To the Editor: The Grade Cap Passes the Advanced Game Theory Exam

Does the grade cap “fail the game theory exam?” Professor Scott Duke Kominers ’09 makes many valid points, but ultimately gives the right answer to the wrong question, oversimplifying the problem.

The oversimplified question: In a world where faculty reliably challenge their students and distinguish the strongest work with the highest grades, does a cap on A’s make sense? No. But, alas, that’s not our world. Kominers is laser-focused on unrepresentative special cases and ignores the proposal’s provision for handling them.

His op-ed says little about the actual problem: grade inflation. Grades are inflated because faculty, like students, are agents in a market who respond to incentives. Game-theoretically speaking, faculty “buy” effort with higher grades. But higher grades also buy higher course evaluations and larger enrollments.

Lower Q scores and enrollments are about more than ego. As professor Steven A. Pinker observes, non-inflation can drive students away from one’s whole department. Inflated grades can be compounded by deflated workloads. The proposal aims to prevent students’ flocking to courses with lower standards. We currently have a market in which the buyers are free to print money ad libitum. We need a cap.

Kominers is right that this involves tradeoffs. The elite mathematicians in Math 55 could all plausibly earn uninflated A’s. With a cap, some may flee to a less advanced course like Math 25 for fear of being “punished” with an A-minus. This is clearly not the proposal’s intent. But that’s why it includes a provision that Kominers doesn’t mention: By mutual agreement, the instructors of Math 55 and 25 can reallocate, migrating a handful of A’s up, just in case some otherwise intrepid mathletes would choose to migrate down.

This solution isn’t perfect, but we shouldn’t let the perfect be the enemy of the good. The proposal’s broader aim is to shift our grading culture so that strong students no longer feel obliged to have perfect grade point averages and can eagerly enroll in courses in which they might get an A-minus.

Math 55 is unusual, but it’s now typical for a course’s top student and median student to receive the same grade. This provides widespread incentives for students to avoid classes with more rigorous evaluation and to shift their competition — which is inevitable — to extracurriculars.

Kominers repeats the familiar argument that a cap will promote unhealthy competition. We disagree. Making it harder to get A’s will, if anything, give students more incentives to form study groups. The idea that capping A’s will promote ruthless competition hinges on the misguided assumption that Harvard students could never view an A-minus as a good grade. (Poor Princetonians had to settle for B’s, prompting revolt.) It wasn’t always this way. We encourage students to imagine how liberating it would be to put the “us” back in “A-minus.”

If we can’t rein in the perverse incentives that produce grade inflation, then we’re playing the wrong game. It’s time to liberate students from the tyranny of perfection and incentivize students to challenge themselves — not just in Math 55, but across the full range of Harvard courses.

— Jason Furman ’92, Aetna Professor of the Practice of Economic Policy, Harvard Kennedy School and Department of Economics;

Joshua D. Greene ’96, Alfred and Rebecca Lin Professor of Civil Discourse, Department of Psychology and Center for Brain Science, member of the Subcommittee on Grading of the Undergraduate Educational Policy Committee;

and Alisha C. Holland, Gates Professor of Developing Societies, Department of Government, member of the Subcommittee on Grading of the Undergraduate Educational Policy Committee.

Letters to the Editor must respond directly and explicitly to either an opinion piece recently published on the Editorial page, or else to The Crimson’s manner of coverage within any section of the newspaper. Letters that respond to the subject matter of a non-opinion Crimson article, rather than The Crimson’s coverage of that matter, will not be accepted.

Letters to the Editor are evaluated at the discretion of the Editorial Chairs. They should be submitted to [email protected] and, with rare exception, should run between 150 and 350 words. We require Letters to the Editor to be signed, with the signatures appearing on the page or as a hyperlinked list at the discretion of the Editorial Chairs. We do not accept Letters to the Editor from organizations or anonymous writers.