Let’s work through what NIST has done with its new rule on assignment of subject inventions by written agreement.
The Supreme Court in Stanford v Roche (2011) ruled that Bayh-Dole’s contracting provisions apply only to subject inventions, and that subject inventions are inventions owned by the contractor. The Court made it clear that Bayh-Dole did not vest ownership of inventions with the contractor, did not compel inventors to assign inventions to the contractor, did not prevent inventors from assigning other than to the contractor.
The Court was adamant that Bayh-Dole’s contracting provisions dealt only with the priority of ownership of an invention made in a research project with federal support only after a contractor had acquired title to such invention. Here, read for yourself. The Court starts with a reference to section 210, which asserts Bayh-Dole’s precedence over all other statutes with regard to subject invention but for Stevenson-Wydler (which concerns federal laboratories):
But because the Bayh-Dole Act, including §210(a), applies only to “subject inventions” —“inventions of the contractor”—it does not displace an inventor’s antecedent title to his invention.
There–Bayh-Dole applies only to inventions a contractor has come to own, not to inventions that an inventor has invented. Remember, Stanford’s argument was that Bayh-Dole vests ownership with the contractor. The Supreme Court rejected that argument.
Only when an invention belongs to the contractor does the Bayh-Dole Act come into play.
Bayh-Dole provides no authority, no mechanism, no procedure for compelling an inventor to give up rights to an invention. And Bayh-Dole claims to be the sole federal authority with regard to rights to subject inventions.
The Act’s disposition of rights—like much of the rest of the Bayh-Dole Act—serves to clarify the order of priority of rights between the Federal Government and a federal contractor in a federally funded invention that already belongs to the contractor. Nothing more.
That “nothing more,” so Poe-ian, like a raven, hammers home the point. So a subject invention is an invention that a contractor has come to own, and Bayh-Dole cannot possibly be involved in requiring an inventor to give up ownership of an invention just because it has been made in a project receiving federal support. Instead, Bayh-Dole’s contracting provisions apply only to those inventions made in a project receiving federal support that a contractor acquires. If the contractor does not acquire such an invention, then–and here’s the big surprise from Stanford v Roche–executive branch patent policy and all those pre-empted laws still apply. The federal government may assert ownership of any such unassigned invention, according to the requirements of executive branch regulations and any applicable statutes.
Now you might see the reasoning for the (f)(2) written agreement. It was intended to cover the case in which an invention is not assigned to the contractor. In such a case, so the idea went, the government would have to deal with inventors directly. In the past, these inventors were those who could be made parties to the funding agreement by executive branch policy or whose inventions were covered by a statute requiring federal acquisition of their inventions. If an inventor wanted to retain rights, the inventor would have to request a determination of rights from the federal agency–but would be obligated one way or another to establish the government’s rights by assignment or by granting a non-exclusive license.
It’s just that the (f)(2) written agreement, as drafted, is specific to subject inventions–ones owned by a contractor, a party to the funding agreement. That agreement would make no sense at all unless we recognize that the written agreement itself makes inventors parties to the funding agreement–they become “contractors” within the definitions provided by Bayh-Dole, their inventions are then subject inventions (but owned by the inventor-contractors not by the employer-contractors) and thus inventors and their inventions then also come within the scope of Bayh-Dole’s contracting provisions. The implementing regulations reflect this fact with a patent rights clause at 37 CFR 401.9 that’s specific to inventors who own their subject inventions–they are to be treated as small business contractors, but with fewer obligations than small business contractors. This outcome is not obvious, but it is clear once you see it, and it makes good sense of what otherwise is nonsense.
There is one condition that may alter this outcome–that inventors become contractors and their inventions are subject to an inventor patent rights clause. That condition involves an employer’s equitable claim to title to an invention–a claim that has not been perfected by assignment but exists in the nature of the circumstances surrounding the invention. And it is to this claim to equitable title that the new assignment regulation applies. If you want technical details–can we add yet another angel to the pinhead?–this is for you.
