Report No.: 97416-CR Central America Social Expenditures and Institutional Review Costa Rica July 2 2015 Education Global Practice Health, Nutrition and Population Global Practice Social Protection and Labor Global Practice Latin America and the Caribbean Region Document of the World Bank Costa Rica Social Sector Expenditure and Institutional Review Table of Contents List of Figures ................................................................................................................................. 4 List of Tables and Boxes................................................................................................................. 6 Appendix ......................................................................................................................................... 6 Acronyms ........................................................................................................................................ 7 Acknowledgments........................................................................................................................... 9 I. Executive Summary .............................................................................................................. 10 I.1 Education ........................................................................................................................ 10 I.2 Health ............................................................................................................................. 12 I.3 Social Protection and Labor ........................................................................................... 14 II. Context .................................................................................................................................. 17 III. Recent Trends in Social Spending in Costa Rica............................................................... 19 IV. Performance and Challenges in Education ........................................................................ 29 IV.1 Recent Evolution of Education Public Spending ........................................................... 29 IV.2 Education Outcomes and Challenges ............................................................................. 35 IV.3 Institutional Arrangements ............................................................................................. 42 V. Performance and Challenges in Health ................................................................................. 44 V.1 Recent Evolution of Health Public Spending ................................................................. 44 V.2 Health Outcomes and Challenges .................................................................................. 47 V.3 Institutional Arrangements ............................................................................................. 51 VI. Performance and Challenges in Social Protection and Labor ............................................ 56 VI.1 Recent Evolution of Social Protection and Labor Public Spending............................... 56 VI.2 Social Protection and Labor Outcomes and Challenges ................................................ 60 VI.2.1 Social Security ........................................................................................................ 60 VI.2.2 Social Assistance and labor .................................................................................... 65 VI.2.3 Active Labor Market Policies ................................................................................. 70 VI.3 Institutional Arrangements ............................................................................................. 78 VII. Conclusion and Policy Recommendations......................................................................... 80 VII.1 Education .................................................................................................................... 80 VII.2 Health.......................................................................................................................... 81 2 Costa Rica Social Sector Expenditure and Institutional Review VII.3 Social Protection and Labor ....................................................................................... 84 Appendix ....................................................................................................................................... 87 References ..................................................................................................................................... 94 3 Costa Rica Social Sector Expenditure and Institutional Review List of Figures Figure 1: GDP growth in Costa Rica and Central America, 2001-2013 ...................................... 17 Figure 2: Poverty headcount ......................................................................................................... 18 Figure 3: Inequality ....................................................................................................................... 18 Figure 4: Costa Rica Social Spending 1950s, 1990s, and 2007-13 .............................................. 20 Figure 5: Social Spending as a % of GDP by sector (%) 2007-2013 ........................................... 21 Figure 6: Per capita social public expenditure by sector (2012 or latest year available) .............. 21 Figure 7: Social Spending as a % of GDP by country 2013 (%) .................................................. 22 Figure 8: Distribution of social spending by quintiles, 2013 ........................................................ 22 Figure 9: Central government (excluding decentralized entities) overall balance, 2007-2015 .... 23 Figure 10. Budget execution 2007-2013 ....................................................................................... 25 Figure 11: Public Sector Performance and Efficiency in Costa Rica and LAC, 2010 ................. 27 Figure 12: Production Possibility Frontier (Data Envelope Analysis) for Total Social Public Spending, Costa Rica and LAC, 2010. ......................................................................................... 29 Figure 13: Public spending on education versus GDP per capita (%) (circa 2011) ..................... 30 Figure 14: Comparison of public spending in education by levels (%) ........................................ 30 Figure 15: Secondary public expenditure as percentage of country’s GDP (%) .......................... 31 Figure 16: Comparison of the share of wage bills over total public educational expenditure (%)32 Figure 17: Teacher's wage premium in CA .................................................................................. 32 Figure 18: Trend in Student-teacher ratios by level...................................................................... 33 Figure 19: Student-teacher ratios in Secondary vs GDP per capita .............................................. 33 Figure 20: Distributional impacts of public spending by education level, 2013 .......................... 33 Figure 21: Demographic trends in Costa Rica 2010 compared with 2050 ................................... 34 Figure 22: Demographic Projections in Costa Rica: Projections on Student Population ............. 35 Figure 23: Gross enrollment rates by educational level................................................................ 36 Figure 24: Age specific enrollment by and geographic area ........................................................ 36 Figure 25: Post-secondary enrollment by income quintile ........................................................... 36 Figure 26: PISA score versus GDP per capita .............................................................................. 37 Figure 27: SERCE standardized score by socioeconomic status index ........................................ 37 Figure 28: Secondary education attainment of the workforce ...................................................... 39 Figure 29: Distribution of graduates in public and private universities........................................ 39 Figure 30: Daycare attendance by household income quintile, 2014 ........................................... 40 Figure 31: Preschool attendance by household income quintile, 2014 ......................................... 40 Figure 32: High school exam pass rate by school type ................................................................. 41 Figure 33: Takeup of transfer and subsidies by household income quintile, 2014....................... 42 Figure 34: Organizational levels of the basic education system ................................................... 44 Figure 35: Public spending on health as a % of GDP ................................................................... 45 Figure 36: Per-capita public spending on health (real, US$ 2007) ............................................... 45 Figure 37: Public health expenditure as a % of GDP vs GDP per capita ..................................... 45 4 Costa Rica Social Sector Expenditure and Institutional Review Figure 38: Increase in private and public health spending shares of total national health spending since 1995 ..................................................................................................................................... 46 Figure 39: Out-of-Pocket health expenditure by income quintile 2004 and 2013 ........................ 46 Figure 40: Costa Rica MDG 4 and MDG 5 results ....................................................................... 48 Figure 41: Outpatient visits by quintiles, 2006 ............................................................................. 49 Figure 42: Inpatient visits by quintiles, 2006................................................................................ 49 Figure 43: Demographic trends 1990 compared with 2005.......................................................... 49 Figure 44: Waiting lists for specialized services: 2007-2011 ....................................................... 51 Figure 45: Satisfaction levels in Central America with the ways public hospitals work: ............ 51 Figure 46: Public Spending on SPL as a % of GDP (%) .............................................................. 56 Figure 47: Public Spending on SPL per capita constant local currency (2007) ........................... 56 Figure 48: Public Spending on SPL as a % of GDP by countries ................................................ 57 Figure 49: Social Security spending by main categories, 2007-2013 as a % of GDP .................. 58 Figure 50: Social Assistance Spending as a % of GDP 2007-2013 .............................................. 58 Figure 51: Share of workers contributing to SS............................................................................ 61 Figure 52: Share of workers contributing to SS by countries ....................................................... 61 Figure 53: Average Pension Benefit (% of average salaries, 2013) ............................................. 62 Figure 54: Pension coverage 2007-2014....................................................................................... 63 Figure 55: Pension coverage elderly by quintile........................................................................... 63 Figure 56: SS coverage of elderly total and by quintiles (%) ....................................................... 63 Figure 57: Social Pension – Distribution of beneficiaries (Targeting Accuracy) % elderly ........ 64 Figure 58: Main social assistance programs - Coverage first quintile 2009-2014 (% households) ....................................................................................................................................................... 67 Figure 59: Distribution of beneficiaries of main social assistance programs, 2014 (%households) ....................................................................................................................................................... 68 Figure 60: Public spending and % of beneficiaries of main CCTs in Central America ............... 68 Figure 61: Avancemos CCT: Coverage and Spending as a % of GDP ........................................ 69 Figure 62: Poverty headcount before and after the CCT, 2014 .................................................... 70 Figure 63: Enrollment rates ages 5-20, by CCT, 2014 (extreme poor) ........................................ 70 Figure 64: Unemployment rate age 15-64, CA ............................................................................. 71 Figure 65: Unemployment rate (%) 2003-2014 ............................................................................ 71 Figure 66: Unemployment rate by age group 2007-2014 ............................................................. 71 Figure 67: Unemployment rate, youth, CA .................................................................................. 71 Figure 68: Unemployment rate by levels of education 2007-2014............................................... 72 Figure 69: Median earnings by educational level relative to total median earnings .................... 73 Figure 70: Education wage premium 2007 vs 2013 ..................................................................... 73 Figure 71: Public Spending in ALMPs as a % of GDP (%) ......................................................... 74 Figure 72: Budget of Public Training Institutions in Central America, 2013 .............................. 75 Figure 73: Distribution of beneficiaries from INA by age, 2014 ................................................. 76 Figure 74: Distribution of beneficiaries from INA by age and income, 2014 .............................. 76 5 Costa Rica Social Sector Expenditure and Institutional Review Figure 75: Premium to training education, 2009-2014 ................................................................. 77 Figure 76: FODESAF and main institutions................................................................................. 79 List of Tables and Boxes Table 1: Selected Human Development Indicators, Costa Rica, LAC, Central America, and Closest Income/Population Comparators, 2000-2014 .................................................................. 19 Table 2: Health outcomes in Costa Rica: 1995-2011/12 .............................................................. 47 Table 3: Increase in annual cost of treatment for selected NCDs ................................................. 50 Table 4: FODESAF: Budget allocated to SPL programs ............................................................. 59 Table 5: Main Social Insurance Programs .................................................................................... 60 Table 6: Main Social Assistance and social care services ............................................................ 65 Table 7: Coverage main social assistance programs, 2014 (% households)................................. 66 Table 8: Main Labor Market Programs ........................................................................................ 74 Box 1: Public Sector Performance and Public Sector Efficiency Indicators ................................ 25 Box 2: DEA Methodology ............................................................................................................ 28 Appendix Appendix 1: Matrix of Short – and Medium-Term Options for Policy Reform ........................... 87 Appendix 2: Household Surveys databases– Source and definition of variables ......................... 91 Appendix 3: Social spending databases– Source and definition of variables ............................... 92 6 Costa Rica Social Sector Expenditure and Institutional Review Acronyms ADePT World Bank’s Software Platform for Automated Economic Analysis ALMP Active Labor Market Program CA Central America CBA Canasta Básica Alimentaria CCCI Consejos Cantonales de Coordinación Interinstitucional CCSS Social Security Fund. : Caja Costarricense de Seguro Social CCT Conditional Cash Transfer CEN-CINAI Centro de Educación y Nutrición-Centro Integral de Nutrición y Alimentación Infantil CEPAL Comisión Económica para América Latina (Economic Commission for Latin America) CNREE Consejo Nacional de Rehabilitación y Educación Especial CONAPAM Consejo Nacional de la Persona Adulta Mayor CONAVI National Commission for the Prevention and Control of HIV DEA Data Envelope Analysis DPT Diphtheria, pertussis (whooping cough), and tetanus EBAIS Equipos Básicos de Atención Integral en Salud ECD Early Childhood Care EDSTATS World Bank Education Statistics Database EHPM Encuesta de Hogares de Propósitos Múltiples ENAHO Encuesta Nacional de Hogares FIS Ficha de Información Social FONABE Fondo Nacional de Becas FONATEL Fondo Nacional de Telecomunicaciones FODESAF Fondo de Desarrollo Social y Asignaciones Familiares GDP Gross domestic product GINI The Gini coefficient HIS Health Information Systems HIV The human immunodeficiency virus HMN Health Metrics Network ICEFI Instituto Centroamericano de Estudios Fiscales (Central American Institute for Fiscal Studies) ICT Information and communications technology IDB Inter-American Development Bank IMAS Instituto Mixto de Ayuda Social IMF International Monetary Fund INA Instituto Nacional de Aprendizaje INAMU Instituto Nacional de las Mujeres INDER Instituto de Desarrollo Rural INEC Instituto Nacional de Estadística y Censos IPM Índice de Pobreza Multidimensional LAC Latin American and the Caribbean LAC7 Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, and Peru 7 Costa Rica Social Sector Expenditure and Institutional Review LMIC Lower middle Income Country M&E Monitoring and evaluation MDG Millennium Development Goal MEP Ministry of Education MIC Middle-Income Country MIDEPLAN Ministerio de Planificación MIVAH Ministerio de Vivienda y Asentamiento Humanoof Housing MMR Maternal Mortality Rate MOF Ministry of Finance MOH Ministry of Health MTSS Ministry of Labor NBI Necesidades Básicas Insatisfechas NCD Non-communicable disease NGO Non-governmental Organization OECD Organisation for Economic Co-operation and Development PAHO Pan American Health Organization PAI Programa de Abastecimiento Institucional PAISS Package of Health Care Services PANI Patronato Nacional de la Infancia PHC Primary Health Care PISA Program for International Student Assessment PND Plan Nacional de Desarrollo PRONAE Programa Nacional de Empleo PSE Public Sector Efficiency PSP Public Sector Performance RNC Non-contributive pensions RVM Disable, old age and death pension SEDLAC Socio-Economic Database for Latin America and the Caribbean SINAES Sistema Nacional de Acreditación de la Educación Superior SINIRUBE Sistema Nacional de Información y Registro Único de Beneficiarios SIPO Registry of social assistance programs. SPL Social Protection and Labor SCD Sistematic Country Diagnostic SSEIR Social Sector Expenditure and Institutional Review TB Tuberculosis UGM Unidades Geoestadísticas Mínimas UNDP United Nations Development Program WHO World Health Organization 8 Costa Rica Social Sector Expenditure and Institutional Review Acknowledgments This country note was prepared by a team led by Pablo Acosta, Rita Almeida, and Christine Lao Pena, and also comprised by Ciro Avitabile, Eleonora Cavagnero, Alexo Esperato, Emma Mercedes Monsalve, Evan Sloan Seely, and Fernando Montenegro Torres, under the overall coordination of Kathy Lindert, as part of the Central America Social Sector Spending and Institutional Review. We would like to thank Margaret Grosh, Daniel Dulitzky, Reema Nayar, and Fabrizio Zarcone for their guidance, comments, and support in the elaboration of this note. 9 Costa Rica Social Sector Expenditure and Institutional Review I. Executive Summary The evolution of Costa Rica’s social sectors over the past decade has been dichotomous. On the one hand, economic growth has remained relatively high, however poverty and inequality have not declined (moreover, they have increased), and persistent employment challenges remain. On the other hand, the country has continued experiences advances in many social indicators, such as pre-primary and tertiary enrollment rates, access to improved sanitation, and labor force participation, though not in others (secondary school completion, immunizations, employment). Higher economic growth and (to a lesser extent) revenues seem to have allowed a substantial increase in public social spending. The increase in social spending, which now accounts for 20.7 percent of GDP, placing the country at the top of Central America region, and among the leaders in the Latin American and Caribbean region, explain at least part of this improvement in indicators. However, the unequal distribution of income and spending (favoring mostly high skilled and formal workers, but not so much the extreme poor) also explain the disappointing performance in poverty reduction. Looking forward, the key challenges Costa Rica faces are related to continuing improving the quality and efficiency in the social sectors, while improving targeting to serve the most in need, in a tight and severe fiscal context. To expand coverage of excluded population, priority would have to be given to reallocations and improvements within the spending envelope for the social sectors to maximize impact. With a fiscal deficit of more than 6 percent of GDP, further expanding public social spending is no longer an option and budget cuts are looming. Improvements in public spending management and budget execution, including the need of institutional reform to consolidate programs and improve coordination among executing agencies is equally important. In a country that has long been the champion in expanding universal welfare state, sustainability concerns will imply that hard fiscal decisions would need to be made to increase the social returns of budget allocation. I.1 Education Public spending on education has increased in Costa Rica, and it is line with what it is expected given its income level. Public spending on education as a share of GDP grew on average 6 percent per year rising from 3.9 percent of GDP in 2007 to 5.5 percent in 2013. The public expenditure on education is perfectly in line OECD standards, and among countries in the LAC region only Honduras and Brazil spend a higher share of their GDP on education. Basic and tertiary education take up the bulk of public spending on education (with 33% and 26% of the total spending respectively, while the spending on secondary attracts 22% of the resources, a 10 Costa Rica Social Sector Expenditure and Institutional Review share that is disproportionately low for the country’s GDP. Overall public spending is progressive but expenditure in tertiary education is highly regressive; households in the fifth quintile of the income distribution receive in total almost 40% of the public spending in tertiary education. The wage bill represents a large share of the total public education expenditure, and salaries of the administrative staff absorb relatively more than in other countries. Enrollment rates are on average high, but there are large within-country disparities at all levels. Costa Rica’s enrollment ratio for primary and secondary education in 2014 was 108.5 and 117.7 