This paper provides novel evidence on the labor supply response to negative income shocks in retirement, exploiting an institutional feature that caused differential and unexpected income losses among otherwise identical individuals in a sharp regression discontinuity design. We conclude that...