Megan McArdle's an intelligent woman, and she makes a good point about looking at effects on institutions at the margins; but in this case I've got to say she fails to come up with a complete and convincing mechanism for the changes in single motherhood. Essentially she points out a correlation (not even in the general population, but in a vulnerable subpopulation), points out some marginal effects at the outset, and asks us to fill in the rest of the causal chain ourselves. I'm pretty sure she's not being intentionally deceptive. But a sufficiently clever activist or social critic can come up with a convincing story about marginal effects linking just about any two temporally close phenomena; if we're really interested in this rationality business, we should remind ourselves of that especially when the story at hand is flattering to our preconceptions.

A simple counterexample: further down the article, McArdle draws similar conclusions regarding the divorce rate in the US. Her argument is essentially that making a divorce easier to formalize led to a cascade of marginal effects that badly weakened the institution. Note that she's not talking about making divorce directly more socially or religiously acceptable, just legally easier.

It would be a convincing argument, if it generalized properly. But counterexamples aren't hard to find, and one in particular comes from a surprising place: divorce is quite easy to formalize in classical sharia. It can be initiated unilaterally by either party, with varying but generally fairly relaxed rules, although the surrounding legal structures make it somewhat easier for men than for women. This has been true for about fifteen hundred years; yet we don't see high divorce rates in Muslim countries. We could speculate endlessly about why, with varying degrees of cultural acumen, but I'd say that's already fairly convincing evidence that there's a more complicated story that needs to be told.