Can game theory explain grade inflation?

Question

According to wikipedia, grade inflation is the tendency of academic grades for work of comparable quality to increase over time. That article also includes plenty of evidence for the phenomenon and lists some potential causes. Has this issue been studied using game theory? What game-theoretic models of the educational grading process exist that could shed some light on the forces behind this phenomenon?

Answer 1

One paper that looks relevant is "A Signaling Theory of Grade Inflation" by Chan, Hao, and Suen (2007). From their abstract, it seems like this is what you're looking for.

When employers cannot tell whether a school truly has many good students or just gives easy grades, a school has incentives to inflate grades to help its mediocre students, despite concerns about preserving the value of good grades for its good students. We construct a signaling model where grades are inflated in equilibrium. The inability to commit to an honest grading policy reduces the efficiency of job assignment and hurts a school. Grade inflation by one school makes it easier for another school to do likewise, thus providing a channel to make grade exaggeration contagious.

Also check out "Comparative cheap talk" by Chakrabortya and Harbaugh (2005). From the introduction:

Are such statements more credible than claims such as “they both look great” or “every student is excellent”? How much information can comparative statements convey? When does it make sense to withhold comparative information? And, are comparative statements still credible when the speaker is not impartial, e.g. when a professor has a favorite student, or a salesperson receives a larger commission on a particular product?

In particular, section 4.2 discusses "Recommendation games". They consider situations where an expert with private information can rank alternatives for a decision maker. Here is their description:

In recommendation games we find that the expert prefers ex ante to reveal a partial ranking rather than the complete ranking. For instance, if there are three students being recommended by a professor and the middle student is unlikely to receive a job based on the complete ranking, an alternative is to put the top two students in a group and not differentiate between them. As the number of issues increases, such groupings can be used more and more effectively to maximize the expert’s payoffs. The gains from partial rankings may explain why highly ranked schools often obscure the relative quality of their graduates, either by grade inflation as in Ivy League undergraduate programs, or by withholding grades from employers as in some elite M.B.A. programs

One of the conclusion they reach is that grade inflation "should be more severe when average student quality is increasing" so "grades should be more inflated in elite schools". Look into their section 4.2 for a detailed analysis.

Answer 2

As a college professor, I can tell you that student evaluations are a major cause of grade inflation. College administrations use student evaluations of professors as a major determinant in promotions, assigning classes, tenure, any form of recognition.

From the professor's perspective, if you start to get too many bad evaluations, your career is in jeopardy. So, why not go with the flow? Call a C an A- and everyone is happy. Of course, the integrity of the educational system is destroyed in the process.

From the standpoint of the university administration, who wants the hassle of dealing with student complaints? The way to get ahead is to grow your program and generate income. This is especially true of MBA programs that are generally funded not by students, but by their employers. One way to compete with other MBA programs is to make the grading easy, but universities also make the degree programs shorter and the experience more entertaining .

Answer 3

Schools want their students to get more than "their share" of jobs. One way to do this is through grade inflation, that may convince employers that the one school's students are "smarter" than those of other schools with "lower" grades.

Of course, when the other schools catch on, they will raise THEIR grades too, cancelling out the first school's advantage, but causing grade inflation.

It's like watching a performance at a standing room only event. Any ONE person can get a better view of it by standing on tiptoes. But if ALL of them do it, this just cancels out. That's what game theory would predict.