Skip to main content Accessibility help

Pension Reform in Latin America: Distributional Principles, Inequalities and Alternative Policy Options*

  • CAMILA ARZA (a1)


During the 1990s a wave of major structural reforms that changed the distributional principles underpinning pension policies spread across Latin America. Outcomes were not always as expected. The implementation of new pension rules in the socio-economic, political and institutional context of Latin America has resulted in a number of inequalities which affect pension system performance and the gains that different income groups and generations may obtain. In order to overcome the distributional drawbacks of reform, Latin American governments may need to afford a new role to non-contributive pensions, as well as consider the application of specific regulatory adjustments to reduce the risks and inequalities involved in the private pillar. Cross-border policy learning may provide useful tools to achieve these aims.

Durante los años 90 se expandió a lo largo de América Latina una ola de importantes reformas estructurales que modificaron los principios distributivos sobre las que se sostenían las políticas hacia las pensiones. Los resultados no fueron siempre los esperados. La implementación de nuevas reglas para las pensiones en el contexto socioeconómico, político e institucional de Latinoamérica ha generado ciertas desigualdades que afectan el desempeño del sistema de pensiones y las ganancias que diferentes generaciones y grupos de ingreso pueden obtener. Con el fin de superar los inconvenientes de distribución en las reformas, los gobiernos latinoamericanos puede que necesiten asumir un nuevo papel hacia las pensiones no-contributivas, así como considerar la aplicación de ajustes reguladores específicos para reducir los riesgos y desigualdades presentes en el ámbito privado. Una política de aprendizaje entre los países de la región puede proveer de herramientas útiles para alcanzar estos objetivos.

Palabras clave: Pensiones, reforma a las pensiones, desigualdad, distribución, beneficios no-contributivos, pensiones privadas, tres pilares, contribución definida, cobertura, Latinoamérica

Uma onda de grandes reformas estruturais espalhou-se pela América Latina durante a década de 1990, mudando os princípios de distribuição que sustentavam as políticas previdenciárias. Os resultados nem sempre foram aqueles esperados. No contexto sócio-econômico, político e institucional da América Latina, a implementação de novas regras para a aposentadoria resultou em diversas desigualdades que afetam o desempenho do sistema previdenciário e os potenciais ganhos de diferentes gerações e faixas de renda. Para superar desvantagens de distribuição das reformas, governos latino-americanos podem ter que dar um novo papel às aposentadorias de não-contribuintes, assim como considerar melhor a realização de ajustes estruturais específicos para reduzir os riscos e desigualdades inerentes do setor privado. Aprender com exemplos de fora do país pode fornecer ferramentas úteis para atingir esse objetivos.

Palavras-chave: Aposentadorias, reforma da previdência, desigualdade, distribuição, auxílios sem contribuição, sistema de três pilares, contribuição definida, América Latina.



Hide All

1 World Bank, Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth (Oxford, 1994).

2 See, for example, Gøsta Esping-Andersen, The Three Worlds of Welfare Capitalism (Cambridge, 1990); Francis G. Castles (ed.), Families of Nations: Patterns of Public Policy in Western Democracies (Aldershot, 1993); Ferrera, Maurizio, ‘The “Southern Model” of Welfare in Social Europe’, Journal of European Social Policy, vol. 6, no. 1 (1996), pp. 1737; Maurizio Ferrera, ‘The Four “Social Europes”: Between Universalism and Selectivity’, in Martin Rhodes and Yves Mény (eds.), The Future of European Welfare: A New Social Contract? (New York, 1998), pp 81–97; Joakim Palme, Pension Rights in Welfare Capitalism: The Development of Old-Age Pensions in 18 OECD Countries, 1930 to 1985 (Stockholm, 1990); Julian Le Grand, Not Only the Poor (London, 1987); and Richard M. Titmuss, Kay Titmuss and Brian Abel-Smith, Social Policy: An Introduction (London, 1974).

