On June 29, 1976 Mick Stadler wrote a letter to Willard Marcy, the Vice President of Research Corporation’s Patent Program. Stadler, at the time was assistant director of the Case Western Reserve technology transfer program. He would go on to work for University Patents and then Research Corporation and then quietly went around the country involved in venture-based organizations spinning out technologies and companies from universities. The general context of Stadler’s letter is a “Patent Awareness Program” organized by the NSF and a proposal to extend the NSF and NIH IPA programs government-wide. Thus, one of those cc’d on the letter is Norman Latker at the NIH. The reason to revisit this letter after so many years is Stadler’s discussion of what is needed in a technology transfer program.
The IPA program required a participating university to demonstrate that it had technology transfer capabilities, and not all–indeed, perhaps, not many–had such programs or even the resources to create them, or even the volume of inventions to justify the expense. Instead, Research Corporation, which was the leading patent development firm at the time, might play a role in setting up regional offices for technology transfer, hosted by a cooperating university to provide services to a number of universities in the area whose inventors then could liaison readily with Research Corporation staff, at a cost of perhaps a half-time administrative position per university. It was a brilliant proposal, but as it turned out, Research Corporation could not execute on the plan.
In the course of his letter, Stadler lays out eight services of an “effective transfer mechanism.” His observations then are just as relevant today. Let’s work through these eight services.
This is fascinating. Stadler’s starting point is the needs of the inventor. Not the needs of the university, not needs of the patent development firm, not the needs of a “technology transfer” office. The needs of university inventors are wide ranging. Some want broad access. Others want research funding. Others want industry use. Others want a startup company. Others want a partnership with a big company. Others want to get rich. Others want to have nothing to do with commercialization. One might create university policy then to threaten inventors with firing or reputation smearing to turn them all into submissive, everyone-who-cooperates-gets-a-trophy “academic entrepreneurs” or one can create university policy that recognizes and requires the university to be responsive to the inventors’ varying needs.
We might broaden the idea of inventor here to supply context: in universities, almost all inventing takes place in research collaborations–research teams, with leaders and colleagues and assistants. An inventor could be any or all of the members of a collaboration. What are the needs of that collaboration? Who represents the interests of the group that is closest to the invention and creates the conditions under which an invention is made?
Stadler points out that any technology transfer mechanism must give the inventor options. This makes sense. But it runs entirely against present-day policy expectations that inventors don’t have alternatives and should not. They don’t get a choice who they will work with–that would be “free agency” and AUTM, among other organizations, has blasted away at that idea, without any good reasons. Stadler’s position comes about from the idea that patent development firms operate as agents for inventors, brokering the inventors’ inventions to industry. University policies have been written to exclude this idea. Some public universities have gone so far as to assert that an inventor attempting to have a say in university licensing commits an ethical violation–attempting to influence a transaction involving the state for personal reasons. It’s an awful position to take, but it’s been done.
Even Senator Bayh argued (wrongly, it turned out) in an amicus brief to the Supreme Court that Bayh-Dole made inventors last in line for the disposition of rights:
In the Bayh–Dole hierarchy, the ownership rights of contractors come first, followed by the government, with inventors last
The inventor has no direct ownership interest in inventions developed with federal research funding, but only a provisional, subordinated ability to obtain title if the contractor and the government choose not to exercise their superior rights. Under the statute, neither the contractor nor the government requires an inventor’s permission to develop their respective statutory rights.
The Supreme Court rejected Bayh’s argument. Inventors, it turns out, are first in line, as they always have been. Only if they contract away their rights to inventions do they give up their rights, and in such contracting they have the opportunity to specify the conditions under which they will give up ownership. They have alternatives until they choose not to. This, too, is entirely consistent with the U.S. Constitution and federal patent law.
Bayh’s interpretation of Bayh-Dole is crushingly antagonistic to inventors, and especially to inventors hosted by universities undertaking work that is determined to be so much in the public interest that it deserves federal support.
Why is a patent necessary for an industrial interface? That’s a fundamental question, and there’s not a single answer. The conventional answer now–and it is generally wrong, even if there are rare instances in which it is right–is that patents provide monopoly positions without which companies will not develop inventions into commercial products and without such products public research money is wasted. That’s a nonsense position, but it sounds good and has been effective in diverting public money to subsidize profit-seeking speculators while blocking access by professionals, by companies seeking to do research or use for internal purposes, by those working to develop standards–by most anyone other than patent speculators.
Another answer–one that gets poo-poo’d but is closer to the truth–is that patents are necessary to mark places in which researchers have been sufficiently diligent to demonstrate that a new thing–invention, discovery–is fundamentally new over the prior art and non-obvious over that art. That effort to obtain a patent confirms the claim that something is new. That counts for something. The effort to document how the claimed invention is new also counts for something. But the exclusive right that comes with the patent counts for very little in the interface between academic work and industry. The right to exclude might help in enforcing standards (so companies cannot claim a standards implementation when it is not) and in pushing back on companies that would take in rights but then sue to block others from practicing their improvements that rely on those same rights. Other than that, patent exclusivity does very little good for universities at their industry interface. Mostly, patent exclusivity isolates a university, teaches industry to ignore university research or avoid it, and leaves universities to cultivate relationships with patent speculators–folks who do expect to derive value from suing others who are using a claimed invention, either to shake them down (the patent did not attract investment and people developed despite lacking even the right to do so) or suppressing them (the patent works against use–exactly the opposite of Bayh-Dole’s express policy).
