education,� approved April 30, 2003, #92 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/1216?cl=ru-ru). Law of the Kyrgyz Republic “On Preschool Education,� approved June 29, 2009, #198 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/202646/20?cl=ru-ru). Law of the Kyrgyz Republic “On the Organization of Meals for Students in Educational Institutions,� approved December 4, 2019, #135 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/111982?cl=ru-ru). Law of the Kyrgyz Republic “On the Licensing and Permitting System in the Kyrgyz Republic,� approved October 19, 2013, #195 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/205058). National Development Strategy of the Kyrgyz Republic for 2018–2040, approved by the Presidential Decree of the Kyrgyz Republic, October 31, 2018, #221 (source: http://www.president.kg/sys/media/download/52135/). “On the Maintenance of Children in Preschool Institutions of the Kyrgyz Republic,� approved by the Presidential Decree of the Kyrgyz Republic, February 12, 1992, #UP-48 (source: http://cbd.minjust.gov.kg/act/view/ru- ru/47879). Regulation “On the Organization of Meals for Students in State and Municipal Secondary Schools of the Kyrgyz Republic,� approved by the Decision of the Government of the Kyrgyz Republic, September 18, 2006, #673 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/57833). Estimated Monetary Nutritional Standards in Social Institutions, approved by the Decision of the Government of the Kyrgyz Republic, January 15, 2008, #7 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/58839?cl=ru-ru). Decision of the Government of the Kyrgyz Republic “On the State and Development of the Network of Preschool and Out-of-School Educational Organizations,� approved July 28, 2009, #478 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/90145). Decision of the Government of the Kyrgyz Republic “On the Conditions of Remuneration for Certain Categories of Workers in the Education System,� approved September 30, 2019, #511 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/157168). Decision of the Government of the Kyrgyz Republic “On the Establishment of a Two-Tier Structure of Higher Professional Education in the Kyrgyz Republic,� approved August 23, 2011, #496 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/92802). State Educational Standard of the Kyrgyz Republic “Preschool Education and Childcare,� approved by the Decision of the Government of the Kyrgyz Republic, June 29, 2020, #363 (source: http://continent- online.com/Document/?doc_id=34474784#pos=0;0). 38 Strategic Directions of Development of the Education System in the Kyrgyz Republic, approved by the Decision of the Government of the Kyrgyz Republic, March 23, 2012, #201 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/92984). Concept for the Development of Education in the Kyrgyz Republic until 2020, approved by the Decision of the Government of the Kyrgyz Republic, March 23, 2012, #201 (source: http://cbd.minjust.gov.kg/act/view/ru- ru/92984). State Educational Standard of Secondary School Education of the Kyrgyz Republic, approved by the Decision of the Government of the Kyrgyz Republic, July 21, 2014, #403 (source: https://edu.gov.kg/ru/schools/gosudarstvennyj-obrazovatelnyj-standart/). Strategy for the Development of Education in the Kyrgyz Republic for 2012–2020, approved by the Decision of the Government of the Kyrgyz Republic, March 23, 2012, #201 (source: http://cbd.minjust.gov.kg/act/view/ru- ru/92984). Sanitary and Epidemiological Rules and Standards, “Sanitary and Epidemiological Requirements for the Device, Content and Organization of the Operating Mode in Preschool Educational Organizations,� approved by the Decision of the Government of the Kyrgyz Republic, April 11, 2016, #201 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/11947); Mechanism for Financing the Training of Students of Educational Institutions of Secondary and Higher Vocational Education, approved by the Decision of the Government of the Kyrgyz Republic, May 28, 2012, #331 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/93855). Instruction on the Procedure for Calculating the Wages of Employees of Educational Organizations, approved by the Decision of the Government of the Kyrgyz Republic, May 31, 2011, #270 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/92689?cl=ru-ru). Temporary Regulation on the Procedure for Licensing Educational Activities in the Kyrgyz Republic, approved by the Decision of the Government of the Kyrgyz Republic, July 23, 2018, #334 (source: http://cbd.minjust.gov.kg/act/view/ru-ru/12222). Concept of Optimization of the Network of Higher Educational Institutions of the Kyrgyz Republic for 2014–2015, lost force in accordance with the Decision of the Government of the Kyrgyz Republic, June 19, 2017, #386 (http://cbd.minjust.gov.kg/act/view/ru-ru/96541). Education Development Strategy in the Kyrgyz Republic for 2021–2040, Draft of the Decision of the Government of the Kyrgyz Republic (source: https://www.gov.kg/ru/npa/s/2639). Draft of the Action Plan for 2021–2023 on the Implementation of the Education Development Strategy in the Kyrgyz Republic for 2021–2040 (source: https://www.gov.kg/ru/npa/s/2639). Teacher Support and Professional Development Program of the Kyrgyz Republic “New Teacher,� approved by the order of the Ministry of Education and Science of the Kyrgyz Republic June 20, 2017, #785/1 (source: https://edu.gov.kg/media/files/39615a6e-b8d5-41e8-b8ff-fa07bc709a9d.pdf). Building Standards of the Kyrgyz Republic “Public Buildings and Facilities,� approved by the order of the State Agency for Architecture, Construction, Housing and Utilities under the Government of the Kyrgyz Republic December 13, 2018, #20-npa (source: http://cbd.minjust.gov.kg/act/view/ru-ru/200296). 39 2. CHAPTER 2: HEALTH The Kyrgyz Republic’s health outcomes have substantially improved over the past two decades: life expectancy increased while maternal and under-5 mortality decreased. The evolving burden of non-communicable diseases (NCDs), however, has already started to exert significant pressures on the health system. Although current health spending is slightly above average for the Kyrgyz Republic's income level, out-of-pocket (OOP) payments represent the majority of health expenditures in the country. There is a substantial gap between the government resources allocated to the health sector and the actual cost of delivering services outlined in the Program of the Kyrgyz Republic Government on Public Health Protection and Health Care System Development for 2019– 2030, “Healthy Person—Prosperous Country.� and the State Guaranteed Benefits Package (SGBP). That gap has been increasing in recent years. In 2017, the government spent 2.6 percent of GDP on health–almost 1.8 percentage points lower than in 2012 and well below the 4.9 percent of GDP average in the Europe and Central Asia (ECA) region. OOP payments represent a significant share of health spending in the Kyrgyz Republic. They are regressive, hurting households in the poorer quintiles that spend a higher percentage of the budget on health than those in wealthier quintiles. The health system's effectiveness needs to be improved, especially given the low and slowly declining healthcare utilization rate. The Kyrgyz Republic has the lowest outpatient care utilization rate in the region, with 3.5 outpatient contacts per person per year. Primary health care (PHC) remains relatively weak and does not play a gatekeeping role. There are notable shortages of physicians and essential equipment and supplies, limiting the extent of services provided in primary care facilities. The COVID-19 pandemic has further exposed the country's health system's weaknesses and inadequate health infrastructure. Universal entitlement to the SGBP needs to be translated into an effective universal access to quality services. The SGBP accounts for 78 percent of government spending in the health sector, while the Additional Drug Package (ADP) represents only 2 percent. To further improve the health system's efficiency, there is a need to minimize hospitalization by strengthening primary care and revising payment methods. The cost savings from avoided hospitalizations could be reinvested to implement cost-effective interventions in the health sector. In the fiscally constrained COVID-19 environment, a reprioritization of resources to the health sector is needed to achieve health outcomes gains while addressing inequities across socioeconomic groups in order to ensure universal health coverage. To further improve the overall health outcomes, the authorities can consider the following recommendations: (1) refine the scope and contents of the SGBP by ensuring that the benefits are well aligned with the country’s health needs and available financial resources; (2) award higher priority to health sector to combat the rising trend in OOP spending by increasing government financing; (3) increase existing health taxes (for example, taxes on tobacco and alcohol) and introduce new taxes (for example, taxes on sugar-sweetened beverages and junk food) to simultaneously reduce the burden of NCDs and increase revenue; (4) improve the healthcare system optimally by increasing priority to preventative and primary health services to avoid hospitalizations; and (5) develop the health system data to enhance the monitoring of quality access and for better use in decision-making. A. INTRODUCTION 2.1. The Program of the Kyrgyz Republic Government on Public Health Protection and Health Care System Development for 2019-2030 “Healthy Person – Prosperous Country� sets out an ambitious agenda for the health system to improve health outcomes, reduce health inequities, and provide financial protection. Among the objectives of the program are an increase in life expectancy to 75 years, a 30 percent reduction in premature deaths from four major NCDs (Sustainable Development Goals target 3.4), and a reduction in the consumption of tobacco, alcohol, salt, and sugars by 2030. The government aims to achieve these goals by strengthening people-centered systems that ensure public health and deliver quality services throughout the entire life cycle. The program highlights the importance of a modernized, high 40 quality, accessible primary health care system to manage common conditions and diseases. 2.2. This chapter provides an overview of the health sector and healthcare utilization dynamics in the country, evaluates current spending trends, and assesses the extent to which current policies promote the achievement of the government’s health system goals. The chapter's primary objective is to analyze health spending efficiency and equity and determine the health system's policy implications. The chapter examines the evolution of health spending in the Kyrgyz Republic since 2014, when the last PER was undertaken. The analysis in this chapter identifies the challenges preventing the country from managing and improving: (1) health outcomes, including NCDs; (2) reprioritization of health within the government budget to combat the rising trend in OOP spending; and (3) reforms aimed at translating the SGBP into universal access to quality services while strengthening primary health care. 2.3. The health chapter is structured as follows. Section A presents an overview of the health outcomes over the past decades. Section B provides an overview of the health sector, with Section C focusing on the health services delivery system. Section D addresses challenges related to health services utilization, particularly outpatient care. Section E centers on health financing related to government spending on health and the role of the SGBP and Mandatory Health Insurance Fund (MHIF). Section F focuses on payment systems, followed by Section G, on the COVID-19 response and the macroeconomic impact on health financing. Section H is on the recommendations informed by the analysis of the health sector. B. OVERVIEW OF THE HEALTH SECTOR 2.4. To boost human capital—a central driver of sustainable growth and poverty reduction— the Kyrgyz Republic has made substantial progress on improving health outcomes over the past two decades. Between 1990 and 2018, life expectancy in the Kyrgyz Republic increased from 68.3 to 71.2 years (Figure 2.1), while under-5 mortality declined from 66 to 20 deaths per 1,000 live births (Figure 2.2). Some improvements have also been observed in maternal health, with maternal mortality declining from 79 to 60 deaths per 100,000 live births during the same period. The Kyrgyz Republic performs better on all three indicators than other comparable lower middle-income countries. Most former Soviet countries, however, perform better on these measures than other low and middle-income countries due to the Soviet legacy and access to health services. The maternal mortality ratio, however, is much higher than in other ECA region comparators – with maternal mortality ratios of 19 deaths per 100,000 births in Moldova and Ukraine and 17 deaths per 100,000 births in the Russian Federation (Figure 2.3). There is still substantial room to improve and the need to reorient the system, particularly considering the growing NCD burden. 41 Figure 2.1. Life expectancy at birth 1990-2017 2017 75 90 80 Life expectancy at birth 70 Turkey Estonia Georgia Belarus UkraineArmenia Kazakhstan (Years) Kyrgyz Republic Russian Federation Uzbekistan 70 Tajikistan Moldova 65 60 60 00 02 04 06 08 10 12 14 16 20 20 20 20 20 20 20 20 20 50 250 500 1,005 2,000 3,995 7,000 12,735 25,000 50,000 100,000 Kyrgyz Republic Georgia Russia Tajikistan Uzbekistan Belarus Moldova Ukraine GNI per capita, US$ Figure 2.2. Under-5 mortality rate per 1,000 live births 1990-2017 2017 Figure 2.3. Maternal mortality ratio per 100,000 live births 2000-2017 2017 Source: World Bank Development Indicators (WDI) (2020). 42 2.5. Significant improvements in child health Table 2.1. Vaccination coverage, 2000-18 outcomes, due in part to high vaccination coverage, have been observed. Coverage of BCG, HepB3 DPT, Pol3, and 2000 2010 2015 2017 2018 measles was above 92 percent in 2018 (Table 2.1) (WDI 2020). BCG 96 98 97 97 97 Nutritional outcomes, however, are poor, and almost 12 percent MCV3 96 98 96 96 96 of children under 5 in the Kyrgyz Republic are stunted PCV3 88 92 (National Statistical Committee [NSC] of the Kyrgyz Republic HepB3 44 96 97 92 92 and UNICEF 2019). The Kyrgyz Republic has the third highest DPT3 99 96 97 92 94 stunting rate in ECA after Azerbaijan (17.8 percent) and Source: WDI (2020). Tajikistan (17.5 percent). 2.6. Despite overall progress, inequities persist across sex, geographic areas, and socioeconomic determinants. For example, life expectancy is more than eight years higher among women than men (75.8 years for women and 67.6 years for men in 2019), and the gap has narrowed only slightly since 1990 (Figure 2.4). Meanwhile, the under-5 mortality rate is more than 60 percent higher among the poorest quintile than among the richest quintile of the population (at 13 and 8 deaths per 1,000 live births respectively) and almost twice as high among the middle 60 percent of the population (25.3 deaths per 1,000, live births) (Figure 2.5). The highest rates of stunting are among rural children (13.1 percent) and children in the poorest households (14.0 percent). The lowest rates have been observed among children in the richest households (9.2 percent) and in urban areas (8.8 percent) (NSC and UNICEF 2019). Figure 2.4. Life expectancy at birth by sex, Kyrgyz Figure 2.5. Under-5 mortality rate per 1,000 live Republic, 1990-2019 births and prevalence of stunting by wealth quintile, 2018 80 Males Females Poorest 75.8 40 75 72.6 Life expectancy at birth (years) 30 70.4 20 70 67.6 Richest Second 10 64.2 65 61.4 0 60 55 Fourth Middle 50 2010 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2012 2014 2016 2018 Under 5 mortality rate Stunting Source: NSC (2020). Source: MICS (2018). 2.7. Despite the recent improvement in maternal and child health outcomes, the Kyrgyz Republic is facing a growing burden of non-communicable diseases (NCDs) as the main public health risks. In 2017, NCDs accounted for 83 percent of all deaths in the Kyrgyz Republic, lower than the proportion of deaths attributable to NCDs in ECA (89 percent), but substantially higher than the average for lower middle-income countries (62 percent). One in four Kyrgyz citizens are at risk of dying prematurely (between the ages of 30 and 70 years) from one of the four main NCDs (cardiovascular disease, chronic obstructive pulmonary disease, diabetes, and cancer). The probability is higher among men than women — 33.4 percent among men compared with 17.3 percent among women in 2016 (Figure 2.6). Given the current rate of decline, the Kyrgyz Republic is not expected to achieve the Government Health Program target of a 43 30 percent reduction in premature mortality from four main NCDs by 2030 (SDG target 3.4). Figure 2.6. Probability of dying between the ages of 30 and 70 from one of the four main NCDs, 2016 (percent) Source: WDI and World Health Organization (WHO) (2020). 2.8. Lifestyle factors and risky behaviors are the main risk factors in the Kyrgyz Republic. The prevalence of smoking is significantly higher (Figure 2.7) than in countries with similar levels of income and is as high as the rates observed in Belarus (26.7 percent), Georgia (28.8 percent), and Ukraine (28.9 percent). Figure 2.7. Smoking prevalence, 2016 (percent of population ages 15+) Source: WDI (2020). 44 2.9. The prevalence of NCDs and Figure 2.8. Percentage of hypertensive persons taking premature deaths from NCDs could be prescribed blood pressure medication regularly, prevented and reduced with effective 2007-15 preventive actions and policies. Regular monitoring of blood pressure at the primary care 40% 35.1% Percentage of hypertensive persons who took ALL or SOME prescribed blood pressure level is critical to reducing the risk of acute 32.2% medication in last 24 hours (%) myocardial infarction by early diagnosis and 30% helping patients to effectively manage 23.6% 23.5% hypertension. Effective monitoring and 20% 17.4% management of risk factors, such as hypertension 13.4% and diabetes, is critical to prevent costly long-term 10% complications and premature deaths, yet the management of NCD patients in the Kyrgyz Republic remains weak and adherence to 0% Rural Urban Rural Urban Rural Urban medication is low. In 2015, almost two-thirds of those prescribed medication did not take it within 2007 2011 2015 the past 24 hours (Figure 2.8). This translates into Source: KIHS (2015) analyzed by World Health Organization poor management of hypertension, with only 48 (WHO). percent of those prescribed antihypertensive drugs having their blood pressure under control in 2019 (MHIF 2020). Moreover, the government has made limited use of fiscal policy to address the growing NCD epidemic, and the use of health taxes has not been widespread. Global experience suggests that health taxes can lower the consumption of harmful products (for example, tobacco, alcohol, sugar-sweetened beverages, and junk food), thereby reducing the prevalence of NCDs and relieving the health system (World Bank 2017; World Bank 2020c). C. OVERVIEW OF THE HEALTH SYSTEM 2.10. The Kyrgyz Republic introduced significant changes in the organization and financing of the health system since gaining independence in 1991. The country inherited the extensive Semashko healthcare system structure of the former Soviet Union. The Semashko system was organized around the guiding principle of universal access to healthcare free at the point of use. Care was focused on in-patient treatment, and, consequently, primary care was weak. There was an emphasis on the continuous expansion of staff and facilities and an extensive system of parallel health services, which were attached to large industrial enterprises and certain ministries. 2.11. The current organizational structure of the health system is a result of three major health reform programs: Manas (1996–2005), Manas Taalimi (2006–11), and Den-Sooluk (2012-16). The health financing reform consisted of the establishment of the Mandatory Health Insurance Fund (MHIF) as a single purchaser of health services and the introduction of capitation payments for primary care and case-based payments (diagnosis-related groups, DRGs) for hospitals. These reforms introduced changes to the health system with the aim of strengthening PHC and restructuring the hospital sector. 2.12. The MHIF manages its funds through a single pooling system, which consolidates the financial resources for health care from the government budget for the exempt categories of the population and mandatory contributions, as well as from copayments from the population. Health financing from local 45 budgets (of oblasts and rayon) has been discontinued, and currently all money is pooled at the MHIF.23 A large pool enables the MHIF to have greater market power for purchasing services, lowers administrative costs (compared with multiple smaller pools), and ensures more extensive and equitable risk sharing among various population segments (for example, sick and healthy, and rich and poor). 2.13. As of 2018, formal workers contribute 2 percent of their salary for mandatory health insurance. Informal workers can enroll in the mandatory health insurance, making a flat rate contribution of 550 soms. Insurance coverage has increased slightly over the past few years, rising from 62.4 percent in 2016 to 65.5 percent in 2018 (Figure 2.9). Regional differences, however, persist, with insurance coverage ranging from 56.0 percent in Chuy oblast to 73.7 percent in Naryn oblast in 2017 (MHIF 2019). Figure 2.9. Share of insured population 2016-18 80% 73.7% Share of population covered by insurance 75% 72.5% 70% 64.8% 65.50% 65% 62.4% 60% 56.0% (%) 55% 49.8% 50% 45% 40% 2016 2017 2018 Bishkek Chuy Osh Batken Dzhalal-Abad Issyk-Kul Naryn Talas Kyrgyz Republic Source: Mandatory Health Insurance Fund (MHIF) annual reports. 2.14. All citizens are entitled to health services under the MHIF-administered SGBP. The SGBP guarantees free basic health services at the primary care level24 and emergency services for the whole population, as well as subsidized inpatient care for the insured and exempt categories. The government pays the contributions for the exempt categories, including 30 groups for which benefits are granted based on their social status and 16 groups for which benefits are granted based on their underlying medical conditions.25 Beginning in 2016, the SGBP also included some costly specialized services (oncology, hematology, cardiac surgery, and psychiatry). In addition, the insured and exempt populations have access to the Additional Drugs Package (ADP), which provides coverage for outpatient medicines. The MHIF is not able to perform selective contracting and contracts all public facilities for the provision of services under the SGBP. 23 Before 2006, each oblast determined its health budget based on its priorities and capacity to generate revenues. This resulted in large differences in financing across oblasts for the same type of services. Bishkek City was the last to transfer its sources to the MHIF in 2018. 24 Individuals must enrol with a family group practice to receive primary care services. Enrolment is not linked to the place of residence, and patients have the freedom to choose their primary care provider. 25 The list of eligible categories is revised occasionally, often expanding the benefits. For example, the Resolution of the Government of the Kyrgyz Republic of September 7, 2018, No. 420, “On amendments to certain resolutions of the government of the Kyrgyz Republic in the field of health care and health insurance,� extended the categories entitled to free medical care at the outpatient level to include the following categories of insured people: (1) unemployed population officially registered in the public employment service; (2) students of secondary and higher educational institutions of full-time education until the age of 21; and (3) military personnel (Nguyen and Strizrep 2020). 