Much of the current, dominant narrative about patents at universities depends on looking isolating single inventions at a single institution with a single profile for use. “Inventions,” so this narrative go, will not be used or developed unless for each a patent position is established so the university grant an exclusive license to a company attracted to the incentive of that exclusionary position.” Without such an “incentive,” no company will spend the tens of millions of dollars necessary (it is claimed) to make the invention benefit the public.
Perhaps that sounds right to you, sounds normal, like there’s something to this narrative about patents. If so, but why? Is it just that’s all you’ve heard about patents at universities?
Let’s add in necessary qualifications to this narrative of the single patentable invention that should be “protected” so an exclusive licensee has an incentive that it otherwise would not have to spend money to develop the invention for public benefit.
First qualification: this patenting for exclusive licensing must be the default. Otherwise, it is darned hard to offer exclusive rights later, once others hold non-exclusive rights. Sure, it can be done–most anyone who knows licensing can figure out how–but then we’d have a different narrative and practice. With exclusive licensing as the default, however, immediate access by companies interested in a new invention is blocked at a critical time in the development of the technology. Well, technically the block comes about when a patent issues, but patent pending is enough to put off what later will be infringing practice. And if a university is rigging for exclusive dealing, then the inventing team will have trouble encouraging use of their invention outside of official university licensing. An inventive team, actually, may help others design around the university’s patent application. I’ve seen this happen multiple times, almost to the point of routine practice.
The default position, then, creates a block to early access to the invention, pushes companies to avoid the invention (because they don’t want or need an exclusive license and so won’t be the ones getting it), and creates a situation in which members of the inventing team may help others design around the patent application.
Second qualification: the scheme doesn’t work. This is a big qualification. There’s nothing to show that by withholding permission to practice an invention in the hope of an exclusive license that a university will attract exclusive licensees, or that those that are attracted turn out to be effective at creating commercial product based on the university’s patented invention. University officials are so thick about this that they cannot show that their patents, held for exclusive license, were the necessary incentive to “draw forth private initiative.” If the private initiative would have been there anyway, then the patent presents a burden on the relationship, not an incentive, a parasite, as it were, a tax.
Now, there are other things that could be in play with regard to patents–“rewards” for inventors, say–but rewards for inventors can come about many ways, including through non-exclusive, royalty-free licensing. The dominant narrative, however, uses rewards for inventors as an incentive for inventors to participate in the dominant narrative: “if you don’t accept our default exclusive position, then we will deny you any reward for your inventive work and may even set out to ruin your reputation and career, if only as a lesson to others.”
But the scheme doesn’t work in a bigger way. While there are lots of patents (universities have got more than 120,000 US utility patents since 1980), and lots of reported licenses (but licenses not connected to specific inventions), there’s hardly any output in the form of commercial products, and there’s even less output of things that the exclusive narrative suppresses, such as standards or rapidly developing technology platforms with contributions from multiple sources at multiple organizations. Organizations such as AUTM don’t even attempt to track outcomes tied to university patenting. Given that AUTM (originally, the Society of University Patent Administrators) was set up to lobby the federal government to give universities ownership of inventions made in federally funded work, it would be outside AUTM’s purpose to ask whether university ownership of inventions–combined with default exclusive licensing–was all that effective. Well, it’s not effective. A tiny percentage of inventions make it through university management, an exclusive deal, to become commercial product. The rest languish–blocked, avoided, designed around, made obsolete. Think about that in the context of qualifications on the default narrative:
“In order to get 1 invention in 100 made into a commercial product by our exclusive licensing approach, we must accept that 99, or even 999, will fail and will not benefit.” Now tell the story from the perspective of the majority of university-held inventions–university ownership is where inventions go to die more quickly than they would otherwise. A university has to isolate many inventions so that a few make it through. Another way: only the very strongest of inventions can survive a university’s licensing program so that a company is willing to pay to take the university’s license and hold off all others from practicing the invention. Companies may be willing to pay to exclude others–especially companies that do not have a dominant market position–but it’s another thing entirely whether those companies then produce new products based on the licensed rights.
In my experience–mostly in engineering and bioinformatics–maybe a third of the time, where there is new product, the companies taking the university licenses design around the university’s patented stuff and sell product that isn’t covered by the claims of the licensed patent(s). There’s product–a success story of a sort–but it’s not product based on the university’s licensed patent. That invention–the one the university aimed to promote–is not developed and used. There’s a patent, a license, even royalty payments, and perhaps a startup, but the invention itself gets buried, burned up and lost in the transaction. AUTM is not about to report anything like this, and if a university has gotten paid and expect to keep getting paid, it’s not about to cancel the license for failure to develop product. That’s the role, then, for milestone payments and university equity interest–forms of payment that allow the university to profit even when there’s no product to benefit the public based on the exclusively licensed invention.
