"The authors examine the relationship between stock market performance and retirement behavior. They first present a descriptive analysis of the wealth holdings of older households and simulate the labor supply response among stockholders necessary to generate observed retirement patterns. Few households, they find, have substantial stock holdings, and these holdings would have to be extremely responsive to market fluctuations to explain observed labor force patterns. The authors then exploit stock market fluctuations since the early 1980s (particularly the boom and bust between 1995 and 2002), along with variation in stock exposure, to generate a double quasi-experiment, comparing the retirement and labor force re-entry patterns over time of those more and less exposed to the market. Any difference in behavior that emerged during the boom should have reversed itself during the bust. The authors find no evidence that changes in the stock market drove aggregate trends in labor supply."

Support: Paper