Status and Challenges of Social Protection Policies in Latin America

By ENRIQUE VASQUEZ AND ENRIQUE MENDIZABAL

People without social insurance in Latin American countries include large numbers of unemployed created during economic reforms of the 1990s, the growing work force in the informal sector, and other poor and vulnerable groups such as women and the elderly. This article assesses various types of socail protection, insurance and pension programs which operate in the region, weighs their effects on the groups which face the greatest social risk, and seeks ways to provide "spring boards,"not just "safety nets." The authors-both researching social policy and human development with the Universidad del Pacifico in Lima, Peru-are Enrique Vasquez, Professor of Economics, and Enrique Mendizabal, Economist.

SOCIAL RISK AND POVERTY

When people need help in preventing or dealing with social riskand economic vulnerability and cannot obtain that help from market mechanisms, government programs such as unemployment insurance, health insurance, income maintenance, or pensions attempt to fill the gap. Social insurace "necessarily involves redistribution from taxpayers in general to people at social risk in particular. "a new broader definition is that social protection is "human-capital-oriented public interventions to assist individuals, house-holds, and communiites better manage risk , and to provide support to the incapacitated poor"
In Latin America, widespreda social exclusion occurred as a result of the structural adjustment and market-economy policies introduced during the 1990s. Mostly designed by organizations like the World Bank and International Monetary Fund, the main concern of these policies was macroeconomic and fiscal stability. To compensate for the social cost of these economic reforms, Latin American governments embarked on a wide ragne of state interventions to protect the most vulnerable. In the end, neither the economic nor the social policies worked out properly. Goverments failed to provide social security systems that could eliminate the social risk and economic vulnerability which accompanied the shocks so common in the region.
The current state of social protection in most Latin American countries resembles that of European countries before the Second World War. But this does not mean that Latin America should take the same path as Europe did after the war. Instead, it should learn from its own painful experience and develop social protection policies suited to its needs and the problems of the most vulnerable in the region: poor workers in the informal sector , adolescents, children and the elderly.
Social expenditure in the Latin American region is less than 10 per cent of GDP, while in industrial countries it ranges from 16 per cent in the United States to 30 per cent or more in continental Eupope. In fact, this comparison understates the gap, because most Latin American countries include educationin their expenditure on social protection;industrial countries do not , since they already guarantee education access for the great majority of the school -app pop- ulation as a human -capital investment. Services are far from universal for Latin Americans(Ganza, Leon and Sanuma, UNDP-UNICEF,1999). Furthermore, public agencies incharge of implementing social policies have very limited effectiveness and efficiency. In some cases like Peru, the basic problem is not only limited resources, but also and most important, the quality of the social management (Vasquez, 2000)
In industrial countries, as many as five categories of risks are covered by social protection systems: old age, health, employment, income and family. The first two risks are covered by social security instruments which are more or less close "substitutes to market insurance", while unemployment, income and family risks are covered by pure social insurance.
Old age insurance and pension systems, which are compulsory in industrial ocuntries, have become extremely complex,. Debate centers on whether pension systems should remain pay-as -you-go, because proportionately smaller working populations in industrial countries make it increasingly difficult to coverthe elderly whose populationshare is growing. In Latin America, on the other hand, debate has centered on whether pension systems should be privately or publicly funded, as wellas whether they should remain pay-as -you-go or move to individual accounts.
Health insurance. Industrial countries have made considerable progress in the postwar period in establishing quast -uni-versal health insurance schemes. There is no consensus onwhether they should be a public or private responsibility, and it is generally accepted that their performance is not satisfactory. The United States remains the only industrialcountry without some kind of national health insurance system.
Rapidly increasing health care expenditures have led some countries to seek better control of the sometimes excessive prices charged by service providers and to assure patients' rights to service. Thos means a two-tier price structure, with hagher tariffs for those able to pay, and reduced tariffs of free service for lowincome families or individuals, but in no case is treatment denied. Health maintenance organizations(HMOs) have become a popular private model, not only in industrial countries lick the United States, but also in Latin America. However, this is not a solution for countries with high levels of informal-sector employment, which has no provision for an HMO health insurance plan. The private option should focus on helping to increase coverage for those currently uncovered and no just compete for those already covered.
Unemployment insurance systems, as experience shows, are incomplete unless they are associated with low-income social insurance to reduce the risk of individual or family income fallingbelow an unacceptable level.. otherwise, some proportion of individuals with no job or noaccess to the formal labor sector could not claim any unemployment benefits. With the rise of unemployment in Latin American countries during the 1990s, this happened frequently. In particular, young workers without prior experience found it almost impossible to get a becent wage and adequate job. Hence, "in some instances, minimum income guarantee schemes have to be set up precisely to respond to this kind of situation". This form of safety net -unemployment and low-income insurance working together-has proven very effective.
Income maintenance programs, on the other hand, have generated a different kind o fdebate. Ensuring a strict minimum income level might, in some circumstances, "significantly reduce labor incentives" and result in a 'poverty trap'. If the benefits are relatively higher than a formal wage, labor becomes more expensive and individuals may prefer to remainunemployed-or be employed at wages below the legal minimum, thus benefiting from the program. This negative byproduct of a well-intentioned program has become an economic burden for States, which have enacted various reforms to avoid it . some of these include earned-income tax credits in the United States and United Kingdom, where individuals who work a minimumu number of hours and remain below some gross income limit receive a negative income ta as an incentive. This increases the participation rate of the target population without increasing the income risk. And efficient tax and revenue system is required to ensure its sustainability and transparency.
Social protection systems affect both the supply and demand sides of the labor market, since" social protection contributes to an increase in the cost of labor and therefore to a fall in employment." This is especially relevant in countries with a minimum wage policy where the contribution necessarily comes from the employer, while in countries with no minimum-wage policy the contribution comes from employees, thus reducing their net salaries. Since 1990, the European Commission's recommended policy has been that social security contributions from low wages should be reduced "so as to increase the demand for unskilled workers" and thus reduce upemployment among those at higher social risk -poor, unskilled individuals.
Reaons for economic nisecurity in Latin America
Economic insecurity creates social risk and the need for social protection. It is important to understand the origin and causes of economic insecurity in latin America in order to assess different ways of managing it . three of its causes are the trauma of the 1980s, high macroeconomic volatility, and inadequate response by social and political institutions.
The 11980s debt crsis created a deep depression in Latin America, leaving the region with high unemployment rates, weak social insurance institutions and weak social insurance institutions and weak publicy provided safety nets. Real income declined around 20 per cent in countries like Argentina, Chile, Mexico and Venezuela, while in Brazil it dropped only 10 per cent. On the other hand, Peru's decline was about 35 per cent, comparable to the United States in the great depression of the 1930s. Comparing the 1930s great depression and the 1980s debt crisis helps to explain the rise of economic insecurity in Latin America. Both events produced a negative perception of free market instruments for social security among the most vulnerable individuals. For the U.S.in the 1930s, the next logical step was the creation of the "welfare State" with strong public social insurance institutions. latin America in the 1980s, however, lost confidence not only in its private sector as a prime suspect in causing the debt crisis, but also and mainly in an out-of-control public sector.
