Grading on the curve

It has lots of virtues, but one major vice: it creates a disincentive for cooperation.

Eugene Volokh defends grading on the curve I mostly agree. I do it myself.

But curving the grades has a major disadvantage Eugene doesn’t discuss: it discourages cooperation among students. In practice, no matter how explicit the syllabus is, grading policies are sufficiently opaque to students that the perverse incentive effect may not be very great. But it’s there. Any student who helps another student learn is moving the curve in what is, from his perspective, the wrong direction.

I don’t go as far as Mike O’Hare does in thinking that the classroom setting should reflect the real-world setting. But surely teaching students that achievement is purely competitive, and competitive at the individual level, ill prepares them for most workplaces, to say nothing of voluntary organizations and households.

There’s a famous experiment in breeding hens to be raised in cages of six to nine hens each. Plan A takes from each cage the hen that produces the most eggs, and breeds the next generation from those hens. Plan B takes from all the cages the cage that produces the most eggs, and breeds the next generation entirely from that cage. Within a few generations, Plan A breeds scrawny, nasty-tempered hens that specialize in breaking one another’s eggs: the avian analogues of pre-meds. Plan B breeds plump, placid, and highly productive hens, and lots and lots of eggs.

Grading on the curve is analogous to Plan A.

Of course that’s a metaphor, not an argument: incentives aren’t fully analogous to selection pressures. But using a grading structure with perverse incentives built in seems like both a bad idea in itself and a bad example to set for students who will, if they wind up running anything, face the problem of how to design a reward system.

Author: Mark Kleiman

Professor of Public Policy at the NYU Marron Institute for Urban Management and editor of the Journal of Drug Policy Analysis. Teaches about the methods of policy analysis about drug abuse control and crime control policy, working out the implications of two principles: that swift and certain sanctions don't have to be severe to be effective, and that well-designed threats usually don't have to be carried out. Books: Drugs and Drug Policy: What Everyone Needs to Know (with Jonathan Caulkins and Angela Hawken) When Brute Force Fails: How to Have Less Crime and Less Punishment (Princeton, 2009; named one of the "books of the year" by The Economist Against Excess: Drug Policy for Results (Basic, 1993) Marijuana: Costs of Abuse, Costs of Control (Greenwood, 1989) UCLA Homepage Curriculum Vitae Contact:

11 thoughts on “Grading on the curve”

  1. I don't know it encourages competitiveness to the point that cooperation is discouraged. Considering the size of the classes involved, the benefit you get from cooperating with two or three other students outweighs in most cases the detriment from similarly assisting those two or three others. In fact, if most other students are cooperating, it could hurt a student's chances to go solo (assuming that's not that student's preferred study method anyway). In this way, the curve (at least in law school) might cause survival of the most cooperative, since the jerks won't find people to work with and will suffer for it.

  2. Grading on the curve is OK as long as it is based on the half-wit scale. The top score defines a perfect result as it is acheived by the student with the most wits. Half that score is an F – suitable for half-wits.

  3. Personally, I would argue that all grading systems are inherently conter-productive (I am heading to grad school in math this fall, so I am not merely someone who is bitter over academic failure). It is extraordinarily difficult to design any form of testing system that rewards understanding of the material and not merely recitation–this has become a rather large problem in mathematics and physics. Not to merely complain, I would argue that a better system would be similar to the British system of recitations. It is much easier for a recitation instructor to probe a student's understanding by asking questions that may be unfamiliar to the student and seeing how that student is able to think through the problem (much like oral examinations are done in American graduate programs) than it is to write a fair test. However, the great drawback in this proposal would be instituting it on a large scale (vis. large lecture classes at research institutions).

  4. The best argument, in my view, is that norm-referencing has been tried on a large scale with skin in the game, and it was called Enron's personnel policy ("Rank & Yank")
    Further to that, the stupidity factor. A few idiots are enough to drag down the average, and hence the mark required to get a top grade. Equally, the occasional genius will have the opposite effect.

  5. 1. The nasty hens are not the avian analogue of pre-meds; they are the analogue of Republicans. Some pre-meds are caring, cooperative people.
    2. Imagine if Olympic events were scored on a curve. Someone would HAVE to get a ten, even if all the performers had spills, errors, etc. And most performers would HAVE to get fives. Of course, in the Real World, people would never stand for this. Grading on the curve is silly, but easy. It happens when the instructor cannot define and recognize clearly the standards of competence. I second Daniel's comment above about recitations. Italian universities use a similar approach.

  6. In my world, Olympic events *are* graded on a curve – the highest scoring individual in a trial wins that trial.

  7. June and Barry,
    You're both right in a way. The Olympics are scored on a curve (for example, the lowest time wins), but there are minimum acceptable standards for Olympic qualification.

  8. "The friendship of students and of beauties is for the most part equally sincere, and equally durable: as both depend for happiness on the regard of others, on that which the value arises merely from comparison, they are both exposed to perpetual jealousies, and both incessantly employed in schemes to intercept the praises of each other."
    -Samuel Johnson

  9. Having received both a M.P.A. and a J.D. from the same university, my experience is that Plan A is the Law School and Plan B is the Public Policy/Admin program. (Sure it is a generalization.)
    As to Prof. Kleiman's comment that: "using a grading structure with perverse incentives built in seems like both a bad idea in itself and a bad example to set for students who will, if they wind up running anything, face the problem of how to design a reward system." He should see how law firms (particularly large law firms) that have resisted hiring non lawyer managers with real authority are run.

Comments are closed.