For those universities that have made it a centerpiece of their practice to start companies as a way to get rich and create jobs, there’s a sobering critique of using the venture capital route to do it in the April issue of Wired.
…The VC model is based on creating wealth for investors, not on building successful businesses. You buy into a company early on and sell out a few years later; if you pick well, you make lots of money. But your profits don’t accrue to the company itself, which could implode after your exit for all you care. Silicon Valley is full of venture capitalists who have become dynastically wealthy off the backs of companies that no longer exist.
The venture model is about turning a profit for investors by buying early and selling as soon after as possible.
Now enter university administrators looking to make money from startups, and looking to create investment funds to for this purpose when private investors won’t bite often enough. Administrators might say company formation is about economic development and delivering university discoveries to the masses, but the *model* is about making money for investors–and the university wants a cut. The model operates from investment to exit. Exit doesn’t create jobs or products; it merely secures a return on investment. The model is designed to create investor wealth, using the advancement of company interests as the driver. In practice, the model is not even designed to advance the interests of each and every company–just one in five investments needs to work out, a few break even, and the reset can go bust. The model aims to spread the risks so as to create a greater likelihood of one or more “big hits”.
A university’s mandate, in Bayh-Dole’s funding agreement, if not in policy, is to promote the use of inventions and other assets created in the University’s research programs. It isn’t a portfolio model of some big hits and a lot of losers. It is a mandate to promote the use of everything, whether the university claims ownership of it or not. When a university does claim ownership, inventors have an expectation that the university, as the self-appointed agent of their work, will make the effort to license their work, along with promoting it. Clearly, if a university doesn’t have an ownership position, then it also has no need to grant licenses.
In this, one should be able to see that an emphasis on “commercialization” is a narrowing of a university’s activities with regard to research inventions. Some inventions will never be commercial products, but rather platforms or standards. Others will be broadly available research tools before anyone sees the need to create a commercial version.
To focus primarily on startups for that commercialization is narrower yet. To focus on startups to make money represents only a tiny portion of the research assets that might be subject to a university ownership claim in any given year. One can make a virtue of the effort to start companies, and a virtue of making money from starting companies, but in this activity a university is also publicly abandoning the idea that its role is to promote the dissemination of knowledge and make the expertise of its personnel broadly available. In its place is the argument that such stuff preferentially happens through startup companies that receive venture backing with the expectation the startups will be sold off.
It’s really important to see the differences. In a venture model, one might look at a 1000 business plans and maybe take a closer look at 50, and invest in perhaps 10, looking for one or two golden exits. The VC doesn’t hold all the businesses until it’s clear the rest are worthless–if they don’t meet basic investment criteria, the VC passes on them. In a university IP model, the university commits to try to place everything that it takes on. If not a startup, then a license to an existing company. In a compulsory ownership program, the inventors and developers are not voluntary participants. They have not chosen the university as their agent, haven’t negotiated an arrangement that sets out their responsibilities, lines of control, and benefits for success. The university, not the inventors, has control over the licensing objectives, partners, and terms. This is not a portfolio model. This is a compulsory agency model. With it comes more responsibility, not less.
Imagine a sports agent who took on 300 players but only cared about signing 10. The other 290 players would be looking for new agents. But if the sports agent had a deal by which he could require all 300 players to deal only with him, then it would look tremendously unfair, if not immoral. How can someone set up as an agent, and then say, “Everything is okay, so long as I, the agent, make money and can show some few “successes”–the rest of you guys should be *happy for me*”. Yeah, it’s nutso. It would never happen. But this is happening in university technology transfer as universities modify their policies based on past voluntary agent dynamics to become compulsory–and tighter yet with present assignments to do away with review and request protocols.
When a university takes ownership of IP, it takes on a responsibility to place the technology subject to the IP, not merely to make money. It’s no good to say, even, that we are trying to make money, too, for our inventors. That is, it’s a sham to say: we take this technology on for management, not so that it can be developed and used for public benefit, but rather so that we can make money on a first licensing transaction and then it’s someone else’s problem what to do with the technology they just acquired. Licensing money comes after public use (royalty on sales, say), not instead of it (equity interest waiting for a flip).
A university takes ownership of IP to facilitate transfer technology. This argument is, in fact, the fundamental argument. Chase it all the way back to Vannevar Bush and the purpose of university research–especially extramurally funded research–is so that the results advance community and industry. This is the anchor of federal funding. One manages inventions so that they get used. It is not acceptable to say, no matter–we made money on the deal, too bad it didn’t work out. Worse and worse if the technology then remains trapped in an exclusive license, undeveloped and unavailable to others.
There may be good models for institutional support for startups based on university research IP. Creating companies to flip them is not that model. It’s an exploitation of university research, not an extension of the university’s commitment to public service. It makes the university’s entire research program appear to exist merely for institutional self-interest in selling off results, and it makes technology transfer out to be a kind of welfare activity for venture investors, supplementing their funds with university and state “funding gap” dollars in the hopes that they will find a buyer, and quickly.
While there is a time and place for venture backed startups, making them the centerpiece of a university’s IP policy and program breaks the social and imaginative contract by which universities and sponsors support faculty research.