Expertinent: The Political Psychology of Race and Gender

Talk about good timing. A week ago, Cornell law student Gregory S. Parks emailed me a law review article that he had just coauthored with university professor Jeffrey Rachlinski. The subject? "Unconscious race and gender bias in the 2008 election." In addition to their legal studies, both Parks and Rachlinski (whose academic efforts have focused on the influence of human psychology on decision-making by courts, administrative agencies and regulated communities) boast Ph.Ds in psychology. On Monday, I decided to call them up for a chat. The next day, of course, race and gender consumed the national conversation (yet again) when Clinton supporter and former vice presidential candidate Geraldine Ferraro told a California newspaper that "if Obama was a white man, he would not be in this position." Revisiting my conversation with Parks and Rachlinski this morning, I realized that many of the questions we covered--who's battling the more difficult biases? is the 'victim pose' politically helpful? what should we expect in the general election?--are precisely the questions that everyone is asking in the wake of the Ferraro flap. Thus, I defer to the experts:

What inspired you to write this article?
RACHLINSKI: There's a growing body of research among social psychologists that normal adults who explicitly embrace egalitarian beliefs--that everyone should be treated equally and that gender and race shouldn't affect their judgments of other people, especially job candidates--nevertheless harbor implicit associations that can hinder their judgment. Something like 80 to 90 percent of adult Americans harbor at least a mild negative implicit bias toward African-Americans, and a good 30 to 40 percent harbor very negative biases.

PARKS: The research on implicit attitudes or unconscious biases suggests that they operate in two different ways, depending on the categories of individuals: blacks or women. With regards to blacks, people tend to have an implicit animus, and it plays out in various forms of behavior. With regards to women, they tend to have these implicit stereotypes in regards to gender roles, particularly in regard to employment--like, who would best fit certain types of roles in the workplace.

RACHLINSKI: There's preliminary data to suggest that this affects ordinary job applicants, and that resumes of black Americans are treated differently than those of whites. It's been proven that credentials help white applicants a lot more than they help black applicants, for example. Because studies are showing that these implicit, unconscious biases affect job candidates, it occurred to us that the 2008 election is really an elaborate job interview. It's a perfect case study. You have two well-funded, very savvy, highly motivated individuals, both of whom stand to suffer from unconscious biases.

How are the campaigns dealing with these biases?
R: Clinton has an easier path in some ways. She faces a straightforward, content-filled implicit bias that women are not leaders. Psychologists often say that there are two kind of judgment. One's the automatic, unconscious system--the intuitive system. And the other is the explicit, slow, deductive, reason-based system. The unconscious biases operate on that first system. So what Clinton has to do--and has done very effectively--is always look like a leader, so when people think of her, they think of her as such. She fights the bias directly, and at really no cost other than the work required to maintain that image. No one in the Democratic Party blames her for looking tough as nails all the time and constantly going on about policy.

How about Obama?
R: Obama has a tougher job. The biases against African Americans are just a raw animus in a lot of ways. What you see in the studies is that people associate black with negative imagery, just wholesale, without regard to specific content. Blacks are bad, whites are good. You see it over and over in the unconscious bias literature. So what does he have to fight? He has to fight against being black in a way. He has to have people look at him and associate him with the positive imagery that Americans tend to associate with whites. It's not surprising, then, that his campaign is about very amorphous goals like hope and aspiration. That's the message that can work, because he can't embrace black issues without activating unconscious biases in white voters. That's very difficult to begin with.

On the other hand, Obama risks raising specific concerns among his core supporters--notably, African-Americans--if he fights too hard against being black. There's a specific in-group favoritism among African-Americans--a favorable, explicit self-image that's stronger than what you see among whites. When a black leader seems to be running away from his image as a black person, that's viewed negatively. In order to keep his base, then, he can't deny that he's black. It's a thin line that he has to toe.

