I’m just now discovering that there is quite a bit of literature out there on signaling in matching systems.
The basic model is dating and suitors using a limited resource to woo the object of their affection.
Muriel Niederle has done some empirical work on this on an online-dating site. In this match, signals are sent with “virtual-roses”. https://web.stanford.edu/~niederle/Lee.Niederle.Rose.ExpEcon.2015.pdf
She reported that in that match, many suitors did not optimally use their roses and some did not even use all of their roses.
I’m a little skeptical about how analogous the online dating market is to the ENT match. The biggest difference I see is that the ENT match has dramatically asymmetric amounts of information. Applicants submit PSs, test scores, research experience, job history and other info. Programs make websites and video ads. Some list fellowship matches, but not all. All locations are known. Number of residents per year. Faculty names and Subspecialty. None list the number of fellows.
Doximity provides two rankings of programs. One is based on reputation. The other is based on research productivity of residency graduates.
Programs routinely have to sort through 200+ applicants.
Applicants have to sort through <200 programs to decide which to apply to and have to pay to apply to more programs.
So far much of this sounds reasonably similar to an online dating experience.
So what are the differences. First, online dating sites do not have a monopoly on romantic connections. The NRMP does. Second, online daters may have had a more serial monogamy approach to dating and so only sent out one rose at a time. This is very different from the ENT applicant cycle where applications are released on a single day. Finally, the problem of interview date competition is pretty much non-existent in an online-dating market.
Additional paper I didn’t have a chance to read yet.