Public pensions and retirement: Evidence from the Railroad Retirement Act

Abstract

This paper estimates how public pensions affect retirement timing by examining the Railroad Retirement Act of 1937, which replaced private railroad pensions with a national program comparable in many ways to Social Security. Leveraging linked decennial census records between 1910–1940, the first part of the analysis compares male labor force nonparticipation in 1940 relative to 1930, between workers previously in railroad versus other industries with broad pension coverage, and by age. Higher benefits led to earlier retirement, largely driven by exit at age 65. The second part of my analysis also exploits the switch from flat to progressive benefits in average wages to estimate the elasticity of nonparticipation with respect to benefits for men aged 65-69. My central estimate of 0.55 indicates a large retirement response. Application of these estimates to Social Security expansions in the 1950s suggests rising benefits was the key driver of earlier retirement among the already-insured male population during this era.