Myths and Facts About Affirmative Action

The case against affirmative action in admissions to institutions of higher education is based on the attractions of a color blind society and is buttressed by a sense some have that affirmative action programs are not just unfair, but pointless. Their intended beneficiaries are seen as being put in situations where they are unable to compete with whites and not only fail, but are destructively demoralized in the process. Common to arguments against affirmative action in educational admissions is a sense that minorities advantaged by it displace whites who are more deserving of admission because they can better benefit from the education they seek and will perform better after graduation. We have surveyed minority and white alumni who graduated from the University of Michigan Law School between the years 1970 and 1996, and compared white admittees with those of African American, Hispanic and Native American heritages, most of whom benefited in the admissions process from affirmative action. We find that at least with respect to graduates of the University of Michigan Law School, empirical claims used to buttress the case against affirmative action are, for the most part, myths. People of good will believe them, but the facts are not what people think.

The data we present is from a study with considerably more detail that appeared in an article in the Spring 2000 issue of Law and Social Inquiry entitled, "Michigan's Minority Graduates in Practice: The River Runs Through Law School." This article not only explores issues not discussed in this paper, but it also describes our methods, cites relevant literature, establishes the reliability of our sample, and indicates why the accomplishments of Michigan's minority graduates in practice are unlikely to reflect continuing post-law school affirmative action. We shall not touch on these matters in this paper. Interested or skeptical readers should consult our larger study.

We recognize that empirical information, however powerful, cannot transform actions that are fundamentally wrong into behavior that is morally right. If making race an element of college and law school admissions decisions is wrong in the same way that segregation and other forms of discrimination against racial minorities is wrong, empirical data indicating the success of such programs would not justify them. However, there are substantial moral distinctions between decisions that treat race as an element which favors members of historically discriminated against minority groups in matters like college admissions and discrimination against members of these groups because of their race.

Discrimination against minorities has always been stigmatizing. Treating minorities worse than whites was justified by ideological, scientific and religious claims that they were less intelligent, less skilled, more dangerous and generally less worthy than whites. No such stigma attaches to white people who lose out to minorities in competition for employment or educational opportunities. Moreover, discrimination against minorities has been pervasive; historically it affected jobs, education, health care, voting and housing opportunities. It has also often been universal. Until recently, for example, no amount of money or education would have allowed a black person to join the Augusta National Country Club or buy a house in certain areas. Affirmative action, by contrast, exists in only a few spheres and affects whites in a limited, sporadic fashion. Most whites will never lose an opportunity to achieve their goals because minorities have been given a slight boost by affirmative action, and affirmative action gives minorities no opportunities that many whites do not equally enjoy. Even an individual white who loses out on admission to one school because race helped a competing minority is likely to have a second choice school that is almost as good as the one that turned him down. Finally, discrimination against disadvantaged minorities knows no limits except restraints that majorities voluntarily imposed. Affirmative action preferences are self-limiting because members of the majority are in control. Perhaps for this reason affirmative action programs, like affirmative action in educational admissions, often do not allocate to minorities even a proportionate share of the pie being distributed.

Despite these differences, it is probably a good thing that we feel a bit uncomfortable about race-conscious decision making even when historically discriminated against minorities are its beneficiaries. This unease testifies to a shared ideal in which inequalities are not linked to race and minorities are not systematically discriminated against.

Some opponents of affirmative action believe that we are close to achieving this world, that racial discrimination is largely a thing of the past and that little of its legacy lingers on. But these claims are implausible when we one looks at statistics relating to the health, housing, occupations and education of historically discriminated against minorities. The end of de jure discrimination reveals no sharp breaks, but instead suggests that current disadvantage is strongly rooted in the past. And often that past is not so distant. Until the 1970s, black families were, as a matter of policy, often denied federally guaranteed mortgages that whites with similar incomes and jobs could get. Hence, blacks were less able than whites to move to communities with outstanding schools, and rising real estate values did not build for them the same kind of nest egg that whites could pass on to their children. Rather, they just raised the rent.

Moreover, substantial discrimination against minorities endures. To stick with the housing example, the Washington Post on August 5, 2001 described research by George Washington University sociologist Gregory Squires, which found that in the Washington, DC area blacks were only half as likely as whites to obtain their first choice of housing, even after controlling for differences in income, education, housing tenure and their place of residence. Squires' results are quite similar to what the Fair Housing Council of Greater Washington and HUD each found when they sent out economically similar minority and white agents to test local housing markets. Discrimination against blacks (and in the HUD study black and Latinos) was rampant. However, 58 percent of the whites Squires surveyed thought that blacks and whites with similar incomes enjoyed similar housing opportunities in the DC area housing market. Whites who underestimate the prevalence of racial discrimination and inequity might naturally question the need for affirmative action.

