At Xerox’s Palo Alto Research Center, a lot of great technology has been created. Postscript, ethernet, graphical user interfaces, the mouse, among others. The problem has been, that these technologies were not able to make it into the host company’s product lines and instead became platforms for others to build on.
The typical way this is described is that Xerox should have taken stronger proprietary positions and extracted value from their research. I see it differently. To my way of thinking, the dilemma for an industry lab is the choice between feeding platforms that feed the industry broadly and holding things to take advantage of their own position. It is also the dilemma for the university lab.
Consider another company in Silicon Valley, Silicon Image, Inc. SII is responsible for, among other things, the DVI standard for connections between computers and monitors. Here is a relatively small company working things out with a bunch of huge players to deliver a new platform technology. What does it do for SII? It gives them a leg up building products that ride on the standard, that anticipate developments in the standard, and therefore can be first in and most capable implementers of the standard. This is no loss for SII, that DVI isn’t held out as a proprietary product for the company. The standard is a strategic asset, even though it is not owned by SII.
If we look at Teece’s development of innovation, the issue is cast as a competition for benefit among the innovator, the imitators, and the infrastructure. The idea of an innovation becoming a standard means one feeds the imitators. The imitators in fact become the innovators. The platform is the innovation.
Geoffrey Moore, in Dealing with Darwin, a book about the challenges of re-innovating in large companies like Cisco, argues there are only three kinds of innovation to consider: innovation that creates separation from the competition, innovation that allows one to catch up, and innovation that reduces costs in doing any of this. The rest of innovation, Moore asserts, is waste. That cuts it down. One can see in Moore’s development much of the schema outlined by Teece.
Separation = innovator
Catch up = imitator
Cut costs = infrastructure
But the PARC Dilemma suggests that there’s more to it. That one can turn the imitators into the change agent, leaving those with separation with their own proprietary innovation outside the emerging new platform. Call it a network effect.
A lot of the talk about interoperability, standards, open source, and the like is about the development of new competitive platforms. University patent licensing operations nearly entirely ignore this kind of thinking. It’s product separation or nothing. Yet one might argue that the particularly special role for university research is to feed platform development rather than product development. If we return to Bayh-Dole, which doesn’t dictate university practice but does now dominate research conditions, it becomes clear that practical application is just as readily achieved through standards as through monopoly products. Just university patent administrators don’t even try.
We might suggest that a significant effect of Bayh-Dole, reflected in the rise of university patenting of research originated inventions is to remove a huge stream of platform technologies and research tools from circulation in industry and community. The more patents the universities claim, the worse it is. Not because of the patenting, but because of the university implementations and fixation on sharing gifts with monopolists. Seen this way, it is not Bayh-Dole that stands in the way of innovation; rather it is the university implementations and their emphasis on “commercialization”. A few years ago, I was riding in a cab with two directors of university tech transfer offices. They were in agreement: they wouldn’t handle an invention if they were required to license it non-exclusively. Yet universities with this attitude continue to take ownership of the majority of inventions they see through their shops.
This development of the theme intends to put a lot of pressure on the idea that the only–or best–or first–role for university research patents is to give monopoly positions to companies. This also puts pressure on the idea that companies will only invest in new products if they have a monopoly position. And on the idea that without products positions, the inventions would not be used.
This addiction to the exclusive license is a big problem in university patent administration. Of course there is room for exclusive licensing, but university administrators really have not worked their way through the PARC Dilemma. They have don’t have effective ways to work across universities and industry to create interoperable platforms, libraries of tools, testbeds, ad hoc standards. You just can’t do this with one-off inter-institutional agreements. Faculty researchers do this by ignoring the tech transfer office, to the disgust of the officious class of administrators.
There are ways to use patents to develop platforms and feed American industry with robust new resources That’s what research enterprise and RTEI is about.