The response at the time was a leap into the free market. Most countries in the region had externally imposed trade restrictions, public enterprises, fiscal deficits and microeconomic mismanagement. Instead, governments thereafter adopted policies aimed at free trade and financial liberalization, privatization of publicly run companies, and deregulation -for which most of the weakly industrialized economies of the region were no prepared. The result: the complete annihilation of the Pervuian midle class, high devaluation risks in Ecuador, Brazil and Argentina, high unemployment rates in Brazil, financial and economic crisis in Mexico, Brazil and Peru, and macroeconomic volatility in the region.
Entering the 1990s, Latin America found itself in, unprepared for, a highly volatile environment driven mainly by erratic capital flows. Newly opened economies left the region's macroeconomic policy and growth potential dependent on the "fancies of short-term foreign investors ". most governments shifted efforts towards attacting foreign capital to fill the void left by capital sources which had fled. Eventually, therir macroeconomic policy divorced itself from the real economy, social expenditure suffered the most in relative terms, social risk and vulnerability increased, and so did economic insecurity.
Weak social and political institutions were responsible for, and further weakened by economic insecurity in the region. The late 1980s and early 1990s served as killing fields for democratic institutions, labor unions and politica parties. Trade unions were unable to develop strategies for dealing with job insecurity, both caused by and contributing to their loss of members and political power. Ironicallly, the very institutions which should have responded to the basic social needs of the population, but could not respond, found their constituencies so weakened that there was almost no protest from the people most affected.
Furthermore, the "washington consensus view of development policy, sharply constrained by the 'requirements' of global economic integration, has prevented the emergence of an alternative (of at least complementary) vivsion of economic reform driven by local concerns and national aspirations." These policy 'guidelines', 'recommended'by international economic and financial agents, drove most countries to a stalemate in economic growth , which they now have to face, along with rising unemployment rates, few international rewseves, high exchange rate volatility, and an unavoidable dependence on foreign capial.
Dealing with social risks
How do governments and individuals respond to the rise of economic insecurity and social risks? More important , when should governments respond, and when should they leave individuals and the private sector to deal with risks? How do individuals or families behave when confronted with risk?
Three approaches to social insurance by individuals have been identified by Gill and Ilahe(2000):market insurance (i.e. unemployment insurance), selfinsurance (i.e.precautionary savings) and self-protection (i.e. investment in human capital ). Arguably, precaution is preferred and there fore social insurance, provided by the public sector, should concentrate on mechanisms for self-protection. For example , when market insurance does not exist for certain risks or does not cover the poor and extreme pooe, the government should step in to provide it . unemployment insurance, for instance, is offered by governments as a substitute for market insurance, for instance, is offered by governments as a substitute fo rmarket insurance, which does not guarantee ndividuals a stable wage. The United States uses employer experience ratings and /or indi-vidual unemployment history to fix unemployment insurance premiums.
Self-insurance, on the other hand, implies that the individual seeks instruments or assets that will maintain or increase their value in the event of a negative shock. However, individuals may not have the necessary information to make the right decision or lack access to good instruments, such as diversified financial assets. As a result , they may self-insure using bad instruments, such as land, which can lose value after a negative financial shock. In this case, the government could step in to promote the development of stronger financial markets, for example, prudentially regulating the the capital market in cluding banking deposits, deposit insurance, non-bank financial intermediaries-or forstering the market for long-term public bonds.
Finally, self-protection involves the use of some type of asset that will lower the probability of risk. Investment in human capital, i.e. education, reduces the probability of unemployment. Therefore, government policies to improve educational access and quality will be and important component of xocial protectionand may help to substantially reduce the risk of future economic downturns. However, this approach provides poor collateral since individuals cannot borrow against their knowledge or skills in case they lose jobs. Because no market exists for human capital, private agents are induced to seek a more easily collateralized form of capiatal . in latin America, this form of social protection is usually the only possible one for the poor and the unemployed who have no access to government programs or fully-funded pension systems based on earmarked payroll taxes.
Education and health
Arguably, the most important human capital initiatives are educationand health. For developing countries, improved educational access is important as a strategy for eradicating poverty across generations, especially in rural areas, which are shor on proper educational services. When families cannot reduce their social risk through education for children, they seek to enhance their protectionby breeding more children as a source of chea labor for economic slumps and as a "househod safely net "for the parents'old age.
Health care should be universally covered by social insurance. If resources are scarce, increased health couverage should be given priority, even over pension benefits, which in any case typically go in large share toward health care for the retired elderly. Fortunately, public health insurance is still the most important way of financing health services for the population; HMOs and private insurance are only accessible to those with the means and who have formal employment. Workers in the informal sector, for instance, would not have access to HMO coverage. In the latin American region, public health insurance is typically provided by at least two health care systems:"one provided by social insurance, which mainly covers the salaried labor force, pensioners and family dependants of both; and another by the Ministry of Health, which offers public health care to the non-insured poor and low-income population". Only Cuba has a universal standardized sys-tem, offered by the Ministry of Health.
Private health insurance is provided by either voluntary contributions(which are extremely low in the region)or ear-marked payroll deductions. Hence, private insurance is only accessible to "formal workers" and wealthy individuals who can voluntarily contribute towards a healthe insurance plan. Few alternatives for reducing or dealing with healthe risks are available to workers who are not part of the formal workforce and to their families and dependants.
Some general policy recommendations are derived from this:
First, it is clear that the availability of all three forms of insurance instruments will improve individual and family welfare, thus reducing social risk and economic insecurity.
Second, demand for social insurance does not decrease with economic development and improvement of welfare, nor is the clamor fo rbetter social insur-ance "an indication of worsening economic circumstances". On the contrary, as societies develop, they require more and better public goods and services, including social insurance.
Third, financial market strengthening should be considered the central element of any social policy since it can augment the availability of the three forms of insurance. Hence, governments ought not to divorce their social protection policy from macroeconomic and microeconomic policies.
Fourth , governments should seek new and better alternatives for the most vulnerable people, whose sole social protection mechanisms are based on the family, the community and their own personal capital accumulationmechanisms.
Even when fiscal adjustments are crafted so that the poor experience minimum effects from cuts, there is no way to know the extent of such effects in advance. In Latin American countries it is usually nonsocial spending that is protected from cuts. As a powerlessminority, the poor have no political influence over the allocation of public spendingand their share of the benefits. Evidence shows that public spending on social protection is procyclical: duringrecessions, when there are more poor people, less is spent on them. Brown and Hunter (1999) found that Latin American democracies are more likely to protect social spending during a recession, while authoritarian regimes tend to reduce social spending during a crisis andexpand it afterwards .
Fiscal adjustments in the region have not been in the best in terests of the poor. This suggests that the safety nets supposedly provided by social spending did not live up to expectations. Specific public insurance programs, well-targeted social funds and poverty relidf efforts have therefore thaken the spotlight in social policy recommendations for social protection in the region. Among the different options put forward in the region should be mentioned social funds and targeted conditional transfer programs, nad unemployment and health insurance.
Social funds
While social funds vary from country to country, some characteristics remain constant. Among the most important operational characteristics they share are that social funds:
? Promote, finace and implement small social projects, but do not participate in the maintenance or operation of such projects.
