If a university fails to patent under Bayh-Dole, nothing ever happens. But that’s not even the meaningful answer. Look, even if a university gets a patent on a subject invention–one arising from federally sponsored research or development–there’s absolutely nothing in Bayh-Dole that requires a university to license each subject invention, license timely, or if it does license the invention, to license exclusively.
The lack of such requirements are utter failings–fatal defects in the law. It’s beyond comprehension that we have tolerated gibberish about how wonderful the law is. It’s a hot mess of drafting ignorance and incompetence. Beyond that, it doesn’t work. It has never worked, like the IPA program before it didn’t work. It has never fully operated, never been complied with, never enforced (but for one instance of federal bullying of a small company over the failure to put a patent disclosure in the form of a written report rather than as a patent application).
Well, the paper-pushing parts of Bayh-Dole have been enforced–the stuff that has absolutely nothing to do with promoting use of inventions, free competition, or US manufacturing–while all the substantive parts are ignored, waived, or hopelessly bungled. So even if somehow one hoped that the hot mess of drafting incompetence worked brilliantly despite all its failings–it doesn’t. Only if you are a gourmand of fake history, fake law, and fake metrics would you swallow the idea that Bayh-Dole has any merit whatsoever.
The whole point of Bayh-Dole is to enable private dealings on patent monopolies, despite express legislation that required–as a condition of allocating funds–that federal agencies take ownership of any inventions and manage them in the public interest–with the default being, make everything available, don’t sue US citizens for using what public money has supported, don’t play favorites with companies, and don’t take a financial stake in a company relying on patent monopoly pricing to take advantage of public welfare. Bayh-Dole in its brilliance instead insists that the public interest is best served (by default) if every time a federal contractor gets ownership of any such invention, the contractor gets to decide what’s in the public interest, not a federal agency, and not with a default of making the invention available to all (including the inventor).
No. Bayh-Dole’s presumption is that contractor dealings over patent exclusivity are essential for the public to benefit at all from federally supported research. Otherwise, we would all do best going with the federal default of non-exclusive licensing. Yet Bayh-Dole doesn’t care one lick whether a university, having got patents, pops off and licenses non-exclusively or doesn’t manage to license at all. Oh, sure, there’s march-in, but as we all know, march-in does not operate.
Bayh-Dole delegates march-in to federal agencies, and for forty years federal agencies have refused to march-in, even for university patented inventions that have been kept from use for the entire term of the patent. Bayh-Dole does not require federal agencies to march-in and does not give any other part of the federal government oversight over agency decisions on march-in.
If a university patents and never licenses, there’s never march-in.
If a university patents and licenses exclusively and there’s no practical application, there’s never march-in.
This, too, is another fatal defect in the law. You would think someone drafting Bayh-Dole would stop to think that federal agencies must be required to march-in, or better that an agency that has no extramural research contracts must enforce march-in for all the rest. Instead, Bayh-Dole march-in procedures were designed not to operate, as Howard Bremer of WARF bragged after he helped to create them.
Worst of all, there’s that problem of the exclusive license. The only reason for Bayh-Dole to exist is the exclusive license. According to the claims made for the law, the exclusive license is the only thing that permits the public to benefit from inventions made in federally supported work. “What’s available to all will be used by none,” so the fantasy mantra goes. “A non-exclusive license is just a tax.” So another fantasy mantra goes.
It’s all mind-numbing nonsense, but let’s say for a moment that these fantasies were true. Then what? Then Bayh-Dole could not possibly be permitted to allow either no licensing or non-exclusive licensing. No licensing would mean universities are sitting on patent rights and preventing uptake. If a university cannot find a licensee willing to pay patenting costs, cash milestones, and a running royalty on sales or an equity stake, or both, then the university should have to offer an exclusive license royalty-free, for $1. Forget recovering the patenting costs–the public purpose is not that the university hold out for money, but that inventions get used, and used timely.
Worse if the university offers an invention non-exclusively. According to advocates of Bayh-Dole, non-exclusive licensing means the death of the invention. No one will use it, no one will develop it, and the public will be harmed. And yet Bayh-Dole does not prohibit non-exclusive licensing.
I argue that Bayh-Dole does forbid patent trolling–suing for infringement while not demonstrating that the patent holder has worked the invention. Technically, Bayh-Dole places in federal patent law a policy that limits the rights of a patent holder to use the patent system only to promote the utilization of subject inventions and therefore cannot bring an infringement action without first establishing that the action is necessary to promote utilization of the invention. See 35 USC 261 and 35 USC 200. But university administrators clearly don’t accept this argument–witness Caltech trolling Apple and Broadcom. Or Stanford trolling Roche. Nothing about those infringement suits had anything to do with promoting use–the companies were using patented university subject inventions (apparently) not only without an exclusive license (impossible!) but also without any license whatsoever. They were using these inventions despite the patents. There’s absolutely no policy support in Bayh-Dole for Caltech or Stanford to sue for infringement on such patents.
