On Super Tuesday, over 14% of voters in the Oklahoma Democratic Primary voted for neither Hillary Clinton nor Barack Obama, despite their being the only two major Democratic candidates left in the race. This was nearly twice the percentage of “someone else” votes as the next highest state (Arizona, with 7.2% of “neither” votes), and four times greater than Super Tuesday primary states over all (3.74% “neither” votes).

John Edwards received the lion’ share (10.14% of the overall vote) of the ‘neither” vote, and his supporters represented a significant opportunity for both Clinton and Obama. By examining where Edwards did well, both candidates could try and appeal to these “neither Clinton nor Obama” voters.

A review of exit polling from West Virginia shows that Hillary Clinton took advantage of the opportunity to appeal to Edward’s voters, and it was her success among those voters that made the difference between her 24 point win over Obama in Oklahoma, and her 41 point margin over Obama in West Virginia. Obama not merely failed to attract the support of Edwards voters, he actually lost support in the demographic categories where Edwards did best.

NOTE: A full set of charts and data for this article can be found at http://www.glcq.com/election08/edwards.htm

THE EDWARDS VOTE IN OKLAHOMA
Based on exit polls, there were a number of demographic/issue categories where John Edwards did better than his overall 10% of the vote in Oklahoma. As CHART ONE shows, Edwards did well among White voters, and especially well among White Female voters, and Female voters in general. He also did especially when among voters earning 50K-75K, White voters 60 and older, and all voters between 45-59, as well as among voters who were high school graduates but did not pursue additional degrees.

CHART ONE:

“Somewhat conservative” and “moderate” voters were also attracted to Edward in Oklahoma, as well as Rural voters, and “non Catholic Christians”. Edwards voters considered Gender an important factor, though the economy was doing well, were concerned with the war in Iraq and health care, and did well among voters who thought that “caring about people like me” was important.

Hillary Clinton did significantly better in all of these categories in West Virginia, while Barack Obama lost support in nearly all of these categories.

GENERAL DEMOGRAPHICS (Race, Gender, Age)
While Obama had failed to attract much support in Oklahoma in those demographic categories where Edwards did well, he did even worse in those same categories in West Virginia despite having more than three months to find a way to appeal to Edwards supporters. Obama’s support among women was down 8% overall, and among “middle aged” voters (those who are 45-59), a category in which he’d done well in Oklahoma (losing that demographic by only 4% to Clinton in OK) his support dropped precipitously (from 39% to 24%).

CHART TWO

Clinton, on the other hand, was able to gain considerable support in all of Edwards key general demographic categories, including the “Female” and “45-59” demographic. But perhaps most notable is Clinton’s success among “white male” voters. Edwards had done especially well among voters who considered gender “one of several” factors in determining their vote (see CHART ONE), a demographic that comprised 16% of Oklahoma voters. Clinton’s efforts to appeal to Edwards voters resulted in the “gender a factor” vote declining to only 11% in West Virginia, allowing her to gain 8% more “white male” support in West Virginia.

CHART THREE

While Edwards support in Oklahoma was disproportionately “White”, this is unlikely to be due to racism. Rather, it is a reflection of Obama and Clinton’s strengths in the 18% of Oklahoma that were “non-White”. (If one assumes that Obama attracted 80% of the African American vote, extrapolating from the exit polling data Clinton did better than Obama among Hispanic and “Other” voters by more than 3 to 1).

SOCIAL DEMOGRAPHICS
In key “social” demographic categories where Edwards did especially well in Oklahoma, Obama was either unable to attract additional support in West Virginia, or actually lost support. Edwards support included 19% of Oklahomans who were high school graduates but had not attended college, and 26% of Oklahomans who considered themselves “somewhat Conservative”. Despite large numbers of voters being “up for grab” in these two categories, Obama did not noticeably improve his support with those voters, while Clinton made considerable gains (+17% among high school graduates, +23% among “somewhat Conservative” voters) in those categories.

CHART FOUR

Obama also lost support in “social demographic” categories where he and Edwards had both done relatively well. The $50k to $75k income bracket is “middle income” in both Oklahoma and West Virginia. (US media income for a family of four in 2008 is estimated at $67,019, in Oklahoma its $53,138, and in West Virginia its $52,292.) Obama attracted 6% fewer of these voters in West Virginia than in Oklahoma, while Clinton gained 14 points in that demographic.

Obama also lost considerable support (-8%) among Moderates, and “non-Catholic Christians”, while Clinton saw considerable gains in both categories (Moderates +13%, non-Catholic Christians +12%).

CHART FIVE

The rural vote in both Oklahoma (39%) and West Virginia (53%) was sizeable, and Edwards attracted 14% of those voters. Obama was unable to make inroads in this key demographic, merely maintaining his very poor 22% of their votes. Clinton, on the other hand, was able to attract 13% more rural voters.

ISSUES
As noted above, Edwards did very well among those voters who considered Gender to be one of several factors that nfluenced their decision in Oklahoma. Overall, 16% of Oklahoma voters considered gender a factor, but in West Virginia,that percentage declined to 11%, and Clinton gained male support while Obama lost male support.

Both Obama and Edwards did relatively well in Oklahoma among voters who thought that the economy was is good shape, but in West Virginia Obama lost considerable support in that group, while Clinton gained in that category.

CHART SIX

Edwards also scored well in Oklahoma among people who thought that the most important quality in a candidate was that the candidate “cared about people line me." Obama lost 40% his support in that category, while Clinton gained 32 points there. And Edwards supporters in Oklahoma identified Health Care and Iraq as their most important issue. Obama was unable to pick up any of the Edwards supporters who were concerned about Iraq, and lost support among those who were concerned about health care, while Clinton picked up additional support in both categories.

CHART SEVEN

CONCLUSIONS
On Super Tuesday, over 600,000 voters in primary states indicated their dissatisfaction with the choice of Hillary Clinton and Barack Obama as thei Democratic nominee by voting for somewhere else. Nowhere was this dissatisfaction more evindent than in Oklahoma, where despite having withdrawn from the race, John Edwards attracted 10% of the vote.

These Edwards voters represented a danger for the Democratic Party – voters who could easily defect to McCain in November regardless of whether the party chose Obama or Clinton. They also represented an opportunity for both Obama and Clinton to increase their support. Only Hillary Clinton appears to have made the effort to attract Edwards voters in three months between Super Tuesday and West Virginia primary on May 13th. Not only did she succeed in attracting many Edwards voters, but also cut into Obama’s support among those who were concerned with the same issues that Edwards voters had been.

Indeed, its not unreasonable to speculate that much of Obama’s support on Super Tuesday came from voters who had supported Edwards initially, and chose Obama over Clinton when Edwards dropped out less than two weeks before Super Tuesday – and that the loss of support for Obama in Edward’s key demographics were Edwards supporters whom Clinton had successfully appealed to.