Wednesday, September 26, 2007

Today's Ed News: The Law School Mismatch Hypothesis Revisited

Vikram Amar and Richard Sander, in a LA Times op-ed, complain that no one takes their mismatch hypothesis seriously. Invoking Grey's Anatomy worthy dramatic flair, they write:

IMAGINE, FOR A MOMENT, that a program designed to aid disadvantaged students might, instead, be seriously undermining their performance. Imagine that the schools administering the programs were told that the programs might be having this boomerang effect -- but that no one investigated further because the programs were so popular and the prospect of change was so politically controversial.

They go on to make the now familiar claim that black law students would be better off had they not attended selective institutions. Does research have anything to say about this issue?

Funny that they forgot to mention it, but rising star and Princeton economist Jesse Rothstein and his colleague Albert Yoon have actually spent a lot of time rigorously looking at this question. (There's something in the eduwater where Jesse grew up, as his dad is Richard Rothstein.) And here’s what they find:

We find that the data are inconsistent with large mismatch effects, particularly with respect to employment outcomes. While moderate mismatch effects are possible, they are concentrated among the students with the weakest entering academic credentials.

By weakest academic credentials, Rothstein means students in the bottom 20% of the LSAT score and GPA distribution. And what would happen to black student representation if we took the advice of Amar and Sander? Rothstein and Yoon simulated such a parallel world, and found the following:

In the absence of affirmative action, we estimate that the number of black students entering law school would fall by about 60 percent, while black representation at the most selective schools would fall by 90 percent. The decline in black representation would extend well beyond the most selective schools; all but the least selective schools would enroll one-third to one-half fewer black students than they do today. This contradicts claims about the centrality of the so-called “cascade effect,” whereby affirmative action affects diversity only at the elite schools.

Rothstein and Yoon conclude:

Many potentially successful black law students would be excluded, far more than the number who would be induced to pass the bar exam by the elimination of mismatch effects. Accordingly, we find that eliminating affirmative action would dramatically reduce the production of black lawyers.

So to Amar and Sander - it's not exactly true that no one has been able to study this because the data are on lock down. That said, I don't see why they shouldn't have access to the data - but please cc Jesse Rothstein so he can check your numbers. His papers on this issue are available here; one is forthcoming in the University of Chicago Law Review.

Bonus prize!!!: for anyone who teaches undergrad methods or statistics, this is a great exam question, i.e.:

In this opinion piece, the authors argue that affirmative action in law school hurts its intended beneficiaries, black students. Use what you've learned this semester to identify some potential problems with their causal claim.

Data from one selective California law school from 2005 show that students who received large preferences were 10 times as likely to fail the California bar as students who received no preference. After the passage of Proposition 209, which limited the use of racial preferences at California's public universities, in-state bar passage rates for blacks and Latinos went up relative to out-of-state bar passage rates. To the extent that students of color moved from UC schools to less elite ones (as seems likely), the post-209 experience is consistent with the mismatch theory.

In general, research shows that 50% of black law students end up in the bottom 10th of their class, and that they are more than twice as likely to drop out as white students. Only one in three black students who start law school graduate and pass the bar on their first attempt; most never become lawyers. How much of this might be attributable to the mismatch effect of affirmative action is still a matter of debate, but the problem cries out for attention.


Stuart Buck said...

IMHO, Rothstein's credibility on this issue is undermined by the hard-line stance he took in this thread. He was asked the following:

Jesse Rothstein, . . . I'll look at the paper, but meanwhile, if you can briefly explain to readers how "there's little evidence that this harms anyone's chance of success" when, by my calculations, over 1/2 of matriculating black law students at the bottom 2/3 of schools either drop out, fail out, or never pass the bar--and the figures are much worse at individual schools--that would be very enlightening.

Rothstein's answer included this:

"We did consider the scenario you suggest, though we don't know whether it should count as a cost or a benefit. . . . So, consider such a student. Without preferences, she would not be admitted to any law school. Because of preferences, she has the opportunity to attend a third- or fourth-tier school. We can predict based on her entering credentials that she has, say, a 60% chance of graduating and passing the bar. It is very hard to say whether we have done her a favor by admitting her--a lottery ticket with a 60% chance of paying off may be worth buying or it may not, depending on what her other options are.

He then added this comment:

Another thing not to overlook: It isn't as if people who attend law school but do not pass the bar are condemned to the poorhouse. A JD (or even a year or two of law school) can be a useful occupational credential even without bar membership, perhaps more so than are similar credentials in other fields. It isn't inconceivable to me that someone could decide to attend law school even knowing that she had zero chance of passing the bar.

Again, I'm not arguing that law school is necessarily beneficial for these students. I'm arguing that we don't know, and that there isn't much basis for presuming that it is harmful.

To which I provided the obvious response:

This seems very implausible to me. Where in American society is a year or two of law school treated as a "useful occupational credential," let alone as exceeding other possible credentials,rather than as evidence that one couldn't hack it?

And no one's arguing that law school dropouts are "condemned to the poorhouse"; that's a complete straw man. The real question is whether going to law school for one or two years and then having to drop out has any benefit that outweighs the $50,000 that the dropout may have had to borrow, in addition to the opportunity cost from foregoing a salary for that time.

Rothstein's response -- both in this thread and on page 18 of his paper -- either avoids the question or consists of hand-waving assertions.

Rothstein had no response, which is not surprising. It just isn't credible to pretend that we don't know whether it is a negative experience to be drawn into law school, to borrow tens of thousands of dollars, and then have to drop out or fail the bar exam. Such reasoning is the mark of someone who has an ideological position that he is trying to defend by any means possible.

Stuart Buck said...

So to Amar and Sander - it's not exactly true that no one has been able to study this because the data are on lock down.

Amar and Sander didn't claim that "no one has been able to study this." They claimed that nervous officials in California have limited access to that state's data. Do you think they're lying about that?