I have updated this article from June 24, 2011 in light of the Stanford v Roche decision. In its previous version, the article sets out the idea that a federal agency has a right to claim title to inventions made in federal funding agreements–and this is indeed the case under various federal laws and the the Kennedy/Nixon executive branch patent policy. Bayh-Dole does not repeal that federal framework. Bayh-Dole preempts that framework (but for Stevenson-Wydler) when a contractor comes to own a patentable invention made under a federal funding agreement. That’s the definition of subject invention. If a contractor doesn’t own it, then it’s not a subject invention, and Bayh-Dole does not apply–but the framework otherwise preempted by Bayh-Dole does apply.
In my earlier account of Bayh-Dole flow of control, then, I have the idea that a federal agency acting under the prior federal framework, may assign to a university (or other federal contractor) the federal agency’s own right to receive inventions as deliverables. That is, the federal agency assigns its side of the federal contract involving inventions to the contractor. If inventors were obligated to assign to the federal government upon request, now they would be obligated to assign to the university upon request–provided that the university “elected” its option under the federal arrangement to receive title.
In that account, then, Bayh-Dole required federal agencies to permit contractors to choose to stand in for the federal agency with regard to inventions as deliverables. The university in essence may request ownership of any invention deliverable that otherwise would be up to a federal agency to request.
It’s a neat scheme–too bad for the drafters of Bayh-Dole that they didn’t think of it. But it’s also not the scheme that is at work in Bayh-Dole, as the Supreme Court made clear. Under Bayh-Dole, a federal agency must use the standard patent rights clause if it cannot justify an alternative clause. But the standard patent rights clause applies only to subject inventions–inventions owned by a contractor–not to any invention made in a project with federal support. Under the SPRC, inventors are obligated (if a university complies) to disclose subject inventions only. That is, inventors agree to disclose those inventions made under federal contract that their employer already owns. And with NIST delegated by the Department of Commerce to look after the SPRC, NIST has added that if the employer already owns an invention made in a project with federal support, then the inventor must agree to assign that invention to the employer. Well, that makes one’s head spin. Apparently, NIST is concerned that if a university has a claim of equitable title in an invention made in a project with federal support, then the inventor must be required to agree to assign legal title to the university, whether the university wants legal title or not.
I have therefore updated the slides. I’m leaving the old article here, to follow, so you can see how I got stuff wrong. I hope that the new slides do a better job. But really, nothing can do a decent job with Bayh-Dole. Its architecture is spaghetti code. It makes no sense diagrammatically. Oh, sure, you can “simplify” Bayh-Dole to whatever university bureaucrats want it to be–but that’s not the law or the implementing regulations.
If we were to “simplify” Bayh-Dole, it would be this:
Inventors own their federally supported inventions and if they elect to retain title to their inventions don’t have to file patent applications on them or assign them to anyone unless they have promised to do so entirely outside federal funding or the use of resources paid for by federal funding.
Federal contractors are required to make their inventive employees agree to protect the federal government’s rights in patentable inventions made in projects that receive federal support. A result of requiring federal contractors to require this agreement, the inventive employees also become contractors, and their inventions become subject inventions–inventions subject to Bayh-Dole-authorized patent rights clauses.
Inventors’ obligations with regard to inventions they own are controlled by the inventor patent rights clause, a subset of the business patent rights clause.
If inventors assign their inventions to their employer, then the employer is bound by the standard patent rights clause, and if the employer elects to retain title, then must file patent applications and take effective steps to achieve practical application of the invention–to use the invention with benefits to the public on reasonable terms.
If an institutional owner of an invention fails to disclose the invention to the federal government or fails to elect to retain title to the invention having obtained that title, or fails to file a patent application, or fails to prosecute that application, or fails to maintain or defend the validity of any patent that issues, then the federal government can request title (but does not have to).
That all sounds nice, but the really simple version of Bayh-Dole is this:
Bayh-Dole re-establishes a patent monopoly pipeline run by university administrators for NIH-supported inventions flowing to the pharmaceutical industry and to speculators on the future value of those patent monopolies. Everything else is a crock up. Oh, and the patent monopoly pipeline is also a crock up, but a different one.
Everything else about Bayh-Dole is just throwing sticks in the air for distraction. In a sense, then, this whole article is a form of throwing sticks in the air–perhaps with the idea some of them will land on university bureaucrats and Bayh-Dole pundits and expose them for their complicity in the scam.
Anyway, here’s the updated article, followed by the original with all its defects.
This would be a good time to map out how Bayh-Dole operates. I’ve put together 12 slides that show the flow of control. Perhaps this will help folks see what Bayh-Dole does require, and what it doesn’t.
First, let’s look at the law and its implementing regulations. The law is at 35 USC 200-212 and is part of federal patent law. Bayh-Dole is directed at both federal agency licensing of inventions owned by the federal government and at federal agency contracting for research services with universities, nonprofits, and companies (which I will shorten the list to universities, since that’s the focus here. There are some differences for other nonprofits and for companies). More particularly, Bayh-Dole establishes the provisions for a default patent rights clause that federal agencies must use if they cannot justify an alternative.
As such, Bayh-Dole does not apply to universities, but rather to agencies.
Here then is the first lesson of Bayh-Dole (35 USC 202(a)):
Each nonprofit organization or small business firm may, within a reasonable time after disclosure as required by paragraph (c)(1) of this section, elect to retain title to any subject invention
A subject invention is a patentable invention that the university has acquired and which was made under a federal funding agreement. If a university acquires such an invention, and elects to keep owning it, then that ownership is made subject to the standard patent rights clause and anything else in Bayh-Dole–notably Bayh-Dole’s statement of policy at 35 USC 200:
The rights of the nonprofit organization or small business firm shall be subject to the provisions of paragraph (c) of this section and the other provisions of this chapter.
