Free Competition and Enterprise in Bayh-Dole

One reason that Bayh-Dole is so difficult to parse is that there are multiple levels at work. It’s easier to just say that Bayh-Dole gives universities ownership of inventions made with federal support, and they can do anything they want with those inventions because whatever they do is done with good intentions and in the public interest. That sounds simple and wonderful. But it’s not the law, and it’s not true, and universities can’t show that what they are doing with inventions is at all in the public interest. Mostly, they acquire patents and sit on them, dragon-like. Every now and then someone does a huge financial deal and these deals inspire all the other universities to try even harder to patent everything. No one bothers to look at why there is such value changing hands, and so rarely. Is it the monopoly pricing? Is it speculative exuberance? Is it misrepresentation of what’s been done and what’s possible?

Let’s look at just one part of Bayh-Dole to show how it has gone so horribly wrong. When Bayh-Dole was passed, it contained the following elements with regard to competition:

35 USC 200

It is the policy and objective of the Congress to …  to ensure that inventions made by nonprofit organizations and small business firms are used in a manner to promote free competition and enterprise

35 USC 202(c)(7)(B)

(c) Each funding agreement with a small business firm or nonprofit organization shall contain appropriate provisions to effectuate the following: …

(7) In the case of a nonprofit organization…

(B) a prohibition against the granting of exclusive licenses under United States Patents or Patent Applications in a subject invention by the contractor to persons other than small business firms for a period in excess of the earlier of five years from first commercial sale or use of the invention or eight years from the date of the exclusive license excepting that time before regulatory agencies necessary to obtain premarket clearance unless, on a case-by-case basis, the Federal agency approves a longer exclusive license….

These provisions go together. The general statement of policy and objective requires patents on subject inventions to be used to promote free competition and enterprise. This is a general restriction on how patents on subject inventions are used. This restriction is placed in federal patent law, not in executive branch procurement policy as was the case with the Kennedy and Nixon executive policies. The definition of “subject invention” is also placed in federal patent law, not in a master agreement between a federal agency and a university. The statutory definition applies regardless of what happens in any specific funding agreement. The statement of policy and objective is law quite apart from who happens to own a subject invention or anything else that Bayh-Dole might require. And a statement of policy is not the same thing as a statement of purpose. A policy binds the policy-maker. A purpose envisions. Federal policy states that patents on subject inventions are to be used to promote “free competition and enterprise.” What does that mean?

Well, the original 35 USC 202(c)(7)(B) gave us an idea. Nonprofits were not allowed to grant exclusive licenses for longer than eight years–about half the term of a patent at the time, but the term could be limited to no longer than five years from the date of first commercial sale or use. And of course, the term could be extended for regulatory approvals (a concession to the drug industry) and could be extended by a federal agency (if the agency determined that a longer monopoly was in the public’s interest).

Why such a restriction on exclusive licensing? The restriction in an even more restricted form shows up in the Kennedy patent policy from 1963. There, a contractor who was permitted to obtain “principal rights” in an invention made with federal support had three years from the date of patent issue “to bring the invention to the point of practical application.” And “practical application” had a specific meaning similar to that given in Bayh-Dole–use with public benefits on reasonable terms. That is, the purpose of allowing a contractor to deal in the principal rights of an invention made with federal support was to realize a public benefit. But why the time limit?

We get a similar treatment in the revived Institutional Patent Agreement program. There–again, the focus is universities and other nonprofits–exclusive licenses are restricted to “three years from the date of first commercial sale . . .  or eight years from the date of the exclusive license, whichever occurs first.” Even then, an exclusive license could only be used after a licensee determined that (among other reasons) an exclusive license “is necessary as an incentive for the development of the invention.” The IPA ignores the Kennedy patent policy requirements and pushes the term of exclusive licenses from three years to five or eight. But the basic gesture remains–the nonprofit controls a patent for a limited time as an exclusive right. Any extension is subject to federal agency approval.

One reason for the restriction on exclusive licensing is evident from the Kennedy patent policy. A federal agency may permit a contractor to acquire rights in an invention made with federal funds “in exceptional circumstances” when (among other reasons) “the acquisition of such greater rights is … a necessary incentive to call forth private risk capital and expense to bring the invention to the point of practical application….”

It’s worth dwelling on this point for a moment. A federal agency funds research with a contractor. The contractor’s personnel invent. The federal agency can acquire the invention and then fund work to develop the invention. Or the federal agency can allow the contractor to develop the invention, using private funds. If we were to put this into a diagram, it might look like this:

The government acquires the invention and then spends its own money (perhaps with other contractors) to bring the invention to the point of practical application, and then releases the invention for public use. Anyone may make, use, and sell the invention. For instance, the Salk vaccine for polio.

