There was just one post yesterday, “Income, Race and Voting:”
Still thinking about the new GOP idea — hey, let’s go for white voters! Why didn’t we think of that before? And I thought I’d do a cleaner version of some stuff I did a while back. I’m venturing into political science territory here,and would be happy to have real experts weigh in; but I’m pretty sure I have the basics right here.
So, let’s look at some exit poll data, and cross-tab it with Census income data. In the figure below, the red lines show the income-voting relationship from the Times summary of exit polls, which also supplies the broad ethnic group data. For incomes, I use Census data on median household income for 2011, which is also available for regions. For voting I use Alabama to represent the South, Ohio to represent the Midwest.
So here’s my picture:
Contrary to what some people keep saying, people with higher incomes, other things equal, tend to vote Republican. Cut through the noise and fog, and it is true that Democrats broadly want to redistribute income down, and Republicans want to redistribute income up — and on average, voters get that (which is why “libertarian populism” is hot air). But race and ethnicity also matter, a lot. What you can see right away is that there are three groups that are fairly anomalous.
1. African-Americans “should” lean Democratic, given their low incomes, but they are much more Democratic than this alone would predict.
2. Southern whites are just as much of an anomaly; they have close to the national median income,and “should” be pretty evenly split between parties, but instead are almost entirely Republican.
3. Asian-Americans are relatively high-income, but also strongly Democratic. Although I don’t have the data, Jews would surely look similar.
There really isn’t any mystery, of course, about these anomalies. Despite occasional attempts to widen its appeal, the GOP has effectively defined itself as the party of white Christians — and there are still a lot of historical memories that go with that definition.
Interestingly, Hispanic voters aren’t that much of an anomaly; they vote Democratic, but given relatively low income and a corresponding reliance on the safety net, that’s not surprising — and they’re not nearly as Democratic as the only somewhat poorer African-American community.
This in turn suggests a problem Republicans may not fully realize with their new strategy, which is in effect to be even more the white Christian party. You can think of this as an attempt to persuade Ohio whites to start voting like Alabama whites, which I guess could happen.
But what if the effect is,instead, to persuade Hispanics to start voting like African-Americans?