Scott Sumner  

Grade deflation, grade illusion, and academic depressions

Causation... Quarters on the Sidewalk as St...

Standard monetary theory says that changes in the money supply and prices are neutral under certain circumstances, such as in the very long run, and also after monetary reforms where all contracts are automatically adjusted to the change in the money supply. In the short run, monetary inflation may have real effects, partly due to the fact that nominal hourly wages are sticky. If workers have money illusion they may confuse nominal and real wage changes.

How about grade inflation? Does that show a similar pattern? Here is C.W. at Free Exchange:

Some colleges have pursued anti-inflation policies of which Paul Volcker would be proud. In 2004 administrators at Wellesley College, a prestigious, women's-only university, mandated that in introductory and intermediate courses (with at least ten students) the average grade could not exceed a B-plus, equal to a grade-point average of 3.33. Three economists look at the impact.

Only courses in high-grading departments in the humanities and social sciences needed to change grading practices: science subjects were unaffected by the policy. That gave the economists a good "control", allowing for a meaningful analysis of the policy.

What happened? Previously generous departments became more tightfisted. Students were 14% points less likely to get an A in the treated departments (though they were no more likely to get a C-minus or below). Lots more Bs were given.

OK, so Wellesley adopted a contractionary, or "tight grade policy." And grade deflation set in for some departments. Before considering the results, let's consider what might have happened. One possibility is that the policy had no real effects. People saw through the lower grades and realized that Wellesley students were doing just as well, it's just that standards had tightened up. Students in the 85th percentile would be perceived as doing just as well, despite lower nominal grades. In fact, that's what I would have expected to happen. But I was wrong:

More interestingly, the cap changed students' course choices. For courses in the treated departments, enrollment fell by about 19%. Students were 30% less likely to major in one of these courses.
So Wellesley's tight grade policy created a depression in the humanities industries departments, just as a "grade illusion" proponent might expect.

I suppose there will be some crazy Austrian readers who complain that I've got it all wrong. It's not the depression that's the problem, it's the preceding humanities boom. The grade deflation is merely wringing the excesses out of the economy academy. The Economist continues:

These results are positive and other universities can learn from them. Before the policy the difference between the profligate and the parsimonious departments could exceed 0.6 grade points. The hope of higher grades could have encouraged some students, who would really have preferred to study sciences, to move to humanities. But by grading more uniformly, Wellesley removed this perverse incentive. Universities should take note and encourage their students to study what they find intrinsically rewarding; not what will give them bloated grades.
OK, so the Austrians are right in this case. But I still think they are wrong about 2006-09.

PS. And I still think if grade inflation were tackled on a nationwide basis the effects would be neutral. But who really knows?

Comments and Sharing

COMMENTS (12 to date)
trent steele writes:

As a "crazy Austrian," I object to the characterization and ... *finishes reading post* ...
Well I'll be...

Scott Sumner writes:

Trent, I'm considered crazy by lots of people, so I always use the term 'crazy' with affection.

Edogg writes:

Anybody want to argue for or against the idea that humanities and social sciences courses, especially "introductory and intermediate courses," are just easier than the science courses? In general, college courses seem to me to be more about meeting standards and being consistent rather than showing creative genius. (I don't consider this necessarily good or bad.) It doesn't strike me as improbable that most students who enroll in an introductory survey course in the humanities can consistently meet standards in writing the appropriate papers and taking the appropriate tests. We don't want grades to artificially magnify differences anymore than we want them to fail to reveal differences. I suppose the problem is if future employers or graduate school recruiters fail to distinguish between A's earned in different departments.

Brent writes:

To really get into analogies, grade inflation is trying to paper over the problems of the bloated higher ed sector and makes the problems worse at the same time. Hard choices and difficult reality adjustments be damned!

Jon writes:

Haha. Funny post Scott.

I think an analogy to money demand would be better. Factors which increase demand for a course include hedonism, accrued human capital, and GPA.

Knock out "good GPA" and you reduce demand for the course.

Roger Sweeny writes:

The grade is a signal, and it can signal two very different things. A high grade can signal, "this student is good" or it can signal, "this course is easy." The Wellesley policy seems to assume 1) outsiders (especially employers) think all courses are of equal difficulty, 2) all courses should be of equal difficulty.

Scott Sumner writes:

Edogg, You said;

"Anybody want to argue for or against the idea that humanities and social sciences courses, especially "introductory and intermediate courses," are just easier than the science courses?"

I certainly think they are easier to get an A in, but I don't see them as being intrinsically easier. Very few students actually master the material in any college course.

In any case, it's an apples and oranges deal, where it's impossible to say what sort of performance in English literature corresponds to a given performance in chemistry.

Grades are presumably a way of conveying information about relative performance. The interesting question (which your comment alludes to) is whether it should be relative to other people in that class, at that school, or in the US as a whole. Thus if Harvard students are compared to all students in America, most would deserve As in every course. If compared to other Harvard students, you might want a wider distribution of grades to convey information about the relative merits of the different Harvard students. It all comes down to what you think the purpose(s) of grades are. Presumably to convey information---but what sort of information?

Anthony Deluca writes:


Employers/Outsiders surely realize that humanities generally programs grade easier than science programs. But employers only have a rough understanding of how roughly different humanities programs grade. And certainly many of them are going to be totally unaware of a radical shift in policy at a single school.

English Professor writes:

Ask any English professor teaching at your typical college or university, and he'll tell you that his students don't write well. Then ask him why he gives those students high grades. You'll hear one of these answers: 1) That's what students expect nowadays; I don't have a choice. 2) If I gave them lower grades, I'd get poor evaluations. 3) Oh, my students are all so bright that they deserve those grades. The last group are muddle-headed enthusiasts who believe all the education happy-talk you get from the "high self-esteem" crowd. Most of the people in the other two groups don't really believe they are giving honest grades; they are either accommodating themselves to the times or responding to the perverse incentives set up by an academic culture that gives a lot of weight to student evaluations of courses. If you give students high grades, you are in effect telling them that they are smart, which they like to hear. If you give lower grades than your colleagues give, you are telling your students that they're not as smart as they think, which they resent. They then take it out on you in their course evaluations, and bad evals can lead to below-average raises. So what would you do?

Thomas Sewell writes:

If you evaluate both the students and the teacher based on the students' grades, you're inevitably going to setup a situation with perverse incentives.

As others have proposed elsewhere, the solution is for the teacher to teach, the students to learn and a different third-party with different incentives to perform the evaluation against a set standard of how much the students know before and after.

English Professor writes:

I completely agree with Thomas Sewell on the basic solution--third party grading with pre- and post-course evaluations. But you won't find many professors in agreement, mainly because it would show up how little students really learn in their courses. I suspect you would find even fewer colleges willing to introduce such practices, for it would reveal how little value they add to a student's knowledge or abilities.

mbka writes:


you should be awarded the "most consistently original thinker on the internet" medal.

You could have added expectations in the mix though. The depression was obviously generated by the market's expectation that the now nominally lower grades would sell for far less than previously (employer's sticky equation of nominal grades with quality), so their production shrank to a lower real output. What started as a nominal loss in grades ended as a real decrease in production of human capital.

Comments for this entry have been closed
Return to top