Essays on Social Security and Labor Supply

Mobile Share Email Facebook Twitter LinkedIn

by David Seif, Harvard University

A common argument is that investment-based Social Security reform will improve economic efficiency by increasing the perceived link between retirement contributions and retirement benefits.  Under this argument, individuals perceive the monies paid into the Social Security system as a pure tax.  They fail to recognize that payments to Social Security will generally increase their future Social Security benefits.  With personal retirement accounts, in contrast, the link between contributions and future income would be clear, and the economic distortions would be reduced.

My dissertation explores this issue by studying individuals’ labor supply response to two forms of payments made to Social Security, OASDI taxes and the earnings test.  All workers in Social Security-eligible work pay OASDI taxes on income up to an annual limit, while the earnings test in its current form reduces the benefits of certain young Social Security recipients in years that their labor income exceeds a certain level.

The data for my dissertation are from the Health and Retirement Study (HRS), a longitudinal survey, much of which can be linked to Social Security earnings records.  The HRS is conducted in biannual waves, with the first in 1992 and the most recent from 2004.  It includes 30,207 individuals who are observed at least once, and 129,553 person × wave observations, of which nearly 70 percent are of individuals aged 55-80.

To examine the earnings test’s effect on labor supply, I exploit the 2000 elimination of the earnings test for individuals over the Full Retirement Age (FRA) with both a difference-in-difference-in-difference (DDD) regression approach and by fitting a Cox proportional hazard model.  In 2000, the earnings test was eliminated for anyone who had attained the FRA in a previous calendar year, and it was vastly reduced for anyone older than the FRA who reached the FRA that same calendar year.  The 2000 change provides a natural experiment with which I can compare two cohorts—one that was affected by the change in the earnings test, and another that reached the FRA too early to be affected by it.

To study the effect of OASDI taxes on labor supply, I, along with co-authors Jeffrey Liebman and Erzo F.P. Luttmer, exploit complexities in Social Security rules that lead to a large number of discontinuities in marginal Social Security benefits from additional work.  These discontinuities are used to estimate the effect of tax-benefit link on both the extensive labor supply (retirement) and intensive labor supply (earnings and hours worked) margins.

Back to 2008 Dissertation Fellowship Program Recipients