Let’s start with some Bayh-Dole basics.
Bayh-Dole preempts all other statutes but Stevenson-Wydler on matters of federal policy on inventions made in research contracts (35 USC 210). Bayh-Dole is the only authority on the matter.
Bayh-Dole requires federal agencies to use as a default a patent rights clause, and stipulates what must be in that clause. There are, actually, four such clauses (37 CFR 401.14(a)-(j), (l) [small business]; 37 CFR 401.14(a)-(l) [nonprofit]; 37 CFR 401.14(b) [nuclear power]; and 37 CFR 401.9 [inventors]).
Thus, whatever either a federal agency or contractor has by way of rights in an invention made with federal support comes by means of the default patent rights clause, unless a federal agency can justify a new clause based on “extraordinary circumstances” or another enumerated condition.
Consider, then, these observations of Bayh-Dole’s default patent rights clause for nonprofits.
The federal government may acquire ownership of a subject invention only as provided by the default patent rights clause.
That clause establishes the conditions on which the federal government may take title (37 CFR 401.14(a)(d))–failure to report, failure to elect to retain title, failure to file a patent application, failure to maintain a patent.
There is no other way for the government to require assignment of an invention, but for the conditions established by the default patent rights clause. There can’t be another secret patent rights clause that says if the first one doesn’t work, then this secret other patent rights clause applies.
Are you with me so far?
Now some definitions, in short hand. Go to the regs for the full text.
Funding agreement: agreement for research with a federal agency, including any assignment, substitution of parties, or subcontract of any kind.
Contractor: any individual, nonprofit, or small business that’s party to a funding agreement.
Note that individuals are included in the definition of “contractor.”
Subject invention: a patentable invention owned by a contractor and made in the performance of work under a funding agreement.
Be really clear: if an invention arising from federally supported research is not owned by a contractor, then it is not a subject invention.
This is what drives some university patent administrators and law professors bonkers about Stanford v Roche. They thought Bayh-Dole was a law that decreed that institutions were better judges of invention management than lousy inventors. Wrong.
Now for the juicy part.
The default patent rights clause requires universities (and other contractors) to require their inventor employees to make a written agreement to protect the government’s interest in subject inventions. That’s the (f)(2) requirement. It’s not in Bayh-Dole. Bayh-Dole does not require the (f)(2) provision, nor its cognate regarding subcontracting (g).
If universities comply with (f)(2), then potential inventors become parties to the funding agreement and therefore are contractors. Because common law provides that inventors own their inventions, when an invention is made in work supported by a federal funding agreement, it becomes a subject invention and Bayh-Dole’s default patent rights clause kicks in.
If, however, universities don’t comply with (f)(2)–and as far as I can tell, none of them do–then an invention is not a subject invention until it is acquired by a contractor, such as the university that hosted the federally supported research. The federal government has no rights in any such invention–there is no requirement to report inventions that are not subject inventions, and Bayh-Dole makes no provision for the government to acquire inventions if a university fails to comply with (f)(2). That would be hard, after all, since (f)(2) is not in Bayh-Dole!
If someone is looking for a “hole” in the swiss cheese that is Bayh-Dole, that’s a big one. The university administrative fixation on requiring assignment of inventions is not based on complying with Bayh-Dole–the universities ignore almost every substantive element of Bayh-Dole. Instead, the fixation arises because universities try to patch over their non-compliance. The argument is–since we got ahold of every invention and turned it into a subject invention, the federal government should not care that we breached the federal funding agreement, abrogated academic freedom and freedom of research, and generally went about it in a legally sketchy way that doesn’t get challenged because most academic inventors don’t have the time or money to challenge university administrators on the matter.
But here’s the problem–if folks actually care about Bayh-Dole compliance. When a nonprofit owns a subject invention, the default patent rights clause requires the nonprofit to assign the nonprofit patent rights clause in any assignment of the subject invention (see 37 CFR 401.14(a)(k)(1)). That is, in particular, any assignee must accept all of section (k) of the nonprofit patent rights clause–share royalties with inventors and–this is the big one–deduct only expenses incidental to managing subject inventions and–this is the far bigger one–use any remaining royalty income or income with respect to the subject invention for “scientific research or education.”
There it is–not just royalty income (from licensing) but also any income with respect to the subject invention. Hold this thought for one more point.
An invention is assigned when the substantial rights of the invention–make, use, and sell–are conveyed to another. An assignment can happen either expressly, by stating that the entire right, title, and interest is assigned, or implicitly, by exclusively licensing the substantial rights to make, use, and sell.
Almost all universities implicitly assign subject inventions in their exclusive patent licenses. They make a big fuss about not assigning title to the patent. But Bayh-Dole has nothing to say about patent assignments by nonprofits; Bayh-Dole is concerned with the assignment of subject inventions by nonprofits.
There it is, another glorious hole created in Bayh-Dole by university administrator non-compliance with the default patent rights clause. Universities are in non-compliance multiple ways: (1) they do not require the (f)(2) written agreement and substitute their own defective assignment requirement to make up for it; (2) they assign subject inventions to companies without federal agency approval (which would be required); and (3) they do not make express that the assignee is also subject to the nonprofit patent rights clause. For companies that accept such an exclusive license–especially one under which they accept their obligations under Bayh-Dole–they cannot pocket the profits–they have to allocate those to scientific research or education.
