[Updated May 2018 to deal with the NIST screwballedness.]
At the heart of the Bayh-Dole Act is the disposition of ownership in inventions made with federal funding at universities. That disposition is intended to provide benefits to the public through the practical application of those inventions. The question for policy and practice, then, is just how this disposition is to come about, and what sorts of behaviors will bring about benefits to the public. For that, we have to look for the soul of Bayh-Dole.
To understand the disposition of invention ownership under Bayh-Dole, one has to see how the Act is structured. The Act applies to federal agencies, instructing them with regard to how they may claim ownership of inventions made with federal support, and when they must allow non-federal parties to claim ownership of such inventions, obtaining instead for the government a non-exclusive license for government purposes. To be clear, Bayh-Dole the law does not apply to universities, nor to faculty inventors (nor to any other inventors).
To show how this comes about, one has to read both the law and the implementing regulations. Bayh-Dole is a law about how the federal government enters into funding agreements that have a patent rights clause. That is, Bayh-Dole is about the structure of federal agreements. The Act (35 USC 206) directs the Secretary of Commerce (originally, the Office of Federal Procurement Policy) to create regulations and standard funding agreements provisions to implement the objectives of the Act. These are set forth in 37 CFR Part 401, with the standard patent rights clauses at 37 CFR 401.14. There are two, one at section (a) used for general situations, and one at (b) for certain situations involving contractors providing services at federal laboratories. [Conflated by NIST’s May 2018 rule change into a single omnibus clause with conditionals–so that the (b) clause, which set out a standard exceptional circumstance, is now just a buried conditional in the standard patent rights clause.] The one at (a) is the generally applicable standard clause [and now cited with the (a) removed, consistent with NIST’s screwball changes]. There is also what amounts to a third patent rights clause, at 37 CFR 401.9, for use when inventors are allowed by an agency to retain title to the inventions they have made.
Agencies may also tailor these clauses in various ways, to include specifics about where to report inventions and the like and to account for various special circumstances. If an agency declares “exceptional circumstances” under 37 CFR 401.3, it may also change the disposition of ownership of inventions. Thus, for any specific research activity involving federal funds, the controlling instrument is the funding agreement with a specific patent rights clause. The Act provides the authority for the construction of such patent rights clauses, and it is by means of federal agreements that the requirements of these clauses are transferred to federal contractors such as universities.
None of this apparatus would be necessary if Bayh-Dole simply decreed where ownership of federally supported inventions was to end up. But Bayh-Dole does not do this. It is a law pertaining to the formation of federal funding agreements for research, on offer by federal agencies.
If we follow ownership matters from Bayh-Dole to the standard patent rights clause at 37 CFR 401.14, we find the key element that makes the whole implementation of Bayh-Dole work, and that is a provision at section (f)(2) in a section titled “Contractor Action to Protect the Government’s Interest”:
(2) The contractor agrees to require, by written agreement, its employees, other than clerical and nontechnical employees, to disclose promptly in writing to personnel identified as responsible for the administration of patent matters and in a format suggested by the contractor each subject invention made under contract in order that the contractor can comply with the disclosure provisions of paragraph (c), above, and to execute all papers necessary to file patent applications on subject inventions and to establish the government’s rights in the subject inventions.
[Here is NIST’s revised first sentence of the (f)(2) paragraph, my emphasis:
(2) The contractor agrees to require, by written agreement, its employees, other than clerical and nontechnical employees, to disclose promptly in writing to personnel identified as responsible for the administration of patent matters and in a format suggested by the contractor each subject invention made under contract in order that the contractor can comply with the disclosure provisions of paragraph (c) of this clause, to assign to the contractor the entire right, title and interest in and to each subject invention made under contract, and to execute all papers necessary to file patent applications on subject inventions and to establish the government’s rights in the subject inventions.
