Here’s a recent search on Research Enterprise: “bayh-dole provisions only apply to subject inventions.” Is it a question? Is it an assertion?
Does Bayh-Dole apply only to subject inventions? No. Some Bayh-Dole provisions do apply to subject inventions, but much of the law does not concern subject inventions. 35 USC 200, for instance, applies to all inventions arising from federally supported research or development. 35 USC 207-209 applies to the licensing of all inventions owned by the federal government and 35 USC 202(c) provisions expressly apply to subject inventions but not to inventions made under a funding agreement but that a contractor does not acquire.
The botched definition of subject invention
Bayh-Dole defines subject invention at 35 USC 201(e):
The term “subject invention” means any invention of the contractor conceived or first actually reduced to practice in the performance of work under a funding agreement:
This definition is fraught with problems. Wicked bad stuff. Don’t ever forget that. We will work through it with a broad brush.
“Invention” is a defined term 35 USC 201(d). Botched definition there.
“of the contractor” means owned by the contractor. Stanford v Roche. The contractor has to acquire the invention by conventional means, such as assignment from the inventors. Or the inventor has to be a contractor.
“conceived or first actually reduced to practice” is from interference practice–who invented first–and now obsolete. Was used in federal contracting to get at scope of deliverables rather than priority. Botched for that purpose, too. “Conceived” is a term of art in patent practice–now also not relevant since the United States went to first to file from first to invent. Think of conceived as “having an entire invention in mind and recognizing it as inventive.” Think of “first actually reduced to practice” to mean testing to demonstrate that something conceived actually works as claimed since past experience has made making claims without evidence sketchy. Another way: some inventions require demonstration that they work to be patentable. Actual reduction to practice is then necessary to make an invention patentable at all. Prove it by testing it or no patent.
“in the performance of work” implies there is a scope of work–the thing to be done. Inventions made outside this scope will not be subject inventions. But what does it mean to perform “work”? Sounds good, but it’s really quite pompous emptiness.
“under a funding agreement”–funding agreement is defined at 35 USC 201(b). Two sentences. First sentence limits funding agreements to “contract” or “grant” or “cooperative agreement”–call it something else, like “award,” and it’s not a funding agreement. Further limitation: for “performance of experimental, development, or research work.” 35 USC 212 then provides that if the purpose of the funding agreement is “education,” then there’s no federal interest in anything inventive. So–make the purpose of the funding something other than experiment, development, or research–testing, say, and anything invented isn’t going to be a subject invention. Then expansion–“funded in whole or in part by the Federal Government.” Oh wow. The “work” is broader than what the government part funds. “Performance of work under a funding agreement” means that the funding agreement may specify work to be done but fund only some part of that work. The contractor then does the rest of the work on its own dime, but inventions made there are every bit as subject as the ones made with the “in part” federal funding.
I’ll come in again. A federal agency has in mind some generalized outcomes from funding research. Any number of things. Discoveries, inventions, tools, applications, test results, data, products. The agency provides funding to contractor w to do some work toward any of those generalized outcomes. Whatever the contractor also does on its own toward any of those generalized outcomes, then, is also part of the funding agreement, since the funding agreement need fund only in part. It’s not what the federal money is spent on. It’s what the federal agency is spending money on by funding contractor w. Anything else contractor w does directed at what the federal agency is aiming to do is also part of the “performance of work.” See 37 CFR 401.1 for an awkward discussion of scope that in its backhanded lizard-like way gets at these points–but you have to think about it unblinking for a while on a warm rock to finally get it.
The second sentence of the definition of “funding agreement” just blows. Funding agreements are extensible. Anyone that’s a party to a funding agreement may add parties to the funding agreement. That is, add contractors. They may do so by “any assignment, substitution of parties, or subcontract of any type.” Assign a subject invention and you’ve added the assignee as a contractor–the question is what patent rights clause controls what that new contractor does not just with the assigned invention but with any other invention the contractor may acquire that’s been made “in performance of the work” “under the funding agreement.” The new contractor may not get any federal funding. Doesn’t matter. The federal funding need be only “in part” to fund some part of the greater “work.” Taking assignment of a subject invention means agreeing to help the federal government to achieve its objectives–volunteer work, even, or at least volunteer the inventions that arise from the volunteer work. It’s not really “volunteer” work since the consideration is in part the assignment of a subject invention that the new contractor now owns.