46 2.15. Individuals must register with a family medicine provider and receive a referral to access inpatient and specialized outpatient care, which are subject to a copayment. Copayments are defined as flat lump-sum payments made upon admission. The level of copayment varies by insurance status and by exemption status. The uninsured are required to pay a higher copayment at the hospital level, providing incentives for enrollment in the mandatory health insurance program. 2.16. Despite early successes, the reform agenda remains largely unfinished and universal entitlement to the SGBP does not translate into effective universal access to quality service that contributes to improving population health outcomes and ensuring financial protection. The SGBP has remained a largely implicit package: it defines general conditions for health care provision for each level and type of care, a list of laboratory tests, and categories of citizens entitled to copayment reduction and exemption (Strizrep and Nguyen 2019). The absence of an explicit positive list of entitlements and covered services creates confusion for the patients about their entitlements and results in implicit rationing of services. Previously, SGBP revisions had to be approved annually as part of the budget processes, taking into account expected revenues and past budget utilization.26 Beginning in 2018, the SGBP has been revised on an “as needed� basis, but the government does not have clear criteria or guidelines for the revision process. Such changes are being planned under the government’s new program, the State Program for Health Development 2030, which is being supported by the World Bank’s PHC Quality Improvement Program. Defining a benefits package is a continuous process, requiring regular revisions to ensure alignment of the package with the changing burden of disease, demographics, technology developments, cost-effectiveness of treatments, and health systems constraints. 2.17. Weaknesses in budget formulation and the lack of processes for SGBP revisions have contributed to a substantial gap between the costs of the SGBP and the available resources. Earlier estimates suggest that this gap is as large as one-third of the cost of hospital care and almost two-thirds of the needs for the outpatient drug benefit (Kaliev and others 2012). Historical spending levels have served as the basis for budget formulations, without major adjustments to the future costs of the SGBP. Moreover, changes in SGBP entitlements have not been accompanied by adequate changes in the budget provision and prices for services. To ensure effective coverage of the SGBP, there is a need to refine its content and scope so that the benefits are aligned with the country’s health needs and available financial resources. 2.18. The Ministry of Health (MoH) has the responsibility to oversee health facilities and to organize the provision of services. Most of the facilities in the country are publicly owned autonomous entities. The provider-purchaser split helps to increase accountability and establish a system of checks and balances. At the same time, due to limited capacity of the MoH, the MHIF also plays an active role, interacting frequently with the facilities to ensure they follow financial guidelines. Importantly, the MoH, as the owner, does not readily have access to the financial documents of the facilities (unlike the MHIF). The MoH continues to monitor provider resources using input-based norms, in some cases leading to misaligned incentives for the providers. 2.19. The availability and use of data to inform clinical management and policy decisions remains limited. Most facilities still use paper-based clinical forms, and available e-data (that is, clinical information form data) are rarely used to inform clinical or management decisions (Ahmedov and others 2020). 26 The Government of the Kyrgyz Republic approved the latest version of the SGBP on November 20, 2015 (Order No. 790), with some amendments since then. 47 2.20. Despite the progress that has been achieved so far, the health system's effectiveness and efficiency remain a challenge. To address this challenge, in 2018, the government prepared a new health sector vision establishing the priorities for the coming years. The State Program for Health Development 2030 of the Kyrgyz Republic Government on Public Health Protection and Health Care System Development for 2019–30 adopted the motto “Healthy Person—Prosperous Country� to emphasize the importance of investing in health to achieve economic development. The program identified key priority areas for the government, including improving PHC and public health, optimizing hospital and ambulance services, and strengthening the building blocks of the health system. D. HEALTH INPUTS 2.21. The health services delivery system consists of three levels of care. PHC services are delivered at feldsher-obstetrical ambulatory points (FAPs), family group practices (FGPs), and family medicine centers (FMCs). FAPs are in rural areas and are usually staffed with at least a feldsher (paramedic health workers), and those in larger villages also have a midwife and a nurse. FAPs deliver basic health services, including prenatal and postnatal care, immunization, and health promotion and education. FGPs are the main providers of PHC services in the Kyrgyz Republic and usually have three to five doctors. FMCs typically employ 10–20 doctors and, in addition to PHC services, also provide specialized outpatient services, including diagnostics and minor surgeries. Usually, there is at least one FMC per rayon, responsible for all FGPs and FAPs located in the catchment area. Oblast hospitals and territorial hospitals provide secondary care. General Practice Centers (GPCs) were formed by merging territorial hospitals and PHC facilities in remote areas with a population of less than 25,000 to provide both primary and secondary care. Tertiary care is provided by republican facilities (national hospitals and centers) and specialized dispensaries, providing highly specialized outpatient and inpatient services. Besides, some ministries continue to have their health facilities (for example, the Ministry of Internal Affairs, Ministry of Justice, Ministry of Defense, and Ministry of Emergency Situations). Private provision of health services remains limited and is mostly concentrated in large cities, focusing on diagnostics and laboratory services. 2.22. Substantial efforts were Table 2.2. Number of health care facilities in the Kyrgyz made to consolidate hospital Republic, 2000–18 infrastructure and to optimize the 2000 2010 2018 delivery network. The country went Hospitals 321 184 187 through an important phase of hospital Clinics 621 163 146 restructuring in 2001–02. The hospital Family medicine centers - 67 64 system was restructured, and excessive In which a group of family 800 592 582 capacities inherited from the Semashko physicians system were reduced, with many small Independent group family 84 19 17 hospitals transformed into PHC facilities. practices and legal entities Since 2000, the number of hospitals has Stations of ambulance services 115 108 141 been almost halved, falling from 321 Children's homes 4 4 3 Source: National Statistical Committee (2020) hospitals in 2000 to 187 hospitals in 2018 (Table 2.2). 2.23. As a result of this restructuring, the number of hospitals and the number of beds were reduced substantially. Currently, the Kyrgyz Republic has 3.9 beds per 1,000 population (Figure 2.10). The bed-to-population ratio is below the WHO Europe regional average of 5.5 beds per 1,000 population and the 48 OECD average of 4.7 beds per 1,000 population. At the same time, there continues to be some variation in the bed ratio across oblasts, ranging from 3.9 in Osh oblast to 2.0 beds per 1,000 population in Bishkek City (Figure 2.11). Figure 2.10. Hospital beds per 1,000 population 12 10.83 Beds per 1,000 population 10 8.2 8 7.46 5.5 5.7 6 4.7 5.0 4.0 3.9 4 2.7 2.9 2 0 Source: WHO Health for All Database (2020). Note: Latest available year for all countries. Data for the Kyrgyz Republic are for 2018. Figure 2.11. Hospital beds per 1,000 population by oblast, 2010–18 5 Beds per 1,000 population 4 3 2 1 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 Batken oblast Jalal-Abat oblast Yssyk-Kul oblast Osh oblast Talas oblast Chui oblast Bishkek city Osh city Source: e-Health Center (2020) 2.24. The hospital sector, however, still lacks basic inputs. Surveys conducted in 2018 by the World Bank suggest shortages in essential equipment and supplies in hospitals. For example, in the Naryn region, only 20 percent of maternity hospitals reported availability of sterile scissors or blades; clamps and sterile material for bandaging umbilical cords as well as oxytocic injections were available in 60 percent of surveyed hospitals. In Chuy oblast, among the five hospitals with cardiology departments, only one had an echocardiography machine—critical for delivering cardiologic care (Ahmedov and others 2020). 49 2.25. Unlike in several other former Figure 2.12. Physicians per 1,000 population, 1990–2017 Soviet countries, the total number of physicians per population in the Kyrgyz Republic has steadily declined since 1990, falling from 3.4 in 1990 to 1.9 physicians per 1,000 people in 2018 (Figure 2.12). This is significantly below the WHO Europe average of 3.2 physicians per 1,000 population (Figure 2.13) and is lower than the ratio found in Uzbekistan (2.5), Moldova (2.9), and Ukraine (3.0). Although the absolute number of physicians has increased slightly, the increase in the production of doctors has not been commensurate with population growth. Source: WHO (2020) Figure 2.13. Physicians per 1,000 population, latest year available 6.0 Physicians per 1,000 population 4.9 5.0 4.1 4.0 3.3 3.3 3.0 3.2 2.9 3.0 2.5 1.7 1.8 1.9 2.0 1.0 0.0 Source: WHO (2020) 2.26. While the number of physicians has declined, the number of nurses and midwives has steadily increased over the past several years (Figure 2.14). As a result, the nurse-to-physician ratio has increased from a low of 2.4 nurses per physician in 2008–09 to 3.1 nurses per physician in 2018. The nurse- to-physician ratio is significantly higher than the regional (2.6 in Central Asia and 2.4 in Europe and Central Asia) and income group (1.8 in lower middle-income countries and 2.6 in upper middle-income countries) averages (Figure 2.15). Given the relatively low physician-to-population ratio, this provides opportunities to implement some task-shifting, with nurses taking on activities previously conducted by physicians (particularly in PHC settings). 50 Figure 2.14. Nurses and physicians per 1,000, Kyrgyz Republic, 2004–14 Physicians Nurses/midwives Nurse to physician ratio 7 2.8 2.9 6 2.8 2.7 Nurse to physician ratio Per 1,000 population 5 2.6 2.7 2.6 2.6 4 2.5 2.6 2.5 3 2.5 2.5 2.4 2.4 2.4 2 2.4 1 2.3 0 2.2 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: e-Health Center (2020) Figure 2.15. Nurse-to-physician ratio, latest year available 14 Physicians Nurses and midwives Nurse to physician ratio 3.5 12 3 Nurse to physician ratio Per 1,000 population 10 2.5 8 2 6 1.5 4 1 2 0.5 0 0 Kyrgyz LMIC Central Asia UMIC ECA Republic Source: e-Health Center (2020) and WHO (2020) 2.27. The decline in the physician ratio has impacted the availability of staff, particularly at the PHC level. Vacancy rates are substantially higher in PHC facilities than in hospitals, but there are large variations in rates across oblasts. In 2018, vacancy rates in PHC/outpatient facilities (FMCs) were 12.7percent for doctors (up from 11.0 percent in 2015) and 5.1 percent for mid-level medical staff (up from 3.4 percent in 2015). Conversely, the vacancy rates are significantly lower in inpatient facilities: 4.8 percent for doctors (up from 4.3 percent in 2015) and 3.3 percent for mid-level medical staff (up from 2.8 percent in 2015). The averages, however, hide important variations across regions (Figure 2.16). Talas oblast had the highest vacancy rate for PHC facilities (20.4 percent), while Chuy oblast had the highest vacancy rate for inpatient facilities (6.1 percent). Republican preventive and treatment facilities had difficulties filling vacancies at every level, including for mid-level medical staff. Vacancy rates are typically much lower for mid-level medical personnel than for doctors. Bishkek only had a 1 percent vacancy rate for physicians at PHC facilities, suggesting that the problem may not be only one of availability of skilled personnel but the ability or willingness of qualified individuals to locate outside of Bishkek. 51 Figure 2.16. Vacancy rates for physicians by facility, 2018 Talass oblast Issyk-Kul oblast Osh oblast Chuy oblast Jalal-Abad oblast Naryn oblast Batken oblast Kyrgyz Republic PHC facilities Osh City Hospitals Bishkek City 0 2 4 6 8 10 12 14 16 18 20 Vacancy rate (%) Source: e-Health Center (2020). E. UTILIZATION OF HEALTH SERVICES 2.28. Access to and utilization of maternal and newborn health services have improved considerably, reaching almost universal coverage in 2018, with few regional disparities. Almost all mothers visit antenatal care (ANC) services at least once, with 94 percent receiving at least four ANC visits. There is some regional variation: only 90 percent of mothers seek ANC at least four times in Jalal-Abad oblast, compared with 99 percent in Batken oblast and Osh City. Moreover, 99 percent of women had their blood pressure measured, urine sample taken, and blood sample drawn during the ANC visits. Coverage of postnatal care for mothers and newborns is also high (96 percent and 98 percent, respectively). Some elements of newborn care, however, are still lagging. For example, only 63 percent of women who gave birth in the past two years reported skin-to-skin newborn care (Figure 2.17) (NSC and UNICEF 2019). Figure 2.17. Coverage of postnatal care, 2018 100 Newborn dried, 95 80 60 Early initiation of 40 breastfeeding, 81 20 Skin-to-skin care, 0 63 Delayed bathing, Postnatal signal 93 care functions, 98 Source: MICS (2018). 52 2.29. Utilization of other health services, particularly outpatient care, is quite low. The Kyrgyz Republic has one of the lowest utilization rates of outpatient care in the region, with 3.5 outpatient contacts per person per year. By comparison, the WHO European region average is 7.5 outpatient visits per person per year (Figure 2.18). The proportion of population using outpatient services increased only slightly, from 9 percent in 2000 to 13 percent in 2014.27 Conversely, at 14.1 inpatient care discharges per 100 people, the Kyrgyz Republic’s rate of hospitalization was also lower than the WHO European region average of 17.8 inpatient discharges per 100 people in 2016. There is significant regional variation in the hospital admissions rate, ranging from 9.1 admissions per 100 in Chuy oblast to 14.0 in Osh City (Annex Figure 3). Figure 2.18. Utilization of health services, latest available year Outpatient contacts per person per year Hospitalizations per 100 people Belarus 12.8 Belarus 30.3 Uzbekistan 9.7 Russia 23.7 Ukraine 9.3 Ukraine 19 Russia 8.8 Moldova 18.4 Turkey 8.3 WHO EURO 17.8 WHO EURO 7.5 Estonia 17.5 Moldova 6.4 Uzbekistan 16.7 Estonia 6.3 Turkey 16.6 Tajikistan 4.5 Kyrgyz Republic 14.1 Georgia 4.0 Tajikistan 12.3 Kyrgyz Republic 3.5 Georgia 9.7 0 5 10 15 0 10 20 30 40 Source: WHO Health for All Database (2019). Latest year available since 2013. 2.30. Strong primary healthcare is critical for effective management of chronic illnesses and prevention/mitigation of modifiable risk factors. At present, primary care remains relatively weak and is not attractive to patients or health professionals. This is due in part to limited diagnostic and treatment opportunities and the perceived low quality of care. Primary care does not play a gatekeeping role, and patients often choose to bypass PHC to seek care from higher level facilities. Due to poor coordination across levels, PHC facilities are not able to continuously monitor patients, particularly post-discharge. In addition, PHC providers are not held accountable for avoidable hospitalizations or patient outcomes. 2.31. The shortages in essential inputs, including human resources, in primary care may partly explain the observed low primary care utilization rates in the Kyrgyz Republic. There is a notable shortage of physicians in primary care in the Kyrgyz Republic. A shortage of essential equipment and supplies also limits the extent of services that can be provided in primary care facilities and may influence patient decisions to seek care at higher level facilities. For example, Chuy primary care center, with a catchment population of almost 110,000 and about 210,000 patient visits per year, has only one electrocardiogram machine and no ultrasound, colposcopy, or blood analyzer machines. Even when basic equipment is available, supplies are often not and, as a result, basic liver and kidney tests, HbA1c, and blood electrolytes are not offered in most primary care facilities in the country. Access to essential pharmaceutical inputs in primary care, which is a key factor in the effective management of chronic conditions, is also extremely limited. 27 Among those who utilized outpatient services, almost 30 percent visited FMCs or polyclinics, 29 percent sought care at family group practices, and 21 percent sought care at FAPs. Only 9 percent of individuals sought outpatient care at private facilities. 53 2.32. There is some room to further improve the efficiency of the health system. While the bed occupancy rate at 86 percent is relatively high, the average length of stay is 8.1 days, slightly higher than the WHO European average of seven days (Figure 2.19). Notably, the average length of stay declined from 10.1 days in 2010 to 8.2 days in 2018. Regional differences remain, with the average length of stay ranging from 8.6 days in Osh oblast to 7.1 in Talas oblast in 2018 (Annex Figure 4). Figure 2.19. Bed occupancy rate and average length of stay, latest available year Bed occupancy rate (%) Average length of stay (days) Belarus 89% Ukraine 9.9 Ukraine 88% Russia 9.7 Uzbekistan 87% Kyrgyz Republic 8.1 Russia 87% Tajikistan 8 Kyrgyz Republic 86% WHO EURO 7 WHO EURO 79% Moldova 7 Moldova 76% Estonia Uzbekistan 5.9 69% Turkey 66% Estonia 5.4 Tajikistan 66% Georgia 4.5 Georgia 36% Turkey 3.8 0% 20% 40% 60% 80% 100% 0 2 4 6 8 10 12 Source: WHO Health for All Database (2018). 2.33. It is possible to further reduce the Table 2.3. Avoidable hospitalizations in the Kyrgyz number of avoidable hospitalizations (that Republic, 2016-18 is, those that could have been prevented 2016 2017 2018 with adequate primary care) and use the Admissions related to hypertension cost savings to support cost-effective (percent of all admissions) 4.3 2.4 2.0 interventions. Conditions such as Admissions related to asthma 11.0 7.6 6.0 hypertension and asthma could be prevented (percent of all admissions) with information and education campaigns and Source: MHIF. managed effectively at the PHC level without requiring hospitalization, except in complicated cases. Encouragingly, avoidable hospitalizations in the Kyrgyz Republic have declined over the past three years. The share of admissions related to hypertension declined from 4.3 percent in 2016 to 2.0 percent in 2018, while the share of admissions related to asthma fell from 11.0 percent to 6.0 percent (Table 2.3). The hospitalization rates for tracer chronic conditions28 in the Kyrgyz Republic, however, are significantly higher than compared with select OECD countries. Moreover, significant variations across regions on the rates of hospitalizations imply regional variations in quality. 2.34. Pneumonia and acute upper respiratory infection were among the top 20 hospitalization cases in 2018, accounting for 56,829 admissions. According to evidence-based guidelines, these conditions, in most cases, should be managed outside the hospital. The MHIF spent 388.8 million soms for inpatient treatment for these conditions in 2018. Reducing hospitalization rates for pneumonia and acute upper respiratory infection by 20 percent could potentially save the government 77.8 million soms—resources that could be used to develop more cost-effective outpatient treatment. 28 Tracer conditions include hypertension, chronic obstructive pulmonary disease, diabetes, and asthma. 54 2.35. The hospital system is not uniformly Figure 2.20. Cost per bed-day per hospital and efficient. There is significant unexplained occupancy rate variation in the cost per day across hospitals, and the cost per day do not appear to be correlated with a hospital’s occupancy rate (Figure 2.20). In addition, there is considerable variation in similar expenses among hospitals. For example, employee costs vary from 245 soms to 629 soms, medicine cost varies from 84 soms to 521 soms, and OT/Cathlab cost varies from 33 soms to 116 soms. The cost per bed-day within the same department across hospitals varies by more than 200 percent. The large variation in resources used to treat a disease suggests that the standards of care are not uniformly applied. Source: World Bank (2020) F. HEALTH EXPENDITURE 2.36. High performance health financing is an investment that benefits the economy through six main channels: (1) building human capital; (2) increasing skills and jobs, labor market mobility, and formalization of the labor force; (3) reducing poverty and inequity; (4) improving efficiency and financial discipline; (5) fostering consumption and competitiveness; and (6) strengthening health security. High- performance health financing for universal health coverage means that funding is adequate and sustainable, pooling is sufficient to spread the financial risk of ill-health, and spending is both efficient and equitable to assure the desired levels of service coverage, quality, and financial protection for all people (World Bank 2019). 2.37. Current health spending (inclusive of government and private sources) is slightly above average for Kyrgyz Republic’s income level. In 2017, Kyrgyz Republic spent approximately $79 per capita on health, or 6.2 percent of GDP (Figure 2.21). The composition of health spending has not changed substantially, and the health system is still predominantly financed by OOP spending (Figure 2.22). Despite relatively high levels, after accounting for inflation, current health spending has been declining since 2013 (falling from $94 to $79 in constant US$ between 2013 and 2017), largely due to declines in government health spending. 55 Figure 2.21. Current health expenditure per capita (US$) and as a share of GDP, 2017 (percent) 10,000 20 2,500 Current health expenditure 15 Estonia (Percent of GDP) Russian Federation 500 Armenia Turkey Georgia Belarus Armenia (US$) Azerbaijan Kazakhstan 10 Ukraine Moldova Georgia 100 Kyrgyz Republic Uzbekistan Moldova Tajikistan Ukraine Tajikistan Azerbaijan Estonia Kyrgyz Republic Uzbekistan Russian Federation 5 Belarus 25 Turkey Kazakhstan 5 0 250 500 1,025 2,000 3,995 7,000 12,375 25,000 50,000 100,000 250 500 1,025 2,000 3,995 7,000 12,375 25,000 50,000 100,000 GNI per capita (US$) GNI per capita (US$) Source: WDI and WHO Global Health Expenditure Database (2020) Figure 2.22. Current health spending per capita (in constant 2017 US$) by source, 2000-17 Government schemes and compulsory contributions Household out-of-pocket payment Other private spending 100 94 89 86 86 Current health spending per capita (2017 90 4 76 5 2 3 79 79 80 75 72 71 0 1 66 6 67 67 1 70 3 2 44 56 4 5 2 39 60 49 33 49 47 30 28 31 45 44 50 9 28 29 33 5 US$) 34 34 36 40 0 0 0 26 30 25 17 18 19 47 20 40 41 42 46 34 34 32 39 36 37 34 33 10 16 16 18 19 21 0 Source: WHO Global Health Expenditure Database (GHED). 