Since the economy went into a tailspin several weeks ago, Barack Obama has been surging in most of the national public opinion polls. His lead has recently stabilized between four and eight percentage points over Republican Presidential candidate John McCain with just three weeks to go in the campaign.

However, a question remains whether all these public opinion polls are accurate. It's a question that must be asked because Barack Obama is the first black Presidential nominee and because of the possibility that social desirability bias will have an effect on the results of public opinion polls .

Social desirability bias is a term used in scientific research to describe the tendency of respondents to reply in a manner that will be viewed favorably by others. It has become a possible explanation for a frequently observed discrepancy between voter opinion polls and actual election outcomes when a white candidate and a non-white candidate run against each other in political campaigns. It is a term that could apply in this election year as voters of both races respond to questions from various public opinion pollsters and polling organizations.

Consider that there is even a political term used to describe the discrepancy between a large lead a black candidate has in pre-election public polls and a result that was much different on Election Day. The "Bradley Effect" refers to a tendency on the part of voters to tell pollsters that they are undecided or likely to vote for a black candidate, and yet, on Election Day, vote for the white opposition candidate instead.

The "Bradley Effect" is named after African American Tom Bradley who lost the 1982 California's Governor's race despite being well ahead in voter public opinion polls. In fact, exit polls that were taken on Election Day, showed a significant Bradley victory by nearly ten percent of the vote.

Based on those exit polls, a number of media outlets even projected Bradley as the winner. The next day, early editions of the San Francisco Chronicle would feature a "Bradley Win Projected." headline. However, Bradley actually lost in a close vote and post-election research indicated that a smaller percentage of white voters voted for Bradley than polls had predicted. In addition, previously "undecided" voters had voted against Bradley in statistically anomalous numbers

Over the years, there have been other races which have been cited as demonstrations of the Bradley effect. These races include the 1983 race for Mayor of Chicago, the 1989 race for Mayor of New York City, and the 1989 race for Governor of Virginia. In each of these races, the black candidates barely won the election after public opinion polls showed them holding large double digits pre-election day leads.

The fact is that we may already seen a Bradley effect in this election in the Democratic primary between Barack Obama and Hillary Clinton. After the Super Tuesday elections of February 5, 2008, researchers from the University of Washington found significant trends suggesting the possibility of public opinion polls overestimating Barack Obama's support in states with black populations below eight percent.

University of Washington researchers also found a possible "reverse Bradley effect" with public opinion polls underestimating Obama's support in states with a black population in excess of 25% of the total population.

Certainly, there are many factors to consider in addition to a "Bradley effect" in election 2008. It has been a volatile election year with an uncertain economy. Voters are still evaluating both Presidential candidates and neither candidate has yet completed their 2008 Election campaign. There is still time for public opinion polls to dramatically change.

Indeed, some public opinion polling even suggests that up to 20% of all voters could still change their minds on or before Election Day. As a result, there could well be many surprises in several of the important battleground states on election night.

However, Barack Obama has held a substantial lead in just about every public opinion poll since September. If this trend in the polls continues, the result on Election Day will either be a Barack Obama victory or a national controversy concerning the effect of a candidate's race on public opinion polling.