A number of university licensing officers have made a big deal about exclusive patent licenses. The exclusive license, they argue, is the key advantage they have over the federal government’s typical approach to invention management. The purpose of the Bayh-Dole Act, in their way of thinking, was to enable the private creation of monopolies based on faculty-proposed research supported by the federal government. Once, I was talking with other directors of university licensing offices about this. One said, “If I was told that I could only license a new invention non-exclusively, I would refuse to accept it for management.” The second director agreed. In their minds, the value proposition for a new invention ran to the monopoly power of the patent, not the potential utility of the invention or the prospect of improvements brought by further research or the public benefit of broad access to practice the new invention.
There are other ways to manage inventions–ways aligned with a university’s public mission, with academic freedom and scholarly exchange–but these ways get suppressed when people get fixated on the value of monopoly. The problem these folks have, however, is that they mistake patent licensing for technology transfer. While a patent license might be included in the transfer of technology, a patent license is almost never sufficient. A bare naked exclusive patent license is a game based on a threat to deny access.
With an invention comes a cluster of both intellectual property rights and NIPIA–non-IP intangible assets. And it’s often the NIPIA that have the most value that’s aligned with a university’s mission and social standing. What is NIPIA? Things like talented people, a list of worthwhile problems, capability to use tools, potential improvements, connections to other labs and people, ability to explain what is going on–and ability to explain what doesn’t work so well.
“You can’t license these things,” a patent administrator might fuss. And the patent administrator would be generally correct–and stupid about the situation. Why? Because the situation that presents is one of technology transfer first, and patent licensing later. By deciding to put the patent licensing first, the patent administrator renounces the fundamental goal of university research technology transfer–so that research discoveries get used. Even though one cannot rightly license someone’s talent, one can put that talent into a position to be valuable to others. Companies may resist being forced to take a license to a patent–they would rather be free. But companies pay all the time for people to help them with their work.
It is a rare company executive who complains that people with talent don’t just show up and volunteer their time. If someone has things that can help a company, then the company is ready to pay for that help. Technology transfer is about enabling instruction long before it is about promising not to keep threatening someone with a lawsuit for infringement.
If a university leads with NIPIA rather than a patent right, the transaction that is proposed is a fundamentally different transaction, handled at a different point in a company’s decision-making structure. A patent license–especially an exclusive patent license that demands a company to make a product that comes within the scope of the patent–requires executive approval at the top of a company, and the review often starts with legal counsel questioning whether the patent will hold up and worrying about the invasive nature of an exclusive license. And university-proposed exclusive patent licenses are almost always invasive power plays. Here are just a few of the things that a university exclusive patent license typically does:
- demands commercial sale of product
- threatens to cancel the license if milestones for development are not met
- makes the company carry all the risk
- requires one quarter of pre-tax income from sales–even if the product loses money
- forces the company to submit to audits of its sales records
- prevents the company from sublicensing the rights
- prevents the company from participating in standards
- takes all settlement income if a company does take the lead in suing for infringement
- refuses to offer to license improvements or workarounds developed by the university
- makes it difficult to deal with any other licensors needed to make a product
- makes it difficult to sell off a product line or division or the company itself
All these things are done in the name of increasing the value of the patent license to the university, but it might take a company’s negotiators six months or more to try to resolve even a few of these demands. A company can spend, easily, more than $20K just trying to negotiate a deal. $20K is usually nothing to a big company, but for a startup, that money might be a major percentage of the company’s seed funding. So we can add the following as well:
- delays rapid adoption of new technology
- costs a substantial amount just to negotiate
- eats up valuable company budget
If the company needing a license is a startup, there’s more bad news. The university will want to see a business plan, and then will in all likelihood try to incorporate the business plan into the license contract–turning a planning document into an enforceable promise. Any change in the business plan then will require the approval of university business officers. I’ll choke down the term “bureaucrat” or worse. The university also will want to own a piece of the startup–5% royalty and 5% equity is a common mantra (make the two add to 10 is a common, unsophisticated rule of thumb). So the company has to shed some equity, just to get a right to practice the invention.
If a company taking a license does want access to further developments made at the university, then there’s yet more. First, the company will have to “sponsor” the research–paying for the research it wants. Meanwhile, the university may go ahead and also get federal funding (or even funding from another company, in another lab) to work on the same stuff. Lots of luck expecting that the developments the company would want will happen to come about with the company’s bit of funding. But it’s worse than this. Even for research the company funds, most universities will not give the company anything better than a first right to negotiate for a license to new inventions–even as the company is already negotiating for a right to the initial invention! Having never understood the idea of royalty stacking–where each patent owner demands its own 5% of income, making sale of the product utterly unprofitable–the university attempts to stack its own patents. If the first invention is worth 5%, then the next invention should be worth another 5%. Craziness.