respectively. Enrollment for postsecondary education increased from 28.9 in 2007 to 39.7 in 2014 and is among the highest in the LAC region. However, there are large within-country disparities. Only 3% of the children in the first quintile attend daycare, as opposed to 16% among those in the fifth quintile. Similarly, for children in preschool age, 27% of those in the first income quintile attend preschool, while attendance goes up to 49% for those in the top income quintil. While almost 60% of the 19 year old living in urban areas are enrolled in school, less than 50% of those living in rural areas do so. Moreover, gross attendance rate in post-secondary for those belonging to the highest quintile of the income distribution is 65.1%, as opposed to a gross attendance rate of 23.3% for those belonging to the first income quintile. High spending in education did not result into a relatively higher level of of human capital of the labor force and learning outcomes are not in line with country GDP per capita. Only 35% of the workforce has completed secondary education, a percentage that puts Costa Rica not only behind the OECD average (64.7%) but also behind the LAC average (35.4%). Results from 2012 PISA assessment show that, while the country did on average better than other countries of the LAC region such as Brazil and Colombia, it lagged behind some its “comparator countries”. For instance Serbia and Bulgaria, countries with a GDP per capita similar to Costa Rica, display an average score of 449 and 439 respectively, as opposed to 407 in Costa Rica. Moving forward it is critical for Costa Rica to improve: targeting of subsidies in secondary education; teacher’s quality; and autonomy and accountability both in Basic and Higher education. Only 29% of the 15-19 year old belonging to the first quintile of the household income distribution receives Avancemos, a conditional cash transfer aimed at reducing dropout in secondary among the poorest. A better targeting of subsidies and transfers can significantly weaken the liquidity constraints that affect poorest households. In Costa Rica, the large supply of university degrees to prepare teachers led to an excess supply of teachers and a huge variation in teacher quality. Establishing quality standards, possibly through an entry exam, might raise the quality of teachers at all levels. The monitoring and evaluation (M&E) system can be improved by introducing standardized test scores for students and evaluation systems for teachers. The Basic Education system is a characterized by a high degree of school autonomy but more needs to be done to increase the quality of assessment and accountability. Both public and private universities can rely on large autonomy that was granted as part of the constitution. However, 11 Costa Rica Social Sector Expenditure and Institutional Review there is very little quality assessment and accountability, especially among private universities. Only 68 of the 1,165 university programs have been evaluated by the Sistema Nacional de Acreditación de la Educación Superior (Sinaes). I.2 Health Public spending on health in Costa Rica has grown significantly in the past seven years in absolute terms and as a percentage of GDP. Costa Rica’s public spending on health has increased from less than 6 percent of GDP in 2007 to 7 percent of GDP in recent years. Its health spending as a share of GDP is greater than other countries in Central America and the Latin America and Caribbean average. It is comparable to the OECD average, which includes high income countries such as Sweden and the United Kingdom. Costa Rica has life expectancy and mortality rates comparable or even superior to several developed countries. Its high public health spending together with its emphasis on universal coverage and primary health care interventions has contributed to positive health outcomes. For example, Costa Rica’s life expectancy is 81.5 years for women and 76.7 years for men, while its child (9.9 deaths /1,000 births) and maternal mortality (40 deaths /100,000 births) rates are very low, partly a consequence of a universal institutionalization of births (99.1 per cent). Non-communicable diseases, however, pose a serious challenge to Costa Rica’s hea lthcare system. The rise in NCDs and greater demand for long-term care due to longer life expectancies have presented challenges in terms of costs (costs of treatments have grown dramatically), clinical (new diseases that require new treatments), and declining patient satisfaction (surveys show a decreasing perception of hospital quality and surging waiting lists times at CCSS, particularly for key surgery services). Institutional challenges also create pressures on the health system. The Costa Rican health system also faces the following four main issues: (a) a fragmented health system with the CCSS’s six managers at the central level tending to work in silos; (b) an inadequate health information system; (c) historic budgeting; and (d) the high financial burden imposed by three CCSS national hospitals which duplicate services and absorb considerable resources. In addition, private spending share has increased over time, raising concerns about equity. Private spending’s share of total health spending rose from 23.5 per cent in 1995 to 31.9 per cent in 2010. Although both private and public health spending have been increasing over the last fifteen to twenty years, the burden of spending has not been distributed evenly. In particular, household health expenditures are increasing for the poorest two quintiles while they have been decreasing for the rest. 12 Costa Rica Social Sector Expenditure and Institutional Review The Government prepared a CCSS Institutional Strengthening Proposal which served as the basis for the Bank’s recommendations. The Strengthening the Equity and Sustainability of Health Insurance Proposal (Propuesta Programa FESSS) seeks to achieve a timelier, higher quality and efficient form of care. It focuses on three main areas of improvement: (a) Health Service Delivery: Comprehensive Care Networks and Hospitals; b) Funding and Resource Flow, and c) Management. Based on its review of the above proposal, the Bank proposes the following set of recommendations:  In the short term (1 to 2 years): (a) Health Service Delivery: (i) training of primary care personnel in chronic disease management and (ii) strategies to reduce wait list including a plan prepared to improve the management of Major Ambulatory Surgery to improve wait list management; (b) Funding and Resource Flow: keep historic funding initially in hospitals but gradually introduce prospective payment mechanisms that allow for increased transparency of services such as DRGs and similar payment mechanisms; and (c) for Management: (i) create a General Manager position who would run the institution based on technical criteria and strike a balance between the various CCSS functions, (ii) establish an Institutional Strategic Intelligence Group that would report to senior management (General Manager and Division Managers), and be composed of experts in planning and analysis; and (iii) integrate clinical information across the CCSS organization to manage the care of chronic patients through the expansion of the EDUS platform (Unique Patient Health Record) throughout the organization.  In the medium-term (3 to 5 years): (a) Health Service Delivery: (i) articulate integrated networks which include creating network management teams, stratifying population by healthcare needs, testing of performance based mechanisms such as prospective capitations, and expanding implementation of information technologies [family health (SIFF) and clinical history (SIES) modules of the EDUS platform (Expediente Único Digital en Salud)] to all facilities; (ii) convert the three CCSS national hospitals into excellence centers that specialize in high complexity-high cost services, with each center specializing in its respective area of expertise; (iii) create sub-acute units to provide specialized geriatric and chronic disease services (chemotherapy, etc.) at lower cost and higher quality than traditional hospital forms; (b) Funding and Resource Flow: introduce capitated payments to networks to improve equity and efficiency; and (c) Management: implement Enterprise Resource Planning (starting with administrative and financial modules) IT platform. 13 Costa Rica Social Sector Expenditure and Institutional Review I.3 Social Protection and Labor Costa Rica has made important efforts to expand spending and coverage of its Social Protection and Labor (SPL) system, through a mix of both universal and selective programs. The SPL sector in Costa Rica accounts for the largest share of social spending (40 percent) —and as a 8.7 percent of GDP, overall SPL spending in Costa Rica is by far the highest in the Central American region and among the highest in the LAC region. Most of it obeys to a generous social security system (pensions, health insurance) that has large coverage per regional standards, but also to increasing efforts to reach out the poor through cash transfers and other social assistance to the poor. Despite social security has high coverage, it is overly subsidized, in particular for public sector employees; better resource allocation and improved targeting could allow further expansion of the social pension component. In Costa Rica, the vast majority of contributory pension funding comes from government transfers, by subsidizing both public sector (RPN) and private sector (RIVM) pensions, the former one to a larger extent. And while more than 50 percent of the poorest elderly have access to the pension system, which is high compared to the rest of CA countries, leakages persist in the system. Coverage of the elderly among the poorest quintiles could be improved with better targeting accuracy of social pensions. There are still some questions about the long term sustainability of the pension system. Through these parametric reforms, the RIVM system estimated that the point when contributions would no longer cover expenses had been postponed to 2041. However, recent increases in benefits to the teachers’ pension scheme and opposition to increase contributions have brought this deadline nearer. Recent IMF estimates imply that RIVM is projected to turn a cash deficit over the medium and long term due to system maturation and population aging, with an adjustment equivalent to about 1.5 percent of GDP required to ensure actuarial equilibrium of all pension systems for the next 100 years. It is critical to improve incentives and enforcement so that high-income workers would contribute its fair share, rather than evade or underreport income. It would also be necessary to draw on collective insurance programs already available to expand contributory coverage and family insurance. To improve the quality of coverage, incentives could be introduced so that more people make voluntary contributions to the system. On the other hand, and despite a wide portfolio of social assistance programs, coverage of social assistance is still low, and especially for the moderate and extreme poor. Of all 46 social assistance programs managed by 22 different institutions, only the universal school feeding program has the largest coverage of the poor population. Strikingly, the Avancemos CCT program, launched in 2006 with the objective to promote the retention and reintegration into the formal education system of children, and is poverty targeted, only covers a fifth of the extreme poor, and shows important leakages to higher quintiles. While programs like 14 Costa Rica Social Sector Expenditure and Institutional Review Avancemos are helping mitigate poverty and inequality (in a context of high unemployment), this impact could be maximize by improving coverage among the poor through enhanced targeting accuracy (without increasing budget envelope). In terms of active labor market programs (ALMPs), Costa Rica spends important resources, more even than a large number of OECD countries; however its efficiency needs to improve. Costa Rica has high unemployment (per Central American and Latin American standards), especially among the youth and those who cannot finish secondary or tertiary education. As a consequence, it spends a relatively large proportion of the budget in implementing job skills training interventions (through the national training institution, INA), as well as labor intermediation. However, a thorough assessment and evaluation of targeting and quality of courses delivered, in particular through INA, is needed. While INA reached in 2013 around 13% of the labor force, the share of out-of-school youth is less than 20% and most beneficiaries belong to the higher quintiles levels. Moreover, estimated returns to INA training are negative: beneficiaries actually earn less than those with no training, while attending other public institutions (e.g. public universities) and private institutions have positive gains. In terms of Government policies and goals, one of the pillars of the National Development Plan 2015-2018 is to reduce extreme poverty and inequality, and includes an important institutional reform. The government’s objective is to lift 56,000 families (half of total) out of extreme poverty. The strategy includes the provision of financial support to satisfy the family’s basic needs, and creating conditions for economic independence as a result of the incorporation of these families into the economy. The strategy is led and directed by the newly created Ministry of Human Development and Social Inclusion (who is also, the Executive President of IMAS, the main government executing agency of social assistance programs), but as this is a minister without portfolio yet, its operational strategy is still heavily dependent b on human and logistical resources from IMAS. This is the right approach, since the current institutional setting of the SPL sector is quite complex in Costa Rica and limit discretionary spending. The SPL sector in Costa Rica is comprised by several and uncoordinated institutions (more than 28) without clear leadership or accountability, with programs and actions developed in isolation. Services and benefits provided even within the same institution are not necessarily comprehensive and complementary; instead, they are dispersed, duplicated and fragmented. The creation of the Ministry of Human Development and Social Inclusion is thus good news in that regards. But for such an institution to work effectively, it should have some control over the programs’ plans, budgets and budget allocation. It also needs to ensure that disaggregated information on program costs and beneficiary profiles are collected and analyzed on a timely basis, in order to better monitor and evaluate the programs in search of improved efficiency and effectiveness. 15 Costa Rica Social Sector Expenditure and Institutional Review The SPL sector also needs to strengthen its monitoring and information systems, ideally through the institutionalization of a single beneficiary registry from existing program registries and census of the poor. In order to have an integrated SPL system, the country needs to strengthen its information system and consolidate the information on programs articulating the institutional offer with the social demand. A single registry of beneficiaries would improve targeting and diminish overlaps. As of today, CCSS registries (pensions, health insurance) do not communicate with SIPO (registry of social assistance programs), for instance, nor with a census of the poor population. As a consequence, there is no “graduation” strategy from programs, to derive beneficiaries from one to another. A single registry of beneficiaries with a uniform PMT formula would help prioritize interventions for the extreme poor. In addition, the government has announced the promotion of family guides whose basic function is to provide and guarantee families in extreme poverty, access to the government services 16 Costa Rica Social Sector Expenditure and Institutional Review II. Context Costa Rica registered decent economic growth above the Central America (CA) average for 2000-2013 (except during the global crisis). Costa Rica growth track record has been quite impressive within the CA context, growing at 5 percent per year between 2001 and 2008, and 4.3 percent in 2010-2013 after the global crisis (Figure 1). Only Panama registered higher economic growth in the period. In terms of GDP per capita, it has also experienced important growth of 2.7 percent per year, as well only behind Panama in the CA region. Figure 1: GDP growth in Costa Rica and Central America, 2001-2013 Source: IMF, World Economic Outlook Database, October 2014 However, poverty was less responsive to growth and indeed, increased in the last few years.; inequality increased as well, in contrast to the trend of falling inequality in the rest of LAC. Costa Rica has one of the lowest poverty rates in the Latin American and Caribbean (LAC) region, only above Chile and Uruguay. But what puzzles from the Costa Rican case is that the decent economic growth in GDP and GDP per capita has not translated into poverty or inequality reduction; in fact both have increased, especially after the global economic crisis. Poverty headcount was 23.6 percent in 2003, 21.3 percent in 2010, and 22.4 percent in 2014. Extreme poverty, at 5.1 percent in 2003, rose to 6.7 percent in 2014 (Figure 2). Indeed, the poverty elasticity to growth has declined: while between 2001 and 2007, a one percent increase in GDP per capita was associated with a 0.60 percentage reduction in the official poverty rate, after the global financial crisis, the growth-poverty elasticity decreased to 0.44. The Gini coefficient, that measures inequality, also rose from 0.47 in 2003, well below the Latin American and Caribbean (LAC) average (only above Uruguay in the region), to 0.52 in 2014, now on par with the LAC average (Figure 3). While several hypotheses have been put to test to explain recent poverty and inequality evolution, including sector growth composition and skill shortages 17 Costa Rica Social Sector Expenditure and Institutional Review that favor the returns to high-skilled workers, the limited capacity to redistribute income effectively through taxes and transfers seems to have played a role as well.1 Figure 2: Poverty headcount Figure 3: Inequality Note: *Change in survey methodology. Note: *Change in survey methodology. Source: SEDLAC Source: INEC While not manifested in poverty and inequality indicators, progress has continued to take place in social indicators. Table 1 compares Costa Rica’s evolution of selected indicators in the areas of education, health and nutrition, and social protection with three groups of comparators: i) the top 7 economies in the LAC region; ii) the remaining countries in the CA region; and iii) a set of 8 countries around the world that can be considered “comparator countries” when judged after a series of criteria (see Table note). In order to show progress, the 15-year period (2000- 2014) was split in two intervals (2000-2006 and 2007-2014). Costa Rica shows a consistent trend of improvement in social indicators across the board (enrolment in pre-primary, secondary, tertiary, sanitation facilities, labor force participation), and in many of them it has reached optimal levels (primary enrolment, pupil-teacher ratio, nourishment, institutionalized births, improved water sources) and are at higher level than closer comparator countries. But in some of them it deserves attention a closer look that will be discussed in more detail in the report: secondary school completion is still low as it would be expected for Costa Rica’s level of income and public spending. Progress in some health indicators has stalled (prenatal care, immunizations, and hospital beds). And employment and unemployment challenges remain and are more acute than in other countries in the region. 1 World Bank (2015), “Costa Rica: Systematic Country Diagnosis”. 18 Costa Rica Social Sector Expenditure and Institutional Review Table 1: Selected Human Development Indicators, Costa Rica, LAC, Central America, and Closest Income/Population Comparators, 2000-2014 Costa Rica LAC 7* Rest of CA Closest Comparators** Indicator Name 2000-2006 2007-2014 2000-2006 2007-2014 2000-2006 2007-2014 2000-2006 2007-2014 Education School enrollment, preprimary (%gross) 62.1 70.1 65.8 85.7 45.8 57.1 59.6 66.2 School enrollment, primary (%gross) 111.5 109.2 111.7 109.7 109.1 111.9 105.9 102.8 School enrollment, secondary (%gross) 73.0 96.7 78.8 87.9 58.3 66.8 88.4 88.2 School enrollment, tertiary (%gross) 25.6 45.6 37.2 48.4 25.7 28.9 43.6 54.3 Primary completion rate, total (%) 91.5 95.2 98.4 102.1 78.4 90.7 97.5 95.0 Pupil-teacher ratio, primary 22.5 18.4 24.5 23.2 32.4 28.7 17.7 16.2 Secondary completion, age 25+ 31.2 35.0 36.0 41.8 22.6 26.3 52.9 66.0 Health Pregnant women with prenatal care (%) 91.7 89.9 93.7 96.0 86.0 94.0 97.4 97.4 Undernourishment (% of pop) 5.1 5.9 11.9 9.8 20.9 17.4 13.4 8.3 Immunization, measles (% 12-23m) 87.7 86.7 95.2 94.5 94.7 94.7 88.0 91.1 Improved sanitation facilities (% of pop) 92.0 93.4 79.4 83.9 64.7 69.8 90.3 90.4 Improved water source (% of pop) 95.4 96.3 90.2 92.6 86.2 89.7 96.8 97.5 Hospital beds (per 1,000 people) 1.4 1.2 1.8 2.0 1.2 1.0 5.8 5.0 Births attended by skilled health staff (% of total) 98.3 99.2 93.2 94.9 76.6 80.6 98.3 97.7 Social Protection and Labor Employment to population, 15+ (%) 57.6 58.6 58.6 61.3 58.3 60.4 46.2 48.3 Labor force participation, female (%) 41.6 45.7 49.3 52.8 43.3 46.3 42.2 43.4 Unemployment, total (%) 6.2 6.7 8.7 7.0 6.4 5.3 14.6 12.0 GINI index 49.3 49.2 53.6 50.0 53.4 50.1 38.8 37.0 Poverty headcount ratio, rural (%) 26.2 60.1 52.6 65.1 57.9 32.8 33.6 Poverty headcount ratio, urban(%) 18.1 37.4 23.0 42.5 37.4 13.0 10.9 *Argentina, Brazil, Chile, Colombia, Ecuador, Mexico Peru. ** In terms of GDP, GDP per capita, population, population density and percentage of rural population: Bulgaria, Lebanon, Lithuania, Montenegro, Panama, Serbia. Source: World Bank Development Indicators (2014) III. Recent Trends in Social Spending in Costa Rica Costa Rica’s tradition for a strong welfare state has deep historical roots. The construction of Costa Rica’s welfare state goes back to its early days. It was consolidated during the 20th century with the creation of the universal health and social security system Caja Costarricense de Seguro Social (CCSS) in 1941, and the establishment of a rights-based legal system (for instance, the Labor Code in 1942, and the Constitution of 1949). 