Early on, federal research support was debated in terms of a dichotomy between procurement and subvention. As a procurement agent, the government purchased research services and the things that those services created. The government paid contractors to do work and deliver the results for government use and management, inventions and patent rights included. By contrast, as a subvention agent, the government provided “grants-in-aid” to support research without any particular focus on something that the government needed for its use. The idea behind subvention was that the conduct of the research served a general public purpose–to advance the frontiers of science, to find new ways of improving people’s health, to discover new things about the world. These were general goods, and there were big questions about how the government should get involved, how government funds might bias research, how the government should choose who to support, and the formal instrument for that support.
There were plenty of options–the government could gift money to individuals, such as in the form of scholarships or fellowships. Or gift money to universities, in the form of block grants, to be used however the university decided. Or the government could fund research projects–proposals for research, in which the statement of work rather than the potential of the individual became the primary focus. And the government could fund as a gift, or a subvention contract, or as a procurement contract. Today, of course, most everything coming to a university from the federal government is a subvention contract, a “funding agreement” that carries a host of conditions for how money is handled, how findings are reported, and within all this apparatus, how intellectual property and data are handled–including patentable inventions.
In the debates on how the federal government should support faculty research, the idea of the contract to support a project came to dominate. Thus, subvention-funded research came to look a great deal like procurement-funded research, except that in subvention research the work was specified by the investigator rather than by the government. There were still deliverables in subvention research–it’s just that the expectation (at least the administrative theory anyway) was that the deliverables would go to the public in addition to going to the government. Put another way, the researchers and their institution were not “in it for themselves.” It is sorting out what it means for inventors and their host nonprofit institutions to make inventions available that lies at the foundation of public policy for the government’s subvention-funded research.
In subvention research, the government would receive reports of the research findings and, like the public, could use the information in those reports any way it chose. For inventions, then, the government expected at least a royalty-free license to practice any invention. Similarly, the baseline expectation was that the public, too, would receive such a license unless there was some reason why the public interest was better served through the use of the patent system, in which case the license would still be non-exclusive, but now with a “reasonable” royalty requirement (as distinct from a monopoly-based royalty requirement). Much of what becomes the Bayh-Dole apparatus has to do with ending the policy debate with regard to what constitutes a basis on which a subvention-funded invention should be patented, and if patented, when non-exclusive licensing should be abandoned in favor of exclusive licensing.
I don’t know if there ever has been a policy debate regarding when exclusive licensing might be deferred in favor of speculating on the future value of the patent monopoly, by assigning patents on subvention inventions–yeah, posi–rather than licensing for use. Much university patent management practice presently relies on speculative assignment of patents. One way is through exclusive licensing deals that are in fact assignments (because they grant all substantial rights in an invention to the exclusive licensee, including the right to enforce the patent). Another way is by assigning an invention to a startup that the university owns, and hoping that the startup is acquired by investors or another company. In either case, the exclusive license agreement (=assignment) is acquired along with the startup, and thus the value realized is primarily due to the future value of the patent, not the value of products created under the authority of the license.
There appears to have been no debate on when speculative use of patents might be appropriate in subvention inventions because it was outlandish to think such thoughts. Research findings were to be used. Patents were to be used only when necessary. Exclusive licenses were to be used only when dedication to the public or non-exclusive licensing would fail to create the conditions under which a subvention-funded invention might be used with public benefits. And speculating on the value of patents just never came up.
The idea that the patent monopoly itself might be tradeable worked against the idea that the point of the patent was to induce private “risk capital” to support what otherwise would not get developed so quickly or so well. Once the thing was developed, so the subvention thinking went, the developer ought to get a fair return on its “risk capital” and after that things should open up for all to practice, with some form of “free competition”–either the term of the exclusive license ended, or others were sublicensed to practice in competition with the first licensee on reasonable terms. The risk capital that was to be called forth was in the service of a public enterprise, to be used to develop a product and having done so, to retire the debt, gain a “reasonable” return on the investment, and then dedicate the product to a private, competitive market.