2.38. Although government health expenditures in the Kyrgyz Republic are above average for its level of income, the share of GDP allocated to health has declined substantially since 2012 (Figure 2.23). In 2017, the country spent 2.6 percent of GDP on health, compared with 2.9 percent in Uzbekistan, 3.1 percent in Ukraine, and 3.6 percent in Moldova. Notably, the share has declined significantly since 2012, when it reached a peak of 4.4 percent of GDP. Prioritization of health within the government budget remains relatively low, with health representing less than 11 percent of government spending in the Kyrgyz Republic. Conversely, education and social protection receive almost 25 percent of the government budget each (Figure 2.24), and the education share has increased by almost 9 percentage points since 2010, when it represented only 16 percent. Importantly, the government has not been able to maintain the 13 percent budgetary target for health proposed under the Manas Taalimi program and supported by development donors under the Sector-Wide Approach Project. 56 Figure 2.23. Government health spending as a share of GDP, 2017 (percent) 10 8 (Percent of GDP) 6 Estonia 4 Belarus Moldova Ukraine Georgia Turkey Kyrgyz Republic Uzbekistan Russian Federation 2 Tajikistan Kazakhstan Armenia Azerbaijan 0 250 500 1,025 2,000 3,995 7,000 12,375 25,000 50,000 100,000 GNI per capita (US$) Source: WDI (2020) and WHO Global Health Expenditure Database (2020). Figure 2.24. Government expenditures, 2014-18 124910.3 129098.1 115961.0 120000 105252.4 21.5% 23.4% 95584.6 21.5% Social Protection 100000 23.6% 24.4% 25.8% 25.9% Education Million Soms 80000 26.2% 24.2% 60000 23.5% 10.5% Health 11.5% 12.9% 12.1% 40000 12.6% Other government spending 40.8% 39.8% 40.2% 20000 39.6% 40.1% 0 2014 2015 2016 2017 2018 Source: Ministry of Finance (2020) 2.39. As the administrator of the SGBP and the ADP, the MHIF is the main spending authority for the health budget, with the MoH financing only limited vertical programs. The MHIF relies on the republican budget for most of its revenues. The republican budget constitutes more than 70 percent of the MHIF revenues, with social insurance contributions representing less than 16 percent. Almost 4 percent of revenues comes from copayments (Table 2.4). Table 2.4. Mandatory Health Insurance Fund revenues, 2018 and 2019 2018 2019 Social insurance contributions (million Soms) 2328.4 2320.0 16.0 percent 15.2 percent Republican budget (million Soms) 10466.7 10739.2 71.8 percent 70.5 percent Copayment (million Soms) 550.3 551.7 3.8 percent 3.6 percent 57 Special funds (million Soms) 744.5 543.6 5.1 percent 3.6 percent Other (voluntary premiums, carryover revenues) (million Soms) 489.5 1079.1 3.4 percent 7.1 percent Total revenues (million Soms) 14,579.4 15,233.6 Source: MHIF Annual Report (2020). 2.40. The amount of funds allocated for the exempt categories, however, is insufficient to cover the cost of treatment. The government transfers only 150 soms per person belonging to an exempt category to the MHIF. In 2019, the MHIF received only 429.6 million soms, while it paid 1,945.4 million soms for inpatient treatment for the exempt categories. This resulted in a deficit of 1,415.8 million soms that had to be subsidized from other revenue streams (Table 2.5). Table 2.5. MHIF revenues and expenditures for exempt categories of the population, 2019 (million soms) MHIF expenditures for Difference MHIF revenues Exempt categories inpatient treatment (million (million soms) (million soms) soms) Unemployed 2.6 7.4 -4.8 Individuals receiving social benefits 9.4 133.2 -123.8 Students 20.9 51.4 -30.5 Military personnel 0.4 3.9 -3.5 Pensioners 73.4 421.9 -348.5 Children 322.9 1,227.6 -904.7 Total 429.6 1,945.4 -1,415.8 Source: MHIF Annual Report (2019). 2.41. In 2018, the SGBP represented the Figure 2.25. Distribution of government health bulk of government spending in the health spending by program, 2018 sector (78 percent), while the ADP represented State Guaranteed only 2 percent (Figure 2.25). Spending on the ADP 1% Benefits Package remains too low to bring health benefits and be Program (total) attractive to participating pharmacies. The MHIF 15% MHIF Additional spends less than 2 percent of its budget on the 5% Drug Package ADP, using a budgetary norm of $1 per insured 2% person. These funds are not sufficient to cover the cost of providing ADP medicines. Public Health Care 78% 2.42. The list of MHIF reimbursable drugs is extremely limited, and even for drugs Medical-social rehabilitation, included in package funding is insufficient to administration, cover the population in need. A recent analysis education conducted by the World Bank shows a substantial Source: MHIF gap between the number of reimbursed prescriptions and the estimated need for three conditions (hypertension, diabetes mellitus, and anemia) and contraceptives. In 2017, the MHIF reimbursed on average 1.7 prescriptions per registered hypertensive patient. While the prescription dosage is not specified, the amount per prescription was only $4.16. For diabetes and anemia, the two high-burden conditions in the country, coverage is even lower (Nguyen and Strizrep 2019) (Table 2.6). Notably, the population in need was estimated using the number of registered 58 patients with hypertension and diabetes. Given the large number of undiagnosed cases, the actual population in need is likely to be larger, indicating significant gaps in funding. 2.43. The limited funding for the ADP results in rationing at the provider level (without clear guidance or monitoring). This includes budget caps for ADP medicines being set at the FMC level, FMC distribution of special prescription forms for ADP medicines to primary care physicians on a monthly basis, and, if need exceeds the number of forms, primary care physicians have the discretion to decide which of their patients will receive medicines under the ADP and which will have to pay the full price. Table 2.6. MHIF’s reimbursement under the ADP compared to estimated need, 2017 Amount per person in Reimbursed Amount (US$ Estimated Prescription per Conditions need (US$ prescriptions equivalent) population in need person in need equivalent) Hypertension 284,434 705,153 169,412 1.68 4.16 Diabetes mellitus 1,140 13,328 56448 0.02 0.24 Anemia 131,323 568,575 812,800 0.16 0.70 Contraceptives 5,959 50,279 1,569,594 0.00 0.03 (injection and oral) Source: Nguyen and Strizrep (2019). Source of data: (1) for MHIF’s reimbursement: ADP database; (2) for population in need: number of people with hypertension and diabetes represents the registered cases recorded by eHealth center; number of people with anemia is a sum of estimated number of children ages 0-5 with anemia and estimated number of women ages 15-49 with anemia (source: WDI); people in need of oral and injection contraceptives are women ages 15-49. All data are for 2017 except the estimated population in need of anemia drugs and contraceptives (2016). Exchange rate: US$1 = 70 Kyrgyz soms. 2.44. Approximately 60 percent of the SGBP funds are spent on inpatient care and 36 percent are spent on PHC (Figure 2.26). Notably, the share allocated to PHC has increased by 2 percentage points since 2017, in line with the government’s commitment to focus on PHC. The continued reliance on inpatient care, however, is inefficient. 2.45. Despite the low number of physicians and low wages in the health sector (relative to other government sectors), the relatively high share of expenditures allocated to salaries reflects the limited funding of the health sector. Personnel costs are the largest expenditure item in public health expenditure, absorbing almost two thirds of the total budget (Annex Figure 6). The share of personnel expenditure, however, cannot be reduced by freezing salaries, as these are still relatively low compared with other public sector wages in the Kyrgyz Republic (Figure 2.27).29 Since 2017, average health sector salaries have stood at 63 percent of the average national salary. Meanwhile, the salaries in public administration and education have slightly increased relative to the average national salary during this time period. 29 All employees of public health facilities are public sector employees, with corresponding rights, privileges, and benefits. 59 Figure 2.26. SGBP spending by level of care, 2018 Figure 2.27. Salaries in health relative to national average and government salaries in other sectors, 2014-19 Inpatient care Public administration Education 2% 1% Health and social services 1% 1.4 Wage relative to national average wage General centers of 1.26 practice 1.3 1.22 1.21 1.20 1.17 1.16 1.2 28% Family medicine centers across all sectors 1.1 1 Family group practices 60% 0.9 0.75 0.73 0.72 8% 0.8 0.68 0.71 0.69 Emergency care 0.67 0.7 0.6 0.70 0.66 0.63 0.63 0.63 Dentistry 0.5 2014 2015 2016 2017 2018 2019 Source: MHIF Source: National Statistics Committee (2020) 2.46. Budget execution rates have been relatively high for the health sector (that is, more than 94 percent), indicating that money allocated is effectively spent. This, however, may also be due to the way the budget processes are implemented, including the use of frequent amendments to revise the budget throughout the year that could ensure the achievement of certain budget execution rates. 2.47. In 2018, the Kyrgyz Republic introduced program-based budgeting in the health sector (Figure 2.28). The 2020 Program Budget for Health has eight programs (four programs in the MoH and four programs in the MHIF budget). Since appropriations can only be made within ministries or government entities, the MoH and the MHIF have separate programs for their expenditures. The MHIF budget represents the bulk of government health spending and contains the following four programs: (1) planning, governance, and administration; (2) delivery of PHC services; (3) delivery of inpatient care; and (4) additional services outside the SGBP. The program-based budget fluctuates substantially over the three-year period, ranging from 14.7 billion soms in 2021 to 17.2 billion soms in 2022. Program budgeting has yet to fully reflect the stated priorities of the government’s health program. For example, despite the declared focus on PHC, the share of the MHIF budget allocated to PHC will decline from 37.8 percent in 2020 to 31.8 percent in 2022. Figure 2.28. Program-based budget for MHIF, 2020–2022 20,000,000 18,000,000 17,215,619 16,009,830 16,000,000 14,661,032 Kyrgyz Soms (thousands) Program 1. Planning, governance, 14,000,000 5,480,871 and administration 6,058,040 12,000,000 5,829,096 10,000,000 Program 2. Delivery of primary health care services 8,000,000 6,000,000 8,410,753 11,044,747 Program 3. Delivery of inpatient 4,000,000 8,151,936 care 2,000,000 1,383,037 511,000 511,000 Program 4. Additional services - outside the SGBP 2020 2021 2022 Source: MHIF 60 2.48. Public financing is a key factor for countries to make progress toward universal health coverage. Countries that are in the highest quintile of the WHO-World Bank universal health service coverage index30 are those that have higher levels of government financing for health in levels, as a share of GDP, the government budget, and total health spending (Table 2.7) (Tandon and others 2020b). Table 2.7. Public financing and the universal health coverage index of service coverage, 2017 Public Health share Public spending OOP spending Per capita UHC index of spending of public share of total share of total public service coverage share of GDP spending health spending health spending spending (US$) (%) (%) (%) (%) Lowest quintile 32 2.3 8.5 39.2 41.9 Second-lowest quintile 108 2.8 9.5 51.5 32.2 Middle quintile 252 3.2 10.3 53.1 39.9 Second-highest quintile 566 3.8 12.2 61.3 30.4 Highest quintile 2,512 6.1 15 69.1 20.8 Source: Tandon and others (2020b). To further improve the health system's efficiency, there is a need to avoid hospitalization by using the cost-savings to support cost-effective interventions with a view to achieve the government's broader development objectives. 2.49. Prioritization of health in budget decisions is important. Low levels of domestic government financing in the Kyrgyz Republic mean that there is currently a substantial gap between the costs of financing an essential package of quality services for the population and the available resources. Efficiency gains as described in the previous section are important, but a low level of public financing is itself also a source of inefficiency (Figure 2.29). Figure 2.29. Inefficiencies of low public spending on health 30 The universal health coverage index measures the proportion of a country’s population with access to health services and prevalence of health risk factors. 61 2.50. The Kyrgyz Republic continues to rely on Figure 2.30. Out-of-pocket spending as a share OOP payments as a main source of funding for the of total health spending vs. public spending on health system—an inequitable and inefficient way health as a share of GDP, 2017 (percent) of financing (Figure 2.30).31 In 2017, OOP payments Armenia represented 56 percent of all current health spending 80 Azerbaijan (Percent of current health spending) (WHO GHED, 2020). Notably, the share has been 70 Out-of-pocket health spending steadily increasing since 2009, when it was at the lowest Tajikistan 60 Kyrgyz Republic Georgia (38 percent). The upward trend is alarming, as it 50 Uzbekistan Ukraine prevents the country from ensuring financial 40 Russian Federation Moldova protection. 30 Kazakhstan Belarus Estonia 20 2.51. In 2014, individuals on average spent Turkey 10 2,689 soms (in constant terms) compared with 1,539 soms (in constant terms) in 2009 (WHO 0 0 2 4 6 8 10 2018). The rate of increase in annual health spending has Public spending on health also been higher among poorer households than those in (Percent of GDP) the wealthiest quintile. Between 2009 and 2014, average Source: WHO Global Health Expenditure Database (2020) annual health spending increased by almost 76 percent among the poorest quintile compared with 49 percent among the richest (WHO 2018). Most of the OOP spending is on pharmaceutical products. Medicines and medical products account for more than 60 percent of OOP spending, followed by inpatient care (16 percent in 2014) (WHO 2018). Figure 2.31. Composition of out-of-pocket health spending in the Kyrgyz Republic, 2000-14 Source: Jakab, Akkazieva, and Habicht (2018). 2.52. As a share of household consumption expenditure, OOP spending increased from 3.9 percent in 2009 to 7.2 percent in 2014 (Figure 2.31). Almost 13 percent of households experienced catastrophic spending on health32 in 2014. OOP spending is also regressive, with households in the poorer quintiles spending a higher share of their budget on health than richer quintiles (Table 2.8). On average, households spent about 5 percent of their non-food consumption on health in 2019 (Kyrgyz Integrated Household Budget Survey). 31 OOP payments are defined as direct payments by households and individuals for healthcare goods and services at the time of the purchase of goods or use of the services. These are different from household contributions in a prepayment form, such as premiums for insurance, and include both formal and informal payments (for example, under-the-table payments). 32 Catastrophic OOP spending is defined as the share of households with OOP payments that are greater than 40 percent of household capacity to pay for health care. Capacity to pay for health care is measured as the difference between total household consumption and a standard amount to cover basic needs (foods, housing, and utilities) (WHO 2018). 62 Table 2.8. Household health expenditures, 2000–19 2000 2005 2010 2015 2016 2017 2018 2019 Percent of expenditures 1.0 % 2.1% 1.7% 2.5% 2.1% 2.2% 2.8 % 2.8% Percent of non- food 2.2% 4.5 % 3.7% 5.3% 4.4% 4.1% 5.1% 5.0% expenditures Source: NSC. 2.53. High levels of OOP spending are a barrier to utilization and to improving health outcomes. In addition, large OOP health expenditures can have a direct impact on poverty. Many people facing financial catastrophe due to health shocks sell assets, go into debt, or reduce their consumption of other necessities, while to avoid such consequences, others forego health services from the outset. It is important to draw on funds from prepaid and pooled sources, with subsidies for people who cannot afford to contribute, to ensure that financial protection for a set of guaranteed services is universally available. 2.54. There are several drivers of the high OOP spending. Government health spending in general is low and has been declining in real per capita terms since 2012. Low funding allocated to providers necessitates extra payment collection (for drugs, supplies, and services) and informal payment to supplement health workers’ income. In addition, coverage of pharmaceuticals is extremely low due to limited funding. This, in part, explains the high OOP spending on medicines and medicinal products. At the same time, however, there are several challenges related to pharmaceuticals, including high levels of inappropriate prescriptions resulting in overuse and misuse, weak regulation of the overall drug market, and higher prices for drugs than in wealthier countries. In addition, the share of OOP spending on non-prescribed drugs has increased sharply, pointing to the need for population education on rational drug use. Also, given the weak PHC system, patients often bypass PHC to seek care in facilities with high OOP payments (for example, private facilities and hospitals). In addition to increasing government funding for health, revisions to the SGBP and the ADP are needed to improve financial protection. Meanwhile, pharmaceutical reforms and regulations can address unnecessary prescriptions and lower the market prices. G. PAYMENT SYSTEMS 2.55. The Kyrgyz Republic was among the first former Soviet countries to introduce new provider payment mechanisms beginning in the 1990s, including the introduction of DRGs and capitation payments for PHC (Figure 2.32). The government, however, has not revised these methods to account for changing population demographics, utilization patterns, and available budgetary space. As a result, hospital effectiveness and efficiency remain a major challenge. The payment systems are not sophisticated enough to deliver desirable quality, fairness, and efficiency. Pure capitation runs the risk of under-provision of services in both quantity and quality, especially in the context of budget constraints, low population awareness of entitlement, and the MHIF’s lack of data to monitor provider behaviors. 2.56. The DRGs are rather rudimentary, and the payment rates have not been adjusted to reflect the cost of delivering services. A recent hospital costing study supported by the World Bank shows that in 29 diagnoses for which costing was performed, government case-based payment rates were below actual costs for 18 conditions. The rate per bed-day in the Kyrgyz Republic of $102 is significantly lower than the rates in Georgia ($157), Ukraine ($236), Kazakhstan ($317), and Russia ($322). At the same time, for some conditions, the MHIF rates exceeded the actual cost of delivering the services (World Bank 2020). Further investigation is warranted into hospital operations and MHIF’s pricing mechanism. 63 Figure 2.32. Estimated cost vs. MHIF rate for select inpatient services Acute upper respiratory infection, unspecified Single live birth Bacillary enteric infection unspecified Acute bronchitis, unspecified (general dep., short-stay dep.) Spontaneous vertex delivery Hypertensive heart disease without (congestive) heart… Excision of varicocele and hydrocele of spermatic cord Hypertensive encephalopathy Bronchopneumonia unspecified Open repair of indirect inguinal hernia Encephalopathy, unspecified Cost Other specified bacterial intestinal infections Government rate Atherosclerotic heart disease Chronic obstructive pulmonary disease with acute… Classical cesarean section Other operations on skin and subcutaneous tissue Appendectomy Unstable angina Pneumonia, unspecified Cholecystectomy Intracranial injury Hepatitis A without hepatic coma Laparoscopic cholecystectomy Type 2 diabetes mellitus with multiple complications Allergic purpura 0 2000 4000 6000 8000 10000 12000 14000 16000 Source: World Bank (2020). 2.57. Within the current case-based payment for hospital services, the DRG system assigns patients to the DRGs based on the main diagnosis and procedure only. This means that it does not account for comorbidities and complications. As a result, hospitals treating more complicated cases bear financial risk because they are paid the same amount per episode of care as the lower level hospitals that treat less complicated cases. For example, in the Kyrgyz DRG system, cardiac surgery is represented by one group, while in the Australian DRG system, it can be assigned to 12 groups (without interventional cardiology). Moreover, the current reporting system in the Kyrgyz Republic only allows providers to enter one additional diagnosis for complications and two additional diagnoses for comorbidities. This level of detail is insufficient for planning and future DRG refinement (Nguyen and Strizrep 2019). 2.58. The MHIF has initiated several schemes to incorporate performance indicators into the PHC payment, including the rollout of the PHC Balanced Scorecard, payment for treated tuberculosis cases, and a stimulation package for family medicine doctors (Figure 2.33). In 2018, the MoH and MHIF introduced a new pay-for-performance program to improve the quality of care in primary care facilities and hospitals. The new initiative stemmed from the World Bank-supported Results-Based Financing Project and has two main components: (1) balanced score care-based payments to health facilities; and (2) indicator-based salary top-ups to primary care physicians (Ahmedov and others 2020). While pay-for- performance at the primary care level has been found to improve quality of care, it is important to ensure alignment between the various payment mechanisms for the incentives to be effective. The MHIF should consolidate the various initiatives and revise the payment methods to incentivize quality and coverage of priority high burden conditions. 64 Figure 2.33. Components of MHIF’s Results-Based Financing schemes Detection and Prescribing of Monitoring the Antenatal care treatment of Detection and health of children medicines Patient for pregnant hypertension, treatment of according to under age 5 women prevention of type 2 diabetes satisfaction additional CVD program of MHI complications Share of newborns registered at FPG (FAP) Number of newly within the first three Share of women diagnosed and days after discharge registered for registered from the hospital (in pregnancy at GA patients with the case of home ≤ 12 weeks arterial delivery without hypertension Number of Share of Percentage of subsequent newly prescriptions satisfaction with hospitalization - diagnosed and issued the quality of within the first three registered according to AP medical services days after birth) patients with of MHIF, by at the outpatient type 2 diabetes generic names level Share of Share of children successful pregnancies Share of patients age under the age of 5 in women who ≥ 40 years of age who regularly observed received the basic have a risk of CVD by FPG physician ANC package in mortality (FAP feldsher) accordance with CP H. COVID-19 RESPONSE AND THE MACRO-FISCAL IMPACT OF COVID-19 ON HEALTH FINANCING 2.59. The government of the Kyrgyz Republic was quick to react to the COVID-19 pandemic. On February 3, 2020, the government issued a decree on the allocation of additional funds in the amount of 30 million soms (approximately $430,000) for the COVID-19 response33, and in March 2020, the government of the Kyrgyz Republic approved the COVID-19 Contingency Response Plan. Overall, the government approved a budget of 5,692.2 million soms ($71.4 million) for the COVID-19 emergency response. In the amended budget for 2020, an additional 2,061.6 million soms ($25.93 million) was allocated directly for health (an increase of 13 percent over the approved budget). Of these additional funds, 21 percent came from budgetary funds and 79 percent from government investment. As of August 15, 2020, almost 75.6 percent of the allocated funds for COVID-19 was spent. Funds were mobilized through emergency budget plans (17 percent), reallocations from previously approved budget, budgetary reserves, additional funds, and savings from servicing external debt. 2.60. Figure 2.34 shows the breakdown of expenditures by government entity. Funds were used to procure medical equipment, goods, and supplies (including personal protective equipment, tests, and disinfecting supplies); rehabilitate health facilities, provide food to low-income households; award bonus payments to health workers; transfer students from abroad; and establish quarantine facilities. Almost 2,337 million soms ($29.8 million), or 54 percent, of funds were spent directly by the health sector (MoH and MHIF). Of these, almost 46 percent was spent on procurement of personal protective equipment, 29 percent on bonus payments to health workers, and 25 percent on procurement of medical equipment, goods, and supplies. In 33 In addition, the MoH opened a special account to accumulate voluntary contributions from the public and private companies. By May 12, 2020, almost 137 million soms (US$1.73 million) had been raised, with the majority of funds used for supplemental pay to health workers. 