Other perceptions also lead people of good will to question the need for and desirability of affirmative action programs. Opponents argue that these programs have no educational value, that they put minorities in situations where they can't cope, and that they turn out graduates who are unfit for the jobs they trained for or, at the least, substantially less fit than their white competitors. These claims are fed by stereotypes about minorities, but they also appear plausible to people who try to resist stereotypical thinking because they seem consistent with two facts about affirmative action admissions: One, except for Asians, minority applicants to selective colleges and professional schools often have lower high school or college grades than the white applicants with whom they compete and, on average, they score considerably worse than white applicants on standardized admissions tests. Two, students admitted to selective schools in part because of affirmative action considerations receive grades that are lower on average than the grades white and Asian admittees receive. But what a student offers a school and what he can take from his education turn on more than the capacities that grades and test scores imperfectly measure. Indeed, whether there is any link is an empirical question, and we undertook our study in part to answer it. We found that among students attending the University of Michigan Law School there is no clear link. Our findings are consistent with what others have found, looking at medical school graduates and at graduates of selective undergraduate institutions. We call the perceived link between admissions test scores and career success as well as other perceptions that some people have about affirmative action programs myths because the facts they assert are contradicted by our empirical investigation.

Universities often justify their affirmative action programs by claiming that the diversity they bring benefits all students. Opponents see little educational value in racial and ethnic diversity. Our data support the university's position. As can be seen in Table I, many white alumni looking back at their time in law school believe that racial diversity contributed substantially to the value of their classroom education, and data not in this table reveal that most other white alumni believe that racial and ethnic diversity made at least some educational contribution.

Table I also indicates that the perceived educational value of diversity increased markedly over time, but this trend disguises a more nuanced story. Only the views of white male graduates change markedly over time. The proportion of white women graduates attributing substantial educational value to ethnic diversity has always been close to 50 percent. We believe this is because when white males constituted the majority of all Michigan law school students, diversity meant less to them than it did to women and minorities, since their classmates were mainly people like themselves. By the 1990s, however, white males were less than half the student body. This change in student body composition brought with it an increase in the proportion of classroom contributions by women and minorities and, most likely, an increase in the proportion of white males who engaged in cooperative cross-race and cross-gender interactions in moot courts, law journals and similar settings. Although a correlation does not prove causality, we believe that these factors are the most likely explanation for the attitude change among white male graduates over time.

Another myth that fuels the sense that affirmative action programs are unfair to whites is the perception that minorities receive financial aid packages that largely pay their way while needy whites must take on substantial debt to pay for law school. This view is starkly contradicted by our data. In every decade, minorities are more likely than whites to have graduated with educational debt, and if in debt, to owe substantially more than indebted whites, both in dollars and relative to first year incomes. In the classes graduating between 1990 and 1996, for example, 96 percent of minority alumni but "only" 71 percent of white alumni graduated with some educational debt. Among those with debt, the average minority graduate owed $59,651 (in 1996 dollars) at graduation, and the average white graduate owed $48,404. Far from getting a free ride, Michigan's minority students are literally mortgaging their futures on their legal educations.

Some people believe that civil rights laws and the substantial progress minorities have made means race no longer is important to the delivery of legal services and consequently should play no role in law school admissions. Table II, which looks separately at different racial groups indicates that the market for legal services is far from color blind. Regardless of ethnicity, lawyers and clients with the same racial backgrounds gravitate toward each other. Black lawyers have a far higher proportion of black clients than do lawyers from other backgrounds, Hispanic lawyers have by far the highest proportion of Hispanic clients, and so on. Patterns are similar if we look at the ethnicity of principal client contacts for lawyers representing organizations.

Race plays a smaller role, particularly among Michigan's most recent graduates, in the kinds of clients served. Alumni of all races are found in all kinds of practice settings. Nonetheless, white alumni in private practice tend to render proportionately more service to small- and medium-sized businesses than minority alumni, while minority alumni are more likely to serve middle class and poor individuals.

It is clear from these data that Michigan's minority alumni disproportionately serve minority clients. It is possible that if Michigan abandoned affirmative action, whites who replaced minority admittees in law school would also replace them in representing minority clients. But this is unlikely. "Replacement" whites would probably look like whites now admitted, which means that few would have many minority clients. The service gap is more likely to be made up in part by an increased concentration of same-race clients in the practices of minorities admitted to law school without affirmative action, thus enhancing segregation within minority law practices, and in part by underemployed white lawyers more aggressively seeking minority clients. It is likely, however, that the full gap would not be made up, that the quality of the legal services minorities receive would diminish, and that it would be harder for minority clients to find lawyers they felt comfortable with.