? Help establish the formal parameters to support investments benefiting the poor.
? Enjoy some kind of operational autonomy, even though they are part of the public sector.
? In a highly centralized government, however, such autonomy might be subordinated to the central government's own policy desires.
In 1987, the world's first social fund was created in Bolivia-Fondo Social de Emergencia, since then, their number has increased ocnsiderably and, today, almost all countries in the Latin American region have one. All of them sharem certain objectives and trends from which various lessons may be obtained.
The main objectives of most social funds include improving the infrastructure for the provision of social services, such as sanitation, roads, bridges, schools and health centers. In the absence of an emergency-such as a natural disaster -unemployment is typically the prime objective of social funds in the initial stage . Other objectives include:community development , commonly reachedby a "learning-by -doing" process where the social funds finance projects promoted, managed and implemented by the communities; improved delivery of basic social services, with emphasis on training service providers; and support for decentralization of social protection -as is the case for social funds in Chile, Honduras and Bolivia.
Among the most important general trends observed in social funds worldwide is that they are becoming more integrated into countries' overall social and economic efforts. In the absence of strong social and political institutions, as in most Latin American countries, this might become a force against the development of an efficient welfare state. As social funds pay more attention to popular participation, the individuals and families who benefit tend to ask for more social services- such as health and nutuition survices, occupational education, etc. Peruvian experience in FONCODES has shown that popular participation has positive impact on the results of the projects implemented by the fund. Participation of beneficiaries helps correct mistakes in the desing and focus of projects and make them more effective,implying that the poor are better at deciding what is best for them.
Some other trends include increasing decentralization of social funds and demand for more imcome-generating subprojects. These trends arise not from the search fo rbetter management and operation of social funds, but from the wider politica and economic environment which favors, for example, more popular participationand more emphasis on microfinance in many deviloping countries.
Some lessons about what works and does not work in sociaa funds should be carefully considered when Latin American governments are formulating social policy. In general, they have done well inadjusting rap9idly to changing circumstances, promoting community participation to enhance sustainability, and targeting poor communities -though more effort is needed with the poorest nad marginalized groups. Compared to other public sector agencies, they have done well in achievingcost efficiency and accountability and in generating trust among communities toward the public sector-though in some countries corruption and public funds embezzlement is common.
Not all lessons have been positive. The most important deficiencies include poor performance in microcredit programs; poor beneficiary rates and gaps in coverage for the "poorest of the poor";inability to integrate with collaborative programs inother public institutions, so as to improve efficiency; incosistency with the social policy framework, which confuses beneficiaries and detracts fro capacity buklding; inadequate impact indicators; and inability to provide assistace with mass impact, especially with regard to unemployment. For example, the Peruvian social fund, FONCODES, has experienced mixed outcomes in education, and very high degrees of leakage (50 per cent) and gaps in coverage (68 per cent) in nutrition programmes.
Even so , social funds in Latin America have proven to be a cost-efficient way of relieving social risk and economic insecurity, with few exceptions. While their deficiencies can be easily corrected some critics mainti=ain that social funds do not accomplish what is expected, do not reach the poor, build infrastructure that is not staffed or maintained, do not produce long term impact, etc. While this is true for some coumtries in the region, their true potential should not be overlooked. If well administered, social funds could represent one of the main sources of social insurance for the poor and vulnerable. If well targeted, social funds could reduce social risk by buildingthe capacity to manage risk.
Beyond social funds: a universal social protection system?
While social funds are indeed good safety nets, the poor and the unemployed need more than that: they need a "springboard" out of their risky and vulnerable situation. One such springboard in the long term, for instance, could be high-quality education, which breaks the chain of poverty carried generation after generation. The uneducated poor represent one of strongest problems that all social protection systems facein the region. Poor education levels may be the main cause of low income, lack of access to social protection services, and lack of political power to press for government intervention.
Beyond the safety net myopia of social funds, expanded coverage is the big issue. A good social protection system would have the potential to bring in previously exclded groups, such as the unemployed, women, the elderly, people with disabilities and workers in the informal sector. Do social funds have the right toools to identify and incorporate those who remain uncovered/ evidence tends to suggest that a better alternative is needed.
Social security in industrial countries is universal, and employment and contribution to a fully funded system are not required for inclusion. These pay -as-you-go (PAYG) systems obtain their funds from general revenue and do not need criteria to establish eligibility for benefits. However, in Latin America, universal programs hav emany political and practical difficulties.
First, they are fiscally unsustainable. Individuals have many incentives to avoid paying and ge "a free ride " in the use of services. Since universality means using general revenues to finance social protectionwith uniform benefits regardless of income, people who contribute more are likely to oppose such a redistributive system-hence the shift towards HMOs and private insurance. Even when fully funded systems are fiscally sustainable and include efficiency standards in their practice, they have the "side effcet of creating a class of uninsured".
This happens when insurance companies, seeking high returns, try to reduce risk and avoid insuring the elderly, women in reproductive ages and disabled individuals whose levels of dependency are extremely high. This practice, known as "cream skimming", implies that in the absence of universal public insurance only the rich would be protected. In Chile, a poorly regulated public/private mix with the stete acting as 'lender of last resort' ended in a dual model: private insurers covered the yound and rich, while the state took care of the other 65 per cent of the population.
Private market insurance is an opeion only for those who have the means to afford it and toose who are given no choice by their employees. HMOs, for instance, coverthe entire workforce of a given institution, giving employees no option but to accept the payroll deduction as a contribution that entitles them to health benefits provided by the insuance company . according to many physicians, this practice damages the quality of the health care service provided. It is also out of reach for those employed in the informal sector or selfemployed, whose only alternatives are public insurance, selr-insurance and self-protection. In the Latin American context, this means that uncovered individuals and families will seek self-insurance and self-protection inside the traditional social protection system -their families and communities.
Social protection must therefore be led by the public sector and complemented by private initiatives that help increase coverage , not only in social insurance (protecting against negative events), but also in wider social protection. The government must administer resources and programs that provide poor and vulnerable families with a minimum decent standard of living. One of the most innovative programs implemented in the region to satisfy these basic needs are the targeted conditional transfer (TCT) programs.
Targeted conditional transfer (TCT) programs
Basically, TCT programs in Latin America involve cash grants to poor families with young children on the condition that they keep their children on the condition that they keep their children in school and, in some cases, visit health centers. They represent a shift in social policy away from subsidies (mainly food and fuel) to income transfer programs targeted to the poor that require certain conditions from the beneficiaries. TCT programs do not cover workers in the informal sector, seasonal and self-employed workers, though they could cover the unemployed if carefully designed to avoid creating incentives to remain unemployed. Representative programs in the region include Bolsa Escola in urban Brazil, PETI in rural Brazil, and Progresa in Mexico, the first being a decentralized institution, while the others are managed by the federal government.
