Bayh-Dole in another simple diagram

Bayh-Dole conflates three distinct forms of federal contracting for research and then moves one form–university-hosted research–into the category of another, procurement from commercial firms. Here’s the diagram:


The effect of Bayh-Dole is depicted by the blue arrows.

The box on the left is labeled “CRO”–contract research organizations. When one contracts for research with a CRO, typically all IP developed in the contracted work is a deliverable. The CRO makes its way based on winning contracts and doing great work, not on trying to sell product to consumers. The research is the CRO’s product. For federal government work, CROs may take the form of government labs (GOCO), cooperative research centers (FFRDCs), non-profits, and for-profit companies. What defines the relationship is the nature of the procurement and the organization’s mode of operation, not the tax standing of the organization. Thus, universities, when they operate a federal lab (as in the case of, say, the JPL) or set up a research institute specific to federal contract, operate in the mode of a CRO. In a CRO relationship, inventions are another deliverable. The government would reasonably expect to own any such inventions created under a CRO contract.

The center box is labeled “COM”–commercial firms. When the federal government procures products or parts of products from a commercial supplier, it recognizes that the company may be in business of developing and selling such products. Under the Kennedy executive branch patent policy, the federal government permitted companies with “technical competence” and an “established non-governmental commercial position” to retain ownership of inventions, provided the government received a royalty-free license to practice the inventions and the contractor agreed to limitations on the patent property right “in the public interest.” In this center box we find for profit companies that sell products and might be expected to sell products based as well on work performed for the federal government. Here, too, we might find “dual use” technologies–developed for the military, say, but also having a civilian market, such as depth finders for ships and boats.

The box on the right is labeled “U”–universities and their research foundations that front for research. This is the box that Vannevar Bush sought to establish in Science the Endless Frontier. In this box goes research that is judged to be in the public interest–that the research is conducted at all. Bush envisioned that this research would be undertaken by the “free play of free intellects” without the constraints of either commercial product development or government agency mission-specific requirements. This new form of contracting would be along the lines of a “grant-in-aid” rather than a procurement contract, and the results would follow whatever pathway that the inventors chose, from dedication to the public to filing patent applications and pursuing a commercial venture, so long as the government obtained its non-exclusive, royalty-free license.

The lower box is “FED”–the federal government. Bayh-Dole goes out of its way to make it clear that federal agencies can grant exclusive licenses, even permitting the licensee to enforce the patent–meaning, that the exclusive license may operate as an assignment of the invention. Essentially, Bayh-Dole permits federal agencies to operate a shadow patent system in which they have patents issued to themselves to be offered for sale to companies and speculators for “commercialization.” While Bayh-Dole spends considerable effort laying out the bureaucratic conditions upon which such exclusive licenses/assignments might be granted, it offers far fewer of the “protections” set out for university and small business contractors. There’s a watered-down preference for American manufacturing. There’s a stipulated “first” preference for licensing to small businesses as “applicants,” but nothing obligating federal agencies to seek out small businesses (as there is in 37 CFR 401.14(a)(k)(4) for nonprofits.

Bayh-Dole forces federal agencies to abandon the university box and treat university research as if it were “early stage” commercial product development, allowing university administrations to take control of inventions and behave as if they are the start of a commercialization process. Doing so fits perfectly with the idea that university biomedical research is really just research done for the pharma industry and for speculative investors looking to sell to the pharma industry an early-stage product or company. Thus, we find university patent policies and licensing offices filled with assertions about the importance of “commercialization,” to develop commercial products from inventions using a patent monopoly to ensure that no university research invention gets used (legally) unless there is first a commercial implementation (and payment to the university).

Bayh-Dole also forces federal agencies to abandon the idea of contract research organization research. CROs, too, must be given the right to take ownership of inventions rather than deliver them to the government. But a CRO generally has no product intention, so Bayh-Dole allows CROs to accumulate patent rights that otherwise would be made available to everyone.