2The creation of the main welfare institutions in the mid-20th century drove a significant increase in social spending. With the abolition of the army, the country had more resources to invest in the provision of basic 2 World Bank (2015), ibid. 19 Costa Rica Social Sector Expenditure and Institutional Review public and social services to the entire population. As a result, public spending in education, health and social protection started to rise in the 1950s, reaching over 3 percent of GDP by 1958 (Figure 4). It then increased steadily over the decades, reaching almost 10 percent by the end of the 1990s and over 20 percent by the first decade of the 2000s. Investment in education was initially the highest category of social spending, but by the 1990s health and pensions had both surpassed education. Figure 4: Costa Rica Social Spending 1950s, 1990s, and 2007-13 Source: World Bank, Costa Rica SCD. In the last few years social spending as kept it increasing trend, reaching 20.9 percent of GDP in 2013. The additional resources channeled to social sectors grew in parallel among all components, with social security/pensions still explaining the largest shared, but followed closely by health (Figure 5). However, education and social assistance and labor increased the most in comparative terms, by 50 percent in its share of GDP between 2007 and 2013. Most of the increase in GDP shares took place between 2008 and 2009, after that year shares remain more or less constant. 20 Costa Rica Social Sector Expenditure and Institutional Review Figure 5: Social Spending as a % of GDP by sector (%) 2007-2013 Source: World Bank SSEIR / ICEFI social spending database Costa Rica is by far the country with largest share of social spending with respect to GDP in CA. Costa Rica’s spending of 20 percent of GDP in social sectors stands among the largest in Latin America, and by a wide margin in CA (average of 13 percent) (Figure 6 and Figure 7). Most of the difference with other countries in the region obeys to spending in social security (7.5 percent of GDP, vs 5.4 percent in Honduras, which follows) and in health (6.7 percent of GDP, vs. 4.7 in Nicaragua, as follower). In education, it is second in the region (5.5 percent of GDP, vs. 5.8 percent in Honduras), and only in social assistance it is placed at average levels (1.1 percent of GDP, below El Salvador, Panama, and Honduras in the region). Figure 6: Per capita social public expenditure by sector (2012 or latest year available) Source: ECLAC – CEPALSTAT 21 Costa Rica Social Sector Expenditure and Institutional Review Figure 7: Social Spending as a % of GDP by country 2013 (%) Source: World Bank SSEIR / ICEFI social spending database Social public spending is not overall progressive mainly due to health and old age benefits. Figure 8 shows the distribution of social spending by sectors and quintiles. Public spending on education is progressive, highest for the poorest quintiles and lowest for the highest quintile. Public spending on health is highest for the middle income quintile, followed by the two higher income quintiles. Social security spending benefit more the highest quintile than the poor. Other social assistance and labor interventions (cash transfers, sickness and disability, etc.) are progressive since most of the spending is allocated to the first two quintiles. Total social spending is not progressive due mainly to the high amount allocated in social security. Figure 8: Distribution of social spending by quintiles, 2013 Source: World Bank SSEIR / ICEFI social spending database Note: Distribution of spending was calculated based on the distribution of beneficiaries per sector. For education, the distribution of public students enrolled in each level of education by income quintile was taken into account. For health, the distribution of the utilization of public health providers by income quintiles was considered (based on ENV 2008). For old age, we considered the distribution of pension beneficiaries by income quintiles and, for social assistance, the distribution of social assistance beneficiaries. 22 Costa Rica Social Sector Expenditure and Institutional Review Looking forward, fiscal deterioration poses severe challenges to continue financing social sector expenditures at current levels. The adoption of countercyclical policies, which were adopted to cushion the impact of the crisis on domestic economic activity, led to growing fiscal deficits.3 In particular, the wage bill and direct transfers rose steeply in 2009, and stayed at the higher level thereafter. The wage bill increase from 5.5 percent of GDP in 2008 to 7.4 percent in 2014 as a result of a policy that aimed to increase pay in the Central Government to the 50th percentile of pay for similar jobs in the rest of the public sector. On the other hand, government revenues, which had risen steeply prior to the crisis, dropped back to the 2000-05 average level. Thus, the overall fiscal balance switched from an average surplus of 0.3 percent of GDP in 2007 to a deficit of 6.3 percent of GDP in 2014 and projected to reach 6.5 percent of GDP in 2015 (Figure 9). The deterioration in fiscal accounts will drastically limit the possibilities for further increases in social sector spending, and more likely demand cuts to achieve fiscal sustainability in the coming years. Figure 9: Central government (excluding decentralized entities) overall balance, 2007-2015 Source: IMF, World Economic Outlook Database, October 2014 The budget formulation process is also quite complex and poses severe rigidities that limit resource allocation according with developmental priorities. Costa Rica has three distinct budgetary processes in the public sector.4 First, the Central government budget or National budget, which is approved by the Legislative Assembly. Second, the budgetary process of the institutions outside the Central government, whose budget is approved by the Comptroller General of the Republic (CGR), and which doubles the size of the National budget. Of the total public sector budget, the National budget is about 33 percent, and the budget of other institutions is 66 percent. The third budget process is for municipality budgets, which account for less than 2 3 World Bank (2015), ibid 4 World Bank (2015), ibid 23 Costa Rica Social Sector Expenditure and Institutional Review percent of public sector expenditures. The CGR verifies the legality of proposed expenditures by the municipalities. Budgets are also increasingly constrained by constitutional mandates and rules, such as directives on minimum spending for education, municipalities, housing subsidies, and community development. Additional laws stipulate that portions of taxes and fees must be allocated to certain activities and institutions, such as FODESAF and CONAVI. More importantly, the executive has no power to direct or contain spending in autonomous institutions, since they operate with wide margins of budgetary and administrative independence, and are constitutionally protected from political interference or changes in government. Thus, a large part of public expenditures is not under the direct control of the Executive or the Legislative powers, which limits discretionary spending according to Government priorities. Budget execution in Costa Rica is also a challenge and would need to be improved, though this has not happened in the last seven years. One of the salient features of Costa Rica is that its public institutions, including in social sectors where current expenditures predominate, consistently underexecute, on average by 10 percent of the budget (Figure 10). This ratio is typically lower, of 15 percent, in education, and to a lesser extent in health. Among the main reasons for low budget execution is related to the nature of the budget process formulation and the numerous earmarked expenditures that reduce the margin for the executive to control public investment and current expenses. Since central government tax revenue is still insufficient to fund all these legal and constitutional mandates, it is not surprising that the government has often resorted to cutting back on public investment, putting pressure on service delivery. 5 At the same time, limited capacity of project implementation units in the various ministries, and cumbersome processes in the Public Procurement Law (Ley de Contratacion Administrativa), also affect budget execution. Despite the current fiscal situation, increased implementation capacity would increase investment but not necessarily put more pressure on public finances. 5 World Bank (2015), ibid 24 Costa Rica Social Sector Expenditure and Institutional Review Figure 10. Budget execution 2007-2013 Source: World Bank SSEIR / ICEFI social spending database Overall, Costa Rica’s social spending is effective but less efficient compared to other LAC countries. An analysis of the relationship between social outcomes and spending through the concepts of Public Sector Performance (PSP) and Public Sector Efficiency (PSE) presents several interesting results. PSP is assessed by constructing composite indicators based on observable socioeconomic variables that are assumed to be the output of pursued public policies. PSE relates PSP scores to their cost in terms of public spending. The overall assumption behind the assessment of public sector performance and efficiency employing PSP and PSE indicators is that the observed outcome indicators are solely the result of public spending policies (Box 1 provides additional information on the PSP and PSE analyses). Box 1: Public Sector Performance and Public Sector Efficiency Indicators We analyzed the relationship between social outcomes and spending using the Public Sector Performance (PSP) and Public Sector Efficiency (PSE) approaches developed by Afonso, Schuknecht, and Tanzi (2005, 2010).6 PSP is measured by constructing composite indicators based on observable social variables that are assumed to be the output of pursued social public policies. Specifically, the PSP for country = 1, … , with = 1,2,3 social sectors (education, health and social protection and labor) is determined by: = ∑ ; = 1, … ; ℎ = ( ), = 1, … , . (1) =1 6 The methodology follows Afonso, Schuknecht, and Tanzi (2005, 2010) for OECD countries, replicated later on in Afonso, Romero, and Monsalve (2013) for LAC. 25 Costa Rica Social Sector Expenditure and Institutional Review where ( ) is a function of k observable social indicators (for education, we take gross secondary enrollment and literacy rate; for health, we take maternal mortality and immunization rates; and for social protection and labor, inequality (measured by the Gini coefficient) and extreme poverty headcount (percentage of population earning less than $1.25 a day). To obtain PSP indicators we assign equal weights to each sub-indicator, computed as the average of the corresponding outcome indicators, each one of them normalized by its sample mean. The PSP indicator for each country is then obtained by averaging the values of all sub-indicators. Resulting PSP scores are then related to the average value of one of the normalized output indicators. Hence, countries with PSP scores in excess of one are seen as good performers, as opposed to countries with PSP values below the mean PSP Secondary School Gini Index Enrollment Social Education Protection Literacy rate and Labor Poverty headcount (percentage Maternal of population Mortality earning less Health than $1.25 a Immunizati day on rates PSE relates PSP scores to their cost in terms of public spending. PSE weights public sector performance in each social sector by the amount of relevant public expenditure that is used to achieve such performance. To compute PSE scores, public spending in each sector is normalized across countries, taking the average value of one for each of the expenditure categories ( j). This is, for each country = 1, … , with = 1,2,3 social sectors, the PSE is defined by: = ∑ ; (2) =1 The analysis indicates that social spending in Costa Rica is effective (in improving outcomes) but not necessarily efficient (“value” per dollar spent), overall but in particular in the education and social protection sectors, while it is neither effective nor efficient in the health sector (Figure 11). 26 Costa Rica Social Sector Expenditure and Institutional Review Figure 11: Public Sector Performance and Efficiency in Costa Rica and LAC, 2010 Overall Education Health Social Protection and Labor Source: World Bank SSEIR team’s, authors’ calculations using CEPAL and WDI databases A LAC “production possibility frontier” analysis shows that Costa Rica could increase its social performance by as much as 3 percent with the same level of public social spending. Figure 12 shows the production possibility frontier for total social public spending for LAC, applying the data envelope analysis (DEA) using the PSP scores as an output and social-public spending-to-GDP ratios as an input. Based on the DEA analysis (explained in Box 2), Costa Rica could move toward the LAC “production possibility frontier” and increase its social performance by 10 percent, with the current level of public social spending. 27 Costa Rica Social Sector Expenditure and Institutional Review Box 2: DEA Methodology The DEA methodology, developed by Farrell’s (1957), assumes the existence of a convex production frontier to construct an envelope around the set of observations. DEA compares each unit with all other units, and identifies those units that are operating inefficiently compared with other units' actual operating results. DEA presents two approaches: 1) input-oriented shows by how much input quantity can be proportionally reduced without changing the output quantities; 2) output-oriented assess how much output quantities can be proportionally increased without changing the input quantities used. Efficiency for each unit can be measured by computing the distance to the theoretical efficiency frontier (or compared to the best practice units). DEA provides an efficiency rating that is generally denominated between zero and 1, which will interchangeably be referred to as an efficiency percentage between the range of zero and 100%. The best practice units are relatively efficient and are identified by a DEA efficiency rating of = 1. The inefficient units are identified by an efficiency rating of less than 1 ( < 1). The Figure illustrates the single input single output DEA production possibility frontier. Countries A, B and C are efficient with output scores equal to 1. On the other hand, country D is not efficient, since its score [d2/(d1+d2)] is below unity. 28 Costa Rica Social Sector Expenditure and Institutional Review Figure 12: Production Possibility Frontier (Data Envelope Analysis) for Total Social Public Spending, Costa Rica and LAC, 2010. Source: World Bank SSEIR team’s, authors’ calculations using CEPAL and WDI databases The next sections describe the main challenges in sustaining social sector expenditures while improving its effectiveness and efficiency in improving sectorial outcomes in Costa Rica. The analysis presented so far will be expanded and complemented by more detailed expenditure and institutional analysis in each sector that reviews progress achieved, as well as main challenges that need to be addressed to improve social sector spending efficiency and effectiveness. IV. Performance and Challenges in Education IV.1 Recent Evolution of Education Public Spending Public spending on education has increased in Costa Rica, and it is line with the country income. Public spending on education has been increasing steadily over the period between 2007 and 2013. It represented 3.9 percent of its Gross Domestic Product (GDP) in 2007, and 5.5 percent in 2013. The public expenditure on education is perfectly in line OECD standards, and among countries in the LAC region only Honduras and Brazil spend a higher share of their GDP on education (Figure 13). However, the country shows significant imbalances that make it different from most of its neighboring countries. Basic and tertiary education take up the bulk of 29 Costa Rica Social Sector Expenditure and Institutional Review public spending on education (with 33% and 26% of the total spending respectively), while the spending on secondary attracts 22% of the resources (Figure 14)7, a share that is disproportionately low for the country’s GDP (Figure 15). The imbalance is not explained by differential changes in the demographic composition across the educational levels. While the number of students in primary has been steadily declining, it has been constantly increasing in secondary, going from about 437,000 in 2007 to 491,000 in 2011. Figure 13: Public spending on education versus GDP per capita (%) (circa 2011) Source: World Bank SSEIR / ICEFI social spending database for Central America, EdStats for the rest of the countries. Figure 14: Comparison of public spending in education by levels (%) Source: World Bank SSEIR / ICEFI social spending database 7 Due to ICEFI data limitations, spending on pre-school and primary education could not be disaggregated in Costa Rica. 30 Costa Rica Social Sector Expenditure and Institutional Review Figure 15: Secondary public expenditure as percentage of country’s GDP (%) Source: World Bank SSEIR / ICEFI social spending database for Central America, EdStats for the rest of the countries . The wage bill represents a large share of the total public education expenditure, and salaries of the administrative staff absorb relatively more than in other countries. In line with other countries of the LAC region, Costa Rica displays a large wage bill that accounts for almost 87% of the total public education expenditure. However, unlike most countries inside and outside the region, there is a very large amount of resources devoted to the salaries of the administrative staff: 24.8% of the total public education expenditure (Figure 16). In Mexico, that is comparable to Costa Rica in terms of both GDP per capita and share of public spending in education, salaries of the administrative staff only account for 10% of the total public education spending. While the average salary of a primary education teacher in Costa Rica is comparable to the average entry level salary of a primary education teacher in the OECD countries (approximately 30,000 US$ per year), the average salary of a secondary education teacher is close to the top range of the salaries of secondary education teachers in the OECD (about 50,000 US$ per year). A career as a teacher offers a record high return in the Centro America region (Figure 17), with an average monthly premium of 9% compared to those employed in similar occupations. This premium becomes even larger on an hourly basis (25%), since teachers on average work less hours than postsecondary graduates in other occupations. 31 Costa Rica Social Sector Expenditure and Institutional Review Figure 16: Comparison of the share of wage bills over total public educational expenditure (%) Source: Edstats Figure 17: Teacher's wage premium in CA Student-teacher ratios have been declining over the last years and are low compared to the country GDP. Both in primary and secondary education student-teacher ratios have significantly dropped in the period between 2007 and 2011. While in primary the ratio declined from 19.5 to 17.3, in secondary it went down from 17.9 to 14.9 (Figure 18). The latest poses Costa Rica below the average student-teacher ratio in secondary education among countries with similar GDP per capita (Figure 19). The decline in the student-teacher ratio came from an increase in the number of teachers hired, rather than a reduction in the number of students. While there is a large supply of graduates who have been trained as teachers, it is unclear whether they meet the quality requirements to be good teachers. The evidence presented so far suggests that, while an increase in secondary level expenditure seems necessary, this should not result either in an increase of teacher salaries or the hiring of new teachers. 32 Costa Rica Social Sector Expenditure and Institutional Review Figure 18: Trend in Student-teacher ratios Figure 19: Student-teacher ratios in by level Secondary vs GDP per capita Source: EdStats Source: EdStats Overall public spending in progressive but expenditure in tertiary education is highly regressive. Overall public spending in education is progressive, with students in the first quintile of the income distribution receiving 258 millions of dollars and the fifth quintile receiving 157 millions (Figure 20). The progressive nature of the education public expenditure is driven mostly by the high progressivity of spending in primary education. Unlike for primary and secondary, public universities are perceived to have better quality than private ones. 63% of the students attending public universities belong to the fourth and fifth quintile of the household income distribution. Given the high share of public expenditure in tertiary education, spending in tertiary education is highly regressive. Households in the fifth quintile of the income distribution receive in total almost 40% of the public spending in tertiary education. Figure 20: Distributional impacts of public spending by education level, 2013 Source: Source: World Bank SSEIR / ICEFI social spending database 33 Costa Rica Social Sector Expenditure and Institutional Review Moving forward, Costa Rica can also leverage aging population trends and transfer resources towards improving quality and reducing inequality. UN demographic projections (United Nations, 2013) estimate that Costa Rica will continue the transition towards a stationary population pyramid in line with higher levels of development (Figure 21). For youth populations, this trend implies a decrease in the number of potential students and a likely gradual reduction in student enrollment (Figure 22). This population aging process and reduction of projected enrollment is, in fact, an observable trend throughout most countries in the region. Bruns and Luque (2015) project that between 2010 and 2025. Costa Rica will experience an 8.7 percent decline in the student population aged 4-18 and an analogous 8.1 percent reduction in the number of teachers needed to maintain the current student-teacher ratio (assuming constant enrollment ratios).8 This demographic trends allow the country to harness the decrease in student enrollment to improve quality by simply maintaining or reducing the number of teachers and transferring these resources towards quality-improving and inequality-reducing strategies articulated in this report. Figure 21: Demographic trends in Costa Rica 2010 compared with 2050 Total Population by Age group and Sex Source: ICPD, Country Implementation Profiles: Costa Rica (2012). Population Pyramids are based on medium variant of the 2010 revision of the World Population Projections (WPP) by UN Population Division. 8 Bruns, Barbara and Luque, Javier, Great Teachers: How to raise student learning in Latin America and the Caribbean (Washington: 2015) The World Bank Group. 34 Costa Rica Social Sector Expenditure and Institutional Review Figure 22: Demographic Projections in Costa Rica: Projections on Student Population Demographic Projections in Panama: Total Population Aged 5-14 and Share of Total Population 900 18 800 17 700 16 600 Percentage Thousands 15 500 14 400 13 300 200 12 100 11 0 10 2010 2015 2020 2025 2030 2035 2040 2045 2050 Axis Title Estimated Population Aged 5-14 Percentage of Total Population Source: Data from United Nations (2013) World Population Prospects: The 2012 Revision, Volume II, Demographic Profiles (ST/ESA/SER.A/345) IV.2 Education Outcomes and Challenges Enrollment rates are on average high, but there are large within-country disparities at all levels. Enrollment rates are on average high for both primary (108.5) and secondary education (117.7). Enrollment ratio for postsecondary education rose from 28.9 to 39.7 (Figure 23) and is among the highest in the LAC region. 