Grey Divorce and labour Supply

ABSTRACT

We use data from the Health and Retirement Study (HRS) to examine how labour supply changes around divorces that occur later in life. We find that the probability of work and hours worked increase for women, but decline for men, with evidence of an anticipation effect for men. We find weak evidence of a post-divorce decline in per-capita wealth and stronger evidence of a decline in per-capita non-own-wage income for women, but not for men. While not causal, these findings are consistent with income and possibly wealth effects driving the post-divorce increase in women’s labour supply.

KEYWORDS:

• Divorce
• ageing
• retirement
• labour supply
• income effect
• wealth effect

JEL CLASSIFICATIONS:

• J12
• J14
• J16
• J26

Acknowledgments

We thank Tarush Gupta for excellent research assistance. We thank Bryan Stuart and Hector Tzavellas for helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

¹ Johnson and Skinner (1988) provide further discussion

² The wealth results are consistent with Sharma’s (2015) finding of a larger post-divorce decline in total wealth for women than for men.

³ We drop all person-wave observations with age below 50 or missing.

⁴ Our methodology is similar to Goda and Streeter (2021), who examine wealth changes around events that include widowhood and divorce.

⁵ We also estimate saturated models, with dummies for all pre- and post-divorce waves. Results (in Figures A1-A7 in the appendix) are similar.

⁶ Around 22 (73) percent of individuals are not married five waves before (five waves after) divorce. By construction, all are married one wave before divorce and divorced in the next wave.

⁷ One individual reports working 196 hours, which we recode to 168 (the number of hours in a week).

⁸ Alimony and child support (components of ‘other household income’) are not asked about for some respondents in 1994, and for all respondents after 2000. Results are similar across observations that include and exclude alimony/child support (see Figures A6 and A15 in the appendix).

⁹ Brown et al. (2019) estimate smaller remarriage probabilities. Excluding those who do not experience divorce (as they do) brings our estimates closer to theirs (see appendix).