New paper: New Evidence on the Labor Supply Effects of the Social Security Earnings Test

Leora Friedberg of University of Virginia and Anthony Webb of the Center for Retirement Research at Boston College have released a new paper entitled New Evidence on the Labor Supply Effects of the Social Security Earnings Test, which will be presented at the NBER's September 25^th Washington, DC event on Tax Policy and the Economy. Here's the abstract:

This paper investigates the impact on labor supply of changes in the Social Security earnings test in 1996 and 2000. We highlight how inertia in labor supply choices influences responses to policy changes. We do this in two ways. First, we show that taking account of last year's employment status is important in estimating responses to current earnings test changes, a step that has not always been taken in the literature. Second, we test the effect of not only actual but also anticipated earnings test parameters which cohorts faced at earlier ages. This approach demonstrates that past and anticipated future rules influence current employment and earnings. Thus, we identify an impact of earnings test changes on employment at ages that are younger than those directly affected and in years that follow the direct change. Lastly, we show that earnings test changes that were initiated in 1996 had an effect, in addition to the changes in 2000 that have been recently studied by others. Overall, we predict that the elimination of the earnings test in 2000 raised employment among Health and Retirement Study respondents by around two percentage points at ages 66-69 and 3.5 points at age 65, with gains persisting as exposed cohorts aged and also being observed at younger ages due to the shock to anticipated earnings test rules.

One important issue with how the earnings test affects labor supply is whether Social Security participants correctly understand that benefits lost to the earnings test will be made up with benefit increases once they reach the full retirement age. Friedberg and Webb argue (correctly, I believe) that most participants don't understand how the earnings test works. I wrote on this aspect of the earnings test here.