These TCT programs have five main objectives. First, to achieve an increase in educational attainments and substantial health improvements for young school-age children, which would reduce future poverty. Second, to reduce shortterm poverty by targeting grants to the poor. Third, to decrease child labor by requiring a minimum school attednance by children in beneficiary households as an explicit or implicit objective. Fourth, to provide income support to poor families acting as a safety net against adverse shocks that could throw them deeper into poverty. Finally, to achieve integral development of child education and health by providing supply-side financial support to schools and health centers. These programs are badly needed in Latin America. They address educational, health and nutrition indicators which are particularly low for the poor in the region, especially for children, the most vulnerable group. They bring together supply and demand in a rational way so that, for example, the number of schoolrooms is increased in line with hegher school attendance. Political acceptance, support and sustainability are high, though fiscal sustainability, on the other hand, is sadly not secure.
The effectiveness and success of TCT programs can be measured by looking at three of their main objectives:
(1) targeting the poor. At the design stage, the poorest regions of the country are selected in federally administered programs, and decentralized programmes target the municipalities where the conditions for their interest groups- the poor, rural child workers, etc. -are the worsst. Actual selection of beneficiaries, however, differs substantially . in urban Brazil, Bolsa Escola rates various aspects of the beneficiaries' living standards, while PETI in rural Brazil looks for per capita income lower than half the minimum wage and for working childrenof school age (7-14). Kprogresa in Mexico requires beneficiaries to have total family income less than the cost of a standard food basket and uses a specific index to score the household's characteristics. Others like Honduras' PRAF-BID II simply include all families in the chosen poorest municipalities.
Successful outcomes of Bolsa Escola hav ebeen attributed in part to the rating system. Other less successful programs, mostly the decentralized ones, have been unable to serveas mech as 43 per cent of their potential beneficiaries, mainly for financial reasons. Centralized programs like Progresa, on the other hand, show low gaps in coverage rates, 14-16 per cent, but high leakage rates, of about 35 to 38 per cent.
(2) Increasig human capital. The programs have had successful outcomes in both education and health/nutrition. Substantial positive ffects in both fields were observed among the beneficiaries of the Progresa program in Mexico. Bolsa Escola had success in bolstering school attendance, but its program does not include the health component.
(3) Reducing child labor. These results were not as positive. Progresa reduced paid child labor by 25 per cent among children aged 12-13, but the results fo rchildren aged 14-15 were substantially negative; in fact, the incidence of paid child labor significantly increased.
From these experiences, it is possible to put forward some policy recommendations for TCT programs in Latin America.
First, involvement by municipalities helps achieve efficient operations. Programs should be decentralized, with a local share in funding as well as federal or central government financing.
Second, TCT programs should not be seen as substitutes for social funds, other social protection policies or education related interventions.
Third, it is important to determine in advance a fair and clear grant amount that families will receive. The question is whether to award grant amount that families will receive. The question is whether to award grants on a per child or per family basis. Grants on a per family basis means that families with more children may not satisfactorily increase their income and reduce their social risk. On the other hand, grants on a per child basis could be an incentive to have more children, which would increase both the total costs of household social services and the social risk.
Fourth, TCT programs could be designed to work as safety nets to those living inpoverty and high unemployment risk.
Finally, all programs should provide explicit financial support to guarantee a minimum service quality, thus augmenting the role of the "supply side."
However, the key question in the entire debate on social protection reform is : are TCT programs fiscally sustainable in the long run? To succeed, they need long term commitments by public authorities; that would require political as well as fiscal stability, which are far from possible in the region. Seekingfiscal stability through loans or aid from the World Bank or International Monetary Fund means accepting economic, trade and fiscal policies that are beyond the scope of social policy. Funding is thus a crucial issue to resolve before it is safe to consider TCT programs as alternatives to a universal public insurance system.
Social risk mangement: an option?
The World Bank has introduced the concept o fsocial risk management (SRM) into its defintion of social protection. SRM involves using prevention strategies, mitigation strategies and coping strategies to "manage" risk and thus decrease economic insecurity. Prevention and mitigation strategies are to be put into practice before a negative shock occurs. Prevention strategies include economic and macroeconomic policy, public health policy, and educationand training strategies, among others. Mitigation strategies, on the other hand, seek to protect against variations in income -for example, by diversifying its sources, and by providing informal or formal insurance. Social funds play na important role in providing such insurance to the poor and vulnerable. Finally, coping strategies seek to reduce the impact of a shock once it has occurred. For example, an individual could relieve the situation by drawing on savings, unless a lifetime of poverty has progibited an accumulationof assets that could offset a sudden loss of income.
By no means an alternative to social protection, SRM can help individuals manage risk and vulnerability by strategies that avoid the consequences of being uninsured or partially insured.
The instruments used ofreach strategy will vary among individuals and families. These can be informal/personal arrangements, such as marriage, mutual community support and real state assets; formal/market based arrangements, such as financial assets and insurance contracts; and formal/ publicly mandated or provided arrangements, such as rules and regulations, protection of property rights, social insurance, transfers and public works. The following table summarizes the interaction between strategies and arrangements under the concept of SRM and the mechanisms available for individuals to deal with risk.
As shown in the table above, the absence of market-based and publicly provided social insurance has a negative social effect. Risk coping, which by far is the most popular way of dealing with risk, induces households to send children to work, thus augmenting their vulnerability and economic insecurity. The implication is that filling this void should have high priority for the government.
The aim of the SRM approach is to change households and governments from the usual coping strategy they adopted in response to shocks, economic insecurity and probable risk, towards a risk management strategy that prevents or at least mitigates against negative effects. Preventing problems is preferable to coping with them, and cheaper, but until now social funds have mainly contributed to coping. Turning those funds in the direction of SRM is similar to designing TCT programs which try to reduce the probability of future risk and to mitigate it in the mid term. Using the SRM approach, social funds should target according to vulnerability and not only poverty. The ones who suffer the most during economic crises and budget cuts are not only the poor, but also those above the poverty line who have no permanent income, are unskilled, have large young families and lack access to market based insurance. Furthermore, there should be more community -participation in projects to develop organizational capacity, conduct training, offer legal assistance, provide preventive health services, and generate income.
Another look at the table shows that several personal and market based methods of risk coping leave individuals and families in worse situations than before. A family's vulnerability is increased if it sends its children to work, sells real or financial assets, or borrows money from banks or neighbors. Public instruments in this case only alleviate the problem in the short term, serving as a safety net when they should act as a springboard. Under consideration in Peru is a solution to this apparent flaw in the SRW scheme-creation of a social protection net, omnipresent and designed to satisfy individual needs and prevent, mitigate and cope with specific risks.
PENSION REFORM
Background
Old-age insurance or pensions seek to secure a minimum standard of living for the elderly after they leave their jobs. Three alternatives for dealing with old-age pensions are : (i) A universal social security system which obtains its funds from general revenue and covers everyone without any qualifying criteria;(ii) a contributory social security system-or fully funded system-which is funded by earmarked payroll taxes as required contributions towards pension benefits in the future; and (iii) resourcesof the family and the community, which remains the most traditional system in the region to deal with risk and social protection.
During the 1990s there was a shift from the first to the second alternative in Latin American countries. This change, initially promoted by Chile, called for the elimination of the publicly run pension system (PPS) to be replaced by a fully funded new private pension system (PrPS). However, most countries in the region did not drop the first alternative and created a multi-pillar system; the first being the public "universal' security system, and the second the private contributory social security system. In many cases, an informal third pillar also exists in the form of private investment in land, homes, tools or education.
