Is the Democratic Party the party of the rich? Many conservative commentators think so.

Or is the Republican Party the party of the rich? Many liberal commentators think so.

The conservatives tend to rely on evidence at the state level. Michael Franc of the Heritage Foundation recently wrote that "Democrats now control the majority of the nation's wealthiest congressional jurisdictions. More than half of the wealthiest households are concentrated in the 18 states where Democrats control both Senate seats."

This view squares with maps drawn after the 2004 election showing a remarkable correlation between states that have above-average median household incomes and states that voted Democratic. Minnesota, of course, fits the pattern: above average household income, blue state.

The liberals, on the other hand, rely on evidence at the individual voter level. Paul Krugman, for example, in a blog posted last month cited exit polls from 2006:

Among voters earning less than $100,000 (78 percent of voters), 55 percent said they voted Democratic, 43 percent Republican. Among those earning $100,000 or more, 47 percent voted Democratic and 52 percent Republican.

And the fact that people with higher incomes are more likely to vote Republican has been consistently true since 1972, Krugman wrote.

The one thing both sides seem to agree on is this: being the party of the rich is a bad thing, something you try to pin on your opponents.

Actually, both sides' arguments are correct. But how could this be? If the rich tend to vote Republican, why do the rich states vote Democratic? It seems like a contradiction.

A fascinating explanation is presented in a paper published recently by four statisticians, including Andrew Gelman, a Columbia professor. Gelman writes a blog with the snappy title, Statistical Modeling, Causal Inference, and Social Science.

The title of the paper is a bit livelier: "Rich state, poor state, red state, blue state: What's the matter with Connecticut?"

Gelman and his colleagues show that in presidential elections since 1976, "richer states have increasingly favored the Democrats." On the other hand, "higher-income people have been consistently more likely to vote Republican, especially since 1970."

Even in blue states, they find, the wealthier people favor Republicans. But here is the surprising finding that explains the contradiction:

In poor states, like Mississippi, wealthy people are overwhelmingly voting Republican. In rich states, like Connecticut, the link between wealth and voting Republican is much weaker, barely correlating at all. (Minnesota falls pretty close to halfway between these extremes.) Differences in racial composition among the states explain about half the phenomenon, the researchers found. The other half remains unexplained.

I exchanged emails with Gelman, asking him why wealthy people have different voting patterns, depending on whether they live in a poor state or a rich one.

"We're still thinking about it," Gelman wrote. "One issue is that, in poor states, it's the rich people who are more religious, but in rich states, the rich people are less religious. Thus in Mississippi, religion and income go together, but not in Connecticut." Being more religious also correlates with voting Republican.

Got a theory? Why are wealthy people in poor states so likely to vote Republican, while wealthy people in wealthy states are almost equally divided between the parties? You don't need to be a statistician to weigh in, just a registered commenter.

Joel Kramer is MinnPost.com editor and CEO.