When Marriage Doesn't Work

When it comes to divorce, it’s easy to look back and pick out all the portents of doom one should have paid closer attention to. But Ioana Marinescu, an assistant professor at the Harris School of Public Policy, argues in a new paper that it’s possible to predict the likelihood of divorce. Marinescu’s model, based on economic data, can’t anticipate certain events that could alter the course of a marriage, but it can identify which marriages are more likely to end.

Marinescu’s paper, recently published in the journal Labour Economics, grew out of her previous research on job loss. “Jobs end because they’re not the right job, or maybe they were the right job and stuff changed,” she says. “That’s intriguing. It could also explain marriage: why people stay together or don’t.”

Marinescu began with the assumption that most people who enter into marriage find their partner a good match at the time. So then what happens? There are two theories, she says. The first is the “learning model,” which posits that marriages begin with a lack of information, and over the years one or both parties make a series of negative discoveries about the other.

The second model is the “change model.” Marinescu finds this theory much more plausible, given the number of couples who live together before they get married and have a chance to get to know each other—and their bad habits—quite well before making a lasting commitment. “When you got married,” she explains, “you knew what you were getting into. You thought it was a good match. Then things happen. There are new developments, like losing your job. If a marriage is so-so, this could put it over the edge.”

In order to test the impact of job loss on marriage, Marinescu needed an enormous data set. Divorce and job loss are not, after all, common occurrences in the course of one’s life. First she identified the married couples in the US Census Bureau’s Survey of Income and Program Participation and zeroed in on the first ten years of the marriage, when couples are most likely to get divorced. Then the fun began. She discovered that when one spouse was fired from his or her job, the couple became 65 percent more likely to get a divorce. (More precisely, they were 57 percent more likely if the husband was fired, 90 percent more likely if the wife was fired.)

“When someone is fired,” Marinescu posits, “it’s more indicative of how their character has changed. What you’re doing at work is not good, and you might be doing the same thing at home. A layoff is economic. But if you’re fired, you personally messed up.”

This is not to say that couples that are forced to endure a layoff together always stay together: a layoff increases the likelihood of divorce by 57 percent. Men may feel insecure because their position as breadwinner is threatened; previous studies have shown that if a woman earns more than her husband, the couple is more likely to divorce. (Marinescu’s study only examines heterosexual couples because the data come from the 1990s and 2000s, before same-sex marriage became legal.)

There are a few variables that might alter the influence of job loss on a marriage, Marinescu found. The most notable is homeownership, a sign of a strong marriage: buying and selling property, and dividing it in the event of a separation, is too big a hassle to enter into lightly. Children, though, are more of a mixed bag, because parenthood makes people change. (It also interferes with their sleep patterns.) Previous research has found that the risk of divorce increases in the first year after the birth of a child.

What are the policy conclusions? “If this is all about learning,” she says, “you can never prevent divorce because you’ve already made a mistake. The best thing is not to rush. Learn first. You can help people by teaching them how to face negative changes. People who are more resilient in the face of change are likely to have more stable marriages.”