Bryan Caplan  

Sociology and Signaling

PRINT
What I'm Reading... The State of the European Cris...
Back in 1995, I attended an IHS seminar for graduate students.  We heard some lectures, practiced our public speaking, and did mock interviews.  The last activity was pretty traumatic.  It's hard for a second-year grad student to role-play someone who's wrapping up his dissertation.

Part of the process was writing up a mock c.v. - which led to a moment I still remember.  One of the students wrote his GRE scores on his c.v.  During the denouement, the mock interviewers raked him over the coals:
You don't put your GRE scores on your c.v.  It's makes you look like a grad student!  It doesn't matter how high your scores are.  Schools want to hire creative assistant professors - not stellar grad students.
Good advice, no doubt.  But why is it good advice?  As usual, the signaling model sheds a lot of light.  If the average candidate who puts his GRE scores on his c.v. is professionally clueless even given impressive scores, it's a bad idea to include them.  So only professionally clueless candidates do so, reinforcing the equilibrium.

Of course, we can easily imagine a world where revealing your GRE scores is the height of probity, and failing to do so is professionally clueless.  In fact, we don't have to just imagine this possibility.  It's how grad school admission actually works.

So there are multiple equilibria.  But what determines which equilibrium actually occurs?  In all honesty, any plausible answer is going to have to appeal to customs, interpretations, and other concepts that economists usually associate with sociology.  If you're a narrow-minded economist, of course, this is just another excuse to sweep signaling under the rug.  But if you're a serious social scientist, you'll take a deep breath instead.  Do you really want to search for your keys under the lamppost because it's brighter there?  Or do you want to wander into the darkness of sociology and discover what's out there?



COMMENTS (4 to date)
gabriel rossman writes:

how about this?
people find standardized test scores distasteful and avoid them when possible. with faculty searches, there are fairly clear signals available that are sufficient for making the short list, most notably PhD pedigree and publication records that can be made commensurable by journal prestige. for deciding between the finalists there is the flyout.
in contrast, grad school admissions involves much younger people with fewer accomplishments to distinguish them and there are no flyouts because it would be prohibitively expensive in time and money to host 40 prospective grad students. test scores are therefore useful because they are useful at revealing potential when other information is lacking.
my argument is mostly the kind of stuff that (i imagine) most economists would be comfortable with. the real question for sociologists is explaining why people find standardized tests distasteful. for that i think a good start is Karabel's /The Chosen/ which explains how we got "well-roundedness" in ivy league admissions.

Robin Hanson writes:

Once economists get into customs and interpretations, they will call it economics, not sociology.

student writes:

so ad coms looking to fund a grad student want to know if he will be a good researcher. so they care about his gre score.

but, departments looking to hire a new prof care about finding good researchers. so they don't care about his gre score?

is the lesson we're suppose to draw from this post that academics don't *actually* believe the gre has anything to say about our ability to do academic research?

if so, why am i studying 2 hours a day for this bloody test now?

Hyena writes:

In order for psychometrics to replace other signals, you're going to need a memetic sweep. It has to happen in the face of mathematical confusion and legal barriers which greatly diminish its benefits. The earliest this sweep could have really began is the early 1990s and I don't think it much increases fitness right now.

I don't think there's been enough time, all things considered, for society to convert to psychometrics. College admissions are the exception which proves the point: it's a case where we have no other good signals to go on, granting immense marginal benefits, and psychometrics could be imposed regardless of public knowledge. It's telling that most people don't recognize SATs as IQ tests.

Comments for this entry have been closed
Return to top