« on: July 22, 2006, 09:26:51 AM »
Red (and other compelling AA defenders)---I've found your discussion enlightening, but I have one major question that nobody has addressed (though i skimmed a few of the posts, so please correct me if i overlooked it). I personally am very torn about AA, though I do believe it is a useful social policy for the moment, so I am a supporter. That said, however, there's a major gap regarding the implementation of AA that I'm not sure how to answer, so I would like to know what others think.
How does a general (or "societal" or "structual" problem as you all like to say) problem dictate particular actions? In other words, what concrete changes should an AA policy enact in law school admissions decisions on particular candidates?
Lets assume that all the arguments that have been provided in defense of AA are true (that there is a large LSAT (I'll ignore GPA) gap between members of different races, that the "stereotype threat" is prevalent and is the best conception we have for explaining this gap, that merit and fairness are the proper criteria for law school admissions, etc.). Now if all of these are true, it seems to follow that, as Red suggested in her very first post, AA is a good policy because it offsets the psychologically (or societally) induced LSAT gap that obscures the equal talents of members of different races and it thereby increases the meritocracy of law schools (which we have assumed is good). AA thus seems to be a good general solution to a general problem.
What, though, are the concrete effects of this policy? On the decision of any particular candidate with an LSAT score lower than the average accepted by a schoool, how should the admissions committee determine that this candidate suffered from the "stereotype threat" as opposed simply to achieving a score indicative of their apptitude (and yes I know this assumes without argument that GPA\LSAT indicate apptitude, but there does not seem to be a better suggested method for evaluating apptitude so i will not argue this point here)? If, as Red said, black students score on average 6 points lower because of the "stereotype threat," does that mean we should add 6 points to every black student's LSAT score and from there judge all candidates equally and blind to race?
That might be the systematic solution, but it does not seem fair or meritocratic. Certainly, some students are affected more than others by the "stereotype threat" and the 6 points is just an average. How should we account for particular students who score well on other standardized tests (ACTs, SATs) but bomb the LSAT--do they get +6? What about students with high GPAs (let's even say a 4.0) but bad LSATs, do they get +6? Why would the "stereotype threat" affect these students in certain acadmic situations but not, seemingly, in other comparable ones? Should we give the +6 to students only when they have low ACTs\SATs and low LSATs, or only when they have low GPAs, low SATs\ACTs\ and low LSATs? And what about students who are affected by more than 6 points because of the "stereotype threat," but only get a 6 point boost? What is fair and meritocratic?
It would certainly be fair and meritocratic to give a boost to students that corresponds with the damage done by the "stereotype threat" when it is apparent, but there does not seem to be any way to determine this in a rigorous manner. Without such a determination, it seems unfair and unmeritocratic to boost everyone's scores equally (indeed, such rationale seems to me to be the continuous problem with AA: people agree with it on a general level, but in any individual instance where AA might seem to have helped a particular student gain admission, that student will get attacked for not deserving it, because how can we know when lower numbers obscure the higher talent of a student and when lower numbers actually indicate lower talent?). The idea of boosting scores equally for a particular race seems to lead, logically, to the idea of a quota system (because it assumes that all students are inherently equal and different races should be represented equally in proportion to the number of applicants, which should equal the proportion of the different races in society since all people are assumed equal).
Instead, it seems that, ideally, LSAT scores should be understood as flexible because of the possibility of the "stereotype threat," but this response undermines the use of LSATs (which i find problematic because it is the only common criterion of admissions for all students) and also neglects to mention how AA would be consistently implemented fairly and meritocratically.
It seems to me that, in the end, with AA in admissions decisions, there is generally a +6 boost (or +x), since there does not seem to be any way to determine when the "stereotype threat" occurs or to what extent it affects individuals, and so students get equal boosts which are usually too large or too small (though, on average, it is equal, it is not in particular situations). This result is highly unsatisfactory to me because although it does deservedly benefit some, it also undeservedly harms some. Perhaps it is best (and trite) to say that AA has lots of problems but that it is the best solution available to this "systemic" problem, but that is not much of a defense of AA.
This is the conclusion I reach, and my intuition is that, at the moment, the benefits of AA are greater than the harms, but this is not for logical or certain reasons (the benefits of AA are not deducible from principles of fairness or meritocracy as some think), and i do not think that the gap between benefits and harms is very great.
So I would like to hear why people think I am right or wrong, but, more importantly, what concrete suggestions anyone has for implementing AA policies in evaluations of particular candidates.