All meetings Tuesdays, 17.15-18.45 at LSE. The first three are in Room 
 S.600; the last is in the Graham Wallas Room. 
 9 MAY: Rosemary Thorp (Queen Elizabeth House, Oxford), 'Poverty over the 
 twentieth century: changing facts and approaches in Latin America'. 
 23 MAY: Colin Lewis (LSE) and Peter Lloyd-Sherlock (East Anglia), 'Social 
 insurance "reforms" in Brazil and the Argentine: lessons from past 

 13 JUNE: Kent Deng (LSE), 'Poverty under Mao's regime, 1949-78: 
 4 JULY: Jérôme Destombes (LSE), 'Nutrition and poverty in northern Ghana, 
 Financial support is provided by the Suntory and Toyota International 
 Centres for Economics and Related Disciplines. 


Paper Summary - 23/v/2000

Social Insurance Regimes in the Argentine and Brazil, c.1900-1999: social policy `reform' and some lessons from history

Colin M. Lewis and Peter Lloyd-Sherlock



Poverty alleviation or political co-option; protecting the economically and social disadvantaged or capturing resources: the debate about social policy has never been ideology free, nor has social policy provision been economically and politically neutral. Yet some proponents of current social policy initiatives in Latin America advance arguments of equity and neutrality. This paper will test these assertions by focusing on pension `reform' in the Argentine and Brazil, locating present debates about the need for change - and the origin of pension funds `crisis' - within an historical context. The paper opens with a survey of the principal issues that have shaped the reform agenda and comments on the relevance of the two country case-studies. It then examines pension regimes. Particular attention is given to (i) the financing of pensions and events and processes that affected fund solvency (ii) the debate about public/private provision and (iii) the implications of past problems for current initiatives.


Perceptions of Crisis and Models of Reform

Social security `reform' (essentially pension reform) is now firmly on the political agenda and likely to remain so. It is regarded as crucial to the viability of neo-liberal projects of structural adjustment. All-too-frequently, social security reform is presented as a component of fiscal reform, capital market deepening and financial sector de-regulation, rarely as a topic of intrinsic interest. Yet social security systems can play a critical role in compensating for the social costs of economic reform. There are now several models of reform: (i) a competitive regime in which private and state funds vie for business; (ii) a complementary arrangement where state cajas provide a basic - usually very basic - pension and the private funds a top-up; (iii) a residual/transitional arrangement where, as in Mexico, the state maintains a public system solely for those currently enroled, workers now entering the labour market being provided for by the private sector. Only Chile has evolved a fully-funded capitalisation (capitación) system in which the final pension paid to an individual is determined by the investment income generated from premiums paid by the insured during his or her working life into a specific account. Most of the previous (and un-reformed) systems provided a flat rate pension with ad hoc provision for a capitación supplement (or for the award of an additional pension on a case-by-case basis).

Although the rhetoric of reform is dominated by several sets of assumptions and ideologies, all start from the premise that the previous system has failed, both as an agent of social welfare and of economic development. Structural problems attributed to the old systems include low coverage (only a relatively small proportion of the economically active population was enroled), a failure of delivery (driven by growing insolvency) and lack of fairness (amongst groups and between generations). Proponents of the neo-liberal model also stress the economic and financial failure of the old system. Namely, the inadequacy of an under-funded pay-as-you-go system which, increasingly dependent on state hand-outs, became a source of fiscal instability. Others perceive failure in terms of a lack of equity. The old, contributory arrangements, which were fragmented and highly stratified, did not address the requirements of the needy and simply replicated existing social imbalances. All critics accept the necessity of de-politicising pension administrations. Partiality, evasion, delinquency and over-blown, expensive bureaucratised administrations typified pre-reform systems.

Wide-ranging adverse criticism has, unsurprisingly, resulted in differing expectations of reform. Neo-liberals (or neo-conservatives) emphasise the social and economic benefits of `choice', `responsibility' and `individual empowerment'. These are the sentiments that characterise the social philosophy under-pinning the broader model of development. In macroeconomic terms, reform is presented as strengthening national capital markets and increasing domestic savings rates - both vital to the consolidations of the new development paradigm. State retreat and the de-politicisation of the administration of social insurance constitute the political dimension of the neo-liberal discourse about pension privatisation. Those concerned about equity and outreach, while advocating the continuing role of the state and the importance of harmonisation (of benefits), accept the need for greater efficiency and resource flexibility in order to secure the delivery of retirement and invalidity pensions capable of sustaining recipients with dignity. Hence, from all quarters, the language of reform is about administrative efficiency and pension sustainability within the context of macroeconomic stability. It is also accepted that regulation - of the private sector - and transparency - in the public - are equally important to the viability and vitality of any new arrangement. It is not easy to reconcile all these objectives - the economic, social and political - in the short-term. Large segments of the population remain uninsured and many groups cannot afford to take out the new private pensions now available.

