Report No. 89022 Kyrgyz Republic Public Expenditure Review Policy Notes Social Assistance May 2014 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank KYRGYZ REPUBLIC - GOVERNMENT FISCAL YEAR January 1 – December 31 CURRENCY EQUIVALENTS (Exchange Rate Effective as of April 30, 2014) Currency Unit = Kyrgyz Som (KGS) US$1.00 = KGS 53.9615 Weights and Measures Metric System ACRONYMS AND ABBREVIATIONS DCS Disability Certification Service MBPF Monthly Benefit for Poor Families DFID UK’s Department for International MoSD Ministry of Social Development Development MSB Monthly Social Benefit EPL Extreme Poverty line PER Public Expenditure Review GDP Gross Domestic Product SECO Swiss Economic Cooperation GMI Guaranteed Monthly Income Organization KIHS Kyrgyz Integrated Household Survey UNICE United Nations Children’s Fund Vice President: Laura Tuck Country Director: Saroj Kumar Jha Country Manager: Alexander Kremer (Acting) Sector Director: Roumeen Islam Sector Manager: Ivailo V. Izvorski Task Team Leader: Evgenij Najdov Policy Note Author: Erkin Mamadaliev Acknowledgements This note is part of the Kyrgyz Republic Public Expenditure Review (PER) Policy Notes led by Evgenij Najdov (Task Team Leader) and K. Migara O. de Silva (co-TTL). The PER work was initiated by Orhan Niksic. Faruk Khan took over as task team leader from September 2012 to June 2013. K. Migara O. de Silva’s co-TTL-ship was uninterrupted with primary responsibility for the sector notes on wage bill management, public investments management and intergovernmental transfers. This note has been prepared by Erkin Mamadaliev (Social Protection Specialist, ECSHD), with inputs from Franziska Gassmann (consultant) and Marina Petrovic (Social Protection Economist, ECSHD). The team has benefitted from advice and contributions of Bakyt Dubashev, Evgenij Najdov and Helen Edmundson. Zakia Nekaien-Nowrouz, Sarah Nankya Babirye, Ewelina Lajch and Lilia Saetova provided technical and administrative support. The PER analysis has been conducted in close coordination with Government counterparts, with earlier drafts, power-point presentations, and workshops used to support a dialogue on public expenditure policy priorities facing the Kyrgyz Republic. The team is grateful to the UK’s Department for International Development (DFID) and the Swiss Economic Cooperation Organization (SECO) for co-financing this programmatic Public Expenditure Review. iii Contents Acknowledgements ............................................................................................................................................................... iii 1. Executive Summary ............................................................................................................................................ 1 2. Poverty in the Kyrgyz Republic...................................................................................................................... 2 3. Overview of Non-contributory Social Protection Programs and Spending ................................. 5 4. Performance of the Social Assistance Programs................................................................................... 11 Targeting Accuracy And Benefit Adequacy............................................................................................. 12 Poverty Impact.................................................................................................................................................... 14 5. Other Issues.......................................................................................................................................................... 16 6. Conclusions and Recommendations .......................................................................................................... 19 References ................................................................................................................................................................................ 30 List of Annexes Annex 1. Tables and Figures ............................................................................................................................................. 22 Annex 2. Main social assistance programs .................................................................................................................. 28 List of Boxes Box 1. World Bank definition of Social Assistance..................................................................................................... 5 Box 2. Social Protection system in the Kyrgyz Republic and financing of social benefits ......................... 7 Box 3. Reform of the system of privileges ..................................................................................................................... 8 Box 4. Establishing welfare quintiles ............................................................................................................................ 11 Box 5. Four roles for Social Assistance ......................................................................................................................... 12 List of Figures Figure 1. Absolute and extreme poverty in rural and urban areas, 2011 ........................................................ 3 Figure 2. Poverty rates by the number of children in a household..................................................................... 4 Figure 3. Social assistance spending as share of GDP, 2010 – 2012,.................................................................. 5 Figure 4. Selected Program Spending Dynamics ........................................................................................................ 8 Figure 5. Benefit levels and evolution of the guaranteed minimum income vs. extreme poverty line ........................................................................................................................................................... 10 Figure 6. Distribution of program beneficiaries and benefits by quintile, 2011 ......................................... 13 Figure 7. Adequacy of last resort social assistance programs ............................................................................ 14 Figure 8. Relative reduction of extreme poverty incidence rate for different years ................................. 15 Figure 9. The extreme poverty gap, 2011 ................................................................................................................... 15 Figure 10. Cost-benefit ratios of social protection benefits, 2011 .................................................................... 16 iv List of Tables Table 1. Poverty estimates for the Kyrgyz Republic, 2005-2012 ........................................................................ 3 Table 2. Poverty indices for individuals by demographic characteristics, 2011 .......................................... 4 Table 3. Expenditure on Social Protection and Selected Programs, 2008-2012........................................... 6 Table 4.Trends in number of beneficiaries ................................................................................................................... 9 Table 5. Benefit coverage of social protection benefits and private transfers, 2011 ................................ 12 Table 6. Share of benefits in total household consumption, beneficiary households only, 2011 ........ 13 v 1. Executive Summary 1. Enhancing the effectiveness of social assistance expenditure, including its impact on poverty, is the key challenge facing the system of social assistance in the Kyrgyz Republic. Non-contributory social protection expenditure doubled between 2008 and 2011. However, performance of the system in terms of coverage of the poor, targeting accuracy, and impact on poverty did not improve. The increase in spending has favored programs largely benefitting individuals and households that are financially better off. At the same time, the Monthly Benefit for Poor Families with Children (MBPF), the program with greatest potential to alleviate extreme poverty and contribute to equalizing opportunities in the years of childhood (critical for success in adulthood), has seen drops in coverage and has been underfinanced. While the MBPF is well targeted to the poor, more than 60 percent of children in the poorest quintile remain uncovered. The benefit provided by the program is inadequate to meet the most basic needs. Expanding coverage of the bottom quintile by social assistance and improving its adequacy would require a gradual shift from a largely categorically targeted system to a poverty-targeted one, and re- allocation of expenditure toward pro-poor programs. 2. The Kyrgyz Republic needs to reform its system of social assistance to better protect the vulnerable against economic shocks. These include, but are not limited to, food and energy price shocks resulting from instability in the global economy and markets, the potential negative impact of energy tariff reform on the poor. Further spending increases will not suffice, since a large part of the poor remain excluded from the system, while much of the spending is going to the non- poor. Reform of the system should include refocusing it towards the poor and making it responsive to existing challenges and potential shocks. The government has already taken some steps in this direction. However, more decisive action is needed to improve the effectiveness of the system. 