Look at NIST’s new language in the (f)(2) written agreement requirement. Keep in mind that nothing in (f)(2) is expressly required by Bayh-Dole:
The contractor agrees to require, by written agreement, its employees, other than clerical and nontechnical employees . . . to assign to the contractor the entire right, title and interest in and to each subject invention made under contract . . .
Whatever authority there is for NIST to require contractors to require inventors to assign inventions “made under contract” to the contractor, it does not arise in Bayh-Dole. But NIST does not require contractors to require inventors to assign just any inventions–NIST requires contractors to require employees to agree to assign subject inventions. That is, contractors must require inventors to assign inventions that the contractor already owns and these inventions must have been made “under contract”–meaning under a contract that meets the definition of a funding agreement (and not a private contract such as patent agreement or employment agreement between the contractor-employer and the inventor-employee; and not a grant and not a cooperative agreement).
If we rule out the idea that the NIST assignment requirement is nonsense or is redundant or reaches to inventions the Supreme Court ruled it could not reach to–“once we have eliminated the impossible, whatever remains, no matter how improbable, must be the truth” (to quote Sherlock Holmes on the matter). Improbable as it may seem, the new NIST assignment requirement states, “If a contractor has equitable ownership of an invention made with federal support under a contract–and therefore technically the invention is owned by the contractor and thus a subject invention–then the contractor must have required its employee-inventors to make a written agreement at the time the contract was awarded to assign formally that invention to the contractor.” In other words, if a contractor has an equitable claim of ownership, it must perfect that claim with an assignment, and to ensure that outcome, it must for each federal contract require those specified employees involved to make a written agreement to assign such inventions where the contractor does have an equitable claim to ownership.
That is, the contractor must establish some agreement or fact circumstances outside the scope of the federal funding agreement–one not preempted by the written agreement requirement itself in the patent rights clause–under which the contractor has an equitable ownership of inventions made within the scope of that funding agreement. If a contractor comes to have equitable ownership, then the contractor must perfect that ownership interest by obtaining an inventor’s assignment.
That’s still a strangely improbable requirement, but at least it makes sense of NIST’s requirement. Even so, this new requirement is ham-handed. Why should NIST force a contractor to take assignment of an invention that a contractor may not want to own, even if the contractor might have an equitable claim to own the invention? What purpose is served by the federal government dictating how contractors must engage their employees with regard to inventions? None, really. It’s a huge intrusion into a company’s relationship with its employees and its decisions about what inventions it will own, and on what basis.
There’s additional support for the equitable ownership idea in the (f)(2) provision. The written agreement requirement is restricted to employees who are not “clerical or non-technical” employees. Why exclude clerical or non-technical employees from the requirement? Clearly, in a university environment, a highly skilled worker (a post doc, even) could be hired to perform a clerical or non-technical job. A doctoral student could provide services as a graduate staff assistant or even as an hourly worker, perhaps washing glassware or managing stocks of reagents or tending to laboratory animals. Such a worker could easily also contribute to technical discussions and participate in inventive work–even though as an employee the worker is in a clerical or non-technical position. Why, then, does the (f)(2) written agreement requirement exclude such workers?
Here’s a reason. If the federal government were using a patent rights clause to claim inventions as deliverables directly from individuals (even as employees), then the federal government would have a claim to equitable title based on an agreement to comply with federal regulations only from those participants who were hired to conduct research, to experiment, to invent–those participants who might be reasonably considered to have work assignments that might result in inventions. Someone washing glassware certainly may assist in the overall project–and very clean glassware may even be important to that work–but there’s no basis for an equitable claim based on providing financial support to a project to such a person’s insights or inventions. In a similar manner, there would be no basis to claim the inventive insights of someone who was not an employee–someone working as a visitor or volunteer or a participant in a conference discussion.
