I investigate the factors that fostered rising retirement rates prior
to social security and most private-sector pensions by estimating the
income effect of the first major pension program in the United Sates,
that covering Union Army veterans. The elasticity of nonparticipation
with respect to Union Army pension income was 0.73. The findings sugge
st that secularly rising income explains a substantial part of increas
ed retirement rates. Comparisons with elasticities of nonparticipation
with respect to social security income suggest that the elasticity of
labor force nonparticipation may have decreased with time, perhaps be
cause of the increasing attractiveness of leisure.