Sunday, 24 March 2013

Knot bothered about causality

The report Knot Yet, on the increasing age of marriage in America, has gathered plenty of media attention.
It's a really important topic, but I can't help be worried that in 2013, we still accept social science that makes no attempt to distinguish causality from correlation. The report rightly contains comments like: "... we cannot rule out the possibility that some of these associations are simply due to the type of young adults who marry in their twenties." Indeed they cannot. But media reports don't bother to mention this, instead throwing around headlines like "Late Marriage and its Consequences", or phoney statements like "Upper-class women reap a large wage premium from delaying marriage", where all we can say is "Upper-class women who delay marriage have higher wages" - or, exactly equivalently, "Upper-class women who have higher wages delay marriage".
Is proper social science really so hard to do on this topic?