35 Costa Rica Social Sector Expenditure and Institutional Review Figure 23: Gross enrollment rates by educational level Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations based on ADePT software However, urban and rural areas of the country present large disparities in enrollment in secondary and upper-secondary. While almost 60% of the 19 year old living in urban areas living are enrolled in school, less than 50% of those living in rural areas do so (Figure 24). Equally worrying, the disparities in post-secondary enrollment associated with household income have been dramatically increasing over time (Figure 25). In 2014, gross attendance rate in post- secondary for those belonging to the highest quintile of the income distribution was 65.1%, as opposed to a gross attendance rate of 27.3% and 23.3% for those belonging to the second and first income quintile respectively. In 2007, the gross attendance rate for those in the top and the bottom quintile of the income distribution was 52.4% and 14.8% respectively. Figure 24: Age specific enrollment by and Figure 25: Post-secondary enrollment by 36 Costa Rica Social Sector Expenditure and Institutional Review geographic area income quintile Source: World Bank SSEIR team’s analysis of household Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations based on ADePT software surveys, authors’ calculations based on ADePT software Learning outcomes are not in line with country GDP per capita and there are large socioeconomic differences. Costa Rica took part in the 2009 and 2012 PISA assessment. Results from 2012 show that, while the country did on average better than other countries of the LAC region such as Brazil and Colombia, it lagged behind some its “comparator countries”. For instance Serbia and Bulgaria, countries with a GDP per capita similar to Costa Rica, display an average score of 449 and 439 respectively, as opposed to 407 in Costa Rica (Figure 26). Moreover, the latest results from the TERCE examination of Latin American students by UNESCO show that Costa Rica is the only country in the region where students in both 3rd and 6th grades performed worse in TERCE (2013) than they did in SERCE (2006) in reading and in mathematics. Consistent with the findings for enrollment, there are large differences in learning outcomes associated with socioeconomic status. Data from the Second Regional Comparative and Explanatory Study (SERCE) show that one standard deviation (sd) increase in a socioeconomic conditions index are associated with a 0.3sd increase in test scores (Figure 27). Figure 26: PISA score versus GDP per Figure 27: SERCE standardized score by 37 Costa Rica Social Sector Expenditure and Institutional Review capita socioeconomic status index Source: SERCE (2006) Source: EdStats High spending in education did not result into relatively higher level of the human capital of the labor force, and the supply of skills produced by tertiary education does not match with the labor market needs. Only 35% of the workforce has completed secondary education (Figure 28), a percentage that puts Costa Rica not only behind the OECD average (64.7%) but also behind the LAC average (35.4%). Although 26% of the public education spending is directed towards tertiary education, the higher education system does not sufficiently contribute to generate the most needed skills. The distribution of graduates in public and private universities is heavily biased towards social sciences and education9, whereas basic science and engineering are very distant in terms of number of graduates. Students still prefer to study careers that had growing demand during the 1980s and 1990s, rather than those that have high demand (and growing wages) today (Figure 29). This obeys in part to the fact that for private universities (which produce more than half of the graduates), social sciences demand lower infrastructure investments, and thus are more profitable. Moreover, those who did not complete secondary education and attend the National Learning Institute (Instituto Nacional de Aprendizaje, INA) do not experienced any gains from training (Jimenez, 2014). These results suggest that the education spending is not successfully providing future generations with the skills that high value added sectors search for, such as foreign languages, IT knowledge, analytical thinking, interpersonal skills (Costa Rica, SCD). 9 Evidence presented in Bruns (2014) shows that only 55% of the teacher graduates are working as teachers. 38 Costa Rica Social Sector Expenditure and Institutional Review Figure 28: Secondary education attainment of the workforce Source: EdStats Figure 29: Distribution of graduates in public and private universities Gaps in the access to education services start early in life. There is well established evidence that high quality Early Child Development programs act as equalizers, since they can reduce the effect of household socioeconomic differences on the child cognitive and non-cognitive development and, therefore, the ability to perform well in school. Evidence from the US shows that the quality of the kindergarten services has a long lasting effect on school and labor market outcomes (Chetty et al. 2014). In Costa Rica, there is a large differential in daycare (age 1-3) and preschool (age 3-5) attendance by household income quintile. Only 3% of the children in the first quintile attend daycare, as opposed to 16% among those in the fifth quintile (Figure 30). Similarly, for children in preschool age, 27% of those in the first income quintile attend 39 Costa Rica Social Sector Expenditure and Institutional Review preschool, while attendance goes up to 49% for those in the top income quintile (Figure 31). The data do not allow to assess the quality of services attended, but it is plausible to speculate that children belonging to higher income households can attend better premises, while those belonging to the poorer strata, especially for the children age 3-5, are completely precluded an important source of socialization. Figure 30: Daycare attendance by Figure 31: Preschool attendance by household income quintile, 2014 household income quintile, 2014 Source: World Bank SSEIR team’s analysis of household Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT surveys, authors’ calculations using standardized ADePT software (Education Module) software (Education Module) Parental education seems to play an increasingly determinant role in children’s education investment. Children form households with a low educational environment that is, children whose parents have low education) have a much faster dropout rate than children from more educated households (Costa Rica SCD). Moreover, a recent study of the factors behind human capital accumulation finds that, although in the 1990s context variables such as place of residence explained a significant share of the educational attainment inequality, at present only household and individual level variables, such as parental education, explain such outcomes (Trejos and Murillo, 2012). Both in primary and secondary education students from richer households have strong preference for private schools that are better equipped and display better results . 40 Costa Rica Social Sector Expenditure and Institutional Review According to the 2012 ENAHO, among those that attend public schools 33% of the students belong to the lowest quintile of the household income distribution, and only 6% to the top quintile of the distribution. A similar picture emerges in secondary education, where 25% of the students belong to the lowest quintile and 10% to the highest. Private schools can rely on more resources: for instance on average for each student in the public sector there is 0.19 computers available, as opposed to 0.40 in the private sector. As shown in (Figure 32), private schools display on average a higher pass rate (85%) for the final high school exam than public ones (55%). Figure 32: High school exam pass rate by school type Source: ENAHO(2013) It is not possible to establish the extent to which this dramatic difference in outcomes is the result of differences in the students’ ability or differences in school inputs. The income polarization across public and private schools exacerbates pre-existing socio-economic differences in attainments and learning outcomes. Different explanations might be behind this socioeconomic gradient. For instance, primary enrollment rates and the share of children who are over-age (i.e., who are a year or older than the correct age for their grade) is higher for children in poorer quintiles. Taken together, these statistics suggest that even though enrollment is high among poor children, repetition rates are also high, and therefore the quality of their learning might be lower, and their incentives to drop out after completing primary may be higher. On the other hand, among children age 15-19, when asked about the main reasons for not studying, 24.5% mentioned reasons related to liquidity constraints, while 27.2% mentioned the lack of interest, an answer that is potentially related to insufficient information about the returns to education. We note that Costa Rita has seen an increased returns to education particularly for secondary and tertiary education due to the skilled-biased development the country has experienced. 41 Costa Rica Social Sector Expenditure and Institutional Review IV.3 Institutional Arrangements The diagnostic, targeting and design of subsidies in secondary education can be improved. There are large socioeconomic differences both in the transition probability from lower secondary to upper secondary, and from secondary to tertiary. Liquidity constraints can partly explain this difference. Evidence from the 2014 ENAHO shows that subsidies directed to students are poorly targeted (Figure 33). For instance only 29% of the 15-19 year old belonging to the first quintile of the household income distribution receives Avancemos, a conditional cash transfer (CCT) aimed at reducing dropout in secondary among the poorest. On the contrary, among those in the fourth quintile of the income distribution, 8% report receiving this monetary transfer. A recent evaluation (Hidalgo and Romero, 2013) shows that this CCT has a positive impact on dropouts and enrollments.10 In other words, the CCT helps students stay in school helps those dropping out getting back to school. Although some students remain in education by other factors, between 10% and 16% of the students do so solely because of the CCT and would, otherwise abandon their studies. Likewise, re-entering the educational system do so for various reasons, but even higher percentage (77% or more) did so because of the transfer were given. Although the sample in the case of reintegration is much smaller than in the case of desertion, it cannot ignore the positive impact of the transfer is much higher in reintegration. Therefore, a better targeting of subsidies and transfers can significantly weaken the liquidity constraints that affect poorest households. Figure 33: Takeup of transfer and subsidies by household income quintile, 2014 Source: World Bank SSEIR team’s, authors’ calculations using Household Surveys: ENAHO 2014 Improvements in teacher quality can help to reduce the socioeconomic differences in outcomes. Improvements in teacher quality can help to raise the average quality of learning, and 10 The study explores propensity score matching and a differences in differences approach. 42 Costa Rica Social Sector Expenditure and Institutional Review at the same time reduce the differential associated with family background. In Costa Rica, the large supply of university degrees to prepare teachers led to an excess supply of teachers and a huge variation in teacher quality. There are very few courses accredited to prepare kindergarten and preschool teachers, where instead the country would need significant investments. Public preschool programs with high quality teachers have the potential to reduce the gaps that children from poor strata display at school entry. Establishing quality standards, possibly through an entry exam, might raise the quality of teachers at all levels. However, quality cannot be preserved in the absence of effective in-service training, an area where there are large gaps. Finally, there is need for an effective evaluation system. Evidence from the US (Dee and Wickoff, 2013) shows that an effective evaluation system can lead to significant improvements in relatively short period of time. Three elements that are currently missing in the Costa Rican system are essential for a good evaluation system: a) well defined teaching standards, b) multiple measures of teacher performance, possibly combining subjective criteria with objective ones; c) monetary incentives related to the teacher performance. While teacher salaries in Costa Rica are relatively high, they are mostly unrelated to the performance. The monitoring and evaluation (M&E) system can be improved by introducing standardized test scores for students and evaluation systems for teachers. The introduction of an adequate M&E system would benefit both the school level planning and the decision making at central level. At school level, a comprehensive M&E system would allow: 1) data driven school planning, 2) teachers to track their students’ gaps and progresses, 3) principals to set targets for their teachers and to monitor them, 4) parents to be better informed on their children’s learning needs. At system level, better and more timely data would allow to redirect resources to geographic areas and education levels that show larger gaps. Autonomy and Accountability can be increased both in Basic and Higher Education. Patrinos et al. (2013) propose a framework to understand the factors that can increase education quality. Improvements in learning outcomes require a systematic approach where both elements of structural quality and institutional factors are considered. Among the latter, quality of the assessments, stakeholders’ autonomy and an accountability system in place have been proven to have a positive effect on learning. The Basic Education system is a characterized by a high degree of school autonomy (Figure 34), more needs to be done to increase the quality of assessment and accountability. As previously discussed, evaluation systems for students and teachers represent a necessary condition. In order to increase accountability, providing parents’ associations with more decision power seems an important first step. The framework used by Patrinos et al. (2013) can be useful to understand some of the issues in the Higher Education system as well. In Costa Rica there are 5 public universities, 52 private ones, and 5 international. Both public and private universities can rely on large autonomy that was granted as part of the constitution. However, especially among private universities there is very little accountability. While institutions that receive public funding have to submit reports to the Contraloría Nacional de la República once a year, private institutions have no similar obligation (Francis, 2014). 67.7% of the programs that have been created between 1990 and 2010 are private, and rather 43 Costa Rica Social Sector Expenditure and Institutional Review than addressing the needs of the productive sector of the country, they have been mostly focused in the areas of Education and Social Sciences: 45.6% of the 546 private programs are focused in these two areas. There is also limited quality assessment. Only 68 of the 1,165 university programs have been evaluated by the Sistema Nacional de Acreditación de la Educación Superior (Sinaes). Figure 34: Organizational levels of the basic education system Source: Francis, (2014) V. Performance and Challenges in Health V.1 Recent Evolution of Health Public Spending Public spending on health in Costa Rica has grown significantly in the past seven years not only in absolute terms but as percentage of GDP. Public spending on health increased from 5.6 per cent of GDP in 2007 to 6.7 per cent of GDP in 2013 (Figure 35). Costa Rica’s public spending on health is greater than all the other countries in Central America (Figure 35 and Figure 36). 44 Costa Rica Social Sector Expenditure and Institutional Review Figure 35: Public spending on health as a Figure 36: Per-capita public spending on % of GDP health (real, US$ 2007) Source: World Bank SSEIR / ICEFI social spending Source: World Bank SSEIR / ICEFI social spending database database Despite a relatively small GDP per capita, Costa Rica’s public spending on health is comparable to the OECD average. Costa Rica’s health spending as a share of GDP is greater than other countries in Central America and the Latin America and Caribbean average (Figure 37). It is comparable to the OECD average, which includes high income countries such as Sweden and the United Kingdom. This demonstrates the country’s strong commitment to public services. Figure 37: Public health expenditure as a % of GDP vs GDP per capita Source: World Bank SSEIR / ICEFI social spending database for Central America except Nicaragua. WDI for the rest of the countries. 2010 45 Costa Rica Social Sector Expenditure and Institutional Review Private spending on health has increased over the last two decades. Private spending share has increased from 23.5 per cent in 1995 to 31.9 per cent in 2010 (Figure 38). Although both private and public health spending have been consistently increasing over the last fifteen to twenty years, the burden of spending has not been distributed evenly. In particular, household health expenditures are increasing for the poorest two quintiles while they have been decreasing for the rest. Figure 38: Increase in private and public health spending shares of total national health spending since 1995 Source: National Health Accounts (NHA)-WHO 2012 Households’ out-of-pocket health expenditures also suggest rising inequality of financing. The increase in private spending also raises equity concerns, as private expenditures have increased for the poorest two quintiles of population, while decreasing for the rest (Figure 39). In addition, this spending is mainly on services that are also provided by CCSS, hence raising concerns about quality and effective coverage. Figure 39: Out-of-Pocket health expenditure by income quintile 2004 and 2013 46 Costa Rica Social Sector Expenditure and Institutional Review Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Health Module). Chronic disease control is the key challenge to Costa Rica’s healthcare system. Costa Rica has sustained excellent health outcomes over the last decades. As a result of this success, the population is now aging rapidly and chronic diseases are on the rise. Costa Rica is well positioned to continue offering strong health care within the country, yet has room for strengthening health insurance offerings. The resulting challenges, summarized in the next section, will require a realignment of Costa Rica’s healthcare system. V.2 Health Outcomes and Challenges Costa Rica’s health outcomes are comparable to those in high income countries. In spite of its middle-income status, Costa Rica has life expectancy and mortality rates comparable or even superior to several developed countries’ (Table 2). This high public health spending coupled with an already robust health care system has had a positive impact on health outcomes in Costa Rica. This has been achieved through universal and equitable health coverage and the focus on primary health care interventions. Table 2: Health outcomes in Costa Rica: 1995-2011/12 Indicator Description or categories 1995 2000 2005 2009 2011/2012 3,479 3,931 4,328 4,579 4,805 In thousands of people Total population 2.47 2.29 1.68 1.32 1.41 Annual growth rate Live births per 1,000 women aged 2.76 2.43 2.1 1.94 1.81 Fertility rate 15 to 49 years old (in a given year) 79.2 80.2 80.9 81.5 82.0 Women Life expectancy at birth 74.5 75.5 76.2 76.7 77.5 Men Number of deaths of infants under 14.4 12.9 11.6 10.6 8.6 Infant mortality rate one year per 1,000 live births in a given year Maternal mortality Number of deaths in childbirth per n.d. n.d. 36.3 26.7 rate 100,000 live births 23.1 47 Costa Rica Social Sector Expenditure and Institutional Review Deliveries attended by As a percentage of all births in a skilled health n.d. 98.2a 98.5b 99.1c 99.1 year professionals Source: World Bank and WHO. Notes – a. 1990 data; b. 2004 data; c. 2008 data Costa Rica has low under-5 and maternal mortality rates. Costa Rica’s life expectancy is 81.5 years for women and 76.7 years for men. Figure 40 reflects the country’s excellent performance in child mortality (9.9 deaths /1,000 births, World Bank) and maternal mortality (40 deaths /100.000 births in 2010, WHOSIS), which is partly a result of a universal institutionalization of births (99.1 per cent). Costa Rica’s very good performance, even in 1990, makes it quite difficult to meet its Millennium Development Goal (MDG) four (child mortality) or MDG 5 (maternal mortality). For Costa Rica to meet its MDGs, it would have had to reduce its rates to some of the lowest in the world as it is already one of the top performers in terms of these indicators (Figure 40). Figure 40: Costa Rica MDG 4 and MDG 5 results Part of Costa Rica’s success in positive health outcomes can be attributed to the high utilization rates across all quintiles. Despite the rise in private spending, utilization rates remain high, and public institutions provide over 80 per cent of services (Figure 41). Utilization of public outpatient services is similar for all quintiles, and surveyed individuals from all income groups consider the CCSS as the most important ambulatory provider for ambulatory health services (primary health care or EBAIS and specialized care provided by the CCSS). Yet, CCSS inpatient services tend to be utilized by the better off (Figure 42). 48 Costa Rica Social Sector Expenditure and Institutional Review Figure 41: Outpatient visits by quintiles, Figure 42: Inpatient visits by quintiles, 2006 2006 Source: World Bank SSEIR team’s analysis of household Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT surveys, authors’ calculations using standardized ADePT software (Health Module). software (Health Module). Non-communicable diseases pose the greatest challenge to Costa Rica’s healthcare system. Improvements in income, environment, lifestyles, and medical services have resulted in higher life expectancy and a rapid increase in the percentage of the population represented by the elderly (Figure 43), leading to a rise in NCDs and greater demand for long-term care. The increased demand introduces new pressures: clinical (new diseases that require new treatments), financial (higher and continuous use over time), and also related with patient satisfaction. As for the latter, surveys show a decreasing perception of quality of hospitals and surging waiting lists times at CCSS, particularly for key surgery services. Figure 43: Demographic trends 1990 compared with 2005 Costa Rica, 1990 Costa Rica, 2005 90-94 90-94 75-79 Male Female 75-79 Male Female 60-64 60-64 Age Age 45-49 45-49 30-34 30-34 15-19 15-19 0-4 0-4 15 10 5 0 5 10 15 15 10 5 0 5 10 15 Population (%) Population (%) Source: World Health Organization (WHO): Country Basic Data, 2013. 