Bayh-Dole authorizes the Secretary of Commerce to create implementing regulations, and the Secretary has delegated the responsibility to NIST.
Bayh-Dole requires Commerce to produce standard patent rights clauses (35 USC 206). The standard patent rights clauses are at 37 CFR 401.14 (three of them packed into one omnibus clause with conditionals–one for companies, one for nonprofits, and one for certain DOE nuclear propulsion and weapons research) and 37 CFR 401.9 (for inventors as contractors, a subset of the company version of the standard patent rights clause). We will refer to the nonprofit standard patent rights clause as the Standard Patent Rights Clause, or SPRC. The implementing regulations provide for variations from the SPRC to accommodate certain agency-specific matters, treaties, foreign work, and exceptional circumstances. For this discussion, we will assume the SPRC is tailored only for agency-specific matters like the agency’s name or whether the funding agreement is a grant, cooperative agreement, or contract.
Agencies are required to use the SPRC in all funding agreements with universities. Funding agreements may be in the form of grants, which are regulated by 2 CFR 200) or contracts, which are regulated by the Federal Acquisition Regulations (FARs) and their defense equivalents, the DFARs. The SPRC is inserted into these regulations, along with federal agency variations, which the agencies then use to form funding agreements with universities.
As you can see, the flow of control in Bayh-Dole means that universities come to be affected by Bayh-Dole not directly by action of law but as terms and conditions in a funding agreement–essentially, a federal contract into which universities voluntarily enter. If a university acquires ownership of an invention–makes it a subject invention–then the university may use the SPRC to preempt federal agency objectives with regard to that invention, but on the condition that the university comply with the SPRC–with the provisio that federal agencies do not have to enforce the SPRC or act on any rights reserved for the federal government.
Bayh-Dole expressly grants universities the right to keep ownership of patentable inventions that they have acquired and which were made under a funding agreement–but Bayh-Dole is silent on how a university might come to acquire such inventions. Once acquired, however, a subject invention is subject to the SPRC and everything else in the “chapter”–that is, 35 USC 200-212.
The SPRC is drawn almost entirely from 35 USC 200-204, but includes multiple provisions that are not in Bayh-Dole proper. One of these provisions is section (f)(2) of the SPRC. Section (f)(2) requires, among other things, that universities require of their employee investigators a written agreement under which they agree to protect the government’s interest in inventions by a) disclosing subject inventions–ones already owned by the university–to personnel designated by the university, b) signing paperwork to permit patent applications to be filed on these inventions already owned by the university, and c) signing paperwork to establish the government’s rights in these inventions already owned by the university. For c), this could mean assigning ownership or granting a non-exclusive license to the government, or assigning title or granting a license to others that permits granting a non-exclusive license to the government. We will discuss this more in a bit. In May 2018 NIST added to this (f)(2) agreement an assignment requirement for subject inventions–it is a strange requirement–a university must require employee investigators to assign inventions that the university has already acquired–that is, apparently, a university must obtain legal title for all subject inventions for which the university has acquired equitable title.
For now, it is important to see that the SPRC is part of a federal contract by which agencies manage invention rights in funding agreements where a university has acquired ownership. The (f)(2) agreement is the means by which agencies manage how a university deals with its employee investigators to ensure that the agencies obtain rights in inventions owned by the employee investigators when they invent.
Here is the funding agreement situation, divided with the agency-university relationship on the left, and the agency-research employees relationship on the right:
The SPRC makes no assumptions about the relationship regarding inventions that might exist between the university and its employees investigators, other than that they are employees. It does not require assignment agreements between the university and research employees other than in the case of equitable title, per NIST in May 2018, nor that the university have a patent policy, nor that the university even have a technology transfer function. All this ought to be really concerning–but folks let it all slide.
We will work through the flow of control by which a funding agreement is established, and then examine what happens when an invention is made.
The start of a funding agreement is on the right side of the relationship, between a faculty investigators and a federal agency. A faculty member–almost always–decides to conduct a project and writes a proposal, often in response to a request for proposals issued by the agency. Faculty members are not assigned to research projects and are not required to write research proposals (though in some departments, if they don’t get sponsored research, they also won’t get tenure, and if they are unsalaried faculty, they won’t get paid without extramural funding–it’s just that the university doesn’t assign them to do research and has no right to control that research. Why is that important? Because universities have to resort to bullying to establish a claim to equitable title in faculty inventions).
The proposal is submitted to a federal agency formally through the university. The agency then reviews the proposal and makes a decision whether to fund the work. It is rare that a university refuses to submit a proposal from a faculty investigator. It may happen, for instance, when only one proposal on a given grant program is allowed for any one university and multiple proposals are prepared. A proposal also may not be allowed to go forward if it involves classified research or other research for which the university lacks the resources, such as laboratories with security to handle dangerous substances.
In general, however, the university does not censor proposals, nor does it seek funding itself and then assign faculty researchers to conduct research not of their choosing. It is the principal investigators who decide to seek federal assistance. It is their deal, their choice. Furthermore, it is the investigators that the funding agency seeks to work with. It is the investigator’s qualifications and expertise that are reviewed as part of the grant application. It is a matter of special performance. If the principal investigator bails from the deal, the award if off unless a substitute investigator is found who is acceptable to the agency. The university’s role is subordinate to the process–to provide technical services to assist in the application process, and if an award is made, to manage the financial accounting and other regulatory compliance matters on behalf of the agency and the investigators.
The university, then, is something of a steward. 2 CFR 200 makes this express. In section 316, dealing with intangible property (including inventions), 2 CFR 200 requires inventions (and other intangible assets) “to be held in trust by the recipient as trustee for the beneficiaries of the project or program.” The university, as a condition of the award, is to act as steward, not self-interested owner. An understanding of this role shapes much of what follows for invention ownership and innovation practice. Disagreement over this role is the source of many of the problems that arise in the present, now conventional, approach university administrators take to technology transfer.
