My only follow up would be this: what else can explain Obama’s 40-point deficits in West Virginia and Kentucky? The states are lacking in some of Obama’s most reliable constituencies, but so are states like Nebraska, South Dakota, Idaho, Wyoming, and Alaska, but Obama won each of those contests easily. ~Steve Benen
Benen is kidding, right? First, South Dakota hasn’t voted yet–North Dakota moved up to 2/5, but not S.D., which votes on June 3. That’s why Clinton was campaigning in Sioux Falls the other day (it worries me that I know that). I know that it is probably considered unconscionable pro-Clinton shilling to say this, but the caucuses that Obama won by such ridiculous margins on February 5 are not representative of the broad majority of Democratic voters in those states. That’s just the nature of a caucus system. Caucuses go to the candidate with the superior organisation, funding and GOTV efforts, which is why Romney performed so much better in these venues than in most primaries, and why Ron Paul doubled and tripled his normal 8-10% percentage of the vote in some of these same caucus states. Romney had the money and organisation, and Paul had money and loyal, zealous supporters. It is a credit to Obama’s political operation (and a lasting mark of shame on Clinton’s) that he cleaned up in these caucuses, but it is not evidence that he used to win the sorts of voters he is now losing.
In Oklahoma, Clinton won the primary by 22 points, and the electoral map looks a lot like Kentucky’s will in a little over a week: an island of Obama voters (OKC) in a sea of light blue (except that he is unlikely even to win the Louisville area). My guess is that, if Nebraska, Wyoming, Alaska and Idaho had held primaries, Obama’s share of the vote would have been at Oklahoma-like levels, and we would not now be talking about Obama’s victory lap. Instead, they held caucuses–those are the breaks. This is not to argue that those caucus results “don’t count” or “shouldn’t count as much,” but it is true that a caucus format disproportionately attracts certain kinds of voters (those with more income, more education and more information), and these tend to be the voters who are more likely to prefer Obama. Obviously, it helps even more when only one campaign actively competes and the other pretends that these elections don’t matter, which is another reason why Obama’s margins in some places were so huge. A caucus format does not involve “disenfranchisement,” as some lame Clintonites have tried to argue, but it rewards the campaign that can mobilise better-informed, highly-motivated supporters and punishes the campaigns that have supporters who are either less activist (and generally less obsessed with politics) or too busy to participate. We don’t know whether race was a factor in Oklahoma voting, for example, because no one even bothered to investigate the question, but it is not necessarily obvious that race, much less race alone, explains the size of Clinton’s lead in Kentucky and West Virginia. Caucus results from Super Tuesday definitely don’t tell us what we want to know, because a caucus is an entirely different kind of process.
P.S. South Dakota and Montana are both holding their primaries in June, so at that point we might be able to compare apples to apples when we see the results next month.
Update: Oklahoma’s exit polling seems to confirm that the patterns people started obsessing over in March and April were quite evident by early February in all their particulars.