3 Notable exceptions include Carmelo Mesa-Lago, Social Security in Latin America. Pressure Groups, Stratification and Inequality (Pittsburgh, 1978), Changing Social Security in Latin America: Towards the Alleviation of Social Costs of Economic Reform (Boulder, 1994), and Alberto Barbeito and Rubén Lo Vuolo, La modernización excluyente (Buenos Aires, 1992).

4 Income smoothing refers to balancing out income over the course of a person's life, transferring resources from periods when income is higher to periods when income is lower, in order to attain the highest possible living standard over the entire life course.

5 For a more detailed discussion see Camila Arza, ‘Welfare Regimes and Distributional Principles: A Conceptual and Empirical Evaluation of Pension Reform in Europe’, European University Institute Working Papers, RSCAS 2006/30 (2006).

6 On conceptions of citizenship in social security design, see Plant, Raymond, ‘Citizenship and Social Security’, Fiscal Studies, vol. 24, no. 2 (2003), pp. 153–66.

7 Beveridgean systems typically provide universal flat-rate benefits which are independent of the work-nexus.

8 Bismarckian pension systems are characterised by state provision of earnings-related benefits to eligible workers.

9 See Gøsta Esping-Andersen, Social Foundations of Postindustrial Economies (Oxford, 1999); and Peter Baldwin, The Politics of Social Solidarity. Class Bases to the European Welfare State. 1875–1975 (Cambridge, 1996).

10 Baldwin, The Politics of Social Solidarity.

11 For an excellent assessment of the risks involved in different types of schemes see Barr, Nicholas, ‘Reforming Pensions: Myths, Truths, and Policy Choices’, International Social Security Review, vol. 55, no. 2 (2002), pp. 336.

12 The distinction between ‘broad’ and ‘some’ risk-pooling reflects different degrees of coverage of the risks of old-age financing. There is no absolute dividing line between the two. While in some cases, for instance, inflation is fully covered (for example, in DB systems with good indexation formulas), in other cases it is not (for example, unindexed annuities). Labour market risks can also be fully covered (when benefits are independent of working histories), covered partially (for instance when the system requires contributions for a period shorter than the whole working life, or when some periods out of the labour market, such as maternity, are insured), or they can be left uncovered (when there is a full link between contributions and benefits).

13 See Barr, ‘Reforming Pensions’; Nicholas Barr, The Welfare State as Piggy Bank: Information, Risk, Uncertainty, and the Role of the State (Oxford, 2001); and Robert Holzmann, Richard Hinz, Hermann von Gersdorff, Indermit Gill, Gregorio Impavido, Alberto R. Musalem, Michal Rutkowski, Robert Palacios, Yvonne Sin, Kalanidhi Subbarao, Anita Schwarz, Old-Age Income Support in the Twenty-First Century: An International Perspective on Pension Systems and Reform (Washington DC., 2005).

14 Cohort longevity, in contrast, if properly projected by the insurance company, is fully borne by the individual who receives a lower benefit in the face of higher (projected) life expectancies.

15 Mesa Lago, Carmelo and Müller, Katharina, ‘The Politics of Pension Reform in Latin America’, Journal of Latin American Studies, vol. 34, no. 3 (2002), pp. 687715.

16 On the origins and development of social security in Latin America, see Carmelo Mesa-Lago, Social Security in Latin America. Pressure Groups, Stratification and Inequality (Pittsburgh, 1978).

17 Two other countries (Ecuador and Nicaragua) also approved structural reforms, but never applied them. Implementation was postponed indefinitely in Nicaragua due to the fiscal costs of the transition from one system to another, and in Ecuador due to an appeal in the Constitutional Court, which ended in 2005 with the declaration of the unconstitutionality of several articles of the new pensions law. The Dominican Republic's law has been partially implemented in only one of the three schemes it stipulates.

18 See Mesa-Lago, Carmelo, ‘Private and Public Pension Systems Compared: an Evaluation of the Latin American Experience’, Review of Political Economy, vol. 18, vo. 3 (2006), pp. 317–34.

