The objection may be made that I am basing my conclusions on polls rather than actual election results. In the races for governor in California in 1982 and Virginia in 1989, preelection polls seem to have understated the percentages ready to vote against black candidates Tom Bradley and Douglas Wilder. But those elections were held 26 and 19 years ago. And we did not see a similar effect in most Democratic primaries this year: It was Obama's vote that was understated in pre-primary polls in New Hampshire.
Exit polls taken on Election Day did tend to overstate Obama's percentage in many states. But that could result from respondent self-selection. Only about half of those approached to take the exit poll do so. Obama voters, with higher levels of enthusiasm for their candidate, may have been more likely than Hillary Clinton voters to go to the trouble of filling out the exit poll. That's consistent with the greater propensity of Obama supporters to participate in caucuses in the four states that held both caucuses and primaries.
On balance I think Obama's race has been a political asset. I believe that most Americans think it would be a good thing, all other things being reasonably equal, for our country to elect a black president. I know I feel that way myself. I think that impulse has inspired many voters, ever since his speech at the 2004 Democratic National Convention, to give Obama a sympathetic look-over, to be readier perhaps to appreciate his strengths and to overlook his weaknesses than they might be with an otherwise similar non-black candidate. The refusal of a very small number of voters to support a black candidate does not, I think, offset this significant advantage. The Obama candidacy is indeed a test -- a test not of American voters, but of Barack Obama.