Skip to main content

Obama’s racial penalty

Claude Fischer, professor of sociology | October 17, 2012

Barack Obama has run his presidential races with an extra weight on his shoulders: being black. Sure, there are some pundits who claim that he benefits from his race – black loyalty, white guilt, and such – but serious scholars understand that his race has been, in net, a notable disadvantage. My rough sense from looking at some of the political science analyses of the 2008 campaign is that he may have gained about 1 percent in the final vote by garnering more black support than a white Democrat would have gotten, but that he lost about 5 percent of the vote by getting less white support than a white Democrat would have gotten – for a net minus of about four. Four points in a presidential race is a lot.

Obama's image lightened, unaltered and darkenedAs an historical matter, it is striking that Obama’s racial penalty has not been higher. As a political matter, the question of whether the penalty will still be that high in 2012 may make all the difference in who wins the White House. Or maybe, one study suggests, the penalty might be even higher.

(The picture here, by the way, is from a 2009 study showing that the less favorable on-line respondents were to Obama, the more likely they were to pick the darkened picture of him as the accurate one.)

New days

That a black man was elected president in 2008 would be jaw-dropping news to an American of 1908. Aside from the general racism in American society, laws throughout the South kept blacks from voting and cultural repression elsewhere dampened their political participation. Yet, a century later Obama carried even Virginia and North Carolina.

Survey researchers have tracked attitudes about a potential black president since the mid-twentieth century, asking Americans a question on the order of, “If your party nominated a (Negro/ black/ African-American) for President, would you vote for him if he were qualified for the job?” The graph below shows the vast change. In the 1950s, most respondents told interviewers “no”; in 2010, about four percent of white respondents said “no.”

graphScholars debate what to make of answers to this sort of question. All the trend may show, some argue, is that over the years whites learned to say the right things to interviewers. (See here, here, and here for examples of researchers’ discussions.) Certainly, putting on appearances is part of the story. But so is real change. And Barack Obama is the proof.

Is, then, the glass 96 percent full or is it 4 percent empty? Empty in that, while a black man sits in the Oval Office, 4 percent of whites would tell interviewers face-to-face that they would rule out a black candidate because he is black? Even were it only 4 percent, that 4 percent matters.

2008 and 2012: Less race?

Between 2008 and 2012, voters who thought that Barack Obama, because of his race, would fail or would favor blacks had a chance to watch him operate. Most observers, even critical ones, do not describe his as a radical or Afrocentric administration, although some on the far right may. Has Americans’ experience with a black president made a difference?

Political scientists Charles Tien, Richard Nadeau, and Michael S. Lewis-Beck recently reported a survey study that compared early-2012 estimates of Obama’s racial penalty to the one they calculated just before the 2008 election (here). In 2008, focusing on those survey respondents who said that they expected candidate Obama to favor blacks if he was elected, the researchers generated a 5-point estimate of the racial penalty – the percentage of voters who would have voted Democratic if the Democratic candidate had been white. That calculation seemed prescient. Although Obama won the 2008 election with 53% of the popular vote, given the economic situation and the large victory that Democrats scored in the 2008 congressional elections, history suggests that Obama should have had won in a landslide, with perhaps 58% of the vote.

Replicating their 2008 survey in 2012, Tien and colleagues predict that the “price” Obama will pay in November will be about 3.3 points – less than in 2008 but consequential nonetheless. (By the way, perhaps Romney faces a “Mormon penalty.” I have not come across serious statistical analyses of that possibility, but they may exist.)

One important question for close followers of the campaign is whether this racial penalty is already displayed in the pre-election polls. Or do racially-biased poll respondents hide their bias until election day – the “Bradley Effect”? In 2008, the sophisticated survey aggregators (Silver’s “Five-Thirty-Eight” and Wang’s “Princeton Election Consortium”) provided predictions that were within about 0.3 points of Obama’s final percentage (see here), suggesting that the racial penalty is probably already showing up in the polling data.*

Or more race?

Some researchers have argued that the election of Barrack Obama has paradoxically rekindled the racial animus previously suppressed in American politics. For example, in one study, telling subjects about Obama’s support of health care reforms stimulated the racially prejudiced among them to oppose those reforms; not as many responded that way when experimenters described the reforms as Bill Clinton’s ideas (here).

Brown University political scientist Michael Tessler recently analyzed the revival of what he calls “old-fashioned racism” in American politics (pdf). Drawing on variety of surveys, Tessler shows that straightforward racist thinking – as in, for example, expressing opposition to interracial dating (in contrast to “new-fashioned racism” such as objecting to political demands by blacks) – became more consequential during and after Obama’s election. In the prior several years, white respondents who objected to interracial dating were, once their backgrounds and political ideologies were taken into account, not much more likely to say they voted Republican than did white respondents who answered that interracial dating was OK. In 2008, however, the old-fashioned racist respondents were much less likely, other things being equal, to vote for Obama than McCain – or than to support Hillary Clinton. And in 2010, the old-fashioned racists, all else being equal, became much less likely to vote for Democratic candidates generally than they had been before.

While many observers hailed the election of Barack Obama as ushering in a “post-racial” age, it appears that having a black face be the face of the Democratic party has helped to further polarize, on a racial basis, Americans’ division by party.

If Obama’s presidency has rekindled racial animus among a small but swing group of voters, perhaps his “race penalty” might be even greater, not lesser, this year than it was in 2008.


* Wang and Silver differ, at least this year, on this major methodological point: Silver uses polls plus a lot of other, mainly economic indicators to predict the election. Wang insists on using only the polls, arguing that all these other factors already manifest themselves in the polls. Whatever the general issues, to the extent that the racial penalty appears only in the polls, using other indicators to predict the election may underestimate the racial effect.

Cross-posted from Claude Fischer’s blog, Made in America: Notes on American Life from American History.