Essays on retirement and health
MetadataShow full item record
I exploit unanticipated reforms to the Texas Teacher's pension plan to estimate the effect of pension incentives on retirement decisions. In 2000 and 2002 the Teacher Retirement System increased the benefit levels of all employees covered by the pension system. The reforms provide plausibly exogenous variation in the incentives to work - which differentially impacted workers due to non-linearities in the pension's benefit schedule. I leverage the reforms coupled with the non-linear benefit schedule in an instrumental variables framework to estimate the effect of pension related incentives on the decision to retire. I find substantial heterogeneity between men and women in their response to a one-year incentive to remain in the labor force. Additionally I find that a 10 percent increase in forward looking incentives decreases the probability of retirement by 1.84 percentage points from a baseline of 11 percent. In the second chapter, I shift the focus to the estimation of the elasticity of lifetime labor supply. I use a nonparametric approach based on the excess bunching of retirements around kinks and discontinuities in the lifetime budget constraint that arise out of nonlinearities in the pension benefit schedule. The method of using excess bunching at budget constraint nonlinearities to estimate elasticities is relatively new and has only recently been applied to retirement. I find that high service individuals face a kink in their lifetime budget constraint which leads to excess retirements around the nonlinearity in the pensions benefit schedule. Furthermore, I find that for low service individuals, the TRS pension benefit schedule creates a notch in the individual's lifetime budget constraint. This notch leads to a distinct shape in the distribution of retirements, where there is noticeable missing mass in the distribution just prior to the notch point. This happens as individuals delay retirement until they reach the notch point.