Slate’s Matthew Yglesias has a nice post on grade inflation at top universities.  He observes that because top colleges and universities have become much more selective, part of why grades are getting higher may be that better students are attending them.  One way to think about it is that the grades are exchangeable by university–that same student that got an A in a Harvard course would have gotten an A anywhere–it’s just that the student who would have gotten a B- in that Harvard course didn’t get into Harvard in the first place.

I’ve been pondering this issue lately, because I’ve recently changed jobs–from a highly selective liberal arts school to a large state university.  I’ve had the opportunity to see first-hand the selection effects Yglesias describes, and to observe some of the behind-the-scenes assumptions that seem to go into grading by professors.  I agree with the main thrust of his point–that selection effects drive these grade distributions more than do coddling of students at top universities.  You find great students at every school; but you many find fewer students eking out C’s and D’s at more highly selective colleges.  Other factors play a role in determining grade distributions too, of course.

What makes grade inflation a problem?  I think most perspectives on this suffer from a severe case of the money illusion. People think an A means something, when it is actually just a signifier whose purpose and interpretation are culturally set at any given moment. This is a major misunderstanding that I observe among faculty especially, but also students and maybe parents.

Grades are a means of communicating value–like a price on a good; it makes sense to think of grades as a kind of currency.  From this perspective, almost everybody agrees that within a single course in a particular semester, grades should represent something fungible: if you and I got the same grade, then while our brilliance and effort may be distinct, it should be roughly of the same caliber. Where it gets confusing is when we extend this outside a particular course.  Is an A in an anthropology course fungible with an A in a physics course?  Is an A in an intro psychology course from 50 years ago fungible with one now?

This question may be more clear if we imagine students’ transcripts as price tags, specifying a particular value for an underlying good (this is probably something like the perspective of an employer).  Now imagine cutting everyone’s transcript into individual lines, and swapping lines with the same letter grade with each other–subject to the constraint that you only exchange lines with the same letter grade. If an exchange doesn’t change the value of the transcript, then the grades were exchangeable.  If a class of grade swaps doesn’t change the value, then the class represents a fungible asset. Something like a GPA presumes that these grades are fungible everywhere within a university–that that anthro C and that physics C feed into the same equation in the same way, and so do that senior seminar and that big intro class taken freshman year.

From this perspective, one ‘problem’ with grade inflation is when graders treat grades as fungible across institution, while purchasers (employers) do not.  So an employer says "That student got a 3.62 GPA at Yale–they must be doubly great!", when in fact the same student would have gotten a 3.5 at nearly any school (Okay, maybe not Princeton). Yglesias assumed just this kind of fungibility in his essay; I believe that I certainly do in my grading. When I first went to the University of Richmond, I found my grades went through the roof–because I kept finding myself amazed by how well each student handled the "tough" assignments I was handing out.  My personal grade distribution stayed pretty stable after that, suggesting that I treated the grade as fungible across schools.

The only real problem, of course, is that we don’t as a society agree about what we want this signifier that is so important in career success and documenting college outcomes to, err, signify.