Hillary Clinton, Barack Obama, The “Woman Factor” and the Value of Empiricism in the 2008 Race

The 2008 presidential race among Democrats is distinctive because for the first time, a woman is seen as one of the leading contenders for nomination. This has led to a whole lot of supposition amongst the pundits and consultants who make their living by making claims up out of thin air and then sounding authoritative when proclaiming them to be true. James Carville, a political consultant long associated with the Clintons, recently wrote an editorial proclaiming the advantages of a Clinton 2008 presidential run. In that editorial, he made the following claim about the role of gender in politics, and how it would affect Hillary Clinton’s chances:

Pundits and fundraisers and activists may be unsure of whether Hillary can get elected president, but Democratic voters, particularly Democratic women and even independent women, are thrilled with the idea.

The X factor for 2008 — and we do mean X — is the power of women in the electorate. Fifty-four percent of voters are female. George Bush increased his vote with only two groups between 2000 and 2004: women and Hispanics. Bush got 49 percent of white female voters in 2000 and 55 percent in 2004. Of his 3.5-percentage-point margin over John Kerry, Bush’s increase with women accounted for 2.5 percentage points. The rest came from a nine-point increase among Hispanic voters: from 35 percent in 2000 to 44 percent in 2004. We believe that Hillary is uniquely capable of getting those swing voters back to the Democratic column.

Hillary’s candidacy has the potential to reshape the electoral map for Democrats. Others argue they can add to John Kerry’s 20 states and 252 electoral votes by adding Southern states, or Western or Midwestern, depending on their background. Hillary has the potential to mobilize people in every region of the country.

Certainly she could win the states John Kerry did. But with the pathbreaking possibility of this country’s first female president, we could see an explosion of women voting — and voting Democratic. States that were close in the past, from Arkansas to Colorado to Florida to Ohio, could well move to the Democratic column. It takes only one more state to win.

Carville’s not alone. Robert Kuttner of the Boston Globe agrees with James Carville when he writes, “Presumably, America’s women will flock to Hillary.” Sheesh — someone needs to tell James Carville and Robert Kuttner that her name is “Hillary Clinton,” not “Hillary.” But to get back to the point, I’m interested in the word “presumably.” Presumably? What makes Kuttner’s a reasonable presumption?

Carville and Kuttner assume that Hillary Clinton’s candidacy will gain the support of women because she descriptively represents women. Descriptive representation, as Hannah Pitkin noted in 1967, is the sense in which a politician represents a class of citizens because he or she resembles them. But does descriptive representation lead to substantive representation, the match between a citizen’s policy priorities and a politician’s policy priorities? And how do people decide who to support — primarily by the descriptive appearance of a candidate, or more by the candidate’s positions?

These are empirical questions, and Carville and Kuttner aren’t being empirical when they make their claims. They’re supposing and assuming and presuming. Rather than play pundit and pull a conclusion out of my posterior, I’d like to look at some actual data. Since November 3, 2004 Irregular Times has been offering Election 2008 bumper stickers, shirts and campaign buttons expressing support for the possible Democratic presidential contenders. And since we’ve been selling these items, we have also been paying attention to the number of items that we sell for each campaign. Instead of the weak and easy indicator of an opinion given over the telephone to a stranger, our own system for tracking candidates measures the kind of support that counts – whether ordinary Americans are willing to spend money to support a particular candidate in a visible way.

Yesterday, I went back through the data for the months of February through June, 2006, and looked at the gender (measured by first name, and omitting ambigous first names — apologies to the Pats of the world) of those who indicated their support for the various 2008 Democratic contenders. What percent of items supporting each candidate were bought by women? What percent were bought by men?

To ensure enough numbers for a reliable result, I looked only at candidates garnering at least a 5% share. Here’s what I found:

For all the fuss people make about Hillary Clinton being the “woman’s candidate” for 2008, there isn’t much empirical evidence of it in our results. The distribution of her supporters into men and women is pretty much the same as the sex distribution of all people buying 2008 electoral gear: about 3/5 men and 2/5 women. Al Gore actually does slightly better in attracting support from women than Hillary Clinton does, and Joseph Biden and Barack Obama attract women’s support much more strongly than Hillary Clinton does. Only a single case, Russ Feingold, fits the “descriptive representation” pattern, attracting his supporters disproportionately from men; the other four candidates either show little gender pattern (Clinton and Gore) or the opposite gender pattern from what Carville and Kuttner predict. Just being a woman does not mean that one’s supporters will be primarily women, and just because a candidate is a man doesn’t mean that women will not be particularly drawn to him.

In conclusion, if James Carville really means what he says about the importance of a “woman’s candidate,” he should support a White House run by Barack Obama, not Hillary Clinton. If Carville didn’t really mean what he said, and is just shilling for a political patron, then he shouldn’t be believed or even paid attention to. Instead of supposition, we should pay attention to empirical information when thinking about the 2008 candidates and the 2008 election.