3. This note proposes a reform agenda to address the challenges facing the system of social assistance This would require: i) a progressive shift of public expenditure toward pro-poor programs; ii) improving coherence and complementarity among social protection programs, and iii) making social assistance programs more inclusive and responsive. More specifically, the analysis in this policy note points to the following options:  Integrated program administration, including unified benefit application, stronger information systems, and better coordination. This will reduce the duplication of functions and generate fiscal savings.  Develop a strategy on safety nets with clearly articulated objectives, expected results, and key performance indicators. To ensure broad-based support, this strategy should be developed with public discussion, stakeholder information, and communication.  Increase MBPF budget outlays, while improving program design and implementation.  Redesign the portfolio of social programs and introduce complementary and coherent programs that protect the poor, and provide opportunities to social assistance beneficiaries to become self-reliant.  Improve the administration of disability benefits by streamlining access to certification services for disabled children and adults, addressing incentive issues and providing opportunities for rehabilitation and earning income. 1  Reform the current regressive allocation of energy compensations and subsidies A shift towards a poverty-targeted approach to mitigate social impact of energy tariff reform could help protect the poor within an affordable spending envelope.  Review pre-independence policies related to early retirement and pension bonuses to assess the rationale of public spending on these programs.. 4. The structure of this chapter is as follows. The next two sections provide an overview of poverty and the composition and trends in non-contributory social protection expenditure in the Kyrgyz Republic. The fourth section looks at performance of the social assistance expenditure in the Kyrgyz Republic and shows that while spending has increased significantly in recent years, its performance in terms of coverage and targeting of the poor, as well as in terms of its impact on poverty remains modest. Other issues, which are important to take into account in the process of improving effectiveness of social assistance expenditure, are discussed in the following section. The last section concludes by summarizing key challenges and proposing recommendations for system reform, as well as areas requiring further research and analysis. 2. Poverty in the Kyrgyz Republic 5. A large share of the Kyrgyz Republic population continues to be poor, despite significant decline in poverty over the past decade. Absolute poverty declined from 43.1 percent in 2005 to 31.7 percent in 2008 and 2009 (Table 1).1 Extreme poverty, defined as the monetary equivalent of total consumption insufficient to secure 2,100 calories, declined at a faster rate from over 11.1 percent in 2005 to 3.1 percent in 2009.2 However, due to exposure to external and internal shocks, poverty has risen in recent years. The combination of rising global food and fuel prices, the global economic crisis and the internal conflict took a heavy toll on the Kyrgyz economy and on the living standards of the Kyrgyz population. 1 The headcount index of poverty reflects the percentage of the people whose consumption is lower than the established poverty line (Table 1). The poverty line is estimated by specifying a consumption bundle with food (including own production) and nonfood items (excluding housing costs). The poverty line was established in 2003 (updated in 2008) and it is being adjusted for inflation on an annual basis. 2 The food/extreme poverty line is empirically derived from the KIHS data. It reflects the actual food consumption habits of the Kyrgyz population, and, more specifically, it looks at food consumption of low- income households. 2 Table 1. Poverty estimates for the Kyrgyz Republic, 2005-2012 (In percent) Measure of welfare 2005 2006 2007 2008 2009 2010 2011 2012 Absolute poverty Headcount index of 43.1 39.9 35.0 31.7 31.7 33.7 36.8 38.0 poverty Poverty gap index 10.5 9.1 6.6 7.5 6.1 7.5 7.5 7.7 Poverty severity index 3.6 3.1 1.9 2.6 1.8 2.5 2.2 3.4 Extreme poverty Headcount index of 11.1 9.1 6.6 6.1 3.1 5.3 4.5 4.4 poverty Source: NSC website http://212.42.101.124:1041/stat1.kg/, World Bank calculations for 2011. 6. Poverty incidence is higher in rural areas. More than 40 percent of people living in rural areas find themselves below the poverty line compared to 31 percent in the urban population (Figure 1). At the same time, the rural poor constitute almost 70 percent of all the poor, which is notably higher than their share in the total population. There are also significant regional disparities, which directly translates into differences in the oblast-level poverty rates. Figure 1. Absolute and extreme poverty in rural and urban areas, 2011 (In percent) Incidence Poverty distribution 50 100 40.4 77.4 40 36.8 80 68.6 31.1 30 60 Absolute Urban 20 40 31.4 Extreme 22.7 Rural 10 5.3 4.5 20 2.6 0 0 Urban Rural National Absolute Extreme Source: World Bank calculations based on KIHS 2011. 7. Children face a higher risk of being poor or extremely poor. Data suggests that 52.9 percent of the under-six year old and 43 percent of the six to 15 years olds live in poverty in the Kyrgyz Republic (Table 2). Seven and a half percent of the under-six year old children live in extreme poverty, which is significantly higher than the national average of 4.5 percent. Males and females are not significantly different in terms of poverty status. 8. With regard to household composition, the highest poverty rates are observed for households with three or more children. Two thirds of all persons in households with three or more children younger than 16 live in poverty (Figure 2). Poverty estimates further suggest that every tenth person living in households with three or more children younger than 16 is extremely poor. This is consistent with the higher poverty incidence observed in larger households. 3 Table 2. Poverty Indices for Individuals by Demographic Characteristics, 2011 (In percent) Poverty headcount Share in total poor by sex: Male 37.6 48.2 Female 36.1 51.8 by age (seven groups): Under 6 years 52.9 16.2 6-15 years 43.0 25.2 16-20 years 33.4 9.3 21-40 years 39.5 26.9 41-60 years 25.8 16.9 61-70 years 28.2 3.2 71 years or older 24.3 2.4 Extreme poverty headcount Share in total extremely poor by sex: Male 4.8 50.0 Female 4.3 50.1 by age (seven groups): Under 6 years 7.5 18.4 6-15 years 4.7 22.3 16-20 years 5.2 11.8 21-40 years 5.1 28.3 41-60 years 3.0 16.1 61-70 years 1.6 1.5 71 years or older 2.1 1.6 Source: World Bank calculations based on KIHS 2011. Figure 2. Poverty rates by the number of children in a household (In percent) Number of children younger than Three or more 10.1 66.7 Two 4.3 41.4 1.3 Extreme 16 One 22.3 Absolute None 0.6 9.2 0 20 40 60 80 Poverty rate Source: World Bank calculations based on KIHS 2011. 9. Summing up, the positive trend of declining poverty came to a halt in 2009. In 2012, absolute poverty rates were back to pre-2007 levels, while extreme poverty remained stable at around five percent. Children face higher poverty risk, both in terms of absolute and extreme poverty. Large households and/or households with many children are particularly at risk of living in poverty. 4 3. Overview of Non-contributory Social Protection Programs and Spending 10. Social protection system in the Kyrgyz Republic is complex. Non-contributory social protection includes a mix of cash benefits, allowances, and subsidies, which are targeted to certain population categories. For child benefit however, a means-test is applied in addition to categorical eligibility. The system is comprised of a mix of old programs inherited from the Soviet past and of some new programs and features introduced since Independence. The Government spent on average more than 2.5 percent of GDP on social assistance in the past three years. This is relatively high compared with other countries with similar income levels, and is slightly higher in comparative terms than the average spending level in Europe and Central Asia region. Spending sharply increased between 2008 and 2010, when the bulk of subsidies and in-kind benefits were monetized and the Government implemented a set of social protection measures following the doubling of electricity tariff for residential consumers in December 2009, which was subsequently rolled back in 2010. Figure 3. Social assistance spending as share of GDP, 2010 – 2012, (In percent of GDP) Latvia 11 Azerbaijan 11 Bulgaria 10 FYR Macedonia 11 Turkey 10 Armenia 12 Kosovo 12 Montenegro 10 Albania 11 Belarus 11 Serbia 10 Moldova 10 Ukraine 11 Kyrgyzstan 11 Estonia 10 BIH 10 Romania 10 Croatia 11 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% Source: Europe and Central Asia Social Protection Database World Bank. 11. The analysis in this Box 1. World Bank definition of Social Assistance chapter is focused on several Social assistance programs, also referred to as Safety Net, provide key programs. Non-contributory non-contributory transfers financed from general revenues and social protection in the Kyrgyz generally targeted in some manner to the poor or those vulnerable to Republic includes a number of shocks, such as: cash transfer programs; energy  Cash transfers, targeted or not, conditional or not; compensations and discounts;  Food or other in-kind distribution; benefits, such as early retirement  Public workfare jobs; and pension top-ups paid to  General price subsidies (e.g., for food or fuel); some categories of pensioners;  Fee waivers for essential services such as health or scholarships; exemptions from education; health service co-payments;  Social services provided free of charge and institutional care. 5 school feeding; limited scale public works; social