Had enough? This is not even a deep dive. Just enough to persuade you that the definitions here are a botch job as in “Rank idiots used boxes of dynamite to prop up the door to the mine. Looks safe enough. Let’s take a look inside! Got a match?”
The botched definition of invention
“Invention” is a defined term in Bayh-Dole. This is in itself weird, because federal patent law otherwise doesn’t define invention. Bayh-Dole is an amendment to federal patent law. Go figure. Here’s 35 USC 201(d):
The term “invention” means any invention or discovery which is or may be patentable or otherwise protectable under this title or any novel variety of plant which is or may be protectable under the Plant Variety Protection Act (7 U.S.C. 2321 et seq.).
This definition is also a royal botch job. I can’t begin to explain all the hurt. Invention is any invention or discovery. Well whoop. That’s a restriction based on usage. “Is or may be patentable” leaves open who decides what may be patentable. And what’s patentable is itself a wide open topic. If an inventor does not recognize inventiveness, then an invention is not patentable. So where are we then? Running circles. It’s a pseudo restriction that tries to become general by adding “or otherwise protectable”–the result is “invention is any [invention or discovery] protectable under Title 35.” But what the heck does “protectable” mean? Why not just recite 35 USC 101 and be done with it? Patent law–a person who invents or discovers any new and useful process, machine, manufacture, or composition of matter can obtain a patent. If invention is “any new and useful process, machine, manufacture, or composition of matter” then a subject invention would be some subset of such stuff. Do you see how botched things are from the start?
Worse, Bayh-Dole’s definition of invention tacks on plant varieties, which are not patentable and are treated by a separate body of law. This is just a clever botch for plants that are patentable under 35 USC 161. So recite 35 USC 101 and 161 and be done with it. But no. So we get plant varieties included with plant inventions, and inconsistently at that. You can chase down the inconsistencies if you have nothing better to do.
At the root, this is a problem in bureaucrats trying to create in words an abstraction of the deliverables that ought to be specific to a particular contract for research or development services. They want a “uniform policy” rather than have each contract negotiated to meet the needs of the agency and the contractor. It’s bizarre when you think about it. Why would a “uniform policy” be high on the list of things to have in hand in “unleashing” innovation through federally supported research? Sure, if research funding was pork and needed to be distributed geographically to ensure every state got their share of it. But to create such a policy and then try then to improve on its uniformity to somehow spur research discovery or technology transfer or economic development is just bonkers. May as well wear grass skirts and dance for rain–which would at least be more entertaining than what bureaucrats normally do.
Think of it this way. Consider a specific funding situation. A federal agency wants a contractor to work on a particular problem–say, to study compound x for biological effects. The deliverables from such study would then include any discoveries having to do with compound x and any inventions having to do with studying compound x. Deliverables could also include anything that the contractor’s personnel think of while leaping ahead from biological effects of compound x to how those effects might be used–yes, just what the federal agency would intend to do by funding the work in the first place. It’s not like a federal agency thinks–“let’s fund compound x for its biological effects but let’s not justify doing so because if there are effects they might be used in treating disease or anything.” If you want the deep dive, start with Mine Safety Appliances. So now you have one class of deliverables for one research contract. Now do that again for each of a few thousand research contracts. Now try to construct an abstract definition that accounts for all those deliverables and doesn’t toss out any of them.
But Norman Latker at the NIH was more determined to get Bayh-Dole than he was competent to draft it and so we get a botched definition of invention to use in a botched definition of subject invention all made up into federal law that everyone has to treat as if it were intended and not botch through and through.
Where have we got? If a contractor does not acquire a given invention, it is not a subject invention and Bayh-Dole in certain of its contracting parts doesn’t apply. Nothing requires a contractor to acquire inventions, even inventions made “in performance of work under a funding agreement.” Those unacquired inventions fall outside of “subject invention” but not necessarily entirely outside of Bayh-Dole’s scope.
