The Impact of Chile’s 2008 Pension Reform on Labor Force Participation, Pension Savings, and Gender Equity

In an individual-account based system and in the absence of state-provided non-contributory pensions, women can be particularly vulnerable to old-age poverty. This is due to their lower wages, interrupted careers, typically younger retirement ages and longer life spans. Reducing the pension gender gap was a significant objective of the 2008 pension reform of the Chilean Pension System, which significantly changed features of the existing contributory pension system. The main objective of this study is to examine whether the reforms to the pension system promoted gender equity and also whether the insurance features of the new pension system, designed to guard against low pension accumulations, will generate unintended behavioral responses by altering incentives to work and save. These responses are captured by a dynamic model of labor supply and saving decisions estimated on longitudinal data from the Encuesta de Proteccion Social of the Microdata Center from the University of Chile and administrative data from the Superintendencia de Pensiones. The behavioral model takes into account dimensions of individual heterogeneity which allow us to capture distributional aspects of the impact of the reform. We document that the fit of the model to the data is reasonable and use the model to simulate the differential, 5-year ahead impact of the reform on women’s pension levels relative to men’s, labor supply, poverty levels, contribution densities, participation in the formal sector and age of effective retirement. According to our simulations, the reform will dramatically improve pension saving levels for women (see figure 1), bridging a sizeable part of the gap between male and female pension benefits (see figure 2), and reduce poverty levels at older ages (see figure 3). This results from the very large coverage1 and high level2 of the basic solidarity pension (“Pension Basica Solidaria”, or PBS) implemented by the reform. We anticipate some negative behavioral responses in the form of lower labor force participation at older ages and lower participation in the covered sector, resulting in lower contribution densities. In our simulations, attempts to make the reform more incentivecompatible by tapering-off non-contributory benefits do not offset income effects resulting from higher benefit levels. As retirement nears, incentives to contribute to the pension system are lower than before the reform due to higher expected income in retirement. This tends to reduce participation in the labor market, particularly in the covered sector and for women, relative to before the reform. These predictions are qualitatively consistent with the available post-reform (2009) data but the magnitude of these effects remains to be validated with more recent data that can capture the full impact of the reform. The report develops as follows. Section 2 describes the Chilean Pension System, focusing on aspects that are relevant to gender equity. Section 3 provides an overview of the methodology which is then detailed in sections 4 to 9. Section 10 discusses model fit, section 11 reports model simulations with and without the reform and section 12 concludes.