If speculating about electability is a mug’s game, speculating about electability on the basis of primary and especially caucus wins is simple insanity.  Every campaign promotes this speculation, but it is very misleading to make any claims for the general election based on such outcomes.  After all, Bush was destroyed in New Hampshire in 2000 by John McCain, but he then carried that state in the general against Gore.  His relative weakness with independents against McCain was obviously irrelevant when he had to run against Gore.  Florida, meanwhile, was a Bush romp in the primaries and was not, to put it mildly, nearly so easily won in November.  Likewise, McCain has won New Hampshire twice in primary season, but I don’t think any sane strategist believes New Hampshire will vote Republican in November, McCain or no McCain.  So when McCain talks rubbish about being competitive in New York in the fall, or Mark Penn blathers about competitiveness in “swing constituencies,” or Obama fans become excited because their man won an Idaho caucus, we should acknowledge this as nothing more than very unpersuasive spin.  It is interesting that Obama has won a lot of caucuses and relatively few primaries (except where he has a certain natural demographic advantage), which we might call the political Mitt Romney Disease.  National polling shows Obama as the more competitive Democratic candidate nationwide, while the same polling showed Romney as a hopeless disaster, so there may be no relationship between electability and reliance on caucuses.  Still, I would suggest that if the nominating contests tell us anything about the general they may tell us that the candidate who relies more heavily on caucuses is probably the one who excites and mobilises activists and not the one who wins over large swathes of the electorate.  But this is really every bit as unreliable as the national polling of candidates about whom most voters know nothing.