NIST’s other changes to Bayh-Dole’s implementation, 2

We are working through NIST’s other changes to the implementation of the Bayh-Dole Act. It’s a rather strange exercise, since the standard patent rights clause required by Bayh-Dole as a default is not enforced and universities comply only with the paperwork requirements and not with the substantive provisions. For that matter, federal agencies do not act on their rights under the standard patent rights clause–don’t exercise their license, and don’t march-in to protect the public from nonuse and unreasonable use. So it’s all rather sad to watch NIST “fine-tune” such a disastrous situation, arranging deck chairs on the Titanic.

Let’s look at a third significant change NIST has made.

3/The 60-day window for government requests to own

Bayh-Dole section 202(c) sets out the conditions upon which the federal government may “receive title” to a subject invention. These conditions come at the end of paragraphs setting out provisions that are to be placed in each federal funding agreement with a small business or nonprofit contractor. Each reads as [provision]; [remedy]. Thus, the contractor will be required to disclose each subject invention within a reasonable time; the government “may receive title” to any subject invention that is not so disclosed (202(c)(1)). This pattern repeats for election to retain title, and filing a patent application. There is no such remedy for failing to report on invention utilization, failing to include in patent applications a government funding and rights notice, preference for US manufacture, or any of the nonprofit-specific requirements, such as restrictions on assignment of subject inventions and on use of income earned with respect to subject inventions, as well as a preference in licensing for small businesses.

At the design level, then, the breaches for which a contractor might be required to give up ownership of a subject invention are the bureaucratic ones–failure to disclose, failure to elect to retain title, failure to file a patent application. These are all but empty requirements. If a contractor files a patent application on a subject invention, that invention will be public in 18 months. If the application carries a federal funding notice, then the disclosure requirement is met without any additional paperwork. If the contractor does not file a patent application, then the U.S. patent rights are generally blown one year after public disclosure or public use or sale or offer to sell–for universities, a bar date is almost a given. What then is the point of the government having the right to “receive title” to a property that is generally won’t carry patent rights.

Put it this way, if a contractor blows past a patent application bar date, then the invention is no longer a subject invention–it is no longer patentable–and therefore it is not subject to Bayh-Dole’s patent rights clause, and the government has no right under Bayh-Dole to “receive title.” There’s no title to receive. It’s a dead parrot. Similarly, there’s no particular point in the notice to elect to retain title. If the contractor has gone to the trouble to obtain title but then suddenly doesn’t want that title and lets the invention skate, then we are back to blowing the patent rights and the invention is no longer a subject invention–no remedy possible. If the contractor goes ahead and files a patent application, then in 18 months a federal agency will have public notice of the invention and if there’s a problem with the federal funding notice or uncertainty about whether the federal government should have a license, the agency can follow up as it chooses. In any of these scenarios, the actual communication required by Bayh-Dole is pointless bureaucracy. Nothing happens as a result of these communications that pertains to practical application, innovation, commercialization, public benefit, global technology leadership. Nothing. It’s just a paper show–and that’s where Bayh-Dole (written by the NIH) parks the potential for a contractor’s loss of ownership.

For anything substantive–failure to report on subject invention utilization, failure to achieve practical application, failure to make a subject invention adequately available, failure to prefer US manufacturing in exclusive licenses in the US, failure to make the terms for public benefit reasonable, failure (for nonprofits) to include the nonprofit patent rights clause with any assignment of a subject invention and failure to use properly income with respect to a subject invention or prefer small businesses in licensing–none of these failures carry the possibility of loss of ownership of a subject invention. Only for nonuse and failure to make a subject invention adequately available is march-in even permitted, and there with a procedure that is so easily delayed and disruptable that it has never been successfully used.

Everything substantive in Bayh-Dole is waivable, or the federal agency does not have to act. Even the paper requirements are waivable or don’t require federal agency enforcement. If one were serious about Bayh-Dole as meaningful public policy, the rules would be flipped. Loss of ownership would be a remedy for the substantive breaches, not the paper reporting breaches, and federal agencies would be required to enforce the substantive breaches, or the law would operate to make the remedies happen by operation of law rather than through a bothersome process of more paperwork forced on federal agencies. Truly, Bayh-Dole is a gawd-awful law by design–no matter, really, one’s views on the desirability of having public money subsidize speculative profit-taking patent monopolies that exploit public suffering identified by the federal government of most in need of government intervention. If one believes that Bayh-Dole is a powerhouse law for the public good, then why would one also then place the draconic remedy of loss of ownership in the paperwork section rather than in the provisions dealing with the failure to achieve the public good? Bayh-Dole advocates must believe that the publicly subsidized speculative betting pool on the future value of patents is itself the public good. What trash.