When a grant is awarded, the federal funding agreement is established. The agreement has a number of components. The primary one is a statement of work, which is generally the proposal that has been submitted by the principal investigator, the faculty member responsible for the conduct of the work. A statement of work, however, may be broader than the proposal–especially if the proposal incorporates other university documents (such as policy statements) by reference. The principal investigator chooses staff and collaborators, decides whether to subcontract, creates the work plans and designs experiments, assembles the tools, gathers and reviews data, directs reporting and publication, and authorizes expenditure of federal funds.
In addition to the statement of work, the funding agreement consists of the appropriate set of regulations that govern the relationship. If a grant, then 2 CFR 200. If a contract, a set of clauses from the FARs or DFARs. In either case, these will include the SPRC with its (f)(2) agreement requirement.
When the university accepts the funding on behalf of the principal investigator, it also agrees to the SPRC. The university in accepting the SPRC therefore displaces any university policies or agreements that would conflict with the SPRC. Put another way, the university cannot accept the SPRC’s conditions while asserting at the same time requirements for its research employees that are contrary to the SPRC.
This is an important point. It is not Bayh-Dole the statute that dictates the university’s practice at this point, but rather it is the university’s own affirmative acceptance that the SPRC is how rights will be managed. Each time a federal grant is awarded, the university must require under (f)(2) written agreements for all employees who would be involved in the research. Bayh-Dole does not require inventors to disclose or assign any invention. The university must require inventors to disclose and assign–but only for subject inventions–inventions the university already owns. (If the assignment obligation were extended to all inventions made with federal support, Bayh-Dole would be back to being a vesting statute, and the Supreme Court ruled it was not).
All Bayh-Dole does is to assure the university that if it gains ownership, a federal agency cannot invoke federal statutes or regulations outside of Bayh-Dole to require assignment of the invention to the federal government. But the university must comply with Bayh-Dole’s requirements, or for some non-compliance, a federal agency may still require assignment of a subject invention, but federal agencies don’t have to require any assignment. Got all that?
As required by the SPRC’s section (f)(2), the university requires its employee investigators to make a written agreement to protect the government’s interest. In directing its employee investigators to make this agreement, the university necessarily releases them from any conflicting university agreements or policies (including IP policies and conflict of interest policies) that the university may otherwise require them to follow. The (f)(2) agreement permits employee investigators who invent to sign away patent rights or grant licenses as may be required by the funding agency.
Now the research gets going and a employee investigators make a patentable invention. In terms of the SPRC, this means that an invention is conceived or first actually reduced to practice in performance of work under a funding agreement. What happens then?
Inventions that are made in performance of work under the funding agreement by university employees investigators are not necessarily “subject inventions.” The key characteristic of a subject invention is that it is owned by a contractor–by a party to a federal funding agreement. If an inventor is not a party to a federal funding agreement, then the invention is not a subject invention, Bayh-Dole and the SPRC don’t apply, and the invention falls under the framework of invention practice that Bayh-Dole otherwise preempts.
However, if a university complies with the (f)(2) requirement, the university’s act of delegating various actions to employee investigators in their personal capacity means that these employee investigators become parties to the funding agreement–they become contractors. They own their inventions–and so their inventions are subject inventions and Bayh-Dole applies. But wait–the SPRC does not apply–instead, the inventor patent rights clause applies–37 CFR 401.9–unless the inventors assign their invention to the university or other institutional contractor.
Here then is another important lesson: in Bayh-Dole, “contractor” is almost always a collective noun, singular in form but plural in meaning. Cognitive dissonance? Yes, of course. This is Bayh-Dole.
As provided by the (f)(2) agreement, employee investigators agree to disclose subject inventions to the personnel the university designates to receive such disclosures. The university does not necessarily own such inventions, and if the university makes assignment a condition of disclosure (such as is now done at various universities–including the University of Misery), then the university violates the (f)(2) requirement.
These personnel could be technology managers in a tech transfer office, or legal counsel, or an external foundation or other invention management organization. (The personnel could also be principal investigators or department chairs, but I don’t know of any university that has done this.)
Per the SPRC, the university is responsible for reporting subject inventions to the funding agency. Then someone–the agency, the university, and the inventors–has to decide what to do with the invention. There are various options.
The effect of Bayh-Dole now becomes evident. Rather than agencies insisting on a first right to obtain title to inventions made with federal support (if the agency had such a policy in place), Bayh-Dole requires the agencies to use the SPRC. The SPRC, in turn, allows the university to “elect to retain title” to the subject inventions the university has acquired.
The phrase “elect to retain title” has been the subject of a lot of sloppy university bureaucrat guidance–and I include myself back before 2008 or so in that group. “Elect to retain title” means that the university may forestall the agency from requesting title. For subject inventions that a university does not acquire, “elect to retain title” means that the inventors also may forestall the agency from requesting title. Indeed, under the inventor patent rights clause (37 CFR 401.9), inventors don’t even have to file a patent application if they elect to retain title. If, however, an inventor owns an invention that is not a subject invention but was still made under a federal funding agreement, then the federal agency may agree to allow the inventor to retain title–but we are back to the old-style “determination of rights” practice, which most federal agencies have retired–often under the illusion that Bayh-Dole covers everything.
Bayh-Dole relieves agencies of the administrative burden of case-by-case review of each request by a university to manage subject inventions. Or, put another way, any private contractor gaining ownership of an invention made in a project deemed to be in the public interest and so supported by federal funds can preempt the federal agency’s public purpose. Isn’t it strange to see the same situation cast two different ways? Federal agencies are relieved of the administrative overhead of considering the public interest, even for research that they determined was in the public interest for some reason. Or that public interest is simply preempted by a private contractor, who is free to go all willy-nilly with patent monopolies, provided the contractor does the requisite paperwork handwaving.