The opportunity presented by subvention funding was for private control of posi to attract private risk capital to perform a public service–to make something available to the public and to industry generally. The opportunity was limited by this federal public covenant to develop something more quickly than otherwise and then release proprietary claims once that work was accomplished and one had recovered one’s investment. That goes for both the university (investing in patents and the cost of patent administration to pursue this public covenant) and for any private developers of the invention (investing in development). The patent license agreement was to represent a mutual commitment to this public purpose. The federal patent policy in turn was to establish the conditions under which such mutual commitments might take place between university inventors, university patent managers, and private investors and companies willing to serve the public interest.
This idea of a public covenant is distinct from a competing idea, that of exploiting the patent system in any way possible to extract value from patents, short of illegal antitrust behaviors. In this way of thinking, a patent could be used in any number of ways that might have value–to suppress all use of an invention (in favor of other products, or to control the timing of introduction of a new product); to wait until others had adopted the invention and then sue them for infringement when they have no viable alternative but to pay; to use the invention only for one’s own benefit and never allow anyone else to practice during the term of the patent; or to trade on the value of the patent property for others who might find even better value from one or more of these practices. Of course, one might also license a patent for “commercial development” of products, and here maximum value (so the thinking goes) is obtained when the licensee can price a product for what the market will bear, for the full term of the patent–that is, using the patent monopoly in any legal way short of antitrust law requirements. Any of these uses of a patent are legal, even reasonable in their way, and one might argue that the patent system permits these uses, and since the patent system is in the public interest, so also these uses also serve the public interest, even as patent owners pursue their own best self-interest.
If this idea that the use of the patent system is a good thing is not sufficient, then university patent brokers added their own public covenant to it–that when a university made money from its patents, that money would go back to the university for public uses, as well as to be shared with inventors to encourage other inventors to come forward and contribute their inventions to this same program of patenting, called “technology transfer.”
This public covenant–that money from patents is in the public interest because it benefits universities and it is in the public interest for universities to have more money–is decidedly unlike the public covenant that has run with executive branch patent policy for research funding for the past 75 years or so. In the federal public covenant, patents are used (via non-exclusive licenses) to create a competitive environment for product development. If no such environment forms or could form, then the patent owner and a licensee could mutually agree to develop a product anyway, and then dedicate that product and the patent rights to the public or to competitive uses (again, via non-exclusive licenses).
In the nonprofit private covenant, any use of a patent is fine, so long as a nonprofit makes money from the effort and there’s a repeated statement that the intent of any patent management is to benefit the public. In the nonprofit private covenant–one might call it a faux public covenant–a patent is a patent, and exploiting that patent any way one can for value at a nonprofit is limited only by the constraints on the nonprofit by its own charter, mission, tax exemption, and delegation of authority. If the federal government allows a university to manage patents exactly as any other patent owner might manage patents, then anything is fair game–it’s just a matter of will in senior administration whether to sue for infringement or do a big monopoly deal with a company with the prospect of making a lot of money.
The Kennedy/Nixon executive branch patent policy–bipartisan in a sense–expressed the federal public covenant in the form of government ownership of patents unless a contractor had a commercial position and capability to develop a given invention. When the government owned, it dedicated inventions to the public or licensed them non-exclusive (and royalty-free). When the contractor with a commercial position owned, it had three years from the date a patent issued to develop the invention, and the government had a royalty-free license to practice (including selling and having sold) the invention. After that time, monopoly rights in the invention were to be released unless the contractor could persuade the government otherwise. For all other contractors–ones without commercial positions or capability–they had to make a case that their management of inventions would result in practical application with public benefits on reasonable terms. Even then, they had only a limited time to exploit monopoly patent rights to get to practical application. And of course, the government also got its non-exclusive license to practice.
