Two white students denied admission by the University of Michigan have filed a class-action lawsuit against the university. If they win their case, Michigan may find itself in the same position as the law school at the University of California at Berkeley, where 270 new students enrolled this past fall -- only one of whom is African American, and only seven of whom are Hispanic. Law-school officials attribute the small numbers of black and Hispanic students this year to a new university policy banning race and gender preferences in admissions decisions. This policy obviously harms the minority-group students who would have joined the class under the old admissions criteria. To the extent that Berkeley's law school is a stepping stone to the judiciary, politics, or merely six-figure salaries in corporations, the absence of those students will have an impact on the racial composition of the population of elite Americans.
       Many opponents of the University of California's new policy have cited this argument. Less well explored -- though often cited rhetorically -- is the idea that racial integration on campus may be of benefit to society as a whole. Although leaders of workshops on diversity assure us of the advantages of multiculturalism, and other educators offer anecdotes about the importance of the black perspective in the classroom, the debate is usually general, not cast in terms of what specific advantages integrated classrooms produce
       Based on recent research in political science and my own experiences teaching that subject, my belief is that even if black students do not speak in class directly about their experiences as members of a minority group, their physical presence alone is good for both white students and, eventually, the larger society, by opening up political debate and understanding.
       One problem that troubles political scientists is known as the "race of interviewer" effect. Briefly put, survey respondents' attitudes toward race and public policies related to race vary depending upon whether those respondents are interviewed by white or African American pollsters. When interviewed by whites, respondents tend to express conservative views; when interviewed by African Americans, respondents generally say they hold more-moderate opinions. For many years, this difference irked political scientists, who assumed that people's "true" opinions surfaced when they were interviewed by pollsters of the same race as themselves. Scholars thought, in other words, that white people lie when speaking to African American pollsters, to gain the pollsters' approval.
       But over the past 10 years, political scientists have questioned whether, in fact, individuals possess fixed opinions about many specific issues. Instead, John Zaller, a professor of political science at the University of California at Los Angeles, has argued that people keep many different "considerations" about any given issue handy in their heads. When they are asked to express a political opinion, they call some of these considerations to mind and use them in constructing a response. For instance, when asked whether government should provide aid to blacks, a typical white American may think about social programs such as affirmative action, abstract ideas about equality of opportunity, and attitudes toward blacks as a group; one of these considerations may be dominant, or the individual may use all of them in forming a response. Individuals engaged in this process are highly suggestible: The considerations they call to mind may be a function of how the question is phrased, what they saw on the news last night, even personal characteristics of the interviewer, including race.
       Lynn Sanders, an assistant professor of political science at the University of Chicago, has argued that, if people don't have fixed and lasting positions on most public policies, this might help to explain the race-of-interviewer effect: When whites interviewed by an African American express more support for programs intended to help minority groups, it is because whites in that circumstance are more likely to use considerations about blacks' interests in forming their opinions than they are when interviewed by another white. In other words, white people are more likely to consider the interests of blacks when they talk to a black pollster.
       I had mulled over these ideas, but I must confess that they weren't at the forefront of my mind as I taught my first college classes last year -- two discussion sections of Michigan's introductory class in American government. My students ranged from C-SPAN junkies to political know-nothings. As I ransacked the literature on American government to construct my part of the syllabus, I kept the know-nothings in mind. I decided that I needed to include readings on civil rights, race-based redistricting, and welfare. From the start, I noticed a dramatic difference in the reaction of the two classes to these parts of the syllabus. My morning class, when alert enough to register a partisan response, leaned left. My afternoon class tilted far to starboard. Even after the professor giving the lectures in the course explained one day that, contrary to common perception, Social Security, rather than welfare, accounted for the biggest chunk of the federal budget, my next afternoon class -- on the federal deficit -- turned, against my will, into theme and variation on "My father's a doctor, and I resent that he works his fingers to the bone to pay for those people on welfare." The discussion was tinged with racial misunderstandings and stereotypes -- not surprising, since political scientists have found that people's attitudes toward welfare are closely related to their attitudes toward African Americans as a group. Similar discussions in the afternoon class followed throughout the semester. On the other hand, I could not have sparked a racially conservative response in my morning class if I'd tried.
       It is possible that the two groups of students simply had very different ideologies to start with. But on social issues not related to race -- women's rights, say -- both classes embraced liberal values. As I pondered my students' different stances on racial issues, another explanation occurred to me: Lynn Sanders is right. My morning class, the liberal group, included several African American students (as well as an Asian and a Middle Eastern student); my afternoon class had none. Just by looking around them, students in the morning class were prompted to think about African Americans and their interests. The students then used those thoughts as they selected evidence and formed the opinions that they voiced in classroom discussions. Even though the African American students were not outspokenly liberal themselves, the result was a class in which racial issues were dealt with thoughtfully and more generously than in the afternoon section.
       The effect of this mere presence of African American students relates to debates about affirmative action. My classes introduced many students to issues of public policy and political justice that they had not thought about before. Students in the afternoon class did not hear the perspectives of African American students, and they were not prompted by the presence of black students to take black interests into account. If Sanders is correct, having even one or two black students in that class -- even if they did not vocally defend racially progressive policies -- could have inspired some of the white students to articulate a perspective sympathetic to African Americans. This, in turn, could have helped shape the political thoughts of a good number of other students; and since many students in this class were thinking substantively about welfare for the first time, what they heard that day took on extra importance.
       If my university were to lose the lawsuit against affirmative action, some minority-group students would still enroll, but their numbers might drop. Fewer such students would mean more classes like my afternoon section. And in the end, whites would be hurt. We should not overlook the tangible gains that come from having classes in which not everyone is the same color. For whites, those gains include being prompted to think from a perspective not one's own -- a critical skill that needs to be learned during the college years. In my view, the country also gains when classrooms contain diverse students. It has not escaped political scientists that the major fault lines in American politics have tended to run along racial lines throughout the nation's history. It appears that many whites and blacks have profoundly different views -- not only on public policies including affirmative action, busing, and aid to inner cities, but also on larger questions, such as social responsibility for the less fortunate, the impact of racism on individuals, and even whether or not there is now a race problem in this country. Without minority-group students in my classroom, the possibility that I can foster consensus or even intelligent disagreement on these issues is greatly diminished -- as is the possibility that the country can do the same.