Extension of Social Security Coverage, Labour Market and Lessons from Latin America

1. Extension of Social Security Coverage, Labour Market and Lessons from Latin America Carmelo Mesa-Lago ISSA 5th International Research Conference on Social Security: “Social Security and the Labour Market: A Mismatch?” Warsaw, 5-7 March 2007

2. Summary I. Pensions and Healthcare Systems in Latin America II. The Transformation of the Labour Market in Latin America III. Legal and Statistical Coverage of Groups Difficult to Cover • Self-employed • Domestic Servants • Employees of Microenterprises • Agricultural Workers and Peasants IV. Social Security Pension and Health Coverage, Trends and Impact of the Reforms V. Inequalities in Coverage by Income, Gender, Geography and Ethnicity VI. Lessons Learned from Latin America

3. Healthcare Systems in Latin America Most countries in Latin America have three health sectors: • Public, which legally should protect the uninsured population (the majority in 12 countries) but in practice seldom does it, making extremely difficult to estimate the population with access. • Social security sickness-maternity (comprising a principal program and often separate schemes), which covers 41% of the total population ranging from 8% to 88% and is the main provider in eight countries. • Private, the smallest sector that covers 11.5% of the population stretching from 1% to 25%. • Brazil and Cuba don’t have social security but a public national system; and Haitian social security is incipient. • Health reforms began in Chile in 1981 and have been implemented in virtually all countries albeit with different scope, depth and speed; most reform pursue an expansion of the private sector.

4. Pension Systems in Latin America • All 20 countries have social security pensions for old-age, disability and survivors. • Structural reforms have totally or partially privatized pensions (shifting from defined benefit to defined contribution) in ten countries, albeit with three diverse models: Argentina, Bolivia, Chile, Colombia, Costa Rica, Dominican Republic, El Salvador, Mexico, Peru and Uruguay. • The remaining ten countries maintain public systems: Brazil, Cuba, Ecuador, Guatemala, Haiti, Honduras, Nicaragua, Panama, Paraguay and Venezuela, several of them have implemented parametric reforms or are considering them. • The distribution of the labour force in 2004 was: 66% was affiliated to public systems and 34% to private systems.

5. Table 1. Population Groups Difficult to Cover by Social Security in Latin America, 2001-2004 (%)

6. Transformation of the Labour Market and Legal-Statistical Coverage • The Latin American insured formal sector has shrunk, while the uninsured informal sector has expanded and averages 47% of the urban labour force (EAP). • Self-employed: 30% of EAP; legal voluntary coverage or exclusion in 16 countries; actual coverage ranges from 0.2% to 15% (Chile’s coverage in private insurance is 2% but social security provides free affiliation to the poor and subsidies to low income); a great barrier to coverage is a contribution equal to the sum of the percentages paid by salaried workers and employers (Costa Rica had voluntary coverage until 2006 but covered 45% of them because the state subsidizes the low-income self-employed). • Domestic servants: 3-10% of EAP; legal mandatory coverage in 13 countries; actual coverage from 3% to 31% because many of them lack contracts or are unable to denounce evasion or conspire with the employer to evade. • Employees in micro-enterprises: 7% to 16% of EAP (probably underestimated); legally excluded in virtually all countries; coverage in microenterprises is between one- third and one-thirtieth that of large enterprises (in Honduras coverage in microenterprises is 4% versus 46% in large enterprises). • Non-salaried rural labour force: 56% of total rural EAP but exceeds that average in the least developed countries; excluded from legal coverage in half of all countries; actual coverage from 1% to 6% (in Brazil, Ecuador and Mexico coverage increases to 18%-50% of rural EAP because they have special social-security schemes for peasants).