In fact, almost all minority students who attend Michigan graduate and pass a bar exam. Substantial proportions (almost one in five from the classes graduating from 1990 through 1996) are chosen for judicial clerkships, another marker associated with law school success. Differences in graduation rates, bar passage rates and clerkship rates between Michigan's minority and white students are small and almost always statistically insignificant. After passing the bar, virtually all of Michigan's minority graduates then find legal practice or other law-related jobs. Although minority graduates are less likely than white graduates to take jobs with the nation's largest law firms (100 or more attorneys) more than half the minority graduates in the last cohort we examined (1990-96) secured such jobs. Given the tremendous competition for positions with these large, high-paying (and almost always overwhelmingly white) firms, this is further evidence of the capacity of Michigan's affirmative action admittees to achieve.

These measures are anything but racially neutral in their impact on minority applicants. If Michigan and similar schools did not look beyond these measures and consider race, minority enrollments, especially black student enrollments, would drop precipitously. This happened when affirmative action was banned at the University of Texas, UCLA and Berkeley law schools.

The disparate impact that these measures have on minorities does not, however, mean that they are not measuring merit in a race-neutral fashion. It might be that minorities admitted in part because of affirmative action are less likely than whites to benefit from a legal education or to succeed in the practice of law. They do less well on their first-year law school grades than whites and, without more, this could be taken as a sign that they are less well-suited to the practice of law. To check this, we created three measures of post-law school success: earned income, self-reported assessments of success, and serving the community. Using multiple regression analysis to hold constant such factors as gender and time since graduation, we find that LSAT scores and college grades have no relationship to any of our measures of success. The same is true, incidentally, if we just look at white students or just look at minorities. This means that within the range of students Michigan admits, LSAT scores and college grades have nothing to do with merit, if merit is defined not as doing well on tests, but rather, as fitness for law practice. To exclude minorities from top law schools because their LSAT scores and college grades are lower than those of white applicants is to allocate positions in law schools using measures that eliminate minorities who would be as successful as whites if they were allowed access to the same high-quality legal education.

We found no evidence that Michigan's minority graduates are less successful than the school's white graduates, and in one respect, they seem to do better. Table III reveals that among graduates in law firms there is no statistically significant difference between minorities and whites in the likelihood of making partner, and among graduates not in law firms the one statistically significant difference in the likelihood of being a supervising attorney favors minorities.

Income is another indicator of the success of Michigan's minority graduates. Among those in private practice, in 1997, the median minority graduate of the 1970s was earning $120,000 (average $168,000), the median minority graduate of the 1980s was earning $95,000 (average $125,000) and the median minority graduate of the 1990s was earning $70,000 (average $74,000). After controlling for time since graduation, there is no significant difference between the earnings of minorities and those of whites. Lawyers in businesses tend to earn more than graduates in private practice while those in government earn somewhat less, but Michigan's typical graduates, whatever their niche, do quite well financially by national standards.

Figure I illustrates for whites and minorities graduating from 1990 through 1996 both the similarity in incomes of Michigan's minority and white graduates, and the lack of relationship between an index composed of LSAT scores and college grades and career success as measured by income. The solid triangles represent minority alumni and the unfilled circles represent white alumni.

Note first placement on the horizontal access. The further to the right a person is, the higher his admissions index score, a statistic that combines relative placement of these alumni on college grades and LSAT scores. Minorities are on average further to the left on this axis, meaning that their admissions indices are, for the most part, lower than those presented by white applicants.

But now look at the vertical access. The higher up a person is the more he earns relative to national norms. Vertically there is no obvious difference between minorities and whites. They earn the same. Moreover, there seems no relationship between where a person is on the admissions index (horizontal dimension) and how much he earns (vertical dimension). Yet the result of abolishing affirmative action is to deny admission to most of those minority applicants to the left of center divide, even though many of them go on to do better financially than many of the whites with index scores to the right of the divide. Income graphs for graduates of the 1970s and 1980s present the same picture except there is substantially more clustering to the top. Indeed, the individual earnings of the median minority graduate of the 1970s were in 1996 in the top 8 percent of total US household incomes, and the total household incomes of these graduates are in the top 3 percent of US household incomes. Mean incomes are even higher.

