Despite the myths you might have heard, half of American first marriages don’t end in divorce. In reality, about a third do, down from the divorce surge of the 1970s and 1980s, though second and third marriages are much more vulnerable. Recent marriages are doing particularly well thus far: Just 15 percent of the Americans who tied the knot since 2000 have decided to get it undone within the first eight years of marriage.
The predictors of divorce, however, remain mysterious. But in a new study published in the American Sociological Review, Harvard sociologist Alexandra Achen Killewald has found that the things that increase the probability of divorce — as they relate to work, at least — have changed over the past couple decades. It turns out that the amount of money that either the husband or wife makes isn’t that important: For contemporary couples, the biggest determinant is whether the husband is working full-time.
The data set is enviably large. She tracked 6,309 married couples between 1968 and 2013, 1,684 of whom divorced or permanently separated during that time. To account for the “quiet revolution,” or the time when women could count on pursuing full-fledged careers, she drew a line between couples married before and after 1975. In the early cohort, wives who did 50 percent of the housework had a 1.5 percent chance of divorce within the next year, while those who did 75 percent of the housework had a 1.1 percent chance. In the later cohort, wives’ paid or unpaid labor had little effect on her chance of divorce, but the husband’s breadwinning mattered a lot. If he was employed full-time, there was a 2.5 percent chance of splitting up in the next year, and if he wasn’t, there was a 3.3 percent chance. These might not seem like gigantic numbers, but they’re measuring year-to-year probabilities rather than the total timeline of a marriage, and for an event that affects the life course as much as divorce does, they’re worth paying attention to.
The results contradict a couple of the leading explanations for why people divorce and why so many people broke up in the 70s and 80s in particular. Drawing from that data, Killewald concludes that the “material circumstance” of the couple has little to do with divorce: neither how much money either partner makes nor the wages they earn relative to one another are determining factors. Also, it doesn’t appear to be the case that the financial dependence — or independence — of the wife is affecting the likelihood that a couple splits up. For couples in the post-1975 cohort, the way they divvy up unpaid labor — household chores, taking care of the kids — had little effect on divorce probability.
“This shows that, for contemporary couples, wives can combine paid and unpaid labor in various ways without threatening the stability of their marriages,” Killewald wrote to Science of Us in an email. ”But, for those same marriage cohorts, the risk of divorce increases substantially when the husband isn’t employed full-time.” While the homemaker ideal has waned in importance, the notion of the breadwinner is still hanging on.
There are a couple limitations to the study, Killewald says. There wasn’t enough data for same-sex couples to do separate analyses for the group, indicative of how much more qualitative and quantitative research is needed to do there. Also, her data didn’t get at why the husbands weren’t working. It could be that they were involuntarily unemployed or that they were “stay-at-home” dads. Other research indicates that, for now, it’s mostly involuntary: Pew found that in 2012, just 21 percent of stay-at-home dads did it because they wanted to care for family (though that’s up from 5 percent in 1989).
Killewald is careful to note that her work here can’t capture the many intricacies of a relationship. Lots of couples have husbands go through periods of unemployment and are perfectly fine, and there are, of course, tons of factors out there that lead to divorce beyond money. Psychologist John Gottman, who’s spent decades studying couples, says that the four big drivers are contempt, criticism, defensiveness, and emotional withdrawal. But learning about the correlates of divorce sheds a light on what brings people together.
Particularly, the money you’re bringing in is less important to your marriage stability than the unpaid or paid work that you do. This is because of the “symbolic content” that work carries for people, Killewald explains. The study suggests that when partners fulfill the roles that are expected of them, marriages are more stable. But what it means to be a “good wife” and “good husband” has clearly changed since the 1960s, and people bring different conceptions of what being “good” at their marital role is depending on their background. “There’s no objective definition of what work behaviors constitute being a ‘good wife’ or ‘good husband,’ and individuals may also vary in how they understand the expectations of marriage,” she explains. “I hope readers might think about what those terms mean to them and to their partners.” In other words, use your words.
*An earlier version of this post stated that half of American marriages don’t end in divorce, when in fact it’s first marriages that have that lower divorce rate; second and third marriages end in divorce much more frequently.