The nature of federal research contracts
Let’s work through how Bayh-Dole might have been structured. We start with the nature of federal contracts. A federal contract is not quite like a conventional contract formed under state laws. The federal government stipulates the provisions of each funding agreement by means of regulations. For grants, there’s not much more. Contracts consist largely selections of alternatives provided by the regulations, with blanks filled in for specifics of each deal. For grants to universities, 2 CFR 200 sets out the requirements. Combined with the proposal that sets out the statement of work, approval by an officer of the federal agency, and acceptance by the university, we have a “funding agreement” in the form of a federal contract.
Here’s an early instance, from a Public Health Service regulation from 1963 (emphasis of the last sentence (shown here as not italic)) is in the original; I’ve added some blank lines for easier reading):
1. Formal reports of inventions. Department of Health, Education, and Welfare regulations (45 CFR, parts 6 and 8) provide that all inventions arising out of the activities assisted by Public Health Service grants and awards shall be promptly and fully reported to the Surgeon General.
In respect to inventions reported, the institution and the principal investigator agree either:
(a) To refer the inventions to the Surgeon General for determination, in accordance with Department patent regulations, of the ownership and manner of disposition of all rights therein and whether patent protection on such inventions shall be sought, and, if so, the manner of obtaining, administering, and disposing of the patents in the public interest; or
(b) where the institution has a separate formal patent agreement with the Surgeon General covering inventions deriving from Public Health Service support, to make a determination of ownership and disposition in
accordance with its policies as approved or as modified by such agreement. In no event, shall patent applications be filed on inventions reported under (a) above, without prior written consent of the Surgeon General.
These regulations form part of each PHS funding agreement with a university. Notice that the funding agreement is actually made with both the university and the “principal investigator”–both are involved. This practice follows a now forgotten issue of how the federal government should decide what individuals to fund. As Charles Kidd explains, deciding who to fund based on personal credentials created the appearance that some people were simply not good enough for federal dollars. Thus, folks came up with the idea that projects rather than people would be reviewed. Someone could accept that they wrote a poor proposal and might do better next time rather than face the judgment that they were inadequate as scientists. So the government funded people–principal investigators, the ones who proposed the work–by means of projects. You can see then in the PHS funding agreement requirements for patents that both the university and principal investigators are bound to the regulation. The principal investigator is getting funded–personally–on the basis of a review of the proposal submitted through the university, which must agree to release the investigator from official duties and provide the resources necessary for the work to be undertaken.
As is clear from the example, the PHS also requires assignment of inventions unless it has otherwise agreement (using, for instance, an IPA–though the PHS had suspended adding institutions to its IPA program by this time). Later in 1963, the Kennedy patent policy reconfirmed the PHS practices, providing authority for federal agencies to institute regulations that required assignment of inventions made with federal support unless the contractor had an established commercial position and capability and the research was not intended to produce a product for public use, or in an area that only the government could use (such as military weapons, atomic/nuclear energy, or space technology), or in an area that the government had the primary knowledge and responsibility (such as certain areas of health care), or were subject to requirements for public use by government regulation. Otherwise, a federal agency could allow a contractor to own inventions (called “principal rights” by the Kennedy policy) if the contractor could show that it would manage the requested inventions to achieve “practical application” in the public interest.
There’s a bunch of details in how all this operates, but let’s skip over them here to get to the main point. The Kennedy patent policy establishes the authority for federal agencies to require assignment of title by whoever has that title–investigator or university. When an agency creates such a regulation, it implements the Kennedy patent policy and forms a federal funding agreement with each grant recipient (individually and institutionally). The federal agency then has the right (under the federal funding agreement) to require assignment, or may waive that requirement (in keeping with executive branch policy) if the agency determines that the public interest is better served by the contractor managing “principal rights” in an invention.
In releasing claim to any such invention, federal agencies retained for the Government (federal government, state government, and municipal governments) the right to “practice and have practiced” each invention for Government purposes–make, use, sell; have made, have used, have sold. Further, federal agencies could require special provisions to limit how patents were exploited by contractors–especially the term of any exclusive license. If a contractor had not achieved practical application within three years of a patent issuing, then the federal policy default was that the monopoly right in the patent should be broken up by means of non-exclusive licensing. These conditions we can call the “public covenant” that runs with any invention that the federal agency releases for contractor exploitation.
