The question to ask is not what was intended by Bayh-Dole but rather what ought to be federal policy regarding inventions made in work for which those involved have gone out of their way to apply for federal funding. As an opening, it would appear more favorable to the public that a contractor retain exclusive rights only for those instances of an invention that have been brought to the point of practical application within a reasonable time of a patent issuing (say, three years), and then only in the areas of application in which those instances operate. If you create a patent monopoly in an actual market, you can keep your patent monopoly in that market, provided that your pricing and other terms are reasonable–that is, comparable to what they would be in a market with competition. “Reasonable” terms in this context are not “whatever you can extract by way of payment based on your being the sole source.”
Put it another way: the research was proposed to the federal government as being in the public interest. Any invention is a door of opportunity, opening up a new frontier. A patent provides a map to the new territory as far as an inventor can document that frontier at the time of filing a patent application–any reasonable variations on the theme can be constructively reduced to practice by describing them in sufficient detail to persuade a patent examiner that they, too, are new, non-obvious, and useful. That’s the start of a map to a new frontier. The inventor gets first dibs on the frontier, but as others arrive, the inventor’s right to exclude others is temporary–the inventor gets to retain exclusive control only for those instances that the inventor uses to provide a reasonably accessible benefit to others. Anything else should go public–either with a reasonable royalty to the inventor or if the inventor can’t be reasonable, then royalty-free.
In this development of the idea, the door of opportunity is created on behalf of the public, using public resources. The bargain, then, is that the inventor gets a period of time–three years, say–of exclusivity to use instances of the invention to provide a benefit to the public. The door of opportunity is public, the map of the new area is public, the opportunities themselves are public. Even any invention holes are public. After three years, whatever the inventor has managed to use or, if needed, to bring to the point of practical application, stays under the control of the inventor. Everything else is returned to the public, either as public domain or as a commons (i.e., non-exclusive licenses on reasonable terms).
There are consequences to a government getting into the business of supporting research. One cannot just fixate on one desired outcome and ignore what else happens. When Vannevar Bush proposed a national research organization, he aimed to keep it independent from politics but accountable to the president, under the control of a director who would exercise wise judgment in what to fund. Instead, research funding was distributed throughout federal government agencies; they built their own procedural apparatuses by which to award funding; university administrators moaned and groaned about the differing requirements; and a need for “uniform” treatment was hatched for the convenience of the poor university administrators attempting to comply with differing agency procedures rather than to support federal agency purposes in supporting research determined to be in the public interest that otherwise would not be undertaken. It has been a botch from day one.
The two options present in the Kennedy patent policy are also present in Bayh-Dole, where they take the form of march-in at 35 USC 203. March-in, in turn, depends on one of four conditions going true:
failure to meet the standard of practical application
failure to satisfy public health or safety needs
failure to satisfy access for public use specified in federal regulations
failure to meet the US manufacturing requirement provisions
The second and third conditions depend on the first condition. If there’s no utilization of an invention, then it cannot possibly satisfy public health needs, say, or access for public use. Let’s look then at the first condition. Bayh-Dole provides a definition of practical application. It is clear that the policy mindset continues to be shaped by the idea that there may be a gulf between reduction to practice and a form of invention suitable to uses that result in benefits to the public. Here’s 35 USC 101(f):
The term “practical application” means to manufacture in the case of a composition or product, to practice in the case of a process or method, or to operate in the case of a machine or system; and, in each case, under such conditions as to establish that the invention is being utilized and that its benefits are to the extent permitted by law or Government regulations available to the public on reasonable terms.
The first part is similar to the Kennedy (and Nixon version of Kennedy) policy, but drops “to the point of” from “practical application.” Something, apparently, is lost–a detail involving time. The last part differs in more substantial ways. Here’s Kennedy:
as to establish that the invention is being worked and that its benefits are reasonably accessible to the public.
Here’s Bayh-Dole:
as to establish that the invention is being utilized and that its benefits are to the extent permitted by law or Government regulations available to the public on reasonable terms.