A subject invention is a patentable inventions made under the funding agreement that the contractor *already owns*. Thus, the assignment requirement here is mysterious. NIST’s chief counsel has made public his belief that Bayh-Dole is a form of vesting statute, apparently ignoring the Supreme Court in Stanford v Roche. Thus, the rule change here appears to be so that Bayh-Dole’s implementation better conforms to what NIST’s chief counsel thinks the law should be rather than what the Supreme Court made clear the law is. Anyway, the short bit here is that the new assignment requirement applies only to inventions that a contractor already owns. If that contractor is the inventor by the operation of the (f)(2) requirement, then the inventor is subject to the inventor patent rights clause at 37 CFR 401.9. If the initial contractor claims the assignment provision requires the inventor to give up ownership, then Bayh-Dole is once again a vesting statute, which the Supreme Court expressly rejected. Congress would have to amend the law to make vesting express–and would have to navigate the Constitution to do it, since the Constitution does not give the federal government to reserve rights in inventions other than for inventors–not for employers, and certainly not for any old hack bureaucrat that happens to handle federal money designated to support research proposed by university faculty, for which they must obtain a release from their university duties to conduct.
Short form: if vesting, it is not legal, per the Supreme Court. If not vesting, then it requires assignment of inventions that the contractor already owns. WTF?
Further, the NIST assignment requirement makes nonsense of the last part of the written agreement, which requires inventors to establish the government’s rights. They won’t have any rights to assign to or license to the government if they are required by the same written agreement to assign those rights (entire right, title, and interest) to their employer. It just doesn’t work. The assignment requirement, therefore, cannot be directed at rights an inventor has (and is expected to have in the future) at the time that a federal award is made and the (f)(2) written agreement requirement becomes operative for a given funding agreement.
What follows is a discussion of the (f)(2) clause but for the NIST addition of an assignment requirement.]
The first sentence of (f)(2) is quoted above. It requires a contractor to require certain employees to make a written agreement that commits them to three things [now four]: (1) to disclose inventions to the contractor; (2) to sign documents to file patent applications; and (3) to sign documents to “establish the government’s rights” in inventions; [and (4) to assign subject inventions to the contractor]. This agreement, which I call the (f)(2) agreement, is at the heart of the implementation of Bayh-Dole. Understand how (f)(2) works, and you understand a lot about Bayh-Dole, and why so much of what passes for university claims about Bayh-Dole and invention ownership is plain wrong.
The (f)(2) agreement is unique to the standard patent rights clause. It does not show up in the Bayh-Dole Act proper. (f)(2) is, however, the enabler of much of what Bayh-Dole seeks to accomplish, and is, further, the core element of the implementation of the standard patent rights clause, which is a procurement clause for invention rights. Since ownership of inventions starts, under federal law, with inventors, that is also the place that federal contracts have to start. The fact that (f)(2) is not in Bayh-Dole means that a lot of commentators and pundits on Bayh-Dole have simply missed it, and therefore do not know how the federal agreements operate with regard to inventions. Reading the contract is Contracts 101. Strangely, many university administrators have not done so with this cornerstone of their practice.
It is important to consider the structure of the (f)(2) clause. Eliminating various qualifying phrases and inserting some markers makes this structure clear:
The contractor agrees to require, by written agreement, its employees,
(i) to disclose promptly in order that the contractor can comply with the disclosure provisions of paragraph (c), above, and
(ii) to execute all papers necessary to file patent applications on subject inventions and
(iii) to establish the government’s rights in the subject inventions
[(iv) to assign subject inventions to the contractor].
There are three [four] clauses each started by infinitives, all in parallel that set out what the written agreement is to require of these employees. The first clause establishes the disclosure obligation, and the additional text lays out who the disclosure is to be made to, what is to be covered, and the format for the disclosure. The first clause is the one that is concerned with the Contractor’s (or employer’s) compliance, and in particular with the obligation to report inventions to the funding agency (the subject of paragraph c). The next two clauses are independent of this concern.