Bayh-Dole’s statement of policy and objective is broader than subject invention
Consider: 35 USC 200–Bayh-Dole’s statement of policy and objectives–is more general than subject invention:
It is the policy and objective of the Congress to use the patent system to promote the utilization of inventions arising from federally supported research or development
35 USC 200 applies to any invention “arising from federally supported research or development.” There is no restriction to inventions owned by contractors. A contractor could refuse to acquire an invention. It is not a subject invention, and under Bayh-Dole, federal agencies have no right to request title to such inventions under Bayh-Dole, and contractors have no obligation to disclose such inventions to the federal government (see 35 USC 202(c), where federal agencies get their rights to acquire title–to subject inventions, not to just any inventions made in the performance of work under a funding agreement).
For an invention “arising from federally supported research or development” to be a subject invention, it must be owned by a contractor–a party to the federal funding agreement. If no contractor acquires the invention, then it cannot be a subject invention, even if the invention was made in performance of work under a federal contract. The Supreme Court was clear about all this in Stanford v Roche.
Thus, it is inaccurate that a contractor must disclose every invention made with federal support or made in work receiving federal support. The contractor’s obligation to disclose is limited to those inventions that the contractor owns, that the inventors have disclosed in writing to the contractor’s patent personnel, and that are or may be patentable, and have been made in performance of work under a funding agreement.
Subject invention then is a special class of such invention–there’s that botched definition tree at 35 USC 201–(e)[of a contractor, conceived or tested] then (d)(patentable or otherwise, and plant varieties] then (b) [funding agreement in part, additional contractors] then (c) [any party to a funding agreement is a contractor]. A subject invention is one owned by a federal contractor and made under a federal contract. All other inventions, whatever they are, are not subject inventions.
The incompetence of NIST was on display when it changed the standard patent rights clause at 37 CFR 401.14(f)(2) to include a requirement that contractors require inventor employees to assign subject inventions to the contractor. The contractor has to already own the invention for it to be a subject invention. The Supreme Court in Stanford v Roche made clear that there is no authority in Bayh-Dole to force inventors to give up the rights secured to them by Congress under the Constitution’s patent and copyright clause. And if the inventor employee is also a contractor, then the institutional contractor has no right to require assignment under 35 USC 202(a), 37 CFR 401.14(g), and 37 CFR 401.9. Indeed, by complying with 37 CFR 401.14(f)(2)–a provision not in Bayh-Dole–an institutional contractor gives up claims to inventions of inventors and at the same time makes those inventors parties to the funding agreement–i.e., contractors. Go figure. This is beyond the mental capacity of most university bureaucrats, so if you get it, don’t get all uppity about it. They won’t believe you. You are mere thunder or a cold breeze or a crow squawking.
Beyond this drop out of inventions not acquired but which 35 USC 200 still applies to, fully half of Bayh-Dole deals with the licensing of federally owned inventions that do not have to be subject inventions at all. Those inventions include inventions acquired from federal employees–and those inventions are not subject inventions. Those inventions are acquired under 37 CFR 501, not 37 CFR 401.14.
Even on the federal contracting side of Bayh-Dole–35 USC 207-209–there are inventions arising from federally funded research or development, and then there are subject inventions. Both exist. Nothing in Bayh-Dole indicates a Congressional intention to force all inventions arising from federally supported research or development to be subject inventions, nor even all inventions made by a contractor’s employees. Bayh-Dole’s contracting provisions apply to inventions arising from federally supported research or development which are acquired by a contractor. If no contractor acquires such inventions, then they are not subject inventions. 35 USC 200 applies–but not 35 USC 202, 203, or 204, which are specific to subject inventions.
35 USC 200 is more than just a statement of Congressional intent. It also announces a policy and that policy is embedded in federal patent law–Title 35–and as 35 USC 261 has it, patents are subject to the provisions of Title 35. Patents on inventions within scope are subject to 35 USC 200’s policy even if they are not subject inventions. Bayh-Dole’s policy runs with each patent, much as a covenant might run with ownership of real property.