19 In Argentina, the system is both parallel – workers can choose between private or public coverage – and mixed; those who choose private coverage are eligible for ‘first-pillar’ benefits provided by the state.

20 It was argued, for instance, that ‘Some of the problems [of PAYG systems are] the inevitability of intergenerational transfers and low rates of return to later cohorts’; World Bank, Averting the old-age crisis, p. 236.

21 Remaining special regimes include, among others, schemes for public employees in Mexico, Peru and Colombia, for oil industry employees also in Mexico, and for the armed forces in most countries. See Mesa Lago and Müller, ‘The Politics of Pension Reform’.

22 See Barr, ‘Reforming Pensions’; John Eatwell, ‘Pensions, Fiscal Policy and the Distribution of Risk’, paper presented at the Conference Pension Fund Capitalism and the Crisis of Old-Age Security in the United States, New School University, New York, 10–11 Sept. 2004; and Peter R. Orszag and Joseph E. Stiglitz, ‘Rethinking Pension Reform: Ten Myths about Social Security Systems’, in Robert Holzmann and Joseph E, Stiglitz, New Ideas about Old Age Security: Towards Sustainable Pension Systems in the 21st Century (Washington D.C., 2001), pp. 17–89.

23 See Barr, ‘Reforming Pensions’.

24 See World Bank, Averting the Old-Age Crisis; and Olivia S. Mitchell, ‘Building an Environment for Pension Reform in Developing Countries’, World Bank, Social Protection Discussion Paper 9803 (Washington D.C., 1998).

25 Indermit S. Gill, Truman Packard and Juan Yermo, Keeping the Promise of Old Age Income Security in Latin America (Washington D.C., 2003).

26 In the scholarly literature, however, the issue had been raised on a number of occasions before and during the reform processes. See for example Mesa-Lago, ‘Changing Social Security’, and ‘La Seguridad Social y el Sector Informal’, ILO-PREALC (Santiago de Chile, 1990); Barrientos, Armando, ‘The Changing Face of Pensions in Latin America: Design and Prospects of Individual Capitalisation Pension Plans’, Social Policy and Administration, vol. 31, no. 4 (1997); and Vuolo, Rubén Lo and Barbeito, Alberto, ‘La Reforma del Sistema Previsional Argentino: El Mercado de Trabajo y la Distribución del Ingreso’, Estudios del Trabajo, vol. 6, no. 2 (1993), and La Nueva Oscuridad de la Política Social. Del Estado Populista al Neoconservador (Buenos Aires, 1994).

27 In 2000–2001, non-contributory pension benefits represented only 0.2% of GDP in Argentina, 0.4% in Chile, 0.3% in Costa Rica and 0.6% in Uruguay. In contrast, non-contributory pensions concentrated about 1.3% of GDP in Brazil, where quasi-universal rural pensions exist, and 1.2% in Bolivia (year 2004), as a result of the universal pension ‘Bonosol’ (Fabio Bertranou, C. Solorio and W. van Ginneken, Pensiones No Contributivas y Asistenciales: Argentina, Brazil, Chile, Costa Rica y Uruguay, International Labour Office (Santiago de Chile, 2002), p. 19, and Larry Willmore, ‘Non-Contributory Pensions: Bolivia and Antigua in an International Context’, CEPAL Series Financiamiento y Desarrollo, no. 167 (Santiago de Chile, 2006), p. 27). See also Barrientos, Armando, ‘Poverty Reduction: The Missing Piece of Pension Reform in Latin America’, Social Policy and Administration, vol. 40, no. 4 (2006), pp. 369–84.

28 Further methodological details are provided in the note to the table.

29 See, for example, Brittain, John, ‘The Incidence of Social Security Payroll Taxes’, The American Economic Review, vol. 61, no. 1 (1971), pp. 110–25. For a combined cost-benefit analysis for the Argentine case, see Arza, Camila, ‘Distributional Impacts of Pension Policy in Argentina: Winners and Losers within and across Generations’, International Social Security Review, vol. 53, no. 3 (2006), pp. 79102.

