Supreme Court Justice Antonin Scalia made news yesterday when he commented that black college students might be better off going to “less-advanced,” “slower-track” schools where they would be more likely to do well because the classes are not “too fast for them.” These comments are not just indelicately worded—they also reference a theory that has been largely discredited by social science. (Evidence that claims to support Scalia’s viewpoint withers upon closer examination.)
Scalia made these comments during oral arguments for a prominent affirmative action case, in which the court is considering (for the second time in three years) whether the use of racial preferences in university admissions violates the Constitution. His remarks reference the so-called “mismatch hypothesis,” which posits that minority students are harmed by policies that allow them to attend competitive colleges for which they lack adequate academic preparation.
Mismatch is possible in theory, but it presents an empirical question as to whether selective colleges admit students who would be better off at less challenging institutions. Straightforward comparisons of students with similar academic credentials who attended different colleges consistently find that students are more likely to graduate from more selective institutions. This finding holds for all groups of students examined, including underrepresented minorities and students with weaker academic preparation.
Critics of this line of research correctly note that these comparisons may be biased by unobserved differences among students. But studies that take advantage of random variation in college choice, such as that induced by cutoff-based scholarships, also find that students are more likely to succeed at more selective colleges than at less demanding institutions. For example, one recent study found that a Massachusetts scholarship program caused students to attend lower quality, in-state colleges and be less likely to graduate as a result.
Proponents of the mismatch theory cite studies that claim to find evidence that minority students would be more successful if they were redirected to less demanding colleges. But this evidence rarely survives scrutiny. One study claimed to show that California’s affirmative action ban increased the graduation rates of minority students. But I reanalyzed the same data and found no compelling support for this claim, suggesting that the findings derived from the methodological weaknesses of the original study.
Another study based on the same California data claimed to show that minority students are more likely to be successful in science fields if they attend a less selective college—more or less Scalia’s point. But my reanalysis found at best very weak evidence for this claim, and no evidence of any connection to affirmative action policies.
The mismatch idea will persist as long as the political controversy over affirmative action remains. But mismatch should be seen for what it is: a political strategy and not an argument based on credible evidence.