It is important to recognize that the university has no mandate nor obligation under the SPRC to acquire title, retain title, to manage inventions, to file patent applications, to seek out licensees, to start companies, to attempt to develop commercial products, or to make money through licensing and infringement litigation. Any decisions that a university makes on these matters arise entirely from the motivations of university administrators–that is, from the motivations of the individuals that the university delegates to the task of making such decisions.
If the university does not acquire title to a subject invention, here’s what happens:
If the university declines to acquire title, then the relationship with the agency shifts to the right side of the diagram, the investigator side. University administration is then all but out of the picture–but for agency consultation in the matter. The agency deals directly with university inventors under the (f)(2) agreement and the inventor patent rights clause.
If a contractor fails to disclose a subject invention or no contractor elects to retain title, or for other reasons having to do with patent applications and issued patents, then the federal government may request title and the (f)(2) agreement ensures that the inventors will establish the government’s rights (such as by assignment) and assist the government in filing patent applications.
The procedures for allowing inventors to retain title are set forth in 37 CFR 401.9, which forms something of a mini-Bayh-Dole Act for inventors working as subcontractors and dealing with the government. Inventors are required to comply with a subset of the company version of the SPRC, acting in the place of the contractor as a subcontractor: offering rights if they don’t continue with a patent application in the US or any foreign country, marking patents with a federal funding legend, reporting progress toward utilization, requiring substantial manufacturing in the United States for exclusive licenses, and accepting government march-in rights if the inventors or their licensees fail to make progress or are unable to meet certain other conditions.
Most importantly, under 37 CFR 401.9 inventors are not required to use the patent system–they are not required to file patent applications. Their obligations to the federal government end with disclosure and election to retain title if they do not file a patent application. The rest of their requirements arise only if they do file a patent application.
Nothing in Bayh-Dole requires any university to require assignment of any invention made with federal support. Inventors could voluntarily choose to assign their inventions to the university. Only with the NIST requirement that inventors must assign subject inventions (ones that the university already owns) do we see any sort of requirement to assign–and there it’s a matter of equitable title–and for that there’s a whole other discussion. Upshot–universities don’t have a claim to equitable title in inventions made in federally funded research conducted by faculty. Yes, a whole deposition of university lawyers will disagree with this assertion, and it would take about $200,000 to get them to back down–but there it is. Screw the truth. It’s a stupid game played by university bureaucrats.
Nothing in the SPRC requires the university to have agreements in place to ensure assignment of any inventions other than subject inventions that the university already has acquired. Inventors start out with ownership of inventions, and nothing in the SPRC requires the university to disturb that ownership. If the university is unable to obtain an assignment agreement on terms it finds acceptable, then the university must turn the invention management discussion over to the federal agency and inventors.
If a university acquires title and elects to retain that title, then the picture changes to left side of the flow diagram and the relationship the university has with the funding agency.
After retaining title, the university (or any assignee of the university) must file patent applications and grant a non-exclusive license to the government. The SPRC provides that a university may assign its SPRC rights, provided that the assignee take on the SPRC responsibilities that the university has agreed to–that is, the nonprofit variation of the SPRC, with its paragraph (k). It is important to note that Bayh-Dole’s nonprofit assignment requirement is specific to inventions, not to patents. An invention may be assigned by granting an exclusive license in all substantial rights in the invention–such as an exclusive license to make, use, and sell. A university may retain title to a patent on a subject invention and still in the same exclusive patent license also convey ownership of the underlying subject invention on which the patent is based. I will let you consider the implications for a moment–and then help you relax with the knowledge that universities don’t comply with this SPRC obligation, even though they routinely grant exclusive patent licenses that assign the underlying invention. Federal agencies don’t care, either. So it’s all good. Bayh-Dole doesn’t operate once again as drafted. But what good is a law that everyone has to comply with? Isn’t it so much better to have a public policy with a convoluted goofball law that no one has to follow? And if that’s the case, why are we even bothering with Bayh-Dole flow of control? Oh, heck. On we go, as if driven by some weird compulsion for statutory logic.
Once the university and the inventors have worked out an assignment, the university can elect to retain title, file patent applications, grant the government a non-exclusive license, and set about using the patent system to promote the utilization of subject inventions, collaborate with industry, help small businesses, stimulate American manufacturing, promote free competition and prefer US manufacturing, prevent adverse effects on further research, protect the public public from non-use or misuse of inventions, and reduce administrative overhead–all those good things in 35 USC 200.
One more variation: the university goes to the trouble to obtain title to an invention, and then elects not to retain title. Hairball!
In this situation, the federal agency may use 35 USC 202(d) to allow inventors to “retain title.” Of course, it doesn’t make much sense, given that for an invention to be a subject invention, a contractor must own it, and if the university is the owner, then in not electing to retain title, the inventors still don’t own the invention and so have nothing to retain, having assigned away their ownership to the university. If the inventors are made contractors when a university complies with (f)(2), then of course they may “retain” their ownership if they elect to retain title and otherwise comply with the inventor patent rights clause. But if the inventors aren’t made contractors because the university does not comply with (f)(2) (or university lawyers assert that (f)(2) doesn’t involve making inventors into contractors), then their inventions aren’t subject inventions and so 35 USC 202(d) doesn’t even apply:
If a contractor does not elect to retain title to a subject invention in cases subject to this section, the Federal agency may consider and after consultation with the contractor grant requests for retention of rights by the inventor subject to the provisions of this Act and regulations promulgated hereunder.
It would appear that if an inventor contractor elects not to retain rights to a subject invention, then the federal agency may consult with the inventor contractor to allow the inventor contractor to retain rights anyway. Whoop. Let’s say this part is badly drafted. Perhaps what was meant was that if an institutional contractor does not elect to retain title to an invention that it has already acquired, then the institutional contractor may reassign the invention to the inventors, and they will be subject to 37 CFR 401.9–the “regulations promulgated hereunder” for inventors. But perhaps what was meant was some fantasy of bureaucratic patent attorneys given a free hand to draft their dream legislation without significant debate and this is the best they could do.