30 Total payroll taxes increased from 18% to 20.5–22% in Peru; from 17.8% to 33.8% in Colombia; from 20% to 26% in Mexico; from 19% to 24% in Bolivia; from 11.8% to 13.5% in El Salvador; from 22% to 26% in Costa Rica; from 17% to 21.5% in Nicaragua; and from 9.25% to 20% in Dominican Republic; Gill et al., Keeping the Promise, p. 21.

31 For year 2001, pension deficits financed with government transfers were estimated to be about 4% of GDP in Uruguay, 2.5% in Argentina, 0.5% in Mexico, 3.5% in Bolivia, 7.2% in Chile, 1.4% in El Salvador and 0.7% in Peru. For year 2050, the deficits were estimated at 2.8% of GDP in Uruguay, 4.4% in Argentina, 0.6% in Mexico, 0.9% in Bolivia, 5.4 in Colombia, 0.8% in Chile, 0.5% in El Salvador and 1% in Peru, see, Asta Zviniene and Truman G. Packard, ‘A Simulation of Social Security Reforms in Latin America: What Has Been Gained?’, Background paper for regional study on social security reform, Office of the Chief Economist, Latin American and the Caribbean Region, The World Bank (Washington D.C., 2004), p. 21.

32 Lars Calmfors, Giancarlo Corsetti, Seppo Honkapohja, John Kay, Willi Leibfritz, Gilles Saint-Paul, Hans-Werner Sinn and Xavier Vives, ‘Report on the European Economy 2005’, European Economic Advisory Group at CESIFO (Munich, 2005), p. 76.

33 For a discussion, see Camila Arza, ‘Aims, Outcomes and Prospects of Pension Reform in Argentina: An Assessment Ten Years After’, New School for Social Research, Argentine Observatory, Policy Paper No. 6 (New York, 2005).

34 Gill et al., Keeping the Promise and Holzmann et al., Old-Age Income Support.

36 For a survey of existing models and their application around the world see Willmore, ‘Non-Contributory Pensions’.

37 Larry Willmore, ‘Universal Pensions in Low Income Countries’, Task Force on Pension Reform and Social Insurance, Initiative for Policy Dialogue (Columbia University, 2001).

38 Comisión Económica para América Latina, ‘La Protección Social de Cara al Futuro’ (Santiago de Chile, 2006), p. 141.

39 Data for year 2004. Costs are estimated to grow to 2.9% of GDP by 2050 as a result of population ageing (2% if the eligibility age is raised to 70 years of age); Willmore, ‘Non-Contributory Pensions’, p. 27.

40 Lloyd-Sherlock, Peter, ‘Simple Transfers, Complex Outcomes. The Impacts of Pensions on Poor Households in Brazil’, Development and Change, vol. 37, no. 5 (2006).

41 Comisión Económica para América Latina, ‘La Protección Social’, p. 133.

42 See HelpAge International, Non-Contributory Pensions and Poverty Prevention. A Comparative Study of Brazil and South Africa (London, 2003); Helmut Schwarzer and Ana Carolina Querino, ‘Non-Contributory Pensions in Brazil: The Impact on Poverty Reduction’, Extending Social Security Working Paper, ILO Social Security Policy and Development Branch (Geneva, 2002).

43 Sebastián Martínez, ‘Pensions, Poverty and Household Investment in Bolivia’, University of California at Berkeley, 2004, unpublished.

44 HelpAge International, Non-Contributory Pensions; Schwarzer and Querino, ‘Non-Contributory Pensions’; Barrientos, Armando, ‘Poverty Reduction: The Missing Piece of Pension Reform in Latin America’, Social Policy and Administration, vol. 40, no. 4 (2006), pp. 369–84; Lloyd-Sherlock, ‘Simple Transfers’; Martínez, ‘Pensions, Poverty’, among others.