It is useful to assess whether the problems addressed by the recent wave of reforms are recent or of long-standing. Some proponents of reform draw attention to conjunctural factors, such as the sudden acceleration of population ageing and the sustained fiscal impact of the 1980s debt crisis. These are certainly contributory factors but are not necessarily the sum of systemic difficulties intrinsic to many social insurance models. Also, it is useful to establish whether the basic concepts and rhetoric of reform are new or whether they can be traced back to previous efforts to stabilise social insurance regimes. This can only be done by careful historical study of the development of pension systems of particular countries.

Coverage Growth: ideology, benefits and administration

Ideas informing social security and welfare policy have shifted more than once over the last hundred years. Private provision, largely administered by charities and mutual societies, was the norm approximately a century ago. Subsequently, Bismarkian precepts gained ground and influenced the emergence of contributory social insurance for particular groups of workers. Later still, social insurance was subsumed within larger developmentalist projects - the politics and economics of `stabilising development'. More recently, state retreat and privatisation can be depicted as a return to assumptions of the past. Has the ideological wheel come full circle? There are, of course, differences. A much larger proportion of the economically active workforce is now covered by social insurance than that envisaged in the late nineteenth century and present projects pretend to universality.

Tables II and III, which detail the growth in state social insurance, confirm that the proliferation of individual social insurance funds did not lead to the enrolment of a large proportion of the economically active population before the mid-1940s. Early Peronist administrations saw a sudden extension of coverage in the Argentine. Official figures record that the number of affiliates rose from 2,771,000 to 4,892,000 between 1944 and 1954. But the value of benefits and quality of services may be questioned. The initial impact of increased insurance affiliation on general welfare was, at best, patchy and as inflation took hold, the real value of pensions lagged, notwithstanding periodic adjustments. By the 1980s they only covered a small fraction of the basic income needs of older people. There was a similar up-grading of public welfare institutions in Brazil during the corporatist Vargas period. Social insurance coverage grew steadily in the 1930s, increasing from around 142,500 individuals to almost 190,000 affiliates between 1930 and 1932. With the shift towards institutos from 1937, there was a marked surge in membership. Table II shows that in 1930 less than one percent of the economically active population was socially insured in Brazil. In 1940 this figure was 8.7 percent. By 1947 approximately 3,000,000 Brazilians were enroled in social insurance schemes. As in the Argentine, the inclusion of affiliates was still very selective and benefits varied greatly over time and from occupational group to occupational group. Moreover, the Bismarkian ethos that under-wrote coverage growth for much of the period further limited outreach, as captured in the different coverage rates displayed in Tables II and III. The proportion of the population of retirement age in receipt of a state pension was consistently lower than the proportion of the economically active population enroled in state insurance funds. The differences are more structural than demographic.

Through the 1950s and 1960s there was a steady growth in coverage in the Argentine, though much less in the case of Brazil. Again as Table II shows, there was an unprecedented rise in pension coverage in Brazil during the 1970s: by 1980 it had overtaken the Argentine with 87 percent of the workforce protected. This was largely due to the establishment of a non-contributory scheme for rural workers, the Fundo de Assistência ao Trabalhador Rural (FUNRURAL) without which it is estimated that only around half of workers would have social security entitlements. Less than a quarter of the total population was covered until the rural scheme was set up in the 1970s.

FUNRURAL had been established in 1963. It had a very limited role, providing only basic medical protection. It did not pay pensions. In part this was due to the level of funding. In the mid-1960s, the financial base was extended to include a 2.5 percent wage tax on urban enterprises. This entailed a substantial redistribution from the urban to the rural sector and permitted FUNRURAL to offer affiliates retirement pensions. As implied, FUNRURAL differed from other social insurance funds in several respects. Not only did it provide medical assistance before becoming responsible for pensions, it broke with the tradition of tripartite funding - contributions by beneficiaries, employers and the state. In the Argentine, no such arrangement for rural workers was forth-coming at this stage and, as demonstrated in Table II, total coverage levels have never exceeded 70 percent of the economically-active population.