This then is an overview of the flow of invention management control under Bayh-Dole, a law applicable to federal agencies that then use a standard patent rights clause in funding agreements with universities. Universities, in turn, agree to the SPRC when they obtain funds on behalf of their investigators, and universities agree to require their employee investigators to have their own relationship with each funding agency within the scope of the funding agreement. In all of this, the (f)(2) agreement is pivotal, as it provides the link between the university and its employee inventors with regard to those employees’ obligations to the federal government.
It is a shame, then, that I still don’t know of a single university that implements the (f)(2) agreement as it is specified in the SPRC. When I asked about this agreement early in my work at the the University of Washington, the Director of Technology Transfer put it off, saying, “We won’t do it until an agency specifically requires us to comply.” Instead, universities substitute policy statements and assignment requirements directed at the *university* for the *university’s benefit*, not to protect the *government’s rights* as the SPRC directs. No wonder administrators and university attorneys are increasingly strident about assignment agreements. They are attempting to find a way to appear to comply with the SPRC without actually following the requirements that the SPRC lays out for them.
I wonder if it might not be a better relationship all the way around if university administrators were to follow the SPRC, pull out of their increasingly ineffective bureaucratic approach ownership and control, and return university research to the course that made it so highly attractive to the government and the public, and the envy of the rest of the world–that of providing the greatest freedom possible to university investigators to conduct research of their own choosing (“the free play of free intellects” to quote Vannevar Bush), to invent, and to act on opportunity, and where institutional resources are desired, to create an arrangement that works for the local circumstances as they present. It takes a pretty special group of researchers and administrators to be ready to take on the capabilities and mutual respect assumed by Bayh-Dole. An approach to federally funded research that opts for freedom makes just this assumption and stands to present to the world the most effective approach to advancing science, research, innovation, and community development.
******
[Here’s the original version of the article–things change, there’s new information, we learn, and like that. If folks had adopted this option to the federal government’s rights, then Bayh-Dole would have been *really* simple–any federal agency may permit a non-profit or small business to elect an option to receive invention deliverables that otherwise would be obligated by federal statute or agency regulation to the federal government, subject to compliance with the provisions of a standard patent rights clause and whatever.]
This would be a good time to map out how Bayh-Dole operates. I’ve put together 10 slides that show the flow of control. Perhaps this will help folks see what Bayh-Dole does require, and what it doesn’t.
First, let’s look at the law and its implementing regulations. The law is at 35 USC 200-212 and is part of patent law. Bayh-Dole is directed at federal agency contracting for research services with universities, nonprofits, and small companies (which I will shorten to universities). More particularly, Bayh-Dole makes “uniform” federal agency treatment of inventions as deliverables in funding agreements. As such, Bayh-Dole does not apply to universities, but rather to agencies.
Bayh-Dole authorizes the Secretary of Commerce to create implementing regulations. In particular, it requires Commerce to produce standard patent rights clauses (35 USC 206). The standard patent rights clauses are at 37 CFR 401.14. There are two, one for general use, section (a), and one for use under exceptional circumstances, and rights to inventions must be negotiated, section (b). [These clauses were revised by NIST into a single clause with multiple conditionals in May 2018].We will deal only with (a), which we will call the Standard Patent Rights Clause, or SPRC. The implementing regulations provide for variations from the SPRC to accommodate certain agency-specific matters, treaties, foreign work, weapons and naval propulsion, and exceptional circumstances in general. For this discussion, we will assume the SPRC is tailored only for agency-specific matters.
Agencies are required to use the SPRC in all funding agreements with universities. Funding agreements may be in the form of grants, which are regulated by Circular A-110 (now in 2 CFR 215 [now 2 CFR 200]) or contracts, which are regulated by the Federal Acquisition Regulations (FARs) and their defense equivalents, the DFARs. The SPRC is inserted into these regulations, along with federal agency variations, which the agencies then use to form funding agreements with universities.
As you can see, the flow of control in Bayh-Dole means that universities come to be affected by Bayh-Dole not directly by action of law but as terms and conditions in a funding agreement–essentially, a federal contract into which universities voluntarily enter.
[It’s actually a bit more convolutedly nuanced that this. Read this bit only at your own risk. Bayh-Dole applies only to subject inventions–patentable inventions that have been acquired by a party to a federal funding agreement and made in performance of work under that funding agreement. Thus, the funding agreement must come first–a university accepts the funding agreement and its SPRC–and then must acquire a patentable invention made under that funding agreement. Now Bayh-Dole kicks in as well. Here’s 35 USC 202(a):
Each nonprofit organization or small business firm may, within a reasonable time after disclosure as required by paragraph (c)(1) of this section, elect to retain title to any subject invention…. The rights of the nonprofit organization or small business firm shall be subject to the provisions of paragraph (c) of this section and the other provisions of this chapter.
This part of Bayh-Dole expressly grants universities the right to keep ownership of subject inventions–the patentable inventions that they have acquired and which were made under a funding agreement–but only subject to the SPRC and everything else in the “chapter”–35 USC 200-212. Thus, Bayh-Dole reads something like this: federal agencies must use a standard patent rights clause in all funding agreements for research or development work unless they can justify an alternative clause; a university when receiving federal support on behalf of a faculty investigator agrees to the SPRC; if subsequently the university acquires ownership from a faculty inventor or employee inventor of a patentable invention made under that funding agreement, Bayh-Dole and the SPRC both permit the university to choose to keep ownership of that invention, but only on the condition that the university complies with the SPRC and with everything else in Bayh-Dole.
