Who Determines When You Retire? Peer Effects and Retirement

John M.R. Chalmers, Woodrow T. Johnson, Jonathan Reuter

NBER Retirement Research Center Paper No. NB 08-13
Issued in September 2008

We study the factors that affect the choice of retirement date for a large sample of retirement-eligible Oregon state employees between 1992 and 2003. Given the complexities of the Oregon Public Employees Retirement System, we hypothesize that individuals will infer their optimal retirement behavior, in part, from the retirement behavior of their coworkers. Indeed, controlling for individual retirement incentives and characteristics, we find that individuals are economically significantly more likely to retire in months when more of their coworkers retire. The facts that this positive correlation is robust to the inclusion of numerous fixed effects and controls and survives estimation using two distinct sets of instrumental variables lead us to conclude that it reflects peer effects in the choice of retirement date. With respect to the possible welfare consequences of peer effects, our preliminary evidence is mixed, but points to modest costs and benefits. Interestingly, we find little evidence of peer effects amongst employees whose salaries are in the bottom 25th percentile, where financial literacy is likely to be the lowest.

PDF (134 K)Abstract (12 K)Executive Summary (30 K)