The idea, then, was that the federal government should own inventions made with public support, and “dedicate” these inventions to the public unless the public interest might be better served by licensing or by contractor efforts to achieve practical application. Contractors, then, obtained rights in such inventions to serve the public interest in place of the federal government. Private “risk capital” might be be “called forth” to support such work, but only in the service of a public objective–to get an invention into use faster and with less government money than others, for the public benefit. Once that use was achieved, and the risk capital recovered with a reasonable return, the patent monopoly would be broken up and everyone would have competitive access to the invention. Private capital, under the public covenant, was offered in the service of public interest, with returns limited so that once practical application had been achieved, the invention could be dedicated to the use and benefit of all. That was the policy at least.
Advancing the frontiers of science
There’s another way to look at things. The government, led by Vannevar Bush, had established the idea that sponsoring research was a proper role for government, a proper use of public money. Science fumbled along, with some work done by industry but directed to industrial problems and profits; some work done in government but directed to the needs of government operations, especially the military. Support for science was otherwise lacking. There were some foundations, of course–The Smithsonian Institution, Research Corporation, and the like. But most public charities focused on health education and prevention of disease and did not concern themselves with advancing scientific knowledge generally. At universities and in scientific societies were people who studied things primarily to find things out, to observe the world with enough attention to observe what others had missed, what others had not thought to look for. They stumbled on things by accident–radioactivity, penicillin; they worked out crazy schemes that then proved true and potentially useful–relativity, Turing machines. It was to support these people that Vannevar Bush sought federal funds. If they had more funding, and stayed free to do what they were doing, then the frontiers of science could be pushed out more quickly, providing more opportunities to combine new with old and create things yet barely imaginable–communicating devices, cures for disease, digital watches with two-way communicators.
There were problems that came with the movement of the federal government into this area of independent scientific research. One was selection. Who should the federal government support? If money was given to people on the basis of who they were, then those not receiving money would be judged, by implication, to be unworthy of government funding–a pretty damning assessment. So the “project” was substituted for the “person” and people competed for federal support based on the project proposals they submitted. A second problem then became choice of project. How might the government choose among competing proposals? What proposals were “best”? What topics were “best”? What desired outcomes were “best”? Bush proposed that an independent, capable director lead an public research organization to decide.
The federal government, however, spread research support over all federal agencies, so that the agencies decided, and hence could expand their reach into the same pool of independent scientists that Bush hoped to assist. What’s best turned out to be “whatever it is that the government appears to want.” If the government formed committees of scientists to review proposals, then the best turned out to be “whatever it is that committees of scientists working for government approval (and for funding their own research later) appear to want.” Consensus science starts with consensus scientific review committees. If one is looking for the unexpected, for outliers, for things that once confirmed toss whole traditions of scientific theory out the door, then consensus selection of scientific research proposals is a poor way to go about it.
There was a third problem, that of accountability. How could the federal government know that scientists receiving funding were doing great work? How could the federal government know that money was being spent well and wisely? What if scientists just settled down to a newly created set of entitlements and went through the motions, publishing authoritative-sounding rubbish that no one could decipher or use? What if scientists sold their souls to industry and proposed research that industry otherwise would support, but now that research was subsidized by the federal government–so the government was merely subsidizing shareholders in industry? Bush argued that a society had two choices–to work with confidence or to work by fear. Confidence meant that one had to trust in the integrity of those doing science, that they would have that sort of honesty that Richard Feynman later wrote about–not to fool themselves, to provide context so others could see the limitations and errors of their work. Fear meant that one distrusted such an idea of integrity and in that fear sought to make others fear–fear of loss of funds, fear of sanctions, fear of humiliation for failing to produce significant results.
Federal attorneys and accountants and research program officers chose fear, and with it federal procurement contracting as the basis for the support of independent science–and thus, the federal funding agreement directed at approved projects, conducted by investigators. The contract would hold the institution and investigator accountable for reports, data, and–and this was the real stickler–inventions.
The problem of patents on scientific discoveries
The problem of patents on discoveries was the fourth and most difficult problem of all. In the absence of federal support, scientists generally published their results and everyone had access to them. Priority was the status-maker, not profit or secrecy. Even when a scientist patented an invention, they tended to leave science and pursue that invention–and that was a good thing in its way, another inventor seeking to make good. But if federal money was going to pour in, then what ought to be done with patentable inventions? Bush recommended that virtually nothing be done–the government should get a royalty-free license for supporting the work, and otherwise should stay out of it, letting the patent laws do what they were established to do. Research to expand the frontiers of science was unlikely to produce patentable inventions anyway–the laws of nature aren’t patentable, and to expand the frontiers of science is to stumble on new phenomena that has yet to be characterized or to provide an explanation for such new phenomena, or to provide new explanations for well known phenomena. None of this was patentable. Some definitions of “basic” or “fundamental” research went so far as to argue that if something was patentable then the research wasn’t nearly basic or fundamental enough.