49 Costa Rica Social Sector Expenditure and Institutional Review Costs, particularly for chronic and other non-communicable diseases, are growing exponentially while productivity seems to be lagging. A brief review suggests that the cost of treatments at CCSS have grown dramatically in the past eleven years. (Table 3). Therefore, costs need to be closely monitored and analyzed to better understand their main drivers. Table 3: Increase in annual cost of treatment for selected NCDs Annual cost of Treatment* Condition/Disease (in millions of colones) 1998 2000 2005 2009 Diseases of the Circulatory System (Total) 11.506 15.935 33.605 73.525 1.Hypertension 4.917 7.35 14.969 30.74 2. Rest 6.589 8.585 18.636 42.785 Tumor (neoplasms in general) 5.775 7.448 17.661 43.151 Diseases of the Respiratory System 27.679 32.937 66.549 114.702 Digestive System Diseases 10.812 16.313 34.073 73.308 Conditions of the Endocrine System, Nutrition and 6.059 10.62 21.36 41.592 Metabolism (Total) 1. Diabetes 4.148 6.288 12.509 24.982 2. Rest of conditions 1.912 4.333 8.851 16.61 Source: Department of Health Statistics, CCSS, 2011 The challenges posed by the broad demand for services and the increase of non- communicable diseases are already exerting important pressures on the health system. These new demands impose pressures on the CCSS’ delivery of care and its financial management. This trend is likely to continue and intensify over the next decades. The CCSS is the main provider of ambulatory and inpatient services but the current model of health care, and financial-administrative management model shows signs of stress as suggested by patient dissatisfaction mainly because of waiting lists. Waiting lists for highly specialized services continue to negatively affect patient satisfaction and public confidence in the responsiveness of the system. Waiting lists in surgery and external consultations have grown considerably over the last decade (Figure 44). As a major source of dissatisfaction, they are systematically noted by the media and public leaders, thereby resulting in an erosion of public confidence in the CCSS. 50 Costa Rica Social Sector Expenditure and Institutional Review Figure 44: Waiting lists for specialized services: 2007-2011 Source: Monge et al, 2013 Dissatisfaction with the services being provided by the public hospitals has increased. From 2007 to 2011 the percentage of people who said they were not at all satisfied or not very satisfied with the way the public hospitals work increased (Figure 45). This is a trend that is not seen in the majority of the other Central American countries. Not only are people in Costa Rica less satisfied with the public hospitals, but a higher share of the population is reported to be satisfied with private hospitals and clinics. For a country with such commitment to public service, it is essential that Costa Rica improve service delivery at CCSS. Figure 45: Satisfaction levels in Central America with the ways public hospitals work: 2007 and 2011 Source: World Bank SSEIR team’s analysis of Latinobarometro 2011, authors’ calculations V.3 Institutional Arrangements Since it was established in 1943, the Costa Rican Social Security Fund has been an exemplary health system model. The CCSS is a unified health system. It owns a wide network of hospitals, clinics, and primary health care services and purchases all medicines, laboratory 51 Costa Rica Social Sector Expenditure and Institutional Review reagents, equipment, and other inputs needed to provide individual health care services. The CCSS insures 88 per cent of the population (Saenz, 2011) and is responsible for the vast majority of health services in Costa Rica. As Figure 41 shows, the public sector provides over 80 per cent of outpatient consultations in the country, with a combination of CCSS and EBAIS (Equipos Básicos de Atención Integral en Salud). In addition, the CCSS provides indirectly about 7 per cent of consultations through its contracts with cooperatives (“others” category). In contrast, the private sector covers about 12 per cent of outpatient consultations. The CCSS is a public but autonomous entity that is financed primarily financed through payroll taxes. CCSS’s budget is raised mainly through payroll taxes (87 per cent), which the CCSS collects directly from formal workers. Another 11 per cent of the budget comes from the central government, as a payment to insure the informal employees. The rest comes from other sources, including the National Institute of Health. The budget is approved by the Board of Directors under the supervision of the general comptroller’s office, but does not require congressional approval. Costa Rica’s universal coverage has been achieved through the expansion of a single social insurance scheme. The CCSS currently covers about 90 per cent of the population. This is partly a consequence of a successful integration of the social insurance (for the formally employed) and Ministry of Health (informally employed), whereby the former absorbed the facilities of the latter during the 1990s. In addition Costa Rica has a long primary care tradition, so the CCSS manages an extensive network of primary and basic specialized services. There is a clear separation of functions between the Ministry of Health and the CCSS . The Ministry of Health is responsible for health sector stewardship and key public health services, while CCSS provides individual health care services. At the same time, the CCSS is responsible for selected public health interventions, such as vaccinations and data collection of public health risks at the community level. The private sector provides small scale ambulatory and diagnostic services to all income groups, and inpatient care mostly to the well off. In moving forward, the CCSS will need to harness its strengths and address its challenges. As any institution, the CCSS has its own set of strengths and weaknesses. These include the following:  Robust primary care: the CCSS has 1,014 primary basic teams, which are always staffed by at least a doctor, a nurse, and nursing agents. In addition, nursing agents collect a census of community health data covering virtually the whole country.  High burden of national hospitals. The CCSS has three national hospitals in San José; each of which a center of reference for a catchment area. The national hospitals absorb considerable resources and do not have harmonized service offers, so there is extensive duplication. 52 Costa Rica Social Sector Expenditure and Institutional Review  Fragmented management at the central level: At CCSS there are currently six Division Managers (Medical, Financial, Administrative, Infrastructure, Logistics, and Pensions) who tend to work in silos. These are presided by the CCSS Social Security President, who has both political and technical responsibilities. Decisions tend to be based on partial information, and it has been difficult to achieve a balance between clinical efficiency and finances.  Health system fragmentation: at CCSS, the clinical information platforms (primary care and hospitals) rely often on information collected on paper, which is then manually entered and analyzed at the central level. Such procedures result in data limitations and large time lags to obtain indicators. The administrative and financial platforms are automatized, but based on outdated programming languages (Cobol and similar), thereby limiting their functionality and raising their costs.  Historic budgeting is preventing the equity of resources. In a historical budgeting system, the CCSS assigns resources to health centers based on what they received in the previous exercise. Thus, this allocation does not correspond to the population needs; a simple calculation of the resources allocated to each of the three national regions shows inequalities. Addressing the historical budgeting system would be therefore essential to attain fair and efficient services. The CCSS is preparing an institutional strengthening initiative based on three major areas. Over the last year, the Board of Directors of the CCSS has reviewed the Strengtheing Equity and Sustainability of Health Insurance Proposal (Propuesta Programa FESSS) to strengthen the equity and sustainability of the CCSS. The proposal has identified three key areas for improvement, to achieve a timelier, higher quality and efficient form of care: (i) Health Service Delivery: Comprehensive Care Networks and Hospitals; ii) Funding and Resource Flow, and iii) Management, which provided the bases for the Bank’s recommendations Health Service Delivery Articulate integrated health services networks (RISS). According to the WHO, various European and Latin American countries are using RISS to improve chronic disease care; salient examples include the Basque Country (Spain), Catalonia (Spain), and Brazil . Articulating RISS implies integrating the health facilities (primary care, hospitals, etc.), as well as the health facilities with their ancillary functions (laboratories, ambulances, etc.). In addition, the RISS manage high complexity patients proactively through case management, disease management, and others. The articulation of RISS necessitates various elements, some of which are attainable in 3-5 years. These include a) creating network management teams; b) stratifying the Costa Rican population to identify medium and high risk patients; c) testing performance-based payment mechanisms for health facilities (i.e. prospective capitation), and d) expanding information technologies. The latter entails finalizing the implementation of the family health (SIFF) and clinical history (SIES) modules of the EDUS platform (Expediente Único Digital en Salud). The financial and administrative modules of an Enterprise Resource Planning (ERP) 53 Costa Rica Social Sector Expenditure and Institutional Review platform would also need to be implemented (these are discussed below in the management section). It is suggested that these interventions are first tested in a pilot project. To articulate the networks, Costa Rica can build on and further strengthen its robust primary care. All integrated network models rely on strong primary care, which can serve as the health system’s gatekeeper and also reduce unnecessary referrals. The CCSS can, therefore, harness its robust primary care platform and further strengthen it by introducing family medicine, and/or the continuous training of first level personnel in the most common chronic diseases (diabetes, cancer and others). The first level will also need direct access to health technology, which includes digital medical records, telephone consultation with specialists, and others. Conversion of CCSS tertiary hospitals (particularly the three national hospitals) into excellence centers. These hospitals could specialize in high complexity-high cost services, with each center specializing in its respective area of expertise. Through a new hospital management unit, the CCSS could rationalize the services of these hospitals and facilitate their transformation into centers of excellence, and assure that they offer complementary, coordinated services. A plan is needed to improve the management of Major Ambulatory Surgery (CMA) to promote sustainability and reduce waitlists. CMA is commonly used in OECD countries to enhance the efficiency of surgery services and patient satisfaction, who get to spend less time at hospitals. Though on the rise at CCSS, CMA is still mainly confined to national hospitals. Therefore, a plan would be needed to extend CMA to the rest of the health facilities. For example, El Salvador, has published a central resolution specifying which surgical procedures should be carried out as CMA and the requirements for implementation. The creation of sub-acute units can further improve the timeliness and quality of hospital care. Sub-acute units provide specialized geriatric and chronic disease services (chemotherapy, etc.) at lower cost and higher quality than traditional hospital forms. The experience in the Basque Country shows that sub-acute cases represent 20 per cent to 25 per cent of hospitalizations, whereby their management through sub-acute units can incur considerable savings. The Bank recommends that the CCSS create a sub-acute unit in the geriatric and other tertiary hospitals. Home specialization can be also used to provide more affordable and quality of care to selected patients B. Funding and Resource Flows Capitated payments to networks can enhance equity and efficiency. In prospective capitation, the networks receive funds based on the characteristics of the population they cover. The funds are calculated based on a formula that takes into account population size and characteristics (burden of disease, geographical dispersion, etc.). As the international literature supports that prospective capitation favors the equity and efficiency of services, the Bank suggests assigning funds to networks based on this method. 54 Costa Rica Social Sector Expenditure and Institutional Review Keep the historic funding initially in hospitals but gradually introduce prospective payment mechanisms. Prospective payments that maintain adequate funding but allow for greater transparency in the services offered be gradually phased. These payments could be based on Diagnostic Related Groups (DRGs) and similar payment mechanisms. C. Management The creation of a General Manager position would be a quick win to improve CCSS management. The CCSS President has both political and technical responsibilities, and the CCSS Division Managers (Medical, Financial, Administrative, Logistics and Infrastructure) work in silos. Hence, the key institutional functions are not integrated. It is recommended that CCSS creates a General Manager position, who would run the institution based on technical criteria and strike a balance between the various CCSS functions, particularly between healthcare delivery and finances. A Strategic Intelligence Unit can help streamline senior management decisions. Currently, information is highly fragmented at CCSS, so senior management is unable to make decisions based on evidence. The Bank therefore advises the creation of a Strategic Intelligence Unit, composed of a multidisciplinary team that will integrate the information existing at CCSS to prepare the needed information (indicators, reports, and others) for senior management decisions. The Institutional Strategic Intelligence Group would report to senior management (General Manager and Division Managers), and be composed of experts in planning and analysis. Improve the functioning and coordination between key information systems. Two information systems are key: Enterprise Resource Planning (finance and administration) and EDUS (healthcare delivery). Currently, the CCSS operates various IT platforms that are disconnected. As a result, it is unable to draw administrative (human resources, drug availability, etc.) and financial information (budget executed, etc.) for management decisions. Such vital information can be generated through the implementation of an ERP (Enterprise Resource Planning) platform, prioritizing the basic accounting, financial management and human resource modules. In the short term, the Bank recommends that clinical information be integrated across the organization to manage the care of chronic patients. This can be achieved through the expansion of the EDUS platform (Unique Patient Health Record) throughout the organization. The EDUS platform is already under implementation, and two of its modules are considered priorities: Clinical Electronic Basic Module - SIES and Family Records – SIFF. The Bank also recommends the expansion of EDUS to emergency services at hospitals. 55 Costa Rica Social Sector Expenditure and Institutional Review VI. Performance and Challenges in Social Protection and Labor VI.1 Recent Evolution of Social Protection and Labor Public Spending The Social Protection and Labor (SPL) system in Costa Rica is composed of both universal and selective programs. The promotion of universal social programs in mid-twentieth century gave rise to an extensive network of services for the entire population in the areas of health, social security, education, housing and basic services (water and electricity). In the 1970’s, the SPL system was expanded with targeted programs aimed at the poorest and most vulnerable population. Since then, the most significant developments of the Costa Rican social protection system in the beginning of the 21st century have been programs providing incentives for education (school feeding, scholarships, CCTs), social assistance pensions and social services for children and adults elderly (Vega, 2012). Public spending in SPL is high and has increased over the last few years both in real per capita terms and as a share of GDP. SPL spending as a share of GDP grew on average 3 percent per year, rising from 7.1 percent of GDP in 2007 to 8.7 percent in 2013 (Figure 46). In per capita terms, SPL spending also rose importantly in real terms during that same time period, (Figure 47). The SPL sector in Costa Rica accounts for the largest share of social spending— roughly 40 percent—but its share has decreased over time on average 1% per annum (40% in 2013 vs 44% in 2007). As a share of GDP (as well as in real dollar terms), overall SPL spending in Costa Rica is by far the highest in the Central American region (Figure 48), mostly accounted for social security spending, and among the highest in the LAC region. Figure 46: Public Spending on SPL as a % Figure 47: Public Spending on SPL per of GDP (%) capita constant local currency (2007) Source: World Bank SSEIR/ICEFI Social Spending Database Source: World Bank SSEIR/ICEFI Social Spending Database 56 Costa Rica Social Sector Expenditure and Institutional Review Figure 48: Public Spending on SPL as a % of GDP by countries Source: World Bank SSEIR/ICEFI Social Spending Database Social Security accounts for the bulk of SPL spending, but this share has slightly decreased in the past few years. The social security sector in Costa Rica is very large—mostly due to generous public sector pensions. Social security spending has increased in recent years, from 6.4 percent of GDP in 2007 to 7.5 percent in 2013 (Figure 46). Despite the increase in spending, the share of SPL spending attributable to social security has decreased (87% in 2013 vs 91% in 2007). The social security sector is administered by the Costa Rican Social Security Institute (CCSS) and includes health insurance, contributory and non-contributory pensions11. The CCSS alone accounts for almost 4% of GDP and more than 50% of social security spending (Figure 49). The amount spent on special regimes was 2.5% of GDP in 2013 and has remained constant in recent years. Teacher’s pensions are also quite generous and represent around 0.6% of GDP in 2013. 11 For the purpose of this section, we exclude the analysis of health insurance since it is covered in the health section of this document 57 Costa Rica Social Sector Expenditure and Institutional Review Figure 49: Social Security spending by main categories, 2007-2013 as a % of GDP Source: World Bank SSEIR/ICEFI Social Spending Database Social assistance spending rose significantly from 2007 to 2009 but has remained stagnant since then. The non-contributory components of Costa Rica’s SPL system include a wide range of social assistance benefits, labor market programs, and social services. Spending on non- contributory programs increased as a shared of GDP from 0.7 percent in 2007 to 1.1 percent in 2013. Cash transfers explain the majority of the increase since its allocation almost doubled during that time (0.6 percent in 2013 vs 0.3 percent in 2007). The Avancemos CCT is the largest SP program. All of the other components of the social assistance system have remained constant in terms of percentage of GDP. Active labor markets programs account for the second largest share of social assistance spending, averaging 0.4 percent of GDP between 2007 and 2013. Compared to its neighbors, social assistance spending is in the medium to lower range (average of 1.4 percent of GDP for the rest CA countries) (Figure 50). Figure 50: Social Assistance Spending as a % of GDP 2007-2013 Source: World Bank SSEIR/ICEFI Social Spending Database 58 Costa Rica Social Sector Expenditure and Institutional Review The Fund for Social Development and Family Allowances, FODESAF, is the main source of financing of several social security, social assistance and labor programs. FODESAF was created in 1974 (Law 5662) and is managed by the Department of Social Development and Family Allowances (DESAF), which belongs to the Ministry of Labor and Social Security (MTSS). It manages resources equivalent to around 2% of GDP and it is mainly funded from two main sources (with almost equivalent weights): a) a fixed central budget contribution, and b) contributions from public and private employers equivalent to 5 percent of their total wage and salary expenses. FODESAF provide grants to finance around thirty programs spread over twenty public and private entities (Estado de la Nación, 2013). Around 63% of its budget is allocated to three large programs: Non contributive pensions (CCSS), IMAS (CCT and other cash transfers programs), and BANHVI (Housing Voucher). (Table 4). Table 4: FODESAF: Budget allocated to SPL programs Source: DESAF (2014). Informe de Gestión 2010- 2014. 