Post-2008 PVIs (For States)

You may remember back in July I tried to predict how the PVIs of the various states would look after the 2008 election. (PVI is a tool generally applied just to congressional districts, but you can use the formula for any unit of analysis: states, counties, whatever. Click here to learn a bit more about the method of calculating it.) That was quite the exercise in speculation (nevertheless, if you go back to that diary, I was extremely close on many of them… although that speaks more to Nate Silver’s predictive abilities than my own, as I was relying on his mid-July projections for each state, many of which were quite accurate on their own).

Now we have actual state data for the presidential race, so I can return to this topic with more authority. In most people’s minds, this was a sea change election, a total map-changer… but if you look closely at the underlying data and not just the colors on the TV screen, it wasn’t. Most of the states behaved exactly as you’d expect them to, coming in a few points more Democratic in a year where the Democratic candidate performed a few points better than the previous few Democratic candidates. In other words, most states’ boats were lifted the same amount by the one overall rising blue tide.

There were some big shifts and drops, though; where were they? The states where the PVI most notably shifted to the Democrats were Colorado (+3), Hawaii (+6), Indiana (+3), Montana (+4), Nevada (+3), New Mexico (+3), North Dakota (+3), South Dakota (+3), and Vermont (+5). With the exception of Hawaii (favorite son effect) and Vermont (large 2000 Nader effect falling out of the equation), the explanation for these states seems to be a combination of two factors: Obama’s greater appeal (maybe personality-wise more so than policy-wise) to midwestern and western states, and the fact that the Obama campaign actually put a lot of ground game effort into these states instead of treating them as an afterthought. (Like the saying goes, “80% of success is just showing up.”) Interestingly, in July one other state projected to swing big to the left in PVI was Alaska, but that was prior to Palinmania.

States going the other way were Arkansas (+5), Louisiana (+4), Oklahoma (+4), Tennessee (+4), and West Virginia (+3). (A number of northeastern states had a smaller shift, not because they moved to the right, simply because they were already pretty pro-Kerry and thus didn’t move to the left as fast as most other states.) These would tend to suggest that Obama did have at least something of an “Appalachian problem,” or at least that he underperformed notably in the states with a high white evangelical/”American ancestry” population.