services and institutional care, provided free of charge; and aid to population affected by natural disasters, etc.3 All these programs and transfers are financed from general revenues and are not linked to past contributions, unlike social insurance transfers, such as old-age pensions, unemployment, health, and disability insurance (see Box 1). This Chapter focuses primarily on selected non-contributory social protection programs and spending due to the following reasons.4 These programs represent the bulk of social assistance spending and account for the sharp spending increases between 2008 and 2010. They are also the most important due to their relatively large size, and present interest from the point of view of potential reforms in the sector. Early pensions and top-ups are generally part of social insurance policy, but they are financed from central budget revenues (rather than from the pension scheme member contributions), similarly to key social assistance programs reviewed in this Chapter. Table 3 presents a summary of spending on the key non-contributory social protection programs. Table 3. Expenditure on Social Protection and Selected Programs, 2008-2012 (Kyrgyz som million and percent of GDP) 2008 2009 2010 2011 2012 Total spending on Social Protection* mln KGS 4,680 5,599 11,102 14,274 17,353 as percent of GDP 2.49 2.78 5.04 5.23 5.70 MBPF mln KGS 654 778 1,168 1,348 1,488 as percent of GDP 0.35 0.39 0.53 0.49 0.49 Monthly Social Benefit (MSB) mln KGS 552 599 1,147 1,513 1,811 as percent of GDP 0.29 0.30 0.52 0.55 0.60 Cash compensations mln KGS 688 1,017 2,191 2,540 1,927 as percent of GDP 0.37 0.51 0.99 0.93 0.63 Electricity compensations mln KGS 475 526 1,720 1,750 1,772 as percent of GDP 0.25 0.26 0.78 0.64 0.58 Early retirement and top-ups** mln KGS 803 1,341 1,869 2,116 2,563 as percent of GDP 0.43 0.67 0.85 0.77 0.84 All selected programs as percent of GDP 1.69 2.12 3.67 3.39 3.14 Source: Treasury of the Kyrgyz Republic and WB staff calculations. * Total spending on Social Protection here includes all spending financed from the state budget, but does not include spending by the Social Fund financed by the pension scheme member contributions. Detailed information on all social protection spending is presented in table A1-1, Annex 1. **Composition of spending on early retirement and pension top-ups is presented on figure A1-1 in Annex 1. 3 See Annex 2 for the list of key programs with information on agencies responsible for their administration, budget allocations and average benefit values. 4 These programs are: (i) Monthly Benefit for Poor Families with Children (MBPF); (ii) Monthly Social Benefit (MSB); (iii) Cash Compensations (former privileges, in-kind benefits and price subsidies); (iv) Energy compensations to pensioners; and (v) Early retirement benefits and pension top-ups paid from general revenues. 6 12. Spending on the selected programs significantly increased in 2010, with the pace of increase moderating since. The largest increases were registered in Cash compensations and Electricity compensations (see Table 3). Monetization of privileges contributed to the increase in social assistance spending. The system of privileges inherited from the past, was reformed in 2007 – 2010 and replaced by the program of cash compensations paid to 25 groups of beneficiaries (see Box 2). The increase in spending was also driven by social protection measures intended to compensate vulnerable groups in view of the electricity tariff increase in late 2009. The measures included increases in all main social assistance benefits, including: energy compensations to pensioners, reinstatement of a lifeline tariff subsidy scheme for population in high altitude regions, and top-ups on wages of public sector employees earning less than KGS 5,000. The tariff increase was fully rolled back by the interim government following the April 2010 events; however, the compensation measures were retained. Spending on early retirement and merit top-ups has also been on the rise due to significant growth in the number of workers retiring early. The number of pensioners receiving early retirement pensions, fully paid for by the budget, increased 1.7 times between 2007 and 2011. Box 2. Social Protection system in the Kyrgyz Republic and financing of social benefits Social protection in the Kyrgyz Republic is multi-faceted. It includes social insurance programs, such as old-age pensions, unemployment benefits, death, disability, and health insurance. These contributory programs are administered by state entities including the Social Fund, the Ministry of Youth, Labor and Employment, and the Mandatory Health Insurance Fund. Social assistance benefits and non-contributory benefits (early retirement and pension top-ups) account for the largest part of social protection spending from the central budget (3.14 out of 5.7 percent of GDP in 2012). They include a number of cash benefits, administered primarily by the Ministry of Social Development and by the Social Fund. The third key component is social care. It is largely limited to institutionalized care for children, people with disabilities, and the elderly and is administered by the Ministries of Social Development, Education, and Health. The main social assistance benefits are financed from the central government budget. Financing of other types of non-contributory cash transfers such as energy compensations, early pensions, and pension top- ups, which are channeled through the Social Fund, also comes from the central government budget. Only 6 percent of total expenditure on social benefits, other than those channeled through the Social Fund, was financed by local budgets in 2011 (Table A1-5 in Annex 1). Benefits for temporary work incapacity, funeral costs, public works, and certain privileges are the types of benefits where the local level contribution is relatively significant. It is important to note that besides payments of certain benefits, local level authorities have the obligation to finance salaries of social workers at ayil okmotu level. Also, they have the discretion to provide additional benefits such as housing subsidies, income supplements and grants, food and solid fuel, as well as other types of in-kind benefits. 7 Box 3. Reform of the system of privileges The Kyrgyz Republic inherited from the Soviet times a safety net consisting of a large number of categorical poorly targeted and costly benefits (“privileges�). Prior to 2010, 39 categories of the population (e.g. , World War II veterans, Chernobyl catastrophe liquidators, people with disabilities, etc.) were entitled to over 40 types of subsidies and in-kind assistance. These included free transportation, discounted utility bills, free medicines, free dentures, and so on. Entitlements to such benefits did not necessarily mean that the benefits were fully financed and used. Many people have been de facto excluded from the system since they did not have access to the benefits (e.g., people with limited ability to travel could not use free transportation, or there were no supplies of free medicines in the stores). In 2007 - 2010 the Government carried out a reform of the program monetizing most of the in-kind benefits. As of now, 25 categories of beneficiaries continue to be entitled to the benefit that is provided as flat monthly cash payments ranging from KGS 1,000 to 7,000 per month. This reform has achieved higher transparency of public expenditures, expansion of consumer choice through provision of cash instead of notional but often unavailable benefits, simplified administration of the program. 13. Among all non-contributory social protection programs, only the Monthly Benefit to Poor Families with Children (MBPF) is explicitly targeted to the poorest. It is the only program that uses a means test in addition to categorical identification to define eligibility. The program provides monthly cash benefits targeted to children to bring their per capita income up to the Guaranteed Minimum Income (GMI) level, which was equal to about one-third of the extreme poverty line in 2011.5 To qualify for the MBPF, applying families must have per capita income lower than the GMI as assessed by the district social protection departments using the guidelines set in the Law on State Benefits. In all the other programs, beneficiaries may claim eligibility if they belong to a certain category (categorical programs).6 Eligibility in these programs is not linked to welfare status. Most common among more than 25 categories of the population eligible for categorical social assistance are pensioners, the elderly, veterans, mothers of many children, the disabled, decorated citizens, and survivors, among others. The categorical programs accounted for more than 80 percent of the total social benefit spending in 2011. 14. Expansion in spending in Figure 4. Selected Program Spending Dynamics recent years has favored the (Percent of GDP) categorical programs. Figure 4 4 Early retirement and shows how spending on the five 3.5 top-ups programs evolved in the review 3 Electricity period. Since 2010 spending on Percent of GDP 2.5 compensations selected programs gradually declined 2 Cash compensations as percentage of GDP. In absolute 1.5 terms, spending on all programs MSB 1 continued to grow, with exception of 0.5 cash compensations. The MSB MBPF 0 program continued to grow in both 2008 2009 2010 2011 2012 absolute and relative terms. The rate of increase in the MBPF, the only Source: Ministry of Finance. 5 Children under three get a flat rate benefit equal to the GMI. Twins and triplets get higher flat rate benefits. Children under 16 are eligible, in addition to youth between the ages 16–18 in schools. 