59 Costa Rica Social Sector Expenditure and Institutional Review VI.2 Social Protection and Labor Outcomes and Challenges VI.2.1 Social Security The social security and insurance system covers a large and growing portion of the population, and its coverage is one of the highest in Latin America. Social security and insurance programs in Costa Rica fall into five categories: contributory health insurance, non- contributory health insurance, unemployment insurance, contributory pensions, and non- contributory pensions. Illness and maternity includes the universal health insurance (Seguro de Enfermedad y Maternidad) and the health insurance for the poor and extreme poor State (Asegurados por cuenta del Estado – FODESAF). Unemployment insurance and benefits (Auxilio de Cesantia) is constituted by a scheme contributing employers, employees, and the government. The non-contributive pension (Regimen No Contributivo de la CCSS, financed by FODESAF) is a basic pension for old age individuals in situation of poverty who have no access to regular pensions. The contributive pension system consists of both mandatory (financed equally by employers, employees, and government, only available to formal sector) and voluntary contributions. Contributive pensions entail an old age, disability and death fund (RIVM), mostly for private sector employees, and a national budget program (RPN), which itself comprises fourteen programs providing pensions to public sector employees, mostly teachers. Table 5 shows a breakdown of the main social insurance programs including description, target population, expenditures and coverage. Table 5: Main Social Insurance Programs Main Social Insurance Description Target Population Coverage Spending Programs per year % of GDP Invalidez, Vejez y Muerte Disability, old age and Public and private 3.19 (IVM) de la CCSS survivors pension (CCSS) employees 200,639 RNC de la CCSS, Non- contributory regime Poor and 0.47 financiado por FODESAF CCSS financed by Extreme poor 97,838 FODESAF Seguro de Enfermedad y Sickness and Maternity Health Insurance 0.38 Maternidad (SEM) -CCSS security (CCSS) – Universal 2,375,379 Asegurados por cuenta del Insured by the State Health Insurance 0.07 Estado (indigentes) - (indigent) - FODESAF - Poor and Extreme 204,423 FODESAF poor Régimen del Magisterio Magisterio Nacional scheme Private and Public 0.05 Nacional teachers 38,609 Régimen del Poder Poder Judicial scheme Judicial 0.15 Judicial employees 3,084 Régimen a cargo del Ministry of finance scheme Public employees 0.02 Ministerio de Hacienda Auxilio de Cesantía - Unemployment Relief - Employees 0.17 60 Costa Rica Social Sector Expenditure and Institutional Review Fondo capitalización labor capitalization fund laboral Seguro de Riesgos del Labour Risk Insurance - Employees Trabajo - INS INS 1,130,856 Source: World Bank SSEIR, authors’ calculations using SSEIR/ICEFI Social Spending Database, FODESAF and LAC SP database In terms of pension coverage, it has increased importantly in recent years, and it is now close to the top coverage countries (Chile and Uruguay) in the LAC region. The share of employees contributing to the system has increased in recent years from 63% in 2007 to 69% in 2014. In addition, around 15% of employees contributed voluntarily during the same period (Figure 51). Compared to CA countries, Costa Rica is by far the country with the highest share of contributors to pensions for all income levels (Figure 52) and among the highest in LAC countries, only comparable with Chile and Uruguay (Rofman, Apella, & Vezza, 2015). In CA, only Panama is closer to these figures (60 percent coverage). Even among the poorest (lowest income quintile), half of them contribute to the system, a percentage among the top in the region. Figure 51: Share of workers contributing Figure 52: Share of workers contributing to SS to SS by countries Source: World Bank SSEIR Source: World Bank SSEIR However, Costa Rica’s contributive pension program is a highly regressive social insurance intervention. Public spending in contributive pensions accounts for about 30% of SPL spending. Government-subsidized contributive pensions are regressive by nature since financial contributions and employment in the formal sector are both eligibility requirements. In Costa Rica, the vast majority of contributory pension funding comes from government transfers. The government subsidizes both RIVM (CCS) and RPN pensions (Figure 49). The latter, in particular, are overly generous thanks to this government subsidy (Figure 53). Private contributions (by individuals) in the public sector scheme (RPN) accounted for just 14 percent of payments received by beneficiaries in 2006; this large subsidy that is only available to 61 Costa Rica Social Sector Expenditure and Institutional Review individuals who have formal sector jobs represents a highly regressive government intervention (World Bank, 2008). Figure 53: Average Pension Benefit (% of average salaries, 2013) 100 90 80 % of average salaries 70 60 50 40 30 20 10 0 Public Pensions - Public Pensions - Public Pensions - CCSS Civil Servants Teachers Judiciary Source: Ministerio de Hacienda (2013), “En Ruta hacia la Consolidación Fiscal: Agenda para un Diálogo Nacional”. The share of elderly covered by pensions is on the rise, even among the poor, due to non- contributory pensions. In Costa Rica, the elderly benefit from one of the best pension systems in Latin America and the Caribbean. The pension system covers more than 60 percent of the population over 65 years old (Figure 54), which is among the highest out of all other LAC countries.12 Moreover, pension coverage increased for all income quintiles, the poorest benefit from social pensions while the richest benefit from contributory pension (Figure 55). In addition, compared to its neighbors, Costa Rica has the least inequality in terms of access to pensions among the elderly; the coverage rates for the poorest and richest quintiles differ by 8 percentage points (14%) (Figure 56). In Costa Rica, pensions are responsible for 10% of the reduction in poverty from 2000 to 2008—a share that is slightly above the Latin American average of 9% (World Bank). 12 See Ribe, Robalino, and Walker (2010). 62 Costa Rica Social Sector Expenditure and Institutional Review Figure 54: Pension coverage 2007-2014 Figure 55: Pension coverage elderly by quintile Source: World Bank SSEIR team’s analysis of household Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculationsusing standardized ADePT surveys, authors’ calculations using standardized ADePT software (Social Protection Module). software (Social Protection Module). Figure 56: SS coverage of elderly total and by quintiles (%) Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Social Protection Module). The non-contributory pensions cover around 40 percent of the poorest, but coverage can be improved by improving targeting accuracy. More than 50 percent of the poorest elderly have access to the pension system, which is high compared to the rest of CA countries (Panama 45%, El Salvador 14%, Guatemala 6% and Honduras 3%) (Figure 57). The coverage of the poor is explained mainly by the social pension system; around 40% of elderly in the first quintile are covered by the social pension. However, some leakages persist in the system; 31% of social pension beneficiaries are in the top three quintiles of income distribution. Coverage of the elderly among the poorest quintiles could be improved with better targeting accuracy of social pensions. 63 Costa Rica Social Sector Expenditure and Institutional Review Figure 57: Social Pension – Distribution of beneficiaries (Targeting Accuracy) % elderly Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Social Protection Module). Costa Rica has undergone several parametric reforms in recent years to strengthen the financial sustainability of the system. The parametric reforms were implemented with the objectives of improving the distribution of benefits across the Costa Rican population and decreasing evasion of contribution requirements. To avoid a collapse of the RIVM, several reforms were enacted in 2005: (i) the monthly contributions needed to retire were increased from 240 to 300 colones; (ii) the percentage contribution was increased by 0.5 percent every five years until reaching the maximum of 10.5 percent of the salary; (iii) the reference period to estimate the pension’s value was increased from the last four to the last ten years of salary; (iv) benefit payments were changed so that they are no longer a fixed percentage of the salary but decreased as the salary increased, (v) early retirement options were introduced to reduce incentives for fraudulent disability claims, and (vi) efforts were made to enforce universal coverage. Despite these parametric reforms, there are still some questions about the long term sustainability of the pension system. Through these parametric reforms, the RIVM system estimated that the point when contributions would no longer cover expenses had been postponed to 2041. The CCSS accounts are currently roughly balanced, but the demographic trends in the country, where population average age is raising, imply an imbalance in the foreseeable future. On the other hand, the special pension regimes (education, civil service and judiciary) are a more urgent problem. The special regimes are far more generous than the CCSS regime, which raises the question of equity, apart from their burden on the fiscal accounts. For instance, in 2012, the CCSS spent 2.2 percent of GDP in general pension payments for 190,000 pensioners, against 2.6 percent of GDP for 62,500 pensioners of the special regimes. And while the average annual pension of a retired worker of the general system of the CCSS amounts to around 50 percent of GDP per capita, the average pension of a retired civil servant is about 170 percent of GDP per capita, and the average pension of a retired judge is about 270 percent of GDP per capita. A recent study by the IMF projects that central government on special regime pensions will remain 64 Costa Rica Social Sector Expenditure and Institutional Review flat at 2.4 percent of GDP through 2020, and will slowly diminish after that. Moreover, the special regime for the judiciary will start to run deficits very soon, and will deplete its assets by 2030. Overall government support to special pension regimes will cost Costa Rica about 59 percent of (2012) GDP over the next four decades.13 VI.2.2 Social Assistance and labor Costa Rica has several social assistance programs managed by a myriad of institutions; both universal social benefits as well as programs targeted only to the poor and vulnerable. Social assistance and social care services includes around 46 programs managed by 22 institutions that provide support mainly for children, women, the elderly and disabled. The largest of these programs are: school feeding (Comedores escolares-MEP), subsidies (Bienestar y promocion familiar, IMAS), housing vouchers (Bono Familiar de vivienda), the conditional cash transfer (CCT) program Avancemos, scholarships (Becas FONABE), childcare services (Centros de Educación y Nutrición - Centros Infantiles de Atención Integral (CEN-CINAI), and other social care services (Red Nacional de Cuido, PANI) (Table 6). Table 6: Main Social Assistance and social care services Social Assistance and Social Description Target Coverage per Spending Care services Population year % of GDP Comedores escolares - MEP School Feeding and All public nutrition program- MEP education centers 673,445 0.59 Bienestar y promocion Subsidies Poor and Extreme familiar (IMAS) poor families 201,526 0.43 Bono Familiar de la Vivienda Family Housing Voucher Poor families with - (Banco Hipotecario de la - (Housing Mortgage housing needs 9,978 0.32 Vivienda) Bank) Avancemos - IMAS Conditional Cash Poor families. Transfer, Moving Ahead Children between 174,196 0.20 IMAS 12-25 years old Becas FONABE Scholarships FONABE Poor and extreme poor students 131,075 0.16 Centros de Educación y Education and Nutrition Poor and extreme Nutrición (Cen) y Centros Centers (Cen) and poor children 0-13 125,562 0.05 Infantiles de Atención Integral Children's and women (Cen-Cinai) Comprehensive Care Centers (Cen-CINAI) Construyendo Lazos de Building Solidarity Links Poor elderly Solidaridad/Red Nacional de / National Network of 4,610 0.05 Cuido - CONAPAM Care – CONAPAM 13 World Bank (2015), ibid 65 Costa Rica Social Sector Expenditure and Institutional Review Protección integral de la niñez Comprehensive All children y la adolescencia - PANI protection of childhood 7,248 0.05 and adolescence - PANI Pensiones de guerra War pensions Participated in war between 1948- 0.04 1955 Programa mujeres jefas de Program women heads of Poor women hogar en condición de pobreza households in poverty 0.03 (INAMU) (INAMU) Atención a la discapacidad - Attention to disability - Disability, poor CNREEE CNREEE 3,458 0.03 Pensión vitalicia a personas Lifetime pension for People with con parálisis cerebral people with deep cerebral cerebral palsy 2,380 0.03 profunda - RNCP palsy - RNCP Gracia (ex presidentes y otros) Grace (former presidents Elderly (former and others) public employed) 2,310 0.02 Atencion y educacion jovenes (Ciudad de los ninos) 486 0.01 Distribución de alimentos a Food distribution to Poor children 0-13 familias (DAF) - Ministerio de families (DAF) - Ministry years 4,909 Salud of Health Centros de Cuido y Desarrollo Care Centers and Child Poor and extreme Infantil (CECUDI) - Red Development (CECUDI) poor 6,062 Nacional de Cuido - National Network of Care Atención integral a familias en Comprehensive care for Families, mainly situación de pobreza (IMAS) families in poverty women and 3,273 (IMAS) teenagers Source: World Bank SSEIR, authors’ calculations using SSEIR/ICEFI Social Spending Database, FODESAF and LAC SP database Despite a wide portfolio of social programs, coverage is still low—especially for the poor. Of all programs, the universal school feeding program has the largest coverage of the entire population to reach 33 % in 2013, and 53% among the poorest income quintile (Table 7). (Figure 58). Avancemos (CCT) is the second largest program, with coverage increasing from 1.1% in 2007 to 12.9% in 2010, after which coverage stabilized at that level. Despite this increase, the CCT’s coverage is still especially low among the extreme poor and has remained around 28% for the lowest income quintile since 2010 (Table 7). With respect to the rest of the programs, none of them cover more than 15 percent of the extreme poor. Table 7: Coverage main social assistance programs, 2014 (% households) Quintiles of per capita consumption Poverty Status Total Q1 Q2 Q3 Q4 Q5 XP MP NP School feeding program 32.6 53.9 45.0 34.4 20.4 9.1 56.5 49.6 26.8 Avancemos CCT 10.5 21.0 17.2 10.6 3.4 0.4 20.9 19.6 7.7 Food pension 10.0 13.5 11.8 11.1 8.2 5.4 15.9 12.5 8.9 66 Costa Rica Social Sector Expenditure and Institutional Review Fonabe 8.5 18.7 14.3 6.9 2.3 0.4 19.7 15.7 6.0 Social pension 6.3 13.0 7.7 5.7 4.0 1.0 12.5 12.5 4.4 CEN – CINAI 5.8 15.1 7.9 3.9 1.8 0.2 18.1 11.1 3.5 Transportation subsidies (education) 5.2 10.6 8.1 4.7 2.0 0.8 9.5 8.3 4.2 Other SA IMAS 2.7 7.1 4.3 1.3 0.7 0.1 8.1 6.1 1.5 Other scholarships 2.7 2.9 3.4 3.2 3.1 0.9 2.9 3.0 2.6 School supplies 1.3 1.9 1.3 2.1 0.6 0.5 2.1 1.5 1.2 Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Social Protection Module). Figure 58: Main social assistance programs - Coverage first quintile 2009-2014 (% households) Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Social Protection Module). Coverage among the poor can be enhanced by improving targeting accuracy. Despite recent improvements, targeting is still inadequate and there is a big opportunity to improve targeting and increase coverage among the poor. Most social assistance programs like school feeding and scholarships are universal by design and thus not progressive since beneficiaries are spread almost equally across income groups. For instance, scholarship programs are distributed mostly to students in the three highest quintiles (53 percent of scholarships beneficiaries) (Figure 59). The childcare centers CEN-CINAI are the most progressive of the major programs, followed by FONABE (scholarship) and the non-contributive pensions. Avancemos CCT, with 73% of its beneficiaries in the first two quintiles, is disappointingly poorly targeted as compared to other social assistance programs in Costa Rica. In addition, compared to other countries, Costa Rica’s conditional cash transfer program is the worse targeted; measured by the percent of the beneficiaries classified as extreme poor, total poor, or beneficiaries belonging to the first quintile covered, Avancemos CCT ranks at the bottom compared to other CCTs in the region. For 67 Costa Rica Social Sector Expenditure and Institutional Review instance, in Costa Rica only 44% of the CCT’s beneficiaries are poor compared to Honduras with 89%. Guatemala 85%, Panama 82% and El Salvador 57% (Figure 60). Figure 59: Distribution of beneficiaries of main social assistance programs, 2014 (%households) Source: For beneficiaries: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Social Protection Module). Figure 60: Public spending and % of beneficiaries of main CCTs in Central America Source: For beneficiaries: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Social Protection Module). For spending: LAC SP database Avancemos CCT was launched in 2006 with the objective to promote the retention and reintegration into the formal education system of children from families who are struggling to keep their children in the formal education system. 14 From 2006 to 2015, the government 14 In the seventies and early 2000, similar but smaller programs were developed by the IMAS. The program “Superémonos” was performed by the IMAS between 2001 and 2005, and aimed to contribute to the satisfaction of 68 Costa Rica Social Sector Expenditure and Institutional Review established an increasing amount in accordance with the year completed, so that it becomes an incentive to educational attainment. Since 2015, there are only two categories of benefit amount: 22,500 colons for 7th to 9th grades and 35,000 colons for 10th to 12th grades. 15 The transfer is paid monthly and maintained for the rest of the year, provided the student remains in the educational. The coverage expanded from 8,137 students to 185,314 in 2011 and declined to 174,196 in 2014 (Figure 61). Spending has increased from 0.004 percent of GDP in 2006 to 0.27 % of GDP in 2009 and since 2010 has been declining to 0.18 as a % of GDP in 2014. Figure 61: Avancemos CCT: Coverage and Spending as a % of GDP Source: IMAS Still, preliminary estimations suggest that Costa Rica CCT Avancemos has an impact on poverty and in secondary school enrolment. Some preliminary estimation suggests that the CCT has an important impact in poverty especially in extreme poverty. Estimates using 2014 household survey data and national poverty lines showed that the CCT can attribute a reduction of 5.3 percentage points in total poverty and 4.9 percentage points in extreme poverty among beneficiaries. At the national level, the CCT seems to have reduced the extreme poverty and total poverty headcounts by 0.4 percentage points (Figure 62). Moreover, estimates using the same source show that the CCT have an impact on enrollment and attendance, especially at the upper secondary level. Enrollment rates oat the upper secondary level, are higher among children that benefited from the program compared to those that did not receive the benefit. (Figure 63) basic needs of families and encourage children from 6-18 years old to remain in the education system, by providing a monthly bonus (CEPAL, 2012). 15 The idea of this change to promote secondary school enrolment by providing a transfer greater than the one provided by FONABE. 69 Costa Rica Social Sector Expenditure and Institutional Review Figure 62: Poverty headcount before and Figure 63: Enrollment rates ages 5-20, by after the CCT, 2014 CCT, 2014 (extreme poor) Source: World Bank SSEIR team’s analysis of household Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT surveys, authors’ calculations. software (Social Protection Module). VI.2.3 Active Labor Market Policies Costa Rica has the highest unemployment rate in Central America, especially among the youth. Costa Rica has the highest unemployment rate in CA countries, 8.7% compared to 5.2% in CA (according to the ILO Labor definition) (Figure 64). The international crisis that took place in 2009, severely hit the unemployment levels of the country. Overall unemployment rate almost double in 2009, it rose from 4.9 percent in 2008 to 7.8% in 2009, and though it declined in 2010 to 7.3%, it has been increasing since then, to 9.7% in 2014 (Figure 65). Among the youth, unemployment rates were substantially higher, 17.9% in 2009 and have risen more, being 21.4 percent in 2014 (Figure 66). Compared to the CA region, Costa Rica is the country with the highest youth unemployment rate (Figure 67). Moreover, unemployment is still disproportionally high among those who could not finish secondary education (Figure 68). 70 Costa Rica Social Sector Expenditure and Institutional Review Figure 64: Unemployment rate age 15-64, Figure 65: Unemployment rate (%) 2003- CA 2014 Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT Source: INEC software (Labor) Figure 66: Unemployment rate by age Figure 67: Unemployment rate, youth, CA group 2007-2014 Source: World Bank SSEIR team’s analysis of household Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT surveys, authors’ calculations using standardized ADePT software (Labor ILO Module). software (Labor ILO Module). 71 Costa Rica Social Sector Expenditure and Institutional Review Figure 68: Unemployment rate by levels of education 2007-2014 Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Labor ILO Module). Evidence points towards a shortage of high skill workers. Changes in the labor market did not favor the low-skilled and job creation has concentrated in high skilled workers. Previous work have indicated a raise in labor demand for those with higher levels of education and skills from 1994 to 2004 (World Bank, 2007). This increasing demand for high skilled workers seems to have continued. From 2007 to 2014, total employment increased for workers with tertiary or complete secondary education in all economic sectors. In contrast, labor demand for low skilled workers decreased in that same period especially in sectors with low skill jobs as construction and manufacturing sector and agriculture, fishing and mining. As consequence, in the last few years there has been an increasing and constantly higher returns to education for skilled workers than for low skilled workers (Figure 69).16 In fact, from 2007 to 2013, the wage premium for workers with secondary incomplete education or less, decreased while, for workers with higher levels of education increased (Figure 70). This trend has contributed to the increase in inequality measured by the Gini coefficient. Efforts to address the apparent skill shortages are needed like investments in both training and education. 16 Returns to education are especially higher for those in the public sector. The wage bill is the largest line item in the economic composition of Costa Rica’s non-financial public sector. It has averaged 7.4 percent over the last 10 years, which is higher than the Latin American average of about 6 percent. 