With this context, let’s look at what NIST has done. The existing standard patent rights clause (37 CFR 401.14(a)(d))) puts a time limit on federal agency action to request title if a contractor fails to disclose a subject invention or elect to retain title to a subject invention:

provided that the agency may only request title within 60 days after learning of the failure of the contractor to disclose or elect within the specified times.

This limitation is not in Bayh-Dole, and NIST’s new rule removes this time requirement. On the face of it, NIST has made the patent rights clause consistent with the statute. But the change is significant. Going forward, a contractor and any assignees or exclusive licensees are exposed to the government stepping in and claiming that there’s been a paperwork failure to disclose or to elect to retain title. These could be paperwork or procedural problems–that the disclosure was not technically complete, that the notice to retain title came in two days late. Five or ten years later, a federal agency could call in its title, apparently regardless of the reliance of the contractor on the federal agency’s inaction. The 60 day window, then, was useful in setting a standard for how long a federal agency had to make a choice whether to require assignment of the invention or not. Up or down, agency–just don’t leave everyone hanging in uncertainty for years.

Perhaps this was the sore point with federal agency behaviors prior to Bayh-Dole. Under the Kennedy/Nixon executive branch patent policies, federal agencies were to make a determination of whether a contractor might retain ownership of a given invention after the invention had been disclosed and the contractor had requested a determination. These situations were specific to contractors without commercial positions (like universities) and for research directed at government product development, public health, and areas of primarily government use and expertise, such as atomic energy, space exploration, and military weapons systems. The determinations might drag on. Federal agencies could have just said “no–you haven’t made a clear case here for a private patent monopoly doing a darn thing for the public–go away.” But no, they dragged things out, and the Bayh-Dole advocates used that against them. Thus, the 60-day window is a gesture in Bayh-Dole to a situation in which a federal agency could once again drag things out. Of course, where it is in a contractor’s interest to drag things out, ostensibly acting against the public interest in such things as a determination of whether an invention is a subject invention or march-in for nonuse or unreasonable use, Bayh-Dole sets up procedures that allow a contractor to drag things out interminably. Again, an awful steaming brown mess of law.

NIST has done nothing with the other elements in the patent rights clause section (d) that have to do with a federal agency receiving title. For instance, in (d)(2), a federal agency may request title if a contractor fails to file a patent application in the time required, but if the contractor then files a patent application before the federal agency’s request for title is received by the contractor, then the request may be refused. The effect is–so long as the contractor corrects the breach before the agency completes its request, the request doesn’t matter. Why not have that, as well, for (d)(1) matters of disclosure and notice of retention of title? In other words, what do the time requirements for disclosure and election to retain title actually do? Not much of anything. The implementing regulations undermine them. But then, this is all in the paperwork section of the law anyway–there’s nothing substantive about the paperwork, though there is about the government’s remedy in taking ownership for paperwork mess up. We can see what is super important in Bayh-Dole–proper paperwork! Public benefit–meh.

The standard patent rights clause also adds a third area in which a federal agency can request ownership–(d)(3)–if a contractor fails to continue prosecution of a patent application, fails to pay the patent maintenance fees, or fails to defend the patent. None of this is in Bayh-Dole–filing a patent application is sufficient. If the application is rejected, according to (d)(3), a contractor must attempt to overcome the rejection or lose title (potentially) to the government. This is all weird. It is as if those implementing Bayh-Dole decided that the focus of the law was insisting that the patent system must be used rather than the patent system should be used only when doing so promotes the use of inventions arising from federally supported research or development. Such a huge difference.

All this would not matter if Bayh-Dole had been restricted only to federal contracting. But Bayh-Dole also sweeps aside executive branch patent policy with regard to federally owned inventions–expressly authorizing the use of exclusive licensing to the point of assignment of the invention (as an exclusive license gone too far). If, once a federal agency received title to a subject invention, the agency released the invention for open domestic access, then the government receiving title to an invention would differ from the public domain only in those cases in which the federal agency determined that it would be in the public interest to grant non-exclusive licenses rather than dedicate the invention to the public domain. Why? One reason would be to ensure the quality of product manufactured using the invention. Historically, that was important in the less-regulated patent medicine industry. Presently, this is important in the enforcement of standards, where a company taking shortcuts and failing to fully implement a standard harms consumers and the marketplace for emerging technology. Another reason is that the federal government can use a patent to control the import of product that practices a given invention, thereby offering some protection to the formation of a US expertise in a new technology.