But let’s say that the federal agency does not, for whatever reason, decide to develop an invention with its own funds. It could release the invention for public development or it could use the patent system to “call forth private risk capital”:

Private capital develops the invention, so to “recognize the equities” of those providing the funds, the government may permit the contractor (or assignee or licensee) to have a limited monopoly on the invention, during which time the contractor may recover its investment, plus a reasonable profit as the reward for working with the government to achieve a government purpose, a public purpose that takes precedence over private interests, but does not exclude private interests. Once the contractor has recovered its investment and a reasonable additional amount, then things return to a state of competition and anyone may make, use, and sell the invention.

This is the theory. But of course it doesn’t quite work this way. Most inventions, to be developed, end up requiring other inventions, and so it is the rare case that a single invention, developed, becomes a product. No, a product will likely involve multiple inventions, and so the “private risk capital” that funds those additional inventions can establish patent positions in them as well, effectively blocking competition from making the same product once the exclusivity in the federally supported invention has expired. The invention might be used for other purposes, but then we are back to not having the invention developed to the point of practical application–but now development has to go in another direction. What else might a polio vaccine do, besides prevent polio?

Despite the limitations of the theory, it is easy to see the public policy gesture. Federal agencies can permit contractors to hold patent rights that the agencies otherwise could contract to acquire, so long as the contractors develop the underlying inventions (and not just sit on the rights looking to exit with a big payment), and that if a contractor is successful in developing an invention, it has a limited time to recover its investment (sell at whatever price the market will bear), and after that, it will compete with everyone else for sales (and prices will, in theory, fall below those of the limited monopoly). In essence, such government permissions amount to a shadow patent system, in which the term of exclusivity is limited, and after which a contractor may still charge for licenses, but has to license on reasonable terms to everyone who wants a license–that is, one may still profit from licensing, and have the advantage of say, the 4% royalty that the competition must pay, but the contractor cannot continue to exclude others, and therefore cannot sue to enforce the patent, other than to secure non-exclusive licenses. That is, a contractor is obligated to offer non-exclusive licenses on reasonable terms first.

Now come back around to Bayh-Dole. We see this same structure at work. A nonprofit contractor is expected to license a subject invention to one or more companies. If an exclusive license, then to a single company. That company, if not a small business, then has a limited time to develop the invention and use it or sell product as it will to recover its costs of development. After that time–five or eight years–the license becomes non-exclusive, and we have the “free competition and enterprise” set forth in Bayh-Dole’s statement of policy and objective. But for the efforts of the exclusive licensee, there would not be a viable implementation of the invention. But now there is, and so there can be competition to see who can best use it or sell product based on it. The patent system has created the conditions for “free competition and enterprise.” A company has willingly served that role–accomplished a public purpose on behalf of the federal government by developing an invention with its own capital, and accomplished a second public purpose by creating the conditions for competition in a term shorter than that of an ordinary patent.

There are additional elements having to do with how the government uses its own non-exclusive license to practice and have practiced, and how the government might intervene in a contractor’s use of patent rights for public or regulatory or treaty matters. These all add sophistication to the problem. But let’s keep to the basic theme.

Again, this is all theory, because most products require multiple rights. A car has 300 or more patents. A printer has 50 or more. Even prescription drugs can involve 5 or more. Products are often platforms, and around each platform may be scores of patents that prevent others from developing anything similar, and scores more patents held by others that block the platform from applications or improvements in areas controlled by those others.

But now look what happened with Bayh-Dole. Four years after the law passed, folks were back with amendments. One of these amendments removed the restriction on exclusive licensing. Gone was the preference for small businesses (which, if they licensed exclusively from a nonprofit, did not have any restriction on the term of exclusivity). But more importantly, gone too was the mechanism by which the federal government acknowledged the use of private capital by providing a limited time under which the contractor could charge monopoly prices and so recover its investment. Any competition comes at the end of the patent term, not sooner (unless the competition is infringing, or the competition has been motivated to design around the subject invention and thus make it obsolete or its patent meaningless–and how is that an objective of Bayh-Dole?).

By removing the term on exclusive licenses, the amendment to Bayh-Dole left hanging the policy and objective regarding “free competition and enterprise.” We might read that statement of policy and objective to mean–use the patent system in various ways, with various restrictions on use of subject inventions, so that there will be “free competition and enterprise”–and one such way is by restricting the term of exclusive licenses. Federal agencies use the patent system to attract private investment, but part of the risk that investment runs is that it will recover its expenditures within the term set by the law. If an investor does not believe it is possible to recover the investment, then perhaps it will not make the investment. The same is true, of course, for inventions that run the full term of their patent–even 20 years may not be sufficient to invest, develop product, and recover the investment. And for those inventions, some other approach to investment other than monopoly control is indicated.