You said before that "credentials help white applicants a lot more than they help black applicants." Does that mean that Obama shouldn't recite specific accomplishments and resume points?
R: The data suggests that it doesn't help black job applicants, and that it wouldn't help him. According to the research, adding resume credentials helps white applicants much more than black applicants. So if his campaign starts to be about what he's done, it won't help.

How do you know that unconscious bias is affecting voters?

R: It's tough to collect data in one election--psychologists like to have multiple, multiple experiments to support their results. But this is a case study. What we say in the paper that you see among white voters is a tendency to sort of flinch when voting for Barack Obama. That's how unconscious biases work. They're that first emotional, unconscious, affective, rapid system that we don't even always have conscious access to. People don't always know why they're doing what they're doing. In a vague sense, maybe--but it's very ill-defined. So it's at the last minute that you see white voters flinching.

How do you measure the flinch?

R: We tie it to the Bradley effect--the tendency for poll numbers to overstate support for a black candidate in a black vs. white election. What we picture is a white voter who sort of favors Obama but goes to the polls and just can't do it at the last minute. Then he's embarrassed about it and he lies to the exit pollsters. How can we tell this is going on? It's a little hard from the data we have. But there's a correlation between the tendency to see a Bradley effect in the 2008 primaries and the percentage of white voters in a given state. In largely black states, you tend to see the opposite--a fair number of African-Americans who show black preferences on implicit associations.

Where are you seeing the Bradley effect?
R: The states that showed the paradigmatic Bradley effect are New Hampshire, California, Massachusetts and Rhode Island. The states that showed the reverse effect are Virginia, South Carolina, Alabama and Georgia.

Let's talk about the future. Will this gender and race dynamic change in the general election if Clinton is the nominee?
R: It changes quite a bit. In the general election, you'll see more concern--if Clinton gets the nomination--with her not being a traditional homemaker. You'll see that explicit bias more among Republicans and Independents than you do among Democrats, because more Democratic women tend, relative to the general population, to be professionals. They've encountered the same kind of stereotypes that she's facing. They're sympathetic when she tries to look tough and not show emotion. Come November, then, Clinton will be forced to appeal to a lot more voters who explicitly embrace the idea of women in the home--which means she may risk undoing her earlier work to fight the implicit bias that women aren't leaders. She'll be the one forced to walk that tightrope.

What about Obama?
R: He faces fewer white voters who like or care about the idea of a post-racial future. Liberal Democrats like the idea that someday race won't matter; Independents and Republicans, not as much. There's good data showing that Republicans harbor stronger negative implicit biases towards African-Americans than Democrats. So he's got to fight those biases a good deal more than he does among Democratic voters, and liberals are no longer enough. The other problem for Obama in the general election is that strong link between "black" and "foreign."

P: There was a study that came out a couple of years ago titled "American Equals White." And what it showed was that at the implicit level people tend to correlate whiteness with Americanness as opposed to blackness with Americanness. What's more, studies of the 2008 election have shown that when you prime individuals with images of the American flag--at a subliminal level, so you just flash is for a millisecond--it has a tendency to make white individuals show less liking toward Barack Obama. This harkens back to question of Obama not wearing the American flag pin and the accusations that he failed to put his hand over his heart during the singing of the national anthem. This stuff is tricky for him, especially considering that some opponents are questioning his patriotism. If images of Americanness make white Americans see Obama as less American at the implicit level--while at the explicit level rivals are questioning his patriotism--then he's damned if he does and damned if he doesn't.

R: And that's more of a problem in the general election than in the primary because he'll be running against a war hero. Hillary Clinton looks nowhere near as "American," in a psychological sense, as John McCain. So the implicit biases that Obama has to fight are a lot harder. One thing that gets easier for him, though. Black voters worried very early on about whether Obama was electable--would whites really, truly support him?--and whether he was "black enough." I think winning a long primary obviously makes him electable. So he gets past that. As far as whether he's authentically black, it's a long primary season. Occasionally showing he's "black" and walking that tightrope seems to be doing the trick. So in the general election, perhaps he can focus more on counteracting implicit biases and not worry as much about proving his authenticity.

