2008 in review - The Harvey Gantt Problem (June)

The national media called it the Bradley effect - but I always disagreed, because that hypothesis was based on people actually lying to pollsters and that showed less respect for people than I think is deserved.  Still, it seemed a valid question to ask.  This is my favorite thing that I was wrong about in 2008.

Harvey Gantt was once a rising star in the Democratic party in North Carolina.  He was the first African-American admitted to Clemson University, graduating with a degree in architecture and then went on to earn a Master’s Degree from MIT.   After a decade on the City Council of Charlotte, North Carolina, he was elected and then re-elected mayor.  He was a dynamic speaker and ran optimistic campaigns promoting jobs and education.  In 1990 he swept to the nomination for the US Senate, running against Jesse Helms - the conservative icon and unreformed darling of the “Fergit, Hell” wing of southern politics. 

Gantt led in the polls, by as much as double digits, but without a majority, right up to the election.  And then, on election night came the surprising news that Jesse Helms had pulled out the win and was re-elected.  The final polls all showed Gantt with about 47-48% of the vote, and that is where he ended up.  Helms, though, rose from the low forties in polling to pull out a narrow victory, as all of the “uncommitted” votes pulled the lever for Jesse Helms.

In 2008, the question is whether Barack Obama has a Harvey Gantt Problem.  During the Democratic primary season, a debate went on about whether America is more sexist than racist.  The Harvey Gantt Problem is a test of whether America will elect an African-American as President, or whether there remains a subtle and hidden racial divide.  

The thesis of the Harvey Gantt Problem is this - until the polls reflect that an actual majority of people say they will vote for Obama, he will not actually prevail.  The uncommitted votes are going to go against him.

It wasn’t that people lied to the pollsters in North Carolina, saying they would support Gantt and then didn’t.  The 47% or so who said they would vote for Gantt did so.  The Problem is that in the privacy of the voting booth, everyone who hadn’t committed to voting for Gantt voted the other way.  Racism?  Perhaps.  But the psychological analysis of the electorate is for another day.  This question is presented because it is about winning an election.  Or more accurately, about watching to see whether what was once true in America still is.

Since the presidential election isn’t simply a popularity contest, Obama’s Problem isn’t merely about the percentages in the polls.  National polls are entertaining to read, and fodder for the commentariat, but the popular vote isn’t the issue.  Obama could get to 50% in the national polls, and in the election, without winning.  The Electoral College further sharpens the features of the Problem.  

In some ways, the Electoral College makes the thesis easier to track and easier to test in the end.  Current polling data places Obama in a national lead - but not a substantial one, and without a majority.  Recent polls, however, also show that about 35 states can be considered fairly certain to vote for one candidate or the other.  The remaining “swing states” are what this election will be about.  Applying the thesis of the Harvey Gantt Problem, if McCain can prevent Obama from gaining a majority in those states - even where the polling shows McCain behind - then McCain prevails in the election.

Another departure from the North Carolina race of 1990 is the presence of third-party candidates, who may draw votes from one candidate or the other, and so the uncommitted vote may actually go to a third party candidate instead of McCain.  It is possible therefore, for the thesis to remain correct - that Obama receives essentially the same percentage of votes as percentage in the polls - and for Obama to win a state where the uncommitted vote goes to Libertarian candidate Bob Barr.

One of the things to watch for will be attempts - whether by the Republican party, or by 527 groups working to defeat Obama - to subtly inject race into the election.  With an African-American running for the presidency, it is unlikely that we will see a “Willie Horton” ad.  The backlash would be swift and the condemnation universal.  But in a close election, where a few votes based on race could swing a state, something less directly confrontational may air.

In 1990, a TV ad ran in North Carolina called “Hands”   The ad showed a pair of white hands holding a sheet of paper - “you needed that job” intoned the voice-over, “but they had to give it to a minority.”  Crumple paper.  Harvey Gantt favored affirmative action - certainly no surprise for a Democrat in the South in 1990 - and the unsubtle point of the ad was that Jesse Helms would protect white folks from losing their jobs.  But “Hands” wasn’t just about affirmative action.  The social dynamics of 2008 are different enough that such an appeal would have a very limited audience - illegal immigrants are now the bete noir.  The ad was about reminding people to think about race.

It isn’t clear that the Harvey Gantt Problem is real.  No one can hope that it is.  But if it exists, it says all too clearly that we still have a ways to go in overcoming racism.  Seldom do we have such a bright test to measure how far we have come.  

I spent the summer and fall tracking the polls to see what the prediction would be using this hypothesis.  The next post, written the day before the election set up the final question.