Thoughtful piece by Columbia's John Huber about the obsession with the identification of causal effects tending to drive other equally legitimate questions out of the ball-park. I happen to think he is right.
On the other hand many sociologists are not clear about the kind of question they are asking in the first place - "effects of causes", "causes of effects" or something else? And, believing that the be all and end all of what we are supposed to be doing is not just discovering the "effects of causes" does not mean that when your question is of the "effect of causes" type you can just ignore identification problems, endogeneity etc and in best hand waving style, hope for the best. At the other extreme don't get me started on the first find a cool instrument, then think up a question, school of thought.
Wouldn't be surprised though if there was a large group of sociologists out there who believe that the identification problem is a hitherto unrecognized aspect of intersectionality...
Posts
No comments:
Post a Comment