Population aging has put a fiscal strain on the Social Security system of many countries. One of the main policy responses has been to move from a Defined Benefit (DB) to a Defined Contribution (DC) Social Security system. How this change would affect saving and labor supply is crucial for evaluating the desirability of such a reform. While there is substantial existing work on how the size of Social Security benefits impacts economic outcomes (e.g. Feldstein 1974; Attanasio and Brugiavini 2003; Attanasio and Rohwedder 2003), there is little convincing empirical evidence on how the type of Social Security benefits impacts outcomes of interest, in particular on the effects of moving from a DB to a DC system. This project aims to develop an empirical strategy for convincingly and precisely estimating how a DC system impacts saving and labor supply behavior, using extremely accurate data.