Andrew F. Newman‡
We study US divorce rates, which despite the continuing rise in female labor force participation (FLFP), have been falling since the mid-1980s, reversing a two-decade trend. A cross section of U.S. states for the year 2000 displays a negative relationship between the divorce rate and FLFP.
We present theory and evidence in support of the view that these recent trends are the product of two distinct economic forces: relative to their non-career counterparts, career women display greater selectivity in the search for marriage partners and greater ﬂexibility in sharing the beneﬁts of a marriage with their partners. Greater selectivity implies that career women will be older when they ﬁrst marry and that their marriages will be of higher average “quality,” possibly making them less prone to breakup. Greater ﬂexibility implies that it is easier for two-earner families to re-adjust the intrahousehold allocation to compensate for changes in outside opportunities, making marriages more resistant to “shocks.” Our evidence shows that both eﬀects may be playing a role in generating the trends the trends.
Acknowledgement to be added.
Department of Economics, Boston University, 270 Bay State Road, Boston, MA 02215, and the Eitan Berglas School of Economics, Tel-Aviv University, Tel-Aviv, Israel 69978; Email zvika@BU.edu, URL http://people.bu.edu/zvika, http://www.tau.ac.il/~zvika/.
Department of Economics, Boston University, 270 Bay State Road, Boston, MA 02215, Email firstname.lastname@example.org, URL http://people.bu.edu/afnewman.
Department of Economics, Boston University, 270 Bay State Road, Boston, MA 02215, Email email@example.com, URL http://people.bu.edu/olivetti/.
From the early 1960s through the early 1980s, the divorce rate and the female labor force participation rate (FLFP) in the U.S. both trended upward. A large, multidisciplinary literature has oﬀered many suggestions as to why this should be so. In the past two decades, however, the divorce rate has fallen, while FLFP continues to rise. Moreover, a cross section of U.S. states for the year 2000 displays a negative relationship between the divorce rate and FLFP. This evidence suggests that something is missing from existing theories that connect FLFP and divorce.
In our empirical analysis, the most important variables that are negatively correlated with divorce are the median age at ﬁrst marriage (MAFM) and FLFP, even when both are present in the same speciﬁcation. After exploring the evidence in greater depth, showing for instance that these two variables’ contributions persist controlling for education, income inequality, and a number of demographic characteristics, we go on to suggest two economic forces that may be generating the trends. Relative to their non-career counterparts, career women display greater selectivity in the search for marriage partners and greater ﬂexibility in sharing the beneﬁts of a marriage with their partners.1 The selectivity eﬀect can be expected to be strongly associated with delaying the age at ﬁrst marriage, but is less successful at explaining the additional contribution of FLFP.
One distinguishing feature of a career woman is that she values a marriage of given quality less than a non-career woman, because she has her own means of ﬁnancial support. We show that this implies she will be choosier in selecting a partner; on average she will marry later and the quality of her marriage will be higher, leading to a lower chance of divorce. Countering this “ex-ante” selectivity eﬀect, however, is an “ex-post” one: a career woman will be less tolerant of low-quality marriage; this could lead to higher divorce rates for career women relative to non-career women who marry at the same age. The other distinguishing feature of the career woman is that the earnings she derives from the labor market...