I make this claim: The most important innovation metric of a university technology transfer program is the number of unlicensed inventions for which patent rights are claimed.
I argue that unlicensed patents suppress collaboration and innovation, work to the advantage of the status quo, and create uncertainties that do more damage than licensing does good. If all this obvious to you, then no need to read further. Consider the differences between invention and patent. An invention is a research event. Inventions take place in the minds of practitioners. An invention is an epiphany, a recognition of how things can come together, various parts relate, function is enabled, stuff gets done. An invention represents progress over the known, public ways of doing things. An invention is an indicator of the awareness of the practitioners to the context of their work–what can be done, what has been done in the past, and what would be a good thing to be able to do. The realization might be directly in line with the area of study, or it might involve an improvement to equipment, or it may be entirely skew, arising because someone made a mistake with reagents or left a petri dish by an open window or let the mind wander from the things at hand.
A patent is a different thing altogether. A patent legally excludes practice, but for the approval of its owner, backed by governmental authority. A patent is the result of a deliberate economic action—the application for patent.
An invention is a cool thing, something potentially attractive to consider. A patent, by contrast, is at its core a threat. A patent says: “You may not practice this invention without permission, which doesn’t have to be given. Furthermore, permission comes by way of a license. And that license is to be embedded in a formal contract. Under that contract you must agree to be diligent, report, and pay, or we will sue you for infringement and damages, pull the license, and hand things off to your competitors. The contract is subject to all the requirements our legal counsel can think of to shift all the risk to you, and to give us a direct line to your resources to take you down in court if you should breach the agreement.”
Given that the whole point of university technology transfer is to encourage use of inventions arising in research, one would think that someone wanting to use such an invention would be greeted warmly. The exchange would go, “Would you show us how to use your invention?” “Yes, of course!” But no, that is not it. The point of university technology transfer most places is to make money for the university, and in particular, to make money from a license deal (rather than from research grants or services or donations or subscriptions), and preferably a lot of money. The use of the invention is narrowed to money-making activities, and those further reduced to money-making activities involving a patent license. This is generally called “commercialization” by university technology transfer offices, but it does not mean products sold in a marketplace–it means, rather, “making money by licensing with the premise that someone may make products sold in the marketplace”.
Patenting and licensing create significant transactional costs that ride on top of any effort to test or adopt a university invention. Despite the assertion that a patent is a necessary step in the “commercialization” process, and even if this step is indeed important, patenting introduces serious overhead–not the least of which is the expense of obtaining a patent, which often runs to more than $10K per issued US patent, and often much more. Foreign patents, where there are translation expenses as well, can easily be $20K each. Expense, however, is not the greatest overhead. That honor goes to the contractual relationship proposed by which patents are to be licensed to promote “commercialization.”
Consider the university standard practice of granting exclusive patent licenses. Some universities, notably MIT, have made a big thing about how important exclusive licenses are. But think about it. If a university proposes to grant an exclusive license, then just one company gets the benefit of the license (if “benefit” is the right word here). Everyone else is cut out. What’s the point of working with such an invention, if one isn’t going to jump at being the exclusive commercial proprietor. No point in asking for research use, or trying to developing something for in-house use. Eventually the university will license exclusively and everyone else will be cut out. The clear message is, don’t both with the invention at all unless you intend to make a product from it, and take an exclusive relationship with the university.
University “commercialization” rhetoric therefore creates significant uncertainties for research, for internal applications (the “maker” community), and for industry in its production and sale of products. Nearly everyone has to wait it out to see who the university does a deal with and what the terms of that deal are (if the terms ever become public). The terms will indicate what rights are left available—territory, field of use, research purposes, and the like. Without that information, the conservative advice—and the very best advice—is to avoid university inventions.
And that’s the advice company attorneys give their scientists and technicians: “Don’t work with university inventions under the shadow of patents held for “commercialization”. Don’t work with university labs that could have patents held by administrators. Don’t seek out their publications, because if the tech transfer office licenses exclusively to someone, then our company is exposed to charges of willful infringement—triple damages and attorney’s fees–if we happen to be practicing the invention in our R&D programs.”
Think about it the way most any company has to think about it. If the university follows through on its aspiration to license exclusively, then it will license to some company that will try to shut down any use we might make of the invention. So don’t bother with the invention unless it is going to make it into our product line. Furthermore, even if the university fails to find a licensee, it can still turn troll on industry and go after any users with the claim that it is “protecting its rights”.
When a university claims patent rights in an invention, for almost everyone (except the desired exclusive licensee), the university has hung out a sign that says, avoid, design around, ignore, suppress, cut off, do not acknowledge, do not collaborate, do not ask for assistance, do not show interest. The number of unlicensed patents indicates the extent to which the university has been willing to hang out this sign.
For exclusively licensed patents, one can see there’s a similar story for almost everyone. There is a difference, however, in that it is possible that if the exclusive licensee really does create product, then at least that product can be purchased by most anyone and things can get back somewhat to normal for the bit of the patent that covers the product being sold. But the time from license to product is every bit as uncertain for others as if the patent had never been licensed. And for any other part of the patent that is not being exploited by the licensee, there won’t be any getting back to normal.
For individual inventors, perhaps it is tolerable to hang out a sign that says avoid or design around. Companies do this, however, at some risk, especially in areas where everyone has patents and everyone is also loath to use them. In those circumstances, there’s a lot of cross-licensing and standards setting to deal with claims before something goes nuts. Universities aren’t individuals and aren’t companies. Their approach to patents cannot follow the individual path of waiting, with latent threats, for a buyer, and cannot be modeled on patent trolls who simply exploit their rights for payment. Universities don’t generally need cross-licenses and don’t do so well setting standards.