We might add by way of observation–why should a university that puts *nothing* into a federally funded project that it isn’t compensated for have any *better* claim than the federal government to the inventive work of clerical or nontechnical employees–even inventive work that’s relevant to a federally supported project? If the federal government has no claim on the inventions of such participants, neither do the university administrations that host such work and do nothing more than handle federal money on behalf of the participants. There is no claim whatsoever available under Bayh-Dole.
We might say more generally that Bayh-Dole–via the patent rights clause–excludes institutional contractors from having any claim of ownership of inventions made in projects receiving federal support based on the circumstances of that federal funding–participation, payment, use of university resources, area of expertise, nature of the results, potential for “commercialization.” If an institutional contractor is to have an equitable claim, that claim will have to come about because the institutional contractor had the right to assign and direct a given set of employees to invent or experiment or made their continued employment expressly conditioned on their agreement to assign inventions within the scope of the institution’s business. For companies, such arrangements are common and reasonable. For universities, and especially for faculty and students, just the opposite.
To put it coarsely: if a university cannot establish an equitable claim of ownership in an invention made with federal support, then it cannot require the inventor to assign that invention to the university. If, however, the university can establish an equitable claim, then–now, according to NIST–the university must require the inventor to assign that invention to the university. For a university, such an equitable claim would come about because a faculty or student inventor has, for a specific federal contract, voluntarily agreed to assign the invention to the university in exchange for something of value–not employment or continued employment, not release to work on the federally supported project, not a general statement of patent policy that conflicts with the policy on freedom of research and freedom of publication. There must be something else, such as an agreement to augment federal research support with special university support not reimbursed by the federal government.
Let’s put it another way. In German invention law, there’s a distinction between “service” inventions and “free” inventions (summary here). Service inventions are ones made by an employee in work directed by an employer or arising from the experience and activities of the employer. Free inventions are everything else. Bayh-Dole deals with a middle ground–work that may be specified by the government (as in a contract) or specified by a faculty or similar investigator (as in a nonprofit grant), and who may be assigned and directed by the employer contractor (as in a company working under a contract) or may be released to conduct without employer control (as university faculty and students working in a university with federal grant support).
These are very different situations, all bundled up in the same set of patent rights clause generalizations to give the appearance of a single arbitrary federal policy. Under Bayh-Dole, almost all inventions made by an employee in a company (but for clerical and nontechnical employees), will be service inventions if in line with the company’s business. And almost all inventions made by university faculty and students will be free inventions unless for some reason certain faculty and students have given up their freedoms and agreed to accept university control over their work and their publications. The new NIST rule on assignment asserts that if an employee creates a service invention–an invention to which the employer has an equitable claim–then the employer has an obligation to obtain assignment of the invention and to that end must make employees make a written agreement in which they agree to assign such inventions to the employer and not to the federal government.
An employer cannot assert any other right by written agreement to ownership in any other invention made with federal support. The written agreement requirement preempts an employer from simultaneously complying with the written agreement requirement and also requiring something else that conflicts with the required written agreement. If inventors are delegated to sign papers to establish the government’s rights, then other than in the situation that an employer has a claim to equitable title to an invention made under a federal contract, an employer can make no claim to ownership in any subject invention owned by a specified employee-inventor. An employer cannot require two conflicting obligations from employees. The federal obligation takes precedence. Anything else is void.
Think about it. If the scope of federal interest in an invention made in a project with federal support does not extend to clerical and nontechnical employees, and does not extend to volunteers, collaborators, and visitors, then why should the university assert a claim on that same inventive work that’s more expansive than the federal government’s claim? Or, perhaps more to the point–where is the basis for equitable title in such inventions? Or, put another way, why should a university assert ownership of inventions for which it has no equitable claim to title? Bluntly, why should a university assert ownership of inventions to which it has no right? More bluntly, how can a university assert ownership of inventions to which Bayh-Dole’s standard nonprofit patent rights clause obligates the university to have no such special right? It is not permitted. But of course university administrators don’t let mere laws stop them.