Jerry Bowyer is a conservative radio host in Pittsburgh who was nice enough to have me on his show a couple of times to talk about Spinsanity and All the President's Spin. But his latest economics column for National Review Online is wrong-headed at best.
He writes that "states that went for George W. Bush in the last election are considerably poorer than the ones that went for Kerry. The notion that the GOP is the party of the rich simply doesn't match the economic reality."
But you can't make reliable inferences about individual behavior from aggregate data; this is what is known as the "ecological fallacy". And the obvious counterpoint to Bowyer is that the bivariate correlations between party and income are pretty clear.
Let's consider the simplest measure -- which party's candidate people voted for in 2004. Using data from the (flawed) 2004 exit poll, we can see an obvious pattern. As voter income increases, their likelihood of voting for President Bush tended to increase, while the opposite pattern generally held for John Kerry:
The fact that Bush received more support in states with lower incomes does not imply that the pattern of increasing GOP support at higher levels of income doesn't hold within each one of those states (in mathematical terms, the relationship could be approximately linear with differing intercepts by state). Bowyer also neglects many other factors that would have to be considered in a proper analysis, including variations in turnout by income level. In short, the article is a mess, and National Review Online should be embarassed to have published it.
Hmm. Can't you read that graph as evidence that the economy is doing fine, as there are enough rich people to vote Bush in?
Posted by: rone | July 12, 2005 at 10:24 PM
Thank you for the ecological fallacy link. Professor Ratcliffe's web site looks very interesting, and I intend to spend some time going through it.
Posted by: Michael Koplow | July 13, 2005 at 09:35 AM