JSTOR

Affirmative action in higher education admissions is now prohibited in several states and more legal challenges are pending. This situation adds urgency to the question of whether "race-neutral" reliance on the LSAT is fair to students of color. In this comment, a database of law school applicants from fifteen highly selective institutions is used to test whether the LSAT produces racial and ethnic differences among students with equivalent performance in college after controlling for institution attended, grade-point average, graduation date and field of study. Based on the data and a review of the legal education literature, it is argued that the LSAT is culturally biased in the sense that it creates an artificial barrier to entering the legal profession. In other words, racial and ethnic gaps on the LSAT are found to be larger than differences in undergraduate grades, law school grades or measures of subsequent success in the legal profession. Causes of test bias, including stereotype threat in the testing milieu, are investigated. Next, recurring problems are examined regarding the traditional predictive validity approach to assessing fairness on the LSAT. Several alternative hypotheses for these findings are explored. It is concluded that test bias can be a rationale for justifying affirmative action programs, and that traditional law school admission criteria, which are driven by LSAT scores, should be significantly altered.