For supporters of Democratic presidential candidate Barack Obama it is a nightmare scenario—his apparent lead in the battle for the White House suddenly evaporates on Election Day. The cause? Race.
As Republican rival John McCain celebrates victory, it emerges that a small but decisive percentage of white voters who had declared to opinion pollsters they supported Obama actually chose differently in the privacy of the ballot booth.
With opinion surveys making Obama the favorite over McCain less than three weeks before the November 4 election, attention has turned to the question of how many white Americans might be lying to pollsters about their willingness to vote for a black president.
The phenomenon is known as the “Bradley effect,” after Tom Bradley, an African American who narrowly lost the 1982 California governor’s election despite leading in polls.
His defeat surprised observers who concluded many white voters had not been honest about their intentions. Ever since, pollsters have tried to factor in the Bradley effect in elections featuring black candidates.
Analysts counter that since the 1970s, surveys have shown a sharp decline in the number of voters who say they would not vote for an African American for president.
Recent surveys have indicated that Obama’s race is less of an impediment to voters than McCain’s age. At 72, McCain would be the oldest person to assume the U.S. presidency.
FACTORING IN RACE
The analysts also note that few people who are turned off by a black candidate’s race would likely vote Democrat anyway.
Politicians rarely address it directly and both the Obama and McCain campaigns downplay the idea the some whites might not be speaking honestly about their voting plans.
The diversity of the U.S. electorate makes it hard to quantify the phenomenon, pollsters say. An older voter in South Florida may be of the same race as a young union member in Virginia or a businesswoman in Colorado but share little else.
The Democratic primaries in which voters in the states picked the party’s presidential candidate painted a mixed picture of the potential for a Bradley effect. In some states, Obama got fewer white votes than expected but in South Carolina and elsewhere it was the opposite.
Commentator Earl Ofari Hutchinson said there was a potential for the Bradley effect to be decisive, and that Obama might need a 10 percent lead in opinion polls to be sure of overcoming it.
Pollster John Zogby said he took the prospect of overstating white support for Obama very seriously.
Zogby tries to factor any potential Bradley effect into his polling but said he believes it will be negligible, in part because estimates of Obama’s support had proved accurate during the primaries.