65 addition, the Department of Construction spent 12 million soms ($0.15 million) to renovate seven government hospitals. Figure 2.34. Breakdown of COVID-19 emergency response expenditures by government entity Ministry of Labor Other and Social 1% Development Ministry of Foreign 5% Affairs Ministry of 4% Health 16% State Material Reserve Fund 16% Mandatory Health Ministry of Insurance Fund Emergency Security 38% Situations forces 6% 14% Source: Ministry of Finance of the Kyrgyz Republic. Data as of August 15, 2020 2.61. Much remains unknown as to how the health and economic impact of COVID-19 will affect levels and sources of health financing. As of September 28, 2020, the Kyrgyz Republic reported 46,355 confirmed cases of COVID-19, of whom 1,064 individuals had died. Depending on the dynamics between new infections and the extent of lockdowns across countries, the economic shock may continue into 2021 or even longer. 2.62. Given the pressing needs, the government was able to increase funding to the health sector in 2020 despite a 5.9 percent contraction in real GDP between January and August 2020 and substantial declines in revenues (World Bank 2020). The macroeconomic outlook remains pessimistic for 2020, with slower recovery in 2021. The latest International Monetary Fund (IMF) projections suggest that real GDP per capita in the Kyrgyz Republic will decline by 13.8 percent in 2020, representing the seventh highest decline in the world. This is significantly higher than the declines projected in other Central Asia countries and above the averages for lower middle-income countries (-4.9 percent) and those in ECA (-6.0 percent) (Figure 2.36). At the same time, government expenditures are expected to increase. Due to declining revenues, this will result in a widening of the fiscal deficit. The fiscal deficit is expected to increase from 0.3 percent of GDP in 2019 to 7.3 percent of GDP in 2020 (Figure 2.35). Although GDP growth is projected to bounce back in 2021, the magnitude of recovery will largely depend on the government’s ability to address the COVID-19 pandemic and the availability of a vaccine. 66 Figure 2.35. Real GDP per capita growth actual (2000-19) and projections (2020-25), (percent) Lower middle income average Europe & Central Asia average Kyrgyz Republic Kazakhstan 15 Russian Federation Tajikistan Real GDP per capita growth (%) Uzbekistan 10 5 0 -5 -10 -15 Source: IMF October 2020 World Economic Outlook. Figure 2.36 Government revenues and expenditures in the Kyrgyz Republic, actual (2010-19) and projections (2020-25) Fiscal balance Government revenues Government expenditures 50 40 Percent of GDP (%) 30 20 10 0 -10 Source: IMF October 2020 World Economic Outlook 2.63. Four scenarios were used to model the macro-fiscal impact of COVID-19 on government health spending in the Kyrgyz Republic. Using the October IMF projections for GDP, government revenues, and government expenditures and the WHO Global Health Expenditure Database for health spending, government health spending was projected until 2025. Scenario 1 estimates the elasticity of government health spending per capita relative to GDP per capita, the general government expenditure to GDP ratio, and the debt service-to-general government expenditure ratio, accounting for periods of GDP per capita declines and increases. The elasticities are used to forecast government health spending per capita from 2018 onward. Scenario II assumes that the government health expenditure to general government expenditure ratio will remain at the same level as in 2017 while Scenario 3 assumes that the average government health expenditure per capita growth rate between 2009 and 2019 will remain the same. Scenario IV applies average consumption-to-GDP ratio between 2015 and 2017 to forecast OOP spending per capita from 2018 onward and total health expenditure per capita (using government health expenditure per capita from Scenario 1) to calculate how much government health spending per capita is needed to maintain the trend from 2018 onward. 67 2.64. If public spending on health responds to the current economic shock the same way it has in previous years, real per capita public spending on health can be expected to decline as a result of the economic contraction. This would hold true even after controlling for changes in the government spending share of GDP and accounting for the debt servicing share of government spending. Given current projections of per capita GDP, the government spending share of GDP, and the debt service share of government spending (Scenario 1), per capita public spending could contract by almost 18 percent in 2020 and would not reach the pre-crisis level by 2025 (Figure 2.36). Even if the government protects health’s share of the budget by keeping health’s share of government spending at pre-crisis levels, real per capita spending on health will grow slower than in pre-crisis years. Figure 2.37. Projections of government health spending per capita (2017 US$), 2020-25 Actual Scenario I Scenario II Scenario III Scenario IV 60 Government health spending per 50 capita (2017 US$) 40 30 20 10 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Source: Authors’ calculations using data from the IMF October 2020 World Economic Outlook and the WHO GHED. 2.65. Despite the bleak macroeconomic outlook, moving forward, it will be important to ensure that the government reprioritizes health within the government budget. To continue the delivery of essential services and avoid reversal of progress to date, the government must continue to secure adequate resources for the health sector. 2.66. Health taxes present a win-win opportunity for the government. Not only can they help to increase the fiscal space for health spending they can be an effective instrument for reducing the consumption of harmful substances and thereby decreasing the burden of NCD. This includes taxes on goods and services that have harmful health effects, such as on tobacco, alcohol, sugar-sweetened beverages, and carbon emissions. Given the tight fiscal environment, such taxes are less likely to face political opposition and may be easier to collect than broader consumption taxes. For example, in tobacco, not only can the increase in the tobacco excise tax bring in more revenues, but it can alsoresult in considerable health and financial benefits to Kyrgyz households and be pro-poor (Box). Box 2.1: Increasing Tobacco Taxes in the Kyrgyz Republic: Results from an Extended-Cost Effectiveness Analysis At present, tobacco taxes in the Kyrgyz Republic are among the lowest in the region. Yet smoking is one of the country’s top causes of mortality, implicated in more than 20 percent of deaths among males. An extended cost-effectiveness analysis was conducted for the Kyrgyz Republic to assess the health, financial, and distributional consequences of raising cigarette taxes, with a focus on financial risk protection. In particular, the authors estimated the number of premature deaths averted, OOP expenditures on tobacco-related diseases saved, and the number of individuals who did not fall below the poverty line as a result of tobacco tax increases. Estimates were made using the WHO-recommendation of a 75 percent tax rate. Given that smoking 68 is heavily concentrated among men in the Kyrgyz Republic, the extended cost-effectiveness analysis was conducted for the male population only. The results indicate that a higher tobacco tax would bring large health and financial benefits to Kyrgyz households and would be pro-poor. Under the 75 percent tax, 104,000 premature deaths would be averted, along with $2.4 million in OOP expenditures and 12,100 new poverty cases. In addition, government savings on tobacco-related health expenditures would amount to $7.3 million. The benefits of tobacco tax increases are concentrated among the bottom 60 percent of the population, with almost 50 percent of averted deaths and 45 percent of OOP expenditures averted accruing to the bottom two consumption quintiles. Given, however, the higher smoking prevalence among the richer quintiles, the price increase would also bring about benefits for the highest quintile in terms of reductions in OOP expenditures related to tobacco-related diseases. I. RECOMMENDATIONS 2.67. Substantial efforts need to be made in the next 10 years to meet the ambitious targets set forth in the government Health Program for 2030. The declared focus on primary care has not yet materialized into significant improvements, including availability of equipment and diagnostics, at the PHC level. Moreover, the program budget for 2020–22 does not reflect the stated priorities of the government’s Health Program. Despite the declared focus on PHC, the share of the MHIF budget allocated to PHC will decline from 37.8 percent in 2020 to 31.8 percent in 2022. Concerted efforts need to be made to reorient the health system toward primary care. 2.68. To ensure effective coverage of the SGBP, there is a need to refine its content and scope so that the benefits are aligned with the country’s health needs and available financial resources. The SGBP revision process has to be supported by well-established structures. Changes to the SGBP should be informed by changes in the burden of disease, the demographic profile of the population and the evidence on cost-effectiveness and efficacy of health treatments, interventions and technology development locally and internationally. The MOH has initiated work on establishing the procedure and process for SGBP revision, supported by the World Bank under the Primary Health Care Quality Improvement Program and also explicitly in the government Health Program. Given the limited fiscal space caused by COVID-19, the government should consider improving the targeting of the SGBP, for example by incorporating explicit poverty targeting within the copayment exemption categories for the SGBP. Revising the benefits package commensurate with the available resources, burden of disease, cost-effectiveness, and financial protection would not only improve the efficiency of government health spending, but also ensure that individuals do not suffer catastrophic health spending. In addition, the ADP for the insured population should be revised and its budget increased to improve effective coverage for priority conditions and related medicines (for hypertension medication) at the primary care level. 2.69. Reprioritization of health within the government budget is needed to combat the rising trend in OOP spending and maintain adequate levels of spending given the macro-fiscal impact of COVID-19. Low levels of domestic government financing mean that there is currently a substantial gap between the available resources and the actual costs of delivering an essential package of quality services for the population. Government health spending has been declining in real per capita terms since 2013, while OOP spending has continued to rise. This threatens government’s progress toward universal health coverage and limits financial protection. To achieve the government Health Program’s objective of reducing the population’s financial burden when seeking health care, the government must prioritize health within the budget and ensure real per capita growth in government health spending. Efficiency gains are important, but a low level of public financing is itself also a source of inefficiency. The government should ensure stable increases in real per capita financing for the health system and avoid large year-on-year fluctuations. 69 2.70. Increase existing health taxes (for example taxes on tobacco and alcohol) and introduce new taxes (for example, taxes on sugar-sweetened beverages and junk food) can be a way to raise revenues despite the bleak macroeconomic outlook due to COVID-19, while at the same time reducing the use of harmful products, curbing the growing NCD burden, and improving health outcomes. These taxes will generate positive health effects in the medium- to long-run, as higher prices are expected to lower consumption and thereby reduce the incidence of non-communicable diseases (for example, diabetes and cardiovascular disease). In addition, this can lead to significant savings from government funds as well as avert OOP expenditures (that is, due to lower hospitalization costs given the decline in the prevalence of NCDs). 2.71. There is some room to further improve the efficiency of the health system. By further investing in strengthening PHC, it is possible to reduce the number of avoidable hospitalizations (that is, those that could have been prevented with adequate primary care). Conditions such as hypertension and asthma could be prevented with information and education and managed effectively at the primary care level without requiring hospitalization, except in complicated cases. The government should continue to prioritize the financing of PHC services and shift away from an over-reliance on inpatient treatment. As recommend in the 2014 PER, the introduction of modern technologies, such as day surgery and laparoscopic surgery, could reduce the number of hospital beds, decrease unnecessary hospitalizations, benefit patients, and improve health outcomes. The installment of modern technologies, however, would require investments in equipment and training. Incentivizing outpatient procedures and day-surgeries through the initiation of new payment methods could further help to improve the efficiency of the system. For example, the revision of the DRG- based reimbursement rates to remunerate day surgery and new technologies with favorable reimbursement rates could help to reduce unnecessary hospitalization. 2.72. Going forward, it is important to maintain the independence of the strategic purchaser while improving its capacity and refining provider payment methods. This includes the revision of PHC capitation payment to incorporate consolidated performance-based payment schemes for quality and incentivize preventive services. In addition, the DRG system should be revised to improve fairness and transparency in hospital payment. It is important to align the incentives created by provider payment methods and those arising as a result of the PFM processes. Enabling providers to use greater autonomy may also improve the efficiency of health spending and quality of care. 2.73. Further investments are needed to develop data systems to monitor quality, access, and efficiency, and build capacity to use these data for decision-making purposes. The government should invest in and build the capacity to analyze the use of resources in the health system to adjust the payment rates and benefits packages. The Kyrgyz Republic has national data collection for health statistics and health insurance, both of which are partially automated and have had limited use for quality improvement. Setting up a new quality of care data collection and monitoring system will require several key elements, such as an e- platform, automated data collection from facilities, and data exchange systems/capabilities. Under the Primary Health Care Quality Improvement Program, the government plans to establish a quality of care e-platform to collect, monitor, and report of data at the primary care level. Such systems should be scaled up in the future to include all levels of care. 70 Annex: Progress of policy recommendations for health sector proposed in 2014 Public Expenditure Review (PER) Health Progress made/emerging issues Reduce the scope and improve the targeting of co-payment exemptions provided No progress in the revision of the scope and by the State Guaranteed Benefit Package (SGBP). The current policy of co- targeting of the SGBP. Remains relevant and payment exemptions based on 28 privileged social categories and priority medical critical. conditions does not favor the poor: 57.5 percent of beneficiaries are not in the bottom two quintiles of the income distribution and 51.3 percent of the bottom two quintiles are not exempted from co-payment. A much stronger focus on the poor, and potential consolidation of exemptions could be achieved using a mean tested approach—such as the one used by the Monthly Benefit for Poor Families— or a proxy means test methodology Optimization and modernization of hospital services. The introduction of modern The number of hospitals has been declining. technologies, such as day surgery and laparoscopic surgery, could reduce the The government has completed a master plan number of hospital beds, unnecessary hospitalizations, benefit patients, and for the optimization and modernization of improve health outcomes. However, costs would not necessarily be reduced, as the hospital services in 2020. Important to ensure introduction of modern technologies would require investments in equipment and steadfast implementation of the master plan training. In addition, other measures should be introduced to better manage and recommendations (also one of the pillars hospitalizations and control costs. On the supply side, introduction of volume under the government Health Program for ceilings at hospital level, as well as the revision of the DRG-based reimbursement 2030). Payment reforms not yet introduced rates to remunerate day surgery and new technologies with favorable but are planned under the Health Program, reimbursement rates could help. On the demand side, the selective reintroduction and primary care payment reforms are of co-payment for children, pensioners, and pregnant women that were abolished supported by the World Bank under Primary in 2006 would also benefit the situation. Health Care Quality Improvement Program. Improve cost-effectiveness of management and prevention of priority diseases. Limited progress. Tobacco taxes have Hospital-based care for CVDs is expensive and ineffective. On the other hand, increased but remain below the improved management and prevention of CVDs can reduce costs and save lives. recommended 70 percent rate. Mass media educational campaigns, risk factor screening at Primary Health Care (PHC), and pharmaceutical secondary prevention are cost-effective interventions. Tobacco taxation is a particularly effective measure to reduce morbidity related to CVDs and bring additional revenue to the state budget. Self-financing of dental and balneology services. About three percent of total Not explored in this PER. government health expenditure is used to subsidize dental services, balneology, and rehabilitation services. However, free services are enjoyed by a limited number of population groups. Therefore, private provision of these services and public provision of targeted subsidies would be a more efficient and equitable alternative. Harmonization of the rules governing the use of the various financing sources. The SGBP is financed by four sources: budget funds, MHI/Social Fund payroll contribution, patient co-payment, and special means. Each of these sources has different spending rules. The harmonization of spending and reporting rules would produce some immediate, but small saving from reduction of staff positions in accounting and more efficient use of liquidity. In addition, important benefits and potential sources of efficiency would derive from the improved flexibility that would be gained in the use of budget funds that at the moment are only allowed for MHI/Social Fund contribution. Even if it is difficult to quantify the monetary value of these benefits, this measure could lead to significant fiscal space for the health sector. Reallocating savings from unfilled positions to other use. Health facilities have a Salaries in the health sector remain low and large number of vacant positions and the practice is to redistribute savings from are significantly lower than the average salary unfilled positions to other staff. The practice of redistributing savings from unfilled in public administration (and education). positions to other staff has raised a number of efficiency concerns. The salary increases introduced in 2011 have squeezed the health budget, severely reducing resources for other key inputs, such as food and medicine. The revision of staff norms to reduce the number of unfilled position could reduce the problem. In 71 addition, training should be provided to the Chamber of Accounts and facility managers to improve the use of the resources and flexibility provided by the single payer system. Improve the procurement of hospital drugs and medical supplies. The use of Framework agreements not yet introduced. “framework agreements� could produce the efficiency saving and quality Hospitals can directly purchase supplies improvement of centralized procurement, while maintaining the flexibility of the through the government procurement site. current system where each facility purchases its own drugs. Under the proposed scheme, the MHIF would determine the demand, in order to negotiate price and quality with suppliers, so that facilities could purchase under these conditions defined in the “framework agreements.� However, the introduction of “framework agreements� would require time and enhanced capacity, both at the MHIF and Drugs Supply and Medical Technology at the MoH. Table 2.9. Summary of policy recommendations Priority Policy Recommendations Time Frame Expected Impact Health Increase existing health taxes (for example, tobacco Short term Positive effects on health and and alcohol) and introduce new taxes (for example, fiscal savings taxes on sugar-sweetened beverages and junk food) to simultaneously reduce the burden of NCDs and increase revenue. Refine SGBP (State Guarantee Benefit Package) content and Short to medium term Effective coverage of the SGBP scope so that the benefits are aligned with the country’s health needs and available financial resources. Prioritize health within the government budget to maintain Medium term Stable increase in real per capita adequate levels of spending to respond to COVID-19, ensure financing for health continuity in service delivery, and combat the rising trend in OOP spending. Reduce the number of avoidable hospitalizations. Medium term Improved quality of primary healthcare service Invest in primary healthcare by prioritizing financing of PHC Medium term Enhanced efficiency of the health services with the introduction of new payment method. system Maintain the independence of the strategic purchaser; while Medium term Improved efficiency of health improving its capacity and refining provider payment methods. spending and quality care Develop data systems to monitor quality, access, and efficiency, Medium term Improved capacity to analyze the and build capacity to use these data for decision-making use of health system resources purposes. 