The Problem with ex post facto explanations

The New York Times recently published (as an editorial piece) a dozen experts' views on why Hillary Clinton did not win the Democratic nomination. They views are all different. A couple suggest that her problem was that Barack Obama was more successful as part of their arguments. None mentioned that more delegates to the convention are going to support Obama than would support Clinton, which is perhaps the one demonstrably correct explanation of why she lost.

I think it is obvious that there were many factors that led directly and indirectly to the outcome of the primary process, and that there can be no adequately valid short answer as to why Clinton lost and Obama won. I suspect that the professional pollsters and campaign managers are working very hard to analyze what happened in order to learn lessons that may be useful in future presidential campaigns. They may have success, at least in that they will get jobs and perhaps in that they will do better in those future campaigns than they would have without the analysis.

Still, when a dozen very smart people who have studied the campaign full time for 18 months come up with different short answers to the same basic question, it should make us think. Each of the answers taken alone seems rational, and we can assume that at least eleven of them are less valid than one would desire. Indeed, the one answer that has the nature of a tested hypothesis is that the Clinton campaign strategy and candidacy, which was planned on the basis of prior understanding of American presidential politics, failed against the strategy and candidacy of Obama. Of course, we don't know how much of the outcome to attribute to the differences in the strategies, how much to attribute to the candidates, and how much to attribute to the combination of strategy and candidate.