Research

This paper investigates the forces behind a shift in the timing of work across two cohorts of men without a college degree in the United States. The 1960s cohort (born 1955–1970) participates less at prime ages and more at older ages than the 1930s cohort (born 1930–1945), so the two age profiles cross. The prime-age decline is three times larger among men with a work-limiting health condition, even though health distributions are similar across cohorts. I estimate a life-cycle model of men's labor supply and saving using PSID data for the 1930s cohort, introducing wages, family income, health risk, and retirement institutions that shape participation over the life cycle. The estimated model matches the cohort's participation profiles, and I use it to account for the cross-cohort shift and the contribution of each channel to the shift. The model is able to reproduce both the prime-age decline and the late-career reversal. At prime ages, higher spouse income among married men and wider disability access finance nonparticipation, especially for men in bad health. At older ages, the wage profile becomes the largest force raising participation, as the 1960s cohort's wage disadvantage closes with age. The transition from defined-benefit to defined-contribution pensions and changes in Social Security rules add to the rise by weakening the incentives to leave the labor force, together contributing over half of the rise the model produces at ages 65–69. Overall, the 1960s environment is a welfare gain (2.9% of lifetime consumption). This gain is concentrated among married men, whose wives' higher earnings contribute to an improvement of 13.7%. The other five channels have a negative effect on welfare, with the largest loss coming from wages (−8.9%).