However necessary was the reform, the shift towards a private system increased the number of "uninsured and partially insured individual who have contributed little or nothing in their working years and may be left in poverty in old age". This newly created group does no exist in countries where" they have universal flat (uniform) or very broad means tested old-age security benefits". It consists mainly of the following:
n Individuals who worked mainly inside the household (primarily women).
n Individuals with disabilities which have kept them from workin gin the formal sector.
n Individuals who worked mostly in the informal sector where contributions are no tcollected.
n Self-employed individuals. Farmers in particular are very likely to live in poverty or extreme poverty, and have little or no access to social insurance or pension systems.
Furthermore, the new system raises a new issue: the coverage problem which results when private pension providers seek to reduce risk and therefore avoid the most "difficult groups". These groups include, for instance, farmers, self-employed workers and disabled individuals, whose contribution is not mandatory and would require large pension benefits in old age. Instead of developing incentives to promote their participation, private providers engage in "cream skimming," the same practice as in health insurance (disussed above). This consists of leaving the high-risk poor portion of the elderly without insurance and only covering those with higher incomes and lower vulnerability and poverty risks.
This raises the question of whether private insurance concepts contribute to the welfare of the most vulnerable and difficult groups. Most of the Latin American population is unemployed, employed in the informal sector or sel-employed, and not part of the "formal"workforce. Therefore, it is safe to say that moving away from a universal system only increasesthe number of the uninsured. Even if private pensions in all their variations did cover all formally employed workers, they would not provide for those in the informal sector, including women workers in the househole and sporadic owrkers, since they are not contributory programs based on earmaked payroll taxes.
Pension reform in Latin America
"Young or old, Social Security affects eveyone. And virtually everyone believes that it won't be there for future generations like it has been in the past". That belief in itself indicates there are reasons for pension reform. Chile's reform took place over a decade before other Latin American countries and set the standards for changes made in the PPS of Colombia, Mexico and Peru. These reforms, including Chile's, are still very recent, and real results have not yet been observed. It is thus difficult to know how well the new private pension systems are covering the poor and most vulnerable people.
Broadly speaking, during the first half of the 1990s, the PPS in most Latin American countries found themselves on the verge of financial breakdown. In Peru, for instance, hyperinflation and price distortions sent the real value of the system's assets plunging, while simultaneously the overall obligations of the system expanded as a result of the popular policies adopted udring the 1980s. Following Chile's experience, new pension systems were put into practice with three main objectives: first, to establish a solid private pension system that would create indivedual accounts and provide workers a reasonable pension after retirement, directly proportional to their contributions. Second, to generate a considerable investment fund through which the new system would contribute to the devilopment of the financial and capital markets, as well as increase domestic savings. Third, to eliminate inefficiency in the use of resources by the PPS, in which all retired workers receive the same pension regardless of their contributions.
Developing countries, unlike industrial countries, have a large young workforce that could easily finance retirement pensions for the elderly. However, this workforce is mostly unemployed or in the informal sector. Therefore, reform should aim at increasing the number of people contributing to the system by providing incentives such as individual accounts.
Lessons of reforms
Colombia. The Colombian pension reform of 1993 is unlike the reforms in other countries such as Chile or Peru. In Colombia, a system of competiton was established between the existing pay as you go (PAYG) system and a new fully funded system (FFS), similar to the privately run Chilean system. Competition means that affiliates may choose between the FFS and the PAYG, based on their own assessment of expected future returns in either system. To maintain certain stability, affiliates must wait three years to switch between systems and six months within the FFS. In any case, the PAYG system ends up operating as a last resort option for workers who would otherwise take the risk that a higher than average portfolio return implies.
The system of competition immediately increased th econtribution rate and effectively reduced the benefits by delaying retirement age by two years. This reduced fiscal expenditure by about 1 per cent of GDP annually. However, the fiscal costs of the transition are still too high, in part because of additional concessions to public servants, delay in making new retirement conditions effective, and the costs of operating the PAYG system as insurance against a loss in the FFS, which induces high mobility between the systems.
Following the reform, pension contributions increased from 6.5 to 14.5 per cent of earnings. The PAYG system which seems inefficient, and one of the causes of the reform, was kept"basically for those who were risk averse and preferred to have defined benefits as an alternative to a defined benefits as an alternative to a defined contribution system. Thus, the defined benefits system remained public and the defined contribution system, private, although switching is allowed between them."
An important aspect of the reform is that competition between the PAYG system and FFS added to the last resort quality of PAYG and forced the FFS to ensure at least a minimum expected return thus promoting efficiency and caution. Lower than expected returns would not only have considerably negative effects on the private pension administrators, but also on the affiliates and the economy as a whole.
The Colombian pension reform needs a series of "second generation" reforms to address the fiscal burden mentioned before. First, concessions granted to special groups of public servanted to special groups of public servants should be elimiated. Second, retirement two years later for both male and female workers should be made effective sooner than the 20-year period established. Third, the retirement age should be further increased by three years (60/65, female/male), in accordance with the rise in life expectacy. Finally, high payroll taxes need to be substituted by regular taxes to avoid damaging effects on employment and international competitiveness, since the current payroll deductions only create incentives for informal employment.
Mexico. The 1995 Mexican pension reform was prompted by demographic tendencies, PPS financial health, and evasion. Though Mexico, like most developing countries, has the demographic advantage of a large young workforce, the 1990s seemed the right time for reforms to avoid future problems-especially given rapidly decreasing child mortality and birth rates, and increasing life expectance. The strongest reason for reform, however, was the severe financial disequilibria in the public system, which predicteda deficit o faround 141 per cent of GDP in current value (1994)for a horizon of 75 years. Finally, evasion was high mainly because of the lack of proportionality between contributions and benefits, tempting employees to declare lower salaries or cross over to the informal labor sector.
The Mexican pension reform was built around three pillars: a first public run pillar with a redistribution objective; a second private run pillar based upon mandatory contributions capitalized in individual accounts; and a third pillar of voluntary savings. The second pillar involves active management of individual accounts, pension funds and withdrawal mechanisms by private agents. This pillar creates a direct relation between ocntributions and pension benefits, thus eliminating the incentives for evading contributions. On the second pillar rests all the structure of the pension system.
The reformed Mexican system has unique interaction between public and private agents, unlike the Colombian (competition) and Chilean (fully private) systems. It was designed to:
n Augment equity, efficiency and sustainability of the pensionsystem.
n Build the foundations of a financially stable pension system
n Limit the fiscal impact.
n Strengthen the development of the financial system and reduce its volatility by increasing private financial intermediaries and instruments available.
n Increase domestic long term savings and overall aggregate savings.
Peru. During the first half of the 1980s, the Peruvian PPS began to show clear signs of financial weakness. Rising unemployment rates reduced the mumber of contributors, while new retired workers claimed their pension benefits every day . this created a heavy burden on the system. It was not able or not interested to increase returns on pension fundinvestments. It was on the verge of financial breakdown by the start of the 1990s and was draining resources from other areas of the budget. Hyperinflation, low employment and high informal employment prompted its reform. Whether the reform was the right one is not very clear.
