Research Grant Summary - Trends in Labor-Force Participation
Anderson, Patricia M., Gustman, Alan L., Steinmeier, Thomas L., Social Security Bulletin
Authorization for extramural research in the vast area of Social Security is awarded under sections 702 and 1 10 of the Social Security Act. The following summary, "Trends in Labor-Force Participation and Retirement: A Nontechnical Summary of a Final Report to the Social Security Administration, September 1994" (Grant No. 10-P-98343-1-01), is based on the grant topic of "trends in labor-force participation and retirement of older workers." The summary was 1 coauthored by Patricia M. Anderson, Assistant Professor of Economics at Dartmouth College in Hanover, New Hampshire; Alan L. Gustman, Loren M. Berry Professor of Economics at Dartmouth College in Hanover, New Hampshire; and Thomas L. Steinmeier, Professor of Economics at the Texas Technical University in Lubbock, Texas. (An announcement of the research grant subject areas for fiscal year 1995 appears on pp. 74 in the Summer 1994 issue of the Bulletin.
Trends in Labor-Force Participation and Retirement: A Nontechnical Summary of a Final Report to the Social Security Administration, September 1994
The research summarized here investigates trends in labor-force participation and retirement among older workers. Disagreements remain in the literature attempting to explain trends in retirement in the post-World War II period and the eventual leveling of that trend, and basic questions have remained unanswered. Are trends to earlier retirement explained by trends in demographic characteristics, educational attainment, occupation, industry, wage, and other observable characteristics often measured in microdata sets? If not, are the trends in retirement the result of changes in pensions, Social Security, and related policies? If changes in pensions and Social Security are an intermediate cause of trends in retirement, what ultimately are the causal factors?
The analytical framework distinguishes approaches at three different levels. First, trends in labor-force participation and retirements of older workers may be analyzed by focusing on labor-supply behavior at the level of the individual, who takes the pension and Social Security benefit structures as given. Second, these trends may also be analyzed from a labor-demand perspective at the level of the firm, which determines pension structure as part of its personnel decisions that affect employment and the structure of compensation. Or, third, these trends may be analyzed at the level of the market, again treating compensation levels and structure as endogenously determined, and focusing either on a reduced form or structural approach. This study adopts elements from each approach.
Our investigation first confirms the presence of trends in labor-force participation and retirement in data for men from the late 1960's through the middle of the 1980's, but not in data applicable to retirement trends of women. The analysis then indicates the relation of these trends to components of labor-force participation, finding that the trend in the retirement hazard, that is, the probability of retirement by those remaining in the labor force, dominates the trend in participation for those entering the retirement-age range, taken as age 55. The analysis then considers the relation of trends in labor-force participation to causal variables. On the one hand, our findings suggest that the observables collected in microdata surveys, such as the Current Population Survey (CPS) and the Panel Study of Income Dynamics (PSID), add only modestly to our understanding of the trend to retirement and the recent leveling of the trend for men. On the other hand, within the context of a structural retirement model, it is possible to explain about a quarter of the trend to earlier retirement among men observed from the early 1970's to the early 1980's as resulting from trends in incentives created by retirement programs.
More specifically, results from three types of empirical analyses are reported in this summary. First, using the CPS and PSID, we investigate trends to reduced labor-force participation and to earlier retirement observed from 1968 through 1991. …