Should a patent on an invention made at a university be managed any differently from a patent on an invention made in a company or made by an independent inventor?
That’s a fundamental question, and one that shapes university patent policies with implicit answers. For most American universities, that answer appears to be “no”–there’s no difference. A patent is a money-making asset, and any way one can make money is fair game–licensing or trolling, it’s all fair game. There is nothing in university patent policies that places limitations on university administrator exploitation of a patent–nothing that, say, makes the default a royalty-free, non-exclusive license; nothing that limits litigation for infringement; nothing that sets requirements for offering an exclusive license or limits the duration of such a license.
There may be in university patent policy language allowing a company to own inventions arising in research (for a higher fee) or to expect at least a royalty-free, non-exclusive license when sponsoring research (if the company pays for the entire cost of the research). But these sorts of policy statements are restrictions on money-making efforts, not restrictions on the general idea of a university-owned patent. These sorts of policy statements are there to deflect claims that administrators failed to make the money they have claimed to be able to make (to other administrators, to the faculty, to inventors, to the general public–mostly informally, but given how much attention is giving to the royalty-sharing formulas, money is the key thing. There are no rights-sharing formulas; no public-benefits-sharing formulas).
As far as university administrators are concerned, there is no difference between a university-owned patent and any other patent. Patent policies, where they do appear to limit administrators, is to authorize administrators, in certain circumstances, to not attempt to make the maximal amount of money–the implied policy then is “You have a mandate to make as much money as you can from the exploitation of university-owned patents, but in some situations, you will not be disciplined by the university if you allow a research sponsor to own an invention or receive a royalty-free license.”
The policy implied by the other sections of policy is one of maximum exploitation of patents for profit. Where university practice differs from conventional patent management practice is that the university rhetoric around invention management emphasizes licensing first, and exclusion second. In a company, one might expect exclusion to be the first line of thinking, followed by licensing as a strategic move to build a standard, or exchange technology with a competitor, and the like. But what happens in practice is that universities fail to license promptly and therefore exclude by inaction–no one may legally practice the university-owned invention but for a license, and university administrators, in seeking an exclusive licensee prepared to “commercialize” the invention, necessarily exclude all possible licensees who merely wish to practice the invention–in their research or in their other internal operations.
There are two likely outcomes from this initial condition. First, administrators do grant an exclusive license. In this case, all other possible users of the invention are excluded not as a matter of administrative choice but rather as a consequence of an enforceable contract. Access for use is now subject to the exclusive licensee’s control, usually governed by sublicensing and infringement clauses in the exclusive license with the university. If the exclusive licensee rapidly produces a commercial product that may be sold to interested users, allowing them to practice the invention, then things work according to the public rhetoric. But often, the exclusive licensee does not produce a product in a timely manner–producing a product is often much more difficult than simply practicing an invention. For a product, one must deal with budget justification, scaling up production, testing, meeting quality assurance standards, building in features anticipated by a substantial number of potential customers, managing sales force training and technical support, setting up distribution logistics, creating sales literature, obtaining any governmental approvals. In such cases, users remain locked out of access to the university invention for their own use. If they use anyway, they risk being sued for infringement.
In the second case, administrators fail to grant an exclusive license, and sometime later–often years–they discover that companies are using the invention anyway. At this point, administrators are faced with a dilemma. If multiple companies are using the invention without the need for exclusivity, then the university strategy of seeking an exclusive licensee was unfounded from the start–and in fact the patent served no public purpose, as it was not needed to encourage private development of the invention for practical application. Clearly, companies did their development despite the presence of a patent–if the companies were foolish enough even to check for the patent (and by doing so expose themselves to claims of “willful” infringement and treble damages). At this point, administrators could simply notify each company of the patent, grant each company a royalty-free, non-exclusive license outright, and offer additional services (such as access to data, expertise, and related tools and rights). But I don’t know of a single university that has pursued this route. Instead, with increasing frequency, administrators sue the companies practicing within the scope of claims of the university’s unlicensed patent.
