What Are The Benefits (If Any) Of “Diversity”?

The always perceptive Peter Schmidt has an interview today in the Chronicle of Higher Education with James Sidanius, a professor of psychology and of African and African-American studies at Harvard University, whose research interests are “the political psychology of gender, group conflict, institutional discrimination and the evolutionary psychology of intergroup prejudice.” Prof. Sidanius led a team of researchers who conducted a long-term study of about 2000 students who entered UCLA in 1996. The results have been published in a new book, The Diversity Challenge: Social Identity and Intergroup Relations on the College Campus (Russell Sage). The students, Schmidt reports

were surveyed in the summer before they entered UCLA, at the end of their freshman year, and again each year until their graduation or the end of their fifth year in college.

Schmidt’s email interview with Prof. Sidanius is quite revealing, although it’s hard to say whether it reveals more about the students, about UCLA, or about Prof. Sidanius and his team. Some excerpts follow.

Q. Your book says you had expected to find that “cultural diversity and the multicultural practices that universities have put in place as a response to it have some profound effects on students.” What was your basis for having such expectations? Instead of finding profound effects—good or bad—your book reaches the bottom-line conclusions that cultural diversity and multicultural practices failed to either significantly change students’ ethnic identities and views on race or to have the feared effects of heightening ethnic conflict and separation on campus. At the end of the day, does your study buttress—or does it undermine—the argument that race-conscious college-admissions policies serve a compelling government interest by producing educational benefits? In which direction (and how far) does your study move the ball in the affirmative-action debate?

A. We found that neither the multicultural policies of the university nor the general multiethnic environment had profound effects on the intergroup attitudes of the students. Thus, in broad terms, the students left the university with largely the same social and political views that they entered college with.

However, three facts show that this does not mean that there were no theoretically expected changes. First, in general, students did become less politically conservative, ethnocentric, and racist over four years of college. Secondly, for the most part, increased levels of intergroup contact (e.g., in the form of having roommates from other ethnic groups) did attenuate levels of racism, and increased positive affect for members of other ethnic outgroups. Thirdly, in terms of the effects of affirmative action, we were interested in addressing the issue of whether or not being an affirmative-action admittee affected the academic self-confidence and academic performance of minority students. The negative effects of affirmative action on student performance were complex and conditional. For example, for black students who thought that they were affirmative-action admittees, grade-point average was negatively affected, but only among those with high personal-identity stereotype threat (concerns about confirming negative stereotypes of their ethnic groups). However, among those with no such stereotype-threat concerns, being an affirmative-action admittee had no deleterious effects on academic achievement.

Two things stand out for me here: 1) “Diversity” doesn’t make much difference, and 2) The researchers’ belief that all the surveyed students became “less politically conservative, ethnocentric, and racist over four years of college.” So, they were all “politically conservative, ethnocentric, and racist” when they began — at least to a degree — and they had all become less so after four years. Way to go, UCLA!

It is hard to judge the effect of affirmative action from this comment. Were the “deleterious effects” measured? How was the putative existence of “personal-identity stereotype threat” confirmed (other than by, I suppose, the presence of “deleterious effects”)?

Q. Your book talks about the limits of using “pan-ethnic” classifications—such as “Hispanic” or “Asian”—to lump together people from very different backgrounds. Tell me about the differences you found between people in each of these groups.

A. While white and black students identify with pan-ethnic labels, the use of these pan-ethnic labels was much less prevalent among Asian and Latino students. This was particularly the case among Asian students, who strongly preferred to identify with labels referencing their nations of origin rather than their pan-ethnic identities. This tendency was less strong among Latino students, about half of whom thought of themselves in pan-ethnic terms (e.g., Latino, Hispanic) and half of whom thought of themselves in terms of their national origins. However, unlike [for] the Asian students, university exposure appeared to increase the use of pan-ethnic labels among Latino students. There was no evidence indicating that the use of pan-ethnic labels either increased or decreased ethnic tension or intergroup hostility on campus.

UCLA, in short, created “Latinos” from students who had thought of themselves as Mexican-Americans or Guatemalan-Americans, etc., before arriving there.

Q. You talk about finding important differences between native-born and immigrant students within various racial and ethnic groups. Please talk about these, and whether they point to potential new avenues of research.

A. We found some important differences in the level of political socialization and attitude stability of those ethnic-minority students whom we classified as “old” and “new” immigrants. “Old Asians” and “old Latinos” were defined as those born and raised in the U.S.A. and who spoke English at home. In contrast, “new Asians” and “new Latinos” were defined as those who were immigrants or did not speak English at home. The data showed that old Asians looked very much like whites in their general levels of political conservatism, their levels of racism, and correlates of these attitudes. While old Latinos resembled whites and old Asians in the crystallization, consistency, and stability of attitudes, they were also found to [have] significantly lower levels of racism and general political conservatism. As opposed to “old” immigrants, the “new” immigrants showed relatively low levels of political socialization into American political culture, and lower levels of attitude stability, crystallization, and stability. However, this is the group which showed a systematic increase in political crystallization and stability with increasing exposure to the university environment.

So, whites (standard Americans?), again, are conservative and racist, and “old” Asians are just as bad. “Old” Latinos aren’t as bad, presumably because they aren’t as assimilated as “old” Asians, and the “new” immigrants were apparently the least contaminated with the conservatism and racism of white America. I have no idea what “crystallization” means.

Here’s a really interesting one, concerning the detrimental effect of rooming with Asians, who are described as almost toxic:

Q. … Non-Asian students assigned to live with Asian students generally tended to have more bias afterward. You suggest that the explanation may be that Asian students are relatively conservative, and their views rubbed off on their roommates. But do you have any empirical evidence that this was the cause of the shift? Might some other force or forces be at work here?

A. There has been a slight misunderstanding of our results here. Exposure to Asian roommates does not increase prejudice against Asians. Rather, the data show that exposure to Asian roommates increases prejudice against blacks and Latinos, and decreases friendship heterogeneity and intergroup comfort among students of all ethnicities. While we cannot be certain of the cause for this effect, the fact that Asian students have the most ethnocentric and xenophobic attitudes of all major student categories would suggest that socialization effects (or attitudinal contagion) are the most likely causes of these results.

I suppose the “slight misunderstanding” is that rooming with an Asian doesn’t increase your hatred of all groups, just all non-Asians.

There’s more. Read the whole thing if you have access to the Chronicle. For now I will mention only one other interesting finding: the researchers found that,

prior to college entry, 57 percent of black students and 72 percent of Latino students believed they may have been or definitely were admitted under affirmative action. In comparison, only 10 percent of white students believed that they may have been or definitely were admitted through affirmative action.

I will refrain from commenting further on the fact that prior to college entry half of the “Latino students” didn’t think of themselves as “Latino.” Who, I wonder, were those 10% of the whites who thought they were admitted through “affirmative action?”

Now, these findings would be more useful if only UCLA would reveal the actual numbers of blacks and Latinos who would not have been admitted without affirmative action.

Say What? (2)

  1. Miguelito February 20, 2009 at 12:56 am | | Reply

    Who, I wonder, were those 10% of the whites who thought they were admitted through “affirmative action?”

    Women? Don’t see who else would think AA actually helped them, at least not white males.

  2. J. Luecke February 23, 2009 at 9:30 am | | Reply

    The Chronicle article states that UCLA’s AA policy balances “several factors indicating disadvantage,” including poverty, so that in addition to women, low-income whites may have perceived themselves as benefitting from AA.

Say What?