Non-subject inventions made under federal contract
What’s odd–and we have to assume this was intended, not some gaff–is that Bayh-Dole does not deal with the non-subject inventions made under contract. Why not? Well, because when Bayh-Dole was passed, there was executive branch patent policy to deal with all such inventions. That was the Nixon patent policy of 1971, which modified the Kennedy patent policy of 1963. That policy controlled the disposition of inventions in all executive branch research and development contracts. It was codified as the Federal Procurement Regulation, with stipulations for standard patent rights clauses.
Bayh-Dole was passed in the context of the FPR, and extracted, as it were, nonprofit and small business federal contracting from the FPR and announced a new regime for these contracts when a contractor acquired an invention made under contract. If a contractor did not acquire a given invention made under contract, then that invention was still controlled by the FPR. If inventors wanted the benefit of Bayh-Dole–royalty sharing, say–they had some reason to assign their inventions made under contract to the contractor they worked for.
But strange things happened. First, Reagan issued a Memorandum in 1983 requiring all federal agencies to use Bayh-Dole for all contractors, not just nonprofits and small businesses. That Memorandum in effect amended the Nixon patent policy and pre-empted the FPR. Worse, in 1985, the FPR was replaced by the Federal Acquisition Regulation, and the FAR implements only Bayh-Dole–see FAR 52.227-11.
So the federal government removed the general requirements on research contracts and focused on subject inventions only. Which was really stupid. The folks involved did not think about what “of” meant in an “invention of the contractor.” “Of” means ownership, not merely something made under the auspices of the contractor, or with the contractor’s resources, or by a contractor’s employee. Again, see Stanford v Roche.
Much effort has been devoted to trying to construe Bayh-Dole as comprehensive. University patent administrators and their attorneys insisted that Bayh-Dole vested title in all inventions made under contract with the university-as-contractor, or that the university had a right of first refusal and inventors had to assign to either the university or to the federal government. Stanford v Roche shot that down.
Then folks decided that to make Bayh-Dole work “as intended” they had to demand that all inventors assign all their inventions upfront to the universities–a present assignment. One law professor went so far as to argue that those drafting the law made the assumption that all federal contractors obtained patent agreements compelling assignments from their inventors. That ignored the fact that the primary drafter of Bayh-Dole was Norman Latker at the NIH.
Latker had worked closely with Howard Bremer, patent counsel for the Wisconsin Alumni Research Fund, in drafting the NIH’s Institutional Patent Agreement master agreement. Latker would have known all too well that the University of Wisconsin not only did not require patent assignments, it did not even have a patent policy. The IPA program was designed to create an obligation for inventors to assign at universities that had implemented no such obligation in policy or employment. Latker claimed that Bayh-Dole was based on the IPA program, which in ways it sort of was. But unlike the IPA program, Bayh-Dole, as part of federal patent law, does not require inventors to assign their inventions to the nonprofits that happen to host federally supported research.
Nothing in Bayh-Dole suggests that the law was intended to deny inventors the rights that the Constitution authorized Congress to secure for them. Senator Bayh insisted, much later, that he intended the law to deny inventors their rights–in an amicus brief in the Stanford v Roche case, Bayh argued that inventors were last in line to have any claim to owning the inventions they made. The Supreme Court rejected Bayh’s claim. His account of personal intention long after the fact had nothing to do with the interpretation of the law. And that’s a good thing.
However, university administrators and their attorney cohorts, having been thrashed by the Supreme Court decision in Stanford v Roche picked up on an argument made in a dissenting opinion–that an agreement to assign a future invention ought to be pretty much the same in effect as an assignment now of an invention that’s yet to be made–and twisted that around into the idea that all university personnel should be made to assign upfront all their future inventions before any of those inventions had been invented, so that would make Bayh-Dole work “as intended” (where it’s not stated who indeed was doing the intending, other than university administrators and their attorneys).
If all inventors are made to assign everything they invent before they invent, then, so the argument goes, all such inventions will be owned by university-contractors, and thus will all be subject inventions, and there you go. It’s a bullying position–and so it succeeds only because most university personnel don’t have the quarter million dollars and two or three years to argue the point in court.