But if federal agencies become fixated on dealing in patent monopolies, then the request to receive title reads like a question of whether a federal agency is competing with the private contractor to see who can do a patent monopoly deal with a private company. (NIH apparently has this fixation in the case of prescription drugs, though not for research tools–there must be some heated debates–just wait until someone finds a drug that can be used as a research tool (“take two of these, wait an hour, and then tell me your hypothesis”)).

This apparatus in (d)(3) is rather unnecessary. Any action by a contractor that evidences the contractor no longer wants title to the invention acts as a notice of election not to retain title. Folks fascinated by vesting might here find a place for it. If a contractor-owner fails to maintain a patent application or issued patent, why not vest the ownership of the invention (and any patent rights) with the federal government automagically? Why bother with the federal agency having to be notified, having to decide, with all the uncertainty of what the federal agency might do, and the time frame (often no more than sixty days) in which to act.

There’s another consequence of federal agencies becoming fixated on a trade in patent monopolies: an institutional contractor cannot determine that it is not appropriate to file a patent application on a subject invention, cannot decide that a patent will not promote the use of an invention. That’s been the case with software, for instance, where the code might use free software (GPL) and thus a patent is merely an expense but has no way to be applied to promote use. In pre-Bayh-Dole federal policy, even if the federal government acquired ownership, the invention would be made freely available and the contractor had no worries about access. But if the federal government goes all bonkers on trying to wrap up patent rights to hand off to a favored company as a monopoly, then the patent may block open source development and screw everyone (other than the government-appointed monopolist, who might find a happy way to derive financial value from screwing everyone else). More fine public policy there, Bayh-Dole.

Of course, it is true that Bayh-Dole does not require inventor-contractors to file patent applications, and there’s no consequence if they don’t–the government has no right to receive title if an inventor-contractor does not file a patent application (for which see 37 CFR 401.9–(d)(2) is excluded from the inventor patent rights clause, while (d)(1) and (d)(3) are present). It’s just that no one complies with the (f)(2) written agreement requirement and even if they did, they would reject the idea that inventors have become contractors and that the (f)(2) written agreement pre-empts all other claims on assignment of their inventions but for those arising from equitable ownership. It would be too much for their heads to hold, that Bayh-Dole’s fundamental set up is that inventor-contractors own their inventions and have no obligation to assign to anyone except (1) an employer where the employer holds equitable title; or (2) to a federal agency if the inventor fails to disclose the invention or fails to choose to keep ownership of the invention. In case (1), an inventor is hired to invent or assigned to experiment or is a senior officer with an equitable duty to assign; in case (2), it’s a matter of following paper bureaucratic procedures.

Thus, it is also odd that a contractor, having acquired an invention made in a project receiving federal support, cannot somehow undue that acquisition. There’s no “We took assignment by mistake” or “Now that we see what we have got, we don’t want it.” Once inventors give up their invention, they cannot get it back by assignment from the contractor without consequences, especially if the employer-contractor is a nonprofit and required to include the nonprofit patent rights clause in any assignment of a subject invention. Of course, if a university assigns a subject invention to a federal agency, that invention is no longer a subject invention because it is not owned by a contractor and therefore fails the definition of subject invention.

There’s nothing in Bayh-Dole on federal disposition of inventions that authorizes a federal agency to assign an invention it has acquired to the contractor’s inventors, other than using the ruse of an exclusive license (in response to an acceptable business plan and all that) that authorizes the inventors to enforce the patent. Nothing on assignment to a contractor’s inventors. 37 CFR 401.10 applies only to inventors who are federal employees, and there the regulation contemplates a federal agency assigning a federal employee’s invention, co-invented with a contractor’s employees and acquired by the federal government, to the contractor. Nothing about a nonprofit contractor assigning subject inventions owned by the contractor back to the inventors outside of the restriction in the standard patent rights clause at 37 CFR 401.14(a)(k)(1). Of course, if everyone can ignore the standard patent rights clause when it suits them, then the patent rights clause is hardly uniform at all. It is just a show of uniformity and a swarm of randomness just below the surface.