The experiences of both countries demonstrate the difficulty of creating a universal social insurance programme purely based on Bismarkian, contributory principles. This point is rarely made in the current literature about pension reform.

The pace and nature of coverage extension in both countries suggests that the approach to social security development ad hoc. Figure I illustrates organisational and administrative changes resulting from ad hoc growth.

*** INSERT Fig. I HERE ***

Fund proliferation (detailed in Table I and summarised in Figure I) during the inter-war decades momentarily triggered a debate about the creation of a complete, national systems of social insurance. Later still, the concept of social security - protecting all workers against a wide range of social risks - took root, and was manifest in a substantial growth of the proportion of the workforce covered by social insurance (with some health provision) in the 1940s and 1960s. There was, too, a meritocratic element to the debate about social insurance/welfare coverage in the 1940s - namely the evolution of the concept of productive citizenship. This was most clearly articulated by the Collective of Peronist Intellectuals which observed that a modern state should provide a range of welfare services, including retirement pensions: it was equally the duty of the able-bodied to work to support the state. A feature common to both countries was that `welfare rights' - in essence social insurance and social security - were enshrined in the 1934 `Vargas' Constitution in Brazil and the 1949 Argentinian constitución peronista. Although both Constitutions were later overturned, constitutional reform in the 1990s resulted in the re-instatement of `welfare' clauses, that is, inclusive coverage as a right of citizenship.

Fund proliferation, combined with an emphasis on greater state action in the social and economic spheres, also explain the drive for administrative reform. Figure I confirms that the first major restructuring of the administration of the Argentinian pension system occurred in 1944 with the creation of the Instituto Nacional de Previsión Social (INPS). Set up as an autonomous state agency, the Institute was charged with managing existing funds. It was also granted the authority to redistribute finances amongst the different occupational schemes. The separate pension funds were, however, allowed to continue in existence and this prevented the INPS from exercising many of its powers of co-ordination and control. Approximately twenty years later there was another re-structuring of pension fund administration. The Social Welfare Ministry (Ministerio de Bienestar Social) was created in 1966 and three years later all twelve, non-military social insurance funds were amalgamated in three, decentralised, semi-autonomous `super' cajas - (i) civil servants, (ii) workers and employees in industry, commerce and other activities, and (iii) the self-employed. At the same time, there was some attempt to harmonise contributions and benefits and to streamline administration. Since the late 1980s there has been a change in the language of debate - from the imperative of centralised, co-ordinated administration to the need for greater flexibility and autonomy at pensions fund level. This reflects wider concerns about de-centralisation and the reform of the state, and conforms with attempts to depoliticise and professionalise sections of the government machine.

In Brazil, with the onset of the authoritarian estado nôvo (1937-45), the system was similarly reorganised and centralised. In 1937 the number of Caxias de Aposentadorias e Pensões (CAPs) was reduced to 104. Some ten years later there were only 30 caixas and five larger agencies knows as institutos, which were organised on a sectoral basis. The rapid early growth in the number of Brazilian agencies is explained by the fact that, before 1937, social insurance was principally organised on a company basis. Thereafter, the system was structured along occupational lines broadly as in the Argentine. Sectoral institutos were formed for newly incorporated groups of workers and gradually company caixas were absorbed by institutos or company pension funds catering for groups of workers in related fields were consolidated first into caxias unicas and later `super' institutos. An instituto was created for bank employees in 1934, for industrial workers in 1938, and for transport workers in 1939. Between 1963 and 1971 there was much discussion about harmonising benefits amongst occupational categories and a number of attempts to extend coverage to rural workers and in 1972 to domestic servants.

During the 1960s and 1970s Brazilian social insurance was subsumed within a larger policy framework of state provision of public goods and social services - education, healthcare, social assistance, social security and housing. There was, however, much ambiguity in official policy following the 1964 military coup. Within the context of orthodox stabilisation measures, some members of the regime were enthusiastic advocates of returning worker accident and social insurance to the private sector, reactivating a business ravaged by rapid inflation during the early 1960s. Others pressed for a larger government role along with nationalist regulation of insurance companies. The Ministerio da Previdência Social was established in 1964 and in 1967 the Instituto Nacional de Previdência Social (INPS) absorbed the six existing Institutos de Aposentadorias e Pensões. As with the formation of the Ministry of Labour, Industry and Commerce (charged with the supervision of pensions funds) in 1930, although articulated in terms of harmonisation and administrative efficiency, the under-lying motives were co-optive and meritocratic. Further administrative change occurred in 1974, with the formation of the Ministry of Social Insurance and Social Assistance. Some three years later, the National Agency for Social Insurance and Assistance was established within the ministry to take over the functions of the INPS. Subsequently, and reflecting the increasing politicisation of worker pension and healthcare administration, responsibility for pensions was transferred back and forth between the Labour and Health ministries. Most recently, as elsewhere, the debate has shifted back to private pension provision within the context of financial market de-regulation and a larger role for commercial insurance companies.