Only desperate, determined, or flaky companies will get into such negotiations. The rest keep their distance. It’s like being invited over for dinner only to find out that to get in the door, your car will be impounded, you will be strip searched, and the dinner actually is take out from yesterday, and there’s a cover charge. This is something you don’t do twice. And that’s a good reason why university licensing offices do not make anything of repeat business–there hardly isn’t any. Go ahead, try to find a company that liked its patent licensing experience so much that it came back again and again for more. There is simply no way to make the process of impound, strip search, cover charge, and yesterday’s meal more efficient, more attractive, or more productive. Anyone trying is way, way off track.
In one case, I was at a campus licensing office working with a company that had previously obtained a license from another campus in the university system. “How was that license?” I asked. “Great–it worked out well.” “Then,” I replied, “Let’s do that same license again.” We were agreed in ten minutes. I obtained the license from the other campus, made revisions for the new material, and was ready to sign when the central licensing office got ahold of the deal. “You cannot offer those same terms to that company,” I was told. “Um, why not?” “Because we granted the other campus an exception to policy to do that deal, and if we gave you that same exception to policy, then we would appear to be changing our policy, and we cannot allow that.” “But why *not* change your policy if it is so bad?” Silence. The result was much disruption, fuss, and overall took an easy pathway to agreement and made it stink. But that’s often how university administrators are about this stuff. Their consistency across the years and preservation of authority is more important than getting to a deal in the present. None of this, by the way, is ever addressed in a technology transfer office’s annual report.
It turns out that exclusive patent licenses are just about the least attractive, most difficult form of patent license one can imagine. It’s even worse when the exclusive license is offered by a conservative, risk-worrying, non-profit whose only purpose happens to be profit-seeking. If a university weren’t profit-seeking, it would simply assign ownership of patents and be done with it. (And, yes, despite what you may have heard, the standard patent rights authorized by Bayh-Dole does permit assignment of subject inventions to for-profits, though for nonprofits, there may be an additional step.) Instead, the university licensing office starts with the premise that it will grant an exclusive license. This means, immediately, saying “no” to every company or organization that would want to use the invention but not commit to making a commercial product of it.
The reality is, most university-hosted inventions are bits of things–a new method among thirty new methods for messing around with a carbon nanotube; a new compound among a hundred new compounds that has the potential for making a rat puke when it eats radishes; a new sensor improvement among fifty sensor improvements that uses a new arrangement of quantum dots; a new method among a hundred new methods of removing oil tar from a beach. Those bits of things often have an immediate use, but not only are a long way from being a commercial product. The usual language university folks use is “early stage technology”–as if at some later stage, product-status will come about. It’s a load of bullshit, really–and I use bullshit in its technical meaning, as a statement made without regard for the truth. The use of “early stage technology” is code for “we will hold this until we find an exclusive licensee willing to spend whatever it takes to turn this invention into a commercial product.”
What really happens is this: by claiming patent rights on a little bit of university-hosted invention, a university chooses to demand that the transfer of the technology also involve a formal license to the patent. By fixating first on trying to find an exclusive licensee, the university greatly reduces the possible early licensees. Many companies may want to use the new invention; only a very few–maybe none–want to try to make a product of it. And those companies that do show up wanting to make a product of the invention are often the ones created by the researchers themselves under the university-promoted delusion that this is what entrepreneurs do–figure something out, and then try to sell it off to investors. Yes, sometimes–but mostly what entrepreneurs do is recognize something that people need and then figure out a way to deliver that something in a competitive way. The university version of entrepreneurship is almost 180 degrees from reality. It’s like a mirror reflection, that uses all the same wording, but actually is almost useless.
Consider the case for use, however. There are multiple forms of use that show up with research inventions.
1. Another organization may want to use an invention in order to evaluate published claims about the invention.
But if the university has applied for a patent, then the organization needs a license or later, when the patent issues, the organization could be set up for a claim of infringement. And if the organization can’t get a non-exclusive license, then what’s the point of even evaluating the publication? A university will happily offer a company an “evaluation” license, but typically the university will insist that the evaluation be for the purposes of entering into a commercialization agreement and for no other purpose. So that rather shoots down evaluating a published claim pertaining to the invention without a lot of wrangling. Perhaps you recognize how a university fixation on offering an exclusive and commercialization license disrupts–and transforms–what publication is all about. Publication is no longer an offer to test something–it is an advertisement that something has been done and is reserved for a company that will create a commercial product. That typically means, a speculative investor, a happily misguided researcher, or a company at home trolling the industry.