Thus, there are four typical outcomes of the dual monopoly model (compulsory university ownership, default attempt to exclusively license to create a commercial monopoly):
- Exclusive license with timely commercial product
- Things work as publicly claimed
- Internal uses allowed only by buying the commercial product or by sublicense
- Exclusive license with delayed product or product never produced
- All other possible uses precluded by contract
- The university is paid to keep the invention from use
- No license, followed by infringement
- No patent was needed to encourage development
- No exclusive position was needed to encourage development
- The premise of a public service need for university patent ownership fails
- No license, no use
- A failure of the university’s patent management program
- A waste of university money
In terms of actual practice, of course, there is no data. Universities don’t publish their figures on what is not licensed and what is licensed exclusively but which does not result in a commercial product. For federally supported inventions, the Bayh-Dole Act exempts from public disclosure information provided to the federal government by universities on the utilization of subject inventions. But from my experience, the most frequent outcome by far is 4 (nothing ever licensed), followed by 2 (license into a failed product), and then 3 (no license but later infringement), and then, last and rarest, 1 (an exclusive license that results in a timely commercial product).
If one disregards the licenses to future patent rights granted to research sponsors and looks at only the activity of marketing university-patented inventions for commercial development, universities are unlikely to be licensing even half of their portfolio. The University of California at one point reported that no more than 1 in 200 inventions made it to a commercial product. At Stanford, 72 of 3600 (or just over 1%) inventions made at least $1m–or, $50,000 a year for 20 years (a little more in the time of 17-year patent terms). Now, $50,000 a year is unlikely to be indicative of the payments one would expect on a running royalty based on commercial sales–it’s more like an upfront fee plus annual or development milestone fees to keep the license in place. Again, at even the licensing programs regarded as top of the line–and Stanford’s certainly is–the commercialization success rate is low.
What’s interesting is that a university typically makes money from exclusive licensing to companies that fail to produce a product–there are upfront fees, patenting reimbursement fees, milestone fees, and license maintenance fees–all due typically before there’s any commercial product. If the license is to a startup, then the university may take an equity position. If the startup is acquired, the university may make substantial money–even if the startup or its new owner never produces a product under the patent license. The material effect of most university exclusive patent licensing then is to prevent internal use of the university-owned invention.
The university prevents such use on behalf of a future monopoly partner, from which the university expects to receive compensation. If a university only accepted compensation when a commercial product was on the market, then its reported licensing income would reflect the market activity of those products. A university adopting this approach would limit the duration of any exclusive license, to motivate product development. As it is, university patent licensing income is bi-modal: with perhaps one or two “big hit” deals over twenty years, and most of the rest of the income from non-producing licenses.
If the exclusive licensee has the right to make product, sublicense, or sue for infringement, then the patent license itself is an asset that is valued based on a future expectation of using any or all of these methods of extracting value from the patent. The university’s claim that licensing the patent is in the public interest is, at such a point, no different from licensing the patent for any purpose whatsoever. The university abdicates its public position in favor of reserving for future speculation an invention that otherwise would have been broadly available, had there been no patenting.
If university administrators actually wanted to make it clear they would not exploit patents for money against their public claims regarding the role of patents arising from university research, one might find language establishing the basis on which a university will own a patentable invention (not every time such a thing arises) and for what purpose (not simply to make money with the claim that doing so is in the public interest) and with primary or default practices (such as non-exclusive licensing, on reasonable, non-discriminatory terms, with or without royalty) and with what restrictions on anything that might otherwise appear to be profit-seeking (such as litigation for infringement, exclusive licensing, or the term of exclusive licensing). If there is a difference at all between current university-owned patent practices and everyone else’s practices, it is that university administrators wrap their money-making efforts in a shroud of moralizing rhetoric about universities and public missions and public benefits–but it is the actions, not the rationalizations, that matter. Otherwise, any thug is a saint if the thug says so.
Now, even if university patent policies don’t reflect it, there are differences in practice between how universities and others manage patents–and there should be even more differences. We will look at these next.