72 Costa Rica Social Sector Expenditure and Institutional Review Figure 69: Median earnings by educational level relative to total median earnings Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Labor ILO Module). Figure 70: Education wage premium 2007 vs 2013 Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using. Therefore, not surprisingly, Costa Rica spends important resources in Active Labor Market Programs (ALMPs), more event than a large number of OECD countries. Costa Rica spends 0.39% of GDP in ALMPs which is higher than most Central American Countries (except for Nicaragua, 0.46% of GDP) and more than a large number of OECD countries (Figure 71). The main institutions that provide ALMPs are the the National Training Institution (INA), the Ministry of Labor (Ministerio de Trabajo y Seguridad Social) and to a lesser extent the IMAS. ALMPs consist of several programs aiming the youth, the poor and entrepreneurs, for 73 Costa Rica Social Sector Expenditure and Institutional Review instance: training services (INA), PRONAMYPE - Apoyo a la micro y pequena empresa (MTSS, support for micro and small business) and Programa Nacional de Empleo (MTSS) (Table 8). Most of the spending is allocated to training programs, through the national training institute (INA), 0.33% of GDP, which accounts for 84% of total spending in ALMPs (Figure 72). The Ministry of Labor accounts for around 15.6% for ALMPs spending. Figure 71: Public Spending in ALMPs as a % of GDP (%) Source: Table 8: Main Labor Market Programs Main Labor Market Description Target Coverage Spending Programs Population per year % of GDP Programa Nacional de National Employment Poor and Empleo PRONAE - MTSS Program Extreme poor 7,869 0.03 more than 16 or 18 years old Servicios de Capacitación y Training Services – More than 18 Formación Profesional - INA INA years old 92,878 0.32 PRONAMYPE - MTSS Support for micro and Poor and small business – extreme poor 3,520 0.01 MTSS Source: World Bank SSEIR, authors’ calculations using SSEIR/ICEFI Social Spending Datab ase, FODESAF and LAC SP database The National Apprentice Institute (INA) accounts for the largest share in ALMPs by providing technical training; however, did not benefit the most vulnerable. INA was created in 1960 to provide technical training to workers in the manufacturing sector. Nowadays, INA has expanded its portfolio to other economic sectors and promotes, teaches and certifies technical education through a widespread package of training, certification, and technical assistance 74 Costa Rica Social Sector Expenditure and Institutional Review programs. Offers a wide array of training modules across all sectors, with 251 programs and 1,078 modules in agriculture, manufacturing, and services. INA is financed through a mandatory payroll tax of 2 percent, paid fully by employers, reaching 0.32 percent of GDP in 2013. Among similar institutions in Central America, only INATEC in Nicaragua has a higher budget as share of GDP (Figure 62). In 2013 the National Training Institute (INA) reached around 13% of the labor force however, did not focus on groups that are systematically left out of the labor market or marginalized into low-productivity jobs: the youth did not account for the largest share of beneficiaries, - the share of out-of-school youth is less than 20% (Figure 73) and most of beneficiaries belong to the higher quintiles levels (Figure 74). Figure 72: Budget of Public Training Institutions in Central America, 2013 Source: World Bank SSEIR / ICEFI social spending database 75 Costa Rica Social Sector Expenditure and Institutional Review Figure 73: Distribution of beneficiaries Figure 74: Distribution of beneficiaries from INA by age, 2014 from INA by age and income, 2014 Source: World Bank SSEIR team’s analysis of household Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT surveys, authors’ calculations using standardized ADePT software (Labor ILO Module). software (Labor ILO Module). Returns to training suggest that they are generally valued in the labor market; however, returns to INA’s training seem to be low or negative as compared with other training institutions, highlighting quality deficiencies. INA does not seem to provide significant gains in incomes for those who attend it, in contrast to those who receive training from other institutions. Returns to training education in Costa Rica (including INA courses) were estimated using household survey from 2010 to 2014 for individuals 20-30 years-old living in the parental household, and controlling for parental education and income. Results show that training has small positive effects on income compared to those who did not receive any training, around 3% (Figure 75). Though, by type of institution, those who attended INA actually earn less than those with no training while attending other public institutions (e.g. public universities) and private institutions have positive gains. Moreover, for those who did not complete secondary education and attended INA, the wage premium is even lower than for those with no training education (though the coefficient is not significant). 76 Costa Rica Social Sector Expenditure and Institutional Review Figure 75: Premium to training education, 2009-2014 Source: World Bank SSEIR team’s, authors’ calculations using Household Surveys: Several labor market programs encourage entrepreneurships by providing funds, training and advice. IMAS and the Ministry of Labor and Social Security (MTSS) have several labor market programs for entrepreneurs and offer financial and training assistance for its beneficiaries. PRONAYPE “Apoyo a la micro y pequena empresa” is a program managed by the MTSS that supports the creation of small and medium enterprises, the program increased its coverage from 2,696 in 2007 to 3,251 people in 2012. “Ideas Productivas” is administered by IMAS and provides funds, training, and advice, and then accompanies the entrepreneur in the implementation of the business and provides assistance in marketing the products; also offers guarantees for the entrepreneur to access the credit market. 17 Few other labor market program focus on the poor unemployed. PRONAE “Programa Nacional de Empleo” is national employment program managed by the MTSS. The program either incorporates unemployed individuals into various activities of public interest, such as building local infrastructure, or it provides a subsidy (of up to three months) for the person to take up a training program provided by INA or an NGO, or it provides starter funds for a small business, targeting particularly young entrepreneurs and female household-heads. However, this program is quite small, as it barely reached 2,000 beneficiaries in 2007, which corresponds to slightly over 2 percent of the unemployed.18 17 Currently IMAS has an annual investment of around US$ 3 million for this program, benefitting around 2,000 families per year. See ECLAC (2009) 18 See ECLAC (2009). 77 Costa Rica Social Sector Expenditure and Institutional Review VI.3 Institutional Arrangements One of the pillars of the National Development Plan 2015-2018 is to reduce extreme poverty and inequality. The government’s objective is to lift 56,000 families (half of total) out of extreme poverty. The strategy includes the provision of financial support to satisfy the family’s basic needs, and creating conditions for economic independence as a result of the incorporation of these families into the economy. The new poverty reduction strategy "Estrategia nacional para la reducción de la pobreza, Puente al Desarrollo" seeks to promote the satisfaction of basic needs of poor families through interagency coordination. The new figure is the social workers “cogestores” consider as the "bridge" linking the needs of families in extreme poverty and the institutional offer. This figure, facilitates articulation at the local level and provides direct and personalized support to the family. The coordination will be monitored by the Presidential Social Council (comprising the Presidency of the Republic, the newly created Minister of Human Development and Social Inclusion, the Ministry of Labour and Social Security, the Ministry of Education, the Ministry of Health, the Ministry of Culture and Youth, the Ministy of Housing, the National Institute for Women, and the Costa Rican Institute of Sport and Recreation) and will be implemented by the IMAS. However, the current institutional setting of the SPL sector is quite complex in Costa Rica and limit discretionary spending, which can block efforts to achieve poverty and inequality reduction goals. The SPL sector in Costa Rica is comprised by several and uncoordinated institutions (more than 28) without clear leadership or accountability, with programs and actions developed in isolation. For instance, two separate agencies are implementing nutrition and food security programs. Services and benefits provided even within the same institution are not necessarily comprehensive and complementary; instead, they are dispersed, duplicated and fragmented. As a consequence, there is no “graduation” strategy from programs, to derive beneficiaries from one to another. Among the implementing institutions, the main ones are: The Caja Costarricense de Seguro Social (CCSS) is the institution in charge of handling social security in the country (contributory and social pensions, health insurance); The Instituto Mixto de Ayuda Social (IMAS) handles most social assistance programs for the poor, including the CCT program; The Fondo Nacional de Becas (FONABE) provides scholarships to primary and secondary students; The Centros de Educacion y Nutricion (CEN) and Centros Integrales de Atencion Infantil (CINAI) are in charge of ECD/Nutrition interventions; The Instituto Nacional de Aprendizaje (INA) is the national training institution. Budget setting is also complex and limits flexibility. The Fondo de Desarrollo Social y Asignaciones Familiares (FODESAF), a social fund, collects funds from earmarked sales taxes (45.55% of total Budget in 2014) for all institutions above (except INA, financed through payroll taxes). FODESAF functions as a financing middleman that distributes funds to implementing agencies and diverts part of received resources back to central government ministries. 78 Costa Rica Social Sector Expenditure and Institutional Review Figure 76: FODESAF and main institutions SPL programs do not necessarily respond to social demands of families and the monitoring and evaluation is limited and not systematic. Participation on the programs/services are self- targeted, people have to apply for the benefits generating long waiting times, dissatisfaction and exclusions. Social programs are implemented throughout the country, regardless of the reality of each region and do not necessarily respond to the particular needs of each regions and of the target population. In addition, there is no systematic monitoring of social programs delivered to beneficiaries and no evaluations have been performed. Costa Rica is also moving towards improving its social sector informational system through the development of a national beneficiary registry and the use of poverty maps. IMAS has an information system for the target population “Sistema de Información de Población Objetivo (SIPO)”, to register and identify the population in poverty and social risk in the country. The main instrument used by the SIPO to collect the required data is the Ficha de Informacion Social (FIS). The “Sistema de Atencion a Beneficiarios (SABE) is an integrated system that includes the entire process to the beneficiary, from initial attention to the delivery of benefit. The “Registro Único de Beneficiarios (SINERUBE) is the national registry of beneficiaries which aims to maintain a current database and national coverage data with information of all people requiring services, assistance, grants or financial aid, finding themselves in situations of poverty or need. In addition, the government has announced the promotion of family guides whose basic function is to provide and guarantee families in extreme poverty, access to the government services. 79 Costa Rica Social Sector Expenditure and Institutional Review VII. Conclusion and Policy Recommendations VII.1 Education Although public spending on education has increased and it is line with the country’s income, spending on secondary is disproportionately low. Costa Rica spends import resources on education but it spends too little in secondary; both the share of public spending and the spending per student in secondary are small. While it has reached high enrollment rates, there are large within-country disparities. In particular, access to upper secondary and post-secondary education is exceedingly unequal across income quintiles and regions. Furthermore, results from 2012 PISA assessment show that, while the country did on average better than other countries of the LAC region such as Brazil and Colombia, it lagged behind some of its “comparator countries. Finally, only 35% of the workforce has completed secondary education, a percentage that puts Costa Rica not only behind the OECD average (64.7%) but also behind the LAC average (35.4%). The targeting of subsidies and transfers in secondary education can be improved to promote greater access and completion rates and reduce within country inequalities at this level. Although Costa Rica has achieved high enrollment rates at all levels of education, it still faces low completion rates for upper secondary education. These problems are higher for at risk groups from the lowest quintiles making it more difficult for them to overcome poverty. One of the main reasons for dropping out of school in Costa Rica is lack of resources. Avancemos is a CCT aimed at reducing dropout in secondary among the poorest. However, only 29% of the 15- 19 year old belonging to the first quintile of the household income distribution receives. On the contrary, among those in the fourth quintile of the income distribution, 8% report receiving this monetary transfer. Therefore, improving targeting accuracy of the CCT program and also combining this financial support with other non-financial programs that provide tutoring, socio- emotional learning, information about returns to education and programs that help preventing teenage pregnancy can significantly improve the access and reduce the dropout among upper secondary education students in the poorest households. Improvements in teacher quality are critical to reduce the socioeconomic differences in enrollment outcomes and for learning achievements. In Costa Rica, the large supply of university degrees, that often do not comply with quality standards, led to an excess supply of teachers and also to a large variation in teacher quality. In particular, there are very few courses accredited to prepare kindergarten and preschool teachers, where instead the country would need significant investments. Establishing quality standards, through an entry exam, would raise the quality of teachers at all levels. However, quality cannot be preserved in the absence of effective in-service teacher training, an area where there are large gaps within Costa Rica. Finally, there is 80 Costa Rica Social Sector Expenditure and Institutional Review also need for an effective teacher evaluation system. Essential elements of this system are: a) well defined teaching standards, b) multiple measures of teacher performance, possibly combining subjective criteria with objective ones; c) monetary incentives related to the teacher performance. While teacher salaries in Costa Rica are relatively high, they are mostly unrelated to teacher performance. The monitoring and evaluation (M&E) system can be improved by introducing standardized test scores for students and evaluation systems for teachers. The introduction of an adequate M&E system would benefit both the school level planning and the decision making at central level. At school level, a comprehensive M&E system would allow: 1) data driven school planning, 2) teachers to track their students’ gaps and progresses, 3) principals to set targets for their teachers and to monitor them, 4) parents to be better informed on their children’s learning needs. At system level, better and timelier data would allow to redirect resources to geographic areas and education levels that show larger gaps. Autonomy and Accountability should be increased both in Basic and Higher Education. Quality of the assessments, stakeholders’ autonomy and an accountability system in place has proved to have a positive effect on learning. The Basic Education system is characterized by a high degree of school autonomy but more needs to be done to increase the quality of assessment and accountability. Evaluation systems for students and teachers represent a necessary condition. In order to increase accountability, providing parents’ associations with more decision power seems an important first step. In higher education, there are 5 public universities, 52 private ones, and 5 international. Both public and private universities can rely on large autonomy that was granted as part of the constitution. However, especially among private universities there is very little accountability. While institutions that receive public funding have to submit reports to the Contraloría Nacional de la República once a year, private institutions have no similar obligation (Francis, 2014). 67.7% of the programs that have been created between 1990 and 2010 are private, and rather than addressing the needs of the productive sector of the country, they have been mostly focused in the areas of Education and Social Sciences: 45.6% of the 546 private programs are focused in these two areas. There is also limited quality assessment. Only 68 of the 1,165 university programs have been evaluated by the Sistema Nacional de Acreditación de la Educación Superior (Sinaes). VII.2 Health Costa Rica’s public spending on health has grown significantly the past seven years. Its high public spending on health - comparable to the OECD average, which includes high income countries such as Sweden and the United Kingdom - together with its emphasis on universal coverage and primary health care interventions has contributed to positive health outcomes including high life expectancy rates and low child and maternal mortality rates. Improvements in income, environment, lifestyles, and medical services have resulted in longer life expectancies and, therefore, an increasing elderly segment of the population that 81 Costa Rica Social Sector Expenditure and Institutional Review demand long-term care and contribute to the rise in NCDs. The increase in NCDS has posed serious challenges to the system especially in terms of costs and high demand for care, resulting in waiting lists. Aside from NCDs, the CR health system also faces four main issues: (a) fragmentation - with six managers at the central level tending to work in silos; (b) an inadequate health information system; (c) historic budgeting; and (d) the high financial burden imposed by three CCSS national hospitals which tend to duplicate services and absorb considerable resources. In addition, private spending has also increased over time, raising equity concerns. Private spending share has increased from 23.5 per cent in 1995 to 31.9 per cent in 2010. Although both private and public health spending have been consistently increasing over the last fifteen to twenty years, the burden of spending has not been distributed evenly. In particular, household health expenditures are increasing for the poorest two quintiles while they have been decreasing for the rest. Moreover this spending is mainly on services that are also provided by CCSS, raising concerns about quality and effective coverage Moving forward, the Bank proposes the following set of short–term recommendations: Health Service Delivery Build on and further strengthen its robust primary care especially through training of first level personnel. All integrated network models rely on strong primary care, which can serve as the health system’s gatekeeper and also reduce unnecessary referrals. The CCSS can, therefore, harness its robust primary care platform and further strengthen it by introducing family medicine, and/or the continuous training of first level personnel in the most common chronic diseases (diabetes, cancer and others). The first level will also need direct access to health technology, which includes digital medical records, telephone consultation with specialists, and others. Prepare a plan to improve the management of Major Ambulatory Surgery (CMA) to improving sustainability and waitlists. CMA is commonly used in OECD countries to enhance the efficiency of surgery services and patient satisfaction, who get to spend less time at hospitals. Though on the rise at CCSS, CMA is still mainly confined to national hospitals. Therefore, the Bank recommends that the Government prepare a plan to extend CMA to the rest of the health facilities. For example, El Salvador, has published a central resolution specifying which surgical procedures should be carried out as CMA and the requirements for implementation. 