William C. Kidder

excerpted Wrom: KBRNVWWCUFPEGAUTFJMVRESKPNKMBIPB Mirror or Magnify Racial and Ethnic Differences in Educational Attainment?: a Study of Equally Achieving "Elite" College Students , 89 California Law Review 1055-1124 (July, 2001)

A growing number of American law schools, and flagship public law schools in particular, have recently terminated race-conscious affirmative action plans in order to comply with various court decisions, popular referenda, and actions by public officials. Moreover, it is quite possible that the United States Supreme Court will soon grant review to one of the several ending challenges to affirmative action at public universities. As affirmative action continues to come under fire, high-stakes standardized tests like the Law School Admission Test ("LSAT") have also become the focus of intensified criticism. Much of the debate centers on whether standardized tests like the LSAT and the SAT are neutral barometers of racial and ethnic differences in educational achievement.

In this Comment, I attempt to provide empirical answers to the question of whether students of color with the same undergraduate grades systematically score lower on the LSAT than white students, even when controlling for factors such as which college they attended and what undergraduate major they selected. I also compare differences in law school grades to differences in LSAT scores. This investigation of racial and ethnic differences in LSAT scores among college peers is relevant to policy makers, I argue, because it provides an important piece of evidence establishing cultural bias on the LSAT or in the testing milieu.

In this study, I matched African American, Chicano/Latino, Native American, and Asian Pacific American applicants with White applicants who possessed equivalent undergraduate grade-point averages ("UGPA") from the same colleges during the same time period. I relied on a database of 1996, 1997, and 1998 applicants from fifteen highly selective colleges and universities to Boalt Hall, the law school at the University of California ("UC"), Berkeley. This is the first attempt to replicate the UGPA-matching procedure developed in Joseph Gannon's 1981 pioneering study.

The results indicate that among law school applicants with essentially the same performance in college, students of color encounter a substantial performance difference on the LSAT compared to their White classmates. These gaps are most severe for African American and Chicano/Latino applicants. A second round of matching, controlling for choice of major within each college or university, does nothing to reduce these performance differences on the LSAT. The results of this study therefore counter the claims of several standardized testing enthusiasts and affirmative action critics that the LSAT provides a neutral method of assessing academic achievement.

Part I surveys recent controversies over affirmative action in higher education and introduces the present debate over whether standardized tests mirror or magnify differences in educational attainment. Part II describes the applicants included in the Boalt Hall database and details the matching methodology. Part III reports average differences in LSAT scores by race and ethnicity after comparing applicants with equivalent UGPAs in the same institution. It also describes the results obtained from a second round of matching where undergraduate major was added to these controls. Part IV examines Joseph Gannon's earlier research, explains why the present methodology was adopted over the conventional predictive validity approach, and explores three plausible alternative explanations for the substantial gaps on the LSAT among academic peers. I conclude that the LSAT systematically disadvantages minority law school applicants. I therefore argue that affirmative action can be justified as a corrective for those racial and ethnic biases that use of the LSAT introduces into the admissions process. I finally argue that it is essential to revamp admission criteria to reduce the influence of the LSAT, particularly at law schools that are prohibited from using race in admissions decisions.

* * *

Conclusion

Many people, especially affirmative action critics and testing advocates, assume that standardized tests like the LSAT are a neutral reflection of racial and ethnic differences in educational achievement. This study of elite law school applicants, matched on UGPA within the same institutions and majors, establishes that such an assumption lacks empirical support.

Organizations active in the effort to dismantle affirmative action, such as the Center for Individual Rights (CIR), have developed their litigation strategies around the assumption that the LSAT equals merit. Given the centrality of the testing issue to the affirmative action debate, it is essential that the fairness of the LSAT and other standardized tests be vigorously contested when "reverse discrimination" challenges to law school admission policies are still in the pretrial stage. Unfortunately, universities' institutional interests in student diversity only partly overlap with the interests of students of color in preserving educational access through affirmative action. Thus, in major affirmative action cases like Bakke, DeFunis, and Hopwood, universities' records and briefs were devoid of evidence that standardized tests like the LSAT are culturally biased against students of color.

Grutter v. Bollinger has the potential for a more positive outcome. First, the University of Michigan is presenting scientific evidence that standardized tests unfairly penalize students of color. Second, student of color and pro-affirmative action organizations, who, in contrast to Hopwood, have been permitted to intervene as defendants, are also mounting a spirited challenge to the fairness of the LSAT.