So the scheme doesn’t work. Or, it “works” once a decade or so, if a university is lucky, to take credit for a royalty stream for one “big hit” licensing deal. Everything else, then, becomes chaff, collateral damage, stuff that didn’t make it out of the “invention funnel.” Here’s an example, with pretty colors:
In this diagram you can see the default narrative–disclosure of inventions to exclusive license agreement–see, there’s only one license agreement. But now we see there’s also a funnel. Most inventions don’t make it. And even then, only the orange oval invention was there at the start–the blue oval and green oval just show up all magically at the end. Oh, I know, it’s just a heuristic. But look at it–this heuristic normalizes the idea that the university’s approach wastes most of the inventions it touches to serve up a tiny few. These are the inventions that survive the process. The rest are damaged by the process, have no benefit and produce no benefit. The scheme doesn’t work. Each invention under management does not become a product. Most inventions under management do not become products. Nearly all inventions under management do not become products–and those few that do, do so mostly without any indication that an exclusive deal was *necessary* for those inventions to succeed. Any exclusive deal was *necessary* because university officials insisted on exclusivity and sought out then only those companies willing to consider an exclusive deal (and pay for it).
And this is still just looking at single inventions, as if they are colored ovals and circles in a university administrator’s funnel vision. Now add in multiple inventions covering related subject matter arising at multiple universities and other organizations (and independent of organizations, too–what a thought!). If universities insist on default exclusive deals, these overlapping and related inventions are necessarily isolated, and the cumulative technology that they could otherwise form is instead fragmented and disabled. No current technology commons, no libraries of methods, no cumulative technology, no interoperability, no standards. Even if an isolated, sort of mediocre invention that no one really wants might gain some attention if patented, it doesn’t follow that this same practice will be even better for non-mediocre inventions or for inventions that matter only if taken together to enable all sorts of practice without the bother and delay and uncertainty of having to negotiate licenses–and exclusive licenses at that.
Will multiple organizations and multiple inventions in related areas, the dominant exclusive patent license narrative falls completely apart. It’s crap. Stinky, smelly crap.
By contrast, the Research Corporation approach that the university administrators displaced depended on invention selectivity. Instead of taking everything and making money with just a few, Research Corporation examined inventions in the context of what it was set up to do and accepted only 10% to 15% for management, and then was able to license about one third of those–so maybe 5% of the inventions submitted to it, and maybe only 1% or 2% of the inventions arising at universities.
The dominant university narrative goes just the opposite way: take all inventions, run them through an administrative funnel, and declare success for the “portfolio” if there are a few winners per decade. The scheme doesn’t work, even when a university makes money from its licensing efforts. And the scheme destroys what used to go on with the 90% of university inventions that did not seek out institutional management. No one bothers to track the non-patented stuff, or the patented and released stuff (as federal agencies did), or the patented and contributed to a standard stuff, or the patented and cross-licensed to enable reciprocal access without getting patent attorneys barking at each other for no good reason. All these practices are disrupted when a university defaults to an expectation of exclusive patent deals. The scheme sucks.
When a university defaults to an exclusionary position, it defects on the collaboration and good faith expectations of others. The defection comes to roost anywhere there’s other inventive work necessary to create a technology base/platform/library/commons. If everyone else is contributing to a common platform, and one contributor then defects and excludes all others, the platform is blocked in that inventive development. This situation is much like that of the Prisoner’s Dilemma in which the greatest payout is to defect on true collaborators. If you know they will stay true, then the best deal is to throw them under the bus. That’s the effect of a university normalizing the idea of a technology “funnel” of inventions running toward a handful of exclusive deals.
Anyone who works within the scope of a patented invention then cannot practice that work without permission of the patent holder. Everyone who does such work then ends up working, as it were, for the patent holder, who can suppress that work (or tax it–but then what’s this nonsense about exclusivity?) by asserting its patent right. If the university patent holder is determined to offer only an exclusive license, then the collaboration to develop the platform is pretty much hosed. Multiplying this same reasoning to all the institutions involved, everyone starts defecting on the collaboration with their own patents. The scheme invites the patent fragmentation of emerging new work, led by institutions, each in it for its own benefit. You will not find diagrams with pretty colors showing how a university’s patent practice might be used to help other universities get their inventions into commercial use. No other university’s invention could possibly be that important.
