Friday, September 19, 2008

Will Tar Heel Whites Vote Black?

In teaching my U.S. Campaigns & Elections class this fall, I constantly remind my students that as a social scientist, I (and they) study human behavior, and that's pretty dangerous. Because at some point, our data (i.e., human beings as voters) will lack reliability (read: people lie). I know, I know, how dare someone point out that human beings would be deceitful and not give honest answers when asked by a social scientist. Well, here's some evidence from a recent poll that will give pause to how we are interpreting this fall's election, especially in North Carolina.

To put my argument (that human beings lie) into context, here's the results of two questions asked on a New York Times poll taken over the summer, when both nominees were well known to the general public:

Do you think most people you know would vote for a presidential candidate who is black, or not?

  • Would: 69%
  • Would not: 16%
  • Don't know/no answer: 15%

Would you personally vote for a presidential candidate who is black, or not?

  • Would: 90%
  • Would not: 6
  • Don't know/no answer: 5%
So while "most people" would personally vote for presidential candidate who is black, some of us know of "someone" who would not vote for a presidential candidate. So while we're all willing to vote for a black candidate, there are some of us out there who are not.

In a recently-released Elon University poll, a similar set of questions were asked of North Carolina respondents (411 respondents, with a margin of error of 4.9%):

All things being equal, would you rather vote for a black person, a white person, or would a presidential candidate's race make no difference to you?

  • Vote for a black person: 0.7%
  • Vote for a white person: 2.9%
  • Race doesn't make a difference: 95.9%
  • Don't know: 0.5%

Do you know people that will not vote for a presidential candidate is who is black?

  • Yes: 54.7%
  • No: 43.3%
  • Don't know: 1.9%

So, in North Carolina, half of us know someone who won't vote for a black presidential candidate, but it ain't us. Sound fishy?

Well, in the study of white voting behavior when it comes to black candidates, there are two schools of thought. As noted in his excellent study of Changing White Attitudes toward Black Political Leadership, political scientist Zoltan L. Hajnal wrote that while some scholars point to the public opinion polls that report white respondents willing to vote for black candidates and that the number of black elected officials has increased in primarily white areas (states, cities, etc.), the other side of the scholarly fence point to the fact that when there is a white and a black candidate in the same election, whites vote for their own.

For example, in 2006, the U.S. Senate race in Tennessee featured Bob Corker (white Republican) and Harold Ford (black Democrat). In the exit polls from that race, white respondents went 59% for Corker to 40% for Ford. Of course, this was a particularly nasty election battle (the television ads were particularly fierce against Ford), with Corker winning the seat 51% to 48%.

What Hajnal found in his study is that when white voters see black political officials in office, they find themselves more comfortable and will vote for the black incumbent, as opposed to when a black challenger is first running for elected office. As this is the first presidential election with a black candidate as a party nominee, it will be interesting to see how white voters respond and react.

In North Carolina, one can see a partial test of this scenario, in the guise of the May 5th Democratic presidential primary between Obama and Sen. Clinton. Sen. Obama carried the major metropolitan counties along the interstates with over 55% of the vote, along with the "majority-minority" counties in the north-eastern part of the state and stretching along NC's "black-belt" counties. However, Sen. Clinton won significant portions of the white majority NC mountain counties with 55% or more of the vote in these conservative counties (only Buncombe, with UNC-Asheville, and Watauga, with App State University, counties went more than 55% for Obama).

If Sen. Obama is to win North Carolina, some estimates indicate that he needs at least 36% of the white vote. This will be a critical test to see if there is truth in the statement, "whites vote black."