OECD Conference, Paris, May 19th 21 Session 3: Improving income support and redistribution Ian Walker Lead Economist, Social Protection, LAC Region, The World Bank www.oecd.org/els/social/inequality/emergingeconomies Conceptual framework: anti poverty programs in the context of SP systems The use of targeted subsidies to complement limited contributory social insurance coverage Typology of programs and targeting approaches Targeting and poverty reduction outcomes Downsides and limitations: program quality differentiation, savings effects and labor market behavioral effects Future challenges: Strengthening work opportunities Incentivizing savings Strengthening safety nets and human capital development 1
Social Protection Objectives CONSUMPTION SMOOTHING POVERTY PREVENTION HUMAN CAPITAL DEVELOPMENT INSTRUMENTS Savings/risk pooling Transfers / redistribution Active policies FINANCING ARRANGEMENTS Workers contributions Payroll taxes General revenues Earmarked taxes INSTITUTIONS Governance Monitoring and evaluation Public/private sector INDIVIDUALS BEHAVIOR FIRMS BEHAVIOR PROVIDERS BEHAVIOR PUBLIC SPENDING Formal / informal Job search effort Retirement Job switching Formal / informal Job creation Job destruction Service quality Service costs Service coordination Fiscal sustainability Allocative efficiency Coverage (active members / labor force) to 25 25 to 5 5 to 75 75 to 1 2
Worldwide Comparison of Social Assistance and Social Insurance Spending 14 12 1 8 6 4 2 3.1 1.5 Sub- Saharan Africa 2.6.9 East Asia and Pasific 2 8 Easter Europe and Central Asia 1.3 3.8 Latin America and Caribbean 3.6 3.9 1.4 Middle South Asia East and North Africa 2.5 13.2 OECD SA SI Source: Weigand and Grosh (28) Social pensions Subsidised health insurance Conditional Cash Transfers Unconditional Cash Transfers In-kind transfers (mostly food-based programs) Public works programs / employment guarantees Social investment funds Common goal: provide cash or in-kind support to the poorest households to alleviate poverty 3
4 Types of targeting Cash transfers and targeted social insurance Poverty targeted CCT, CT, health insurance Self targeted workfare Community targeted (eg SIFs) Categorical targeting: (child support, social pensions) In kind transfers Food programs (school, other) Food for work 8 Many MIC programs combine geographic and household targeting to determine eligibility Priority areas selected using geographic targeting (micro-area poverty maps) often based on censal data 6 1 3 4 8 5 9 2 7 CHIAPAS. ZONAS DE OPERACION 4 8 Kilometers S N E W Zonas de Operación I CENTRO COPAINALA (MEZCALAPA) I CENTRO TUXTLA II ALTOS TZELTAL II ALTOS TZOTZIL III FRONTERIZA COMALAPA III FRONTERIZA COMITAN IV FRAILESCA IX ISTMO COSTAS V NORTE BOCHIL V NORTE PICHUCALCO VI SELVA OCOSINGO VI SELVA PALENQUE VI SELVA YAJALON VII SIERRA VIII SOCONUSCO FRONTERIZA VIII SOCONUSCO MAPASTEPEC VIII SOCONUSCO TAPACHULA Zonas de Atención â S N E W 4 8 Kilometers ZONA II ALTOS TZOTZIL SEDE 1 2 Kilo me te rs S N E W Distrito Federal Delegación Xochimilco XOC HIMILC O TLAHU AC M uni ci pal Lim ite C olo nia Hogares en pobreza de capacidades - 2 3-7 8-1 7 18-44 45-12 3 Simbología R ura l Localidad es U rba na Sem i u rb ana
Selecting households in qualified areas Some programs select all demographically qualified h/holds (eg with children under 15) Others also apply poverty or means tests Proxy means-testing (most LAC countries, Turkey, etc.): Eligibility based on a weighted index of characteristics (score) that are easily observed and hard to manipulate Multi-dimensional notion of poverty Fewer labor disincentives But requires agency credibility and can be gamed Means-testing (eg Brazil) More transparent for the beneficiaries But verification of income is a challenge Incentive problems: labor supply / informality effects 9 Larger role given to communities (Bangladesh, Cambodia, Kenya, etc.) In many places communities are asked to gather the information used Often quite specific guidelines given so that it is something of a loose PMT Occasionally more latitude given to communities to decide own critieria 1 5