6 In Electricity compensations program beneficiaries also must have a pension below a certain threshold to receive the benefit. The benefit also varies depending on the size of pension. In 2011, 95 percent of all pensioners were receiving the benefit. 8 poverty-targeted program, has been the lowest among the programs that continued to grow. In cash compensations decline in nominal spending was registered between 2011 and 2012 (Table 3). This decline is explained by natural decrease in the number of WWII veterans (who are eligible for the largest compensation amounts), and the efforts of the Government to keep in check the number of categories eligible for cash compensations and the compensation amounts (which are already relatively generous as will be shown hereinafter). 15. Coverage in categorical programs has been expanding. Trends in benefit recipient numbers in the social assistance programs merit special attention, as these dynamics have a direct impact on spending levels. As was shown above, spending on the categorical programs has grown significantly. Two factors can contribute to increases in spending: change in numbers of benefit recipients and change in benefit values. As shown in the Table 4 below, the number of recipients has been growing in most of the categorical programs. In cash compensations, the monetization and reduction in the number of eligible categories led to a reduction in the number of recipients. However, a large part of the former privilege beneficiaries is still eligible for electricity subsidy and thus have retained the privilege, but were re-assigned to a program different from the cash compensations. This group of beneficiaries is called population living in high altitude areas. It counted about 172 thousand households in 2010. Growth was also registered in other categorical programs, including in electricity compensations and in the number of early retirees. In contrast, the number of beneficiaries in the MBPF program has been falling. In previous years, this reduction could be attributed to the decline in poverty, but the number of beneficiaries continued to decline despite increase and stagnation of poverty incidence during 2010 - 2012. The reduction has also been linked to the efforts to reduce inclusion errors by the Ministry of Social Development (UNICEF, 2011). Table 4.Trends in number of beneficiaries (Persons, in thousands) Number of beneficiaries 2005 2006 2007 2008 2009 2010 2011 MBPF 481 475 434 362 371 362 377 MSB 57 59 59 61 63 68 70 Cash compensations 439 434 303 296 285 54 56 Electricity 455 466 464 503 516 compensations Early retirement 46 64 68 71 78 Pension top-ups 45 47 45 45 48 Source: Ministry of Social Development and Social Fund. 16. Growth in the number of beneficiaries has been uneven by category and geographic area. In the MSB program, growth in the number of beneficiaries has been higher in Bishkek, Chui and Issyk-kul oblasts (Figure A1-2 in Annex 1). It was also higher in categories such as orphans and disabled of the third group (Figure A1-3 in Annex 1). The higher growth in the latter category was linked to a proportionally higher increase in benefit level in this category relative to others in the previous years (Table A1-2 in Annex 1) (UNICEF, 2011). The reasons for higher growth rates in some of the geographic areas relative to others are not clear. In cash compensations, categories such as WWII veterans and their survivors, as well as Chernobyl veterans and their descendants, are shrinking. However, the decrease in these categories has been offset by increases in categories such as rehabilitated citizens, honorable blood donors and people with visual and hearing impairments (Figure A1-4 in Annex 1). 9 17. There were 78,000 early retirement benefit recipients in 2011, 71 percent more than in 2007. For comparison, the total number of old-age pensioners increased by about ten percent over the same period. Mothers of five or more children and mothers of disabled children make 64 percent of all early retirees. Twenty-seven percent are retirees living in high altitude areas, 68 percent of them being mothers of three or more children. The rest of early retirees include people living in remote and hardly accessible areas, those who took part in the liquidation of the Chernobyl accident, and former workers of hazardous industries and reorganized enterprises.7 18. Benefit values have increased in almost all programs. Most significant increases in benefit values were registered in 2009-2010 period. They were linked to monetization of the privileges and the compensation measures following the doubling of the electricity tariff for residential consumers, as mentioned above. The MSB benefit value was de-linked from the GMI in 2009 and since then, the benefit values have increased significantly and unevenly for the 14 categories of beneficiaries (Table A1-2 in Annex 1). In Cash compensations, benefit values have been kept constant since 2010, but they are bound to increase if electricity tariffs are raised. Average Electricity compensation has increased significantly from KGS80 in 2009 to KGS280 in 2010 following the tariff hike in 2009, which was subsequently reversed (Figure A1-5 in Annex 1). 19. Early retirement benefits increased along with the average increase in pensions. For example average pension received by a mother of five or more children increased from KGS 621 in 2005 to KGS 2,570 (Table A1-3 in Annex 1). Pension top-up amounts have also increased (Table A1- 4 in Annex 1) with the highest increase observed in top-ups for special merits, which grew by 268 percent on average between 2007 and 2011. 20. Average benefit size in MBPF is much lower than in the other key income-support social assistance programs (Figure 5). Benefit size in MBPF largely depends on the GMI level. The GMI was set at 50 percent of the extreme poverty line (EPL) at inception in 1998 with an intention to bring it up to 100 percent of the EPL over time. However, it has lagged behind the EPL ever since 1998, not exceeding 30 percent of its level in recent years (Figure 5). The GMI is adjusted irregularly, based on allocations agreed with the Ministry of Finance and projected number of benefit recipients. Adequacy of the GMI and of the MBPF benefit remains one of the key concerns and reduces effectiveness of this program in alleviating poverty. Figure 5. Benefit levels and evolution of the guaranteed minimum income vs. extreme poverty line (Kyrgyz som) Average benefit size, 2011 (Soms) 3500 3000 2500 2000 1500 1000 500 0 MBPF MSB Cash Compensations Source: Ministry of Social Development and National Statistics Committee. 7 Pensions of the latter two categories are paid by the employers or from social insurance contributions. 10 4. Performance of the Social Assistance Programs 21. To assess performance of social assistance programs analysts commonly use indicators of coverage, targeting accuracy, adequacy and impact on poverty. These indicators are assessed based on household surveys and are often compared and contrasted with administrative data collected by offices in charge of social assistance programs. Use of household surveys, which collect data on household incomes and expenditure and allow assessing distribution of social assistance benefits across a population, among other uses. All households, participating in a survey, can be ranked from the poorest to the richest based on their per capita income or consumption (welfare status). Commonly, the population is then divided into welfare groups such as quintiles or deciles. Incidence and distribution of social benefits across welfare quintiles or deciles indicate how well social assistance benefits are targeted to those most in need, and what proportion of them is covered by a certain program. 22. The most recent available dataset from the 2011 Kyrgyz Integrated Household Survey (KIHS) is used for analysis of incidence and distribution in this chapter. Incidence and distribution analysis of cash compensations, energy compensations and subsidies, as well as of early pensions and top-ups cannot be directly derived from the KIHS data due to the survey design. We estimate incidence and distribution of cash compensations on the basis of simulated benefit receipt. Using information on the reported size of the pension and the location of the household, we also estimate the coverage and distribution of energy compensations and subsidies. Pensions and private transfers feature in the analysis for illustrative purposes. Box 4. Establishing welfare quintiles We use the average per capita household consumption, calculated by the National Statistics Committee (NSC), as the main household welfare indicator. The welfare quintiles to assess benefit performance are based on counterfactual household consumption per capita in the absence of transfers. The marginal propensity to substitute consumption in the absence of social transfers is estimated at 33 percent. This means that a 100 KGS more transfer would result in 33 KGS increased consumption. In the absence of transfers, households would substitute 66 percent with other income. We review the distributional impact in terms of beneficiaries (coverage and distribution) and benefits (targeting accuracy and benefit adequacy) for the different social transfers. 