Red State, Blue State: Poor State, Rich State?

Over $100,000 in income, you are more likely than not to vote for Democrats. People never point that out. Rich people vote liberal. I don't know what that's about.

--Tucker Carlson, conservative TV commentator, 2007

Andrew Gelman didn't believe Tucker Carlson. And he didn't believe NYT columnist David Brooks' simple "red state, blue state" political-alignment hypothesis, either. And so, Professor Gelman, a political scientist and statistician at Columbia University, set out to debunk these apparent public misperceptions. His quest is chronicled in a new book, Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do, and reviewed in a Chronicle of Higher Education article, "Peeling Back Paint-by-Numbers Politics."

Gelman covers a lot of familiar ground, venturing into established scholarly work about party polarization, religion, regionalism, and how to define the white working class. But he also found this: in the 2004 election, high-income individual voters tended to vote for Bush, but states with high average incomes tended to vote for Kerry.

Thirty years ago, there was apparently no correlation between states' average income and states' presidential preferences. It has evolved over the last few decades. And Gelman is now more respectful of Brooks' observation, saying he "was doing what thoughtful journalists are supposed to do. He noticed a pattern that increased over a few decades and became very dramatic in 2000."

At the heart of the book is the red-state/blue-state paradox. If rich voters prefer Republicans, why do rich states tend to lean Democratic?

Gelman's answer is this: In states with relatively low median incomes, individuals' income strongly predicts their voting behavior. In Mississippi and Montana, for example, Bush won among upper-income voters in 2004 by vast margins, won among middle-income voters by smaller margins, and lost among lower-income voters. In rich states, by contrast, the relationship between income and voting is much weaker. In states like Maryland and Delaware, Bush won among upper-income voters, but only narrowly. And in four wealthy states — California, Connecticut, Massachusetts, and New York — Kerry won outright among upper-income voters.

It is that anomalous behavior among relatively affluent voters in relatively rich states that needs to be examined, Gelman suggests.

Gelman explores several popular explanations for the phenomenon, finding a few to have some possible influence, but most not enough to explain what is happening by themselves. One, for example, is education levels, which have so-far shown only a weak correlation with red state, blue state voting patterns. But one explanation does seem to him compelling.

Affluent people have sorted themselves into two opposing camps. This, Gelman believes, is close to the heart of the matter. There is reason to believe, he says, that affluent people are self-consciously arranging themselves into insular liberal and conservative communities. Among affluent people, religious participation (or its lack) is a very strong predictor of voting. And — as the journalist Bill Bishop and the sociologist Robert G. Cushing argue in their recent book, The Big Sort: Why the Clustering of Like-Minded America Is Tearing Us Apart (Houghton Mifflin, 2008) — there is evidence that counties are becoming more politically homogeneous. (Imagine a new college graduate from Iowa deciding whether to move to Brooklyn or the Dallas suburbs.)

But the sorting is not just geographical, Gelman notes. It also has to do with political-party identification. Evidence suggests that Americans — especially affluent Americans — have gotten better since 1980 at "correctly" choosing the political party that matches their policy preferences. In part, that is because the parties themselves have become more polarized and ideologically predictable; to take the best-known example, there are fewer abortion-rights Republicans and anti-abortion Democrats in Congress than there were in 1988.

Still, according to Gelman, most Americans hold more complicated, unpredictable combinations of policy preferences. And often they don't fit neatly into either party's standard platform positions. He notes with implied irony that, "Only half of abortion-rights supporters, for example, also support universal health insurance. And only 62% of universal health insurance supporters also support abortion rights." The exception? That's right.

But, significantly, Gelman says, there is evidence that one group of Americans is becoming polarized in ways that the political parties would recognize: the rich. 'Americans who are wealthier look more like politicians in terms of the ideological consistency of their beliefs,' he writes. 'The same is not true for those with college degrees as opposed to those without. Nor is it true for Southerners versus non-Southerners, or religiously observant people versus nonobservant people.'

But Gelman knows that the voter landscape is still changing, and so he awaits the results of November's election--and more, perhaps, the exit poll data that will be released in the days thereafter.

There's some suggestion in polls today that younger voters and highly educated voters will break heavily for Obama," he says. "which would contradict our view that age and education are weak predictors. So that's something we'll be looking at carefully. I'll be staying up late that night, like everyone else.

For the times they are a changin'. They really are, and likely will be for some time to come. And for the good, I believe. Let's hope so.