Think about it this way. When a university starts trolling industry, when it wins cases, these “wins” amount to non-exclusive licenses with court-determined royalties. There we are, back to an ass-backwards non-exclusive licensing program, in which a university insists that it must find an exclusive licensee, fails, and then years later gets around to a non-exclusive licensing program by suing companies for infringement when the university could have offered immediately a public royalty-free non-exclusive license from the get-go and licensed the entire industry.
But per Bayh-Dole advocates, exclusive licensing is essential, and the federal government did it poorly, and university folks (really, at the time, Research Corporation, WARF, UC, Stanford, and MIT) did it really, really well. And if that was the case, why on earth did Bayh-Dole expressly authorize federal agencies to grant exclusive licenses–to the point of assigning government-owned inventions? If the federal government was so bad at it, why let the government continue to try? Worse, why make it even harder for the government to license exclusively when nonprofits and small companies have no such restrictions.
35 USC 209 wallows in a pile of restrictions on federal agencies if they think to grant an exclusive licensing, determining the exclusive license is necessary, documenting the determination, demanding a plan for marketing or development from the candidate licensee, and restricting the scope of the license to only what is necessary, with a bunch of other provisions that look a lot like the standard patent rights clause on the contracting side. But on the contracting side of Bayh-Dole, those provisions form a working requirement on the invention as a condition of retaining principal rights. On the federal licensing side of Bayh-Dole, these provisions turn what ought to be a federal obligation to work the invention it has acquired into a pile of diligence requirements for federal licensing.
The result–under Bayh-Dole it is almost impossible for a federal agency to grant a simple public non-exclusive license. Luckily federal agencies routinely ignore Bayh-Dole, so perhaps they grant non-exclusive licenses anyway, even though these are supposed to be the death of the public use of the inventions so badly mishandled that they are made available without patent exclusivity.
Bayh-Dole is such nonsense. If exclusive licensing were really essential–if patent exclusivity was by default necessary–then Bayh-Dole would insist that nonprofits grant exclusive licenses and force them to justify any non-licensing and any non-exclusive licensing. And Bayh-Dole would insist that federal agencies grant exclusive licenses, make it easy for them to do so, and make it difficult for them not to license or license non-exclusively. Well–at least Bayh-Dole does make it difficult for federal agencies to license non-exclusively.
Without a mandate if not a requirement for nonprofits to license exclusively, Bayh-Dole is fatally defective. Without such a requirement, there is no reason for the law to continue to live on. If inventions might survive or even thrive with non-exclusive licensing, then the federal government might just make them all available under a standard domestic public non-exclusive, royalty-free, non-discriminatory license. Anyone could download and print off an official license any time they wanted. That would be the default. If universities wanted to “transfer technology” they would already be through the fussy paperwork side of it and would be freed from the cost and complexity of dealing with patenting. Federal agencies could hire patent lawyers just as easily as universities. Universities could then focus on teaching the invention to industry or to professionals or to other researchers, and providing services as requested. You know–teaching and research and public service, the stuff that universities are supposed to be good at.
Even if a university, then, manages to “patent” an invention under Bayh-Dole, that means absolutely nothing. Not licensing means something–though there’s no consequence for not licensing. Not granting an exclusive license means something–though there’s no consequence for not licensing exclusively. All march-in means is that the federal government must do the licensing for the university, and we are back around to the fact that the federal government has never bothered to march-in and do university licensing for it, and if it did, it would have to compel the university to grant an exclusive license or again there’s no point whatsoever for Bayh-Dole to exist.
Prior to Bayh-Dole, the government had two approaches to contractor rights in inventions. The first was the IPA programs run by the NIH, NSF, and Department of Commerce. These programs required a contractor to take ownership of any invention made under federal contract that the contractor decided to file a patent application on. The second was the Federal Procurement Regulation, that required contractors to have patent agreements with employees so that if the government requested title to any given invention, either the contractor or the employee would be obligated to assign it over to the federal government.
If either the contractor or the employee-inventor wanted to keep rights and not assign them over, they had to make a public case for it. What would they do that wouldn’t get done if–wait for it–the federal government took ownership and gave everyone absolutely free access regardless of whether the government obtained patents or didn’t? That is, pretty much the only reason for a contractor to seek to retain an invention under the FPR was to set up to grant an exclusive license. By contrast, the IPA program made a big deal about default non-exclusive licensing, but then made a bigger deal about exclusive licensing by making it easy to skip the non-exclusive licensing default by arguing “we tried but couldn’t do it” or “we thought about doing it but decided not to bother.”
Without exclusive licensing, Bayh-Dole is dead. Bayh-Dole does not require exclusive licensing by contractors. And Bayh-Dole goes out of its way to make exclusive licensing hard for federal agencies. Meanwhile no one marches-in to deal with non-licensing.
You can see why then a university failing to patent under Bayh-Dole means nothing. A university failing to exclusively license or a federal agency failing to exclusively license is the real test metric. What happens if a university fails to license exclusively under Bayh-Dole? Absolutely nothing! And that’s a compelling reason to get rid of the law now, before NIST wastes more time bungling it up.