Joint Retirement of Couples: Evidence from a Natural Experiment

Abstract

We estimate and explain the impact of early retirement of husbands on the wives’ probability to retire within one year, using administrative micro panel data that cover the whole Dutch population. We employ an instrumental variable approach in which retirement choice of husbands is instrumented with eligibility rules for generous early retirement benefits that were temporarily and unexpectedly available. The empirical model explains wives’ probability to retire within one year. We find that early retirement opportunity of husbands increased the wives’ probability to retire within one year by 25.4 percentage points. This is a strong effect. Our results are robust to several sensitivity tests.

Keywords: Instruments, retirement, couples

JEL Classification: C26, J26, J120, J140