Rich people vote more Republican than do lower-income people. But there aren’t that many rich voters, so in absolute numbers they are outnumbered by middle-income voters. There have been some changes (see graphs above), but Republican voters have always been a mix of different income levels and social classes.

This is not always understood by journalists, however. For example, in an article for the Atlantic titled “The Class War Inside the Republican Party,” Alex Roarty writes:

The GOP has long ago shed its stereotype of being the party catering to the wealthy. . . . In 2008, according to a tabulation of exit-poll data acquired by the National Journal, blue-collar workers made up 51 percent of all GOP primary voters.

I don’t believe it. There just aren’t that many blue-collar workers out there. Or, maybe it’s true — I’ve been wrong before! — but I’ll only believe it when I see the data. I’m guessing that they just took some broad demographic category that includes lots of people who work in non-blue-collar jobs as well as people who aren’t working at all. I doubt the exit poll data has enough information to classify people by their jobs.  (Here is the Merriam-Webster definition of blue collar. The Wikipedia entry is more comprehensive.)

My larger complaint with what Roarty wrote is threefold.

First, as I discussed in this space last year, “blue collar” is a somewhat loaded term, in that it is typically men, not women, whose jobs are categorized as blue collar, and men are more likely to vote Republican.

Second, regarding income, education, and voting, we have found that, within each education class, the richer people are more likely to vote Republican. At the low end of education, the higher-income people include some of those successful blue-collar workers and business owners. The high end of education includes relatively low-income professionals such as social workers who are likely to vote for Democrats.

In most groups of the population — especially the more conservative and Republican groups — richer people are more conservative. For example, military officers are much more conservative than military enlisted personnel.

Third, Roarty’s presentation of time trends is misleading in that the Republican Party has always (or, at least, since the 1940s, which is when we started having reliable polling data) had the support of a mix of social classes. Sure, the details were different — back in the ’30s and  ’40s a lot of Americans were farmers — but the Republican Party (and, for that matter, the Democratic Party) has for many decades had to balance the interests of lower-income, lower social class voters and higher-income, higher social class leaders and funders. It’s not new. It’s always interesting and always worth writing about, but I think it’s highly misleading to blur the historical context. In 1920, 1940 and 1960, Republican leaders have had to somehow get the votes of millions of non-elite voters; they have needed to convince millions of working-class and middle-class Americans that low taxes are good for them; and so on.