The second clause pertains to the inventor’s assistance in filing patent obligations, most importantly filing the declaration and oath that the employee is indeed a true inventor of what has been invented. Note that this clause does not specify who is filing the patent application–it could be the Contractor, a Contractor’s assignee, an assignee of the inventor, or the government. The first objective of Bayh-Dole is for the patent system to be used to promote the use of inventions made with federal support (see 35 USC 200)–the purpose of this second clause is to ensure that whomever is seeking a patent on such an invention has the cooperation of the inventor(s).
The third clause has to do with the government’s rights–either obtaining title or a license to the invention. This third clause is important because it makes clear that the federal government expects inventors to hold rights and be authorized to transact those rights, either by assignment or license. How could an inventor agree to establish the government’s rights in inventions if the inventor never had those rights to begin with, following those who argued that Bayh-Dole vests title outright with universities. Clearly, Bayh-Dole does not, as the Supreme Court made absolutely clear. Certainly those drafting the standard patent rights clause at the Department of Commerce did not believe Bayh-Dole vested title with universities.
[The fourth clause I have dealt with above.]
The form of this third clause is important for three additional reasons. Consider the situation in which a contractor such as a university fails in its obligations under its federal funding agreement: the contractor might fail to inform employees about timely disclosure, or may fail to notify the government regarding disclosed inventions. In such instances, the standard patent rights clause considers how the government obtains the rights it has bargained for, when those rights start with the inventors and remain with the inventors. For the government to have those rights, by means of a contracting instrument, it must have an agreement that includes those inventors. The inventors, therefore, as a default, must have standing to sign documents that assign title, or grant a license, to the government. Again, this is standard contracting: cover each contingency.
The government obtains the relationship it desires with inventors by requiring Contractors, as a condition of the funding, to require their own employees to make the written agreement. Doing so authorizes the employees to act with regard to the government’s interest without any fear that in doing so the Contractor, as their employer, might object. The effect of the funding agreement requirement on the Contractor is to clear away any Contractor claims on employees that might prevent those that invent from acting to protect the Government’s interest. That is, of course, what section (f) is all about. The written agreement that is made, therefore, supersedes any other claim that the Contractor might make on inventor employees that might conflict with the (f)(2) requirements.
From this it must become entirely clear that the (f)(2) agreement has nothing to do with employment agreements or patent agreements that concern an employer’s interest in an employee’s inventions. Such agreements certainly can be made between employers and employees, as they certainly are in many businesses that could operate with a standard patent rights clause. The situation with universities, however, was much different when Bayh-Dole was passed from what it is today. When Bayh-Dole was passed, most universities did not have internal patent licensing operations, and many did not claim ownership of inventions made by faculty inventors, and certainly did not think to claim ownership of inventions made with federal funding, because a number of federal agencies routinely claimed ownership of such inventions anyway.
Nothing in Bayh-Dole interferes with employer-employee relationships. There is no requirement that universities adopt patent policies, or change employment agreements to demand university ownership of inventions made with federal funds, or inform universities that they will by law own such inventions and must comply with the requirements that come with such ownership. None of that. And for good reason. Bayh-Dole is about how the federal government procures inventions made by university faculty (and others) under funding agreements. It has no reason to be interested in who ends up with title to inventions, other than when the government has a right to require assignment of that title, and when anyone else holds title, to require compliance with the conditions that attend that title.
The written agreement required in (f)(2) is not an employee patent agreement. A university requiring an employee patent agreement is not doing a thing to comply with Bayh-Dole (because Bayh-Dole does not apply to universities but rather to federal agencies!) and as well nothing to comply with the funding agreement or its standard patent rights clause of which (f)(2) is an integral part. Whatever a university wants to do with regard to patent policy, employment agreements, and ownership of inventions, it is not complying with (f)(2) in doing what it does. Thus, the current craze for present assignments as a condition of employment, or use of resources, does not have anything to do with compliance with federal funding agreements. It just doesn’t. It can’t.
