The Boy Bias

We all know that kids are a joy, right? And many of us are so thrilled to become parents that we don't care whether the little bundle is a boy or girl. But by a large margin, American men who do have a preference say they would rather have a son than a daughter, and this boy bias subtly shapes such decisions as whether to marry, divorce and have another child.

That's the conclusion of economists Gordon B. Dahl of the University of Rochester and Enrico Moretti of UCLA, who have found that couples are more likely to stay married if they have sons, more likely to divorce if they have daughters and more likely to have another child if all their children are girls.

Happily, the consequences of boy bias are limited. "For most marriages there is likely no effect at all," Dahl said. "On others, the effect is larger," he said, adding that "you'd probably have to have a pretty poor marriage to begin with for this to have a big effect."

Dahl and Moretti analyzed U.S. Census data collected from 1940 to 2000, as well as California birth records from 1989 to 2001. Using these massive data sets, they were able to isolate the effects of boy bias on dating, mating, marriage and child-rearing practices.

They found some evidence of gender bias wherever they looked. In families with at least two children, the probability of giving birth to another child is greater in all-daughter than all-son families -- a result that "would be hard to explain if parents were completely gender unbiased," they wrote in a working paper published recently by the National Bureau of Economic Research.

The bias showed up in the divorce stats, too. "Over the whole study period, the divorce rate for a three-girl family versus a three-boy family is 5.7 percent higher," Dahl said.

Here's the good news, though: The divorce effect appears to have diminished over time. But that doesn't necessarily mean that American men have seen the light and are losing their preference for sons. Instead, these researchers found that the divorce effect has been replaced by what they called a "custody effect."

In the past, fathers rarely had sole or joint custody of their children. That apparently made some fathers with sons think harder about divorce out of fear that they would lose contact with their sons. But since 1970, the number of children living with their fathers has quadrupled -- giving rise to an apparent boy bias in matters of custody. "If you're the daughter of a divorced dad, you're much less likely to be living with your father than a son is," Dahl said. "Fathers with all-boy offspring are 11 to 18 percent more likely to have boys living with them, depending on how many sons they have, compared to all-girl offspring."

Perhaps the researchers' most compelling evidence for boy bias comes from an unexpected source: data relating to shotgun marriages.

The California birth records note whether or not an expectant mother had an ultrasound, a procedure that is more than 95 percent accurate in predicting the sex of an unborn child. Slightly more than a third of all mothers in the California sample had undergone ultrasounds, which means that they and the father may well have known the gender of their child before the baby was born.

When Dahl and Moretti analyzed this group of mothers separately, they found this disturbing pattern: Those who gave birth to boys were more likely to be married than those who gave birth to girls. While the researchers don't know for sure, the circumstantial evidence strongly suggests that "fathers who find out their child will be a boy are more likely to marry their partner before delivery," Dahl said.

Of course something other than boy bias could be at work, the researchers acknowledged. Perhaps fathers fear girls will be more expensive to raise than boys -- call it the "sticker shock" theory. Then again, years of research suggest that sons need their fathers more than daughters need their dads -- the so-called "role model" theory. Perhaps couples with sons somehow take this into account in deciding whether to divorce or to quickly get married if a baby is on the way.

The problem is, neither theory explains why couples with daughters are more likely to try to have another baby. Kids are expensive, whether they're boys or girls. And while boys need dads, that doesn't explain why couples with girls disproportionately desire to have another child.

Boy bias fits all the data, Dahl asserts. Plus, there's direct evidence from polling that point to the source of this bias: men. Gallup surveys over the past two decades show that among American guys who have a preference, sons are favored 2 to 1 over daughters while women who have a preference are only slightly partial to daughters.

Dahl said the best antidote to divorce is a happy marriage, regardless of the gender of your kids. A good home life overwhelms such emotional eccentricities as men's preference for sons, women's slight preference for daughters, guys' love affair with watching professional sports and that hang up women have about toilet seats that are left up. (Now somebody needs to study that.)

And don't despair if you're blessed with daughters, Dahl said. He speaks from experience.

"I've been married for 11 years and have three girls," he laughed. "I'm very happy. So if the question is, 'If you have daughters, is your marriage doomed,' the answer is no."

Author's e-mail: morinr@washpost.com

For more news, or to subscribe to the newspaper, please visit http://www.washingtonpost.com

Copyright 2004 washingtonpost.com

 Return to News