72 References Ahmedov, M., H.T.H. Nguyen, I. Postolovska, and others. 2020. “Assessing Quality of Care in the Kyrgyz Republic: Quality Gaps and Way Forward.� Washington, DC: World Bank. Jakab M, B. Akkazieva, and J. Habicht. 2018. “Can People Afford to Pay for Health Care? New Evidence on Financial Protection in Kyrgyzstan.� Copenhagen: WHO Regional Office for Europe. “Kyrgyzstan Multiple Indicator Cluster Survey 2018,� Survey Findings Report. 2019. Bishkek, Kyrgyzstan: National Statistical Committee of the Kyrgyz Republic and UNICEF. Nguyen, H.T.H., and T. Strizrep. 2019. Toward a More Pro-Poor and Explicit Health Benefit Package in the Kyrgyz Republic: A Critical Review of the State Guaranteed Benefit Package and Options for Its Revision. Washington, DC: World Bank. Postolovska, I., H.T.H. Nguyen, A. Sargaldakova, and others. 2018. “An Extended Cost-Effectiveness Analysis of Tobacco Price Increases in the Kyrgyz Republic.� Washington, DC: World Bank. Schnell, O., J.B. Crocker, and J. Weng, 2017. “Impact of HbA1c Testing at Point of Care on Diabetes Management.� Journal of Diabetes Science and Technology, 11(3), 611–17. Tandon, A., J. Cain, C. Kurowski, and others. 2020a. “From Slippery Slopes to Steep Hills: Contrasting Landscapes of Economic Growth and Public Spending for Health.� Social Science & Medicine, 259, 113- 171. Tandon, A., T. Roubal, L. McDonald, and others. 2020b. “Economic Impact of COVID-19: Implications for Health Financing in Asia and Pacific.� Health, Nutrition and Population Discussion Paper. Washington, DC: World Bank. Watkins, D.A., J. Qi, Y. Kawakatsu, and others. 2020. “Resource Requirements for Essential Universal Health Coverage: A Modelling Study Based on Findings from Disease Control Priorities.� The Lancet Global Health, 8(6), e829-e839. WHO. 2019. “Governance of the Health Financing and Strategic Purchasing of Services in Kyrgyzstan.� Copenhagen: WHO Regional Office for Europe. ———. 2018. “A Scoping Review on Health Services Delivery in Kyrgyzstan: What Does the Evidence Tell Us?� Copenhagen: WHO Regional Office for Europe. World Bank. 2017. “Tobacco Tax Reform at the Crossroads of Health and Development.� Washington, DC: World Bank. ———. 2019. “High-Performance Health Financing for Universal Health Coverage: Driving Sustainable, Inclusive Growth in the 21st Century.� Washington, DC: World Bank. ———. 2020a. “Costing of Health Services in Kyrgyzstan: An Exploratory Approach.� Washington, DC: World Bank. ——— 2020b. “COVID-19 and Human Capital: Europe and Central Asia Economic Update.� Washington, DC: World Bank. ———. 2020c. “Taxes on Sugar-Sweetened Beverages: Summary of International Evidence and Experiences.� Washington, DC: World Bank. 73 3. CHAPTER 3: PENSIONS Over the past five years, the pension system has been financially stable and provided decent support for pensioners. Pensioners account for just over 10 percent of the country’s population, and the old -age pension system covers 85 percent of people older than the pensionable age, with an average pension higher than the actual subsistence minimum for pensioners. Pension spending amounts to 8 percent of GDP, which is on the high end for countries with similar income per capita. However, the system is starting to come under pressure, as the number of contributors has not grown despite a substantial increase in the working-age population during this period, largely because of growing informality. Moreover, the system relies on contributions with a heavy burden on the public and organized formal sector employees, while those involved in the non-organized sector contribute very small amounts. Some of the key objectives of the government in reforming the national pension system reflect concerns around the growing informality and falling benefit adequacy. Those issues are related, as small pensions do not provide any incentives to participate in the system. The government launched major structural reform in 1997 by introducing a notional defined contribution (NDC) system with the intent of strengthening the link between contributions and benefits and making the system fiscally sustainable. However, some important design elements were weakly defined, as noted in the World Bank’s 2014 Public Expenditure Review (PER) report, which resulted in rapidly declining benefit levels in the new system. To address that issue, important legal amendments were made in 2017, which set forth the frequency and criteria for indexation of both the defined benefit (DB) and the NDC plans of the pension and the parameters for indexing the NDC balances. The indexation of both is still lagging the dynamics of wages, contributing to low benefits and ultimately to growing significant risks of poverty among the retirees. Currently, the new pensions are almost half the average of those accrued in the past. And almost 90 percent of the newly awarded pensions are below the subsistence minimum, compared with 30 percent for the current pensioners. Over the next 30 years, if the pension system parameters stay unchanged, the coverage of the pension-age population is projected to decline to 63 percent due to the low participation rate of the current working-age population. This will result in considerably increasing expenditures on the programs alleviating poverty among the elderly to compensate for the falling efficiency of the pension program. There is a need, therefore, for a parametric and structural reform of the benefit package to ensure that the system is socially sustainable. The National Development Strategy (NDS) 2018-40 outlines the government objectives to improve the pension system. The government priorities include in particular: 1) reform the pension system to ensure its financial sustainability; 2) modernize the insurance component of the pension system to accommodate the needs of different age groups; and 3) streamline the early retirement provisions and increase the efficiency of the funded component of the pension system. The chapter presents results of simulation of the parametric changes in to the old-age pensions and concludes with reform recommendations. The recommendations focus on improving the indexation of the basic and insurance parts of the pension to ensure the pension adequacy and on enhancing the design on the pension system to support individuals outside of the formal sector who lack opportunities to accumulate enough pension rights by retirement for a proper insurance benefit. A. INTRODUCTION 3.1. This chapter evaluates the pension system in the Kyrgyz Republic, building on the 2014 PER analysis. The 2014 PER analysis pointed to several deficiencies in the national pension system and recommended certain parametric and structural reform options. Specifically, to address the issue of benefit adequacy, the analysis recommended: (1) introducing pension benefit indexation at the rate of 50 percent wages and 50 percent inflation; (2) valorizing NDC accounts, going forward, with the same rate of 50 percent wages and 50 percent inflation; and (3) retroactively valorizing all NDC account back to the date of the start of the reforms in 1996. The government adopted benefit indexation in (1), introduced valorization under (2) but with a lower than recommended rate (inflation only), and did not address the legacy issue recommended in (3). In this current PER report, the fiscal projections were updated using a new World Bank tool of pension 74 policy microsimulations and the anonymous individual records, kindly provided by the Social Fund. The analysis focuses on implications in the status quo and under several reform scenarios considered by our counterparts for the reform, assessing long-term fiscal balances, social equity, benefit adequacy, coverage, and administrative efficiency. Findings of this analysis, in particular, confirm these recommendations of the PER work in 2014 and indicate that the changes introduced by the government were indeed not sufficient. There were several other recommendations of the 2014 PER report, suggesting structural changes, which this work confirms in principle but does not explore in detail. 3.2. The chapter is organized as follows: Section B provides a brief overview of the performance of the pension system and the main outstanding issues. Section C presents the results of the fiscal projections for the current system in terms of fiscal balances and social indicators. Section D assesses the effects of parametric changes; five parametric changes have been simulated. Section E offers reform recommendations. B. OVERVIEW OF THE PENSION SYSTEM 3.3. The pension system in the Kyrgyz Republic consists of several pillars (Table 3.1). The reform of 1997 transformed the old pay-as-you-Go (PAYG) system into three separate components: the basic part of pension, the insurance part of pension under the DB pension plan, and the insurance part of pension under the NDC pension plan. Each part is calculated based on its own rules, whereby these parts are added and combined into one payment. The basic part of the pension is a state-guaranteed benefit financed by the republican budget. The insurance pensions under the DB and the NDC plans are financed by contributions paid by employers, employees, and the self-employed. The pensions under the DB plan are calculated using the insured service before 1996, while the pensions under the NDC plan use the insured service starting from January 1, 1996, when the individual record management system was implemented. A supplementary mandatory funded pillar was implemented for all employees of the formal sector in 2010, with a contribution of 2 percent of the wage. Starting from January 1, 2012, men born before January 1, 1964, and women born before January 1, 1969, who have only 15 years before reaching their full retirement age have been allowed to opt out of mandatory enrollment. The detailed description of the rules and parameters of the pension system as well as its institutional setup are described in detail in Annex 2. Table 3.1. Design of the pension system in the Kyrgyz Republic Reform of 1997 Reform of 2010 PAYG System Funded Pillar Basic part of the Insurance pension Insurance pension Mandatory Voluntary pension under the DB plan for under the NDC plan component insurance service periods accrued for contributions before 1996 starting from January 1, 1996 Financed by the Financed by contributions paid by the 2 percent Voluntary additional transfers from the employer at 15 percent, and the worker at 8 contribution rate for contributions paid republican budget percent or 10 percent depending on the date those born after by the enrolled of birth; reduced rates for the self-employed January 1, 1964 persons and some categories (men), and January 1, and the 1969 (women) employers Source: Kyrgyz Republic Law No. 57, dated July 21, 1997, On State Pension Social Insurance, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru- ru/557/410?cl=ru-ru. Note: See Annex 2 for details. 75 3.4. In 2017, important amendments were made in the Law on the State Pension Social Insurance, which clearly set forth the frequency and criteria for indexing the basic and insurance part of the pension. It made the procedures for assigning the pension easier and established more objective parameters for indexing the balances on individual accounts of the insured persons in the NDC plan. 3.5. The retirement age is 63/58 for men/women correspondingly. However, there is an extensive list of categories of citizens who are entitled to early retirement. To receive a full state basic part of the pension, men and women need a total of 25/20 qualifying years of contributions correspondingly. At the same time, any person is eligible to claim a reduced basic part of the pension if contributions were paid at least once. Qualifying conditions for the insurance pension in the Kyrgyz Republic are unique in terms of their flexibility, providing entitlement to a pension for the maximum number of persons. However, there is a downside to this flexibility, as insurance pensions under the DB and NDC plans have no guaranteed minimum amount, while the full basic part of the pension is guaranteed only for those who have the complete qualifying years of contributions. 3.6. Despite the expanding cohorts of the young generation, the number of persons who pay insurance contributions does not demonstrate a tendency to increase. In 2015–18, the number of insurance contribution payers varied between 1.3 million and 1.5 million people (Table 3.2). The only category that showed a steady increase is public sector employees. Table 3.2. Number of insurance contribution payers at year-end (persons) Contributor categories 2015 2016 2017 2018 Total 1,380,156 1,342,200 1,493,621 1,351,556 Non-public sector employees 374,949 295,895 278,832 278,819 Public sector employees 316,719 318,098 325,515 325,547 Workers of agricultural cooperatives 4,376 3,589 3,404 3,179 Members of peasant farms 584,236 624,153 791,605 643,788 Individual entrepreneurs 99,876 100 465 94,265 100,223 Source: Kyrgyz Republic Social Fund. 3.7. The largest category of contributors is members of peasant farms. They pay contributions at symbolic rates depending on the size of the plot and fertility of their lands. Based on the 2018 data from the Social Fund, the average monthly contribution from this category was 26 soms, whereas employees paid 4,046 soms; the share of peasant farms in total contributions collected is around 1 percent. Reduced contribution rates set for peasant farms and other categories of the self-employed categories are intended to lower the fiscal burden. However, under such circumstances, it is impossible to accrue a reasonable amount of pension capital and corresponding benefits under the NDC plan. De facto, the self-employed can rely only on the part of the pension that depends on their insurance service. This means that as the insurance DB plan is gradually phased out over time, the size of their benefit will be determined largely by the basic part of the pension rather than the insurance part. This creates significant risks of poverty among these pensioners in the future. 3.8. The next largest category of contributors is public and non-public sector workers. These workers account for 45 percent of contributors, while the share of their contributions in the total pool of contributions is more than 95 percent. The aggregate rate of contributions paid by employers and employees (23–25 percent) is close to the rates in other former Soviet republics that have contributory PAYG pension plans. The contribution rate to the PAYG system in Azerbaijan, Moldova, the Russian Federation, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan is within the 19–25 percent range; in Belarus, the contribution rate is 76 29 percent. A principal difference, however, is that all or almost all of the contributions in those countries are paid by employers, while the employee contribution rate is set at zero or is rather symbolic (1 percent in Belarus, 3 percent in Azerbaijan, and about 5 percent in Moldova)34. 3.9. In the past 20 years, the contribution rate paid by employers in the Kyrgyz Republic has been gradually reduced, whereas the employee contribution rate has been on the rise. Currently, employers are paying the contribution at a rate of 15 percent, out of which 12 percent is registered in the individual accounts of the employees in the NDC plan, with 3 percent going to the budget of the PAYG system. Employees pay a contribution of 8 or 10 percent to the NDC plan, depending on their eligibility for the mandatory funded pillar. 3.10. There are different ways for members of peasant farms to participate in the pension insurance system. One option is to exclude them from the contributory mandate. However, this decision will have certain negative implications. First, the coverage rate of the insurance contribution will be halved. In 2018, the ratio of contributors to the total working-age population was 37 percent, and to the total employment, it was 57 percent.35 Even though the contribution payments may stop, the number of pensioners from peasant farms will start declining only in the longer term. Second, the reduced coverage of the old-age population with insurance pensions will require higher expenditures on social assistance. Alternatively, the members of the peasant farms could be transitioned to some simple new DB scheme designed specifically for them. 3.11. While the younger cohorts dominate both in the population structure and in the number of the contributors, the participation rate in pension insurance is the highest among persons of the preretirement age (Figure 3.1). The contribution coverage rate among the population ages 20–44 is 35–45 percent, while it increases drastically among the population of the preretirement age, especially men. This characteristic of the participation curve can be linked to flexible pension eligibility rules that provide entitlement to pension with minimal contributions. When approaching retirement, individuals begin to take greater interest in the pension eligibility and benefit rules. 34 World Social Protection Report 2017–19: Universal Social Protection to Achieve the Sustainable Development Goals, statistical annexes. https://www.social-protection.org/gimi/gess/ShowWiki.action?id=594#tabs-3. 35 Calculated with data from the Annual Statistical Book of the Republic of Kyrgyzstan, 2014–2018, National Statistical Committee of the Republic of Kyrgyzstan, Bishkek, 2019. http://www.stat.kg/media/publicationarchive/c341067d-1b0e-43da- a650- d69191f43f4b.rar. 77 Figure 3.1. Distribution of contributors by age Figure 3.2. Share of contributors among population, (persons) by age (percent) Source: Kyrgyz Republic Social Fund. 3.12. The recent increase in the pensioner numbers is largely explained by the population growth. In 2019, the total number of pensioners was 670,435, having increased by 80,081 persons, or 13.6 percent, over the past five years (Table 3.3). This increase was driven mostly by the increase in the number of old-age pensioners and pensioners with disabilities. Table 3.3. Number of pensioners in the Kyrgyz Republic at year-end (persons) Types of pensions 2014 2015 2016 2017 2018 2019 Total 590,354 604,791 617,350 633,867 651,155 670,435 Old-age pension 436,849 448,618 460,542 477,184 493,878 511,577 with incomplete record 9,980 12,682 14,780 19,354 22,129 25,343 early retirement because of 74,648 73,616 71,387 70,068 66,935 61,815 special working/living conditions early retirement because of long 53,345 37,845 50,465 40,589 45,974 42,189 service Disability pension 106,346 110,144 111,443 113,928 116,738 119,776 Survivor’s pension 45,551 44,463 43,836 41,305 39,110 37,692 Military disability and survivor’s 1,608 1,566 1,529 1,450 1,429 1,390 pensions Source: Kyrgyz Republic Social Fund reports, form 94-Sots. Note: Without beneficiaries of long-service military pensions and beneficiaries of the social assistance. 3.13. Pensioners account for 10.3 percent of the country’s population; 69 percent of pensioners live in rural areas, which is consistent with the percentage share of rural population. The share of the old-age pensioners in the total number of pensioners is 76 percent; while their share in the population above the statutory retirement age (58+ for women and 63+ for men) is 88 percent36. A high level of population coverage with old-age pensions is achieved due to the minimal qualifying requirements for the insurance 36 Calculated using the Social Fund individual data on beneficiaries and the United Nations World Population Prospects data for Kyrgyz Republic. 78 pension. Besides the pensioners who reached the retirement age, around 40 percent of new old age pension beneficiaries retire early, since many categories of workers are entitled to early retirement37. 3.14. Over the past five years, the number of early retirees has gone down by 19 percent; their share of the total number of old-age pensioners has dropped from 29 percent to 20 percent. However, in 2019, the percentage of new old-age pensions awarded before the retirement age was still high, at 59 percent (compared with 81 percent in 2014). The impact of early retirement pensions has led to a situation where the actual average retirement age is 1.5 years below the statutory retirement age. Notably, the number of pensioners with incomplete service (fewer than 20 years for women and fewer than 25 years for men) has gone up 2.5 times and is now 5 percent of the total number of old-age pensioners and 16 percent of the new retirees. 3.15. Seventy-two percent of old-age pensioners are women. This high gender imbalance occurs not only because of the longer life expectancy and lower retirement age for women, but also because there are many categories of women with entitlement to early retirement. The most common categories among those with the right to retire before their full retirement age are mothers with many children and mothers with disabled children (29,780 persons, or 48 percent of the total number of people with the right to retire early) as well as mothers with three or more children living in high mountain areas (14,700 persons, or 24 percent of the total number of people with the right to retire early). While the birth rate in the Kyrgyz Republic remains high, the aggregate birth rate is consistently higher than three children per woman38. 3.16. In 2019, the average old-age pension benefit was 5,935 soms (Table 3.4), which is by one third higher than the formal pensioner’s subsistence minimum. Table 3.4. Average size of the monthly pension in the Kyrgyz Republic at year-end (soms) Types of pensions 2014 2015 2016 2017 2018 2019 Total 4,315 4,510 4,835 5,172 5,368 5,553 Old-age pension 4,683 4,900 5,239 5,550 5,756 5,935 with incomplete record 2,786 2,657 2,701 2,826 2,836 2,786 early retirement because of special 3,427 3,637 3,674 3,945 4,048 3,921 working/living conditions early retirement because 3,138 3,247 3,588 4,147 4,006 4,028 of long work record Disability pension 3,381 3,523 3,823 4,218 4,377 4,537 Survivor’s pension 2,955 2,976 3,127 3,394 3,390 3,558 Military disability and survivor’s 4,698 5,787 5,829 6,230 6,351 6,461 pensions Pensioner’s subsistence minimum 4,434 4,637 4,304 4,393 4,283 4,287 Source: Kyrgyz Republic Social Fund reports, form 94-Sots. Note: The numbers do not include supplements and allowances financed by the republican budget. 3.17. In 2017, the rules of indexing pensions were restated more clearly, with the objective of linking benefits to the macroeconomic changes over time. The established rule for indexing insurance pensions (50 percent of inflation + 50 percent of increase in wages), while not uncommon, put pensioners at 37 Calculated using the Social Fund individual data on beneficiaries. 38 Annual Statistical Book of the Republic of Kyrgyzstan, 2014–2018, National Statistics Committee of the Republic of Kyrgyzstan, Bishkek, 2019, http://www.stat.kg/media/publicationarchive/c341067d-1b0e-43da-a650-d69191f43f4b.rar. 79 a disadvantage, especially when the higher increase in wages coincided with a lower inflation rate. Another new rule stipulated that if the total size of the pension exceeds the pensioner’s actual subsistence minimum, the basic part is not indexed. As a result, by the end of 2019, the ratio of average old-age pension to current wages went down to 35 percent from 38–40 percent in 2015–18. 