No company in its profit-seeking right mind would accept this provision. They would demand that the license be rewritten so that not all substantial rights were conveyed–say, non-exclusive license only for making and using, exclusive for selling, with no right to enforce the patent. Like that.
If Jamie Love at Knowledge Ecology International wants to go after Pfizer for the price of Xtandi, there’s an approach. Demand that Medivation and Pfizer account for all income earned “with respect to the subject invention” and show that this income after deduction of allowable costs “incidental to the administration of subject inventions” has been allocated to “scientific research or education.” The hairball will be coughed up.
But that’s not even the hole we are looking for. There’s even better stuff.
Imagine that a university did comply with the (f)(2) written agreement requirement. Now each invention is a subject invention, and the inventor is a contractor. There is no requirement in Bayh-Dole for an inventor to assign to anyone. The Supreme Court in Stanford v Roche was crystal clear on this point.
Furthermore, if a university requires the (f)(2) agreement, then that requirement preempts anything the university would otherwise require (such as assignment to the university). An inventor cannot be delegated by a university to establish the government’s rights and at the same time have the university insist that inventors have no rights whatsoever to establish on behalf of the government.
Just as (g) of the default patent rights clause forbids a contractor from having an interest in the inventions of a subcontractor as a condition of the subcontract, so also (f)(2) forbids a research host contractor from having an interest in the inventions of a subcontractor as a condition of participation in the contract.
With the (f)(2) agreement in place, inventors own their inventions, are parties to the funding agreement, and have no obligation to assign to either the federal government or to the university that hosted the research (or anyone else).
But wait, there’s more. When a subject invention is owned by the inventors, Bayh-Dole requires federal agencies to use a different patent rights clause (see 35 USC 202(d)–37 CFR 401.9). Federal agencies are required to treat inventors as small business contractors, but with fewer obligations than small businesses.
Thus, when an inventor-contractor holds patent rights, federal agencies must shift to a different default patent rights clause. A federal agency may request ownership only if the inventor-contractor doesn’t report or elect to retain title or file patent applications or maintain the patent. More importantly for our search for a big hole, there are none of the nonprofit restrictions in the inventor-contractor default patent rights clause, just as there are none in the default small business patent rights clause. Thus, inventor-contractors, like small businesses, may assign a subject invention without any restrictions–and can therefore license all substantial rights exclusively–just what the pharmaceutical industry wants!
This is not so much a hole in Bayh-Dole as it as huge hole in university patent practice pertaining to subject inventions. What fools these mortals be! If a company wants exclusive rights to a subject invention, it should demand to go through the inventor. The inventor-contractor might then assign (voluntarily) the subject invention to an invention management firm to represent her or his interests (or even a university’s patent licensing operation)–but doing so would be entirely outside the university’s claims to policy (waived by the university when it requires (f)(2) agreements instead).
The inventor would be free to negotiate the terms of the relationship with the university or invention management organization. The university’s “royalty sharing schedule” and policy would not apply. Or, if the university insisted that this stuff did apply, the inventor-contractor, a “small business firm contractor,” could take his or her invention elsewhere. Of course, the inventor-contractor could license or assign directly to the company–which he or she might be happy to do if the company is a startup he or she has founded to deal with the invention.
Think about that: the inventor-contractor could simply assign his or her invention directly to his or her startup company (or license) on whatever terms he or she wanted, without having to go through the step of assigning to the university, waiting for the university to pull out its 30 page template exclusive license agreement, being shocked at how awful it is, and having to try to negotiate revisions, only to find out at the first meaningful interest by venture investors that even with the negotiated revisions the deal is still shockingly awful.
Wouldn’t that be grand? If a university patent licensing office wanted to get business from inventor-contractors, it would have to offer a competitive, accountable service that worked on terms acceptable to inventor-contractors. One can see why AUTM repeatedly has opposed such an idea. Why should an inventor-contractor be allowed to set the terms for the university’s involvement–or even get to choose not to involve the university?
But Bayh-Dole is set up, with the addition of (f)(2), to do just that! With (f)(2) enforced, inventor-contractors are subject to their own default patent rights clause, and as long as they report, elect to retain title, and file a patent application (and stay with it), the federal government has no foundation (because Bayh-Dole preempts it) to demand assignment. Heck, the default patent clause for inventor-contractors does not even expressly require a non-exclusive license to the government.
No wonder folks are pressuring NIST to turn the (f)(2) requirement into an assignment clause instead. But that’s batshit crazy. Every company that wants an exclusive license to all substantial rights should oppose it on the grounds that it prevents the operation of the much more efficient inventor-contractor pathway. Not only that, Bayh-Dole forbids such a move by setting out what must be in the default patent rights clause, and the federal government does not have the authority to require such a thing as a part of federal patent law–the Constitution gives the government the authority to reserve exclusive rights for inventors, not for predatory university administrators. Besides, the Supreme Court just got done making it clear that Bayh-Dole applies only to the rights as between the research host contractor and the federal agency–nothing more. So Bayh-Dole cannot apply to the ownership of inventions between the research host contractor and its own employee-inventors.
But if NIST bothered to lay out the issues, I’m sure that companies and inventors both would oppose changing (f)(2) and would demand that universities comply with (f)(2) and thus put inventor-contractors in a position to convey rights without the ball and chain of the nonprofit patent rights clause.
And then we would see Bayh-Dole actually operate as it is written, as it has been approved by Congress. Not this faux, crappy, non-compliant, non-enforced version of the law that university patent brokers and AUTM promote as wildly successful when it’s not.