Except, of course, that if a university claims that a patent agreement may substitute for the (f)(2) agreement, it is not only wrong, and misleading, but it is out of compliance, and by failing to permit faculty to make the written agreement the faculty are to be required to make, the university in effect is requiring faculty, by contract, to be complicit in beaching their federal obligations. One might think, such a contract would be void. Federal contracts, indeed, take precedence over private contracts enforced under state laws. Thus, a non-compliant university, aiming to force a policy or employment based patent agreement on those performing federal research, in essence *voids* that very agreement through its actions. It is important, therefore, that whatever else a university decides to do, it also complies with the (f)(2) requirement.
The structure of the (f)(2) agreement is distinctive for another reason as well. Consider the definition of “Funding Agreement” and “Contractor” in 37 CFR 401.2:
(a) The term funding agreement means any contract, grant, or cooperative agreement entered into between any Federal agency, other than the Tennessee Valley Authority, and any contractor for the performance of experimental, developmental, or research work funded in whole or in part by the Federal government. This term also includes any assignment, substitution of parties, or subcontract of any type entered into for the performance of experimental, developmental, or research work under a funding agreement as defined in the first sentence of this paragraph.
(b) The term contractor means any person, small business firm or nonprofit organization which is a party to a funding agreement.
A funding agreement is any of a set of relationships between a federal agency and a contractor for research work funded by the government. Notice, however, the second sentence of the definition, which expands this definition to include any assignment, substitution of parties, or subcontract. Assignment here refers not to title to invention, but to the Contractor’s side of a funding agreement. A funding agreement includes as well actions that include other parties in the work. Now look at the definition of contractor. It also allows for the expansion of participation in a funding agreement. The definition could have been written “any person . . . which is *the* party to a funding agreement.” That would make clear that the government anticipated that each funding agreement had one contractor. But that’s not how the text is written. Instead we have “a party,” allowing that there could be more than one such party to a funding agreement.
The (f)(2) agreement, part of the standard patent rights clause of a federal funding agreement, is structured as a requirement a university accepts, under which the university is to require the university’s employees to make. Bayh-Dole imposes nothing on the employees. The standard patent rights clause imposes nothing on the employees. There is no, shall we say, statutory action taken here. It is the university, as contractor, that takes the action to require the (f)(2) agreement in writing. By doing so, the university makes each employee into a conditional party to the funding agreement, by assigning certain obligations under the funding agreement to those employees. The employees, therefore, become conditional contractors as well. The contractor is not conceived as a corporate entity, a “person,” but rather as a collective comprised of the employer and the employees who might invent while working within the scope of the funding agreement. If an employee invents with federal funding, then the employee’s (f)(2) compliant written agreement comes into effect, and becomes part of the patent rights clause that governs the disposition of ownership of inventions under the funding agreement.
Thus, when reading the implementing regulations for Bayh-Dole, it is not appropriate to assume that “contractor” means only “the university administration.” Certainly the university administration is involved, but that is not the only contractor that may participate as a party to the funding agreement. If no inventions are made, then the contractor is but the university. But if an invention is made, then the inventors are also contractors. Again, university administrators routinely ignore this result, but it is at the heart of how Bayh-Dole operates. It should be what administrators take pride in knowing through and through.
Two points will show why contractor includes inventors. First, there is the problem of the definition of “subject invention”:
2) Subject invention means any invention of the contractor conceived or first actually reduced to practice in the performance of work under this contract, provided that in the case of a variety of plant, the date of determination (as defined in section 41(d) of the Plant Variety Protection Act, 7 U.S.C. 2401(d)) must also occur during the period of contract performance.
I have added some boldface. Each element of the definition is important. A subject invention is an invention made under a specific contract, but in addition must be an invention “of the contractor.” The Stanford v Roche case hinged on the little word “of”–which as Sean O’Connor pointed out in his amicus brief to the Supreme Court for the AIPLA, means “ownership” not “agency.” That is, the law says “of” and not “by” or “for.” For an invention to be “subject” to the standard patent rights clause, it has to be owned by a Contractor–that is, in the case of a university, the university (as a legal body) or any of the designated employees who have invented (and not others, who are not made conditional contractors, such as clerical and technical workers (expressly excluded) and volunteers (who are not employees and thus also excluded) nor informal collaborators (such as faculty at other universities, or industry scientists and engineers). The problem for university officials who wish to substitute an employment-based agreement directed at university ownership for the (f)(2) agreement is how can an employee inventor be required to disclose all *subject inventions* when those inventions are not owned by the university when they are made?
