In honor of the summer blogging season, we're going to take a break from the usual parade of tepidly amusing anecdotes and run some tepid legal thinking from a paper I wrote this semester. My professor suggested I submit it for publication. The way I see it, if all 6 of my readers make it through this series (and that's a big IF) that'll be approximately triple the readership I'd get in a law review.

Preamble: I think affirmative action is an entirely worthwhile policy for academic admissions. I don't believe there are absolute scales of so-called 'merit' that schools are morally bound to reward with admissions letters. No one 'deserves' to get into school X. Schools should be able to admit whomever they want, for whatever reason, in the interests of delivering whatever kind of academic / learning culture they aim to create. I do not support Calif's Prop 209 as it applies to the Univ of California system (I have no opinion on it outside that context – is that lawyerly precision or what?)

Not only that, I am a beneficiary of affirmative action. Older (30+) students often receive an admissions preference, in support of non-racial diversity. I got into UCLA law school with an undergrad GPA and LSAT that were at or below the median. (And, for god's sake, an art degree – WTF!?) This is conjecture, but I imagine if I didn't have the age / work experience thing going for me, there were many higher-scoring white males who would've had my seat.

Here's the problem. And let me describe it in a race-neutral manner, because it is. Law schools depend largely on a composite of undergrad GPA & LSAT to assess applicants, called an 'academic index'. Preferences work by giving a particular applicant, let's call her Ashley, a numerical boost to her index. For whatever reason, some elite law school wants to enroll more people whose name is Ashley, and without the index boost, this Ashley wouldn't make the cut.

So Ashley gets an index boost, which lifts her into the target index zone, and she gets in. But in real terms, she's near the bottom of her incoming class. This is logical, right? If Ashley needed a boost to clear the lower threshold of the target zone, that means she started below it.

No one would be surprised to hear that success on the bar exam is highly correlated to one's law school GPA, which in turn is correlated to undergrad GPA and LSAT. So what will happen to Ashley? Statistically speaking, since she's starting out at the bottom of the class, she's more likely to end up there, and more likely to have trouble on the bar as a result.

Two thoughts will spring to your mind. 1) Am I saying preferences make people fail the bar exam? Not really. I'm saying that preferences may be effective at getting Ashley in the door, but they don't address the bigger issue – if Ashley's starting out near the bottom of the class, there's a risk it will negatively impact her legal education. It should be clear that this isn't a reflection on Ashley's qualifications in absolute terms. It's a natural side effect of relying on numerical measures so heavily during admissions, and manipulating these numbers to achieve certain effects.

Thought 2): why is it obvious that Ashley won't do well on the bar? Isn't being at the bottom of this elite law school better than being in the middle of a 2nd or 3rd tier school? That's a logical intuition, but the short answer is: no, it's not.

The bottom 10% of students in any school, regardless of eliteness, fare poorly on the bar. The reason why this is true is much debated. But students who end up at the bottom of their 1L class perform worse on the bar than just their admissions numbers would predict. This effect has been verified by the LSAC-BPS, a study of 27,000 law students completed in the mid-90s.

To bring this all together, any beneficiary of an admissions preference – white, non-white, old, young, Ashley – is conceivably susceptible to this problem. It just happens that preferences tend to be deployed in service of admitting underrepresented racial minorities (African-Americans, Latinos, Native Americans). And what we see is, indeed, these groups perform worse on the bar exam than whites and Asians. Not because they're intrinsically less skilled, but because the boost from the admissions preference is, like Cinderella's carriage, cosmetic & short-lived. In other words, bar underperformance correlates to race, because of who typically receives preferences, but it's not caused by race.

To be fair, there are reasonable people who would disagree with that last paragraph. Race-based theories of bar performance are out there, like stereotype threat (the idea that racial minorities tend to perform worse purely because they are expected to) or racial bias in the bar exam itself. But regardless of the reason, the underperformance is real – no one debates this. And regardless of the reason, schools need to think about the legal & ethical propriety of not disclosing these risks to students receiving preferences.

07 Jun 06