3.18. The gender gap in the benefit rate is far lower than the gender gap in wages (14 percent against 28 percent)39. That is because the size of the pension is determined by the components that are more dependent on the insurance service than on the wage. The average size of the old-age pension is 5,672 soms for women and 6,616 soms for men. However, as the DB pensions have been gradually replaced by the NDC pensions, the gender pension gap may worsen considering a five-year difference in the retirement age and, therefore, a shorter period for notional capital accumulation for women. 3.19. Because of the ongoing transition from the DB plan to the NDC, the level of the newly assigned pensions is already almost two times lower than that of the existing old-age pension population. This signals a big problem. The average size of the pensions assigned in 2019 was 3,096 soms whereas the average pension size was 5,935 soms. Fourteen percent of new retirees received a pension of 2,000 soms or less (not exceeding the size of the basic part40), 50 percent received 2,001–3,000 soms, and another 20 percent received 3,001–4,000 soms (Figure 3.3). It means that almost 90 percent of new pensioners received pensions below the subsistence minimum. To some extent, the gap between the new and the existing pensions is explained by the structural changes in the economy and the fact that new cohorts of pensioners have shorter “Soviet period� service. However, this also indicates the need for further reforms within the NDC component. 39 Ratio of women’s wage to men’s wage, Kyrgyz Republic National Statistical Committee website, http://www.stat.kg/ru/statistics/download/dynamic/643/ 40 The size of the basic pension in 2019 was 1880 som per month. 80 Figure 3.3. Distribution of old-age pensioners by pension size at the end of 2019 (percent) All old-age pensioners New old-age pensioners 10001+ 10001+ 2000 or less 5001- 2001-3000 1% 2000 or less 5% 2% 10000 9% 7% 14% 4001- 5000 8% 3001-4000 14% 3001-4000 5001-10000 20% 56% 4001- 5000 2001-3000 14% 50% Source: Kyrgyz Republic Social Fund reports, form 94-Sots. Note: Without supplements financed by the republican budget. 3.20. Budget of the PAYG system. In 2014–19, the budget of the PAYG system was characterized by stability and fiscal balance (Table 3.5). Revenues were generated from two main sources: collected contributions (almost 5 percent of GDP) and the transfer from the republican budget (around 3.5 percent of GDP). Insurance contributions are used to finance payments of the insurance part of the pension (about 4.5 percent of GDP), funeral allowance (0.1 percent of GDP), and administrative costs, including payments for services to deliver pensions and maintain the Social Fund (0.2 percent of GDP). The proceeds from the republican budget finance payments of the basic part of the pension (about 2 percent of GDP) and other state pension programs administered by the Social Fund. 3.21. Expenditures to finance state pension programs (military disability pensions and military survivor’s pensions, electricity compensations as well as other pensions and supplements) have been gradually decreasing due to the decisions to harmonize rules and reduce privileges. It was one of the factors that contributed to the reduction in total expenditures of the PAYG system from 8.6 percent of GDP in 2014 to 8.0 –8.1 percent of GDP in 2017–18. 81 Table 3.5. Budget of the PAYG system in the Kyrgyz Republic (percent of GDP) Budget items 2014 2015 2016 2017 2018 2019 Total revenues 8.8 8.8 8.6 8.2 8.4 8.4 Collected contributions 4.8 4.9 5.0 4.8 4.8 4.7 Transfer from the republican budget 3.9 3.9 3.5 3.4 3.6 3.6 Other revenues 0.1 0.0 0.1 0.0 0.1 0.1 Total expenditures 8.7 8.6 8.5 8.2 8.3 8.3 Basic part of the pension 2.0 2.0 1.9 1.9 2.1 2.1 Insurance part of the pension 4.5 4.5 4.6 4.5 4.4 4.5 Military disability pension and survivor’s 0.4 0.5 0.4 0.4 0.4 0.5 pension41 Electricity compensations 0.5 0.4 0.4 0.4 0.3 0.3 Other state supplements to the pension 0.9 0.9 0.8 0.7 0.7 0.6 Funeral allowance 0.1 0.1 0.1 0.1 0.1 0.1 Administrative costs 0.3 0.3 0.3 0.2 0.2 0.2 Current balance (revenues minus 0.1 0.2 0.1 0.0 0.1 0.1 expenditures) Source: Kyrgyz Republic Social Fund. Note: Without expenditures to long service military pensions and social benefits. The data for 2019 have been calculated based on the preliminary size of GDP and therefore may be slightly overstated. 3.22. The Social Fund has been operating the system quite efficiently. Administrative costs account for 1.6 percent of all costs of the fund, or 0.2 percent of GDP. The World Bank has developed a systematic tool to assess whether human and financial resources are adequate to address the tasks assigned to the entity that administers the national pension system42. The assessment indicates that before the collection function was delegated to the State Tax Service, the Social Fund staff (about 1,900 employees) had been 90 percent of what was predicted by the tool (with 10 percent the margin of error). It means that total resources were fully consistent with the tasks implemented by the Social Fund. When the collection function was delegated to the State Tax Service, the total number of staff members went down and now stands at 1,580. At the same time, about 300 Social Fund employees continue fulfilling auxiliary functions related to the collection of contributions and keeping records on the individual accounts, which makes sense for such a system. Administrative costs of the Social Fund fall within 90–100 percent of what has been predicted by the tool, without considering the investment function that is in part administered by the Social Fund. The analysis demonstrates that administrative resources are used efficiently. 3.23. In 2014–19, nominal annual contributions to the mandatory funded pillar increased from 1.1 billion to 1.8 billion soms (Table 3.6). The contributions increased faster than wages (by 67 percent and 40 percent, respectively) because of new enrolled persons. However, in percentage of GDP, the collected contributions increased only slightly, from 0.26 percent to 0.3 percent of GDP. At this participation rate and the contribution rate, this relative level is close to the maximum level, as it already covers almost all cohorts of contributors, except for the last preretirement cohorts released from mandatory enrollment because of the cutoff age. Table 3.6. Indicators of the mandatory funded pillar in the Kyrgyz Republic 41 Main military pensions are not included here. This includes only disability and survivor’s pensions, paid from the Social Fund. 42 This tool is a part of the toolkit to study administrative efficiency of pension agencies, Social Insurance Administrative Diagnostics (SIAD), World Bank, 2020. 82 Indicators 2014 2015 2016 2017 2018 2019 Collected contributions during the year Billions soms 1.1 1.2 1.3 1.5 1.6 1.8 percent of GDP 0.26 0.27 0.27 0.28 0.28 0.30 Assets by the end of the year Billions soms 6.3 8.4 10.9 13.8 17.0 20.7 percent of GDP 1.56 1,94 2.28 2.60 2.99 3.51 Payments during the year Billions soms 0.09 0.04 0.05 0.09 0.19 0.25 percent of GDP 0.02 0.01 0.01 0.02 0.03 0.04 Source: Kyrgyz Republic Social Fund budget reports. 3.24. In nominal terms, the pension assets increased from 6.3 billion to 20.7 billion soms and, in percentage of GDP, from 1.56 to 3.51 percent. The main instruments for investing pension funds are predominantly five-year, seven-year, and 10-year government securities with an average nominal return of 14– 16 percent. By the end of 2019, 92 percent of all assets were invested in these securities. Funds in fixed-term deposits of the commercial banks account for 6.8 percent of total assets; these are short-term deposits for one to 1½ years that yield on average nominal returns of 10–11 percent. 3.25. Payments from the mandatory funded pillar began in 2013. The law currently stipulates two payout regimes: scheduled withdrawals and a lump sum. Also, the law has been amended recently to allow contingency withdrawals in certain cases (terminal illnesses, house purchase financing, and so on). So far, 98 percent of expenditures have been used to make payments in the form of lump sums for new pensioners, mostly men born before January 1, 1964, and women born before January 1, 1969, who paid contributions in 2010–11 and then were not allowed to participate in the funded pillar. The amount of payments is so far small, but it has been increasing both in nominal terms and as a percentage of GDP. Benefit payments to the participants who contributed to the funded pillar for an extended period of time are expected to start in 2027, when the first eligible cohorts reach the retirement age. 3.26. In 2019, those enrolled in the mandatory funded pillar were able to choose whether to keep their funds in the State Funded Pension Fund or transfer their pension assets to a private company for the first time. In 2019, 342 persons and 9.4 million soms of their assets were transferred to private pension funds. So far, there are only two funds of this type: the Open Joint Stock Company “Dordoi Garant Pension Fund� and the Open Joint Stock Company “Kyrgyzstan Pension Fund� (which had been the only non -state pension fund in the country for a long time). C. FISCAL FORECASTS OF THE EXISTING SYSTEM (STATUS QUO SCENARIO) 3.27. The results presented in this chapter are unique in that they are based on a new microsimulations pension modeling tool using anonymized individual records. This was made possible thanks to the data provided by the Social Fund. 3.28. The forecast baseline year is 2018, with a time horizon of 2020–50. It does not extend to the long service military pensions and social assistance paid to persons without insurance service, as they are not included in the general pension system. The general pension system provides old age pension, disability and survivorship benefits. The following discussion focuses on old-age pension benefit projections, which are based on the 83 demographic assumptions of the medium variant of the World Population Prospects, published by the Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat43. The macroeconomic assumptions were made using the forecast of the World Bank’s Kyrgyzstan office. The status quo scenario implies that the design and parameters of the pension system will remain unchanged at the baseline period level. This status allows the identification of system weaknesses and provides a benchmark for assessing the effect of various measures related to the pension reform. A detailed list of assumptions appears in Annex 2. 3.29. During the period up to 2050, the population of the Kyrgyz Republic will increase by 42 percent, reaching close to 9.1 million people. The population will grow due to the expansion of the working- age cohorts (up by 41 percent, from 3.9 million to 5.5 million people) and pension-age cohorts (an increase by 2.7 times, from 0.5 million to 1.4 million people). However, the number of children will remain relatively stable (approximately 2 million to 2.3 million), which will result in a slower growth of working-age generations in the future (Figure 3.4). 3.30. The share of the pension age population in the total age structure will increase from the current 8 percent to 15 percent. The ratio between the working-age cohorts and pension-age cohorts (the so-called “support ratio�) will decrease by half. In 2019, it was eight working-age people for one pension-age person; by 2050, it will be four. Nevertheless, demographic aging will not result in the proportionally increasing coverage of elderly population by contributory pension benefits in the future, since significant proportion of the working age population does not participate in the pension scheme. 3.31. Currently, 85 percent of the elderly are in receipt of the old-age pensions (excluding early retirement pensions), which reflects “Soviet period� service, when employment was mandatory and formal. But even among today’s generation of elderly people, 11.7 percent have never been employed or run their own business.44 Since currently the contributors account for slightly more than one third of the working- age population (that is, two out of three persons do not work, or work informally and without social insurance), an increasingly higher number of people will be reaching pension age with a very short length of service or even without any pensionable service at all. 3.32. Taking into account the existing patterns of contribution, a rapid expansion of the pension- age population will be accompanied not only by an increasing number of old-age pensioners (from 512,000 to 917,000 people), but also by a growing number of people who are not entitled to a pension (from 58,000 to 511,000 persons) (Figure 3.5). Consequently, the coverage of the elderly with old-age pensions will go down from the current 85 percent to an estimated 63 percent. The share of pensioners with an incomplete length of service will near 50 percent.45 The average insured length of service at the time of retirement for the new old-age pensioners will decrease from today’s 30 years to 23 years. This will have dramatic consequences for the pension system and the wellbeing of future pensioners. 43 World Population Prospects 2019, United Nations, Department of Economic and Social Affairs, Population Division, 2019. Online Edition. Rev. 1: https://population.un.org/wpp/. 44 Older People in the Kyrgyz Republic, Statistical Survey Report, National Statistics Committee of the Kyrgyz Republic, SIAR research and consulting, Bishkek, 2017. 45 Incomplete length of service (LOS) is defined as less than 25 years for males and less than 20 years for females. Individuals with incomplete LOS are entitled to a proportion of basic pension commensurate with a proportion of completed service years in 25 or 20 years that are required to receive full basic pension. 84 Figure 3.4. Projected number and age structure of Figure 3.5. Projected coverage of older people with population old-age pensions (million) (million) Older than working age 10 Old age pensioners ret.age+ Non-covered ret.age+ Working-age population 1.6 9 Under working age 1.4 8 1.2 7 6 1.0 5 0.8 4 0.6 3 0.4 2 0.2 1 - 0 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 Source: Micro simulation results. 3.33. The number of old-age pensioners will grow at a slower rate than the pension-age population, resulting in a reduction in the contributory pension coverage. The reduction in the old- age pension coverage will bring about a greater risk of poverty among the elderly, as well as social tensions. Lower contributory pension coverage and expenditure will require increasing social assistance expenditure to reduce poverty in old age. 3.34. Given the ongoing transition in the structure of the benefit package, the basic part of the pension plays a role of a stabilizing element in the total pension amount. However, the projections show a reduction in the ratio of the basic part of the pension to average wages (Figure 3.6). For the newly awarded pensions, a reduction in the basic replacement rate from the current 9.5 percent down to 7 percent will be due to an increase in the share of pensioners with an incomplete length of service and a proportionally reduced basic part. Starting from 2018, the annual indexation of the basic part of the pension has only been applied when the total individual pension amount was lower than the pensioner’s subsistence minimum. Today, for 60 percent of old-age pensioners, the total pension amount is over 5,000 soms, while the subsistence minimum is less than 4,300 soms. This projection uses the assumption that the subsistence minimum will grow in line with inflation. But over the past decade, this indicator has been changing less systemically46, so if its behavior remains just as unpredictable, the size of the basic part will be frozen for most pensioners for even longer. A reduction in the replacement rate down to 8.5 percent in the next decade may even be considered an optimistic scenario. Subsequently, the replacement rate will be decreasing at a slower pace (down to 6.8 percent by the end of the forecast period) due to the increasing share of the new generations of pensioners with an incomplete length of service. 46 Kyrgyz Republic National Statistics Committee website: http://www.stat.kg/ru/statistics/download/dynamic/825/. 85 Figure 3.6. Average old age pension replacement rates (percent of average wages) All old-age pensioners New old-age pensioners 32% NDC pension 32% NDC pension DB pension DB pension 28% 28% Basic pension Basic pension 24% Minimum subsistence level 24% Minimum subsistence level 20% 20% 16% 16% 12% 12% 8% 8% 4% 4% 0% 0% 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 Source: Micro simulation results. 3.35. Currently the defined benefit (DB) insurance pension accounts for the largest share in the total size of the old-age pension, with a replacement rate in excess of 16 percent of average wages. The DB component of the insurance pension is a legacy scheme based on the years of service before 1996 (the year when NDC accounts were introduced). The DB formula accrual rate is 1 percent of a DB reference wage for a year of service under the pre-1996 pension scheme. The significance of the DB part is rapidly decreasing due to indexation which is lagging dynamics of wages and due to the natural exit of the very old pensioners from the overall pensioner population. In the newly assigned pensions, the DB share is already small (its replacement rate is around 3 percent of the average wages), because service for the periods prior to implementation of the individual record-keeping system is phasing out from the total stock of pension rights. 3.36. During the transition to a new system, the role played by the NDC plan in the formation of the total amount of benefits increases, but its size is too low to make up for the disappearing benefit of the DB component. The replacement rate of the NDC pension for new pensioners will account for 5 percent of average wages in the short term and will grow up to 9 percent in the long term. Given the existing indexation rules it will be impossible to maintain even that replacement rate for existing old-age pensioners; by the end of the projections period, the replacement rate of the NDC pension among the old-age pensioners will be only 6.8 percent. 3.37. Notably, the total replacement rate of the average old-age pension will gradually go down to 14 percent of the average wages and the share of all pensioners receiving pensions below the subsistence minimum will reach 50 percent, while the subsistence minimum declines over time from the current 24 percent to 14 percent of average wage. Such a decrease in the level of old-age pensions is largely due to the weak design of the existing NDC plan, specifically insufficient valorization of the balances on individual accounts of the insured, especially from 1996 to 2017 and multiple categories of privileged pensioners. Another key problem is a weak economy and weak labor markets, which are unfavorable for the pension system. The growth of employment is lagging significantly behind the growth rate of the working-age population, while the level of minimum wages remains quite low and the scale of vulnerable employment remains largely unchanged. 3.38. Women find themselves in a particularly disadvantageous situation. This is because their 86 employment is concentrated either in the public sector, with low wages (education, healthcare, low- or medium-level public administration positions), or in the informal agrarian sector47. 3.39. These challenges have an adverse impact on all the components of the pension system, but the NDC plan has proven to be particularly vulnerable. This is because pension rights are generated strictly from the contributions received. The basic part of the pension and DB pension are based on the length of service, so the effects of inequality in earnings, contribution rates, pension age, and so on could be more easily mitigated. 3.40. The government obligations of financing the non-contributory part of pension expenditures are forecast to decline over the projection period (Figure 3.7). The transfers from the republican budget are determined by the need to finance the basic part of pensions, military pensions, energy compensation, and other state benefit supplements. Considering a limited indexation, the expenditures for financing the basic part of pension will be gradually decreasing in relative terms48. As a result, the government obligations will decline from the current 3.7 percent to 2.8 percent of GDP by the end of the projection period. 3.41. Despite an increase in the number of pensioners, medium-term expenditures on insurance pensions will also decrease as share of GDP (Figure 3.8) due to a substantial drop in the level of benefits as described above. By the mid-2030s, the structural transition from a legacy DB to the NDC scheme will be completed, and the expenditures on the insurance part of pensions will stabilize at the level of 2.3–2.5 percent of GDP49, while the contributory revenues will remain stable at around 5.1 percent of GDP. Figure 3.7. Projected government obligations Figure 3.8. Projected contributory revenues and (percent of GDP) total expenditures of the system (percent of GDP) 6.0% Funeral & admin. costs Other state pensions 9.0% NDC pensions 5.0% 8.0% Old age DB pensions Basic pensions 4.0% 7.0% Contributory PAYG Revenue 6.0% 3.0% 5.0% 4.0% 2.0% 3.0% 1.0% 2.0% 1.0% 0.0% 0.0% 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 Source: Micro simulation results. 3.42. Due to a reduction in the level of PAYG pensions, the system will generate a significant annual surplus (approximately 2.5 percent of GDP in the second half of the forecast period). But that does not 47 Ajwad, M.I., and S.B. Gonzales (2018), “Jobs in the Kyrgyz Republic,� Washington, DC: World Bank. 48 It is assumed that obligations of military and other state benefit payments are proportional to the expenditures on basic part of pension. 49 The dynamics of expenditures on disability pensions, survivor’s pensions, and other types of payments is assumed to be proportional to the expenditures on old-age pensions. 87 reflect the financial health of the pension system, but rather a need for a structural and parametric reform of the benefits package. The new pensions are almost half the average size of those accrued in the past; 90 percent of the newly assigned pensions are below the subsistence minimum, while this percentage does not exceed 30 percent for the current pensioner population. As such, the NDC plan, with its current parameters, does not fully substitute for the “old� DB plan. Thus, the NDC plan requires an urgent intervention to adjust parameters and ensure meaningful benefit for the future retirees. 3.43. Prospects of the mandatory funded pillar remain limited. The volume of employees’ contributions to the mandatory funded pillar will stay at around 0.3 percent of GDP annually. The accumulated assets will reach 5.8 percent of GDP by the end of the forecast period. Annual expenditures on the payments will gradually grow, and by the end of the projection period will be equal to the annual contribution revenues: 0.3 percent of GDP. Compared to the payouts from the main system, the funded pillar and the payouts it provides will remain small. We note that only a limited number of beneficiaries will likely qualify for the regular payouts from this pillar, given the minimum threshold of 30 subsistence minimums in accumulated savings. D. ASSESSMENT OF THE EFFECT OF PARAMETRIC REFORMS 3.44. The simulation results presented in this chapter assess the effect of various changes related to the old-age pensions. Every parametric reform is assessed separately in comparison with the results of the status quo scenario. All the simulated reforms take effect from 2021. The list of parametric reforms was proposed by the Social Fund, taking into account recommendations generated over the past three years, including in the context of the Interdepartmental Working Group established by a Decision of the Committee on Social Affairs, Education, Science, Culture and Healthcare of the Parliament of the Kyrgyz Republic. 