We are working through reflections of Bayh-Dole made by some iron rings in cows’ noses that claim to give milk–er, “industry leaders.” The exercise is useful not merely to mock them for their nonsense–mockery is here salutory and inclusive–but also to provide a basis for reflecting on the Bayh-Dole that these folks have never known, don’t want, and work to prevent–that is, the Bayh-Dole as it is in the statute, and what Bayh-Dole practice would do, if the law were enforced.
More cow nose ring reflections, then, on Bayh-Dole:
it has been extraordinarily successful in that respect.
The “in that respect” is apparently the claim that Bayh-Dole “facilitates commercialization.” That’s an evidence-free claim, directed to a non-objective of Bayh-Dole. Commercialization is so unimportant that no one bothers to track commercialization with respect to Bayh-Dole. No one says–here are my Bayh-Dole patents, and for each, here’s when, if ever, the claimed invention has been commercialized. The NIH IPA rate of commercialization was 4% of patents. Stanford reported, a few years ago, that 20% of their patents were licensed. And of those, only 257 of 6,400 had made enough money ($100K+) that it was reasonable to expect these had been “commercialized.” Again, 4%. And that’s not considering whether the invention carried federal funding (and federal conditions).
While people coo over the idea that universities have gotten way more patents after Bayh-Dole came into effect (without any consideration for whether Bayh-Dole caused any of it), no one asks whether there’s been more “commercialization” or–and this was something discussed by the federal agencies when they were aiming to put together federal patent legislation in 1976–whether the commercial utilization of inventions made in federally supported work was any faster when exclusive patent rights were held by nonprofits. AUTM does not track “commercialization” in general and does not track “commercialization” specifically for subject inventions–inventions subject to a Bayh-Dole patent rights clause. No one tracks this. No one cares. Not an objective of Bayh-Dole. Heck, commercialization not even a fantasy objective of Bayh-Dole. Commercialization is a fantasy premise for patenting activity but not for compliance with the law. To top it off, Bayh-Dole purports to exempt any reports of invention utilization from public disclosure law (35 USC 202(c)(5))–and does not require federal agencies to ask for reports. So whatever does get reported with regard to subject invention utilization becomes a state secret.
the Act provides universities and private licensees “certainty”
This is the old “title certainty” claim–that if a university got title to an invention made in federally supported work, then the university (or more often, a patent development firm contracted by the university) had to worry that the federal agency that sponsored that work in which the invention was made might not agree to allow the university or its patent development firm to retain that patent position. For nonprofits and contract research organizations, the policy was: unless you are in an IPA program, the default for any invention is government open access. That’s what a federal patent meant, before Bayh-Dole. Under open access, the nonprofit had full access to the invention, as did the inventors, as did the research team and all its collaborators, as did professionals and industry and investors and entrepreneurs. A nonprofit in an IPA program had a deal with federal agencies–NIH or NSF or both–that required the nonprofit to take ownership of any invention made in agency-supported work that the nonprofit chose to patent. The deal itself was corrupt. Universities with patent policies that did not require invention assignment could therefore go to the NIH and agree with the NIH that they would require assignment of inventions. The universities–starting with the University of Wisconsin–altered their patent policies (Wisconsin didn’t even have one) to become compulsory assignment policies without altering the text of their patent policies. They voluntarily agreed with a federal position that they would require assignment. They voluntarily chose to waive their own policy and preempt it with what appeared to be a federal requirement that wasn’t a requirement at all. For everyone else, and non-IPA agencies, the nonprofit had to make a case that its exploitation of exclusive patent rights would better serve the public than would open access.
Thus, “certainty” as used here by a Bayh-Dole reflecteur means: nonprofits now were certain that they did not have to make a case that monopoly exploitation of any given patent position would better serve the public than would open access. As for “title certainty”–that a nonprofit could hold title for the life of a patent without any government claim–that is not in Bayh-Dole. More fantasy.
What was uncertain was whether a federal agency would agree to a nonprofit’s patenting for exclusive licensing unless there was a clear case the public would be better served. Universities made crappy arguments for their patent claims. Bayh-Dole, as they reflect on it, gave them a free pass. What happiness, to be freed from explaining how, for any invention they wanted to suppress access to, how doing so would better serve the public interest than in encouraging access.
Just to beat on certainty a bit more, there are plenty of ways that Bayh-Dole messes with title “certainty.” We might say that Bayh-Dole displaces title uncertainty from requesting to retain title (with a showing that retaining title will be better for the public than open access) to compliance with all the administrative requirements which, if breached, might result in loss of title. Bayh-Dole sets out the patent rights clause provisions under which the federal government may take title, once a contractor has got title. One of these conditions arises when a contractor fails to disclose a subject invention timely and in the form of a compliant written report. This claim on title shifts the “certainty” to the definition of what constitutes a subject invention. If a contractor fails to identify an invention as subject and properly disclose it, and instead goes off and patents it without disclosure, there goes “title certainty.”
