Wednesday, March 07, 2007

Candid Evaluations

It's the time of year to rate faculty promotion applications. (“Promotion” here refers to moving from Associate to Full Professor, or in some cases, Assistant to Associate.) Through a series of political compromises worked out over the years, we've developed a relatively complicated series of steps that a given application has to move through. People with input include the applicant, a peer (chosen by the applicant), the department chair, a faculty union committee, the dean, and the VPAA. (Eventually, the President and Board have final say, but in practice they typically go with the VPAA's recommendation.)

We've also developed a relatively clear set of criteria, using fewer categories than in the past and insisting on more detail for each. (For example, we dropped “community service” as a category when we realized that nobody could define it, let alone explain how to judge or even verify it. One professor always wrote “paid my taxes” under community service. I admire the panache, if not the altruism.)

We've had multiple discussions with the faculty union about processes and criteria, so there aren't any surprises. (They were as happy to drop the “community service” category as we were!) We have a very experienced set of department chairs. You'd think this would be easy.

Ugh.

No self-respecting professor would choose a peer who would trash her. So these inputs become largely meaningless by virtue of being uniform. Everybody is perfect.

The department chairs, as a group, see much more harm than good in giving a less-than-glowing evaluation. There's some logic to their position: they have to live with the applicant, and they generally don't see the applications from other departments. As long as an applicant can put at least something in each category, it's tempting to say “ahh, what the hell,” call it good, and kick it upstairs. Multiply that dynamic by a couple of decades, and you get a weird self-defeating process in which broad participation actually results in centralized decision-making, since all that conflict aversion renders the vast majority of the input irrelevant. The VPAA became, by default, the designated bad guy, whose job it was to overturn all those lower-level recommendations.

(Full disclosure: the deans' responses have largely followed the same dynamic as the chairs'. By the time I got here, the process was improbably long, impressively expansive, and ultimately decided by one person.)

We're trying now to introduce some level of candor to the evaluation process. It's an ugly, uphill battle.

We have some basic structural reasons to avoid candor. An unsuccessful promotion applicant remains on the faculty, still with lifetime tenure. Some people have been denied more than once, and all that grudge-nursing can create some very unhappy corners of the college. In the absence of candor, too, an unsuccessful applicant may or may not have a clear sense of the reasons for denial. Tenure abhors a vaccum, though, so in the absence of real reasons, people invent reasons that make sense to them. These often involve ugly accusations of bias, or personal vendettas, or simply misplaced anger.

To make matters worse, we have no merit pay system, other than promotion 'bumps.' In the absence of a consequence for a good or bad review, the path of least resistance is grade inflation. Multiply that by a few decades, and what was born as a failure of nerve has gradually become an inalienable right. It's tough to introduce candor suddenly at the point of promotion, when routine reviews leading up to it have been uniformly glowing (even if only because they all are). It's even tougher to generate candor when the hostility it would engender is palpable and immediate, and the payoff long-term and abstract.

Finally, there's the “you first” problem. A reasonable department chair might calculate, correctly, that if she decides to bite the bullet and go for candor, and her colleagues in other departments don't, then she will generate all of the downside with none of the upside. Better to take the safe way out, and accede to grade inflation.

The faculty union wouldn't stand for merit pay, and we're reasonably sure that in this political climate, a strike would hurt the college more than any concessions we might win on individual issues would help. So we're pretty much caught in the parameters as they exist, perverse incentives and all. Getting smart people to act contrary to palpable incentives in the name of abstract rightness is an uphill battle.

Has your college found a way to do candid evaluations without sacrificing broad input?