I have been thinking about how university technology transfer is depicted, versus how it actually happens. The depictions are something of a prophetic hope–inventions reported to the university’s licensing office will be evaluated for “commercial potential” and those that look promising will be acquired, patented, and marketed to companies to develop into beneficial products. A win for the public, for industry, and for the university and its inventors (in the form of royalties and prestige). It sounds wonderful, but practice does not follow the prophecy, and there’s a big problem with that–repeating the wonderful depiction in the absence of practice is disingenuous, even false, even fraud. It’s not even just putting a positive spin on things. It’s just not what happens.
Universities expanded their patent policies from voluntary reporting and assignment to compulsory assignment of inventions. They expanded the scope of what is to be reported from patentable inventions made in the course of their official duties to anything a licensing officer describes as inventive even if not patentable made within the field of an inventor’s professional expertise. That expansion means university licensing offices claim an awful lot of stuff that they cannot possibly deal with. This, in turn, means that the most frequent outcome of a university taking ownership of an “invention” is that nothing, nothing ever happens. If this were a Talking Heads song, then it would be “Heaven.” But that heaven, described, is a sort of undesirable hell.
Consider the possible outcomes. Here’s a sketch of what typically happens. No one keeps records to document all this, but from twenty years in the technology transfer world, here’s something to start with, going from most frequent to less frequent, if not rare.
1. University takes ownership of the “invention.” The invention gets logged into the university’s technology database. Nothing happens. In about 50% of the cases, the university does not file a patent application. If a university were to describe its IP practice according to what happens most often, it would be this: “We take ownership of your inventions, and that’s all. Suck on that.”
The university does not return the invention to the inventors/developers/authors/ collectors/cultivators. There’s a happy catch-22 to it. The reasoning goes, if the invention is not valuable, then it is a waste of everyone’s time to go through the extra paperwork of returning the invention to the inventors. If the inventors, however, want the invention to be returned, then the invention must be valuable and therefore the university cannot return it to them–with the implication that the inventors, Fenn-style, have kept from the university important information about how the invention might be developed, licensed, and money paid to the university. The university will hold the invention until the inventors come clean with how to make money from it, or until the inventors no longer want the invention because it is no longer valuable to anyone.
2. University files patent application. Then nothing happens. Perhaps it’s a provisional patent application, maybe a full US utility application, sometimes a PCT application naming the US as one of the countries to be pursued with a national phase filing. Then nothing happens.
One or more patents may issue, or the application rejected, such as for prior art. But no matter, the invention is never licensed even if there is a patent. The patent generally serves to keep people away from the invention who otherwise might use it. It is easier to design around most patents than to negotiate access to them, especially on a non-exclusive basis. University patent licenses are generally company-invasive, paranoid, poisoned (against cross-licensing, use in standards, sublicensing), and expensive relative to prospects for use and advantage. So, for about another third of university inventions, nothing ever happens, other than spending $10K to maybe get a patent, maybe not.
3. University files a patent application on an invention made in industry sponsored research. The research contract grants the sponsor a royalty-free non-exclusive license with a right to negotiate an exclusive license, provided the sponsor pays the patenting costs.
This is a pretty frequent deal, with a few variations all getting at the same thing. The sponsor could be a big tech or biotech firm, a member of a research consortium such as the NSF Cooperative Research Centers program, or an SBIR/STTR company thinking that $100K of federal money is just the thing to prove the concept of a new product. It really doesn’t matter. Other than that a patent application gets filed, and the company sponsor is on the hook for paying the university’s patenting costs, nothing else happens. The university counts a patent (or multiple patents), and a license (often not even formalized beyond the sentence in the funding agreement), and that’s the end of it.
4. University takes ownership of invention and licenses it back to the inventors as founders of a startup. This is not really technology transfer as the inventors know more about their invention than the university does. By taking away the invention and then handing it back with strings, the university does not transfer the technology but rather intervenes to exact a rent and to complicate and delay startup efforts.
The startup may get seed funding (from the university, even, or from the state, or from an angel fund) or may even get venture funding and then get acquired by a big company or sublicense exclusively to a big company (a “strategic partnership”). And nothing happens. Yes, the university gets reimbursed its patenting costs (usually) and there may be a payment (even a relatively big payment) with a round of financing or exclusive sublicense or acquisition, but the licensed invention does not result in a product. Maybe some other product is developed, but the likelihood of anything developed is near zero. Consider, the University of Washington claimed 18 startups in 2014 (really, only 1 qualified as a startup in that fiscal year), while receiving over 200 invention disclosures per year. But of those 18 startups, one is successful, run by the inventor after UW failed to do anything with the invention the inventor was required to assign to UW, one was bought out without any products by a Japanese company, and a third, which did have a run of providing on line services failed when the company that bought it failed. The rest putzed around, maybe got some seed funding or federal grants, and then withered or zombied. At the University of Utah, which made a big deal about its startups, when the state legislature investigated what happened to the $100M that it had allocated to chase such startups, they found a culture of “lies and deception.” And a bunch of paper startups with no operations and business addresses at the university’s licensing office. Nothing, plus public fraud.
When a university makes money licensing exclusively to a smaller company, the income often comes in the form of the value of company stock when there’s a round of funding or the company is acquired. Taking an equity interest in a company rather than relying on a running royalty on sales of licensed product means that the university gets paid if the company is successful in puffing up its future value rather than when and only when the company produces licensed product. You can see the switcheroo, then, in how universities report licensing income but fail to distinguish between all the licensing fees, patenting cost repayments, and realized equity on the one hand, and running royalties on licensed products on the other. A university can make a lot of money from a patent deal without ever seeing things through to product sales of licensed product. The patent deal helps a startup get investment funding, and the investment funding helps the company sell out to another company or develop a product that’s not subject to the patent license. There’s technology transfer in a stubby dysfunctional way, since the patent gets used but the invention mostly doesn’t.