The Bayh-Dole definition is truly strange. Kennedy uses “worked.” Working an invention is not merely using it. To “work” an invention means more than just to use an invention. US patent law has no working requirement–there is no requirement, short of an antitrust concern, for a patent holder to use the patented invention or to allow others to use it. Here executive branch patent policy adds a working requirement as part of the consideration for allowing a contractor to retain principal or exclusive rights in a given invention made with federal support. The federal government could, as a condition of funding, require delivery of all inventions. That in some cases it chooses not to, under Kennedy, involves a basic bargain–the federal government will defer on ownership and instead take a non-exclusive license to practice (make, use, and sell) the invention and accept the contractor’s commitment to, within three years of an issued patent, working the invention–either by releasing it on reasonable licensing terms to all or by using the invention to provide benefits to the public.
Here is a discussion of working requirements:
Broadly speaking, to “work” a patent is to practice, in some manner, the patented invention within the country that issued the patent.
“To practice” a patented invention again means more than just to “use” it:
While patents are seen as a means to create incentives for inventors to share their ideas, working requirements are intended to mitigate the exclusivity of patent monopolies by
requiring the patent holder to disseminate its invention into the local market.[6]
That is, working an invention, practicing it in the country that has granted the patent, means to disseminate the invention, or products based on the invention, or products incorporating the invention, into “the local market.” There’s no requirement that such dissemination must be commercial, or mass-produced, or directed at consumers. Such efforts might well meet the expectations of a working requirement, but they are not the exclusive objective of a working requirement. A working requirement might also be met by royalty-free, non-exclusive licensing of the invention–so long as the invention is available for use by others, it is being worked by the patent holder, even if the invention is not actually being utilized at some given point in time.
And the objective of the working requirement aligns with the broad purpose set out in the US Constitution, “to promote the progress of the useful arts”:
The patent holder thereby imparts knowledge and skills to the local community, enhances economic growth, supports local manufacturing, and promotes the introduction of innovative new products into the local market.[7]
While publication of an issued patent does teach the claimed invention in some manner, the progress (or the diffusion, distribution, as the word was used) of the useful arts is advanced as well by the actual implementation of the invention for public benefit. That is, by working the invention, not merely publishing an account of the invention through the patent literature. We will get to soon enough the problem of whether a working requirement involves just one instance or applies to each instance of an invention claimed by a given patent. Just hang in there. No one reads this stuff anymore, but it matters.
Thus, the Kennedy patent policy imposes a working requirement as one of the conditions under which the federal government is permitted to allow a contractor to exercise the right to exclude others. Given that an issued patent is a publication of the invention, it is clear that the Kennedy patent policy did not see mere publication as a sufficient public benefit. The idea of “working” the invention means that the invention is distributed. When the Kennedy patent policy then writes:
the invention is being worked and that its benefits are reasonably accessible to the public.
There are two things going on. First, distribution of the invention, at least in some instances–as product, as licensed–and second, benefits of the invention are “reasonably accessible” to the public. That “its” is a problem. An invention doesn’t just throw off benefits. The “its” in “its benefits” draws in the entire construction “invention is being worked”–that it is the benefits of the working that are reasonably accessible to the public. This makes some sense–it is not enough that a contractor strews the landscape with product or offers licenses (even on reasonable terms)–but that in addition, there must be benefits to the public for such activity, and that those benefits are reasonably accessible (and not, say, just asserted as benefits though no one drawn from the public could possibly access any such benefits). Even a benefit accessible to a select few is not a benefit accessible to the public.
Again, this Kennedy provision and definition as far as I can tell was never used. Our effort then is to use this provision to help us think about what ought to be the case, what ought to be in policy, what ought to be used. That’s important, especially since what is happening now with public supported inventing is so abysmal, so Baltimore as it were.
Now look at the Bayh-Dole definition again:
as to establish that the invention is being utilized and that its benefits are to the extent permitted by law or Government regulations available to the public on reasonable terms.
Bayh-Dole abandons the “working requirement” language and replaces it with “utilized.” Well, “utilized” has its own ambiguity–does it mean “use” as in one of the substantial rights of the owner of an invention, as in “make, use, and sell” Or does “utilized” have a common meaning “to put to use, to turn to profitable use, to make useful” or some such. It is difficult to tell. Again, since an invention is not a thing, we might add that it is not at clear whether utilizing a single instance of an invention is sufficient to establish practical application for all the instances claimed by a patent on that invention, or whether a contractor should expect to retain exclusive rights for only those instances that the contractor or an assignee or licensee utilizes.