82 Costa Rica Social Sector Expenditure and Institutional Review Funding and Resource Flows Gradually introduce prospective payment mechanisms in hospitals, maintaining adequate funding while allowing for greater transparency in the services offered. The latter can be based on Diagnostic Related Groups (DRGs) and similar payment mechanisms. Management Establish a General Manager position to improve CCSS management. The CCSS President has both political and technical responsibilities, and the CCSS Division Managers (Medical, Financial, Administrative, Logistics and Infrastructure) that do not coordinate systematically. Hence, the key institutional functions are not integrated. It is recommended that CCSS creates a General Manager position, who would run the institution based on technical criteria and strike a balance between the various CCSS functions, particularly between healthcare delivery and financing. Create a Strategic Intelligence Unit to help streamline senior management decisions. Currently, information is highly fragmented at CCSS, so senior management is unable to make decisions based on evidence. The Bank therefore recommends the creation of a Strategic Intelligence Unit, composed of a multidisciplinary team that would integrate the information existing at CCSS to prepare the needed information (indicators, reports, and others) for senior management decisions. The Institutional Strategic Intelligence Group would report to senior management (General Manager and Division Managers), and be composed of experts in planning and analysis. Improve the functioning and coordination between key information systems. Currently, the CCSS operates various IT platforms that are disconnected. As a result, it is unable to draw administrative (human resources, drug availability, etc.) and financial information (budget executed, etc.) for management decisions. In the short term, the Bank recommends that clinical information be integrated across the organization to manage the care of chronic patients. This can be achieved through the expansion of the EDUS platform (Unique Patient Health Record) throughout the organization. The EDUS platform is already in implementation, and two of its modules are priority: Clinical Electronic Basic Module- SIES and Family Records –SIFF. The Bank also recommends the expansion of EDUS to emergency services at hospitals. In terms of Medium-term recommendations, the Government could consider the following: Articulate integrated health services networks (RISS). According to the WHO, various European and Latin American countries are using RISS to improve chronic disease care. The articulation of integrated networks will necessitate various elements, some of which are attainable in 3 to 5 years. These include: a) creating network management teams; b) stratifying 83 Costa Rica Social Sector Expenditure and Institutional Review the Costa Rican population (to identify medium and high risk patients); c) testing of performance-based payment mechanisms for health facilities (i.e. prospective capitation); and d) expanding information technologies, i.e. implementing the family health (SIFF) and clinical history (SIES) modules of the EDUS platform (Expediente Único Digital en Salud) in all facilities. It is recommended that these interventions are first tested in a pilot project. Convert CCSS tertiary hospitals (particularly the three national hospitals) into excellence centers. The Bank proposes that these hospitals specialize in high complexity-high cost services, with each center specializing in its respective area of expertise. Through a new hospital management unit, the CCSS can rationalize the services of these hospitals and facilitate their transformation into centers of excellence, and assure that they offer complementary, coordinated services. Create sub-acute units can further improve the timeliness and quality of hospital care. Sub- acute units provide specialized geriatric and chronic disease services (chemotherapy, etc.) at lower cost and higher quality than traditional hospital forms. The experience in the Basque Country shows that sub-acute cases represent 20 percent to 25 percent of hospitalizations, whereby their management through sub-acute units can incur considerable savings. The Bank advises the CCSS to create a sub-acute unit in the geriatric and other tertiary hospitals. Home specialization can be also used to provide more affordable and quality of care to selected patients Funding and Resource Flows Introduce capitated payments to networks to enhance equity and efficiency. In prospective capitation, the networks receive funds based on the characteristics of the population they cover. The funds are calculated based on a formula that takes into account population size and characteristics (burden of disease, geographical dispersion, etc.). As the international literature supports that prospective capitation favors the equity and efficiency of services, the Bank suggests assigning funds to networks based on this method. Management Implement the ERP (Enterprise Resource Planning) platform. The Bank proposes the prioritization of basic accounting, financial management and human resource modules. VII.3 Social Protection and Labor Costa Rica has made important efforts to expand spending and coverage of its Social Protection and Labor (SPL) system, through a mix of both universal and selective programs. The SPL sector in Costa Rica accounts for the largest share of social spending—and as a share of GDP (as well as in real dollar terms), overall SPL spending in Costa Rica is by far 84 Costa Rica Social Sector Expenditure and Institutional Review the highest in the Central American region and among the highest in the LAC region. Most of it obeys to generous social security system (pensions, health insurance) that has expanded coverage, but also to increasing efforts to reach out the poor through cash transfers and complementary assistance to the poor. Despite social security has high coverage, it is overly subsidized in particular for public sector employees; better resource allocation and improved targeting could allow further expansion of the social pension component. In Costa Rica, the vast majority of contributory pension funding comes from government transfers, by subsidizing both RIVM and RPN pensions. And while more than 50 percent of the poorest elderly have access to the pension system, which is high compared to the rest of CA countries, leakages persist in the system. Coverage of the elderly among the poorest quintiles could be improved with better targeting accuracy of social pensions. There are still some questions about the long term sustainability of the pension system. Through these parametric reforms, the RIVM system estimated that the point when contributions would no longer cover expenses had been postponed to 2041. However, recent increases in benefits to the teachers’ pension scheme and opposition to increase contributions have brought this deadline nearer. Recent IMF estimates imply that RIVM is projected to turn a cash deficit over the medium and long term due to system maturation and population aging, with an adjustment equivalent to about 1.5 percent of GDP required to ensure actuarial equilibrium of all pension systems for the next 100 years. It is critical to improve incentives and enforcement so that high-income workers would contribute its fair share, rather than evade or underreport income. It would also be necessary to draw on collective insurance programs already available to expand contributory coverage and family insurance. To improve the quality of coverage, incentives could be introduced so that more people make voluntary contributions to the system. On the other hand, and despite a wide portfolio of social assistance programs, coverage of social assistance is still low, and especially for the moderate and extreme poor. Of all 46 social assistance programs managed by 22 different institutions, only the universal school feeding program has the largest coverage of the poor population. Strikingly, the Avancemos CCT program, launched in 2006 with the objective to promote the retention and reintegration into the formal education system of children, and is poverty targeted, only covers a fifth of the extreme poor, and shows important leakages to higher quintiles. While programs like Avancemos are helping mitigate poverty and inequality (in a context of high unemployment), this impact could be maximize by improving coverage among the poor through enhanced targeting accuracy (without increasing budget envelope). In terms of active labor market programs (ALMPs), Costa Rica spends important resources, more event than a large number of OECD countries; however its efficiency 85 Costa Rica Social Sector Expenditure and Institutional Review needs to improve. Costa Rica has high unemployment (per Central American and Latin American standards), especially among the youth and those who cannot finish secondary or tertiary education. As a consequence, it spends a relatively large proportion of the budget in implementing job skills training interventions (through the national training institution, INA), as well as labor intermediation. However, a thorough assessment and evaluation of targeting and quality of courses delivered, in particular through INA, is needed. While INA reached in 2013 around 13% of the labor force, the share of out-of-school youth is less than 20% and most beneficiaries belong to the higher quintiles levels. Moreover, estimated returns to INA training are negative: beneficiaries actually earn less than those with no training, while attending other public institutions (e.g. public universities) and private institutions have positive gains. This is the right approach, since the current institutional setting of the SPL sector is quite complex in Costa Rica and limit discretionary spending. The SPL sector in Costa Rica is comprised by several and uncoordinated institutions (more than 28) without clear leadership or accountability, with programs and actions developed in isolation. Services and benefits provided even within the same institution are not necessarily comprehensive and complementary; instead, they are dispersed, duplicated and fragmented. The creation of the Ministry of Human Development and Social Inclusion, is thus good news in that regards. But for such an institution to work effectively, it should have some control over the programs’ plans, budgets and budget allocation. As of today, its operational strategy is based very heavily on human and logistical resources available to the IMAS. It also needs to ensure that disaggregated information on program costs and beneficiary profiles are collected and analyzed on a timely basis, in order to better monitor and evaluate the programs in search of improved efficiency and effectiveness. The SPL sector also needs to strengthen its monitoring and information systems, ideally through the institutionalization of a single beneficiary registry from existing program registries and census of the poor. In order to have an integrated SPL system, the country needs to strengthen its information system and consolidate the information on programs articulating the institutional offer with the social demand. A single registry of beneficiaries would improve targeting and diminish overlaps. As of today, CCSS registries (pensions, health insurance) do not communicate with SIPO (registry of social assistance programs), for instance, nor with a census of the poor population. As a consequence, there is no “graduation” strategy from programs, to derive beneficiaries from one to another. A single registry of beneficiaries with a uniform PMT formula would help prioritize interventions for the extreme poor. 86 Costa Rica Social Sector Expenditure and Institutional Review Appendix Appendix 1: Matrix of Short – and Medium-Term Options for Policy Reform Options for Short term (1-2 years) Medium term (3-5 years) Policy Reform Education Increase the  Rethink the general and  Facilitate access to upper access and technical curriculum to make it secondary education by completion rates more engaging and attractive for providing good quality at upper youth (and to the labor market infrastructure in rural and secondary for technical courses). indigenous population regions. education and  Pilot and evaluate the cost decrease within effectiveness of teacher training country programs targeted for at inequalities in risk/vulnerable groups. enrollments.  Pilot and evaluate a program disseminating to parents/students from vulnerable groups the benefits and economic return to education.  Pilot and test the impact of scale up the stipend on the existing CCT programs based on educational level. 87 Costa Rica Social Sector Expenditure and Institutional Review Options for Short term (1-2 years) Medium term (3-5 years) Policy Reform Increase the  Develop financial or non-  Raising accreditation quality of general financial incentives for top standards for university based and technical performing teachers and programs to increase the education students. quality of teacher training.  Increase the enforcement of  Introduce well defined career more instructional time in the progression standards related classroom. with financial rewards to teachers/schools.  Increase the collaboration between the central government and municipalities, principals and teachers in decision making process.  Strengthen the community and school management involvement of parents accompanied by more training/education. Strengthen and  Ensure all new policy pilots are  Develop a systemic approach Institutionalize a tested and evaluated with to monitoring and evaluation Monitoring and rigorous empirical strategies in the education system, evaluation system (program evaluation). including for Higher in the Education  Strengthen and systematize data Education with a (a) well Sector (including collection efforts. defined teaching standards in Higher and (b) standardized tests for Education) students and evaluation systems for teachers.  Introduce at school level, data driven school planning, including setting targets for teachers and students.  Ensure the dissemination of information on student progress and achievement to parents Health 88 Costa Rica Social Sector Expenditure and Institutional Review Options for Short term (1-2 years) Medium term (3-5 years) Policy Reform Health Service  Training of primary care Delivery personnel in chronic disease  Articulate integrated healthcare management networks (RISS), prioritizing: a)  Strategy developed to reduce creating network management wait lists including a plan to teams; b) stratifying the Costa manage major ambulatory Rican population to identify surgeries medium and high risk patients; c) testing performance-based payment mechanisms for health facilities (i.e. prospective capitation, and d) expanding information technologies, i.e. family health (SIFF) and clinical history (SIES) modules of the EDUS platform (Expediente Único Digital en Salud) in all facilities. Funding and  Gradually introduce prospective  Convert tertiary hospitals into Resource Flow payments for hospitals that allow excellence centers for increased transparency of services such as DRGs  Management  Create a General Manager  Create sub-acute units to provide position who would run the specialized geriatric and chronic institution based on technical disease services criteria and strike a balance  Introduce prospective capitated between the various CCSS payments in networks functions  Implement Enterprise Resource  Creation of a Strategic Planning/ERP (starting with Intelligence Unit to prepare the administrative and financial needed information (indicators, modules) IT platform reports, and others) for senior management decisions  Expand implementation of EDUS (Unique Patient Health Record) IT platform to all facilities Social Protection 89 Costa Rica Social Sector Expenditure and Institutional Review Options for Short term (1-2 years) Medium term (3-5 years) Policy Reform  Ensure training, tracking and  Build up and mandate use of Reduce inclusion/ monitoring of the cogestores: key single beneficiary registry exclusion errors figure to avoid the errors of  Advance to an unified definition inclusion and exclusion and of Poverty. increase coverage of the poor - by an articulated range of services/programs.  Similarly, the costestores figure should be used to conduct a comprehensive support families and prevent them from being merely IMAS’s employees.  Fulfillment of commitments (human and financial resources) in institutional plans Strengthen  Strengthen dialogue and  Revise the supply of courses articulation of agreements initiated between offered by INA for vulnerable programs/institut institutions to ensure provision of populations. ions services to the population  Move toward unified payments submitted by the co-managers. (unified card system) with cash  transfer programs. Improve  IMAS should guarantee the Consolidation of information information access and promote the use of systems of social programs: the system information systems to record system of evaluation of social and monitor the beneficiary programs FODESAF, ii) a fully population. shared SIPO by sector, iii) the  Implementation of a common module for the Ficha de single register of beneficiaries of Informacion Social (FIS). targeted social policy, iv) social and  Advance the process of updating cantonal atlas maps (geocoding), v) the FIS which is valid for seven the National Evaluation System years. (SINE) in charge of MIDEPLAN, vi) measuring income poverty, NBI and muldimensionales indicators (INEC). Improve  Develop a financial sustainability  Implement parametric reforms sustainability of plan for the RIVM  Integrate contributory and non- pension system contributory social pension in multi-pillar system 90 Costa Rica Social Sector Expenditure and Institutional Review Appendix 2: Household Surveys databases– Source and definition of variables Countries Period Household Surveys Educati Social Labor Health on Protec tion Costa Rica 2007- Encuesta de Hogares de EHPM, EHPM, EHPM, ENSA, 2014 Propositos Multiples (EHPM) ENAHO ENAHO ENAH ENIGH 2007-2009. Encuesta Nacional O de Hogares (ENAHO) 2010- 2014. Encuesta Nacional de Salud en Costa Rica (ENSA- 2006). Encuesta de Ingresos y Gastos (ENIGH) 2012-2013. El Salvador 2007- Encuesta de Hogares de EHPM EHPM EHPM EHPM 2013 propositos multiples (EHPM) 2007-2013 Guatemala 2006, Encuesta nacional de ENCOVI ENCOV ENCOV ENCOV 2011 condiciones de vida ENCOVI I I I 2006 and 2011 Honduras 2007- Encuesta Permanente de EHPM EHPM EHPM DHS 2013 Hogares de Propositos Multiples (EPHPM) 2007-2013. Demographic and Health Survey (DHS) 2011-2012. Nicaragua 2005 Encuesta Nacional de Hogares EMNV EMNV EMNV EMNV 2009 sobre medicion de nivel de vida EMNNV 2005 and 2009 Panama 2007- Encuesta de Hogares (ECH) ECH, ECH, ECH, ENV 2013 2007-2009. Encuesta de EML EML EML Mercado laboral (EML) 2010- 2013. Encuesta Nacional de Niveles de Vida (ENV) 2008 Methodology: Classification ensures consistency across countries. Education Classification ensures consistency across educational levels: primary education 6 years and for secondary education 6 years. Social Follows World Bank - Aspire classification. Protection Labor Follows ILO classification Health Follows ADePT - Health classifications. Results: Most tables are produced using the ADePT software - Social Protection, Labor, Education and Health. 91 Costa Rica Social Sector Expenditure and Institutional Review Appendix 3: Social spending databases– Source and definition of variables Social Spending: Corresponds to budget executed by centralized and decentralized entities. Period: 2007-2013 Coverage: Central government + Subnational level. All public sectors Data:Total Spending by levels of government, decentralized entities, funding sources and at some times at program level. Classification: Follows IMF classification but with some modification on education and Social Protection. Health: includes expenditure on services provided to individual persons and services provided on a collective basis CA classification IMF Classification Medical products, appliances and 7071 Medical products, appliances and equipment equipment Outpatient services 7072 Outpatient services Hospital services 7073 Hospital services Public health services 7074 Public health services R & D Health 7075 R & D Health Health n.e.c 7076 Health n.e.c Education: includes expenditure on services provided to individual pupils and students and expenditure on services provided on a collective basis. Breakdown of education is based upon the level categories of the 1997 International Standard Classification of Education (ISCED-97) of the United Nations Educational, Scientific and Cultural Organization (UNESCO). CA classification IMF Classification Pre-primary 7091 Pre-primary and primary education Secondary 7092 Secondary education Tertiary 7093 Postsecondary nontertiary education 7094 Tertiary education Other 7095 Education not definable by level 7096 Subsidiary services to education 7097 R&D education 7098 Eduction n.e.c Excludes: teacher's pensions. Includes: Scholarships Modifications: Excludes the amount spent on training institutions. Social Protection: includes expenditure on services and transfers provided to individual persons and households and expenditure on services provided on a collective basis CA classification IMF Classification Sickness and disability 7101 Sickness and disability Social Security 7102 Old age Cash Transfers 7104 Family and children 92 Costa Rica Social Sector Expenditure and Institutional Review Other Social Assistance 7107 Social exclusion n.e.c 7108 R&D Social Protection 7109 Social protection n.e.c 7103 Survivors Active labor Market Programs Amount spent on training institution + labor affairs Subsidies Energy, gas, water. Modification: Excludes: 7105 Unemployment and 7106 Housing. Includes subsidies and Active labor Market spending. 93 Costa Rica Social Sector Expenditure and Institutional Review References Afonso, A., Romero, A., and Monsalve. 2013. Public Sector Efficiency: Evidence for Latin America”. Washington, DC, United States: Inter-American Development Bank. Afonso, A., Schuknecht. L. and Tanzi, V. 2005. “Public Sector Efficiency: An International Comparison”. Public Choice. 123 (3): 321-347. Afonso, A., Schuknecht. L. and Tanzi, V. 2010a. “Public Sector Efficiency: Evidence for New EU Member States and Emerging Markets”. Applied Economics. 42 (17): 2147-2164 Bruns, B. and Luque, J. 2015. “Great Teachers: How to raise student learning in Latin America and the Caribbean” The World Bank Group: Washington, DC. Bruns, Barbara, & Luque, Javier. 2014. Building Better Teachers in Latin America and the Caribbean. World Bank. Caja Costarricense del Seguro Social (CCSS). 2014. Construyendo una agenda estratégica para el fortalecimiento de la equidad y la sostenibilidad del seguro de salud. Propuesta Programa FESSS. San José, Costa Rica. CEPAL 2012. Sistemas de protección social en América Latina y el Caribe: Costa Rica. Comisión Económica para América Latina y el Caribe (CEPAL) Chetty et al. 2014. “How does your kindergarten classroom affect your earnings? EVIDENCE FROM PROJECT STAR. Department of Health Statistics, CCSS. 2011. DESAF (2014). Informe de Gestión 2010- 2014. Estado de la Nación (2013). Cuarto Informe Estado de la Educación. San José, Costa Rica: Programa Estado de la Nación. Hidalgo and Romero, 2013. Evaluación de Impacto de Transferencias Monetarias Condicionadas en Costa Rica: Programa Avancemos. UNIVERSIDAD DE COSTA RICA FACULTAD DE CIENCIAS ECONÓMICAS ESCUELA DE ECONOMÍA IMAS. (2015). Estrategia Nacional para la Reducción de la Pobreza Extrema. Período 2015- 2018. San Jose, Costa Rica: Instituto Mixto de Ayuda Social. IMF. (2013). Costa Rica 2012 Article IV Consultation. IMF Country Report No. 13/79. International Monetary Fund (IMF). 94 Costa Rica Social Sector Expenditure and Institutional Review International Conference on Population and Development (ICPD). 2012. “Country Implementation Profiles: Costa Rica”. Jiménez, R. (2014). Educación pública en Costa Rica: políticas, resultados y gasto. Análisis Serie 6. San Jose, Costa Rica: Academia de Centroamérica. Ministerio de Desarrollo Humano e Inlcusion Social. (2015). Plan Nacional de Reducción para la Pobreza Extrema. San Jose, Costa Rica: Ministerio de Desarrollo Humano e Inlcusion Social. Ministerio de Hacienda (2013), “En Ruta hacia la Consolidación Fiscal: Agenda para un Diálogo Nacional”. Patrinos, Harry Anthony & Velez, Eduardo & Wang, Catherine Yan, 2013. "Framework for the reform of education systems and planning for quality," Policy Research Working Paper Series 6701, The World Bank. Ribe, H., D. A. Robalino, and I. Walker. 2010. From Right to Reality: Achieving Effective Social Rofman, R., Apella, I., & Vezza, E. (2015). Beyond Contributory Pensions : Fourteen Experiences with Coverage Expansion in Latin America. Washington, DC: World Bank. Thomas S. Dee & James Wyckoff, 2015. "Incentives, Selection, and Teacher Performance: Evidence from IMPACT," Journal of Policy Analysis and Management, John Wiley & Sons, Ltd., vol. 34(2), pages 267-297, 0 United Nations, Department of Economic and Social Affairs, Population Division (2013) “World Population Prospects: The 2012 Revision, Volume II, Demographic Profiles” (ST/ESA/SER.A/345). World Bank (2015), “Costa Rica: Systematic Country Diagnosis World Bank. (2008). Costa Rica Public Expenditure Review Enhancing the Effi ciency of Expenditures. World Bank. World Bank. (2012). Opening the Door to Good Quality Jobs: The Role of Human Capital Investment and Social Protection Policies Costa Rica. Report No. 54925-CR. World Bank (draft). World Bank. (n.d.). Central America Poverty Brief. Chapter 2: Costa Rica 2000-2010: The Evolution of Poverty and Equity. World Bank. World Bank. 2014. HNP Statistics and Core Indicators. Washington, DC. World Health Organization (WHO). 2013. Country Basic Data: Costa Rica. 95