Matthew Yglesias wonders if the decrease in Democratic partisan ID from 2008-2010 is a sign that our politics is moving in a more racial identity politics direction. I tend to doubt it. Over the last ten years, the two party vote among white voters has tended to move within a small range and in response to events and personalities. Bush won 54% of the white vote in 2000 and 58% of the white vote in 2004. The Republican share of the white vote went down to 53% in 2006. It rebounded to 55% in 2008 and went up to 60% in 2010. It looks like about 7% of the white voting population is made up voters who swing between the two parties based on conditions and the personalities and appeals of candidates. There are multiple stories that can be told from the data. Here are two:
1. Five percent of white voters abandoned the Republican Party in 2006 in response to dissatisfaction with the Iraq War, Katrina, Republican corruption etc. Less than half of those persuadable whites moved back into the Republican column in 2008 in response to the Democrats nominating a liberal candidate (though a smaller fraction voted Republican than when the liberal Democratic presidential candidate had been white.) Then the combination of undivided Democratic government, the enactment of an aggressively liberal agenda, and a labor market in which the unemployment rate stagnated near 10% pushed virtually all of the floating white electorate into the Republican column for at least one election.
2. The floating white population switches parties based on randomly recurring racial identity crises that they collectively announce by voting for Republicans. Their voting patterns are not based on the economy, war or peace. Even the racial identity of the President does not matter until it does. For some reason the President's race was not a problem for these voters when they were voting for President in 2008, but it became a problem when they were voting for Congress in 2010. These people are so weird.
This is actually bad news for the Republicans - though possibly good news for America. If the Republican share of the white vote in 2010 was based on a combination of dissatisfaction with economic conditions and recoil from perceived Democratic left-radicalism, then a combination of an improving labor market and branding the Republicans as right-radicals might put the Democrats in position to win back some of those white voters. That is probably one reason why Obama is avoiding coming out with an entitlement reform plan. If he can brand the Republicans as Social Security and Medicare cutters and the labor market is seen to be improving, he has a decent chance of whittling the Republican share of the white vote back down to the mid-50s. Republicans shouldn't take their 2010 share of the white vote for granted and need to be asking themselves what they have to do to improve their share among the various populations of nonwhite voters.