The Mismatch Effect

An “effect” is defined as “a phenomenon that follows and is caused by some previous phenomenon”. We have to be careful when labeling something as an “effect”, to make sure we’re pretty sure we’ve correctly assigned causation.

That being said, we have a phenomenon called the “mismatch effect”. When people are placed in a competitive environment on the basis of something other than their merits or qualifications, they find it much more difficult to keep up and wind up in the bottom rank or drop out entirely.

In the case of programs that aim to increase the number of certain groups admitted to top-rank colleges, we often find that members of these groups have lower average test scores or GPAs than the average among those who graduate from those schools.

We can imagine that of people who enroll in any school, there will be an average score and some variation around that average. It may well be that, say, two thirds of those admitted will do well and graduate. The students who fail to graduate will most likely be in the bottom third of that distribution. Now imagine a mismatched group (we’ll pick on Martians) is admitted on the basis of some sort of affirmative action program. Suppose their average test scores are one standard deviation lower than the average for the rest of the school. In this case, nearly 72% of these students will have test scores that place them below the cut-off for “likely to graduate”. This is not to say that no Martians will graduate from that school, merely that far fewer will, and at a rate that’s not at all similar to the graduation rate for non-Martians.

At the Volokh Conspiracy, Eugene Volokh offers:

The Mismatch Effect: A Danger for Students of All Races


That’s the debate about the “mismatch effect,” which I’ve followed over the years (though from a distance); it has mostly focused on whether race-based affirmative action causes problems (such as lower black bar passage rates) as a result of this effect, but it can also be relevant to many students of all races. I was first exposed to it because of the work of my UCLA School of Law colleague Rick Sander, and Robert Steinbuch at Arkansas / Little Rock has been working in it as well; Rob has been kind enough to pass along these thoughts on the subject:

Analysis of a large dataset containing information on graduates from the law school at which I teach, the University of Arkansas at Little Rock, Bowen School of Law, demonstrates that LSAT scores of students enrolled at the school (1) solidly predicted bar passage, and (2) varied significantly in relation to ethnicity.

Although color-blind admissions should produce roughly 25 percent of both Whites and African Americans in each LSAT-score quartile, over two-thirds of graduating African Americans were admitted with LSAT scores in the bottom quartile, as contrasted with only 16 percent for White students. (For more details, see the recent article I coauthored: Steinbuch and Love, Color-Blind-Spot: The Intersection of Freedom of Information Law and Affirmative Action in Law School Admissions, 20 Tex. Rev. L. & Pol. 1 (2016)). Although almost exactly a quarter of White students were admitted in the top quartile of LSAT scores (as expected), remarkably, only one percent of enrolled African Americans fell into the top quartile of LSAT scores. Predictably, this led to dramatic differences in bar passage: 80 percent of Whites passed the bar (the first time), while only 60 percent of African Americans did.

Given that the African-American cohort in our dataset on average had much lower LSAT scores than the bulk of the student body, it’s fair to conclude that this cohort overall was mismatched. This profile dominated because affirmative-action considerations are designed to consider factors beyond traditional credentials and explains why debates on how to deal with poor bar-passage rates often focus on race-based admissions. However, the ensuing discussion often misses that, while on average Whites will not be mismatched because they have such a large population — putting many at or above the mean of the class, the number of Whites who are mismatched could easily equal or exceed that of any other racial group.


And also from the Volokh Conspiracy, Rick Sander offers this:

An emerging scholarly consensus on mismatch and affirmative action (ideologues not welcome)

Williams’s paper presents equations testing dozens of different combinations of models and outcomes. With impressive consistency, his analysis shows very powerful evidence for law school mismatch, especially for first-time takers. His results are all the more compelling because, as Arcidiacono and Lovenheim point out, the weaknesses of the BPS data bias all analyses against a finding of mismatch. Williams concludes his piece, too, with a plea for the release of better data.

Meanwhile, not a single one of the law school mismatch critics has managed to publish their results in a peer-reviewed journal, though at least some of them have tried. As I will discuss in another post, many of these critics still shrilly hold to their earlier views. But it should be clear now to any reasonable observer that mismatch is a serious issue that the legal academy needs to address.

The above references two survey-scale papers, both taking great effort to eliminate ideological bias. Links are in the cited piece.