Demographic: eg children move out of age category CCTs: Non-compliance with conditions Poverty status: Recertification cycle to verify continued poverty (PMT or declared income Links to social insurance and other databases: informality Workfare: time limits on individual and on program 11 World s largest CCT: > 12 million families, 25 of population, operates nationwide,.4 of GDP All municipalities participate, Social Assistance office (CRAS) registers families self-declared per-capita income in cadaster Social Development Ministry (MDS) administers national cadaster and determines program admission and exclusion 2 cut-offs for benefit entitlement: extreme poor: <US4/person /month and poor:<us8/person (about 1/3 of min wage) Extreme poor get bigger benefit (av. 1 of pre-transfer income). MDS uses national h/hold and censal data to estimate eligible households in each municipality and sets a fuzzy budget quota Municipalities that are soft checking incomes can end up with a queue of apparently qualified but unfunded beneficiaries Beneficiaries must update Cadaster declaration every two years Beneficiary turnover is low but in 21, 5 of beneficiaries nationwide dropped for not updating records. 6
Proportion of total benefits received 2-May-21 9 8 7 6 5 4 3 Bolsa Familia Chile Solidario Chile SUF Ecuador BDH Honduras PRAF Mexico Oportunidades Jamaica PATH Cambodia: JFPR Bangladesh FSSP 2 1 1 2 3 4 5 6 7 8 9 1 Deciles of per capita consumption minus transfer 1 8 6 4 2 Pension Coverage by Income quintiles Bolivia 1 8 6 4 2 Chile Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Costa Rica 1 8 6 4 Non-contributory pensions Both contributory and noncontributory pensions 2 Contributory pensions Q1 Q2 Q3 Q4 Q5 7
1 8 6 4 2 Health Insurance Coverage by Income Decile Colombia Peru 1 8 6 4 2 D1 D2 D3 D4 D5 D6 D7 D8 D9 D1 D1 D2 D3 D4 D5 D6 D7 D8 D9 D1 1 8 Chile Non-contributory Health Insurance 6 4 2 D1 D2 D3 D4 D5 D6 D7 D8 D9 D1 Both Contributory and Noncontributory Health Insurance Contributory Health Insurance CCT impacts on consumption and poverty Median per capita consumption (US ) Average transfer ( of per capita consumption) Impact on per capita consumption () Impact: headcount index ( points) Impact: sqd. poverty gap ( points) Mexico (1999) Nicaragua (22) Colombia (26) Cambodia (27).66.52 1.19.75 2 3 13 3 8 21 1 -- 1.3** 5.3** 2.9** -- 3.4** 8.6** 2.2** -- 8
Health: Targeting success sometimes depends on quality differentiation, creating two classes of system Differentiated mandates of subsidized social health insurance or national health systems Often subsidized services have poor quality and severe queuing Pensions: Social pensions have to be small to avoid undermining savings incentives and limit the fiscal cost (eg Bolivia s Renta Dignidad) Programs targeted with Proxy Means Tests have modest or no reductions in adult labor market participation / work effort for (eg CCTs in Mexico, Ecuador, Cambodia). Universal entitlements (for given demographic groups such as children or old people) will not normally affect work incentives but have higher costs for a given reduction in poverty or inequality, unless the targeted demographic group is uniformly poor. CT programs that are targeted based on declared income, without work requirements, are likely to have informality effects undermining job quality and possibly, growth (the Levy hypothesis) Workfare programs where benefits are set too high which they often are - can displace private employment and have very high per capita program costs, limiting coverage. 9
Urban areas / moderate poor / some savings capacity: Incentivize secondary education, training, work search Subsidize employment Use subsidy to promote savings and social insurance participation by informal sector workers Tapered withdrawal of subsidy to avoid poverty traps Packages (such as free health plus access to subsidized pensions savings) are an attractive option Rural areas / extreme poor / no savings capacity: Strengthen targeted safety nets with significant benefits Income effects can be more important than price effects even in CCTs Human capital accumulation for the poor depends on service quality and access, not just on stimulating demand Strengthen links from income support to other systems (education, health and nutrition; other SP programs) Priorities, intervention mixes and implementation speed will differ across countries and regions What can be done depends on what exists Make sure that each step fits within a long-term vision for the SP system: Universal Integrated and coherent Where redistribution is transparent and progressive Where incentives and opportunities to save and to work and human capital accumulation are systematically reinforced, not undermined 1
11