[M]arket design is an eclectic field, drawing on game theory, experiments, computation, and field observation of all sorts (rules are data!).”
… When Paul Milgrom and I began the class (when he spent a year at Harvard in 2001), he had the FCC spectrum auction experience under his belt, and I had the redesign of the National Resident Matching Program under mine, and we had plenty of ideas.
I entertained a faint worry that, at the end of the decade, those might still be the only major applications we had to talk about. But, as things turned out, we can no longer fit all the newly implemented market designs into one course (and Susan Athey will again teach a second semester of Market Design, focused on many recent auction applications, in the Spring). Among the designs we talked about this semester are other health care labor markets, Kidney Exchange, School choice mechanisms, signaling for new economists, internet ad auctions, and more.
I’ve also been gratified by developments in market design as a field of study. Not only have there been successful applications, there’s starting to be an academic literature focused on practical market design, and the theoretical and empirical questions it raises. [Links in source.]
And speaking of “rules as data,” in the prior post Roth notes that rules governing assignment of kidneys from deceased donors may be changing. (See related story from the Arizona Republic.) In the comments I asked, “Why not favor patients with an unmatched donor, and so use deceased donor kidneys to trigger a exchange chain?” — seemed to make a lot of sense to me — and Roth explained both why it is complicated and a few cases in which something like it has already been done.