Glorfindel of Gondolin

Ever since the Bayh-Dole Act was passed, universities have scrambled to patent everything they could in the hopes of reaping rich rewards from the licensing revenues that are occasionally possible when you hold a lot of patents.

This article from JAMA ($) suggests that that revenue is not in fact very significant for the vast majority of universities, and that in many cases the costs of maintaining a technology transfer office to handle the patent applications and the licensing eats up any revenues that flow from technology licenses.

Only a few universities reap large net revenues from licensing. Surprisingly, these aren't always the universities that spend the most on research. Harvard? Stanford? Neither make a ton of money from licensing, at least relative to their research expenses. The universities that score big usually score big because they get lucky. For example, Florida State pulls in big cash because they have the patent on the anti-cancer drug Taxol. The University of Florida cashes in on the Gatorade trademark (wtf??).

The authors argue persuasively that, as a tool for motivating productive research, technology licensing hasn't had a huge effect in the U.S. They point out that other countries considering this model ought to worry that if they don't have a robust system of publicly-funded research, relying on technology licensing is more likely to slant research toward the benefit of the rich, who can pay large licensing fees. That this hasn't happened in the U.S. is probably because we still have a healthy amount of public research funding available.