Again, back to the text of (f)(2):
. . . to disclose promptly in writing to personnel identified as responsible for the administration of patent matters and in a format suggested by the contractor each subject invention made under contract . . .
One response here is that the only way that such a thing can happen, if an invention must be owned by the contractor to meet the definition of an “invention of the contractor” is if the invention is instantly assigned to the contractor the moment it is made. “Aha!” cries a pod of university administrators, “That’s the reason for a present assignment. Before, we thought that Bayh-Dole simply gave us ownership, but the Supreme Court in Stanford v Roche ruined our picnic and now to save Bayh-Dole we have to fix this problem in the implementing regulations by imposing present assignment obligations on all our employees.” But of course, this is not the only way, and isn’t in fact a way at all, because this approach does not comply with the (f)(2) requirement, but rather aims to circumvent it–to prevent it from operating, so that inventors never have the right to establish the Government’s rights in inventions, even though the university is to authorize its inventors to have just such a right.
The other way, however, makes perfect sense. Inventors, when they invent, become contractors, and of course, as inventors under federal law, they own their inventions. Thus, such inventions become subject inventions–inventions of the “contractor” but from our work with the definition of a contractor, we know that there can be any number of them as parties to a given funding agreement. Without taking into account the definitions of funding agreement and contractor, and without recognizing the structure of the (f)(2) requirement, one would have little chance of understanding how the standard patent rights clause works. But it does work nicely. It all works out. It allows for a variety of arrangements between employer and employees, and between inventors and assignees, while protecting the Government’s interest in inventions made with federal support–subject inventions.
A second point can be made with regard to the inventor’s standing as a contractor. Consider the language of 37 CFR 401.9, the mini patent rights clause that comes into effect when an inventor chooses not to assign invention title to an invention management agent pre-approved under the standard patent rights clause, and the funding agency agrees to allow the inventor to retain title, which the inventor has had all along. Under 37 CFR 401.9:
Agencies which allow an employee/inventor of the contractor to retain rights to a subject invention made under a funding agreement with a small business firm or nonprofit organization contractor, as authorized by 35 U.S.C. 202(d), will impose upon the inventor at least those conditions that would apply to a small business firm contractor under paragraphs (d)(1) and (3); (f)(4); (h); (i); and (j) of the clause at § 401.14(a).
The effect of this provision is that agencies construct for inventors that retain title a patent rights clause a subset of the conditions that apply to small business contractors. The inventors are to be treated as contractors, but with not nearly the obligations of nonprofits, and not even, necessarily, the obligations of small businesses. Inventors, as contractors, get the most remarkably friendly treatment of all. One might say, the Government loves the inventors best of all, or to flip it around, worries most of all about nonprofit and university behaviors. Bayh-Dole is represented by many universities as being made to make them happy, and rich, but really, the text of 37 CFR Part 401 teaches just the opposite.
By means of the (f)(2) written agreement, the Government obtains a direct relationship with the individuals who are working within the scope of the federal agreement. By doing so, the Government covers the default condition under which an inventor does not assign title and the inventor’s employer fails to comply with the standard patent rights clause, or chooses not to accept title to a subject invention. Nothing in any of this requires a university to impose a demand for assignment on all employees, and certainly not a present assignment. For all that, the obligation for employees to make the written agreement at (f)(2) only comes into existence when (1) the university accepts a federal funding agreement; and (2) the university then complies with the standard patent rights clause and requires each employee to make the required written agreement, thus becoming conditional parties to the funding agreement for the purposes of the patent rights clause, and therefore, when they invent, becoming contractors within the definition of contractor in the standard patent rights clause.
