The monopoly meme argument is that no one would have ever received any cisplatin if not for an exclusive license to motivate a big drug company to “develop” the drug as a product. Left out is the idea that the federal government paid the development costs itself, including clinical trials, and could have created an industry consortium to establish a standard cisplatin platform and share data as part of developing products, or could have limited the term of exclusivity after which time, the license would go non-exclusive and others could also explore use of the invention. Actually, the exclusive license for cisplatin was time-limited, but at the end of the term, the licensee and university both petitioned for an extension. So much for open access.
The Kennedy patent policy required nonprofits and companies doing health research for the government to show that their private patent monopoly would better serve the public than would federal open access. Latker dropped that requirement, so that a private patent position might not better serve the public, unless serving the public meant increasing the company’s valuation by holding exclusive rights, or maximizing profits by being the sole source and suppressing competing alternatives.
In place of actually showing that a private patent monopoly on a public health invention would better serve the public than open access, Latker set up a set of public protections in Bayh-Dole and then made sure that these protections would not be enforced. He made them narrow, weak, slow, waivable, optional. The few serious protections, such as a limitation on the duration of an exclusive license offered by a nonprofit to a big company, were amended away in 1984, three years after Bayh-Dole came into effect.
But in all his cleverness, Latker couldn’t figure out how to make a law that forced university inventors to cede title to their universities. He had done it in the IPA. But now he was trying to change federal patent law to, as it were, screw over displace nonprofit inventors in favor of institutional “middlemen.” Latker took the definition of subject invention from the shiny new Federal Procurement Regulation that was finalized in 1975, with its “of the contractor” wording, but didn’t take over the parallel FPR requirement that either the contractor or the inventor assign title to the government unless the government decided otherwise.
In the FPR’s standard patent rights clause for situations in which contractors were not allowed by default to hold title in inventions they acquired, the contractor was required to have an agreement with its research employees to “effectuate” the FPR requirement regarding federal claim to title to inventions. It went: inventor–if you don’t assign to the contractor, then you agree to assign to the government unless it decides otherwise. Latker reversed the requirement to assign title for both the contractor and the inventor, but made it appear that title never vests with inventors to start with by leaving out any requirement for inventors to assign to anyone.
In Bayh-Dole fantasy land, title to inventions made in federally supported work doesn’t vest with inventors but instead with either contractors or, if contractors don’t want it or bungle it, then to federal agencies until they squeak and don’t want it either.
So when the Stanford dispute with Roche heated up, sixty or so little demon university lawyers went running up a wall filing amicus briefs with the Supreme Court in Stanford v Roche claiming that Bayh-Dole vested title with universities. Even Sen Bayh went running up that wall with them. It had to be this way “to make Bayh-Dole work.” Nonsense. Disclosure: I worked on a brief joined by IEEE and AAUP that argued against the university demon lawyers, and our careful reading prevailed.
Bayh-Dole removed authority for federal agencies to required title from inventors. If inventors did not assign to a contractor, then the invention could never be “of the contractor”–meaning owned by the contractor–and so could not be a subject invention. Nothing in the law forced inventors to assign anything to anyone. Latker must have recognized his monster had this vulnerability. He tried to shore things up by getting involved in drafting the implementing regulations. But he was stuck with the limiting definition of subject inventions without a statutory apparatus to require federal ownership in any inventions. So he added a “written agreement” provision in the standard patent rights clause–37 CFR 401.14(f)(2). Contractors were to require their research inventors to make a written agreement to report subject inventions and to sign paperwork to establish the government’s rights in subject inventions. But that still depends on an invention being already *owned by the contractor*!
But Latker had another trick. He didn’t call the “written agreement” a “patent agreement” for a reason–it isn’t. And it can’t be the FPR’s agreement to “effectuate” the government’s right to request title because Latker has “reversed” that, so now priority was with the contractor. A contractor could request title–at any time–but in general, the contractor has to make a deal with the inventor to get that title. And Bayh-Dole doesn’t require contractors to make any deal.
The “written agreement” requirement to require inventors to establish the government’s rights in subject inventions would not make any sense if *inventors never would ever have any such rights* if they were always forced to assign title to their contractor-employers. Perhaps they could retain a financial interest in their invention–that would be a sort of right–but there’s nothing in Bayh-Dole that permits them to do so.