45 Barrientos, ‘The Missing Piece’.

46 The income difference between the first and the second deciles counts the same, in terms of the inequality coefficient, than the difference between the ninth and the tenth deciles. On inequality measures see Frank A. Cowell, Measuring inequality (London, 1995).

47 The idea of a poverty line that divides the eligible from the ineligible population for means-tested benefits entails a ‘sharp’ conception of poverty. For a discussion on the implications of a ‘sharp’ definition of poverty see Anthony Atkinson, Incomes and the Welfare State (Cambridge, 1995), pp. 233–4.

48 Korpi, Walter and Palme, Joakim, ‘The Paradox of Redistribution and Strategies of Equality: Welfare State Institutions, Inequality, and Poverty in Western Countries’, American Sociological Review, vol. 63 (1998), pp. 661–87.

49 Ibid., p. 672.

50 Larry Willmore, ‘Universal Pensions in Low Income Countries’, Working Paper of Task Force on Pension Reform and Social Insurance, Initiative for Policy Dialogue, Columbia University (New York, 2001).

51 Ruth Hancock, Stephen Pudney, Geraldine Barker, Monica Hernandez and Holly Sutherland, ‘The Take-up of Multiple Means-tested Benefits by British Pensioners. Evidence from the Family Resources Survey’, University of Leicester (2003).

52 See Lloyd-Sherlock, ‘Simple Transfers’; and Old Age and Poverty in the Developing World. The Shanty Towns of Buenos Aires (London, 1997).

53 See Atkinson, Incomes and the Welfare State.

54 The ‘headcount ratio’ is the ratio of poor over non-poor people in each given age-group. The ‘elderly’ have been defined as the population of 60 or more years old, and the income variable (with which the poverty headcount is calculated) has been adjusted for adult equivalents and economies of scale. See Leonardo Gasparini, Javier Alejo, Francisco Haimovich, Sergio Olivieri and Leopoldo Tornarolli, ‘Poverty among the elderly in Latin America and the Caribbean’, Centro de Estudios Distributivos, Laborales y Sociales, Universidad Nacional de La Plata, Working Paper 55, July 2007, Table 3.3, p. 41. In addition, see Barrientos, ‘Poverty Reduction’, p. 372, who shows that the old are overrepresented among the poor in most Latin American countries.

55 Kohli, Martin, ‘Private and Public Transfers Between Generations’, European Societies, vol. 1 (1999), and ‘Generational Equity: Concepts and Attitudes’ in Camila Arza and Martin Kohli (eds.), Pension Reform in Europe: Policies, Politics and Outcomes (London, 2008).

56 Lloyd Sherlock, ‘Simple Transfers’; and HelpAge, ‘Non-Contributory Pensions’.

57 Superintendencia de Administradoras de Fondos de Pensiones, The Chilean Pension System (Santiago de Chile, 2003).

58 On the Swedish case, see Edward Palmer, ‘The Swedish Pension Reform Model: Framework and Issues’, World Bank Social Protection Discussion Paper, no. 0012 (2000).

59 Ibid.

60 On the mechanisms behind NDC models see Cichon, Michael, ‘Notional-Defined-Contribution Schemes: Old Wine in New Bottles?International Social Security Review, vol. 52, no. 4 (1999).

* I would like to thank Colin Lewis for his invaluable help throughout the research on which this paper is based. I am also indebted to Nicholas Barr, Peter Lloyd Sherlock and Rubén Lo Vuolo for comments and suggestions, and to four anonymous referees for critical advice on how to sharpen the argument and presentation of the paper. Naturally, any errors of fact or interpretation remain my own.


Related content

Powered by UNSILO

Pension Reform in Latin America: Distributional Principles, Inequalities and Alternative Policy Options*

  • CAMILA ARZA (a1)


Altmetric attention score

Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed.