A common approach to discussions of Bayh-Dole is
(1) accept that what is happening is just what people say is happening–the law is working (as people claim), based on a politicized spun history (as people claim), based on a metrics of success (as people claim), and then
(2) misrepresent the law (or uncritically repeat misreprestations of the law) to account for why, and then
(3) go on complain about the law anyway or propose how to build on its wonderfulness.
It’s a grim way to discuss policy, what with all its trappings of academic sophistication and authority of publication. Such articles gather together, reinforcing each other and vying to move the policy discussion to be about some bit that each posits. As a result, the academic literature on anything Bayh-Dole is nigh unto useless for practice, policy, or thinking about innovation. AUTM data is also useless. Summaries of Bayh-Dole drawn from the academic literature are largely wrong–especially if written before 2011. But so are the political spin statements made about Bayh-Dole when it was debated and which have been repeated, unchallenged and unchanged, for over thirty years. Even Senator Bayh was wrong about Bayh-Dole’s fundamental operation, and the US Supreme Court ruled as much in Stanford v Roche. It’s almost like people want to get Bayh-Dole wrong, regardless of whether they love it, hate it, or just want to gussy it up.
Here’s an article by way of example: “Correcting Bayh-Dole’s Inefficiencies for the Taxpayer” by Michael Sweeney, published Northwestern Journal of Technology and Intellectual Property in 2012. Let’s work through the article’s account of Bayh-Dole. The exercise is not so much to show the defects of this work–it’s someone’s annotated opinion, so fine, and I could pick almost anything–but rather let’s use this work as a springboard to open up what Bayh-Dole actually does, were folks to comply with it, and perhaps from there we might begin to think about what ought to be done about it, if anything. So, consider that the article–forgotten perhaps by all but for the internet and a guy’s law firm bio–has a place of honor here at Research Enterprise and still has a legitimate role to play, though it is badly defective on the law and those defects render its proposals for change mostly useless.
Here’s the start of Sweeney’s article on Bayh-Dole:
By transferring ownership rights of federally funded inventions to non-government contractors and their subsequent licensees, the University and Small Business Patent Procedures Act of 1980 (Bayh-Dole Act) gives private actors unprecedented rights to intellectual property that was cultivated with public money.
I have bolded some things for consideration.
The article was published the year after Stanford v Roche, but here it is asserting that Bayh-Dole transfers ownership rights to “non-government contractors.” Bayh-Dole transfers no rights in inventions. Bayh-Dole requires federal agencies to use a default patent rights clause in their contracting for research or development if they cannot justify an alternative clause that restricts or eliminates a contractor’s right to retain title to inventions arising in federally supported work that the contractor has otherwise acquired. This is precisely what Stanford v Roche was all about. The Supreme Court made it clear that all Bayh-Dole did was establish a default priority of claim as between a contractor and a federal agency to a subject invention–one that the contractor had acquired and which had been made under a federal contract between the contractor and the federal agency that provided funding. The persistence of the “vesting” version of Bayh-Dole, even after being thoroughly squashed by Stanford v Roche, indicates how deeply a faux proxy of Bayh-Dole has settled into the academic literature and the legal community.
Even if a federal agency attempted to implement a vesting or transfer of rights from inventor to the organization that hosted federally supported research using the “exceptional circumstances” provision in Bayh-Dole (35 USC 202(a)(ii)), a federal agency would lack the authority under Bayh-Dole to do it. A federal agency has the opportunity, if it follows proper procedure, to restrict or eliminate a contractor’s right to retain title to any subject invention, but not to hand to a contractor title that otherwise vests by patent law in the inventor.
The closest that a federal agency might come would be to determine exceptional circumstances–show how the purpose of the funding or the nature of the objective is not consistent with the contractor having the right to retain title– and thus eliminate a contractor’s right to retain title, require assignment of the invention to the government, and then exclusively license the invention to the contractor. That is, the exceptional circumstances would have to foreclose a contractor obtaining title from inventors and keeping that title but still end up requiring that the contractor obtain an exclusive license (perhaps to the point of assignment of the invention) from the federal agency. Even for the federal government, that seems over the top.
Even then the federal agency would have to justify granting an exclusive license using the criteria at 35 USC 209(a) and the contractor would have to propose a plan for the development or marketing of the invention under 35 USC 209(f). Those features essentially reprise the requirements of the Federal Procurement Regulation’s determination of greater rights process, but now led by the federal agency rather than initiated by a petition from a contractor or inventor and the contractor’s plan would be kept secret. We would be back to the Nixon executive branch patent policy protocols, but now with the contractor’s plan to better serve the public interest held as a government secret.
The use of “federally funded” in this opening paragraph is more obscurely misleading. As a short-hand, it’s okay to write “federally funded” invention. But “federally funded invention” is not the full scope of Bayh-Dole’s interest.
While Bayh-Dole’s statement of policy is broad, Bayh-Dole’s contracting provisions do *not* apply to just any invention arising from federally supported research or development. The contracting provisions at 35 USC 202-204 go out of their way to separate out from the general category of “inventions arising from federally supported research or development” only those that are “subject inventions.” And subject inventions–whatever else they are–are ones that are owned by a federal contractor, by a party to a federal funding agreement. If an invention is not owned by a party to a federal funding agreement, then that invention is not a subject invention and Bayh-Dole’s contracting provisions do not apply and Bayh-Dole does not preempt federal statutes or executive branch patent policy covering the same ground. Again, put another way, if an invention is owned by someone who has not been made a party to a federal funding agreement, that invention cannot be a subject invention.
It’s a big dark empty space for inventions that arise from federally funded research or development and are not owned by a contractor and not subject then to a special federal statute on the matter. Sean O’Connor has called this a “mistaken assumption” that must have been made in a federal report on patenting in 1947. But no, that’s not a good explanation, given that Bayh-Dole was drafted by an NIH patent attorney who had managed the NIH’s IPA program for a decade and knew exactly how nonprofit patent policies operated, having reviewed over 70 of them as a precondition to allow nonprofits to participate in the IPA program.
The Nixon patent policy ought to apply, but its codification in the Federal Procurment Regulation was abandoned in favor of the Federal Acquisition Regulation, and the FAR implemented only the Bayh-Dole requirements–and so codified only the requirements when a contractor acquires an invention and not what happens when a contractor doesn’t acquire. The Supreme Court in Stanford v Roche was clear, too, that nothing in Bayh-Dole requires a contractor to acquire any invention. NIST, always clueless on Bayh-Dole, has tried to add an assignment requirement to the default patent rights clause–but that addition has no authority in Bayh-Dole, other federal statute, or even executive branch patent policy. It’s just made up out of someone’s posterior cortex. Even then, it is stupid because the assignment requirement applies only to subject inventions–one that the contractor already must own and so has already got the assignment.
The only authority that federal agencies have to acquire inventions arising from federally supported research or development within Bayh-Dole’s contracting provisions involves subject inventions–ones a contractor has acquired. Any other inventions simply are not under Bayh-Dole, and nothing in Bayh-Dole requires contractors or federal agencies to turn those inventions into subject inventions. No requirement that contractors own such inventions, no requirement that they must make inventors assign those inventions, no requirement that federal agencies must require contractors to own or inventors to assign. Not there. Supremes even looked and didn’t find it.
It’s clear however that a contractor can add contractors. Any contractor then comes within the chain of government interest whenever it acquires an invention “made in performance of work funded in whole or in part by the federal government”–even if the contractor’s work isn’t directly accounted for by federal money. It could be doing that work “on the side” or “with its own money” or “with investment money”–doesn’t matter. Not determinative. Is the invention within scope of the project that the federal government has funded some where, at some point? Yes? Is the owner of the invention a party to the federal funding agreement, including by any assignment (such as by exclusive license of all substantial rights), substitution of parties, or subcontract. Yes? Then the invention is within the scope of Bayh-Dole’s contracting provisions.
Bayh-Dole’s scope is broader than inventions that are “federally funded.” That broader scope is just about as easy to indicate–“inventions arising from federally supported work.” Some of those inventions might be “federally funded” but others are not–they are “federally supported” in the sense that they arise in projects that have received federal support. A reader who needs a basic summary of Bayh-Dole won’t recognize a short-hand metonymy like “federally funded invention” and may think that somehow federal money is paying for these inventions when it is instead paying for work done in the public interest, in which inventions may arise in some part or another. All that matters is that some part of the overall work receives federal support. Call it the Midas finger or call it contamination or call it a reasonable expectation of a supporter of work.
Put it this way. Everyone who helps with a theater production is promised free tickets to see the play. That means directors and stage managers and actors and lighting and sound folks all get free tickets. But so do janitors and carpenters and ushers and donors. Clean the stage? Get free tickets. It doesn’t matter that the janitor’s statement of work is to clean the stage and lobby and bathrooms and doesn’t mention the theater production at all. Bayh-Dole’s inventions are like the theater tickets. Any federal support anywhere at the research or development theater means that Bayh-Dole comes into play for tickets. What matters for free tickets then is what play one is supporting in some way. Not whether one is paid from this fund or that fund. To avoid giving the government its ticket, one would have to change the play, not merely exclude government money from paying the actors.
“Federally funded invention” then too narrowly focuses on how federal money is spent and not on the scope of the project to which that money contributes.
It’s not the funding of inventions that matters–it’s the funding of some part of a larger work to which any inventive bits might arise, regardless of how the money used to make those bits is accounted for. That money could be federal. Could be non-federal. Could be no money at all. Doesn’t matter. Are those bits in the larger work? Yes? Was the larger work anywhere supported by federal funds for research or development? Yes? Then those inventive bits arise in federally supported research and development, and when acquired by a federal contractor, become subject inventions under Bayh-Dole’s definition.