Thus, there is absolutely no need for a university to demand ownership of inventions “upfront” before any federal grant has been awarded or accepted, and even then, there is nothing in the federal requirements that requires university ownership. Whatever the university demand for ownership might be, it does not come from Bayh-Dole, or from the enabling regulations, or from the funding agreement, or from the standard patent rights clause, or from the (f)(2) agreement within that clause. Any representation by a university that it does is simply wrong. More than that, it is uninformed, negligent, and after anyone reads this discussion, downright deceptive, if not evil, at least in the Google mantra sense.
Faculty deserve a clear exposition of their obligations under federal funding agreements. Bayh-Dole, too, deserves better treatment than it has gotten at the hands of university patent administrators and the law firms that aim to make their livelihoods doing work for university patent administrators. Reviled for blocking innovation and public access to inventions, Bayh-Dole in fact does nothing more than give inventors the opportunity to make a choice of patent management agent, and federal agencies relief from having to claim every invention or case-by-case review and approve inventors’ choices for assigning title. Such an approach leaves a wonderful amount of freedom to explore how to use the patent system to promote use of inventions–by standards, by open access, by pooling and cross-licensing, by licensing the practice community freely while reserving certain commercialization rights, along with exclusive licenses to support substantial investment where no-one could practice or realize the benefits of the invention without such investment.
In addition, Bayh-Dole allows for diversity in inventor choice of invention management agent, for universities to take no interest in such inventions, to have no policy regarding such inventions, and to have no employment agreement or patent agreement that addresses such inventions, other than the conditional delegation to inventors via the (f)(2) obligation of their right and obligation, in the context of their employment, to protect the Government’s interest. If inventors are delegated this right, then they not only own their inventions–as the Supreme Court affirmed–but they also are protected from an employer’s claim to ownership of those inventions as a condition of the federal funding agreement. The soul of Bayh-Dole, then, protects inventors working on projects with federal support (and thus, in the public interest) from a bureaucracy of control asserted as a consequence of the federal funding itself.
To put it simply, the (f)(2) written agreement provides that default ownership of inventions is with the inventor and nothing in the federal funding agreement authorizes any change in this ownership. If an employer desires ownership of a subject invention owned by an inventor, the employer must find a way to obtain that ownership without any reliance on the fact of federal funding (and by extension the use of any resources committed by the employer to the project or acquired by federal funds). NIST’s addition of an assignment requirement to (f)(2) is therefore all the more strange. Clearly, the intent is to turn Bayh-Dole into a vesting statute and its public policy into an assertion that institutions, not inventors, should own inventions made with federal support–and should be free to hold back every one of those inventions in the hope of striking a monopoly deal.
Compare with the federal policy default before Bayh-Dole of releasing all federally supported inventions for public use except in rare situations in which no use was possible without development that would never happen without a monopoly position and the invention was sufficiently uninteresting to the federal government and to charitable organizations that the only funding possible was from speculative profit-seeking investors. Under Bayh-Dole, with the (f)(2) assignment requirement asserted in this way, the entire situation is reversed. Institutions are required to own inventions even before they decide they want to own–even if then they choose not to own and give up ownership to the federal government or re-assign to the inventors. There is a fight on for the soul of Bayh-Dole. The Supreme Court asserted that Bayh-Dole applies only after a contractor acquires an invention made in work with federal support. NIST makes it appear that Bayh-Dole requires all inventions to be assigned to the contractor, and thus that the public policy intended by Congress was that institutions should own all inventions made with federal support, as a default–just what Senator Bayh argued–that inventors should be last in line. It is exactly this that the Supreme Court rejected, but NIST has never understood.]
Inventor ownership of inventions, those inventions becoming subject inventions, and exclusion of employer claims to ownership of those inventions based on federal funding– all this is enabled by the (f)(2) provision in the standard patent rights clause. (f)(2) is the soul of Bayh-Dole. Like an ordinary soul, you won’t find it in the body of Bayh-Dole, but also like the soul, if you remove it from the body, there’s nothing rightly human left.