The Mismatch Myth in American Higher Education: A Synthesis of Empirical Evidence at the Law School and Undergraduate Levels

Abstract

This paper presents a comprehensive examination of the empirical literature testing the academic mismatch hypothesis as it applies to affirmative action and students of color in U.S. higher education. The primary focus of this paper is the mismatch research addressing American legal education (Part II). This includes a detailed assessment of the empirical basis of claims made by Richard Sander and, more recently, by Doug Williams, showing flaws in their work, including questionable claims and methodological choices. In particular, this paper calls into question all research using the LSAC Bar Passage Study data that treats schools in tiers 2 and 3 in that study as separate and hierarchically ordered, arguing that this treatment is statistically unjustified and due to idiosyncrasies in the data and factors that in fact distinguish these tiers serves to enhance the odds of finding a mismatch effect while lowering the likelihood of finding reverse mismatch effects. The paper also reviews research on mismatch at the undergraduate level (Part III), specifically examining the outcomes of graduation rates and earnings and again finding that the mismatch hypothesis lacks empirical support and is, if anything, empirically less plausible than claims made for a reverse mismatch effect. In examining both legal and undergraduate education, this paper both critiques work that purports to find evidence of mismatch and references numerous studies that find no evidence of mismatch effects or evidence of reverse mismatch effects, including studies that use state of the art methods to control for selection bias. Overall the social science evidence points clearly in one direction: affirmative action as practiced today is not plagued by mismatch effects; indeed the evidence indicates that underrepresented minority students tend to do better over the life course if they attend the most selective school that will admit them.

Keywords: affirmative action, mismatch, law school, legal education, graduation rate, college, undergraduate, bar exam, Bar Passage Study, African American, Latino, earnings, leadership, selection bias, Sander, Williams, Bowen, Krueger, Schuette, Fisher, Grutter, LSAC, empirical research, social policy

JEL Classification: I2, J1, J7