exams, "merit," and the ricci case

Interesting op-ed by Lani Guinier and Susan Sturm in today's NY Times, arguing that the promotion exam was flawed and the "merit" argument accordingly unconvincing in the case. I have always thought this the weakest part of the plaintiffs' argument: it seems odd to me that firefighters, especially veterans, should be evaluated by a written test, and the winners seem to have been better crammers rather than better performers. This, although I am principle opposed to affirmative action, and the case was appealing to me on ideological grounds.

The biggest problem with Guinier and Sturm's argument is one of consistency. Sure, it's foolish to evaluate firefighters by written exams: but isn't it equally foolish to evaluate college applicants in the same way? Didn't Guinier and Sturm get to teach at Harvard and Columbia, in part, because they did well on law school exams, which (like those in New Haven) reward memorization and are eminently crammable? The problem here is that we are sending a message that written exams are fine for determining entrance into the elite, but are irrelevant to working class people, especially if others (themselves drawn primarily from an elite pool) don't like the outcomes. Presumably Guinier and Sturm would agree with this, and prefer to reduce the reliance on exams at all levels: but until this happens it's going to be difficult to convince the Ricci plaintiffs that they don't have the right to cram their way to success like everyone else.