The “shall be” used by Bayh-Dole here in 35 USC 202(a) is ambiguous. It appears to anticipate the inclusion of the requirement of paragraph (c)–the provisions for the standard patent rights clause–as well as to the rest of Bayh-Dole. But this “rest of” bit is very strange. If Bayh-Dole applies directly to universities, then the wording should be “The rights of the nonprofit organization . . . are subject to …” not “shall be” subject to….” The “shall be” must mean something along the lines of “shall be made subject to paragraph (c) and everything else by means of a standard patent rights clause.” That is, Bayh-Dole flows down through the standard patent rights clause, even though there’s this express statement in the statute that universities can keep what they gain ownership of, provided they comply with the certain bits of the SPRC.]
The SPRC is drawn almost entirely from 35 USC 200-204, but includes multiple provisions that are not in Bayh-Dole proper. One of these provisions is found in section (f)(2) of the SPRC. Section (f)(2) requires, among other things, that universities require of their employees (other than non-technical and clerical employees)–I will call them research employees here for short–a written agreement under which they agree to protect the government’s interest in inventions by a) disclosing subject inventions–ones already owned by the university–to personnel designated by the university, b) signing paperwork to permit patent applications to be filed on these inventions already owned by the university, and c) signing paperwork to establish the government’s rights in these inventions already owned by the university. For c), this could mean assigning ownership or granting a non-exclusive license to the government, or assigning title or granting a license to others that permits granting a non-exclusive license to the government. We will discuss this more in a bit. [In May 2018 NIST added an assignment requirement for subject inventions–it is a strange requirement–universities must require employees to assign all inventions that the universities have already acquired–that is, apparently, a university must obtain legal title for all subject inventions for which the university has acquired equitable title.]
For now, it is important to see that the SPRC is the contractual means by which agencies manage invention rights in funding agreements where a university has acquired ownership, and the (f)(2) agreement is the means by which agencies manage how a university deals with its research employees to ensure that the agencies obtain rights in inventions, as might be limited by the SPRC.
Here is the funding agreement situation, divided with the agency-university relationship on the left, and the agency-research employees relationship on the right:
The SPRC makes no assumptions about the relationship regarding inventions that might exist between the university and its research employees, other than that they are employees. It does not require assignment agreements between the university and research employees [other than in the case of equitable title, per NIST in May 2018], nor that the university have an patent policy, nor that the university even have a technology transfer function.
We will work through the flow of control by which a funding agreement is established, and then examine what happens when an invention is made. 
The start of a funding agreement is on the right side of the relationship, between research employees and agencies. A researcher, usually a member of the university’s faculty, decides to conduct a project and writes a proposal, often in response to a request for proposals issued by the agency. The proposal is submitted to the agency formally through the university. The agency then reviews the proposal and makes a decision to fund it. It is rare that a university refuses to submit a proposal from a principal investigator. It may happen, for instance, when only one proposal on a given grant program is allowed from any one university and multiple proposals are prepared. A proposal also may not be allowed to go forward if it involves classified research or other research for which the university lacks the resources, such as laboratories with security to handle dangerous substances.
In general, however, the university does not censor proposals, nor does it seek funding and then assign faculty researchers to conduct research not of their choosing. It is the principal investigators who decide to seek federal assistance. It is their deal, their choice. Furthermore, it is the investigators that the funding agency seeks to work with. It is the investigator’s qualifications and expertise that are reviewed as part of the grant application. It is a matter of special performance. If the principal investigator bails from the deal, the award if off unless a substitute investigator is found who is acceptable to the agency. The university’s role is subordinate to the process–to provide technical services to assist in the application process, and if an award is made, to manage the financial accounting and other regulatory compliance matters on behalf of the agency and the investigators.
The university, then, is something of a steward. Circular A-110 [now 2 CFR 200] makes this express. In section 37 [316], dealing with intangible property (including inventions), A-110 requires inventions (and other intangible assets) “to be held in trust by the recipient as trustee for the beneficiaries of the project or program”. The university, as a condition of the award, is to act as steward, not master. An understanding of this role shapes much of what follows for invention ownership and innovation practice. Disagreement over this role is the source of many of the problems that arise in the present, now conventional, approach university administrators take to technology transfer.
When a grant is awarded, the federal funding agreement is established. The agreement has a number of components. The primary one is the Statement of Work, which is generally the proposal that has been submitted by the principal investigator (PI), the faculty member responsible for the conduct of the work. The PI hires staff, decides on subcontracting, creates the work plans and designs experiments, assembles the tools, gathers and reviews data, directs reporting and publication, and spends the money.
In addition to the Statement of Work, the funding agreement consists of the appropriate set of regulations that govern the relationship. If a grant, then Circular A-110. If a contract, a set of clauses from the FARs or DFARs. In either case, these will include the SPRC with its (f)(2) agreement requirement.
When the university accepts the funding on behalf of the Principal Investigator for the proposed project, it also agrees to the SPRC. The university makes a voluntary decision to conduct its affairs in compliance with the SPRC. The university in accepting the SPRC therefore displaces any university policies or agreements that would conflict with the SPRC. Put another way, the university cannot accept the SPRC’s conditions while asserting practices for its research employees that are contrary to the SPRC. This is an important point. It is not the law that dictates the university’s behavior, but rather it is the university’s own affirmative acceptance that the SPRC is how rights will be managed. All the law does is assure the university that if it gains ownership, if it also complies with the bits of the SPRC that make compliance a condition of ownership, then it can keep that ownership.
As required by the SPRC’s section (f)(2), the university requires its research employees to make a written agreement to protect the government’s interest. In directing its research employees to make this agreement, the university releases them from any conflicting agreements or policies (including IP policies and conflict of interest policies) that it may otherwise require them to follow. The (f)(2) agreement permits research employees to sign away patent rights or grant licenses as may be required by the funding agency, or by another organization that gains standing under the SPRC to direct such activity.
Now the research gets going and an invention is made. In terms of the SPRC, this means that an invention is conceived or first actually reduced to practice. (It does not include constructively reduced to practice–such as by filing a patent application). What happens then?