Coverage growth and associated administrative re-organisations have been explained in a number of ways. Yet debate continues as to whether or not these changes were driven from `above' or `below'. There is similar disagreement about the extent to which ideological and fiscal factors triggered the spreading of the social insurance safety net. Although Mesa-Lago and Malloy present the growth of the system in largely political terms, financial, economic, fiscal and demographic factors all played a part. Lewis maintains that periodic increases in coverage were as much motivated by fiscal considerations. `Generational' financial crises in the 1940s and the 1960s, explain coverage extension at that point in the Argentine and in Brazil in the 1960s and 1980s. Administrative reorganisations of the 1960s, designed to ensure greater central control, were as much about revenue generation as about the integration of social and economic policy. Committed to rapid growth and low inflation, post-1968/9 Brazilian military administrations favoured institutional consolidation and a massive expansion of the social insurance regime in order to expand mechanisms of forced savings. The language of the period also echoed that of earlier concepts of the productive citizen: social rights had to be earned and were awarded by the military regime to compliant groups who did not clamour for political rights. Similarly, if for less `developmental' purposes, the Peronist regime of the late 1940s used the proceeds of sales of Crédito Argentino bonds to social insurance funds to cover the operating deficit of state corporations. Hence, coverage extension served a dual objective. Bringing in new groups of contributors generated extra fiscal income as well as covering deficits in existing funds.

Nevertheless, Mesa-Lago and others continue to argue forcefully that the most powerful labour unions and various privileged groups were able to impose their demands on the state. Conversely, the fact that establishment of the railway caja in the Argentine and railway caixas in Brazil followed particularly bitter, disruptive strikes is interpreted by yet others as indicating that the state itself took the initiative and sought to co-opt strategic sections of the labour movement. (In Brazil, state action may also have been prompted by the widespread abuses and financial mismanagement of the private and mutual sectors. Menicucci maintains that Article 20 of the Civil Code of 1916 specifically brought the funds under state regulation in order to remedy past irregularities.) Malloy certainly depicts social insurance programmes as resulting from state inspiration rather than worker pressure. He draws attention to the relative small size and political weakness of the working classes in cities such as São Paulo. Co-optive social insurance programmes became part of the ideology of modernisation espoused by successive Brazilian regimes, not least the estado nôvo and post-1964 military administrations. Other studies, such as Araujo de Oliveira and Fleury Teixeira, refute this argument, claiming that the implementation of insurance schemes was the consequence of increasing divisions between traditional and more modern elite factions. Privileged sections of urban labour were able to forge alliances with progressive elite factions to force through insurance legislation.


Looking to the Future from the Past

What insights does an analysis of past problems of social insurance regimes in the Argentine and Brazil offer in terms of the future of pension provision? First, comparison of social insurance indicates that differences were more pronounced than apparent similarities. While, the state came to play an increasing role in the delivery of social insurance in both countries until the 1980s, Brazil never experienced the state monopoly that characterised the Argentinian system. The longer history of private pension funding has major implications for the retreat of the state but does not necessarily imply that the politics of privatisation will be less problematic in Brazil than in the Argentine. This also has implications for the administration of the system. Paradoxically, in Brazil the public sector was much more centralised than the Argentinian, where administrative responsibility initially lay in the hands of workers themselves (that is, mutual societies) and was subsequently devolved to trade unions. Brazilian schemes also differed in that many funds assumed an early responsibility for the delivery of medical services and, in the case of FUNRURAL, the principle of worker contributions was breached. In addition, very early Brazilian arrangements differed from the Argentinian in that they were organised on a company, not sectoral, basis. Yet, despite these differences, by the 1980s pensions funds in both countries were experiencing similar problems.

