Sunday, November 15, 2009

Signaling for interviews in the Economics Job Market: Registration opens tomorrow

Here is the American Economic Association's announcement of the 2009 mechanism for Signaling for Interviews in the Economics Job Market

"The AEA coordinates a mechanism through which applicants can signal their interest in receiving an interview at the January meetings. From November 16, until December 1, shortly after the December JOE comes out, each applicant on the economics job market can register and designate no more than two departments (or other employers) to whom to send a signal of particular interest. On December 3, the AEA will transmit these signals to the departments a candidate has chosen. (Signals will not be made public.)

Please see Signaling for Interviews in the Economics Job Market for a detailed description as well as the Terms of Use and Privacy Policy. "

Those already registered may select or update signals until Midnight, EST on December 1, 2009. This is a firm deadline.

In each of the years when signaling has been available (2006-9), about 1,000 signalers have used the system (almost all choosing to send 2 signals). Around 25% of JOE listings have received at least one signal in each of those years. So signaling has been pretty widespread, with lots of senders and receivers.

So, that said, should you signal? Or perhaps there's some reason not to signal?

In a December 2008 survey of those appearing on departmental lists of job candidates, 66% of respondents reported signaling. Of the remaining 33% of respondents, 26% reported that they missed the deadline, 21% reported they didn’t know about the mechanism, 41% thought signaling wouldn’t help their chances of getting a job, and 5% thought signaling would hurt their chances of getting an interview.

On this last point I can offer some reassurance; in our surveys of department recruiting committees, there is no indication whatsoever that anyone regards a signal as a negative indicator in deciding whether to offer an interview.

There is also some indication that signals help, at least sometimes (again from survey data, since we don't have direct access to data on who gets interviews), particularly for signals sent to employers who don't receive an excessive number of signals. (The 21 employers receiving the most signals received about 1/6 of all signals: ten of these employers are in the Boston, NY, and D.C. metropolitan areas.) Overall, a signal seems to increase the likelihood of an interview by around 6 percentage points.

I think it is sensible for most candidates to send the two signals they are allotted. I think the most you can expect a signal to do is to cause someone at the signaled department to take one more look at your dossier, and consider whether it makes sense to include you on their already pretty full schedule of interviews, given your interest. So, choose your signals with that in mind, and they may work for you.

(I hope to have a paper ready for circulation reasonably soon, with my colleagues on the AEA job market committee...)

2 comments:

Awesome0 said...

Of the signals I sent, they both "worked". Would I have gotten the interviews anyways, maybe....but I think in particular when you have a chance to tell an employer otherwise that you are interested in a job for geographic reasons (close to family, friends) it can be a great tool. A signal to MIT, probably useless because everybody would want to go there.

cgar said...

For an interesting study of optimal signaling in job markets: "Signaling in matching markets"

http://www.econ.ucla.edu/workshops/papers/Theory/ColesKushnirNiederle_sigmatch.pdf