As in Colombia and Mexico, the public system created too many incentives for evasion and was unable to produce a strong sustainable fund that would not be affected by demographic, economic and political changes. In 1993, Peru created a private system like Chile's in addition to the public system, with contributors owning individual accounts and receiving pension benefits proportional to their contributons and the return on pension fund investments. However, contributions in the Peruvian system are made entirely by the employee, while in Chile they are shared with the employer.
In 1995, Peru made a series of modifications and "second-generation" reforms in the private system in order to elimiate existig differences between the two systems. Among the most important changeswere measures to promote the affiliation of independent workers, eliminate imperfections in taxation affecting the development of the private system, level the costs in both systems, guarantee a minimum pension benefit (similar to those in Colombia and Mexico), and promote voluntary savings in the private system in 1996, more modifications were made to reduce inefficiency in the systems arising from the high incidence of evasion. The financial crisis and the "El nino"in 1998 helped intensify the existing problems. Also, the backfire by creating insecurity and instability in the system.
One of the most important problems, which the governmenthas not yet addressed, is the high commission that private agents chargeto affiliates. At 3.8 per cent, it is the highest in the region compared with 3.5 per cent in Colom bia, 2.89 per cent in Chile and 2.6 per cent in Uruguay. Whatever the reason, high commissions also create aversion to the system and this contributes toward reduced coverage.
Another limitation is the lack of financial instruments to generate future returns to the pension funds (aside from the ban on investing in foreign assets). This is important because 70 per cent of the private system affiliates are under 35 years of age, meaning that pension benefits will not be paid in the short term. Long term financial instruments or ivestments are rare in Peru, mainly because of political instability.
Another limitation is that the "Classifying Commission" determines which assets and bonds the private agents can and cannot invest in . Present debate centers on proposed reforms in these investment limitations. In view of the current economic recession in the country, private agents will be used to inject fresh long-term investment into the economy and to invest in the real estate and construction industries as part of a government program for housing facilities for poor families.
Human effects of the reform: forgotten again?
Private pension systems have extended all over Latin America. Argentina(1994), Bolivia(1997), Colombia (1993), Chile (19810, El Salvador (1997), Mexico (1995), Peru(1993) and Uruguay (1996) adopted new systems such as the ones discussed in this section. In all cases, privately managed pension funds are the most influential force on the domestic capital markets. However, this entails investment risks, individual longevity risks, and demographic risks.
The investment risks can be systematically decreased by diversifying pension fund portfolios, but inflation and exchange risks cannot be avoided. Individual longevity risks -some individuals require more resources because of unexpected logevity- can bbe managed if affiliates with surplus income voluntarily save more than private pension providers require. The demographic risk is that a group will have different than expected longevity, requirig establishment of State guarantees or reinsurance with domestic or foreign pension funds or insurers. These risk reductions should be coupled with transparent investment strategies, accountability, and clear reform legislation that will last enough to gain the trust of the affiliates, provide incentives forworkers to join the privae system, and broaden it pension fund.
How well are the new private pension systems covering themost vulnerable and poor?unlike the reforms in social insurance, pension reforms have apparently not yet created a new group of uninsured. Since they cover only dependent workers (employees), private pension systems under the reforms have not solved the situation regarding unemployed individuals and those employed in the informal sector, or the needs of women who work at home and make no contributions to any pension system. In 1997, only 49.4 per cent of the economically active population in Lima(Peru's capital city )was employed in the private formal sector. Latin America's large informal sector. Latin America's large informal sector and unemployment rates require a pension system that covers more than half of all workers.
In the year 2000, seven years since Peru's reform, less than 30 per cent of the working population, including 18.9 per cent of the formal workforce, was covered by some sort of pension insurance. It could be argued that, in Peru, the public and private systems compete for the same socially included population, and therefore coverage is low. However, if the PPS was to be eliminated as in the Chilean pension reform, a larger group of uninsured could appear. The informal sector would need to be canalized towards the formal sector and sporadic and seasonal workers would need a special contributory system I order to ensure decent old age benefits.
Reports show that in the entire region, except Chile, not even half the formal labor force has been affiliated to the new systems mandated by the reforms. Not even in Chile are all workers (including those in the informal sector )affiliated to some kind of old-age pension scheme.
A solution seems to be a multi-pillar system. The first pillar, the pps, would ensure retirement benefits for the unemployed by providing temporary low wage jobs in activities of social interest. The PPS would be able to benefit from the management skills and investment port folios of the second pillar, the private pension system, and produce higher returns in its pension funds. Such a soulution may eliminate the competition system and turn it into a "collaboration system" between public and private pension providers, more or less like the reformed Mexican pension structure where both systems interact.
Vulnerable groups
The reforms have not secured coverage for all. Indeed they have created a new class of uninsured among the most vulnerable and poor-mainly informal workers, women and the elderly, but also other groups which are also victims of discrimination such as the disabled, people with AIDS, etc.
The same fiscal unsustainability that prompted the reforms damagedthe formal labor sector inmany Latin American countries. International investment and free trade augmented unemployment and forced many formal workers into the informal sector. Informal workers, under a fully funded system, do not apply for pension benefits when old and are entirely uninsured. Some of these workers are employed on and off in the formal sector and contribute to the private pension system albeit sporadically. Private pension providers promote voluntary contributions, yet where these providers exist, voluntary coverage for self-employed owrkers is the lowest for the region -Chile (11 per cent) , Peru( 4 per cent), and Mexico(1per cent). This is not an alternativeto increase coverage of the self-employed.
Women remain largely uninsured in Latin America. Their work, mainly inside their own households, is not remunerated, and they usually "lack the labor force participation that would entitle them to contributory benefits, and even if covered, they are usually only 'partially' insured because their levels of education, wages and years of service were low". Furthermore, women own little property and hence are highly vulnerable to an old age of poverty.
Undercurrent systems, when marriages break up, most women are left without a decent income or pension, especially when such a break occurs at an older age, although a series of measures could be taken "to extend the implicit family contract into old age".
Finally, the edlerly, whose welfare is the main purpose of pension systems, are one of the most vulnerable groups. Time of service requirements prevent most elderly from seeking better pension benefits in the private system and hence are left in the public one. When pension reform leads to a multi pillar sustem, the elderly may receive public pension benefits. However, if the reform leaves only a private system these elderly contributiors to the former PPS would find themselves unimsured by the private system when it seeks to reduce their client risks.
CONCLUSIONS
A new word should be present in all policy discussions dealing with social protection and pension reform: uninsured. Uninsured persons result from a mixture of causes, including social exclusion, crea skimmin gand permissive incentives for free riding. Poor countries can not sustain large uninsured poulations in the long run because they will eventually require support from old-age or health benefits, no doubt at a cost to taxpayers which drains scarce budget resources.
In designing solutions to this problem countries in the Latin American region should be guided by two basic financial facts:
(1) contributory social security and pension programs are a fiscal necessity.
(2) A large dose of redistribution between the richer and the poorest is inevitable. Any other approach would be financially unsustainable and would not succed in providing social protectionand pension coverage for the poor.
Over time, economic development and increased female participation in the labor force will result in greater social protection coverage. However, to move proactively on increasing coverage, measures are required on a national basis that involve both rich and poor people, and both the public and private sectors:
? To protectthe poor, a national health care system or integrated social insurance health systems should be established.