The second major finding of the research concerns the viability of capitalisation systems. While several factors account for the failure of many of social insurances funds to generate sustainable surpluses, `reform' of the system in the 1940s and the 1960s was triggered by both a financial/fiscal and an accumulationist agenda which prefigure concerns of the 1980s and 1990s. It is this that makes the problems of past capitación systems relevant. There are, of course, differences between the 1940s and 1960s, and the 1980s and 1990s. Previously, solutions were conceived in terms of greater centralisation and the incorporation of new occupational cohorts to generate `savings income' for the state: current remedies emphasise de-centralisation and the investment/allocation of pension fund income by (and in) the market, which is perceived as a more efficient mechanism than the state. Beyond dispute is the fact that capitación systems exprienced `generational' funding crises and resulted the formation of rent-seeking alliances that benefited affiliates at the expense of other social groups.

If a capitalisation system is to be re-instituted, the ability of the system to deliver will depend to a very large extent on the effective use of premium income. Here, the history of the early social insurance cajas/caixas indicates what should not be done. Namely, factors alluded to above such as the channelling of very large proportions of fund income into low-yielding government bonds and the immobilising of a substantial part of income in mortgage advances. Problems of shallow domestic capital markets and narrow investment opportunities remain. Although the limits were subsequently liberalised, Chilean private pensions funds were initially obliged to invest in a restricted series of domestic bonds. Currently, Chilean pension providers are allowed to invest up to 40 percent of premium income in equities. The same maximum is permitted for holding of public bonds. Argentinian funds may hold up to 50 percent of assets in equities, up to 50 percent in federal government bonds (and an additional 30 percent in provincial and municipal bonds) but not more than 10 percent in foreign government bonds or 10 percent in foreign corporation. This compares with an actual investment in equities of almost 70 percent by British pension funds. Moreover, governments are anxious to maximise opportunities for domestic borrowing. The lesson of the recent history of private pensions funds in Chile and the experience of the original Brazilian and Argentinian caixas and cajas is that (subject to the usual requirements of security), the ability of private funds to maximise incomes depends on a liberal schedule of permissible areas of investment.

The third main finding derives from the second, namely that problems encountered by the Argentinian and Brazilian social insurance regimes in the 1980s and 1990s have been present for much of the century. These problems may be categorised under various headings, such as, inadequate funding, ineffective use of resources and administrative `inefficiency'. Some of these problems were specific to the Bismarkian, capitalisation arrangement, others resulted from the almost inevitable lurch towards pay-as-you-go systems. These problems were identified before the 1920s yet continue to confront late twentieth century reformers. Present protagonists who argue that change - either the switch from state to market arrangements or parallel, competitive private and public systems - will ensure efficient administration and a productive use of premium income to generate a stream of future resources sufficient to meet the needs of beneficiaries, ignore the lessons of history. Several problems must be addressed: there must be a closer match between benefits and contributions; the payment of contributions must be rigorously policed; contributions and the period over which contributions need to be made are likely to be substantial in order to yield a reasonable pension.

Fourth, a `realism' gap remains. The rhetoric of pensions reform in the two countries is cast in terms of realistic expectations, not least as regards funding and benefits. Over the last fifty years there has also been a considerable increase in rates of contributions. This did not `close' the benefits/contributions gap. Rather, evasion and delinquency grew. Does this account for current efforts to reduce contributions? An important strand in the current debate about labour `flexibilisation' is the reduction in the employer quota, currently standing at around 16 percent (20 percent in Brazil). This either means a reduction in global contributions or passing to employees a larger share of the total social insurance quota. If total contributions are reduced, will the funds realised at the end of a worker's active life yield sufficient resources to provide an adequate pension? If not, who will cover the cost? Even if viable funding arrangements can be established for workers currently entering the workforce, or existing workers transferring to new private pension schemes, the legacy of past delinquency and evasion remains and the government it still responsible for existing pensions. Like the Chilean government, the Menem administration in the Argentine discovered that encouraging workers to move from the public to private pension providers transfers current fund income to private pension companies but not existing obligations. In the medium-term, pension privatisation compounds financial imbalance in the public system: rising obligations to retired members have to be met from reduced resources. The response, in the Argentine, has been to cap existing pensions (a mechanism that has been challenged in the courts) and a promise to honour future contribution requirements. Will the state prove more virtuous in the future than in the past? Recent experience is not positive.

