Print

Affirmative action is needed to get the best candidates, Stanford psychologist says

The researchers plan on submitting their findings to the U.S. Supreme Court, which is expected to hear arguments next fall on what could become a landmark affirmative action case.

Chris Schmidt / iStock Diverse group of people around computers

The researchers argue that decision makers need affirmative action to make meritocratic decisions – to get the best candidates.

When it comes to affirmative action, the argument usually focuses on diversity. Promoting diversity, the Supreme Court ruled in 2003, can justify taking race into account.

But some people say this leads to the admission of less qualified candidates over better ones and creates a devil's choice between diversity and merit.

Not so, says Stanford psychologist Greg Walton. Diversity and meritocracy are not always at odds.

In fact, sometimes it is only by taking race and gender into account that schools and employers can admit and hire the best candidates, Walton argues in a paper slated for publication in the journal Social Issues and Policy Review with co-authors Steven J. Spencer of the University of Waterloo and Sam Erman of Harvard University.

Walton, an assistant professor of psychology, and Spencer plan to present their findings to the Supreme Court in an amicus brief in Fisher v. University of Texas, a case the justices are scheduled to hear next fall and that many court watchers believe threatens to upend affirmative action. (Supreme Court rules bar Erman, who was a recent Supreme Court clerk, from participating in the brief.)

"People have argued that affirmative action is consistent or is not consistent with meritocracy," Walton said. "Our argument is not that it's consistent or inconsistent. Our argument is that you need affirmative action to make meritocratic decisions – to get the best candidates."

The researchers say that people often assume that measures of merit like grades and test scores are unbiased – that they reflect the same level of ability and potential for all students.

Under this assumption, when an ethnic-minority student and a non-minority student have the same high school grades, they probably have the same level of ability and are likely to do equally well in college. When a woman and a man have the same score on a math test, it's assumed they have the same level of math ability.

The problem is that common school and testing environments create a different psychological experience for different students. This systematically disadvantages negatively stereotyped ethnic minority students like African Americans and Hispanic Americans, as well as girls and women in math and science.

L.A. Cicero Greg Walton portrait

'If you base your evaluation of candidates just on performance in settings that are biased, you end up discriminating,' said Stanford psychologist Greg Walton.

"When people perform in standard school settings, they are often aware of negative stereotypes about their group," Walton says. "Those stereotypes act like a psychological headwind – they cause people to perform worse. If you base your evaluation of candidates just on performance in settings that are biased, you end up discriminating."

The conclusion comes out of research on what is called stereotype threat – the worry people have when they risk confirming a negative stereotype about their group. That worry prevents people from performing as well as they can, hundreds of studies have found. 

As a consequence, Walton says, "Grades and test scores assessed in standard school settings underestimate the intellectual ability of students from negatively stereotyped groups and their potential to perform well in future settings."

Walton gives an example of how stereotype threat relates to preferences in admissions or hiring.

A woman and a man each apply to an elite engineering program, he says. The man has slightly better SAT math scores than the woman. He gets accepted to the program, but she does not.

"If stereotype threat on the SAT undermined the woman's performance and as a consequence caused her SAT score to underestimate her potential, then by not taking that bias into account, you have effectively discriminated against the woman," Walton says.

Walton and his colleagues argue that schools need to take affirmative steps to level the playing field and to make meritocratic decisions. If the SAT underestimates women's math ability or the ability of African American students, taking this into account will help schools both admit better candidates and more diverse ones.

While courts have ruled that diversity justifies taking race into account in admissions decisions, justices have not considered meritocracy as a reason for sorting by race.

"Our argument is that it is only by considering race that you can make meritocratic decisions," Walton says. "It's a separate argument from the diversity argument."

Walton's research provides the justices with another reason for upholding affirmative action.

But confronting legal questions is only part of the issue.

Walton says remedies need to be found in policy, as well. Environments need to be created that are fair and allow people to do well.

"The first step is for organizations to fix their own houses," he says.

Testing officials should look at how they administer tests and ask what they can do to mitigate the psychological threats that are present in their settings that cause people to do poorly, Walton says.

Schools and employers, he continues, should look into their own internal environments and ask how they can make those environments safe and secure so everyone can do well and stereotypes are off the table.

But if stereotype threat was present in a prior environment, hiring and admissions decisions need to take that into account.

"In taking affirmative steps," Walton, Spencer and Erman write, "organizations can promote meritocracy and diversity at once."

Media Contact

Greg Walton, Department of Psychology: (650) 498-4284, gwalton@stanford.edu

Brooke Donald, Stanford News Service: (650) 725-0224, brooke.donald@stanford.edu