Knowing Ourselves

by digby


The Wall Street Journal published this interesting article a few days back about how people respond to pollsters and what pollsters are doing in this election to control for bias. Give it a read if you're interested in these things.

The sidebar to the article actually interested me more than the article itself because I think it speaks to the recent dialog about whether or not there is such a thing as dogwhistles and who they are aimed toward:

HOW THE UNCONSCIOUS AFFECTS THE TRUTH
Pollsters try to get voters to reveal the biases they're too embarrassed or afraid to admit by asking questions like, "Is the country ready to elect an African-American president?" But people also have biases they don't know they have. These implicit biases, as psychologists call them, are picked up over a lifetime, absorbed from our culture, and work automatically to color our perceptions and influence our choices.

A massive study called Project Implicit uses a simple online test to attempt to measure the pervasiveness of dozens of implicit social biases, including those based on race, gender, sexual orientation, ethnicity, weight, age and religion. The project, housed jointly at the University of Virginia, Harvard University and the University of Washington, collects 20,000 responses a week -- and hundreds of researchers are using its data to predict how people will behave based on their unconscious prejudices.

The findings from Project Implicit's six million participants over a decade of testing reveal lingering suspicion of minority groups: Some 75% of whites, Hispanics and Asians show a bias for whites over African-Americans. Two-thirds of all respondents feel better toward heterosexuals than gays, Jews than Muslims and thin people over the obese. Minorities appear to carry some of the same biases. As many African-Americans show a preference for whites as for blacks. A third of Arab Muslims show a bias in favor of non-Muslims, and more than a third of gays prefer straight people. The strongest biases are against the elderly. More than 80% of test-takers showed a bias for the young, and that included respondents older than 60.

Project Implicit -- which is funded in part by the National Institute of Mental Health and the National Science Foundation -- studies how people associate a group of people, shown in photographs, with either positive or negative words. (Demonstrations and registration for the full tests are available online at implicit.harvard.edu2.)

Most of the research around its data is academic. One current project aims to see whether liberals or conservatives are more enthusiastic about the future or more nostalgic about the past. Practical applications are starting to evolve too. Clinical psychologists are studying whether implicit biases affect how doctors care for their patients.

Bias against African-Americans and the elderly will likely play a role in November's presidential election. Presumed Democratic Party nominee Barack Obama's father was black, and presumed Republican Party nominee John McCain is about to turn 72 years old. But researchers say they do not know to what degree bias will play out among voters. "We are not slaves to our associations," says University of Virginia psychologist Brian Nosek, one of Project Implicit's founders and principal researchers. A focus on Sen. Obama's promise of change, for example, could lead voters to forget his race; Sen. McCain's war record could let voters forget his age. Overriding bias requires a concerted effort, Mr. Nosek says. Most people don't see their own implicit bias, which can appear spontaneously as intuition, a gut feeling or a vague doubt about a candidate.

In the anonymity of the voting booth, those feelings could have a significant effect on undecided voters, says Yale University psychologist John Dovidio. With as many as 15% of undecided voters up for grabs, implicit bias could have a big effect in the presidential election.

University of Washington psychologist Anthony Greenwald, who developed Project Implicit's online test, predicts that Sen. McCain will get more votes than the polls currently predict in states with small black populations and that Sen. Obama will get more votes than polls predict in states with large African-American populations. The reason: whites unconsciously understate their pro-white bias by telling pollsters they will vote for Sen. Obama, while blacks unconsciously understate their pro-black preference by saying they don't intend to vote.

"There may be more in us that anticipates what we are going to do than we can report to others or ourselves," says Mr. Nosek. We aren't lying, he says, but we also may not be telling the truth.

I think that people who study political propaganda and marketing understand this stuff very well, particularly conservatives who rely on stoking people's prejudices and fear of change. It's not the conscious they are aiming for, it's the unconscious, where we don't even know we think the way we do.

Dogwhistles work in a number of ways, most importantly giving plausible deniability to those who use them. The people who hear them often don't even know what it is they've heard --- it's simply that they are suddenly discomfited or agitated about a person. (Others are not so subtle.) As the understanding of this phenomenon becomes more and more sophisticated, we can expect its application in politics (and business) to become equally sophisticated --- by creating backlash in unusual ways and playing to people's resentments about being called on their resentments. I think it's important to be aware of it and to discuss it when we can see that people are manipulating others with these methods. As the article states, it takes a concerted effort to override these biases and people can't do it if they don't know what's being done.




*On a more prosaic level, you have to wonder if the right's multi-pronged campaign to paint Obama as some kind of exotic, black "other" in the public unconscious is going to successfully override people's biases against the elderly if the only people stoking it are late night comics. It might be enough --- popular culture is extremely powerful. I guess we'll see.


Update: Drew Weston wrote about this over the week-end. I would hope that Obama does NOT take his advice on how to deal with it however. Disaster.


.