Universities and open innovation

Last month I was at the meeting of the University-Industry Development Partnership (UIDP) held at the National Academies conference center at UC Irvine. I was invited to give a talk about open innovation, and so decided to write a presentation trying to use open innovation to interpret university-industry cooperation in innovation. When I was working on the talk I got feedback on the first draft from David Wood, who also has been thinking about university-industry relations.

Before blogging on my own talk, I was waiting for the slides to get posted to allow people to follow along. Then I got buried. As of today, they have slides for all but one of the UIDP talks, including mine.

In my own talk, I began by summarizing a comparison of open innovation (Chesbrough), user innovation (von Hippel) and cumulative innovation (Scotchmer). This was based on my EURAM talk last year.

Next I noted the large body of research on the economics of university innovations, with work by noted economists like

• David Mowery (UC Berkeley) and his edited book on Bayh-Dole
• Marie Thursby (Georgia Tech) who studies university licensing
• Maryann Feldman (U. Georgia) and her book and various papers on university tech transfer
• Scott Shane (Case Western) and his book on academic entrepreneurs.

I encouraged people to read the original works, since there was no way to cover it in detail. Instead, I oversimplified this research (and the process of university innovation) down to two key choices: who pays for the research (and its associated IP), and who gets access to that research. I then summarized the characteristics of the two main commercialization model.

The familiar model is the university licensing model, epitomized by Bayh-Dole (well covered by the Mowery book) and tech transfer offices. This is quite consistent with open innovation. In a Chesbrough (or Teece 1986) sense, universities lack the resources necessary to directly commercialize their innovations — so they need to contract with private firms to realize the value of the innovations and bring them to market.

In this model, universities patent their technologies. They charge a fee for their innovations; for exclusive licensing or contract research, they effectively surrender rights to practice the associated innovations.

The other model, however, is the model of unmonetized spillovers. This corresponds to the Merton’s model of open science, as more recently elaborated by Paul David. This is best known as the way the DoD-funded Internet technologies got commercialized during the 1980s and 1990s. The idea is that if your research is being paid for, a university can allow spillovers, which greatly facilities the process of cumulative innovation.

My interest in this model was based on my own research on the commercialization of Claude Shannon’s information theory for deep space communications. (The paper will be published in the December 2008 issue of the Journal of Management Studies, as part of a special issue on “Research and Technology Commercialization”). As part of researching our book on the San Diego telecom industry, I found how MIT-trained electrical engineers (including Qualcomm founders Andrew Viterbi and Irwin Jacobs) helped NASA adapt Shannon’s theory to improve communication bandwidth with interplanetary space probes.

I didn’t mean to imply that the technology licensing model is dead, but instead to encourage participants to realize that each model has its role. Formal licensing is always going to be important in some industries (such as biopharma), but for other types of innovations, the spillover model can be more efficient. (Universities can often monetize this with donations from alumni entrepreneurs).

While at the conference, I heard an interesting wrinkle on the spillover model in the talk by Broadcom Chairman Henry Samueli. Samueli is a former UCI professor who’s giving naming grants to the engineering schools at both UCI and UCLA.

Broadcom spends about $1 million a year with universities: all their university grants are as gifts rather than research contracts, to reduce paperwork on both sides. Their goal is to identify prospective Ph.D. students to hire, but I imagine that the gifts also create considerable goodwill with the recipient universities for attracting other bright students.