Since pension schemes-along with health care and education-absorb the largest amount of social expenditure in all countries, their reform has a potentially major impact both on the fiscal situation of the state and on the life chances of citizens who stand to win or lose from new arrangements. This makes pension reform a highly controversial issue; and, except for the addition of new programmes and benefits, major restructuring of existing pension systems has been extremely rare in advanced industrial democracies. It was also rare in Latin America before the 1980s and 1990s. But there has been a great deal of experimentation within the region during the past decade. This paper examines the larger economic, social and political context of Latin American pension reform and compares experiences in different countries of the region with options available in Western European societies during the same period. The authors argue that the type of pension reform undertaken in Latin America has been an integral part of the structural adjustment programmes pursued by Latin American governments, under the guidance of international financial institutions (IFIs). Although there was a range of possible remedies to the problems of pension systems in different Latin American countries, neo-liberal reformers and the international financial institutions preferred privatization over all others. They claimed that privatization would be superior to other kinds of reform in ensuring the financial viability of pension systems, making them more efficient, establishing a closer link between contributions and benefits and promoting the development of capital markets-thus increasing savings and investment. And they were able to push through some of their suggestions for reform in spite of considerable opposition from pensioners, trade unions and opposition political parties. Interestingly enough, their pressure proved least effective in the more democratic countries of the region. In Costa Rica, for example, citizens preferred to reform the public system-eliminating the last pockets of privilege for public sector workers and ensuring that new levels of contribution would be adequate to provide minimum benefits for the aged and infirm. In Uruguay, citizens forced a public referendum, through which they rejected a proposal for privatization. At a later stage, they did permit the introduction of private investment accounts, but not at the cost of eliminating the public programme. In Argentina and Peru, after the legislature refused to authorize partial privatization, this was eventually pushed through by presidential decree. Only in Chile and Mexico has there been a complete shift to private pension funds-but, in both cases, influential sectors of the elite, including the military, have been allowed to keep their previous, publicly managed group funds. Looking at the only privatized pension system in existence long enough to allow for some assessment of its consequences-that of Chile-the authors find that many of the claims made by supporters of privatization are not substantiated by the evidence. The first discrepancy between neo-liberal predictions and the reality of Chilean pension reform has to do with efficiency. All previous claims to the contrary, private individual accounts have proven more expensive to manage than collective claims. In fact, according to the Inter-American Development Bank, by the mid-1990s administration of the Chilean system was the most expensive in Latin America. The second disproved claim involves yield. When administrative costs are discounted, privately held and administered pension funds in Chile show an average annual real return of 5.1 per cent between 1982 and 1998. Furthermore high fees and commissions-charged at a flat rate on all accounts-have proven highly regressive. When levied against a relatively modest retirement account, for example, these standard fees reduced the amount available to the account holder by approximately 18 per cent. When applied to the deposit of an individual investing 10 times more, the reduction was slightly less than 1 per cent. The third discrepancy involves competition. Although it was assumed that efficiency within the private pension fund industry would be associated with renewed competitiveness-while the public pension system represented monopoly-the private sector has in fact become highly concentrated. The three largest pension fund administrators in Chile handle 70 per cent of the insured. And to reduce advertising costs, public regulators are limiting the number of transfers among companies that any individual can make. A fourth unfulfilled promise of privatization in Chile has to do with expansion of coverage. It was assumed that the existence of private accounts would increase incentives for people to take part in the pension sc heme, but in fact this has not happened. Coverage and compliance rates have remained virtually constant. A fifth major claim was that the conversion of the public pension system into privately held and administered accounts would strengthen capital markets, savings and investment. But a number of studies have recently concluded that, at best, this effect has been marginal. And finally, the dimension of gender equity within a fully privatized pension scheme is being subjected to increasing scrutiny. Women typically earn less money and work fewer years than men. Therefore, since pension benefits in private systems are strictly determined by the overall amount of money contributed to them, women are likely to receive considerably lower benefits. Public pension systems, in contrast, have the possibility of introducing credits for childcare that reduce this disadvantage. Sweden is an example of countries that have embarked on this course. In the latter part of the paper, Huber and Stephens widen their comparative framework to include recent pension reforms in advanced industrial countries. There, where economic crisis was not as severe and where pressure from international financial institutions was not significant, much broader options for reform were available. In fact, although long-established systems were under stress, no developed country opted for complete privatization. Complex measures were taken to strengthen the funding base of national pension systems, including changes in investment procedures and changes in rules for calculating pension benefits. Reforms also increased retirement age, as well as the number of years required to qualify for a full pension. But even the most thoroughgoing reforms retained a central role for public schemes in ensuring old-age benefits. In conclusion, the authors consider steps that can be taken to craft pension reforms with more desirable results than those obtained to date in Latin America. They recommend measures that address the problem of an aging population by increasing the ability of each generation to pay for its own pensions-rather than relying primarily on the contributions of preceding generations of insured workers. Pension payments should be invested in a variety of financial instruments and benefits must ultimately be related to the yields obtained. Such a strategy does not require introduction of privately managed, individually held, investment funds. On the contrary, risk is lessened by relying instead on collectively managed funds, in which accounts can either be identified with individuals or-more equitably-with generations of contributors. Reformed public pension systems should also contain minimum "citizenship pensions" that guarantee subsistence income in old age to all individuals as a matter of right. Such a measure, financed from general tax revenue rather than from personal contributions, is not beyond the means of medium income countries in Latin America and the Caribbean. In fact, some Nordic countries introduced citizenship pensions when their GNP per capita was lower than that of most Latin American countries today.