The IPA program, when revived at the NIH under the Kennedy/Nixon patent policy, worked to end-run the federal public covenant by giving select nonprofit organizations outright the right to manage patents on subvention inventions. The IPA presented the usual apparatus for government license and non-exclusive licensing but extended the term of exclusive licenses, changed the date on which the term started (from date of patent issue to date of the license or date of first commercial sale), and most importantly created a pathway directly to exclusive licensing. And every university working with a new IPA went right to that exclusive licensing pathway, especially for inventions directed to the pharmaceutical industry.
Bayh-Dole, when the revived IPA program was shut down, extended the end-run around the federal public covenant by suspending executive branch control over federal patent policy, reducing and finally (in an amendment four years later) eliminating term limits on exclusive licenses. Most importantly, Bayh-Dole failed to provide any requirement that federal agencies enforce the requirements of the standard patent rights clause that Bayh-Dole authorized and provided no form of public appeal that the federal government enforce the patent rights clause. And finally, Bayh-Dole (as amended) made all reports of invention use a government secret. Once the patent rights clause need not be enforced, with no prospect for the public to see how patents were being used by universities, and no opportunity for the public to appeal uses that went against the federal public covenant (e.g., nonuse of inventions, monopoly pricing, unreasonable terms, failure to promote free competition, unduly affecting future research and discovery), it has been an easy step to abandon the federal public covenant altogether and replace it by the idea that the patent system is a good thing in the hands of nonprofit patent administrators, and patents on subvention inventions are ordinary patents to be used to create value for nonprofit organizations. The claim is that any such use is necessarily in the public interest, and thus there is no particular reason for the public to inquire further into the nature of that use. Any such inquiries are simply efforts to unfairly compete, to disrupt public efforts for private gain, and to prevent federally supported research from reaching its full potential through exclusive licenses, which necessarily must be kept secret.
Thus, present university patent practice for federally supported inventions is built on the ruins of the federal public covenant. In its place, and using much of the same vocabulary, is a nonprofit private covenant that recognizes no restrictions on the exploitation of patents on subject inventions, or on the money that comes from this exploitation, other than that a portion must be shared with inventors as an expense. The nonprofit private covenant argues that maximizing the value of those inventions that might have financial value is the expression of the public interest–not just in patents, not just in research inventions, but in the purpose behind federally supported research as a whole. The inventions and discoveries of that research are, in many ways, the fundamental expressions of the outputs of that research–the tools, the protocols, the processes, the prototypes, the applications. If claims to patent rights secure (or encumber) these outputs, then public policy that permits those claims has control of a vast portion of the federal research enterprise.
It is as if, instead of granting a right of way for a railroad through uncharted territory, public policy had got things reversed and ceded to private control everything uncharted that could be claimed, leaving only narrow bands of public domain for those things not patentable (and hence not new or useful) and for those things not worth patenting. Even for things not worth patenting, university IP policies often still claim “ownership” (but without any clear idea about what exactly is “owned”). In the nonprofit private covenant, the “endless frontier” of science has been given over without public oversight to patent brokers to exploit, using ordinary patents and any management method available (and arguably legal) for the purpose.
Of course, which patent covenant controls is a matter of public policy. At present, the nonprofit private covenant controls, and the federal public covenant, while present in Bayh-Dole is not enforced. In a strange reversal, the “exceptional circumstances” referred to in Bayh-Dole, under which a federal agency may introduce a variation on the default requirements for the standard patent rights clause, are in fact the circumstances of subvention research funding, subject to the federal public covenant. Bayh-Dole reverses the defaults. The default requirements for the standard patent rights clause are those that would apply to a federal contractor with a commercial position and the capability to develop an invention. That’s the effect of having Bayh-Dole apply to small companies as well as to nonprofits. Small companies were given the incentive to develop commercial products, and thus so also were universities and their affiliated nonprofit patent brokers. There was no outcry when Bayh-Dole was extended by President Reagan, by executive order, to all companies. Why? Because the standard patent rights clause was written as a default for them. Other than some clunky apparatus for reporting inventions, Bayh-Dole is nothing more than the Kennedy patent policy, section 1(b), stripped of its qualifiers and federal public covenant but for the government license:
Everything else, including subvention funding directed to faculty-proposed and -led research hosted by universities, is an “exceptional circumstance” in Bayh-Dole. It’s not that these circumstances are rare–it’s just that they are exceptions to the default set out by Bayh-Dole. The primary exceptional circumstance–and this is the great, clever irony of Bayh-Dole–is subvention research at nonprofits with a federal public covenant. That’s what Bayh-Dole undid by making it appear that universities should, by default, operate as if they were commercial concerns, claiming inventions as if faculty and students were employees, claiming inventions as if the use of public resources necessitated corporate ownership (and thus federal subvention becomes converted into university procurement), and exploiting patents as ordinary assets to be deployed however the university might best make money, however any patent owner might best make money.