? For the rich, it should be made possible to obtain additional coverage or betterprotection through different providers , but without fiscal subsidy.
? Private and public sector insurance and pension systems should work together to develop new services and products that would increase coverage. Free market competition is of no use when half the country's population does not have the purchasing power to access insurance.
? When embarking on new social protection or pension reforms, the principal focus naturally has to be placed on achieving increased coverage for the poor and vulnerable populations, and on providing a springboard for all those who are currently excluded, not juxt short term relief and a simple safety net. To those ends, experience in the region leads to some suggested guidelines:
? Coverage should extend not only to workers in the formal sector-market-based and public -but also to informal sector workers and the unemployed.
? Though the current exclusion of the poor and the extreme poor needs urgent attention,rent, food or fuel subsidy programs should be avoeded since they do not build up or strengthen social, economic or political capabilities.
? One approach is to encourage associations or cooperatives of low-income individuals, families and groups which can facilitate, and provide incentives for, affiliation with insurance and pension systems.
? A longer term approach is to give priority to children's education, since in later life that improves income and access to social insurance and protection. This priority is reflected in targeted onditional transfer (TCT)progrmas which provide income support to poor familie on condition that children attend school; however, TCT programs can experience unsustainable costs and serious shortcomings in coverage.
? It is almost impossible for independent workers to contribute voluntarily. Low-income workers have relatively short life expectancy, so it may make more sese for them to use their meager incomes to survive at present rather than saving for a distant future when they may not be alive.
However, social protection not only involves insurance, but also realtes to the range of national programs and projects, which are directed at eradicaitng poverty. It is important that these programs draw on the best available experience ofr their design and the best mix of parners for their execution.
Fiirst, as regards design, only a few reserch centers study social welfare on a regional basis, and more such work is needed to draw on Latin American experience and the lessons learned for the design of new social protection programs. Some cross-country comparative analyses have been done by research netsorks, and some studies have been conducted on health , education, poverty and social security issues. But most studies confine themselvesto in-country research, and a lack of financial resources affects their sustainability.
Second, as regards execution, one of the main ocnclusions from the analysis of social policy in Latin America is that the private sector an dcivil society organizations need to play a more leading role. They should be in charge of executing thosesocial programs where they have competitive advantage, while the State concentrates on monitoring and carrying out impact evaluations. Nong overnmental organizations are big players in designing cost effective social programs, but need to strengthen their relationships with public institutions; the most innovative practices in the region relate to civil society participation in such programs as locally administered community health, nutrition improvement through "popular kitchens" and mothers'clubs, communnity banks for micro finance, and local committees for conservation and the environment.
Mobilizing these research findings andthese broader partnerships could also help strengthen lobbying efforts, develop political support, and attract international cooperation, which can bring about the needed redirection and reinvigoration of social policies. In the final analysis, better resourcing, design and execution of social programs is essential if poverty eradication is to succeed among the indigenous people in Ecuador, the landless in Brazil, women in Peru, and other poor and vulnerable groups in Latin America.
N o t e s
1. bourguignon, F,. Social Security in Industrial Countries: Which lessons for LAC countries?, LAC June, p.7.
2. jorgensen Steen Lau and Julie Van Domelen, Helping th ePoor Manage Risk Better: The Role of Social Funds; p.3 in Lustin , Nora, "Shielding the Poor: Social Protection in the Developing World", Brookings Institution/inter American Development Bank, Washington, D.C., 2000, PP. 10, 15, 24-25.
3. Bourguignon, F., op . cit., pp. 16-17.
4. The 'trap' implies that in the event of a shock, individuals and families who have chosen to reduce their labor hours to benefit from the income maintenance programs, stand in a poosition of hegher social risk than those hold-ing a formal job.Idem. pp. 20-21.
5. rodrik, Dani, Why is there so much economic insecurity in Latin America? Harvard University, Cambridge, 1999, PP, 3-5.
6. Gill, Indermit S. And Nadeem Ilahe, Economic Insecurity, Individual Behaviour and Social Security, The World Bank, Washington, D.C., 2000, P. 1. 15, 18.
7.Conning, Jonathan,Pedro Olinto and Alvaro Trigeros, Managing Economic Insecurity in Rural El Salvador, first draft, 2000, p.23-24, 34.
8. Mesa-Lago Carmelo, Old-Age Security and Health Care for the Poor in Latin America and the Caribbean, Inter-American Development Bank, Washington DC. 1994, P.4
9. Ravallion, Martin, Are the Poor Protected from Budget Cuts? Theory and Evidence for Argentina, The World Bank/ Universite Sciences
Sociales, Toulouse, 2000, P.1
10 Vasquez, Enrique, Enrique Mendizabal and Giovann Alarcon, Oferta y Demanda de Servicios Sociales Focalizados en Ninos y Ninas en el Peru, Save the Children Sweden/CIUP, Lima, 2000.
11 Brown, David S.and Wendy Hunter, Democracy and Social Spending in Latin America, American Political Science Review 93(4): 779-90, 1999.
12 Ravallion, Martin, op.cit.p.3.
13 Jorgensen, Steen Lau Julie Van Domelen, op.cit.,pp.7-12.
14 Paxson, Christina and Norbert R.Schady, Do schools facilities matter? The case of the Peruvian Social Fund (FONCODES), background paper for the Social Funds, 2000 ESW, Washington, D.C., 2000, P. 29,table No.2.
15 Calculated from the "Encuesta Nacinonal de Niveles de Vida:ENNIV 2000" by the authors.
16 Mesa-Lage, Carmelo, op.cit.
17 James, Estelle, op.cit.p.3.
18 Mesa -Lago, Carmelo, op.cit. pp.6-7.
19 Uthoff, Andras, Trends in Social Secutiry Reform and the Uninsured, Inter-American Development Bank, Wahington D.C.M1999,P.7.
20 the world bank lo de springboard p.46.
21 Uthoff, Andras, Trends in Social Security Reform and the Uninsured, Inter-American Development Bank, Wshington D.C., 1999,P.7.
22 Titelman, Daniel and Andras Uthoff, El mercado de la salud y la reforma a su financiamiento, (mimeo) CEPAL, Santiago, 1999.
23 James, Estelle, op. cit .p.2.
24 Sedlacek, Guilherme, Nadeem Ilahi and Emily Gustafsson Wright, Targeted Condition Transfer Programms in Latin America: An Early Survey, The World Bank, Washington DC, 2000, PP. 1-13.
25 Leakage implies that some of the beneficiaries didn't belong to the potential population, thus showing poor targeting.
26 Jorgensen, Steen Lau and Julie Van Domelen, op.cit, p.4-5, 21.
27 James, Estelle, Old-age protection for the uninsured what are the issues? World Bank, Washington D.C., 1999,P.2-8.
28 Although its "universality" is doubtful since it is only aplicable for formal dependant workers and still leaves behind those who participate in the informal sector, are self-employed or remain unemployed.
29 Holzmann, Robert et al, Social Protection Sector Strategy: from Safety Nets to spring board, The World Bank.
30 Uthoff, Andras, Trends in Social Security Reform and the Uninsured, Inter-American Development Bank, Washington D.C.,1999 , P. 5.
31 James, Estelle, op. cit.p.15.
32 Jose Pinera, Former Chilean Labour Minister, credited with privatised Chile's state pension system in 1981, Tuesday, June 18, 1996, Portland.
33 Munoz, Italo, La reforma del sistema privado de pensiones, in Abusada, Roberto, Fritz Du Bois, Eduardo Moron and Jose Valderrama (Editors), "La reforma incompleta: rescatando los noventa", tomo I, CIUP/IPE, Lima, 2000, p.451-452.
34 Clavijo, Sergio, Pension Reform in Colombia: Macroeconomic and Fiscal Effects, Draft for Comments, IMF, 1997, PP.4-7 (table No.1), 28.
35 Administradoras de Fondos de Pensiones -- Pension Fund Administrators--in the Chilean PPS.
36 Cerda, Luis and Gloria Gandolini, Mexico: la reforma al sistema de pensiones, Mexico, D.F, 1997,pp.1-32.
37 Munoz, Italo, op.cit, p. 451-472.
38 Edwards, Gonzalo and Salvador Valdes, Jubilacion en los sistemas pensionados privados , Documento de trabajo de la Pontificia Universidad Catolica de Chile, 1997, pp.1-4.
39 If an individual outlives those in hi/her age group, undistributed returns could be redistributed to coverhis/her need for more resources.
40 Gestion, informe, enero 1999.
41 Moron, Eduardo, Estudio sobre el Sistema de Privado de Pensiones y Administradoras de Fondos de Pensiones, informe final,Lima, 2000, P.2
42 Six years after the reform only 14.3% of the formal workforce was affiliated in Colombia, 29.6 per cent after 3 years in Mexico, and 18.9 per cent after 7 years in Peru. Idem.p.6.
43 Having in mind the risk-adverse strategy that the PPS must hav eto ensure minimum retirement benefit payments to its affiliates.
44 Mesa-Lago, Carmelo, Old-Age Security and Health Care for the Poor in Latin America and the Caribbean, Inter-American Development Bank, Washingon D.C., 1999, P.6.
45 Moron, Eduardo, Estudio sobre el Sistema de Privado de Pensiones y Administradoras de Fondos de Pensiones, informe final, Lima, Lima, 2000,p.3.
46 James, Estelle, op. cit., pp.23-24.