Income is, of course, not the only possible measure of career success. Another, and perhaps a better measure, is general satisfaction with one's career. We found no significant differences between minority and white alumni on overall career satisfaction. Nor is there any relationship to index scores. The scatter plots for graduates of each decade are like those in Figure I.

A different kind of success measure involves giving back to the community. Table IV presents information on three of the dimensions we measured, service on non-profit boards, involvement in community politics and pro bono legal services. We see that minorities tend to do more service than whites. When we combine these dimensions with mentoring and other kinds of service and community involvement and employ regression analysis to control for gender, years since graduation, practice sector and other variables, we find a significant difference in the tendency of whites and minorities to serve their communities--minorities do more. The difference is small, and we would not make too much of it, but it is there. Compared to national norms, Michigan's white alumni do substantial service; its minority alumni do a bit more.

Putting our findings on our three success measures together, a clear picture emerges. Differences in college grades and LSAT scores, at least within the range required for admission to Michigan, have no relationship to an applicant's ability to put his or her legal education to rewarding and productive use, or, if they do have such a relationship, it is fully counterbalanced by other aspects of files admissions officers consider.

Hence, we can say, that at Michigan Law School and, most likely, at similarly selective law schools, affirmative action brings with it educational benefits for all students while training minority students for successful careers. Moreover, our data indicate that Michigan's minority alumni are highly successful, competent lawyers, and nothing we found suggests that the white applicants they displaced would have done better.

Our data do not directly address one final argument made by critics of affirmative action, namely that it stigmatizes and so demoralizes all minority students, including those who would have been admitted without affirmative action. However, much in our data and in our lengthy personal experience as teachers and researchers indicate that this is seldom true. Looking back at their law school experiences, minorities are as satisfied as whites in all areas we looked at except one, the social sphere, but it is only minority women who are less happy with their social situations than whites. Minority males are like their white counterparts. No one has ever suggested that affirmative action's stigmatic or demoralizing effects are confined to women. Moreover, in some ways, such as the career value of law school friendships, minorities report having gained more from law school than whites. Similarly, our success data are quite inconsistent with students being crushed or stigmatized by affirmative action. Two of us also have more than 30 years of law school teaching to reflect on. Some affirmative action admittees may have felt stigmatized, but this is more likely to have been due to their minority status than to the fact that they benefitted from affirmative action. More probably felt on the spot as representatives of not just themselves, but of their race. Both these problems seemed far more acute in the early years of affirmative action admissions at Michigan when minorities were present only in small numbers. Moreover, minority group members are consistently among the school's strongest supporters of affirmative action. Complaints about stigmatization are seldom voiced.

Conclusion

Minorities admitted to Michigan through affirmative action do as well as whites do after graduation on a variety of measures of success. They do not get a free ride, but go into substantial debt to secure their education. Their presence in meaningful numbers enhances the classroom education of most white students, according to whites who have studied at Michigan. Finally, after graduation they are far more likely than whites to serve members of their minority communities and somewhat more likely to give back to the community through service activities.

To be color blind but test-score conscious is a formula likely to cut dramatically the number and proportion of historically disadvantaged minorities who benefit from attending the country's most rigorous undergraduate colleges and its top professional schools. This will in turn limit their presence on college and university faculties, in the upper echelons of businesses, and in the most rewarding segments of many professions. Excluding minorities from educational opportunities they would make good use of will, in the short run, provide slightly better opportunities for some whites. In the long run, whites and minorities will be harmed by narrowed interaction between races, increased race-based inequality, and an exacerbation of racial tension. In short, increased integration and affirmative action go hand in hand, and the march toward equality is guided by affirmative action.

This paper may read to some as if it is a brief for affirmative action, and it certainly puts affirmative action, as it has been practiced at the University of Michigan Law School, in a good light. But it need not have done that. At the heart of the paper is a careful empirical study. As readers of our larger study will see, our methods were not chosen to bias our results. We could have found that Michigan's minority graduates seldom passed bar exams; we could have found they earned much less than whites or were much less happy with their careers. We could have found that admissions indices predict career success as well as they predict law school grades. We could have found almost no whites who thought that ethnic diversity added to their law school education. But what we found instead is the story we have just told; a story which shows that minorities who enter Michigan with an assist from affirmative action almost all graduate and then go on to have careers that in every way we could measure appear as successful as the careers of Michigan's white students. The inescapable conclusion is that Michigan's minority students are every bit as capable of succeeding in the practice of law as the white students whom they competed with in the admissions process and sometimes displaced. These are facts; they contrast strongly with the myths that some opponents of affirmative action have propagated.