23. In 2011, almost half of the Kyrgyz population received a social transfer (including pensions). The share of transfer recipients was the largest among the poorest 20 percent of the population – two thirds of them benefited from a transfer (Table 5). The largest share of the population lived in a household receiving a pension (43 percent); while the coverage of the poorest quintile with pensions was 55 percent. Other types of transfers with significant coverage rates include the transfers from relatives (25 percent) and MBPF (nine percent). Overall, an estimated 35 percent of the population benefits from energy compensations for pensioners and ten percent from electricity subsidies allocated to population living in high altitude areas. Coverage rates for electricity compensations and subsidies are rather similar across the welfare distribution. The overall coverage by the MSB according to the KIHS is low, between one and two percent. 24. The MBPF is primarily allocated to the poorest households but many low-income households remain excluded. With 22 percent of the poorest 20 percent of the population benefiting from it, the MBPF is the main social assistance program targeted to the poor. However, almost 78 percent of the poorest 20 percent of the population are not covered by the benefit. This is partly explained by the fact that the program is targeted to children only, and by absence of a social assistance program that would be targeted to all poor. 11 Table 5. Benefit coverage of social protection benefits and private transfers, 2011 (Percent) Type of benefit Quintile I Quintile II Quintile III Quintile IV Quintile V Total Any social transfer 66.2 57.7 48.1 38.6 35.5 49.2 Pensions 55.4 49.0 42.4 33.6 32.5 42.6 MSB 1.2 1.7 1.6 0.5 0.8 1.2 MBPF 22.3 11.3 6.1 4.8 1.4 9.2 Energy compensations 36.3 38.5 35.8 32.7 32.7 35.2 Energy subsidy for high altitude 10.5 9.8 14.1 7.2 7.9 9.9 Cf. Money from relatives 32.3 26.4 23.2 20.7 23.0 25.1 Source: World Bank calculations based on KIHS 2011. 25. MBPF is targeted to children only and, as was shown in section 1.2, poorer households have more children. Estimated 41 percent of children in the poorest quintile were covered with the MBPF. This is higher compared to 22 percent of all people in the poorest quintile, but also means that 60 percent of the poorest one fifth of children is not covered. Various reasons may explain this, including low awareness, insufficient understanding of how the program works, or the incorrect exclusion of those who apply. However, the most important factor accountable for low coverage is the low value of the GMI, at about 30 percent of the extreme poverty line, which is used as the threshold to determine eligibility. Based on the distribution of incomes from the KIHS, less than five percent of the population lived in households with incomes below the GMI in 2010 (World Bank, 2013). Benefit assignment based on the GMI level excludes households with formal and/or traceable incomes above the GMI but below the extreme poverty line. Box 5. Four roles for Social Assistance 1. Well-targeted social transfers have an immediate impact on inequality and extreme poverty. 2. Transfers enable households to make better investments in their future, including:  In children’s human capital (nutrition, education, health)  In their livelihoods 3. Transfers help households manage risk  Avoiding hard to reverse losses  Allowing higher risk/higher return strategies 4. Strong social assistance systems can help governments make beneficial reforms, e.g. subsidy and tariff reforms. TARGETING ACCURACY AND BENEFIT ADEQUACY 26. Almost 50 percent of MBPF beneficiaries belong to the poorest 20 percent of the population. The distribution of beneficiaries receiving selected social assistance across the five consumption quintiles is presented in Figure 6. In 2011, almost 50 percent of MBPF beneficiaries belonged to the poorest quintile and only three percent to the richest. This is a significant change compared to year 2008 when only 38 percent of beneficiaries belonged to the bottom quintile. Almost half of the MSB beneficiaries belong to the poorest 40 percent of the population. Distribution of beneficiaries receiving the other three benefits is different. The majority of the beneficiaries in these programs do not belong to the poorest 40 percent of the population. 12 Figure 6. Distribution of program beneficiaries and benefits by quintile, 2011 (Distribution of beneficiaries) (Distribution of benefits) 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% Quintile V 50% Quintile V 40% 40% 30% Quintile IV 30% Quintile IV 20% 20% 10% Quintile III 10% Quintile III 0% 0% Quintile II Quintile II Quintile I Quintile I Source: World Bank calculations based on KIHS 2011. 27. The MBPF program is the most pro-poor of all programs. The share of the MBPF benefits received by the bottom two quintiles is above 80 percent while only less than two percent of the MBPF is absorbed by the richest quintile. According to the 2011 data, the largest shares of the MSB are received by households belonging to the second and third welfare quintile. Only ten percent is received by the poorest 20 percent of the population. This data needs to be interpreted with caution due to the limitations of the survey data on the MSB program. With respect to other social transfers, Figure 6 shows that cash compensations benefit the richest households to a larger extent. Also, more than half of energy compensations and subsidies are received by the richest 40 percent of the population. They appear to be regressive in terms of their targeting performance. 28. Social transfers constitute almost 30 percent in the total consumption of beneficiary households in the poorest quintile. Table 6 shows the adequacy or relative importance of different public transfers as a share of average consumption in each quintile for all households and beneficiary households only. The MBPF is progressive in relative terms – it contributes a higher share of total consumption for those in the poorer quintiles. However, the size of this contribution is rather limited due to the low coverage and small benefit levels. The MBPF program constitutes only nine percent and the MSB five percent of the total household consumption among recipient households in the bottom quintile. Table 6. Share of benefits in total household consumption, beneficiary households only, 2011 (In percent) Type of benefit Quintile I Quintile II Quintile III Quintile IV Quintile V Total social transfer 29 31 30 32 34 Pensions 34 33 32 35 35 MSB 5 12 5 15 6 MBPF 9 6 4 2 2 Money from relatives 12 18 19 16 18 Source: World Bank calculations based on KIHS 2011. 29. The data further suggest that the pension transfer is very important for poor beneficiary households. Its share in the household consumption is 34 percent. With 12 percent share in the household consumption, money from family and relatives (including remittances) also plays a significant role for the beneficiary households in the bottom quintile. 13 30. Compared to the outcomes in preceding years, the consumption power of both the MBPF and MSB remained low. Not surprisingly, the MBPF adequacy is meager compared to benefit adequacy of similar means-tested programs in the region. In other words, all other countries have social assistance programs of last resort, which provide more adequate benefits. Figure 7. Adequacy of last resort social assistance programs (Benefits as percent of total post-transfer household consumption, 2007–2011) 70 60 50 40 30 20 10 0 Source: Europe and Central Asia Social Protection Database World Bank. 31. The MBPF, the only means-tested program targeting the poor households, achieves good targeting accuracy compared to categorical programs. The MSB is distributed evenly across welfare quintiles and the other categorical programs more often benefit the richer groups in the population. The analysis suggests high exclusion error in MBPF: 60 percent of children in the poorest quintile do not receive the benefits and moreover the value of the benefits is inadequate to cover even the most basic needs. The low coverage by MBPF is primarily related to its design based on the GMI as was discussed above. The World Bank has recently explored and suggested options to increase coverage and enhance poverty impact of the MBPF using different design and targeting approaches to selecting beneficiaries (World Bank, 2013). These options include tying the MBPF to a different living standard benchmark, such as the subsistence minimum, which is updated on a quarterly basis, to prevent its real value from erosion. They also include simulated coverage and poverty impact of the MBPF using alternative targeting approaches, such as a combination of geographic and categorical targeting, and the proxy-means test. The simulations show that a combination of measures, including revised GMI-setting method, improved targeting, and increased budget envelope can lead to improved coverage of the poorest children with MBPF, as well as enhance its impact on poverty. POVERTY IMPACT 32. In the absence of social and private transfers, extreme poverty would be much higher. When we take into account all social transfers for 2011, the extreme poverty is reduced by 2.7 percentage points or 37.3 percent in relative terms (Figure 8). Simultaneously, the extreme poverty gap decreases from 1.0 percent to 0.5 percent (Figure 9). In 2005 and 2008, the relative reduction of poverty was also between 35 percent and 40 percent. Pensions and private transfers are again included in the analysis for comparison purpose. 14 Figure 8. Relative reduction of extreme poverty incidence rate for different years (In percent) 40 %relative reduction 35 30 25 2005 20 2008 15 10 2011 5 0 Any transfer Pensions Money from relatives MBPF MSB Source: World Bank calculations based on KIHS 2005, 2008, 2011. 