Inventions that are made in performance of work under the funding agreement by university research employees are “subject inventions.” As confirmed by the (f)(2) agreement, research employees have agreed to disclose subject inventions to the personnel the university designates to receive such disclosures. These personnel could be technology managers in a tech transfer office, or legal counsel, or an external foundation or other invention management organization. (The personnel could also be principal investigators or department chairs, but I don’t know of any university that has done this.)
Per the SPRC, the university is responsible for reporting subject inventions to the funding agency. Then someone–the agency, the university, and the inventors–has to decide what to do with the invention. There are various options. 
The effect of Bayh-Dole now becomes evident. Rather than agencies insisting on a first right to obtain title to inventions made with federal support, Bayh-Dole requires the agencies to use the SPRC. The SPRC, in turn, allows the university to “elect to retain title” to subject inventions. This phrase has been the subject of a lot of recent dispute. “Subject invention” does not mean that the university must own by automatic vesting all inventions by its research employees made in the performance of work under a funding agreement, and it does not mean that by electing to retain title a university somehow thereby obtains title, nor does it mean that research employees have no other legal options but to assign to the university. Rather, “retain title” means that the university may intervene in the (f)(2) relationship it has been instrumental in establishing, and forestall the agency from requesting title from the university’s inventors. In essence, the university stands in for the agency to undertake the objectives of Bayh-Dole, the primary one of which is to use the patent system to promote the utilization of inventions made with federal support. [If a university inventor does not want to assign an invention to the federal government but does want a patent application to be filed (and share in possible royalties from the exploitation of a patent), then the university inventor may assign the invention to the university.]
By making “uniform” the federal agency procedures for inventions, Bayh-Dole relieves agencies of the administrative burden of case-by-case review of each request by a university to manage subject inventions. In place of such review, crafting of individual agreements, and continual oversight of progress, Bayh-Dole substitutes a set of standard requirements built into the SPRC to govern university behavior. [Of course, the Institutional Patent Agreement program created by the NIH and later adopted by the NSF also relieved those federal agencies of case-by-case determinations. It’s just that the IPA program was shut down as ineffective, and Bayh-Dole was introduced as a Senate bill the next year, claiming to be based on the IPA program. The same federal attorney was behind both the IPA program and Bayh-Dole.]
It is important to recognize that the university has no mandate nor obligation under the SPRC to retain title, to manage inventions, to file patent applications, to seek out licensees, to start companies, to attempt to develop commercial products, or to make money through licensing and infringement litigation. Any decisions that a university makes on these matters arise entirely from the university’s own motivations–that is, from the motivations of the individuals that the university assigns to the task of making such decisions, and to the arrangements they make, as stewards, with inventors and on behalf of the beneficiaries of the project in which the invention has been made.
If the university does not elect to retain title, here’s what happens:
If the university declines to retain title, then the relationship with the agency shifts to the right side of the diagram, the investigator side. University administration is then all but out of the picture. The agency deals directly with university inventors under the (f)(2) agreement [if the invention then is no longer a subject invention and agency regulations permit the agency to request title and the agency decides to request title.] If the federal agency requests title, the agency may file patent applications on behalf of the government, or it may simply let the invention enter the public domain. If the agency does not desire title, then under Bayh-Dole (35 USC 202 (d)), the agency may allow the inventors to retain title if they so request. [“Retain title” here makes it clear–as the Supreme Court ruled in Stanford v Roche–that inventors have title when a university elects not to retain title or does not ever acquire title.]
The procedures for allowing inventors to retain title are set forth in 37 CFR 401.9, which forms something of a mini-Bayh-Dole Act for inventors working directly with the government. Inventors are required to comply with a subset of the SPRC, acting in the place of the contractor as a substituted party: offering title back if they don’t continue with a patent application in the US or any foreign country, marking patents with a federal funding legend, reporting progress toward utilization, requiring substantial manufacturing in the United States for exclusive licenses, and accepting government march-in rights if the inventors or their licensees fail to make progress or are unable to meet certain other conditions.
Notably, there are no default restrictions on the inventors’ assignment of title, sharing of royalties, preference for small business, and other provisions that are required of contractors generally or universities in particular (though such restrictions could be added on a case-by-case basis if an agency desired to do so). [Most importantly of all, under 37 CFR 401.9, inventors are not required to use the patent system–they are not required to file patent applications. Their obligations to the federal government kick in only if they do file patent applications.]
If, however, the university elects to retain title, then the picture changes to left side of the flow diagram and the relationship the university has with the funding agency.
The primary matter for the university when it elects to retain title in a subject invention is that it can indeed obtain the rights that are necessary for it to meet the conditions of the SPRC. The conventional answer is that the university must obtain ownership of the invention. But the SPRC does not say that university ownership is required. The SPRC provides that the university retain title–that is, intervene so that the agency does not act on the (f)(2) agreement with inventors. After retaining title, the university is to file patent applications and grant or have granted a non-exclusive license to the government (SPRC section (b) does not specify who grants the license). Similarly, SPRC (f)(2) provides that the contractor will
execute or have executed … all instruments necessary to (i) establish or confirm the rights the Government has … and (ii) convey title to the Federal agency when requested ….
At each point the language avoids stipulating that the contractor must do these things. The contractor, if it cannot do these things, must be able to require others to do these things, and thus the language is sufficiently broad to allow a variety of situations, not restricted to university ownership.
One of these situations involves the use of an invention management organization, such as an affiliated research foundation or a professional invention management firm. The SPRC provides that a university may assign its SPRC rights to such an organization, provided that organization take on the SPRC responsibilities that the university has agreed to. This situation was very common when Bayh-Dole was passed (though now many universities have adopted an in-house technology transfer model and advocate against the use of agents). It is therefore understandable that Bayh-Dole would be drafted to permit the continued use of research foundations as agents.
