You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Unhappy Marriages May Produce More Daughters

Upending a decades-long assumption that having daughters can lead to higher rates of divorce


Scientists have known since the 1970s that couples with firstborn daughters are slightly more likely to get divorced than couples with firstborn sons, and they’ve traditionally assumed that the blame lay with the baby girls themselves. But new research calls this decades-old finding into question, suggesting that a couple in an unhappy marriage is actually more likely to produce a daughter than a son.

“Sociologists would argue that fathers identify more closely with sons,” said Amar Hamoudi, an economist who conducts research at the Duke Population Research Institute and co-authored the study along with Jenna Nobles, a sociologist at the University of Wisconsin-Madison. “If they have a young son in the house, they’re more invested in the family; that draws them closer to their spouse and that keeps the marriage firm.”

It’s well-established that girls and women have lower mortality rates than men at every stage of life, from birth to death, and epidemiological evidence suggests they’re hardier before birth, too. Hamoudi and Nobles argue that female embryos may actually be more likely to survive the sub-optimal conditions in the womb of a woman stressed out by an unhappy marriage. “The biological system that responds to chronic stress and the biological system involved in gestation are very closely linked to each other,” explained Hamoudi. Hamoudi and Nobles analyzed surveys on levels of conflict in marriages from the early 1990s and data on the gender of the firstborn children of the survey respondents. (Relationship conflict is a strong predictor of divorce.)

In any given eight-month period, the risk of divorce for a couple whose first-born child is male is about 1.5 to 2 percent; if the first-born is female, the risk climbs to 1.6 to 2.1 percent—the sex of the baby would be playing a role in about one in a thousand divorces. According to the models Hamoudi and Nobles created, the effect of marital stress on the baby’s sex accounts for about half of that difference—meaning that daughters would be “causing” divorce in only one of two thousand cases.

The evidence on whether or not fathers in contemporary American society display a preference for sons is “mixed,” Hamoudi told me. “Our study makes it an even more open question than it was before.”

And it could have important implications for sociology itself. Sociologists have traditionally refrained from studying embryos and fetuses, leaving them to biologists and epidemiologists—but they may be missing out. “If we really want to understand the population,” said Hamoudi, “We shouldn’t just start the clock at birth.”