33. While poverty reduction is not the main objective of the pension system, pensions seem to be most effective in reducing poverty. In 2011, pensions reduced poverty by almost 35 percent, which is slightly higher than in the preceding years (Figure 8). The poverty reduction ability of pensions is due to the fact that pension benefits are significantly higher than the non- contributory transfers and absorb the largest transfer from the state budget and the Social Fund. 34. The second most important transfer in terms of poverty reduction concerns private transfers such as remittances and other forms of family support. In 2011, they reduced the extreme poverty rate by 12.7 percent. At the same time, private transfers decreased the extreme poverty gap from 0.6 percent to 0.5 percent (Figure 9). 35. The MBPF has become relatively less effective in reducing poverty. This is attributed to the low coverage and inadequate benefit levels. The MBPF program has a limited impact on poverty. The headcount rate of extreme poverty decreased by 4.4 percent and the poverty gap by 7.5 percent only due to the MBPF (Figure 8). All other social transfers did not appear to have a significant poverty reduction power in 2011. Figure 9. The extreme poverty gap, 2011 (In percent) Any transfer 1.01 0.94 Pensions Poverty gap (after MSB 0.49 transfer) Poverty gap (before 0.53 0.59 transfer) Money from MBPF relatives Source: World Bank calculations based on KIHS 2011. 36. The MBPF program is most efficient in reducing the poverty gap. Figure 10 shows cost- benefit ratios of social protection benefits. The cost-benefit ratio is used to measure how much it 15 costs to reduce the extreme (or absolute) poverty gap by 1 KGS.8 In 2011, the MBPF program needed 20 KGS to reduce the extreme poverty gap by 1 KGS. While the MBPF still shows the most favorable cost-benefit ratio (20 KGS), it is now much higher than in 2005 (4.7 KGS) and 2001 (1.7 KGS). Low coverage and the increasing discrepancy between the GMI and the extreme poverty line are possible explanations for the reduced cost efficiency over time. The latter argument is partly supported by the cost efficiency of the MBPF when considering the absolute poverty gap. It costs 3.9 KGS to reduce the absolute poverty gap by 1 KGS, which is in line with the findings of previous years. It is more costly to reduce the extreme poverty gap and this occurs when the program is not able to identify and reach the poorest among the poor. Figure 10. Cost-benefit ratios of social protection benefits, 2011 (Kyrgyz som) Extreme poverty Absolute poverty MSB 223 Pension 9.6 Other SI 116 MSB 9.2 Pension 65 Other SI 9.2 MBPF 20 MBPF 3.9 0 50 100 150 200 250 0 2 4 6 8 10 12 KGS spent per 1 KGS reduction of extreme poverty KGS spent per 1 KGS reduction of absolute poverty gap gap Source: World Bank calculations based on KIHS 2011. 37. The presented data suggest that pensions are the second most efficient program in reducing the poverty gap. However, important noting is that the total amount of transfers for pensions is much higher. If the total transfer and MBPF benefit level are raised and the exclusion error reduced it would, under ideal circumstances, increase the impact on poverty reduction. We reiterate the fact that the value of the GMI that is used to determine the MBPF eligibility and benefit levels is significantly lower than the extreme poverty line, which impacts on the currently low poverty reduction power of the MBPF. 5. Other Issues 38. The Government of the Kyrgyz Republic approved a medium-term Social Protection Development Strategy for 2012-2014 in 2011. The main problems identified in the Strategy are: the insufficient financial support for poor families and the disabled due to low benefit amounts; the misalignment between pensions and social assistance benefits; and the unmet demand for quality social services for vulnerable families and children, the disabled, and the elderly. The priority policy directions identified by the strategy include increasing the benefit amounts paid to low-income families with children (MBPF) and the disabled (MSB), deinstitutionalization of social care, and development of alternative social services. The issue of equity in allocation of social benefits is not 8 The measure is obtained when the reduction of the poverty gap in KGS is weighted against the program cost. The cost-benefit ratio is calculated by dividing the total value of transfers per benefit (which deviates significantly from the administrative data) by the reduction of the poverty gap. The administrative costs and costs incurred by the beneficiaries are not included in the calculation since they are not available in the survey dataset. 16 clearly addressed in the Strategy, i.e. an assessment of whether the social benefits reach the neediest, to what extent poor people are covered by the system, and what needs to be done to improve this is lacking. It also does not suggest a vision for the future of the non-contributory social protection system by and beyond the 2014. Box 5. Strategy on Social Protection Development for the Kyrgyz Republic 2012-2014 The Strategy sets medium-term objectives aimed at improving the well-being of vulnerable population groups. The main strategic goal is to establish an effective and fair system of social protection. The main objectives of the Strategy are specified as follows:  Improving the access of families and children in difficult life situation to state social transfers, guaranteed social services aimed at providing children with a family environment;  Improving the access to social services and transfers for persons with disabilities and creating conditions for equal opportunities for participation in social life;  Providing the elderly with adequate benefit levels and high quality social services;  Creating the social infrastructure for provision of prevention and social rehabilitation for homeless people. 39. An issue related to the above is the need to more clearly define objectives and performance criteria of individual social assistance programs. In the absence of clear criteria of effectiveness and of performance indicators, the key stakeholders seem to disagree on performance of individual programs and on what needs to be done. Developing a monitoring and evaluation framework based on clear objectives and performance criteria would be an important step towards improving the effectiveness of the social assistance system. This would also help to focus reform measures on the more important medium- and long-term issues rather than on short- term responses. 40. Fragmentation of social protection policy making across the government ministries and agencies without adequate coordination mechanisms leads to inconsistent policies between and within sectors. For example, social insurance and social assistance programs are not aligned. Survivors and pensioners with some contribution history may get smaller benefits under the social insurance scheme, as compared to survivors and pensioners with no contribution history at all under a social assistance program (MSB). This provides an additional disincentive not to contribute to the pension fund.9 41. Social benefits have overlapping eligibility criteria. For instance, households in which recipients of cash compensations live often will also receive the energy compensations paid to pensioners. Households with pensioners living in high-altitude regions often will receive both the energy compensations and will benefit from the lifeline tariff subsidy. At the same time a large part of poor households do not receive any benefit. 42. Administration of social assistance programs is an area that needs an in-depth assessment. Majority of the programs are administered by central government agencies, which commonly have offices at district level. Municipalities and local self-governments also play an important role, for example, local self-governments are responsible for preliminary identification of 9More on disincentives to contribute to the pension fund can be found in the chapter on pensions. The chapter also discusses the risks to the system of social assistance, which provides social pensions to people with no pension contribution history, posed by the currently high under-coverage of workers by the pension scheme. 17 the needy households through the so-called social commissions. 10 However, roles and responsibilities of program managers at the central level, field staff and local governments and commissions need clearer definition and better enforcement. Staff capacity remains a concern across the sector at various levels. The Ministry of Social Development (MoSD) and other agencies are in the process of developing and rolling out electronic beneficiary registries. Social assistance beneficiary registry at the MoSD will allow streamlining administration of programs, improved tracking of financial flows, monitoring and analyzing trends, and taking informed decisions. Data exchange capacity among the registries of social protection beneficiaries, as well as with the other registries, such as tax and property, is yet to be developed. Management of financial flows is another area that needs strengthening. Pensions and benefits are predominantly delivered through the national postal service Kyrgyzpochtasy. Irregularities, corruption and violations have been registered in delivery of payments to the beneficiaries, but the nature and scale of these have not been assessed in recent years. 