Thus, it’s difficult for a contractor to undo an assignment it has received–at least if the contractor is a nonprofit. Since there are no restrictions on a for-profit company’s assignments, it can do anything it wants by way of assignment, with only minimal obligations to the business patent rights clause.

NIST’s removal of the 60-day window for agency action to request title to a subject invention opens up a significant uncertainty in the disposition of title. Yes, it’s only in the paperwork section where nothing much matters. True, there may still be a good laches defense if a federal agency takes much longer than 60 days to make its request. And of course, it’s entirely possible that most of the time federal agencies just don’t care and by the time they do find out about a problem with disclosure or election to retain title, nothing matters anyway.


Each of these changes in Bayh-Dole’s implementation have significance. The new assignment requirement appears to apply to all inventions and brings the patent rights clause into alignment with prevailing practice. Except it doesn’t do any such thing, but Bayh-Dole appearances historically have counted more than actual law, other than for the Supreme Court in Stanford v Roche. So we should be braced for another clever scheme to game the law.

The glumping of a default patent rights clause set up as an exceptional circumstance into the main patent rights clause further effaces Bayh-Dole’s design expectation of multiple patent rights clauses, each suited to its particular circumstances, in favor of an omnibus patent rights clause with numerous conditionals that will rarely apply and are subject to strained interpretations based on the inappropriately wide context in which they now must be used. There ought to be more exceptional circumstances patent rights clauses. Section (b) should stay, and there should by now be a (c) for NSF cooperative research centers and a (d) for industry-led common technology development with nonprofit subcontractors and an (e) for research in medicinal chemistry that might produce therapeutic drugs with mass commercialization potential. Bayh-Dole is designed for just this thing. NIST’s changes make it less likely that Bayh-Dole’s design will ever be used.

The loss of references to small business in the implementation effaces the difference between what Bayh-Dole the law establishes and the limits of what an extension of Bayh-Dole by executive order can do. While the changes make it appear that the implementation is one and the same for all companies, the statute and executive order cannot do the same things. The executive order, for instance, cannot preempt statutes, and Bayh-Dole does not repeal those statutes–only preempts them with regard to subject inventions–and those inventions are ones that nonprofits and small businesses have acquired. Nothing more. In practice, likely, nothing will change. In the design of the law, however, NIST further obscures the statute in favor of a misleading but harmonious text management.

The elimination of the 60-day window for federal agency requests for ownership appears to be consequential. But really, how can it be anything other than adding uncertainty. Yes, the change better tracks the statute–but then why not get rid of sections (d)(3), (e), (f), and (g) of the standard patent rights clause–none of which are in Bayh-Dole. Just pulled from someone’s posterior cortex as the urge arose. In practice, perhaps, contractors will beg agencies to decide, or “just kill our work now.” And agencies will drag their feet and brains because they now can. And who cares–really–since this all has to do with bureaucratic paperwork compliance and nothing at all to do with promoting the use of inventions, public benefit, US manufacturing, and the like.

Even if the federal government does obtain ownership of an invention due to a contractor paperwork failure, we can expect that the federal agency will go about the very same effort to grant patent monopolies to favored company-buddies (“commercialization partners”). In the utterly rare case in which a contractor fails to disclose a subject invention and then gets into a lucrative patent deal with a company, if the federal government were to require assignment of the invention (from the company, no doubt, since most exclusive license deals of this sort involve assignment of the invention), we can expect that the federal government would simply license the invention right back to the same company, and take over the royalty stream. That sounds badish.

But here’s the thing–if a contractor is worried about the government stepping in to demand title to an undisclosed subject invention, the contractor will go sell the future royalty stream, and take the net present value of that stream. Now when the government comes calling for title to the invention, what does it matter? Title and financial return have been separated, and it’s up to the federal government to work out terms with the company-assignee that has lost ownership of the invention. Sure, there can be lawsuits galore for another five or ten years, but the contractor will have millions set aside from its royalty stream sale to pay the legal costs to beat off legal challenges. All this is like fighting to the death over a pot of lentils. Nothing changes but that lawyers get well paid. None of it has to do with public benefit. Everything has to do with how speculators on patent monopolies make money and push the obligation and risk to do anything further down the line. It’s an old story. Same as it ever was.

This entry was posted in Bayh-Dole and tagged , . Bookmark the permalink.