What DOES Divorce Do Anyway?

I haven't read the empirical literature on the economics of divorce very closely, but I get the distinct impression that we're not really sure empirically what effect it has. First, Gruber found that unilateral divorce hurt the children in the longrun. But then Wolfers cast some of this into doubt. Another interesting paper by Mark Hoekstra at Pitt finds essentially the same thing using the decision to discontinue divorce proceedings as a way of identifying the effect of divorce (or in his case staying together) on child outcomes. Not a perfect study, as probably the decision to remain together is endogenous, and I suspect the two are really not controls for one another, but it's pretty clever too even if imperfect. Now I see this new paper in a forthcoming JHR that looks at how marriage breakup affects earnings of divorced women and it finds the support for the classic result that divorce is actually bad for the women. Over at Freakonomics, Wolfers goes into a lot of detail about the paper, which is going to require so much effort on my part that I will almost certainly never read it. All this to say, the empirical economics of divorce is hot, active, and I think in need of a few good (wo)men. Wolfers and Stevenson have pushed hard to have their alternative perspective on divorce get an equal footing in the debate, and I think have been overwhelmingly successful at showing some of the benefits of divorce - such as lowering spousal homicides and suicides (didn't see that coming when I first read the paper! Now I totally believe it). Marriage and divorce are still very important and active areas within the economics of the family.