43. The system of social assistance has played a limited role in protecting the poor against shocks such as rising food and energy prices. The main reason for this is the low coverage of the poor with targeted social assistance. The MBPF coverage remains low while other mechanisms of crises response such as unemployment benefits and scalable public works programs are virtually non-existent. No mechanism has been put in place to augment coverage by targeted social assistance, including financing, in times of crisis. Despite these limitations, the MBPF program did serve a vehicle to channel donor support to mitigate the adverse impact of rising food prices on families with children who were already enrolled in the program in 2008-2009. Specifically, the World Bank and the EU topped-up the benefits provided under the MBPF in an attempt to help offset the increases in the cost of living due to rising food prices. This support was temporary, however, and limited to existing beneficiaries without increasing appreciably the coverage of the program. 44. The poor must be adequately protected in the process of energy tariff increase. In the short-term the Government may consider re-introducing lifeline tariffs. In the medium-term, improving the targeting, coverage and delivery mechanisms of social assistance programs, such as the MBPF, is considered more cost-efficient and should be implemented in parallel to introduction of the lifeline tariffs. Once the social assistance system becomes capable of protecting the poor from a variety of shocks, including caused by energy price increases, the lifeline tariffs can be phased out.11 45. Response to special needs includes certification of disability status, payment of social allowances and institutionalization of people with moderate and severe disabilities, including children. The Disability Certification Service (DCS) procedures and processes are outdated. There are limited or no opportunities for meaningful rehabilitation or social integration for the disabled. The current system creates adverse incentives for people with special needs to seek monetary assistance rather than rehabilitation, training and employment opportunities. 10See Annex 2 for the summary of programs and agencies administering them. 11For more detailed discussion on energy tariffs and possible mitigation measures see the energy chapter of the Public Expenditure Review. 18 6. Conclusions and Recommendations 46. The system of non-contributory social protection has undergone several reforms. The most recent reform at the end of 2009 resulted in monetization of the former in-kind categorical benefits, some revisions to the MBPF targeting criteria and the scaling up of measures to mitigate the impact of rising energy tariff on vulnerable households. 47. Spending on non-contributory social protection transfers has almost doubled since 2008, but performance of the system in terms of coverage, targeting accuracy and adequacy, has not improved appreciably. The MBPF is covering an increasingly smaller share of the poor. Efforts to reduce the inclusion errors of the MBPF in recent years seem to have paid off, as leakage of MBPF transfers to non-poor households has further declined. The categorical transfers, such as cash and energy compensations and subsidies, remain poorly targeted. Overall, the poverty reduction impact of the social safety net remains limited. Over time, the cost-efficiency of the various transfers, including the MBPF, has worsened. 48. Cutting regressive categorical programs rapidly is politically challenging and has not been easy in many countries where such benefits were or remain common. A more practical approach in such situations is to phase out such benefits gradually. This can be done by freezing benefit amounts or keeping increases below inflation rate, and gatekeeping not to allow new entrants in the programs. 49. Given the challenges ahead, the Government will have to consolidate and strengthen social protection measures targeted to poor and vulnerable households. Some challenges include the foreseen increase of energy tariffs and further budget consolidation, which will also impact social protection expenditures. 50. The following challenges still remain:  While targeted and efficient use of public resources is a priority for the Government (as reflected in the National Sustainable Development Strategy and other documents), the pattern of social assistance expenditure still favors richer households, rather than the needy.  The social assistance (SA) system rewards certain categories of population for past contributions, as well as a tool to compensate harm. It underperforms in its core role of providing a safety net.  The social assistance system remains complex (many different kinds of benefits with overlapping eligibility) and fragmented. SA programs are not well harmonized among themselves, as well as with social insurance benefits, such as: basic, disability, and survivor pensions.  Subsequently, the SA system is prone to overlap and duplication. Some beneficiaries receive two or more different kinds of benefits, while many of the poor do not receive any support. Administration of benefit programs with overlapping eligibility criteria is costly and is confusing to beneficiaries.  Coverage of the pro-poor MBPF has been falling and its budget grew at a much slower pace relative to other programs. MBPF covers less than one-third of the poorest 20 19 percent, to a large extent due to its design based on the GMI, which remains significantly below extreme poverty line.  The MSB program dynamics needs more research. There are exclusion and inclusion errors, which could be explained by unequal access to medical certification, as well as non-transparent practices in the system. The current system also provides perverse incentives to moderately disabled people, who could potentially earn income.  Electricity compensations and subsidies cover some selected sub-sets of the population and are not targeted to the poorest. While significant resources are spent from the budget, the system remains inequitable: public resources are spent to support consumption of some poor households, but also benefit a large number of the better-off.  More research is required on the early retirement scheme and the regulations setting pension top-ups to analyze the reasons for rapid growth in spending and whether these programs are still relevant and necessary. 51. Reform of the Social Assistance is expected to build on the established programs and strategic directions set out in the government’s key development documents. The following measures are recommended for consideration by the Government:  Develop a medium-term plan to reform the social assistance system by progressively shifting expenditure from categorical to means-tested programs, and gradually phasing out the regressive categorically targeted programs. Phasing out categorical programs can be achieved by continuing the moratorium on the addition of new categories and preventing increase in benefit levels.  The Kyrgyz Republic could consider building a portfolio of complementary and coherent social protection programs that protect the poor, prevent falling into poverty, and provide opportunities to social assistance beneficiaries to become self-reliant.  The objectives, expected results, and key performance indicators of the system of social assistance need clearer definition and need to be shared by key stakeholders. It is recommended to foster broad based support for a clearly articulated safety net reform in the society through public discussion, stakeholder information, and communication.  To address low coverage and adequacy of the MBPF, it is important to continue increasing budget outlays while improving program design and performance in different dimensions (such as outreach, communication, targeting, enrollment and recertification, and graduation).  Consolidate different benefits and unification of eligibility criteria as a medium- or longer-term reform option. This would lead to reduction in benefit duplication and overlap, which would raise cost effectiveness.  Reduce costs by integrating program administration and reducing duplication of functions (e.g. a single benefit application, stronger MIS systems, reduced errors and fraud, etc.).  Address exclusion and inclusion errors in the MSB program by streamlining disability certification, improving access to certification, addressing incentive issues, and providing opportunities for rehabilitation, and earning income. 20  To address the regressive allocation of energy compensations and subsidies, re- consider the policy of mitigating social impact of energy tariff reform. This includes a shift toward poverty-targeted approach to mitigate social impact.  Review the current policy related to early retirement and pension bonuses. This could be done as part of the overall pension reform or separately from it. 21 Annex 1. Tables and Figures Table A1-1. Social Protection Expenditure from the Central and Local Budgets (Thousands Soms, 2012) Name Republican Budget Local Budget Salaries 339 981.44 81 198.09 Contributions to Social Fund 60 569.99 13 059.44 Expenditures on business trips 11 446.36 1 172.17 Utilities and communication services 33 748.15 8 552.06 Rent of buildings and premises 439.30 0.00 Transport services 7 033.89 1 372.68 Purchase of other services 62 329.46 40 141.81 Procurement of drugs and medical items 4 255.50 863.80 Procurement of food 40 920.85 11 648.70 Transfers to the Social Fund, including: 10 652 290.00 13.01 - Basic part of pensions 5 232 355.48 0.00 - Preferential pensions for working in high- mountain and 800 422.89 0.00 hard to reach areas - Pensions to military 1 022 500.00 0,00 - Pensions to mothers of large families and mothers of 1 345 124.76 0,00 children with disabilities - Preferential pensions for working in high- mountain and 85 646.71 0,00 hard to reach areas - Public mandatory personal insurance of military in causa 17 500.00 0,00 mortis, in case of disability - Reimbursements to pensioners for electricity 1 771 536.43 0,00 - Reimbursements for labor injuries 43 700.63 13.01 - Reimbursements to responders of Chernobyl Nuclear Plant 1 266.58 0.00 post-accident clean-up - Increments to pensions of people with disabilities 133 411.72 0.00 - Increments to pensions for special merits 198 824.79 0.00 Social assistance benefits to population 5 263 850.63 322 050.91 - Benefits to low income families 1 008 157.51 5 480.92 - Lump sum benefit at child birth 32 840.50 13.26 - Benefits to mothers of children until they reach 1,5 years of 442 062.50 4.20 age - Unemployment benefits 4 946.95 4.00 - Social transfers to population 1 811 481.60 0.00 - Benefits in case of temporary incapability 0.00 3.74 - Benefits on pregnancy and deliveries 203 405.00 159.37 - Payment of ritual benefits (funeral) 9 313.85 9 243.16 - Compensation replacing privileges and privileges to 1 620 465.48 306 742.97 population - Vocational training expenditures 26 255.86 0.00 - Expenditures on organization of community services 31 943.79 170.79 - Microcredit expenditure 3 377.60 0.00 - Gratuitous aid to population suffered from natural disasters 69 600.00 228.50 Current miscellaneous other expenditures 7 445.50 40.00 Buildings and facilities 5 312.40 185.96 Machines and equipment 1 781.36 1 120.30 Credits, lending and loans 380 400.00 0.00 Total: 16 871 804.82 481 418.93 Source: Ministry of Finance. 