Are voters irrational?

ELECTIONS ARE SUPPOSED to be about determining the will of the people. Citizens throw their weight behind the candidate who best reflects their own values and interests -- at least, that's the story that civics teachers tell.

But what if voters don't know their own interests? That's the claim of Thomas Frank, editor of the left-leaning cultural magazine The Baffler and author of the new bestseller "What's the Matter with Kansas: How Conservatives Won the Heart of America" (Metropolitan). "The preeminent issue of our day," he writes, is "people getting their interests wrong."

For evidence, Frank points to McPherson County, Neb. -- the poorest county in the country, by one measure -- where voters went for George W. Bush in 2000 by a ratio of better than four to one. Enflamed by issues like abortion and gun control, voters there turned their backs on the traditional party of the working class and voted for the candidate of Wall Street.

Frank's thesis has its critics, but he may be on to something, if the work of political scientist Larry Bartels is any indication. In fact, Bartels finds even more confusion among the electorate than Frank does. Voters, in Bartels's version, know their own interests and articulate firm political views. Then they go ahead and vote against them anyway.

In a paper called "Homer Gets a Tax Cut" (the unflattering reference is to TV's uber-dolt Homer Simpson), which Bartels summarizes in the June American Prospect, the Princeton professor points out that, according to a national survey conducted in 2002, 74 percent of Americans believe (correctly) that the income gap between rich and poor has grown over the last two decades. A majority believes that such growing inequality is a problem. Two-thirds say that government policies cater to the wealthy. Fewer than 15 percent say the rich pay too much in taxes.

But then comes the weird part: Fully 70 percent of Americans went ahead and supported President Bush's repeal of the estate tax -- a tax that affects only estates of more than $1 million, or fewer than two percent of the total.

Even those who think the rich should pay more in taxes supported scrapping the estate tax (66 percent of this group supported repeal), as did those who think government policy contributes to inequality (67 percent pro-repeal). In short: Voters say they want to soak the rich, then they vote to give them a bonus.

For more than a decade Bartels has been collecting examples of public-opinion bizarrities that represent "challenges to the most fundamental assumptions of democratic theory." In one paper, he showed that whatever their interests, voters hold incumbents responsible for bad news -- even acts of God. In 1916, Bartels argues, Woodrow Wilson lost votes in several seaside New Jersey counties because of . . . shark attacks. Similarly, he writes, Al Gore may have lost in 2000 because Midwestern farmers suffered droughts and floods while he was vice president.

In the case of the estate tax, Bartels concludes that it's an example of "misplaced self-interest": Homer thinks his taxes are too high, so he votes to reduce Mr. Burns's.

. . .

THE SAT AND IQ. Dozens of readers wrote in to complain that a formula published here on July 4, designed to convert SAT scores into IQ scores, seemed to go haywire for combined SAT scores greater than 1300.

The formula, applicable to scores from 1996 or earlier, translated a 1300 into an IQ score of about 124. A score of 1350 also translated as 124. Then, oddly, IQ began to slowly drop, until a 1600 registered as 121. "Brain deterioration from too much studying?" e-mailed one wag.

Douglas Detterman, the Case Western University psychology professor and co-concoctor of the formula, says the particular IQ test on which it was based doesn't do well at sorting out test-takers at the very high (and low) ends. Additionally, by their nature the mathematical techniques used to establish correlation tend to underpredict at the high end and overpredict at the bottom.

Finally, the College Board, a nonprofit association of schools and colleges, wants it made clear that it alone is responsible for the content of the SAT. The Educational Testing Service -- to which I attributed statements about the SAT -- is a contractor to the College Board.

Christopher Shea's column appears in Ideas biweekly. E-mail: critical.faculties@verizon.net.