3.45. Five parametric reforms have been simulated: 1. Introducing a five-year minimum qualifying period for old-age pension eligibility, gradually increasing it to 20 years in 2036; 2. Gradual increase of the statutory pension age for women ages 58–60, by four months every year; 3. Restricting early retirement for individuals with long service; 4. Use of gender-differentiated actuarial coefficients in the formula for calculating the NDC pension benefits; and 5. Annual valorization of NDC accounts to 100 percent of the wage growth. 3.46. Case 1: Raising the qualifying period. The current qualifying criteria of length of service are minimal, which allows maintaining the broad coverage of the old-age pensioners at a time of increasing employment instability. Under the status quo scenario, the coverage is gradually going down to 63 percent of the population of retirement age, but the number of old-age pensioners is still growing, following the demographic growth. If a five-year qualifying period is introduced in 2021, with a gradual increase up to 20 years in 2036, the number of old-age pensioners will not grow but rather shrink (Figure 3.9). By the end of the forecast period, their numbers will be nearly half of the number in the status quo scenario, and the coverage level will go down to 33 percent. So, the more stringent qualifying requirements for pension will disqualify half of the individuals applying for pension. 88 Figure 3.9. Forecast results related to the establishment of qualifying period (millions of people) Number of old-age pensioners Number of older people not covered by old-age pensions 1.0 1.0 0.9 Status quo 0.9 Status quo 0.8 Qualifying period 0.8 Qualifying period 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 0.1 0.0 0.0 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 Source: Micro simulation results. 3.47. On one hand, such a policy change would bring the pension coverage of older people into conformity with the contributory coverage of the working-age population. On the other hand, there are several arguments against setting high requirements for the qualifying period. The key implications and risks are as follows: • The rapid growth of the number of older people not entitled to an old-age pension and who will instead apply for a monthly social benefit will result in the projected number of social benefit beneficiaries at around 1 million (up from 30,000 currently). This implies that the number of old-age pensioners will be nearly half the number of social benefit recipients. This will inevitably result in shifting the policy focus from pension insurance to social assistance. Furthermore, given such a broad coverage, it will likely be politically impossible to maintain the social benefit at a fixed low rate (over the past several years, it has been only 1,000 soms per month) so the costs of that program will significantly increase. Effectively, while savings will be generated on the pension side, they will be consumed by the needs of expanding social assistance, for which a separate fiscal assessment will be required. • High qualifying requirements will turn into a barrier rather than an incentive for participation in pension insurance, particularly considering the current state of the national labor market, where opportunities for formal employment are limited. • The formula used to calculate the NDC pension does not consider the insurance period but only the accumulated capital, based on actual contributions, and life expectancy after retirement. It will be necessary to set a minimum size of insurance pension; otherwise, those who have met high qualifying criteria will only be entitled to a pension at the level of its basic part (it is relevant for the self-employed and other contributors with a reduced contribution rate). Alternatively, a special DB plan could be introduced for such categories with a reduced contribution rate. 3.48. In view of the above, establishing high-qualifying requirements for pension is not recommended. Instead it should be limited to five years and introduced over time. To ensure more flexibility, as a transition measure, the missing qualifying years could be “purchased� at the time of retirement with a lump-sum contribution. 89 3.49. Case 2: Increasing the retirement age and further restricting early retirement. The list of parametric reforms under consideration includes two options that impact the actual pensionable age: (1) the gradual increase in the statutory retirement age for women ages 58–60; and (2) abolishing early retirement for those with a long service. Both options aim at the same goal, to encourage people to work longer and defer retirement. While the impact of these changes on the employment rate of older people is difficult to predict (particularly when the younger- population cohorts are expanding), it is possible to simulate a change in the number of old-age pensioners quite accurately (Figure 3.10). Figure 3.10. Projected number of old-age pensioners if retirement age is changed (millions of people) Increase in retirement age for women Abolishing of early retirement based on service 1.0 1.0 0.9 0.9 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 Status quo Status quo 0.2 0.2 Increase in retirement age Abolishing of early retirement based on service 0.1 0.1 0.0 0.0 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 Source: Micro simulation results. 3.50. The proposed increase of the retirement age for women ages 58–60 by 2026 will result in the growth of the average actual retirement age for women by 1.7 years. Compared with the status quo scenario, the total number of old-age pensioners will be reduced by 7 percent. 3.51. Contributors with a long length of service (35–40 years for men, 30–35 years for women) currently have the right to retire three years before the normal retirement age. Abolishing this rule, gradually over time, will reduce the number of new early-retirement pensions by half and will increase the average actual retirement age by 0.9 years. Due to its impact on the total number of old- age pensioners, abolishing early retirement based on length of service would be almost equivalent to an increase in the retirement age for women up to age 60. 3.52. Case 3: Using gender-differentiated actuarial coefficients for NDC benefit calculation. Actuarial coefficients are used to calculate the rate of NDC pension and are inversely related to the life expectancy. Currently, for each age group, the annual averages of life expectancy and actuarial coefficients are calculated by gender, and then the average coefficients for men and women are calculated. There is a proposal under consideration by the government to eliminate the last step and not average the coefficients for both genders. However, because actuarial coefficients consider life expectancy at retirement, their differentiation by gender will be favorable for men but unfavorable for women. For example, in 2019, the actuarial coefficient for age 55 averaged 0.0480 for both genders (Table 3.7). If we apply its own coefficient to each gender, then for men it will be higher than the average value (0.0541) and for women it will be lower than the average value (0.0419), because upon reaching that age, a woman’s life expectancy is 5.4 years longer. It means that, other conditions being equal, the pension benefit will be higher for men and lower for women compared with a pension benefit that uses the average actuarial coefficient. 90 Table 3.7. Actuarial coefficients used to calculate the NDC pension in 2019 Retirement at age 55 Retirement at age 58 Average life Average Average life Actuarial Average Actuarial Gender expectancy actuarial expectancy coefficien actuarial coefficient over the past coefficient for over the past t by coefficient for by gender five years both genders five years gender both genders Men 18.48 0.0541 16.28 0.0614 0.0480 0.0541 Women 23.88 0.0419 21.40 0.0467 Source: Kyrgyz Republic Social Fund. 3.53. If gender differentiation of actuarial coefficients is applied, the replacement rate for the NDC pension of new old-age pensioners will increase by roughly 1.5 percent for men, but it will decrease by the same amount for women (Figure 3.11). Given the fact that the demographic forecast suggests a growth in life expectancy of both women and men, in the long run, the advantages of differentiated actuarial coefficients for men will diminish, while the disadvantages for women will be even greater. Figure 3.11. Projected Level of NDC pensions for new old-age pensioners when using gender differentiation of actuarial coefficients (percent of average wages) Women Men 12% 12% 10% 10% 8% 8% 6% 6% 4% 4% 2% Status quo 2% Status quo Differentiated coefficients Differentiated coefficients 0% 0% 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 Source: Micro simulation results. 2050 3.54. Because women account for 70 percent of all old-age pensioners, a reduction in the level of NDC pensions under this scenario will result in a reduction in the general level of pensions and, consequently, higher expenditures from the republican budget on the basic pension and other types of social assistance. The status quo scenario is already providing evidence of a critical decline in the level of insurance pensions, as well as the challenges inherent in the NDC plan. Gender differentiation of actuarial coefficients will only aggravate gender inequality, particularly while the difference in retirement age between men and women persists. 3.55. Case 4: Valorization of NDC accounts by 100 percent of wage growth. Since 2018, valorization of NDC accounts of insured persons has been carried out based on the consumer price index without consideration for the growth of wages. Over the past decade, the macroeconomic trends in this country 91 indicate a very low inflation and quite rapid growth of real wages. This implies capitalization of individual accounts that significantly lags the earnings of the working population. Given the critically low level of new pensions, we suggest that the annual valorization of NDC accounts be made more generous, specifically by applying an index of 100 percent of the nominal wage growth, which is more in line with international norms. If such a rule is introduced from 2021, then by the end of the forecast period, the replacement rate of the NDC pension of the new old-age pensioners will increase and reach 10.3 percent for women and 14 percent for men (Figure 3.12). However, even this measure will be insufficient in the long term, to ensure adequate pension level. Figure 3.12. Projected level of NDC pensions for new old-age pensioners in case of indexation of accounts by 100 percent of wages growth (percent of average wages) Women Men Status quo Status quo Indexation of accounts by 100% of wages growth Indexation of accounts by 100% of wages growth 16% 16% 14% 14% 12% 12% 10% 10% 8% 8% 6% 6% 4% 4% 2% 2% 0% 0% 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 Source: Micro simulation results. 3.56. The situation with benefit adequacy can be improved by applying a retrospective valorization of accounts for the period starting from 1996, for example, revaluation of the funds in individual accounts of the NDC plan using indexation of 100 percent of nominal wage growth. This recommendation requires additional extensive calculations. E. RECOMMENDATIONS The analysis indicates the following strengths of the national pension system in the Kyrgyz Republic: • The diversification of insurance and governmental programs, and PAYG and funded pillars; • Flexible conditions of participation in pension insurance for different categories of the employed population; • Very minimal qualification conditions that ensure full-scale coverage of older people by old-age pensions; and • The introduction of the mandatory funded pillar, where the pensioners can choose whether to keep their funds in the State Funded Pension Fund or transfer their pension assets to a private company. 92 However, there are internal and external factors that undermine fiscal and social stability of the pension system: • While the number of pensioners is growing due to the demography, the growth in the number of contributors is constrained by the capacity of the labor market; • The structural pension reform with poorly calibrated parameters has put the insurance system on the path of gradual but dramatic decline in the new pensions. The benefit size is projected to be halved in relative terms, which will result in disappointment of the population in the pension reform and growth of poverty among older people; and • The indexation of both the individual NDC accounts of insured persons and the assigned pensions is lagging the dynamics of wages. A summary of recommendations for further reforms, based on this analysis, follows: • Consider further restrictions on early retirement and assess options for aligning the pension age for men and women, bringing it to 63 for both; • Set a guaranteed minimum pension for people with an incomplete length of service, for example, as a percent of the basic part of the pension; • Simplify indexation of the basic part of the pension to make the system more predictable and transparent. Specifically, introduce a universal indexation criteria, linked to macroeconomic indicators, that apply to all beneficiaries irrespective of their insurance pension amount; • Improve the NDC plan: (1) avoid introducing gender differentiation in benefit calculation; (2) index (valorize) the accounts by 100 percent of wage growth; (3) set the rules for compensation for the credit periods in the length of service (childcare time, mandatory military service, and so on); and (4) assess options of indexing historic NDC balances retroactively; • Relax minimum account balance requirements for the regular benefits from the funded pillar to avoid unreasonable spending of the funds on one-off lump-sum payments. Assess options for better coordination in payouts from the PAYG and funded pillars; and • In line with the 2014 PER, consider excluding farmers from the mandate of the NDC plan so their contributions can finance the basic part of pensions only. This should be done in coordination with adjusting parameters of the basic part of the pension. Alternatively, consider switching to a simplified DB formula for that category, for example with benefits linked primarily to the individual service period (incorporating the experience of Ireland, Kazakhstan50, the Netherlands51, and the UK). Table 3.8. Summary of policy recommendations Priority Policy Recommendations Time Frame Expected Impact Pension Revise the indexation of the basic part of the pension, so it Short term Improved, predictable, and reflects minimum income guarantee by the state, linked to transparent indexation of the basic macroeconomic indicators to make the system more part of pension predictable and transparent. 50 Law of the Republic of Kazakhstan On the Provision of Pensions in the Republic of Kazakhstan, Electronic Government of the Republic of Kazakhstan: http://egov.kz/cms/ru/law/list/Z1300000105. 51 European Commission, Directorate-General for Economic and Financial Affairs. The 2018 Ageing Report. Economic & Budgetary Projections for the 28 EU Member States (2016-2070), Institutional Paper 079, May 2018. 93 Improve the NDC plan by avoiding gender differentiation in Short term Improved and fair indexation of benefit calculation, setting the rules for compensation for the the insurance pension to preserve credit periods in the length of service (maternity leave, benefit value over time mandatory military service, and so on) and assessing options of indexing historic NDC balances retroactively. Set a guaranteed minimum pension for people with an Short term Improved adequacy of the pension incomplete length of service, for example, as a percent of the benefit basic part of the pension. Consider an alternative proposal of switching to a modified DB Medium term Improved adequacy of the pension formula, with benefit linked primarily to service. benefit Consider further restrictions on early retirement and assess Medium term Reduced fiscal burden on the options for aligning the pensionable age for men and women. pension system Relax the minimum account balance requirements for the Medium term Improved adequacy of the pension regular benefits from the funded pillar to avoid unreasonable benefit spending of the funds on one-off lump-sum payments. Assess options for better coordination in payouts from the PAYG and funded pillars. 94 References Law of the Kyrgyz Republic dated July 21, 1997 No. 57 “On State Pension Social Insurance�, Centralized Legal Information Data Bank of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru- ru/557/410?cl=ru-ru Law “On Rates of Insurance Contributions for State Social Insurance�, Centralized Legal Information Data Bank of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/1393?cl=ru-ru Law of the Kyrgyz Republic “On Pension Provisions for Military Personnel�, Centralized Legal Information Data Bank of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/690 Law “On State Benefits in the Kyrgyz Republic�, Centralized Legal Information Data Bank of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/111670 Regulation on the Social Fund of the Kyrgyz Republic, approved by the Order of the Government of the Kyrgyz Republic No. 137 dated February 20, 2012, Centralized Legal Information Data Bank of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/93511?cl=ru-ru Order of the Parliament of the Kyrgyz Republic dated June 15, 2006 No. 1115-III “On approval of the accounting index amount�, Centralized Legal Information Data Bank of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/51948 Report on the activities of the State Service for Financial Market Regulation and Supervision under the Government of the Kyrgyz Republic for 2019: https://fsa.gov.kg/docs/file/actreport/447-_2019_-.pdf Rules for payment by the Social Fund of the Kyrgyz Republic of pension savings of the State Accumulative Pension Fund, approved by the Order of the Government of the Kyrgyz Republic dated October 24, 2012 No. 745 (as amended), Centralized Legal Information Data Bank of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/93751/40?cl=ru-ru Elderly people in the Kyrgyz Republic. Statistical research report. National Statistical Committee of the Kyrgyz Republic, SIAR Research & Consulting, Bishkek, 2017. European Commission, Directorate-General for Economic and Financial Affairs. The 2018 Ageing Report. Economic & Budgetary Projections for the 28 EU Member States (2016-2070), Institutional Paper 079, May 2018. World Social Protection Report 2017 – 19. Universal social protection to achieve the Sustainable Development Goals. Statistical annexes. URL: https://www.social- protection.org/gimi/gess/ShowWiki.action?id=594#tabs-3 M.I. Ajwad and S.B. Gonzales (2018) “Jobs in the Kyrgyz Republic� World Bank, Washington, DC. United Nations, Department of Economic and Social Affairs, Population Division (2019). World Population Prospects 2019, Online Edition. Rev. 1.: https://population.un.org/wpp/ 95 ANNEXES ANNEX 1: HEALTH CHAPTER The top four risk factors for disability-adjusted life years (DALYs) are: (1) high systolic blood pressure (10.9 percent of DALYs); (2) alcohol use (8.5 percent of DALYs); (3) high fasting plasma glucose (7.8 percent of DALYs); and (4) high body mass index (7.6 percent of DALYs). Almost 27 percent of the adult population is hypertensive (systolic blood pressure ≥ 140 mm Hg), and 10 percent has raised fasting blood glucose (≥ 7 mmol/L). The prevalence of smoking is high (26.5 percent) and largely concentrated among men—almost 51 percent of the male population smokes, compared with less than 4 percent of women. Annex Figure 1. Kyrgyz Republic’s population pyramid, 2017 80+ 21 40 75-79 19 32 70-74 22 33 65-69 47 63 60-64 87 106 55-59 121 141 50-54 149 164 45-49 152 162 40-44 165 170 35-39 191 188 30-34 255 251 25-29 302 301 20-24 280 269 15-19 251 241 10-14 265 255 5-9 347 331 0-4 399 377 Males (‘000s) Females (‘000s) Source: World Development Indicators (WDI) (2019). 96 Annex Figure 2. Prevalence of BP monitoring in PHC visits by oblast, 2015 1 Share of patients who had 0.8 their BP measured 0.6 0.4 0.2 0 Naryn Chui Bishkek Jalalabad I.K. Osh Batken Talas Osh city Source: xxx Annex Figure 3. Inpatient admissions per 100 by oblast, 2010-18 18.0 17.0 inpatient admissions per 100 16.0 15.0 14.0 13.0 12.0 11.0 10.0 9.0 8.0 2010 2011 2012 2013 2014 2015 2016 2017 2018 Kyrgyz Republic Batken Jalal-Abad Issyk-Kul Naryn Osh oblast Talass Chuy Bishkek Osh city Source: e-Health Center (2020). Annex Figure 4. Average length of stay (days) by oblast, 2010-18 16 14 Average length of stay (days) 12 10 8 6 4 2010 2011 2012 2013 2014 2015 2016 2017 2018 Kyrgyz Republic Batken oblast Jalal-Abat oblast Yssyk-Kul oblast Naryn oblast Osh oblast Talas oblast Chui oblast Bishkek city Osh city Republican organizations Source: xxx 97 Annex Table 1. Top 20 causes for hospitalization by ICD-10 code, 2018 ICD-10 Code Diagnosis Number of cases Z37.0 Outcome of delivery/ single live birth 130053 O80.0 Encounter for full-term uncomplicated delivery 120433 A04.9 Bacterial intestinal infection, unspecified 25786 J18.0 Bronchopneumonia, unspecified organism 22262 I20.0 Unstable angina 18950 S06.0 Concussion 16649 J06.9 Acute upper respiratory infection, unspecified 16592 I25.1 Atherosclerotic heart disease of native coronary artery 13384 O82.1 C-section without indication 11158 J15.8 Pneumonia due to other specified bacteria 10959 K35.0 Acute appendicitis 9389 I20.8 Other forms of angina pectoris 8865 O47.0 False labor before 37 completed weeks of gestation 7605 J18.9 Pneumonia 7016 B15.9 Hepatitis A without hepatic coma 6921 F10.2 Alcohol dependence 6682 K80.0 Calculus of gallbladder with acute cholecystitis 6280 P08.1 Other heavy for gestational age newborn 5708 B34.1 Enterovirus infection 5668 O20.0 Threatened abortion 5130 Source: MHIF. Annex Figure 5. Decomposition of government health spending, 2000-15 98 Annex Figure 6. Government health expenditures by economic classification Note: Expenditure shares include funds from the general budget, MHI, special means, and co-payments. 