Joint Retirement: Evidence on the Heterogeneity of Spousal Effects

Evidence abounds to suggest the existence of retirement spillovers among spouses. Using the Survey of Health and Retirement in Europe (SHARE), this paper not only confirms the existence of joint retirement behavior among dual-worker couples around Europe, but also shows that the intensity of retirement coordination varies a lot. The results of the paper are essentially five fold. First, among spouses there is a gender asymmetry: wives are more likely to be influenced by their husbands' decision to retire. Second, a higher labour market attachment (proxied by education, income quartile or self reported quality of work) translates into a lower propensity of retirement coordination. Especially, for men who belong to the highest income quartile or education level there is absence of joint retirement. Third, being a secondary earner increases the propensity of retirement coordination. Fourth, higher age differences between couples generally reduces joint retirement, but in interaction with eligibility rules. Five, there is evidence on the enhancing role of converging preferences in terms of activities practiced by both partners, whereas convergence in philosophical views or personality traits do not have any significant effect. Among the traditionally discussed determinants of joint retirement, leisure complementarities are important for couples' retirement incentives, nevertheless, they are mostly dominated by income effect and feasibility of joint retirement (eligibility for both partners to retire).