Long Term Research

I was at the marketplace innovations workshop at Stanford. I enjoyed meeting the few faces I recognized and the multitude of new folks in the field. Thanks to the organizers for an excellent program. It was great to see folks attacking two-sided marketplaces in Uber/Lyft.

I left with a question. We have had a few iterations of spectrum auctions. These are supposed to be for pricing and using a common good to benefit the society. Do we know a definition of a good or successful auction? We know the auction has to satisfy some constraints for access to remote areas or signal a good price for the govt. We can prove some properties assuming the bidders know the longterm value of the spectrum.  I am not convinced this is a good proxy for what benefits the society, and there should be independent metrics to evaluate whether a spectrum auction ultimately yielded longterm benefit, even empirically and retrospectively.