But where Stadler is spot on is that any university technology transfer mechanism has to be clear (and we might add, intellectually honest) about both the economics and the problems with putting patents out there at the industry interface. Is exclusive patent licensing a winning proposition? Generally, no. Can universities use the patent system to promote use of inventions, maximum participation of small companies in federal research and development, free competition and enterprise, and the like? Yes, but not generally by adopting exclusive licensing as the opening default position. Exclusive licensing, if there every must be any, must come after non-exclusive access for making and using an invention has demonstrated there’s an opportunity for commercial production and the variations in the inventive work have been sufficiently explored that defining a mass-produced version is relatively clear. Commercialization opportunities arise from use, not from the suppression of use. Whatever garage inventors might think, universities are not garages and the faculty, students, staff, and collaborators that conduct research and think about things aren’t acting in isolation. The use of patents at the university-industry interface is not at all the use of patents at the independent entrepreneur-industry interface. Somewhere, somehow, people have to understand this point.
Cost-effective. I can hear structure in the side bands here. What does it mean for a technology transfer mechanism operated on behalf of a university to be “cost-effective”? It’s not merely that the mechanism makes more money than gets spent on it, although that’s a start. The mechanism cannot cause damage and suppress opportunities for more than it is worth. It’s not acceptable that a transfer mechanism can cause all the damage, bitterness, and lost opportunity it wants so long as it brings in more money through its devices than it costs to operate.
Cost-effective also goes another way, toward selectivity matched to relationships. If one patents every invention one sees, then the problem of trying to match each patent to one or more companies becomes impossible. Each patent runs $10,000 or more, and much more if one starts down the road to filing foreign patent applications. Even in a winners-pay-for-losers mindset, running up hundreds to thousands of losers means that one has to get a huge winner, and the odds of that decrease greatly as the hugeness needed rises. In the present approach of default exclusive licensing, even the best universities (or luckiest) find a lucrative winner once a decade, or once every 2,000 inventions. While spending $20+ (or $40+ or $60+) million on patenting and more on salaries and space in the hope of winning big ten years or more down the road might end up being a net positive, it is hardly “cost-effective.” That’s a defective gambler’s mindset–not the strategist that gambles while working the odds, but the fool loser that gambles not knowing the odds.
By contrast, if a technology transfer mechanism selects for management only those inventions that it has an established expertise in placing, then it can work the odds so that it transfers most of the work that it takes in. That’s being “cost-effective”–that is, not spending wildly, and being effective with what one does spend on.
In the broader of public efforts, there is absolutely no reason for a university to have to make money on its technology transfer any more than it must make money on its libraries or its campus trees and ivy. A highly effective transfer program might make nothing from licenses–because the licenses are royalty-free. How then does it cover the cost of the patents that it acquires? First, it might seek donations, or if it is a state school, an allocation from the legislature–and if you laugh that off, then you already have your critique of whether university patenting is generally in the public interest.
The Research Corporation approach, however, took a different path. The university did pay nothing for patents because inventors were directed to ask Research Corporation to manage their inventions. As an external agent, Research Corporation could be selective and pay for those inventions that it did see a future for within its expertise and relationships. The companies obtaining access via Research Corporation would pay royalties and those royalties would cover the cost of patenting–for each invention, without having to resort to winners paying for losers. Cost-effective means there are very few losers, not that one winner every decade means one can spend wildly without accountability for years on end.
When university administrators decided that they could do everything themselves and followed the lead of the University of California, MIT, and Stanford to take their patent licensing in-house, they ended up isolating themselves. They still haven’t figured out that they cannot do everything. They call technology transfer “early stage” and “high risk” and “difficult”–but they cannot bring themselves to say that, the way they are trying to do it, it is impossible. Their approach makes things worse. Anything that happens, happens despite their presence. Stadler imagines that outside agents with focused experience must be available to provide insight to otherwise self-isolated university patent administrators. At best, those administrators go to AUTM events where they exchange views on their isolation.
With the external agent model, each agent could focus on the areas that it knew best–technologies, industries, regions. If one university office has to handle everything from monoclonal antibodies to new rotorcraft to language learning software, then it’s an impossible task. Stadler requires any technology transfer program to provide engagement between university local administrators and external professionals. Universities do this, often, with outside attorneys, matching counsel expertise with inventions. But they have abandoned doing so for the licensing side of their work–and so have become very cost-ineffective.
When someone goes “out of circulation” to join a university technology licensing program, they have only a few years before they are no longer at the forefront of their field. Why? They don’t get out enough. They don’t see enough stuff in the same area at the same time to have a sense of perspective. And as they all do the same thing with their bits of technology, they end up working against their stated interest of public benefit in favor of pursuing their own little bit–when the public benefit might be aligned with only one or two bits of any new technology having any financial value for the inventors and all the rest must be contributed without strings attached for even those few to realize any “value.”
The concern, then, the trick in it all, is what those few that realize “value” do with their royalties. In the Research Corporation model, as originally conceived, those royalties went to a national research fund and were available to anyone. When the universities took over, they diverted royalties to serve only each their own interests. And with that diversion went the rationale for others to release their inventions so that a few might succeed and generate royalties available to the academic research community generally.