22 Figure A1-1. Composition of spending on early retirement benefits and pension top-ups (Percent) 0% 0% 2% 3% Pension bonuses for WWII particpants, etc. 8% Merit bonuses High altitude worker pensions Remote district worker 30% pensions 54% Pensions for mothers of many and disabled children Workplace accident pension 3% Chernobyl clean-up comp. Source: Social Fund. Figure A1-2. MSB beneficiary dynamics by oblast, 2007 - 2012 (unit) 70 60 50 40 30 % change 20 10 0 -10 Source: Ministry of Social Development. 23 Figure A1-3. MSB beneficiary dynamics by category, 2007 – 2012 (unit) 160 140 120 100 80 60 40 20 % increase 0 -20 Source: Ministry of Social Development. Table A1-2. MSB number of beneficiaries and benefit amount by category, 2009 and 2012 (unit) Average monthly benefit (KGS) persons 2009 2012 Disabled children with CPIP 3975 1164 3000 Disabled children 19235 848 1500 Children with HIV or AIDS 164 948 3000 Children born from mothers with HIV/AIDS 49 948 3000 Disabled from childhood - I category 4273 1164 2000 Disabled from childhood - II category 15755 848 1500 Disabled from childhood - III category 5825 532 1000 Disabled - I category 579 848 2000 Disabled II category 3118 532 1500 Disabled III category 1177 316 1000 Elderly citizens, including those in high- 402 3000 mountainous and distant areas 2027 Hero-mothers 183 848 2000 Children, in the event of breadwinner loss 13476 532 1000 Orphans without both parents 547 848 2000 Source: Ministry of Social Development. 24 Figure A1-4. Change in selected categories of cash compensation recipients during 2010-2012 (unit) 40 30 20 10 0 -10 -20 -30 Source: Ministry of Social Development. Figure A1-5. Dynamics in average energy compensation to pensioners (unit) Average benefit, Soms 300 250 200 150 100 50 0 2007 2008 2009 2010 2011 Source: Social Fund. 25 Table A1-3. Early retirement beneficiaries and average pension amounts, 2005-2011 (unit) Year 2000 2005 2010 2011 Total number of pensioners, thousands of persons 525.8 536.0 530.1 545.9 Number of pensioners receiving early retirement pensions, thousands of persons 38.2 42.0 70.8 77.6 percent of the total number of pensioners 12.2 12.6 13.4 14.2 Out of which: Mothers with many and disabled children, thousands of persons 24.6 26.1 45.4 49.3 Spending, million KGS 123.6 194.4 1,013.8 1,519.8 Average size of pensions, KGS 419.0 621.0 1,861.0 2,570.0 Former workers in the high-altitude areas, 9.4 12.1 19.2 20.9 thousands of persons Spending, million KGS 57.4 107.1 534.1 654.0 Average size of pensions, KGS 510.0 789.0 2,318.0 2,768.0 Source: Social Fund. Table A1-4. Bonuses and top-ups paid to pensioners (unit) Year 2007 2008 2009 2010 2011 Number of pensioners receiving bonuses and compensations to pensions (persons) Pensioners in total 45,406 46,677 44,633 45,018 48,003 Age-based 34,569 35,230 34,101 34,718 36,094 Disability-based 8,228 8,833 7,906 7,742 8,484 For special merits 2,609 2,614 2,626 2,558 3,425 Total amount of bonuses and compensations to pensions (in thousand Soms) Pensioners in total 7,057.1 6,830.0 8,053.3 13,660.8 19,669.9 Age-based 3,770.2 3,407.8 3,057.1 3,794.6 5,573.3 Disability-based 593.4 717.6 663.9 633.6 1,123.5 For special merits 2,687.6 2,704.6 4,332.2 9,232.6 12,973.0 Total amount of bonuses (in Soms per recipient) Pensioners in total 155 146 180 303 410 Age-based 109 97 90 109 154 Disability-based 72 81 84 82 132 For special merits 1,030 1,035 1,650 3,609 3,788 Source: Social Fund. 26 Table A1-5. Distribution of selected social benefit spending between republican and local budgets, 2011 (unit) Republican Local budget, budget, percent percent Benefits to needy families 99.56 0.44 One-off maternity benefit 99.99 0.01 Benefits to mothers of children under 1.5 years 99.99 0.01 Unemployment benefits 99.94 0.06 Monthly social benefit 100.00 0.00 Benefits for temporary disability 0.00 100.00 Benefits for pregnancy and delivery 99.98 0.02 Assistance for ritual services (funerals) 22.28 77.72 Cash compensations and privileges 86.44 13.56 Vocational training 100.00 0.00 Public works 92.29 7.71 Microcredit expenditure 96.95 3.05 Assistance to population affected by natural disasters 99.85 0.15 Stipends 0.00 100.00 Total 93.71 6.29 Source: Ministry of Finance. 27 Annex 2. Main social assistance and non-contributory social protection programs, 2012 (unit) № Program Eligibility Agency Coverage Expenditure Average administering the (thousand (millions benefit program people) KGS) 1 Monthly social benefit (MSB) Children and adults with disabilities, survivor MoSD 73.6 1,771.6 2,500 children, and persons of pension age without Social Fund contribution history Additional monthly social benefit Survivors of people deceased and disabled MoSD 0.79 37.3 3,700 victims of April-June 2010 events 2 Monthly benefit to poor families Children from families with per capita income MoSD 361.5 1,488.0 464 with children (MBPF) below the Guaranteed Minimum Income 3 Cash compensations (former 25 categories of citizens, including WWII MoSD 55.4 1,522.9 From 1,000 privileges, in-kind benefits and veterans, their widows, rehabilitated citizens, to 7,000 price subsidies) and in-kind honorable donors, Chernobyl veterans and benefits, including: their survivors, people with hearing and visual impairments, etc. - Treatment in specialized Persons with disabilities, children from MoSD, Federation 1.0 23.2 19,863 recreational institutions needy family of trade unions - wheelchairs, assistive Persons with disabilities MoSD 0.8 56.5 NA appliances 4 Maternity allowance Women in the event of pregnancy and MoSD 28.3 203.4 1,799.5 childbirth, with a history of contributions to the Social Fund, or officially registered as unemployed 5 Funeral allowance Relatives of deceased persons with a history MoSD, local 3.7 10.6 3,073 of contributions to the Social Fund or authorities officially registered as unemployed 6 Social services: Persons with disabilities, senior citizens MoSD 2.3 230 From 4,000 - in residential institutions to 7,000 subordinated to the MoSD; - in residential institutions Children without parental care under the age Ministry of Health 0.2 25.5 subordinated to the MoH; of 4 - in residential institutions Children without parental care above the age Ministry of 3.73 262.3 subordinated to the MoE&S; of 4 Education and Science - rehabilitation centers, crisis Persons with disabilities, children, families, MoSD No No information centers, day-care centers; women, victims of violence information 28 7 Compensation for electricity Persons of retirement age, disabled and Social Fund 532.3 1,771.5 270 survivors of persons with Social Fund contribution history 8 Electricity subsidies for residents Persons living in high altitude settlements MoF, local 171.0 114.0 60 living in high altitude regions (the list of settlements is approved by the authorities Parliament) 9 Housing subsidies to pay for Owners of apartments or houses connected to Local authorities 18.0 No No housing and communal services eligible utilities, with income not exceeding information information legislated levels and spending more than 16% of their income on utilities. 10 Pension top-ups Pensioners with contribution history and Social Fund 48.0 (2011) 332.3 No belonging to certain categories information 11 Early retirement Several categories of citizens, including Social Fund 78.0 (2011) 2,231.2 No workers in high-altitude and remote regions, information mothers of many children, etc. 12 School meals Children - public school students in primary Ministry of 400 475 7-10 grades (grades 1-4) Education and Science (MoE&S), local authorities Source: Ministry of Social Development, Ministry of Finance, Social Fund. 29 References CASE (2008). Kyrgyz Republic social safety net policy note, background report, CASE Kyrgyzstan. Gassmann, F. (2010). To what extent does the existing safety net protect the poor? Ministry of Social Development (2011). Strategy of social protection development of the Kyrgyz Republic for 2012-2014, Bishkek. UNICEF (2011). Public expenditure review, Kyrgyz Republic 2007-2010, Bishkek. World Bank (2009). Social safety net in the Kyrgyz Republic capitalizing on achievements and addressing new challenges, Washington: D.C. __________ (2013). Kyrgyz Republic: Minimum Living Standards and Alternative Targeting Methods for Social Transfers, A Policy Note. Accessed: http://www.wds.worldbank.org/external/default/WDSContentServer/WDSP/IB/2013/06 /04/000356161_20130604143034/Rendered/PDF/781680PN0P10160h0Kyrgyz0Rep020 12Eng.pdf. 30