7. Table 2. Social Insurance Coverage of Labour Force by Pensions and Total Population by Health: Circa 2004

8. Social Security Pension and Health Coverage • Pension Coverage of the Labour Force: In the six countries of the group with lowest non-salaried labour (informal and rural sectors) it ranges from 45% to 57% (except Argentina due to the crisis); in the three countries of the medium group from 21% to 28%, and in the nine countries of the group with the highest non-salaried labour force from 9% to 20%. • Average regional coverage was 31% (above the ILO minimum norm of 20% of the labour force): 26% in the ten private systems and 39% in the ten public systems. Coverage decreased in all private systems after the reform; their average fell from 38% before the reform to 26% in 2004. • Health Coverage of Total Population: In the lowest group it ranges from 54% to 87% (except Uruguay; Brazil does not have social security but a public system); in the medium group from 38% to 45%, and in the highest group from 7% to 26%. • Average regional coverage for the region was 41% in 2004 (below the ILO minimum norm of 75% of the resident population) down from 52% in 1990. In 75% of the countries coverage before the reform and in 2004 either stagnated or declined.

9. Inequalities in Coverage • A positive correlation exists between income, degree of development, urban location, male gender and non-indigenous ethnicity on the one hand and social security coverage on the other. • As lower the income, lower the social insurance coverage and vice versa. • Gender inequalities in coverage are caused by external and system factors: in 25% of countries the female spouse is legally covered in maternity but not in sickness or vice versa; statistical coverage of women is lower than man, due to indirect insurance as a spouse dependent of the male insured; direct access declines by women exit from the labour force to raise children. • The best-covered geographic areas are the most developed, urbanized and wealthier whereas the worst covered are the least developed, rural and poor; only two countries have compensation funds to reduce geographic inequalities. • Indigenous populations are largely excluded from coverage because they are poor or have low-income, work in the informal sector and/or live in rural areas (the three countries that have special social security programs for rural and peasants have been able to cover part of them).

10. Lessons Learned from Latin America Factors that Determine Coverage • Low and stagnant/declining coverage is shaped by: (1) the labour-market transformation, (2) the lack of adaptation of social security to such change in most countries, (3) the failure of the reforms to design systems capable to expand coverage, largely due to the exclusion of informal-rural workers and the poor, and (4) external factors. • As higher the proportional size of informal-rural groups in the labour force, more difficult to cover them, but some countries have implemented successful inclusion policies, mainly but not exclusively the most developed. • Social security programs were originally designed for urban formal workers, with stable jobs and medium-high salary (most of them males), but the majority of the labour force in the region is informal (and rural in the least developed), with unstable jobs and low irregular wages (especially among women), hence coverage is very difficult to extend. • Segmented systems without coordination, weak regulation and poor solidarity are typical to all countries with low coverage; in the least developed, social security was introduced late, some have not expanded coverage to all geographic areas, and several have large rural and indigenous populations difficult to incorporate. Conversely, integrated systems (Costa Rica, Chile, Panama) had a positive impact on coverage, particularly if they grant free coverage to the poor and subsidies to low income. • External factors in addition to labour-market changes are: underdevelopment, elevated poverty incidence, high unemployment-underemployment, political instability or crisis, lack of government commitment, cultural and ethnic barriers, gender inequalities, large rural population and regions poorly developed, scarce fiscal resources and taxing capacity.

11. Policies • Expansion of coverage. Social security must adapt to the transformation of the labour market, expanding coverage to informal and rural workers and peasants. • Incorporation of difficult groups. Countries that have legal obligatory affiliation have higher coverage than those with voluntary affiliation but legal inclusion by itself is insufficient: fiscal subsidies to the low-income self-employed and similar groups provide incentives for affiliation (Costa Rica); special regimes for rural workers or peasants get better coverage than those with voluntary or restricted coverage but with diverse results due to diverse government commitment and financial support (better in Brazil than Mexico and Ecuador). • Reduction of inequalities require: integration or good coordination between the public and social security sectors in segmented health systems to eliminate overlapping, save resources, and extend coverage to those excluded; fiscal subsidies to poor or low-income groups (as in Chile) or to guarantee the basic package in the entire system regardless of age, gender and risk (as in Colombia and Argentina); financing proportional to income or progressive as income increases; elimination of fiscal subsidies awarded to non-poor (free raiders); extension of legal integral health coverage to the spouse of the insured, granting optional insurance to housewives; and placing priority to coverage of indigenous peoples targeting the geographic areas where they live.