The Effect of Pension Wealth on Female Labor Supply Before Retirement: Evidence From Germany

Demographic trends have increased the potential labor supply of the elderly – women in particular – and its macroeconomic importance, but are simultaneously exerting mounting pressure on the financial sustainability of public old-age support programs. In response to these trends, many countries have implemented – or are planning to implement – pension reforms that aim to encourage longer labor force participation. A rich literature has investigated how the characteristics of public pension systems affect employment at older ages. However, the design of social security can affect labor supply well before retirement. For example, more generous systems might induce workers to reduce their lifetime labor supply. What are the effects of changes in pension benefits on labor supply prior to retirement? And, what is the relative importance of pension wealth (wealth effect) and implicit labor taxation (substitution effect) in affecting labor supply choices? Answering the second question has proven particularly challenging in the empirical literature, due to the difficulty in separately identifying wealth effects and substitution incentives of pension programs. This project will provide novel estimates of the effect of pension wealth on individual labor supply before retirement. The project exploits a German pension reform that changed the generosity of pension contributions credited to parents – mostly mothers – for the time spent raising their children. From July 2014, parents of children born before January 1, 1992 had their pension benefits increased by EUR 336 per year per child; the pension contributions of parents of children born on or after January 1, 1992 was left unchanged. The timing and nature of this policy reform – popularly known as ‘Mütterrente’ – generate a pure pension wealth effect on affected parents. Using administrative social security data on a large sample of German women, who had children in a neighborhood of the January 1, 1992 policy cutoff and were therefore aged approximately 45-60 at the time of the reform, the analysis examines preretirement labor supply responses to a substantial change in pension wealth. To identify the effect of interest, the empirical analysis adopts a difference-in-differences design, comparing the dynamics of labor supply before and after July 2014, for women who had children just before and just after January 1, 1992. The richness of administrative data allows to investigate various dimensions of heterogeneity, such as the work-disutility, income and age gradients of the wealth effect. Well-identified estimates of pension wealth effects in the female population and by various dimensions of heterogeneity are useful to inform models of policy evaluation and long-term projections. In this respect, the project sits within the Center’s research focal areas of ‘informing long-term projections and models’ and ‘requirements of work in the modern economy’.