2. Another organization may want to use an invention in order to study the underlying phenomenon.
Beyond reviewing published claims, a company may want to use the invention to understand how things work–not just the invention, but the strategy described by the method, or the chemistry, or the circuits, or the biological process. The invention, then, is a marker of a line of investigation, and the company may want to get on with this line of investigation. This approach is just what a university patent owner will try to prevent. By using an invention in this way, a company could design around the patent’s claims, could get to new juicy findings before the university’s own lab, and could then file its own patent claims, blocking “open university research” (which, of course, is not open at all, since there’s a pending patent and an assumption of an exclusive license–more of that administrative bullshit). So the university’s patenting and licensing strategy disrupts both collaborative and competitive research using published accounts of the invention.
3. Another organization may want to use the invention as a research tool.
Many university-hosted inventions are actually not realizations of something new in the world, but rather are realizations of something new that can improve a tool for looking at the world. In coarse terms, a better telescope, not the planet Clarion. New tools are of course new things in the world, but their purpose is to assist in looking for the really new things in the world. “Hey, I have a new way of detecting CO2 in the atmosphere” is not the same as “Hey, I have discovered how to use CO2 to make O2 that’s better than photosynthesis.” In the former, we grow more sophisticated in our search for something–a bigger ship to sail on a voyage of discovery, or a better forward head for it, at least. In the latter, we actually find something–what we were looking for, or even better, something we didn’t even know existed.
Research tools have been lucrative sources of income in university licensing, and almost always they have been licensed non-exclusively. Consider the Cohen-Boyer gene-splicing patents–non-exclusive; the Axel cotransformation patents–non-exclusive. The Hall expression of polypetides in yeast–non-exclusive. SPICE, that ace circuit simulation program–non-exclusive. Go down the list–Rosetta protein modeling software, PCR, Cre-Lox (the last two, both company-based research tools)–and most research tools end up non-exclusively licensed. And there are good reasons for that. One–non-exclusive licenses serve the entire industry. That means, one has the largest possible market. It also means, one does not have to “pick a winner” or “play favorites.”
Picking a “winner” is itself a tragedy for most university exclusive licenses–it means saying “no” to everyone else. I remember one university faculty member who had the misfortune of having the key finding of years of his research taken by the university, patented, and licensed exclusively (with fanfare) to a company that then sat on the invention and did nothing with it but dither. As the faculty research commented later (I paraphrase)–“They essentially destroyed my future research. I have companies showing up every day wanting to work with me on my discovery, and the university tells me I have to decline, because we will get sued by our exclusive licensee if it appears we are trying to work around the license.” Again, licensing a patent exclusively to a moribund company or to a company that is moribund with regard to developing the underlying invention is a fate worse that death for a new research finding. But you won’t find in any technology transfer office’s annual report a listing of all exclusively licensed but moribund inventions. Ask them–they won’t even know, since their own monitoring of most exclusively licensed inventions is also moribund.
4. Another organization may want to use the invention in its own operations.
A research finding does not have to pass through a commercial product stage to be used.
“Commercialization” is just one pathway to practical application. And within “commercialization” there are also multiple pathways. The pathway that universities fixated on exclusive licenses choose is that of granting an exclusive right in the patent on the condition that the licensee works to create a commercial product out of some part of the claim language in the patent. There could be five claims in the patent (often more), but the company only has to create one commercial product under one of those five claims to be successful. The rest of the claims become, essentially, moribund, undeveloped.
Another pathway of commercialization, however, starts with internal use, which validates a research finding and demonstrates its potential. From there a company may want to provide that same use to its collaborating companies–suppliers, or joint-ventures, or an industry standards consortium. The pathway for commercial use–a form of “commercialization”–may never pass through a product stage (and go straight to a standard) or may become products that a number of companies provide, each with features and price points and availability matched to whatever emerging markets show up. In this approach, early, relatively unencumbered, non-exclusive access is critical. A university, in fixating on the desire for an exclusive license, essentially shuts down other commercialization pathways, destroying many opportunities in the process.
Consider another consequence of denying companies the opportunity to use, this one from laboratory medicine. The University of Utah developed some lab medicine assays to test for diseases, and licensed them exclusively (or in one case, ended up granting two licenses) to companies that created commercial versions of the tests. The problem for doctors working in lab medicine is that for some of these assays, the commercial version of the test is not as accurate as other versions of the test that most any lab medicine department can run on its own. But those other versions–better medicine practice–cannot be conducted because doing so would also infringe the claims of the licensed patents–moribund claims, perhaps, or simply versions of the assays that the company decided not to offer. So here you have a situation in which better practice is excluded in favor of paying practice. And behind it all is a university that has licensed its patent rights exclusively, but more broadly, than is good for patient care. It’s one thing to license a patent exclusively so that a company may make a product. It’s another to prevent any other use, so that the commercial product version does not have to compete with local implementations. The patent strategy at the university is this: to suppress local implementations in order to increase the payoff from the commercial version.
