Race, Polling and the 'Bradley Effect'
Barack Obama's clinching of the Democratic presidential nomination ensures that race will play an outsized role in the November election.
Or maybe not.
Understanding how large (or small) a factor race will be in the general election this fall is central to any attempt to handicap the campaign. Will a significant number of white voters not support Obama solely because he is black? Or is race as an issue simply a media-created phenomenon that misses the fact that the heyday of racial politics has come and gone?
Answering those questions is made doubly difficult by the fact that polling may be a somewhat deceptive indicator in this case. Polling in past elections involving a black nominee have tended to undersell the level to which race negatively impacts voting -- particularly among whites. That is, a black candidate tends to underperform his or her polls on Election Day, as some voters who may have told a pollster they would support an African American candidate ultimately decide against doing so.
So prevalent is this underperfomance by black candidates that it has been dubbed the "Bradley effect" -- named after former Los Angeles Mayor Tom Bradley (D) who, despite polls showing him with a solid lead over Republican George Deukmejian, lost the California governor's race in 1982. Post-election studies showed Bradley, who is black, taking a smaller portion of the white vote on Election Day than public opinion polls predicted.
In other races involving a black candidate -- most notably Charlotte Mayor Harvey Gantt's (D) candidacies against Sen. Jesse Helms (R-N.C.) in 1990 and 1996 as well as L. Douglas Wilder's (D-Va.) victorious run for the Virginia governor's mansion in 1989 -- the Bradley effect came into play.
Given that past history, it's tough for anyone -- The Fix included -- to know what to make of polling that suggests race will not be a major voting issue this fall.
Take a new nationwide survey by Gallup. Asked whether Obama's race made them more or less likely to back him in November or made no difference in how they would vote, nearly nine-in-ten white voters (88 percent) said it would have no impact on their vote. (Roughly three-quarters of black voters said the same.)
Those numbers are entirely consistent with a slew of other national polling on the race issue.
Eighty-seven percent of white voters said they would be comfortable with a black president, including 66 percent who said they would be "entirely comfortable" with the idea.
Seeking to better understand how race influences voting, Gallup asked a series of other questions in the recent survey that are worth reviewing. Asked whether Obama's being black would gain him more votes than it would lose him, there was less unanimity among the Gallup sample. While the largest portion of white voters (47 percent) said Obama's race would "make little difference", one in every four white voters said it would cost Obama more votes than it would gain him, while 21 percent said the reverse.
Another question asked how likely it was that each party would use race as an issue in the campaign. Fifty percent of white voters said it was either very or somewhat likely Democrats would use it, roughly the same number (49 percent) of white voters who said the same of Republicans.
(Interestingly, black voters agreed with their white counterparts on the likelihood of Democrats using race as an issue -- 48 percent said it was likely. But they broke with whites on whether Republicans would use skin color as an issue; 43 percent of black voters said Republicans were "very" likely to use race as an issue and 27 percent called it "somewhat" likely -- for a total of 70 percent.)
While very few white voters are willing to acknowledge that race may play a central role in their decision-making this fall, the surveys show that a slightly larger percentage admit that not only will race be an issue in the presidential election but that it has the potential to lose voters for Obama.
Gallup polling director Frank Newport writes: "The data reviewed here suggest that while most voters deny that Obama's race affects them personally, many do think it will have an effect on the race overall -- either helping Obama or hurting him. Still, there is no consensus on the part of the average voters on whether Obama's race will be a net plus or a net minus."
The truth is that all polling having to do with race seems inadequate to understanding the true effect of having an African American candidate on the ballot will have in the fall. Someone who won't vote for Obama because he is black is extremely unlikely to tell that to a stranger over the phone. No pollster -- no matter how skillful -- can get around that roadblock.
Because of the inability of polling to provide hard data on this issue, it is almost impossible to accurately answer the question of whether race will be a front and center issue for a significant segment of voters or whether it will be either a minor consideration (or no consideration at all) for most voters.
What we do know is that Obama, by becoming the first black nominee for president from a major party, will force voters to wrestle with their views on race in America in a way that many have not done yet in their lives. Does this very personal debate accrue to Obama's benefit or detriment? We won't likely know until November 4.
Polling analyst Jennifer Agiesta contributed to this report.
June 10, 2008; 12:30 PM ET
Categories: Eye on 2008 , Parsing the Polls
Save & Share: Previous: How Many House Seats Will Be in Play This Fall?
Next: Clinton Donors Urged To Back Obama
The comments to this entry are closed.