99 ANNEX 2: PENSIONS CHAPTER Pension Projections: Demographic and Macroeconomic Assumptions A. Population Structure (thousand) 2020 2025 2030 2040 2050 Population,000 6,524 7,010 7,446 8,307 9,126 Male population,000 3,227 3,467 3,681 4,103 4,511 Female population,000 3,297 3,543 3,765 4,204 4,615 Working age population,000 3,900 4,115 4,474 5,158 5,475 Working age male population,000 2,002 2,121 2,310 2,681 2,897 Working age female population,000 1,898 1,993 2,164 2,476 2,578 Population ret.age+,000 524 663 792 1,008 1,366 Male population 63+,000 151 202 253 323 441 Female population 58+,000 373 462 539 686 925 Source: UN World Population Prospects 2019, Kyrgyz Republic B. Macroeconomic Assumptions* 2018 2020 2021 2022 2030 2040 2050 Nominal GDP, billion 557.1 642.9 703.2 771.2 1,334.9 2,355.8 3,944.6 Real GDP growth, % 3.8% 3.8% 4.2% 4.5% 3.4% 3.8% 2.8% Inflation, % 1.5% 3.5% 5.7% 5.0% 2.0% 2.0% 2.0% Average monthly wage, 12,744 14,156 15,211 16,450 25,870 38,441 57,121 soms Real wage growth, % 3.5% 1.7% 3.0% 2.0% 2.0% 2.0% Number of of 1,352 1,404 1,429 1,449 1,595 1,895 2,135 contributors, 000 *Notes: Nominal GDP, 2018. Source: International Monetary Fund, World Economic Outlook, October 2019. Real GDP growth 2018-22. Source: World Bank, MTI GP, and task team assumptions for real GDP growth post-2022. Consumer price index (CPI) 2018-22. Source: World Bank, MTI GP, and task team assumptions for CPI post-2022. Average monthly wage, 2018. Source: Social Fund of Kyrgyz Republic, administrative data, and task team calculations for post-2018. Real wage growth. Task team calculations based on the assumption that nominal wage bill is a constant fraction of GDP over the projection period. Number of contributors 2018. Source: Social Fund of Kyrgyz Republic, administrative data, and task team assumption that post-2018 number of contributors is a constant fraction of projected working age population. Description of the Existing Pension System Pension age and qualifying period. Men who have reached 63 years and women who have reached 58 years are entitled to receive a pension. To receive a full state basic part of the pension, men need a total of 25 qualifying years of contributions, whereas women need 20 qualifying years. In case of incomplete service, any person is eligible to claim a pension if that person has paid contributions for at least one month or has paid at least one contribution. Early-retirement pensions. There is an extensive list of categories of persons who can retire before reaching the retirement age if they work in a harmful or special working environment (underground works, list 1 and list 2; textile production, aviation workers, artists, and so on) or if they live and work in high mountain regions, remote or isolated areas. This rule also applies to mothers with many children, mothers with a disabled child, 100 and mothers in several other categories. The requirements for age and qualifying years of the insurance service that must be met by these early retirement categories to get a full basic part of the pension are substantially lower. Besides, it is possible to claim an early-retirement old-age pension if the insurance service is very long. If the actual service is at least 40 years for men and 35 years for women, men can retire at 60 and women can retire at 55 without any reduction in the size of the pension benefit. If the insurance service is at least 35 years for men and 30 years for women, a man can retire at 60 and a woman can retire at 55 with a reduced benefit. Until the statutory pension age is reached, the size of the pension will be reduced by 1.5 percent for each month of the early-retirement period. When the person reaches the statutory retirement age, the size of the pension is recalculated based on the rules for calculating a full basic part; however, this benefit is paid in a reduced amount until the expenditures of the early-retirement period are fully offset. This mechanism enables the person to decide at what age to retire within a three-year early retirement period. Rules for calculating and indexing the basic part of the pension. Before 2017, the size of the basic part of the pension was set directly by a presidential decree or a government resolution as a flat-rate fixed amount. Legislation did not set forth frequency and procedures for its periodic adjustment. The only requirement was to maintain the basic part of the pension at the level that is not below 12 percent of actual average wages in the country in the previous year. This requirement is still there, but in 2017 legal changes and additional requirements were introduced. Starting from 2018, the basic part of the pension has been indexed annually along with other benefits in line with an increase in average wages52. At the same time, the size of the basic part has been linked to the size of the insurance pension. If the size of the insurance pension (DB + NDC) is higher than three subsistence minimums, the basic part of the pension is not assigned. If the total size of the pension from the PAYG system (the basic part + DB + NDC) exceeds the actual pensioner’s subsistence minimum in the previous calendar year, the basic part is not indexed. The argument to explain such limitations is that the basic part of the pension is a state guaranteed benefit aimed to protect the most vulnerable groups of the population. The size of the basic part of the pension paid to pensioners with incomplete service is reduced proportionally to the number of missing years of contributions. For such pensioners, if they work for not less than 12 months and pay contributions, the basic part of the pension is recalculated, considering the “new� contribution record. After assigned, the basic part of the pension is recalculated/indexed depending on the qualifying years of the insurance contribution record, the overall size of the pension, the rate of the wages increase in the country, and changes in the subsistence minimum income. Rules for calculating the insurance pension under the DB plan. The insurance pension under the DB plan is calculated as 1 percent of the average monthly wage for each full year of service in the period before 1996. Before 2017, the average monthly earnings had been determined using the method traditional for the DB plan, that is, it was based on actual data for any 60 consecutive months of work in the period before 1996, with adjustments to account for valorization coefficients for corresponding years53. Given the issues with 52 The procedure for indexation of pensions assigned in conformity with the Kyrgyz Republic Law On State Pension Social Insurance, approved by Kyrgyz Republic Government Resolution No. 347, dated July 25, 2018, Centralized Databank of Legal Information of the Kyrgyz Republic, http://cbd.minjust.gov.kg/act/view/ru-ru/12235. 53 Valorization coefficients are set by Kyrgyz Republic Law No. 58, dated June 26, 2001, On Adjustment Coefficients to Wages for Pension Calculation, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru- ru/465. 101 keeping earning records for the periods before implementation of the individual recordkeeping system, an alternative method was developed, whereby the earnings are assessed in proportion to the existing insurance service and the size of the minimum wage on the day the pension is calculated. For those who have the insurance service of up to 15 years, the earnings are set at 200 percent of the minimum wage; if the record is from 15 to 20 years, provisional earnings are set at 250 percent, and if the service is 20 years or more, provisional earnings are set at 300 percent of the minimum wage. A person can select which of the two methods to use to calculate the pension size. According to the Social Fund, the second method is used more often. Apart from its saving a lot of trouble and making pension assignment easier and faster, it is preferable to those who received low wages. Imputation of earnings before the implementation of the individual recordkeeping system is a common method. For example, among some of the post-Soviet countries, Azerbaijan and Tajikistan use this method to calculate notional capital for NDC accounts54 55. The imputation of earnings in calculating the DB pension is also justified, especially if it is combined with a progressive schedule of payouts based on the insurance service. Additional focus on the service record in the PAYG systems is always relevant as demonstrated by the experience of using fixed pensions linked to the insurance service (Ireland, Kazakhstan56, the Netherlands57, and the UK). Regardless of the calculation method, the average monthly wage for calculating DB pensions cannot be more than 150 basic amounts. Starting from January 1, 2006, the basic amount has been set at 100 soms58, that is, the maximum size of the earnings taken into account in calculating the DB pension is 150 * 100 = 15,000 soms. Rules for calculating the insurance pension under the NDC plan. The insurance pension under the NDC plan is calculated as the amount of contributions registered in the individual account divided by 12 and multiplied by the actuarial coefficient (the value inversely related to the expected length in receipt of pension). Before 2017, the NDC contributions to the individual accounts of the insured persons had been indexed every year by at least 75 percent of the increase in the actual wage in the reporting year. However, the indexation coefficient was applied only to the new contributions rather than the cumulative amount of notional capital. According to the new rules, the total NDC balance is subject to indexation. The problem is that indexation is in line with movements of the consumer price index (CPI) without considering wage increases at least in part. It means that the increase in the accrued notional capital in the individual accounts of the insured persons will be slower than changes in the income earned by the working population. In fact, in the individual recordkeeping system, the NDC accounts so far keep only records of paid contributions without indexation. Indexation applies retroactively after receiving a pension claim. 54 Republic of Azerbaijan Law No. 54-IIIQ, dated February 7, 2006, On Labor Pensions (with changes and amendments as of April 23, 2019): http://www.e-qanun.az/framework/11566. 55 Republic of Tajikistan Law On Insurance Pensions and State Pensions (as edited in 2015), the Agency for Social Insurance and Pensions website: http://nafaka.tj/ru/qonunho/qonun-12-01-2010-595. 56 Republic of Kazakhstan Law On Pension Benefits in the Republic of Kazakhstan, E-Government of the Republic of Kazakhstan: http://egov.kz/cms/ru/law/list/Z1300000105. 57 European Commission, Directorate-General for Economic and Financial Affairs. The 2018 Ageing Report. Economic & Budgetary Projections for the 28 EU Member States (2016-2070), Institutional Paper 079, May 2018. 58 Resolution No. 1115-III, dated June 15, 2006, issued by the Parliament On Approval of the Basic Amount, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/51948. 102 Actuarial coefficients are calculated annually, based on the data of the National Statistics Committee, on average life expectancy for men and women, considering changes in the past five years. The coefficient value is linked to the retirement age: the later the person retires, the shorter is the retirement period in life, the higher is the actuarial coefficient, and the higher is the monthly pension under the NDC plan. If after the assignment of the pension benefit the retired person continues to work and pay contributions, the pension under the NDC plan is recalculated annually, considering new contributions received by the system after retirement. Rules for indexing the insurance part of the pension. The insurance pension under the DB and NDC plans is indexed based on the same rules; after important changes introduced in 2017, the level of protection is higher. Previously, the insurance part of the pension would be indexed only if the average wages and the CPI increased in the previous calendar year. The new Law On the State Pension Social Insurance explicitly sets forth two criteria for calculating the indexation coefficient: this coefficient should be not less than the average value of the sum of the CPI and the increase in the average monthly wages, and not less than the increase in the actual pensioner’s subsistence minimum in the previous calendar year. According to the statistical data of recent years, changes in the subsistence minimum have not been keeping pace with increases in wages59. For that reason, in all likelihood, the insurance pensions under both plans are expected to be indexed to the average value of the sum of the CPI and the increase in wages (in other words, 50 percent of inflation + 50 percent of the increase in wages, the so-called Swiss indexation formula). Minimum and maximum pension size. The insurance pensions under the DB and NDC plans do not have a minimum size; any calculated value and the basic part of the pension are summed up. The basic part of the pension plays the role of a minimum guaranteed state pension benefit, but only for people with complete qualifying years of the insurance service. For pensioners with incomplete years of the contribution record, the basic part of the pension is reduced proportionally to the number of missing years in the record, and there is no cutoff to stop further decreases. In this situation, men are put at a disadvantage because their insurance service, which makes them eligible to claim a full basic part of the pension, is five years longer than that of women (25 qualifying years for men and 20 qualifying years for women). The maximum size of the DB pension is regulated by the cap on average monthly wages considered to calculate the size of the pension. The maximum size of the NDC pension is not limited directly; however, there are a lot factors within the system itself that impede the accrual of notional capital: from weak valorization of balances of individual accounts to a lack of mechanisms to compensate contributions during the non-insured periods. Disability insurance and survivor’s pensions. Besides the old-age pensions, the PAYG system also pays disability pensions and survivor’s pensions. The disability pension is assigned regardless of the cause of disability if the person has an insurance contribution record. If qualifying criteria of the insurance service are met, the pension assigned to disabled persons from disability groups 1 and 2 equals the full amount of the basic and the insurance parts of the old- age pension; the pension assigned to persons from disability group 3 is 50 percent of the old-age pension. The insurance service criteria are set proportionally to the age when the person became disabled: one year of the contribution record for individuals under 23 and five years of the contribution record for individuals above 30. If the insurance service has fewer years than required by law, the basic part of the pension is reduced 59 Subsistence income for main social and demographic groups of the population. See the Kyrgyz Republic National Statistics Committee website: http://www.stat.kg/ru/statistics/download/dynamic/825/. 103 proportionally, as is the case with the old- age pension assigned to a person with incomplete service. Surviving dependent family members of a deceased breadwinner can claim a survivor’s pension if the deceased had an insurance service that meets the criteria set for disability pensions. The si ze of the survivor’s pension is calculated taking into account the deceased person’s contribution record and the composition of the family: if there is one dependent family member, it is 50 percent of the survivor’s pension; two dependent members, 90 percent of the pension; three dependent members, 120 percent; and four or more dependent family members, 150 percent. When the disability pension or the survivor’s pension is calculated, only the actual insurance service is considered. Supplements to pensions. Supplements to the old-age pensions and the disability pensions are set for some categories such as disabled war veterans and war participants; ex-prisoners of concentration camps; survivors of the siege of Leningrad; political prisoners in the 1930–1950s; Chernobyl victims; mother-heroines (honorary title for bearing and raising a large family) as well as old-age pensioners and pensioners with a disability pension who have turned 80. The list and the size of such supplements are set by the Law On State Pension Social Insurance; the size is a multiple of a specified value. In addition, an allowance for increased electricity tariffs is paid monthly through the PAYG system. Allowances to compensate for costs of higher tariffs were put in place in 2002 by presidential decree and initially were differentiated based on a specified scale linked to the overall pension size (the higher the pension, the lower the allowance). Starting from 2010, the allowance to compensate for higher tariffs has been fixed at 200 soms and has been paid to the pensioners with pensions that do not exceed 4,000 soms60. Military pensions. Long-service pensions of military and security officers are administered by relevant ministries, they are financed by the republican budget, and they are not part of the general pension system. The only exceptions are disability pensions and survivor’s pensions, which, though assigned according to the rules envisioned by the Law On Pensions to Military Servicemen, are included in the general PAYG system, from an administrative point of view. Military disability pensions and military survivor’s pensions are assigned regardless of the length of service. The size of the pension is calculated using the DB plan formula as 30 –80 percent of the earnings depending on the disability cause and the disability group or the number of the dependent persons in the family of the deceased breadwinner. “Military� disability pensions and “military� survivor’s pensions are distinguished from “civil� pensions because the former have a guaranteed minimum size61. Social benefit for people not entitled to a pension. A disabled person with no insurance service will receive a monthly social benefit according to the Law On State Benefits in the Kyrgyz Republic62. The benefits are assigned and paid without means testing through district and city labor and social development authorities; 60 President of the Kyrgyz Republic Decree No. UP 509, dated November 13, 2009, On Setting Compensation Payments in Case of Changes in Tariff Policy, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/60386?cl=ru-ru. 61 Law On the Pensions to Military Servicemen, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/690. 62 Law On State Benefits in the Kyrgyz Republic, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/111670. 104 these benefits are funded by the republican budget and are not part of the pension system. The size of the benefit is set by the government, based on the budget capacity and the economy, considering the subsistence level. For persons without the insurance service, the eligibility starts at 65 years for men and 60 years for women (two years after the statutory retirement age); for mother-heroines, the benefit can be paid upon reaching 55 years. Today, the size of the social benefit paid to children with a disability is 4,000 soms; for adults with a disability, it ranges from 1,000 to 2,000 soms depending on the disability group; for persons without the insurance service (men reaching 65 years and women reaching 60 years), it is 1,000 soms.63 Sources of financing the PAYG system. In the Kyrgyz Republic, the contribution to the PAYG system is divided between the employers and the employees. All legal entities pay a contribution of 15 percent, out of which 12 percent is registered in individual accounts of the employees in the NDC plan, with 3 percent going to the general budget of the PAYG system. The employers in the private sector can determine what base they will use for assessing the contributions: the actual wages of their employees or 40 percent of the average wage in a given city/district in the previous calendar year. The employees of the enterprises enrolled in the mandatory funded pillar pay a contribution to the NDC plan at a rate of 8 percent. The employees who are not required to make contributions to the funded pillar, due to age, pay contributions to the NDC at the rate of 10 percent.64 The self-employed pay contributions at reduced rates. For peasant farms, the contribution rate is fixed and is set in soms per hectare and depends on the district and the type of agricultural lands (irrigated lands, rain-fed lands, gardens, grasslands, and so on). Out of the paid contributions, 89.75 percent goes to the NDC individual accounts. The contribution rate for individual entrepreneurs is set based on the taxation regime and the type of business activities; it is 3–10 percent of a defined value of the actual “reduced� wage in a given city/district in the previous calendar year. The so-called reduced average wage is calculated by the National Statistics Committee without considering the enterprises for which the average wage is 50 percent higher than the average wage for the entire country. Based on the approved values for 2019 and 2020, the reduced wage is 30 percent lower than the average wage65. Besides insurance contributions, a considerable part of the PAYG system revenues is generated using the proceeds from the republican budget. The transfer from the republican budget includes financing payments under state pension programs: the basic part of the pension and the “military� disability and survivor’s pensions; some other pensions (for example, early-retirement pensions taken for social reasons), categorical supplemental payments to pensions, and allowances to compensate for higher electricity tariffs. Mandatory funded pillar. The mandatory funded pillar was implemented on January 1, 2010. The contribution was set at 2 percent of the wage, in addition to contribution payments to the main system. Initially it was planned to combine payments from the funded pillar with the NDC pensions and enroll all employees 63 Kyrgyz Republic Ministry of Labor and Social Development website: https://mlsp.gov.kg/ezhemesyachnoe-posobie-litsam-ne- imeyushhim- prava-na-pensionnoe-obespechenie-sotsialnoe-posobie/. 64 Law On Insurance Contributions Rates under the State Social Insurance System, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru-ru/1393?cl=ru-ru. 65 Data on average monthly wages in the Kyrgyz Republic for calculating insurance contributions in conformity with Article 14-1 of the Kyrgyz Republic Law On Insurance Contribution Rate under State Social Insurance, Kyrgyz Republic Social Fund website: http://socfond.kg/ru/payers/smzInfo/. 105 of the formal sector from state-owned and private companies, regardless of their organizational and legal form. However, organizational difficulties and too-small amounts of funds to be accumulated necessitated introduction of a cutoff age for enrollment. Starting from January 1, 2012, men born before January 1, 1964, and women born before January 1, 1969, who must work less than 15 years before their full retirement age have been allowed not to enroll in the mandatory funded pillar. The additional 2 percent contribution rate for that category of workers has not been cancelled; this money is channeled to generate pensions under the NDC plan, while contributions paid in 2010–11 have been retained in the funded pillar and are paid together with the earned investment income upon reaching the retirement age as a lump sum. Upon reaching the retirement age, the employees for whom enrollment in the funded pillar remains mandatory are entitled to collect fixed- term benefits or claim a lump sum if the amount of their funds is less than 30 pensioner’s subsistence minimums. Pension System Administration. Starting from 2019, contributions to off-budget social funds have been collected by the State Tax Service. Administration of the PAYG system, including maintenance of individual accounts, assignment of all types of insurance pensions, state pensions, and supplements to the pension, are carried out by the Social Fund. Regarding its role, the Social Fund is an executive body of the state social security and pension system in the Kyrgyz Republic, which acts on the principles of self-administration66. Pensions are paid through commercial banks or branches of the State Post as chosen by the pensioner. A separate legal entity—the State Cumulative Pension Fund—was set up to manage mandatory funded pillar assets67. From 2019 onward, the enrolled persons have received the right to transfer their funds from the State Cumulative Pension Fund to a private funded pension fund or a managing company of the funded pension fund that are authorized to conclude contracts on the mandatory pension insurance of the funded pillar of the pension. By default, the State Cumulative Pension Fund remains the pension fund for those who have not selected a managing company/fund. 4. 66 Provisions on the Kyrgyz Republic Social Fund, approved by Kyrgyz Republic Government Resolution No. 137, dated February 20, 2012, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru- ru/93511?cl=ru-ru. 67 Kyrgyz Republic Law On Investment of Funds to Finance the Funded Pillar of the Pension under State Social Insurance in the Kyrgyz Republic, Centralized Databank of Legal Information of the Kyrgyz Republic: http://cbd.minjust.gov.kg/act/view/ru- ru/203321. 106