If restricting the term of exclusive licenses in subject inventions is one way that nonprofits can promote free competition and enterprise, and if Bayh-Dole has gone out of its way to mandate that the patent system be used so that nonprofits do indeed promote free competition and enterprise, then it cannot be the case that the patent system is just to be used as the ordinary old patent system. There’s no point, if that’s the case, in Bayh-Dole’s statement of policy and objective. If there’s no special conditions to promote free competition and enterprise, then the whole statement is superfluous–it has no statutory meaning. It is a puff of wind, an empty gesture.

When folks removed the restriction on exclusive licensing, perhaps they thought they were giving nonprofits more rights and a greater opportunity to profit from licensing deals. Take the situation with Cisplatin, an anti-cancer drug discovered with federal funding, subject to an IPA (just before Bayh-Dole), and developed under exclusive license at private expense. As the limited term of exclusivity approached, the university pushed to have the monopoly extended, even though profits from sale of Cisplatin had far and away exceeded the private costs of developing the drug. Why? Well, for the royalties, of course–but royalties from the exclusive licensee. The university could, of course, have offered licenses to any competitors seeking to sell the same drug (in theory, but perhaps with variations due to the exclusive licensee’s related patent positions), and collect perhaps just as much from all licensees combined.

There’s another problem with all of this–why should the university be so focused on the money, and not on the public benefit that might come from free competition? After all, if there’s no competition, even when there could be, then things will go on merrily with a non-exclusive license. And if there’s competition, then the university makes money from everyone taking a license. So the university can be happy. But there’s the chance (in theory) that having even a second supplier of the drug could drop the profits so much that neither company can afford (or wants to bother) to make the drug, and thus the drug is no longer available to the public. Sure, in theory that could also happen–but what are the conditions for that? The conflict for the university licensor, when it comes to restrictions on exclusive licenses, is between the money from monopoly positions (which sounds nice to patent brokers) and the public benefit arising from competition–multiple suppliers, diversity of product and services, and presumably lower prices and competitive improvements.

To some degree, then, the restriction on exclusive licenses is actually an institutional conflict of interest provision. Take some share of any monopoly deal, but don’t allow monopoly deals to run the full term of the patent. Do something in the public interest as well–license non-exclusively, license to see broader use, license so that others develop capability with the invention while it is still selling, license so that prices fall, license so there’s competitive innovation, license so that the public has a choice who they deal with. What was lost in the 1984 amendment to Bayh-Dole was this implicit institutional conflict of interest requirement.

But here’s the interesting point. Why you read all the way to here. Just because the restriction on exclusive licensing was removed as a requirement for the standard patent rights clause does not mean that the statement of policy and objective was made meaningless. At the same time that folks were amending the nonprofit restrictions on exclusive licensing, they also changed the policy and objective statement on free competition and enterprise:

to ensure that inventions made by nonprofit organizations and small business firms are used in a manner to promote free competition and enterprise without unduly encumbering future research and discovery

This is very odd. Before, nonprofits could offer only limited time exclusive licenses. After that, there was free competition. Now nonprofits can offer exclusive licenses that run the full term of the patent. But how does either form of licensing affect “future research and discovery”? Does the added language place a new restriction on licensing–but now general to both nonprofits and small business firms? Or does the added language create a more liberal use of patents on subject inventions–so they may even encumber future research and discovery, so long as that “encumbering” is not “undue”?

Does the language here mean to say “nonprofits and small businesses may also exclude other research organizations from using a subject invention, if doing so promotes the use of an invention under an exclusive license?” Or is it more direct to say, “nonprofits and small businesses may exclude research uses if doing so helps them make more money from an exclusive licensee”? That is, if nonprofits get more money, then whatever they are doing with licensing cannot be “undue”? This all seems strange.

We might explore another route. The “without unduly” construction clearly limits what goes before, but does not change the main sense. This is a restrictive clause, not a change to the sense of the objective. Thus, the objective stands, but now with the added provision that whatever could be done before now also must meet this new restriction, which requires licensing decisions to be responsive to the needs of research and discovery–not just at other nonprofits and small businesses, but generally, broadly. Read this way, the restrictive clause turns out to be an expansion of the main sense. It’s not enough that licensing practices promote free competition and enterprise–these licensing practices also cannot “unduly encumber” future research and discovery, wherever that activity might take place. That sounds a lot like a requirement to grant upfront a general license for research and discovery unless there is a bona fide reason not to. There’s a good argument that to comply with Bayh-Dole–and here we are outside the standard patent rights clause, looking directly at how patent law has been changed to reflect the public interest in subject inventions–contractors have to provide upfront a research license to any and all that want it. This is similar to the requirement in the standard patent rights clause that the government gets its license to practice and have practiced, but here there is no quid pro quo of license in exchange for rights. It’s just a condition placed on the patent property right in a subject invention, no patent rights clause provisions required.