The other situation is less discussed and involves allowing the inventors to maintain title to inventions but requiring sufficient cooperation between the university and inventors to meet the obligations of the SPRC–a non-exclusive license to the government, reporting on utilization, preference for small business and US manufacture, and the like. All of this can be accomplished without the university taking ownership of the invention or requiring ownership to pass to a designated agent. Nothing in the SPRC actually requires the university to compel inventors to assign their rights to subject inventions.
Furthermore, nothing in the SPRC requires the university to have agreements in place to ensure assignment. Even 37 CFR 401.9 does not indicate that the only way inventors can end up with ownership is by means of the 401.9 process, at the end of a line of choices by the university and agency. 401.9 merely is at the end of those choices. But inventors start out with ownership, and nothing in the SPRC requires the university to disturb that ownership when it elects to retain title. Assignment, if it is desired, could be entirely voluntary on the part of the inventors. If the university is unable to obtain an agreement that permits it to meet its SPRC obligations on terms it finds acceptable, then the university always has the option to decline to retain title, and turn the management discussion over to the agency and inventors.
Given all this, if a university does seek assignment, it has three routes to it. One involves having an agreement in place with inventors under which they agree to assign such inventions. This could be an employment agreement with an invention clause, a separate agreement that is required for employees to participate in sponsored research, or a policy requirement that is met when an invention is made using university facilities or other resources. Despite claims that such an agreement is required by Bayh-Dole, or essential to compliance with Bayh-Dole, an assignment agreement is not required. The university could leave it as a voluntary arrangement, as described above, negotiated on its own terms to the satisfaction of both the university and its inventors, dealing as equals and with nothing forcing the issue one way or another. If no arrangement can be made, then the university declines to retain title, and the agency takes the lead in discussions with the inventors. While this approach is rather attractive in organizations that are based on respect rather than distrust, love of process, and self-interest, there is a third approach available as well. This approach involves relying on the (f)(2) agreement, which like the language pertaining to government rights in the event of a contractor electing to retain title, is also drafted without specifying who is authorized to request paperwork to permit patent applications to be filed or establish the government’s rights in subject inventions.
[When a university elects to retain title before the university has acquired title, then there’s an argument that the university obtains standing under (f)(2) to request assignment, just as the agency would have, were the university to decline to retain title. There’s another argument that the university has no standing to elect to retain title to a subject invention until the university obtains ownership of that invention and so makes it a subject invention. And there’s yet a third argument that the fact of the university requiring the (f)(2) agreement makes inventors parties to the funding agreement through the patent rights clause and thus the inventions become subject inventions regardless of any further assignment activity.
This article is not the place to sort out these competing arguments–it’s enough to note that Bayh-Dole is drafted not to differentiate these differing accounts. The Supreme Court in Stanford v Roche was clear that nothing in Bayh-Dole disturbs an inventor’s initial ownership of an invention–nothing that gives a university ownership of such an invention, or any special privilege to acquire the invention. But the Supreme Court did not consider the effect of the SPRC–the issue was whether Bayh-Dole vested ownership of subject inventions with universities–not to provide a guide to the entire apparatus that implements Bayh-Dole. Hey–that’s more what Research Enterprise does.]
Thus, even in the absence of an employment agreement or other private invention assignment arrangement, the university has available the (f)(2) agreement on which it can rely to obtain assignment of subject inventions. The university may request assignment be made to itself, or to another organization that it designates (and which meets the requirements of the SPRC). In this way, the SPRC is self-implementing. It does not require any further policy, agreements, or administrative offices on the part of the university. The (f)(2) agreement is sufficient. This approach is remarkable in its simplicity.
Once the university and the inventors have worked out an arrangement that meets the requirements of the SPRC, the university can file patent applications, grant the government a non-exclusive license, and set about using the patent system to promote the utilization of subject inventions, collaborate with industry, help small businesses, stimulate American manufacturing, encourage competition and commercialization, prevent adverse effects on further research, protect the public public from non-use or misuse of inventions, and reduce administrative overhead–all those good things in 35 USC 200.
This then is an overview of the flow of invention management control under Bayh-Dole, a law applicable to federal agencies that then use a standard patent rights clause in funding agreements with universities. Universities, in turn, agree to the SPRC to obtain funds on behalf of their investigators, who have their own relationship with the funding agencies. The SPRC serves to protect the public, inventors, industry, and government from the choices and practices of university administrators in the absence of case-by-case agency review of every subject invention that university administrators or their patent agents might seek to manage. In all of this, the (f)(2) agreement is pivotal, as it provides the link between the university and its research employees with regard to those employees’ obligations to the government.
It is a shame, then, that I don’t know of a single university that implements the (f)(2) agreement as it is specified in the SPRC. When I asked about this agreement early in my work at the the University of Washington, the Director of Technology Transfer put it off, saying, “We won’t do it until an agency specifically requires us to comply.” Instead, universities substitute policy statements and assignment requirements directed at the *university* for the *university’s benefit*, not to protect the *government’s rights* as the SPRC directs. No wonder administrators and university attorneys are increasingly strident about assignment agreements. They are attempting to find a way to comply with the SPRC without actually following the requirements that the SPRC lays out for them.
I wonder if it might not be a better relationship all the way around if university administrators were to follow the SPRC, pull out of their increasingly corporate model of ownership and control, and return university research to the course that made it so highly attractive to the government and the public, and the envy of the rest of the world–that of providing the greatest freedom possible to university investigators to invent, act on opportunity, and where institutional resources are desired, to create an arrangement that works for the local circumstances as they present. It takes a pretty special group of researchers and administrators to be ready to take on the capabilities and mutual respect assumed by Bayh-Dole. An approach to federally funded research that opts for freedom makes just this assumption and stands to present to the world the most effective approach to advancing science, research, innovation, and community development.