Bayh-Dole also provides a mechanism by which federal agencies can implement alternatives to allowing contractors to “retain title” to inventions (see 35 USC 202(a) and (b)). Under the “exceptional circumstances” provision, agencies can introduce alternative patent rights clauses that, for instance, re-establish government ownership and open access. The Department of Energy did just that, before Bayh-Dole’s implementing regulations were even issued, baking an exceptional circumstance into an alternative patent rights clause, once at 37 CFR 401.14(b)–now conflated by NIST into a single, ungainly, omnibus patent rights clause. No title certainty there.
It is worth pointing out that if federal agencies used the alternative patent rights clause authority of 35 USC 202(a) and (b), things would be restored to pretty much the same condition as they were pre-Bayh-Dole. Federal agencies could stipulate any invention made in public health directed research or development default to federal open access, for instance, just as the DOE did to restore its policy for nuclear weapons and related atomy stuff. There would also be an alternative patent rights clause for research tools, for cooperative research center work, for work directed at open innovation and open source software, and for work in agriculture extension (farmers get a royalty-free license to practice, too, not merely the government!).
In short, if federal agencies complied with Bayh-Dole, then we would get a rational set of invention practices directed at the nonprofit research “industry” and its rogue patent fixation. University patent administrators would. be inconvenienced, poor babes. University research administrators would have to ensure that they didn’t mix funds with differing requirements if they wanted to preserve the opportunity for patent administrators to mess with patent exclusivities–just as they have to do so when they allow a company research sponsor to have a first right to negotiate an exclusive license to any research result. Can’t mix those funds with any other funds that might grant even a non-exclusive license to someone else. Oh, I know, research administrators now don’t ensure that funds are not mixed. They live on borrowed time awaiting the lawsuit for double licensing and breach of contract. They only fuss about it when the problem is mixing federal funding with differing patent rights clauses. Then they put on a pathetic show of hurt.
We have not even got to the broader, rational “certainty”: with federal open access, everyone is certain that they will have access, that they will never be sued by the government for infringement, that the government will never get into a deal that authorizes some private entity armed with once-federal patent rights to sue everyone else for infringement, either to suppress use, or just for the money because they can. To be certain such bad behavior will not happen and will not be enabled by the federal government is a good sort of certainty. Bayh-Dole blows up this certainty. Not only do contractors get to keep inventions without public oversight or right of appeal and without showing that doing so will better serve the public but also Bayh-Dole authorizes federal agencies to do exclusive patent deals that give the assignee the right to sue for infringement. Those deals blow up open access certainty, suppress broad, rapid use, and enable trolling. Only bureaucrats operating at a high level of professional stupidity could create such a mess and spend the next forty years unable to recognize what they have done, even when they are given the opportunity to reflect.
One might say, then, that Bayh-Dole introduces way more uncertainty into invention access, and all so that nonprofit patent administrators can rest easy at night that they don’t have to show that their management of patent monopolies better serves the public–that they don’t have to report to the public, don’t have any public accountability, and that the public has no right of appeal under the law. That’s a certainty of a sort, to be sure. But isn’t it odd that the objective of Bayh-Dole would turn out to be to help nonprofit patent administrators get better sleep?
Bayh-Dole is reduced to this fantasy: the United States will regain its global technology swagger if university patent administrators get better sleep, knowing they no longer have to account to the public for their exploitation (or not) of patent monopolies. One can see how university patent administrators might gravitate to this fantasy. But even if this is the consensus view–stated in bureaucratic terms, of course–that consensus does not alter the law.
It’s just hard to see how not having to make a case for public benefit somehow facilitates commercialization. It’s not hard to see how getting a free pass allows patent exploiters to exploit however they might. Mostly that means they make a “good faith” effort to find exclusive licensees, mostly fail at this, and rationalize it all by claiming that licensing is complicated and hard and they need more money to do it well at all. And, they are reduced, then, to talking about patent “portfolios”–meaning, that they take control of many inventions in the hope that one or two patents every twenty years will float the entire licensing operation. That’s hardly the “commercialization” of inventions that one might expect from the opening reflections of the Bayh-Dole reflectors. Amassing a patent portfolio to make sure that nothing lucrative gets a way and damn the rest, a few winners pay for all the losers and cover for all the bad administrative decisions does not appear to be strongly aligned with the commercialization of each invention, faster than by open access, in a way to put the U.S. back in the forefront of global technology.
We will leave off for now reflecting on how the burden of Bayh-Dole–all those public protections, especially directed at nonprofits–facilitate any of utilization, participation, collaboration, competition, or US manufacturing. That’s a central problem for technology diffusion–and to get at that problem, we have to use a vocabulary and mindset that the present “reflections” will never get to.