Of course, folks were not thinking clearly about research tools–equipment, apparatus, devices that permitted observation and analysis. Galileo discovered the moons of Jupiter and so found something that did not orbit around the Earth–all pretty upsetting at the time–but the tool he used was a telescope, which he made because he had heard an account of a lens maker who had reported that combining two lenses magnified objects. With that, Galileo did is own inventing and created a research tool for astronomy. Tools were utterly unlike discoveries. They were intentional, they had immediate use, they could well be inventive, and therefore patentable. The patents then could be used to disrupt other researchers–creating a monopoly around an area of investigation. That didn’t seem right in general to most academic scientists, and it didn’t seem right at all when public money was going to support such work–at the request of the scientists themselves. So one part of the public covenant involving patents had to do with research uses. Patent practice had the idea of a “research exception” that acted as a kind of “fair use” of patented inventions, so they could be practiced to discover how they worked and the like. This research exception, however, does not now apply to university work–primarily because universities have made an “industry” out of seeking research funding and exploiting patents on the results for their own profit (for which see Madey v Duke).
The second area of potential patent use was as an industrial monopoly. This use had its attractions, coming from a non-industrial source. Patents historically were a primary tool to break up regional monopolies. Italian city states offered patents to folks who would reveal the trade secrets of a competing city’s guild. Monarchs offered patents to generate revenue for the government for similar reasons, as well as to stimulate activity (such as mining for gold and silver, hence “royalty” as a payment of a share to the crown). The patent broke up monopolies by creating a competing monopoly. Thus, one might imagine federally supported research producing patentable inventions that could be used to break up existing industrial monopolies, introducing competition between the old and the new, pushing the old to adapt and improve or be displaced by the new.
What was less desirable was allowing patents on research tools (and other discoveries not suited to extending the frontiers of science, but playfully exploiting some of the closer reaches of what’s already known) was simply handing them over to existing industrial monopolies to extend and entrench those monopolies. It’s an odd thought, actually, since those industrial monopolies were the ones that had the most money to offer (if they chose to offer), had the most capability to devote to practical application, and the most reach to be able to put something to good use. From one utilitarian point of view, at least, monopolies were efficient, effective, rich, and offered the greatest good for the greatest number. Why not just trot research inventions to the monopolists, get a good price, and go back for more? Somehow, this arrangement didn’t seem right–for universities, it placed them at the service of monopolists rather than in the service of everyone. For public funding of research, it mean that those government funds were a subsidy for the already wealthy rather than a subsidy for the advance of science. Again, it didn’t seem right. Maybe it still doesn’t, even if you try to think it through, it doesn’t. Is it just moral conditioning? Or is it the corruption of a good idea?
There was a third idea about patents, that they should be held in trust and used to prevent the formation of monopolies around research discoveries. Before there was “copyleft” for copyrights, there was this idea of using the patent monopoly to break up the patent monopoly. A university might then hold patents in public trust to ensure that everyone has access to the underlying claimed inventions–everyone except anyone who by patenting improvements or applications blocks everyone else–the blocker is denied a license and has to design around the whole open platform. Sound like “open source”? Yes, of course, that was the idea around holding patents in trust. Use the patent monopoly not for all of its uses–preventing all use, litigating for infringement, “protecting” an invention from competitors–but to break up the patent monopoly, immediately by default, or after a limited time if there’s a reason (such as calling forth private risk capital to assist in development for the benefit of everyone).
If you look over these various ideas, then you might see that the solution the federal procurement attorneys and policy makers came to seemed reasonable–the government, in supporting science, should own any inventions made and hold patents in trust for the good of all. The patentable invention, then, becomes a deliverable under federal funding agreements, and the government, once the owner of the invention, may decide whether to publish the invention (“dedicate” it to the public) or patent it and manage the patent right in the public interest. It’s this last option that we have to consider.
A patent is a right issued by the government in exchange for the publication of an invention rather than holding that information as a trade secret. Once others learn of the invention, in time they will be allowed to practice the invention, too–sooner if the inventor licenses, and later if not. In U.S. patent law, however, there’s no working requirement. An inventor does not have to practice the invention, does not have to license it, and may yet enforce the patent against others. When a company holds such a patent, for instance, it might prevent all uses of the underlying invention in favor of its own products that use a different method or material. The patent prevents competition from adopting the new technology, to the detriment of the existing product. The exchange is limited monopoly for publication.