Further, the general statement of the “free competition” requirement is unchanged–it is as broad as before, even with the requirement to grant a research exception. Just because the restriction on exclusive licenses has been removed does not mean that the clause is otherwise now inoperable. What might be the general class of obligations or limitations in a patent property imposed by this clause on a nonprofit or small business with regards subject inventions? We might argue–and this will sting–that the requirement to limit exclusive licenses was itself a restriction on the broader obligations stated in the policy and objectives at 35 USC. That is, agencies were authorized by the law to allow a restricted exclusive license to meet the expectations of the statement of policy and objective. But for this allowance, then the management of subject inventions is not protected from the full import of the statement of policy and objective.

This is a reversal in things. Stuff that looks restrictive is in fact permissive–both with regard to licensing exclusively as well as to licensing for research. What happens when Bayh-Dole is amended in 1984? Here’s a possibility. To promote free competition and enterprise, in general, one must license a patent non-exclusively. That’s the base line. If one determines that there will be no use without some investment of capital, then one may license exclusively without term restrictions, but that license must not include research and discovery uses except in exceptional circumstances. One must grant research licenses, and must not demand a grant-back encumberance of rights in any discoveries (or inventions, or patents) that might arise in such research. But one is still required to promote free competition and enterprise, and if that’s not met by the term of exclusivity, then it must be met in some other way. We are not talking here about product availability–that’s march-in stuff. We are not talking about promoting use of inventions–that’s a separate consideration with its own statement of objective. How on earth can a patent owner promote free competition and enterprise, having granted an exclusive license for the life of the patent?

There are answers to this question. It’s not rhetorical.

  1. One may require an exclusive licensee to grant sublicenses to all qualified applicants. That would meet the objective. The exclusive licensee undertakes the non-exclusive licensing program on behalf of the patent owner.
  2. One may require an exclusive licensee to cross-license the invention to others contributing inventions that permit the commercial use or sale of the invention as part of a product or service that requires multiple inventions. That would promote free competition–even though the license was exclusive for the full term of the patent.
  3. One may inform an exclusive licensee that if a product is not on the market within a given time, then the patent owner has no obligation to enforce the patent against infringers. The exclusive license stays in place, but infringers turn out to be “free competitors”–the only problem is that the patent owner foregoes licensing income from these free competitors–but in a way, isn’t that even better “free competition”–free of rights and free of payments. Free both ways!
  4. One may grant an exclusive license but only for a limited “field of use”–such as, for use in a specific product and not in any other products. Others, then, might have an exclusive license for their products, so long as they differ in some material way from those of other exclusive licensees. Thus, for a pill, one might grant an exclusive license for the use of a compound, but for a liquid version, there’s another exclusive license, and for the suppository version, yet another exclusive license. Free competition in how the inventive compound is packaged and used. That works.

The point here is that there are ways to promote free competition even in the context of an exclusive  license running for the full term of a patent on a subject invention. When Bayh-Dole stated a restriction on the term of exclusive licenses, it established a safe harbor by which any patent owner of a subject invention could meet the requirements of Bayh-Dole’s statement of policy and objective. When by amendment that safe harbor was removed, a patent owner could–and can–still license non-exclusively or contribute the invention to a standard or license exclusively for a limited term. But if a patent owner of a subject invention licenses exclusively for the full term of the patent, then there must be some other provision, beyond an upfront research license for everyone, that meets patent law’s requirement set forth in the “free competition and enterprise” objective. The burden is on universities and other nonprofits, in particular, to demonstrate what that other provision is, when they license a subject invention exclusively for the full term of the patent.

Upshot of all this:

Bayh-Dole requires, as a condition of patent law, owners of patents on subject inventions to grant upfront a research license to all qualified applicants without demanding a grant-back that encumbers future inventions.

Bayh-Dole requires, as a condition of patent law, owners of patents on subject inventions to include some provision in exclusive licenses for the full term of the patent that promotes free competition and enterprise.

These are two further provisions of Bayh-Dole that universities routinely disregard. No doubt their IP counsel will reject the arguments here–and would do so even if the Supreme Court ruled against them, as in Stanford v Roche. But what gets me is that university administrators and their legal counsel don’t aim to promote use broadly anyway, even if not required by Bayh-Dole, even for inventions that aren’t subject inventions. Where was it that they lost their way? Why should Bayh-Dole have to teach them what they should do? Why on earth should I have to teach them what Bayh-Dole teaches them–and which they repudiate? Why is doing the right thing when it comes to patents now so very difficult for universities?


This entry was posted in Bayh-Dole, Freedom, Innovation. Bookmark the permalink.