Today, pension privatisation is being actively pursued. Assessments of the outcome of the process will be determined by the principal objective of `reform'. Are the changes designed to achieve social/welfare, administrative or larger macroeconomic objectives? Capitalisation schemes, as the example of the past - and the current experience of Chile - indicates, can generate substantial amounts of investment income, particularly in the early life of funds. However, as the history of early funds in the two countries also shows, generous pension entitlements, limited investment opportunities and rising administrative charges can rapidly induced financial crisis and a systemic change from capitalisation to pay-as-you-go. When the previous systems in Brazil and the Argentine experienced generational financial crises - as the mass of `early' contributors became pensioners and deficits loomed - they were bailed out by an extension of the system and by state aid and `nationalisation'. New groups of contributors were brought into the system, permitting cross-subsidisation. If the new schemes are designed to be universal and stand-alone, `historic solutions' - successive sectional extensions of coverage - cannot apply.

All other things being equal, fund financial viability will depended on contributions - premium levels and regularity of payment - and investment income. To date, most of the private funds have managed to secure regular payments but, again, the lessons of the past suggests caution. Affiliates usually experience greatest difficulty sustaining contributions during an economic downturn and funds seen to be under-performing can loose income. Notwithstanding the recent currency crisis, these are two areas in which the private sector is still largely untested, though there is historic evidence of member resistance to increases in contributions. And, as in the past, not all workers are in a position to make contributions. If the objective is fund financial stability and equity, without a continuing state presence, there is little prospect of realising either. The new system will be as exclusive as the old.


In sum, with a decline in the rate of inflation, the accumulationist objective of pension re-organisation is being realised even in the short run. And administrative changes must almost certainly yield positive results. It is equally obvious that the privatisation of pension provision will not deliver the social objectives of equity and that for some time previously under-funded provision means that residual state pension schemes will remain a significant factor in fiscal instability. Those arguing the case for change must learn from history: deficits are not a new phenomenon, nor is the concept of pluralist administration novel. History shows that there are no easy solutions to resolving deficits. The success or otherwise of private pension administration depends on the relationship between the public and private sectors. Are they to be complementary or competitive? The `public-private mix' in the two countries has been quite different and the ideological underpinnings of social security system have changed dramatically over time. How much longer will the theory of the market dominate social policy thinking in Latin America. At the beginning of the twentieth century, social insurance provision was atomised and largely unregulated. By the middle of the century, centralisation and greater state involvement were thought to offer equity and efficiency. Pension sustainability, delivery and `dignity' are now promised by pluralistic, competitive systems. Basic differences between Brazil and the Argentine, despite apparent similarities in the evolution of social insurance regimes, cautions against universalist solutions - the `one-size-fits-all approach' implicit in much of the policy recommendations emanating from international agencies. Those now entering the labour market, and falling within the purview of `reformed' social insurance regimes must look to economic growth and effective fund performance for their future pension needs.






The Argentine



































Notes: * In 1980 only 50.0% of the Brazilian economically active population were contributing. The FUNRURAL scheme for agricultural workers was non-contributory.

Sources: The Argentine - E. Isuani & J. San Martino La reforma previsional argentina: opiniones y riesgos (Buenos Aires 1993) p.; International Labour OrganisationUse of uninitialized value in concatenation (.) or string at E:\listplex\SYSTEM\SCRIPTS\filearea.cgi line 455, line 307. The Role of Social Security and Improved Living and Working Standards in Social and Economic Development (Geneva 1966) p.33; C. Mesa-Lago `Social Security in Latin America' in Inter-American Development Bank Economic and Social Progress in Latin America: 1991 Report (Washington, D.C. 1991) p.186.

Brazil - Ludwig Brazil: a handbook of historical statistics (Boston 1985) p.; C. Mesa-Lago `Social Security in Latin America' in Inter-American Development Bank Economic and Social Progress in Latin America: 1991 Report (Washington, D.C. 1991) p.186.






The Argentine



















Notes: Multiple pension holding means that the data exaggerate the true level of coverage. For example, in 1990, Isuani and San Martino estimate that 14 per cent of Argentinian beneficiaries held more than one pension (E. Isuani & J. San Martino La Reforma Previsional Argentina: opiniones y riesgos (Buenos Aires 1993) p.



Sources: The Argentine - P. Lloyd-Sherlock xxx (1998) p. International Bank for Reconstruction and Development Averting the Old Age Crisis: policies to protect the old and promote growth (Oxford 1994) p.

Brazil - Ludwig Brazil: a handbook of historical statistics (Boston 1985); C. Mesa-Lago Social Security in Latin America in Inter-American Development Bank Economic and Social Progress in Latin America: 1991 Report (Washington, D.C. 1991) p.186.