5. University licenses invention exclusively to a bigger company. Then, most often, nothing happens. Or, the company uses the licensed patent to beat on its competition (as Ciba-Geigy did with its UCLA patent, to delay a rival getting to market with its nicotine patch first–Ciba-Geigy didn’t need the UCLA patent for its own nicotine patch product–just to pound its competition. Later, a court found most of the UCLA patent’s claims invalid, but it was a good run for UCLA nevertheless, as far as the patent racket goes. The University of Washington tried this same sort of deal in a dispute involving laser scalpel companies, giving one company a patent to countersue the other with.
Every so often–once a decade or longer–a patent licensed to a company ends up paying royalties, and somethings big money. So, UCLA licenses a series of compounds with anti-cancer effects to a venture firm (Medivation) prospecting for drug candidates (and misleads the company about a second series that it keeps secret–lawsuit later on that). The company outsources most development and testing, creates a successful product (Xtandi), and then sells out to Pfizer. Big money for UCLA, especially since UCLA sells its royalty rights to Royalty Pharma and so gets a lump sum up front (there is strangeness here, but for another time).
Where there’s big money in a university patent deal, usually that comes with a second hop–a startup is acquired or (really rare, as with Google, goes public) before a product hits it big. (Here is a recent example, at Rice University, involving 3d bioprinting).
6. University gets patent, fails to license, and goes patent trolling to force licensing and settlements. Here’s a case where it is clear that holding a patent has nothing to do with development or use of an invention–companies go off and implement the technology on their own (and may not have a clue about what a university has done or claims). Certainly there was no need for patent exclusivity, other than to collect rents from companies for what they have been able to do on their own, and without patents, and without technology transfer. But universities then come after industry for royalties anyway. That’s what Stanford tried to do to Roche, resulting in Stanford v Roche, which Stanford lost, and rightfully so. It’s what the Washington Research Foundation did with its bluetooth patent. What Caltech is doing with its wi-fi patents.
Again, not technology transfer, clearly patents aren’t needed to attract investment, but the university sues for infringement anyway. This is not so much a case of nothing happening–even though in terms of technology transfer, nothing indeed has happened–but rather the university makes something bad happen, something contrary to the university’s public mission and the depiction of university-based technology.
7. University licenses non-exclusively, creating an ad hoc standard. This is rare, but it happens. The big example is Stanford’s gene-splicing patent licensing program in the early 1980s. Research tools can go this route. But generally, as Stanford’s Niels Reimers argued, patents aren’t needed, other than to extract payments from companies–“it’s just a tax.” There’s an argument that patents in such cases can be used to shape standards and draw in other contributed technology, but I don’t know of anyone in a university transfer program that makes this case (me, not being in any such program being a different problem). Despite Stanford making over $200M in licensing income from its gene-splicing patent licensing program, universities generally do not try this approach with patents (more so with software, but even there the big classic examples–SPICE, X-Windows–have run outside the university’s licensing program).
8. Invention released back to the inventors. Ha. Hardly ever happens. And when it happens, the release is with conditions–university retains a license, university gets paid its patenting costs, university might get share of any income, inventors cannot use the university for development, university owns any improvements (and then this cycle might repeat if the improvements aren’t valuable and the inventors don’t appear to want them, but for some reason the university is willing to waive its claims).
9. Invention is offered to the federal government agency sponsor. This does happen. The university takes title to the invention and then gets cold feet about filing a patent application (runs out of money, gets in a fuss with the inventors (problem inventors!). Typically this path involves a misinterpretation of the Bayh-Dole Act, but it’s okay you see because everyone does it. Anyway, the idea is that if a university takes title to an invention arising in federally supported work, then if the university doesn’t file a patent application, it has to offer the invention rights to the federal agency that supported the work. An inventor might be permitted to petition to receive title, but then it is all weird because the university could just license back to the inventor and avoid the whole offer rights to the government thing. But freedom from administrative confusion is not a pre-condition to running a university licensing office. Nothing ever happens when rights go to the government. “It’s dead, Jim.”
10. Invention licensed exclusively to company, which develops a product under license, and sells it, paying royalties.
This does happen. Maybe once in a few hundred licenses, once in a 1,000 invention disclosures. At Stanford, 53 of 6,400 inventions over a 36-year period paid $1M or more cumulatively. Keep in mind that $1M over the 17 or 20 year life of a patent is about $50,000 per year–more along the lines of repaying the costs of foreign patent filings, paying licensing maintenance fees and milestone development fees, than anything to do with a running royalty on sales of actual licensed product. Even with this low threshold, Stanford’s licensing rate through to product is less than 1% of its invention portfolio.
There are plenty of ways to make money with university IP, but there are a lot fewer ways to transfer technology for public benefit. Universities depict their technology transfer programs using a less-than-1% pathway as if that pathway is the only pathway, that it happens all the time, and that their licensing programs are wildly successful, with broad public benefit, driving–yes, even driving–the innovation economy. In practice, universities end up serving primarily as rent-seekers and forum creators for speculation on future value of patents. But no university is about to write policy and public-facing descriptions that describe what the universities are doing in practice. It is difficult to see how to discuss university or federal IP policy without excluding the university lies and deception and unwillingness to own what they do. And to call this stuff out, well, that’s seen as a lack of civility, an offense against the honor of public servants. Bah. Dolts. Mostly, nothing happens, at some great expense and with many lost opportunities.