When a university patent administrator pipes up with something along the lines of “without an exclusive patent license, this company would never have had the incentive to create a commercial product” you might think again about how the incentive to create a commercial product can be gamed so that research findings that never would require a commercial version to be well-practiced are turned into a profit-seeking venture that depends on denying just that local, non-commercial practice. It’s not so complicated if you get used to seeing something real rather than accepting bullshit from patent administrators. Even Neils Reimers–a person not known to bullshit anyone–made it clear that patents on Cohen-Boyer weren’t necessary for the commercial use of gene-splicing technology. What’s fascinating in the story of Cohen-Boyer is that Reimers when to the emerging biotech industry and asked whether Stanford should file patents, and the industry provided the guidance–yes, and license non-exclusively–and in turn Stanford kept the patent prosecution open, so everyone could see the application as it progressed. A great approach, one that no university I know still practices–but all should.
Finally, one more instance, this one a scenario. Imagine that a university has a patent on a method for removing oil tar balls from a beach, and grants an exclusive license to a company that’s working on a commercial version. Now, somewhere, there’s an oil spill and tar balls start washing up on a beach. A city, to save its tourist industry, wants to use the method to remove the tar balls. But no–they can’t because they have to buy product (if it even exists) from the exclusively licensed company, and use that. Again, the local use is suppressed so the university will get a royalty from each commercial sale–even if the commercial product isn’t available. This is what the old-style federal agency march-in rules were for in the Institutional Patent Agreements that preceded Bayh-Dole and did a much better job of dealing with exclusive licenses that went too long, failed in their purpose of encouraging rapid development of commercial product, and/or denied the public a benefit on reasonable terms that the public should expect. An agency could just step in and make good things happen. Bayh-Dole was written to make march-in so difficult no agency could possibly use it. Another crappy little detail about Bayh-Dole that makes it lousy eating.
These four forms of local use of a research finding can be summarized as
- validation of
- research on
- research with
- local use
All of these uses are suppressed when a university holds a research invention in an effort to find an exclusive licensee willing to commit to developing a commercial product. That effort reduces by orders of magnitude the number of organizations that may have an interest in the research finding, drives up the administrative overhead of securing permission, and tends to attract the business low-life–the speculators, trolls, and gullibles who are willing to try to turn the invention into a product, so long as all other uses are suppressed. Where there is no local use possible without hundreds of thousands or millions of dollars of development, then an exclusive license may well be the way to go. But if local uses are possible–and if a university lab can make the tool, then so can most anyone else–then exclusive patent licensing gets its value from suppressing practical application, not from encouraging it.
There is no mandate in Bayh-Dole that the patent system has to be used to promote monopoly speculation, fragmentation of research across multiple universities, or the suppression of local practice. The emphasis throughout Bayh-Dole is on use, on practical application–not on commercial products, and not with a requirement that every user of a subject invention must pay up or be sued. It’s only a bit ironic that use and sue are a simple anagram pair. It’s just that university exclusive patent licensing sets up the conditions for “sue” much better than it does the conditions for “use.”
There are a few instances in which exclusive licensing of university patent rights makes sense. And there are many instances in which such licensing makes no sense, creates no value, and obstructs rather than advances the public objectives for university research. The mandate to use of the patent system to promote the use of research findings made at universities is not a mandate to suppress local use, force people to use commercial products less effective than what they could do themselves, exclude many companies from collaboration and competitive development in favor of finding one that’s willing to pay a pile to exclude everyone else.
The parasite of exclusive patent licensing has found a compatible host and is getting fat on the juices of discovery. But that parasite is at the same time disfiguring university reputations and destroying the underlying social commitments that have made university research so valuable. Fixing this situation is not a matter of finding a more pleasant way to shake down industry. Fixing this situation requires a rejection of the pretty vocabulary used by a large number of university patent licensing professionals and their advocates–and a rejection of the destructive, unproductive way of thinking that backs up that vocabulary and gets expressed in exclusive license agreements and suppressed in accounts of technology transfer “success.” At this point, no vocabulary is better than the strange, dark castle that gets lit up as the great hope for economic development based on university-hosted research. No policy on patents is better than what most universities have been duped into adopting.
It’s not a matter of starting over–but it is a matter of starting with those things that matter, seeing the situation clearly, and building programs responsive to needs.