R e f e r e n c e s

Alives, Desinard and Timmins Christopher, Social exclusion and the two-tiered healthcare system of Brasil, LACEA, Montevideo, 2001.

Bourguignin, F.,Social Security in Industrial Countries; Which Lesson for LAC Countries?, LAC , June 2000.

Brown, David S. and Wendy Hunter, Democracy and Social Spending in Latin America, American Political Science Review 93(4): 779-90, 1999.

Cerda, Luis and Gloria Gandolini,Mexico: la reforma al sistema de pensiones, Mexico, D.F,1997.

Clavijo, Sergio, Pension Reform in Colombia: Macroeconomic and Fiscal Effects, draft for comments, IMF, 1997.

Conning, Jonathan, Pedro Olinto and Alvaro Trigeros, Managing Economic Insecurity in Rural El Salvador, first draft, 2000.

Forteza, Alvaro, Electoral Competition and the Unfunding of Public Pension Programs, LACEA, Montevideo, 2001.

Gill, Indermit S.and Nadeem Ilahi, Economic Insecurity, Individual Behaviour and Social Security, The World Bank, Washington, D.C., 2000.

Jack, William, The Evolution of Health Insurance Institutions: Theory and Four Examples from Latin America, The World Bank, Washington, D.C., 2000.

Jack, William, The Evolution of Health Insurance Institutions: Theory and Four Examples from Latin America, The World Bank, Washington, D.C., 2000.

Jorgensen, Steen Lau and Julie Van Domelen, Helping the Poor Manage Risk Better: The Role of Social Funds, in Lusting, Nora, "Shielding the Poor: Social Protection in the Developing World", Brookings Institution/Inter-American Development Bank, Washington, D.C.M 2000.

Moron, Eduardo, Estudio sobre el Sistema de Privado de Pensiones y Adminstradoras de fondos de Pensiones, Informe Final, Lima, 2000.

Munoz, Italo,La reforma del sistema privado de pensiones, in Abusada, Roberto, Fritz Du Bois, Eduardo Moron and Jose Valderama (Editors), " La reforma incompleta: rescatando los noventa", tomo I, CIUP/IPE, Lima, 2000.

Paxson, Christina and Norbert R. Schady, Do schools facilities matter? The case of the Peruvian Social Fund (FONCODES), background paper for the Social Funds, 2000 ESW, Washington, D.C., 2000.

Ravallion, Martin, Are the Poor Protected from Budget Cuts? Theory and Evidence for Argentina, The World Bank/Universite des Sciences Sociales, Toulouse, 2000.

Rius, Andres, Why Do Polititians Tie their Hands? The Bounded Rationality of Policy Making and the Reform of the Uruguayan Pension System, LACEA, Montevideo, 2001.

Rodrik, Dani, Why is there so much economic insecurity in Latin America? Harvard University, Cambridge, 1999.

Sapalli, Caludio and Bernarder Vial, Incentives for What/ Are Moral Hazard and Adverse Selection Quantitatively Important ?The Case of Chile, LACE, Montevideo, 2001.

Sedlacek, Guilherme, Nadeem ilahi and Emily Gustafsson-Wright, Targeted Condition Transfer Programs in Latin America: An Early Survey, The World Bank , Washington, D.C.M,2000.

Vasquez, Enrique, Enrique Mendizabal and Giovann Alarcon, Oferta y Demanda de Servicios Sociales Focalizados en Ninos y Ninas en el Peru, Save the Children Sweden/ CIUP, Lima, 2000.

Woodon, Quentin and Mari Minowa, Training for the Urban Unemployed: A Reevaluation of Mexico's Probecat, preliminary draft, World Bank, Washington, D.C., 1999.
World Bank Institute, Social Funds: Myths and Reality, in " International Conference on Social Funds: Evolving into the New Millennium", World Bank, Washington, D.C.,2000.