The effect of university claims on patent rights extends well beyond commercialization and also hammers technology transfer programs at other universities, which in turn hammers research collaboration between industry and universities. Consider the ripple effect of unlicensed or exclusively licensed (without product) patents.
Related research at another university will be potentially blocked by an exclusive license. If university A has developed patent rights in an invention, and investigators at university B have got something that relies on A’s invention, then folks at university B are up S. Creek. They can license but they can’t enable use without everyone who wants B’s stuff also getting a license from A. But A isn’t going to grant a license to everyone just so they can all go off and work with B’s stuff. And if A does think about a license, then it will want a cut of B’s action. That leads to what is called royalty stacking, where a company is forced to pay royalties to multiple organizations eventually cutting its margins to nil or driving the selling price to uncompetitive levels.
Now multiply this ripple across many universities, each with research programs and each with a technology transfer office committed to exclusive licensing to create “commercialization” of research results. It’s a recipe for system-wide failure. With the federal government spreading work around to many institutions, rights fragment, the Avoid Us Each and Every One signs are hung out, and only the highly motivated, well funded company can think to try to obtain all the rights necessary to have freedom to practice. Even then, universities are unwilling, generally, to commit future rights—in improvements, work arounds, and related inventions.
Universities by implementing programs of technology commercialization rather than rights management to enhance research and industry application have turned a highly productive program of federally sponsored research into an utter quagmire of fragmented rights, threats, administrative red tape, and provincial—if not merely selfish—behaviors, all masked over with language about public benefit, innovation, commercialization, jobs, and extra spending money. The actuality is *anything but*. No matter how *excellent* these programs get, their *combined effect* is to destroy each other’s operating plan. No one talks about this, but it ought to be front and center in national policy discussions. If everyone’s tech transfer model is “like Stanford’s” then it *ruins Stanford’s operating model*.
It is no wonder, then, that no university shows a diagram for an invention in the context of multiple companies, research at other universities, technology transfer at those other universities, or other inventions at the university or other university or elsewhere. If it did, it may as well title it: “this is how we have done our part to destroy American innovation.” There is no way to “improve” the system by “improving” local programs of technology transfer. If they are all operating on the same premise of money making from exclusive licensing, the system itself is all but dead. The better each gets, the worse the system overall.
The system does not need to grow, or improve, or have more Valley of Death funding, or more business folks each trying make the hot deal happen. The system needs to diversify, to build major centers, and to create commons. In Harrison White terms, what we have here is a problem with the choice of social discipline that is to prevail in technology transfer. The competition for resources and status means most technology transfer offices attempt the same thing—make money from patents—and compare themselves on the same measures—just look at the AUTM Licensing Survey to see this. It is always number of inventions, patents, licenses, and income. Never companies served, students placed, donations received, or standards set. The university technology transfer monoculture–compulsory ownership, linear model commercialization, exclusive licensing–is suppressing opportunity. Some universities’ tech transfer offices can be centers of start ups and commercial activity. Others can promote industry collaborations and standards, not making a proprietary grab for every invention that comes along. Still others can be referrers of opportunity and handle no IP themselves. Nothing in Bayh-Dole says every technology transfer office must be the same. Nothing in Bayh-Dole says that a university even has to have a technology transfer office. Imagine a university without a technology transfer office. It’s easy if you try. It is not that there should not be people involved in moving research to practice, or practice to research; rather, that these people are not necessarily best organized into an “office”. Perhaps a network, or a coalition, or a club would be a better social structure.
If there were a diversity of approaches to research IP, universities would find more ways to be interdependent, to collaborate on the administration of research. Each would not have had to formalize–and rationalize–a selfish, provincial approach in the guise of a policy that claims ownership and a lion’s share of the profits from some hoped-for big deal while touting “public benefit”. Some universities would gain their resources from industry agreements, others from donations, and others from government grants. It’s the mix that matters, not the monoculture.
Universities involved in technology transfer should seek to reduce the number of their unlicensed patents. To do this, they must be much more selective in what they choose to claim in the first place, and for what they do come to own, they must grant more general licenses that permit a wider range of activity. They may also mitigate the effect of exclusive licenses that have not resulted in products across the entire range of licensed claims by creating a much broader reservation of rights for research and applications. The one thing that clearly does not work is to polish up their marketing materials and spend more on advertising “successes” without showing the negative effects of the model that has (so it is claimed) produced them. Technology change is not a matter of creating a sales funnel based on accumulating a huge portfolio of patents just in case. That is all but declaring that the business model of choice is to troll industry for infringers after the fact rather than cultivate relationships that lead to new products—that is, make money on patents but *not* by doing technology transfer.
We come to a final point about unlicensed patents. In a given jurisdiction, the unlicensed patents represent a fundamental barrier to innovation. Thus, in the United States, every unlicensed US patent is an advantage to anyone operating outside the US, who does not need to take a license in order to practice, and does not have to worry if the university will be a decent business partner, and does not have to worry about willful infringement if its scientists and engineers read the scholarly literature, and does not have to worry that the university will license exclusively to one of its competitors or to a skunk of a company, or go trolling. The number of U.S. university unlicensed patents is the flagship metric for the stifling of research innovation by university administrators. The only success story that matters for a university is that it has reduced its unlicensed patent portfolio to none.
Accumulating patents is rather easy, though somewhat expensive. Licensing exclusively is relatively difficult. Building product under an exclusive license crammed with diligence obligations is actually very difficult. And the whole activity of aspiring to license a vast portfolio of patents exclusively or else (the threat) trolling industry for infringement means that the threat is way more likely than the aspiration. The current standard approach to university inventions all but cries out “We will hurt you”. That’s the fundamental theme, and it won’t change so long as there are vast numbers of unlicensed patents, and the dominant licensing mode remains exclusive.