Rules and Incentives

I continue to have a variety of conversations with faculty about how to make students more academically ambitious and adventurous and less grade-conscious, and also about how to get faculty more interested in teaching and less grumpy about the choices students make.

I am more and more certain that we are too reliant on rules and not enough on incentives, for both students and faculty. I'll bet we'd get better results at Harvard and similar schools, with strong and motivated students and faculty, if we regulated less and motivated or encouraged desirable behaviors in other ways.

This is a followup on an idea I threw out earlier, that we should change the way we award honors. To review, I suggested that instead of using GPA as the primary metric, we should award honors by faculty vote, on review of transcripts and other relevant academic achievements. Reducing honors to GPAs provides perverse incentives: better to get an easy A than a B in a more challenging course, better to take a course in which you know most of the material than a course in which you will start behind the preparation level of most of the class, etc. A faculty committee could decide honors subjectively, on the basis of breadth and ambition as well as achievement. Now you couldn't get a committee to do that for the entire senior class, so I suggested having the votes taken by the departments, where the faculty, hopefully, know the students. That could lead to variations from discipline to discipline (but those variations exist already, because of different grading practices). It would also create the problem of allocating "slots." If there were 80 summas to be awarded to 1600 seniors, what's the fairest way to dole them out? Again, perfect fairness already does not exist. You need a departmental recommendation to graduate summa, whatever your GPA, and some departments are traditionally stingy. I think these problems would be manageable, if there were any support for the broad purpose: to get rid of rigid rules that create bad incentives and replace them with softer rules providing good incentives.

In a recent discussion of grade inflation, a colleague asked me what the incentive was for faculty to give anything but As. Now I know: we shouldn't need an incentive, our job is to grade our students fairly and with discrimination, etc. But we all have lots of jobs, and time spent talking to unhappy students is not only time spent unpleasantly, it is time taken away from the rest of our jobs. And there are few counterincentives. Faculty here used to be given data comparing their grading practices with those of other faculty, normalizing for the quality of the actual students taking our courses. But it was never clear whether this information provided more incentive to soft graders to get harsher than it did to harsh graders to get softer. It's been years since any dean or president has lectured the faculty about tough grading. Again, I think a faculty peer group discussion (at a departmental meeting, with everyone's grade distributions out in the open) might have a stronger dis-incentivizing effect than the kinds of rules Princeton has tried to impose.

The real trick would be to come up with incentive systems for both students and faculty to replace the baroque General Education requirements, the with eight categories to be checked off and an orthogonal "Study of the Past" requirement. The granularity constrains choice, forces awkward compromises (Ec 10, the biggest course in the college, doesn't fit neatly into any category so it can count for either of two, but not both), and yet limits what faculty can actually teach within the curriculum (interdisciplinary courses, I'm told, can have a hard time getting approved, even though interdisciplinary in scholarship is all the rage).

Rules become a distraction, especially in the absence of any larger discourse about educational purpose. Students come into my office during freshman week with grids showing how the various rules will get satisfied over their eight terms, before they have talked to anyone about what they are trying to accomplish by going to college. The rules become the advising system and reinforce a sort of myopia to which students are anyway subject. Satisfying requirements is the sort of game they are good at. And naturally, while they are at the task of satisfying rules they find meaningless, they will try to optimize other measures, such as workload and GPA.

We need some narrative device, or some different kind of feedback and example-setting, to lift the eyes of students from the ground right in front of their feet as they walk through college. One colleague suggested to me the idea of a "beautiful transcript," which of course would not be an unblemished transcript but a transcript with some character and strength. Or maybe we just need more talking cures, some way to make filling out the study card less an act conducted in the dark and in private and then signed off on, hastily, by an academic "advisor." And we need something for the faculty too, since the incentive and reward system for good, caring, creative, innovative teaching is also not very robust, by comparison with the reward system for other faculty behaviors.

I wonder if there is really any hope of getting rid of the regulatory frame of mind, where we tend to think mostly in terms of "What are the degree requirements?" As I find myself more and more advising students, "You should't let satisfying the requirements interfere too much with getting an education!"