Is there anything to be gained by either campaign accusing their opponents of being sexist or racist? It seems to happen every day now. Does the 'victimhood pose' help in any way?
P: Obama, for one, cannot afford to address these things head on. If he gets up and says XYZ is racist and calls people on the carpet about race issues, it will only hurt him. The data supports this view. Studies suggest that when you press people on their gender-stereotypical biases, they kind of laugh it off. Because it's not such a hot issue. They're like, "Whatever. I'm not sexist." But if you press them on their racial biases, particularly in regards to blacks, one of two things happens. If they're low on explicit racial prejudice, they become contrite, apologetic, they want to know what they can do to overcome it. But if they are high on explicit racial bias, they become angry and antagonistic. When you accuse whites who harbor certain levels of racism of racist behavior, it actually makes them angry towards you. And that's why Obama can't afford to push back. He has to acknowledge and affirm that he's black so as not to alienate black voters, but he can't do it in such a way as to raise anxieties among white voters.

Is the calculus different with Clinton? Her campaign has been pretty explicit about pushing back in a way that's centered on her gender, as in the incident with David Shuster at MSNBC.
R: Of course, there are more women then there are black voters, right? It doesn't make blacks angry to point out that blacks are disadvantaged by bias. It makes whites angry. The same is true of gender. In the Democratic primaries she's dealing with a more sympathetic audience among women and to some degree among men. I don't think you'll see that in the general election at all, because she'll be fighting the implicit associations between women and nurturing domestic roles rather than leadership roles. At that point, any effort to play the gender card, if you will, is going to alienate some of the voters she needs--the voters who think it's a good idea for women to stay home.

There's a real split here about implicit associations and explicit ones. The efforts to articulate concerns about racism in the way you described are explicit efforts. Look at yourself, think about it, examine the data--that's a deliberative process meant to get people to reason through the problem and confront themselves in a different way. But you can't fight implicit biases with reasoned argument. It's not how they work. They work on an intuitive, affective, emotional level. Pushing back just makes people angry. You don't see that working very well in the research. And it wouldn't work in this campaign either. Instead, the candidates should combat implicit bias implicitly--Hillary has to look like a leader all the time; Obama looks inspirational. You fight fire with fire.

Do you expect the race- and gender-baiting to get worse in the general election?
P: Even though the RNC has indicated that they are kind of scared about how to attack Hillary Clinton without charges of sexism being leveled against them, and Barack Obama without allegations of racism, you'll still have ancillary individuals and groups who will make these attacks--that, for example, Obama used drugs at one time. There's ample evidence that, at least with regards to juries, they tend to view defendants more harshly when they've committed a crime that seems racially congruent, like a black person committing a more blue-collar crime--robbery, drug dealing and so forth. If they play that up, it could be problematic for him. If they question his patriotism, again, that could be problematic for him, because it raises these implicit biases about whether he's American enough. Republicans will probably play on these things, and perhaps his relationship with his pastor Jeremiah Wright, who openly espouses a black value system, to raise implicit biases in the electorate. And I think that poses some significant challenges for Obama.

What about the "Hussein" issue? McCain himself has already said that his allies should not use Obama's middle name as a political jab.
R: But it doesn't cost McCain anything to disassociate himself from it. The unconscious bias works automatically, quickly and deductively. So you hear the name three times and the context afterwards where McCain carefully explains that this is not something he endorsed is sort of irrelevant. To the extent that saying Hussein over and over again is at all effective on voters, McCain disassociating himself doesn't undo that effect. Because it's that first system, that affective, intuitive one, that's at play.

P: It's the benefit without the burden. He can distance himself after the fact. The RNC has said that they're not going to officially make attacks on race and gender, but you can have other groups raise these concerns and it works to McCain's advantage. The other question here is how Obama and Clinton may tear themselves apart heading into the convention and the general election by raising all these questions about each other. They're provoking these implicit biases among the general electorate as we speak--and the Republican Party may not have to do much next fall.