But there’s a way the written agreement works to turn inventions into subject inventions. Bayh-Dole authorizes contractors to add parties to the funding agreement. 35 USC 201(b). The written agreement in the patent rights clause requires contractors to make inventors parties to the funding agreement. Inventors become *contractors* too. And the next definition, 35 USC 201(c), provides that any party to a funding agreement is a contractor. So the written agreement requires contractors to make potential inventors become contractors. Inventor holds title to invention when he or she invents. Invention then is “of the contractor”–the inventor contractor–and so the invention becomes a subject invention!
It’s just that this new subject invention is not one “of the contractor-employer” under the contractor’s patent rights clause 37 CFR 401.14. It’s a subject invention under the inventor’s patent rights clause 37 CFR 401.9–inventors are treated as sole proprietor small businesses. Now NIST is actively trying to get rid of the inventor’s patent rights clause. The Department of Commerce was in cahoots with Latker, and Senator Dole got the law changed when Latker took a position with Commerce so Latker could keep monkeying with it.
NIST carries on the effort to try to fix Latker’s cleverness, or screw up. In 2018 NIST added an assignment requirement to the written agreement at (f)(2). Now the written agreement requires inventors to promise to assign subject inventions to their contractor-employer. But wait–these subject inventions could be only the inventions that the contractor-employer already owns. Why? Because if the authority to vest title with the contractor-employer was already within the scope of Bayh-Dole, then the Supreme Court would have pointed that out. But the Supreme Court rejected that idea. It was up to contractors to negotiate for rights with their inventors, outside the framework of Bayh-Dole and federal contracting. To negotiate those rights doesn’t involve inventions that are subject inventions. Those inventions are not even under the contractor’s patent rights clause because the contractor-employer does not own them!
On the other hand, if the contractor-employer has made its potential research inventors parties to the funding agreement, then the contractor is blocked by the patent rights clause from demanding an interest in the inventor-contractor’s subject inventions as a condition of any subcontract. See 37 CFR 401.14(g)–another provision copped from the Federal Procurement Regulation’s patent rights clause but not in Bayh-Dole. The idea is that prevails is that one contractor does have a right to the inventions owned by another contractor, and cannot use the fact of being a party to a federal funding agreement to leverage a claim against another contractor’s rights.
In practice, no university complies with the standard patent rights clause written agreement requirement. Instead, they uniformly substitute a demand to own inventions as a condition of the federal award. Which violates the patent rights clause. But no one cares. Bayh-Dole is the sort of law that is only law if someone ever bothers to enforce it. The rest drops away as historical fluff, a soft packing material to ensure the safe shipment of the statutory monster through Congress.
Bayh-Dole is a botch job that federal officials and university-affiliated patent administrators–especially in the form of AUTM, an organization initially created by Latker to show grassroots support for the IPA scheme going government-wide (instead, it was blocked by Congressional oversight and then shut down in 1978 as ineffective and contrary to public policy). The only thing “delicate” about it is the terrifying prospect that someone will read through the tangle of law and regulations and patent rights clauses and history and fake metrics and fake success stories and see that Bayh-Dole is a mess, universities don’t comply with the patent rights clause, federal agencies don’t enforce it but for the stupid, useless paperwork (which they are all fine with), and the law isn’t even saved by it somehow “working” anyway. Its outputs are so bad no one is willing to report them. No one will break out university technology transfer for subject inventions. Take a subject invention. Has the benefits of use of that invention ever been available to the public on reasonable terms? How long did it take? What sort of license was used? Most universities receive almost half of their research funding from non-federal sources. How about, then, reporting how the non-federal inventions are doing? Nope, it all gets mixed together, along with non-inventions, and abstracted as inventions, patents, licenses, startups, and money all in the same year, from which almost nothing can be understood, given that inventions disclosed in a year will have nothing to do with patents issued in that year, and are unlikely to have much to do with licenses in that year, or money.
Bayh-Dole has been forty years of bureaucratic fakery, non-compliance, and scheming to make look good a federal law drafted by bureaucrats for other bureaucrats to enable a pipeline of exclusive control over patents on inventions made in federally funded health research at universities to biotech and pharma companies. That scheme screws over inventors, research tools, technology platforms, and standards. That scheme screws over entrepreneurs and small companies. That scheme even screws over much of the biomedical industry, playing favorites while fragmenting needed rights across many nonprofit patent holders, each holding out for an exclusive deal or nothing–because Bayh-Dole. But really, it’s not Bayh-Dole. They don’t have to take ownership. They don’t have to file patent applications. They could leave inventors alone, or could offer to help without demanding ownership to do it. Bayh-Dole opened an ineffectual door. Universities did not have to follow the monster through it. But they have. How will they ever get back out?