Cohabitation and Divorce: A Case of Complicated Connections

Thanks to PAA President Wendy Manning for the link to a recent article in The Atlantic discussing two academic papers that seem to come up with different conclusions about the relationship between cohabitation and divorce after looking at the same dataset.

Late last month, the Journal of Marriage and Family published a new study with a somewhat foreboding finding: Couples who lived together before marriage had a lower divorce rate in their first year of marriage, but had a higher divorce rate after five years. It supported earlier research linking premarital cohabitation to increased risk of divorce.

That study was conducted by Michael Rosenfeld at Stanford and Katharina Roesler at Quora. 

As evidence of the importance of this topic to family demography, I noticed that the authors referenced four different Past Presidents of the PAA--Larry Bumpass, Andrew Cherlin, Arland Thornton, and Linda Waite. 

But just two weeks later, the Council on Contemporary Families—a nonprofit group at the University of Texas at Austin—published a report that came to the exact opposite conclusion: Premarital cohabitation seemed to make couples less likely to divorce. From the 1950s through 1970, “those who were willing to transgress strong social norms to cohabit … were also more likely to transgress similar social norms about divorce,” wrote the author, Arielle Kuperberg, a sociology professor at the University of North Carolina at Greensboro. But as the rate of premarital cohabitation ballooned to some 70 percent, “its association with divorce faded. In fact, since 2000, premarital cohabitation has actually been associated with a lower rate of divorce, once factors such as religiosity, education, and age at co-residence are accounted for.”

That report is actually based on a paper just accepted for publication in the journal Marriage & Family Review, and in that paper the author references six different Past Presidents of the PAA--Larry Bumpass, Andrew Cherlin, Paul Glick, Daniel Lichter, S. Philip Morgan, and Arland Thornton. 

Both papers utilize data from the National Survey of Family Growth and of considerable interest is that both papers include the following graph illustrating the genuinely revolutionary change in cohabitation patterns in the U.S. The chart below, taken from Kuperberg's article, graphs the percentage of first marriages preceded by cohabitation with the current spouse, according to the year of marriage.

Given the recent rapid rise in cohabitation in the U.S., it should not be surprising that researchers are finding what seem like conflicting findings. It is more likely the case that things are so complicated that it is very difficult to pin down patterns in and causes of the observed changes.