The larger problem with the Teixeira-Abramowitz piece is that when you cut through the rhetoric, my core thesis still stands. Even taking every word in their piece as true, it remains the case that there were well over 5 million fewer white voters than would have been reasonably expected in 2012. This analysis is based on 2008 turnout and population growth. That’s not really in doubt.

Nor is it a mystery which type of white voter stayed home last year. These no-shows fit a profile. They turn out to be the downscale whites whom Teixeira has previously insisted Democrats must woo. If these voters had turned out, they probably would have improved Romney’s share of the vote. This is the crux of my argument, and the only real mystery is why some people find this conclusion so upsetting.

Teixeira and Abramowitz focus on the fact that there were also “missing” non-whites, which is true. There’s nothing “misleading” here. It is made clear by the chart in the middle of Part 1’s first page. Someone could construct a reasonable argument from this that Democrats can counter any Republican surge with missing white voters by bringing these non-whites back. I don’t think that’s likely the case, as I think it would be hard to find a candidate with a substantially stronger appeal to non-whites and better get-out-the-vote organization than Barack Obama. But it is plausible.

Teixeira and Abramowitz, however, try to go a step further. They argue that these “missing” non-whites negate the importance of the missing whites outright. They look at a different data set — the CPS data, which I’ll just call the “census data” — and conclude that the decline among all groups was due to the same factor: lack of interest in the election. If this were true, the only way these “missing whites” might return is if the “missing non-whites” also return — and that would result in no net improvement for Republicans.