The second part of this provision is simply bizarre. The provision imagines that some benefits of utilizing an invention may be prohibited by law or regulation from being made available to the public on reasonable terms. In the general case, Bayh-Dole’s definition here imagines that an invention might be utilized but law or regulation prohibits the public from realizing a benefit, or prohibits that benefit to be available on reasonable terms.
What is going on here? It would be odd if this is just an “oh, gosh, we shouldn’t have to do it if there’s bothersome government regulation.” Perhaps for some research inventions there might be defense Classification–in which case the inventors might not be able to talk about their invention, let alone make the benefits of its being utilized available to the public. But this would appear to be a tiny nit that might warrant a sentence in 35 USC 203 regarding march-in not applying in such circumstances rather than being treated as a broad, ambiguous limitation on the definition of practical application. It’s entirely possible, given the origin of Bayh-Dole with NIH’s patent counsel, that this qualifying provision is aimed at preventing march-in for pharmaceutical inventions that might have requirements for how they can be distributed as products, as well as any delays that might arise as a result of the FDA approval process to demonstrate that a proposed therapeutic has the benefits claimed for it and is safe to use, at least relative to the advantages of those benefits. Again, this is stuff that should have been laid out expressly in the march-in provision, not buried strangely as a qualification in the definition of practical application.
Now let’s deal with the third big change in Bayh-Dole’s definition of practical application–“available to the public on reasonable terms” instead of “reasonably accessible to the public.” Sure, there’s some common elements to both constructions. The words indicating reasonableness and public are in both. But “accessible” is not quite “available” and “reasonably accessible” is not quite “on reasonable terms.” If the Kennedy patent policy was clear on this aspect of the definition of practical application, then there would be no need for Bayh-Dole to make any changes. One would think then that Bayh-Dole either addresses an ambiguity in the Kennedy definition or is directed at a different practice.
Let’s work this a bit. “Accessible” means “easily obtained.” That has to do with distribution and supply meeting demand. If someone wants the benefit, they may easily obtain it. They may have to pay a premium price, but that’s not an issue for “accessible.” “Available on reasonable terms,” by contrast has to do with the terms on offer–who may acquire, price, stipulations on use, requirements to acquire other things along with the benefits of the invention, and the like. Something is “available” when it is on offer, it is possible to obtain it. “Accessible” and “available” share the idea that a benefit to the public is on offer and possible. “Accessible” though moves toward supply; “available on reasonable terms” moves toward price and other conditions of the transaction. The use of “reason” words–“reasonably accessible,” “available on reasonable terms”–indicates that there are degrees of accessibilities and of terms, some reasonable and some not.
What’s left entirely unaccounted for, however, is who decides what is reasonable. The reason words are both directed at benefits that the public should receive–one would think then that the public would determine what is reasonable. But that’s not all that likely. We might think then the federal agency determines what is reasonable. After all, Bayh-Dole delegates the determination for march-in to the agency that funds the research (and that in itself is really strange–what does the agency that funds nonprofit “basic” research have as specialized knowledge regarding utilization of inventions, markets, and the like? What forms of market enforcement do they have? No–makes no sense for Bayh-Dole to delegate to funding agencies anything having to do with march-in. But, of course, if Latker at the NIH wanted to make sure that he (or others at the NIH) got to decide when to march-in (i.e., never), then it would be a good idea to make this sort of delegation, so the NIH could run its own pipeline of inventions to pharma without interference from anywhere else in the executive branch).
Thus, if the funding agency does not march in, then we might expect that the agency has determined, at least with regard to practical application, that the terms of availability are reasonable. Bayh-Dole has no option for public input on march-in or practical application. Folks can protest all they like, but Bayh-Dole effectively shuts them out. The Supreme Court in Stanford v Roche found the lack of third party protections as potentially “troubling” but for the Court’s finding that Bayh-Dole narrowly applied only after a contractor had acquired ownership of an invention by conventional means. Perhaps the Court would also find troubling Bayh-Dole’s delegation to federal agencies making grants of matters of public availability on reasonable terms of resulting downstream inventions without any public input or right of appeal and without any obligation to enforce the march-in provision as placed in any given funding agreement.