USC News

Survey Finds Who Tolerates Discrimination

03/04/08
A large-scale USC report reveals that people are more willing to accept discrimination against poorly educated immigrants.

By Suzanne Wu
USC Gould School of Law professor Edward J. McCaffery

Photo/Spider Photography
A new USC study has revealed that tolerance of discrimination varies by sex and across broad contexts of social, political and economic relevance.

In a survey of more than 3,300 people, men displayed a greater willingness to tolerate discrimination than women, with the most people willing to tolerate discrimination against poorly educated immigrants and the fewest willing to tolerate discrimination against the genetically disadvantaged.

“An individual who sees nothing wrong with certain forms of discrimination will often find others objectionable and favor their prohibition,” wrote USC Gould School of Law professor Edward J. McCaffery and former USC College professor of economics Timur Kuran, who led the research at USC and is now at Duke University.

“Many political struggles of our time, in the United States as elsewhere, amount to clashes over the appropriate boundary between permissible and impermissible forms of discrimination.”

Respondents to both a phone and an online survey were presented with five scenarios, each of which dealt with a form of discrimination targeting a distinct class of individuals: Arab-American airplane travelers, seriously overweight people, the genetically disadvantaged, poorly educated immigrants and African-American motorists.

All questions used the same format, first explaining a controversy and then providing a utilitarian statement in favor of discrimination followed by a consideration of justice. An overwhelming percentage of the respondents chose the equality position in every category.

Of the significant minority who chose to allow discrimination, the highest percentage of people in both the phone and Web survey accepted discrimination against “poorly educated immigrants” (27.7 percent and 32.3 percent, respectively), followed by acceptance of discrimination against Arab-Americans (26.4 percent of phone respondents, 17.8 percent of online respondents).

On the other end of the range, respondents were least likely to accept discrimination against the genetically disadvantaged (6.7 percent of phone respondents, 3.2 percent of online respondents). Tolerance of discrimination against African-Americans (13.7 percent phone, 13.2 percent online) was statistically insignificant from acceptance of discrimination against seriously overweight people (15 percent phone, 13 percent online).

The study, forthcoming in Political Research Quarterly, also revealed that, across all categories, a larger percentage of men than of women accept discrimination. For example, men on the phone were 7.6 percent more likely than women to tolerate discrimination against the obese and 8.9 percent more likely to accept racial profiling of African-American motorists.

The magnitude of the sex difference was even higher among Web respondents. In the Web survey, men were 19.6 percent more likely to tolerate discrimination against the obese and 17.4 percent more likely to accept racial profiling.

These results suggested that live interaction on the phone – even with an interviewer trained to withhold reaction and remain impartial – encouraged respondents to adjust their answers. Specifically, women seemed to overstate and men seemed to hide their tolerance of discrimination on the relatively public medium of the telephone, as compared to the anonymity of the Internet.

The tolerance disparity between men and women narrowed in contexts where a willingness to accept discrimination might appear socially desirable.

The “sex gap” – that is, the difference between the percentages of men and women accepting discrimination, as a percentage of their average – was smallest for Arab-American airplane passengers (8 percent) and greatest in regard to African-American motorists (28.9 percent), among those surveyed by phone.

“As a matter of practice, people morally opposed to discriminatory policies based on reviled forms of prejudice do not insist on equal treatment for everyone, in every context,” Kuran and McCaffery wrote. “The surveys reported in this article validate this poorly appreciated fact in reference to several matters of political importance. In addition, they show that expressed attitudes toward discrimination vary by sex as well as context.”

The study was funded by the USC-Caltech Center for the Study of Law & Politics.