Fathers and daughters

Marginal Revolution and Newmark’s Door have noted two papers by Andrew Oswald (University of Warwick) and Nattavudh Powdthavee (University of London’s Institute of Education) on the effect of the sex of children on parents’ political preferences. The primary paper is on Britain and there is a brief follow-up paper finding confirmation of their results for Germany.

The core of these papers is a curious empirical finding. Parents who have daughters tend to move to the left, and parents who have sons tend to move to the right. More specifically, in Britain, parents who have daughters tend to subsequently switch from voting Conservative to voting Liberal Democrat or Labour (and vice versa for parents who have sons), and in Germany, parents who have daughters tend to subsequently switch from voting Christian Union Party or Christian Social Democrats to voting for the Social Democratic Party (and vice versa for parents who have sons.) Craig Newmark is skeptical, perhaps because raising two daughters have made neither him nor his wife particularly left wing. I am less surprised. There is evidence that women tend to favor bigger government, and I am tolerably persuaded by Oswald and Powdthavee’s suggestion that parents take account of their children’s preferences. (There is, by the way, an excellent survey by Shelly Lundberg on how the sex of children affects parents’ choices.)

My problem with Oswald and Powdthavee’s papers is that I have no idea why or how women’s preferences for political party should be different from men’s, and they offer no clue. They offer the feeble suggestion that women prefer a bigger supply of public goods, such as public safety. Plausible, because women are more vulnerable to violence, but it is not clear to me what that has to do with British women preferring Labour or the Liberal Democrats. Those parties’ attachment to big government is heavily about redistribution, not about more police on the streets. Damned if I know what is going on.