This, then, is the prevailing public policy. There could be a federal covenant. Bayh-Dole authorizes the creation of patent rights clauses that vary from the one specified in the law. It’s just that federal agencies have to be directed to do so, to create under the “exceptional circumstances” provision a patent rights clause that implements the federal public covenant for subvention funding hosted by nonprofits, including universities. The NSF, in particular, should take the lead in doing so, for all non-engineering work. Under a modified patent rights clause, universities would be forbidden to treat subvention funding as a university procurement (thereby eliminating all obligations by inventors to assign inventions to the university that hosts the research). Inventors would have the opportunity to assign inventions to the government, to be dedicated to the public or licensed non-exclusively, royalty free (and hence made available to the inventors as well on these terms), or if the inventors chose, they could accept a federal public covenant and develop their inventions privately, either directly or through an invention management firm that accepted the federal public covenant.
It’s a matter of what policy the federal government will endorse. Bayh-Dole allows multiple federal policies on the matter. The issue is whether federal agencies in dealing with any given contractor might assert conflicting patent policies. That is unlikely to happen if a given federal patent policy is made government-wide. Thus, if for subvention funding, a policy restoring the federal public covenant is used, it will be used in all federal agency subvention funding with nonprofits.
Since most universities receive almost all their federal funding in the form of subvention grants (procurement contracts are typically restricted to special units or the operation of government-owned, contractor-operated federal labs), it is unlikely that a university will encounter conflicting requirements between the standard patent rights clause in Bayh-Dole (rarely appropriate for university support) and an “exceptional” patent rights clause specific to subvention funding at nonprofits (which is almost always appropriate). Even if an “exceptional” patent rights clause is developed and made optional–that, say, the investigator can choose which clause to operate under, standard or exceptional for subvention–the exceptional clause can carry a statement of precedence, so that if any standard patent rights clause conflicts, the exceptional clause prevails.
All this can be done without any changes to Bayh-Dole. The architecture is already there, in the law. All that’s necessary is the political will to re-assert the federal public covenant for inventions made in subvention funding. If the political will is not there, then we might expect further efforts to make Bayh-Dole simply require the institutionalization of all inventions made with federal subvention funding. For that, almost all of the apparatus left over from the abandoned federal public commons may be eliminated–use reports, march-in rights, limitations on licensing, and for nonprofits limitations on assignment and use of licensing income. All that’s left is Kennedy’s 1(b), but without the qualifications. When the federal government provides support, the research host owns the inventive results, subject to a government license. Any patents that result are ordinary patents, with the only requirement that the patent carry a federal funding statement. The rest of Bayh-Dole’s apparatus is, in this view, rot.
If one thinks there is some merit to the federal public covenant, that the huge expansion of research funding should not create a land rush for speculators on the future value of patents on research outputs, then it is Bayh-Dole’s fundamental gesture that’s rot–that universities should operate like corporations with regard to patents, but like tight-fisted, monopoly-minded corporations, seeking “commercialization” partners that share their views, seeking returns from the monopoly value of each patent on a subject invention. If a federal public covenant still matters–I think it does–then it’s time to exploit Bayh-Dole to restore that covenant and eliminate the use of the standard patent rights clause in all subvention funded research hosted by university and other nonprofit organizations.