Over at Slate, Steven Landsburg has a piece on the finding that the parents of daughters are more likely to divorce than the parents of sons:
In the United States, the parents of a girl are nearly 5 percent more likely to divorce than the parents of a boy. The more daughters, the bigger the effect: The parents of three girls are almost 10 percent more likely to divorce than the parents of three boys.
The article goes through a number of mechanisms that might explain the difference, though none are entirely convincing. The language of the article is egalitarian, talking mainly about the preferences of parents. But two of the three hypotheses put forward suggest that the preferences of the father drive the outcome rather than those of the mother. More importantly, the emphasis on parental preferences is ultimately a bit restricting.
We don’t have much formal evidence about parents’ actual motives or preferences, so Landsburg suggests three supporting bits of evidence to support the hypothesis that parents do in fact have a preference for sons. First,
divorced women with girls are substantially less likely to remarry than divorced women with boys, suggesting that daughters are a liability in the market for a husband.
This says that fathers don’t like marrying women with daughters. The inference is that fathers prefer to have sons. Second,
parents of girls are quite a bit more likely to try for another child than parents of boys, which suggests that there are more parents hoping for sons than for daughters.
This one could apply to both parents in principle. Finally,
Take a typical unmarried couple who are expecting a child and have an ultrasound, which more often than not reveals the child’s sex. It turns out that such couples are more likely to get married if the child is a boy. Apparently, for unmarried fathers, the prospect of living with a wife and a son is more alluring than the prospect of living with a wife and a daughter.
Here the father’s preference is driving things again. Yet Landsburg concludes:
But the most natural way to interpret their data is that parents, on average, prefer boys to girls.
Two of the three bits of evidence presented concern the choices of fathers. The five per cent difference in divorce rates between the two kinds of families is not huge. The most natural way to interpret the data, it seems to me, is that most American parents manage perfectly well with children of either sex, but there are enough fathers with a sufficiently strong bias against daughters to tip the divorce ratio in the direction shown by the data.
Inferring motives from aggregate patterns is a difficult business. My own bias is against treating explanations of that sort as final. Landsburg himself brings up other countries—like China and India—where the killing of daughters is common. But of course talking in terms of preferences misses the bigger picture. People in these countries don’t just happen to prefer sons so much that they’ll kill their daughters. In these societies, property, status and social mobility are tightly integrated with patriarchal gender relations. Family wealth and position are much more difficult to manage when you have daughters rather than sons. Individual preferences are generated in the context of institutionalized gender relations that are tied to other sources of social power. That’s what patriarchy is all about.
A natural comparative hypothesis, therefore, would be that the divorce rate for parents of daughters tends to converge with that for parents of sons as gender discrimination (e.g., in the labor market) declines. The idea is that there’s no need to have a bias for boys when daughters and sons can do equally well. It’d be tricky to study, because divorce rates differ for other reasons, but I’d say you could at least compare many OECD countries in this way.
A different way to approach the question—restricting ourselves to the narrow question of individual preferences again—would be to allow parents to choose the sex of their babies and see what happens. This may be on the cards, at least if the engagingly-named Microsort Corporation continue to improve their methods. As it happens, most of the people who avail of Microsort’s services want to have a girl. (This may be due to the success rates of the method itself.) I heard about this company at a talk by a biologist who was arguing that in the future, rational parents will invest in the genetic endowment of their children (height, IQ, looks, etc) in the same way they invest in their education today, and for the same reasons—to maximise the return on their investment in terms of the child’s earning potential. I asked him why rational parents would ever choose to have a girl, given what we know about gender and the wage gap but he seemed to have trouble grasping the premise of the question.
 If the word “patriarchy” gives you hives, well, tough shite. In any event, go read Gerry Mackie’s “Ending Footbinding and Infibulation: A Convention Account”, American Sociological Review 1996 v.61 pp.999-1017 for an elegant treatment of one issue in this area from a rational-choice perspective. [J-STOR link.]