The labor force participation rate of older men under age 65 has shown a significant recent decline. Cross-sectional studies linking early retirement to increased Social Security income have also made explicit or implicit temporal projections of changes in participation in response to changes in benefits. However, use of cross-sectional estimates for projection purposes may run into several problems, including temporal dependence of the participation decision. This paper uses 2-year

more »

... data for men aged 58-62 in 1969 in order to trace changes in labor force behavior near retirement age. Results indicate significant effects of poor health, initial assets, and initial pension eligibility on the probabilities of exit and entry from the labor force. Social Security benefits are found to have insignificant or unexpected effects. The results also indicate evidence of temporal dependence of participation, suggesting caution in interpreting projections of cross-section estimates.