In normal industry settings, once everyone is defecting, folks have to set up cross-licensing or standards or a patent pool in order to prevent an all-out infringement war in which no one can build anything because someone else holds some bit of patent in what’s needed. But with universities, nothing is normal like that. Each university demands an exclusive license to commercialize its bitlet of patent in what once was or could have been a platform of technologies, and so no university is ready to offer non-exclusive cross-licensing to anyone actually expecting to use the platform under development. It doesn’t really matter if a university reserves rights for non-profit use of its patent bitlet–what’s the point if those other universities cannot license both their bitlet and the bitlets of others necessary for anyone in the commercial world to actually use the platform? Unless everyone licenses non-exclusively, the whole thing fragments into each university hawking its bitlet for an exclusive deal, as if its turdly invention will become a diamond under the magical pressure of a patent used to exclude all others.
The reality is that none of them succeed, because they are all doing the same thing against the progress of common effort, and they all defect too soon and too often, and disrupt the formation of the really important non-IP intangible assets–the collaboration, the good will, the formation of a platform, the testing and validation, the variations and freedom to vary, the ease of uptake and adoption, the expectation of availability for commercial use when the time is right.
Defection on collaboration with a single advance that both blocks competition and focuses resources can be remarkably effective, and all the more so if one works from a major research university such as Stanford, MIT, and the like. At these institutions, because they tend to win the really big grants, are in the best position to defect on all the others, and do so. Lesser institutions will never be these leaders, and so for lesser institutions, their gestures to defect on collaborations using patents are cargo cult behavior–their patents are coconut headphones, not the real things, and the whole effort is mostly for pretend, though with real world damage no one wants to talk about.
When something is new and uninteresting to companies in the context of the already common, the patent exclusion idea somewhat works. Even then, the overlay of public interest–and even institutional self-interest–argues for non-exclusive licensing and even there to regulate quality and proper marketing claims, such as the claim of implementing a standard. This defection strategy does not work if there is not first a breakaway cooperation that gets far enough ahead to define a product class that can alter broad practice. If everyone defects at the outset, no one gets ahead and there’s gridlock, like the early airplane industry in the U.S. So many bits patented no one could build an airplane using the latest technology, while in Europe they could.
The idea that a new thing won’t be used unless we all agree to allow one company to exclude the rest of us, spend whatever money it must to get to any saleable product and then sell that product for whatever the market will bear is almost entirely nonsense. Nonsense because the argument matters only for mediocre inventions for which there is no pressing public need (even if there’s perhaps desire for), and nonsense because if there is a pressing need, then there is also a basis for the public to contract for whatever development is needed and then release for mass production when the invention has reached “the point of practical application.” As was done by the Dept of Agriculture to develop new fertilizers, for instance. 100% success rate in introducing to commercial production.
The Research Corporation approach as it diffused into universities becomes all the more exclusionary as it turns attention from engineering to pharma. The Wisconsin Alumni Research Foundation made the case for provincialism so each institution affiliates with its own licensing operation and so creates an institutional conflict of interest in overseeing the deployment of inventions. Instead of seeing inventions move to where they might be best practiced or developed, an institution wants to see income from a patent position, and only with that stipulation should any invention move anywhere at all. The shift then of the patenting interest to biomedical–as early as warfarin through to cisplatin–then sets the stage for the exclusive pharma patent model to suppress all other licensing approaches not only for pharma but for biomedical generally, and–because administrators do not have the capacity for using judgment or allowing choice–for all inventions in all areas of practice regardless of the role of patents in those areas of practice.
The pivot to non-selective invention ownership tied to default exclusive licensing seems to have started in the early 1960s with the pharma boycott of federally supported work at universities where pharma-paid faculty inventors looked to take federal money to subsidize their relationships with their chosen pharma companies, essentially working outpatient for pharma using university and federal resources. All for the good, of course, as a way to create new medicines, public-private style, but the monstrous version of that in which the public pays the subsidy to offset the risk and cost to the profit-takers rather than the virtuous version in which the profit-making side sets aside its expectation of profit in favor of helping with the public effort, asking only for break even with overhead to reinvest in the public effort. Like in times of war, or as the government declared it, a war on disease. Pharma didn’t buy in to the public version and adopted the monstrous version.