Efforts to establish bias on standardized tests like the LSAT and SAT are particularly timely in the present litigation environment since the diversity rationale for affirmative action may soon be rejected or curtailed by the Supreme Court. Supporters of affirmative action would be wise to cover heir bases by supplementing their defense of the diversity rationale with proof of prior discrimination or evidence of bias in standardized tests and other traditional admission criteria.

If race-conscious affirmative action is declared constitutionally impermissible, the debate over the fairness of the LSAT and other standardized tests becomes more, not less, important. Professor Daria Roithmayr, for example, has identified a formidable network of anticompetitive institutional relationships that has locked in standards of merit governing entrance to law school that favor Whites. Without the momentum of an energetic appraisal of traditional merit criteria like the LSAT, there is little hope of altering this maze of feedback loops, which range from the US News & World Report rankings to ABA accreditation requirements, in order to promote more equitable opportunities to enter the legal profession.

Concerns over ethnic bias in standardized tests need to be linked to a more far-reaching and transformative critique of the conventional higher education testocracy. For instance, Professor Lani Guinier, in reviewing he recent comprehensive study of 1970-96 Michigan Law School alumni by Lempert, Chambers, and Adams, concludes:

The study confirms the benefits of affirmative action to all Michigan graduates. It tells us that affirmative action critics' much-touted reliance on objective measures of merit have little to recommend them over the life span of a lawyer. After all, it is the life's work of the graduates that is the big test. Thus, rather than ban affirmative action, the law school might do well to expand its practice and to revamp the admissions criteria for all incoming law students. The Michigan study, to the degree that it can be generalized to other peer institutions, supports the position that law school admission criteria should be revamped. The authors found no relationship between LSAT/UGPA index scores and subsequent success in the legal profession, as measured by income or career satisfaction. Moreover, the minority graduates of Michigan, who had appreciably lower average LSAT scores, nonetheless went on to serve as leaders in public service at higher rates than their White classmates. The hidden societal costs of selection criteria heavily dependent on the LSAT extend beyond the Michigan study. Research shows a negative correlation between social activism and performance on the LSAT for the national pool of test takers.

Given the results presented in this study, it should not come as a surprise that in the four years since Proposition 209 and SP-1, 25% of White applicants were accepted to Boalt Hall, UCLA Law School and UC Davis Law School, compared to 21% of Asian Americans, 18% of Chicanos and Latinos, 15% of Native Americans and only 11% of African Americans. Yet in California's stark post- affirmative action landscape there is a possibility worth exploring. It is always important to recognize the myriad historical and institutional barriers that have inhibited the educational attainment of students of color in the United States. However, the data reported in this study suggest that the LSAT decisively stratifies opportunity by race even among law school applicants who have (sometimes poignantly) overcome obstacles to achieve equivalent academic success over four or more years of college. Given that the LSAT extracts such a substantial "pioneer tax" on many of the most accomplished minority law school candidates in the country, it might be argued that the UC law schools violate Proposition 209's prohibition against racial and ethnic preferences by relying so heavily on the LSAT.

One final point concerns Professors Jencks and Phillips's suggestion, mentioned at the beginning of this Comment, that reducing the Black-White test score gap would do more to promote racial equality than any other feasible strategy. There may indeed be less racial inequality if performance differences on standardized tests suddenly shrank, but it does not follow that America's best chance to decrease racial inequality lies in closing the test score gap. Since, at least in the law school context, the Black-White and Latino-White gaps are greater on the LSAT than on undergraduate grades or law school grades, the continued emphasis on the LSAT acts as an artificial barrier for students of color aspiring to enter the legal profession. Thus, organizing social policy around test scores on the false assumption that tests like the LSAT are a neutral measure of educational differences can exacerbate, rather than lessen, racial and ethnic inequality. Jencks and Phillips's observation that it will take "several generations" for African Americans to adjust to "spending more time studying" implicitly amounts to an "all deliberate speed" approach to ending racial stratification. Even worse, it adds insult to injury to the large number of students of color who have already matched the performance of heir White classmates in college, and yet are suddenly and systematically made to appear "less qualified" after taking a four-hour LSAT.

. Researcher, Testing for the Public, Berkeley, California; J.D. Candidate, School of Law, University of California, Berkeley (Boalt Hall), May 2001.