The dominant exclusionary patent licensing approach then baked in this monstrous version as if it were the public good everyone wanted. What’s strange is that Cohen-Boyer, Axel, and Hall patents were all old-style non-exclusive biomedical style licensing programs. Something changed, though, to make the patent exclusionary model drive out all the rest. That appears to have been the work of Latker and Bremer, first with the IPA program and then with Bayh-Dole.
There were two rationales for either IPA or Bayh-Dole–1) that an institution closer to the inventors could do a better job with non-exclusive licensing for actual adoption than could a federal agency with disinterested open access; and 2) institutions could grant exclusive licenses where federal agencies believed they could not or were not properly equipped to do so–to evaluate licensees, to draft tight exclusive license contracts, to police those contracts for compliance, and to threaten to enforce the government’s patent rights on everyone else in favor of its chosen licensee–all things that the 1947 Attorney General’s report argued against the government doing.
The pivot to exclusive licensing to pharma then got infused into all the new university licensing operations that were formed to cut out the RC/research foundation middlemen everywhere those agents were highly selective. That exclusive licensing model then drove out non-exclusive licensing as the default everywhere except the NSF CRC programs (which NSF finessed without using Bayh-Dole exceptional circumstances), NIH tools (after a fight in which the NIH declined to declare exceptional circumstances under Bayh-Dole), and open source software.
The result is a default compulsory, non-selective, exclusionary licensing practice that’s all in or none.
I dealt with this for over a decade in university licensing. We ran non-exclusive default tech transfer where the IP rights were subordinate to transfer services–delivery, instruction, demonstration, research review, updates, consulting, and the like. The patent folks hated us, argued that inventors were confused by two offices with different models–meaning that inventors were confused if a university had two models and the inventors could choose which model they wanted to participate in–meaning that inventors weren’t confused at all but the patent-exclusionary folks were pissed off if any other approach proved to be effective and with a better financial return than their claims about their approach. We outperformed the patent exclusionary approach at UW for a decade–3:1 to 6:1 returns compared to the patent approach’s 1:1. We cost less, created less liability, had 100x more industry relationships, had repeat business, could do deals in hours rather than months, licensed much of what we agreed to work on, and most of what we licensed was used immediately.
But all this was denigrated by the patent exclusionary folks. Pharma and biotech VCs demanded exclusivity (not true, but there it is), and so that model had to be made to appear to be legitimate and the way they chose to do that was to abstract it as the only viable model that served the public. “If this is how new medicines must be created, then it is also how pretty much anything else ought to be created, too.” Thus, public-private partnerships gets co-opted in its compulsory monstrous form and the default voluntary non-exclusive + services gets driven to the edges.
Bremer and Latker did this. SUPA/AUTM was one organization they formed to diffuse exclusionary practice to universities, first breaking Research Corporation’s monopoly on patent practice (with its old-style voluntary non-exclusive default) and then using the IPA program and Bayh-Dole to introduce compulsory institutional ownership + exclusionary licensing for all because Latker thought his exclusionary licensing tactics worked at the NIH.
I have come to see this whole activity of university licensing as a story about how two men–Latker and Bremer–with technical knowledge–both patent attorneys–positioned in flagship research organizations–were able to impose their ideas about patent practice on the university and federal research enterprise. What we have is to their greater glory–as witnessed by their “founding fathers” talk (and exclusion of Howard Forman and Betsy Ancker-Johnson). For everyone else following along uncritically, it’s an exercise in how administrators can never admit they have been wrong, or fooled, or stupid. They would rather lie and deceive and double down than go, “oh, gosh, we have been wrong and we need to revisit things and re-learn how to do this.” Instead, it’s always “Wild success! but we can tweak the regulations to get even more success!” Which means, without openly admitting it, that there’s no wild success.
Most university patent administrators are not about to own up to their lack of success–it’s pesky regulations that throttle their ability to do secret exclusive patent licenses with no right of public oversight or appeal. But it isn’t even federal regulations that stymie the dominant scheme’s narrative. That scheme was dead on arrival, and even if universities were given outright ownership of every invention made in federally funded work, and were issued patents for free, and were relieved of any public interest conditions, and were told that they could misuse patents any way they wanted–their scheme would still fail. The problem is not all the things that are said to stand in the way of the scheme–the problem is that the scheme itself is stupid, crappy bureaucratic fantasy that fails to recognize the conditions of inventive–and other–work, especially those conditions in the context of research undertaken at universities, and often with federal support.