In effect, Bayh-Dole directs all inventions made with federal support toward monopoly patent licensing for commercial product development. There is no guidance with regard to dedication to the public domain, to the development of standards, to open access on a non-discriminatory basis, to non-exclusive licensing as a primary method of supporting access to new technology by small companies, to the importance of availability for use in research and internal use in industry–uses that often do not require a product version of an invention. While non-exclusive licensing is of course possible (and is essentially required in programs such as the NSF’s Cooperative Research Center grants), the apparatus of Bayh-Dole is built to motivate all contractors as well as federal agencies to conduct a trade in patent monopolies.

Only two industries regularly require such monopolies–the pharmaceutical industry and venture investors. In most other industries, an invention may be as likely to be used by a company rather than transformed into a product to be sold–methods of manufacturing, software controls, sensors to monitor, and the like. For the pharma industry, such inventions are called “tools” and the NIH goes out of its way to beg universities to license tools non-exclusively, and yet even here the NIH acknowledges that Bayh-Dole gives university administrations (and the NIH itself) free reign to license research tools exclusively.

If we look at the executive branch patent policy that Bayh-Dole supplanted, we can see the differences in approach used by Bayh-Dole.

The Kennedy patent policy involved three regulatory layers for the private management of inventions made in federally supported research. First, it identified four classes of research purpose in which the government should take ownership of inventions, regardless of the boxes.

  1. Products for commercial use. When the government is developing a product intended for commercial use. If the government is developing a new fertilizer, say, or a tomato-picking machine, then it might use multiple contractors. Why should these contractors have any right to patent their little bit of the overall product under development? How could such fragmentation ever lead to a product available to all for commercial use? Every user would have to deal with each federal contractor to obtain rights. Worse and worse if a federal contractor refused to grant a non-exclusive license. So if the government was developing something for commercial use or public availability, inventions were also deliverables.
  2. Research directly concerning public health. When the government supports research directly concerned with public health or public welfare. In such an instance, it makes no sense for the government to allow private contractors to establish monopoly positions to exploit public welfare. Why? These are matters of government, undertaken by the government with the assistance of private contractors. Allowing the contractors to assert monopoly positions in what they discover denies public access to the results. This bit is precisely where NIH’s patent counsel and university patent brokers fought back on behalf of the pharmaceutical industry, arguing that only with monopoly positions on research discoveries leading to commercial development might the public benefit from the research. The core of their argument is that no invention that’s not commercialized with a patent monopoly can benefit the public. It’s an argument with almost no truth in it, but there’s a smell of truth left in the jar, and that’s been enough.
  3. Government is principal developer. When the government is the principal developer of the technology and allowing any one contractor to hold a patent monopoly on a piece of the technology would interfere with competitive bidding and would give a contractor that won a government contract a basis to create a monopoly by which to exclude others from work with the government and a government-created advantage as technology moved into public use. The key technologies here were space-related and atomic/nuclear power.
  4. Operation of a federal lab. When the government contracted for the operation of a federal laboratory or to manage the work of others–the contractor should not have an independent right to inventions made in such circumstances.

For these four circumstances, the government should take ownership of inventions and then dedicate them to the public domain or grant non-exclusive licenses. In this way, everyone has access to the inventions, contractors and the public alike. Any proprietary positions come from adding new inventive features to what the government has supported. If there are no new inventive features to add, why, then, the development is pretty much already done. The argument is the flip of the one that the patent brokers used in getting Bayh-Dole passed: people will invest in development as they see new inventive work that will secure them a commercial market for their work, rather than that no-one will invest in further development unless they already have a patent monopoly.

Bayh-Dole eliminates these distinctions in the executive branch patent policy and forces the government to rely on march-in procedures which have been drafted to be so cumbersome that they have never been used.

The second regulatory layer added by the Kennedy patent policy is one of determining when a contractor that lacks capability or an established non-governmental commercial market should have a right to hold a patent monopoly on inventions made outside the four circumstances outlined above: “the determination of rights shall be made after the invention has been identified, in a manner deemed most likely to serve the public interest as expressed in this policy statement….” Agencies could, however, “prescribe by regulation special situations where the public interest in the availability of the inventions would best be served by permitting the contractor to acquire at the time of contracting greater rights than a non-exclusive license.”

The problem with determining what is in the public interest for any given invention is non-trivial. It is easier to resolve if a university research foundation makes the case that it will license an invention royalty-free to all qualified companies, never excluding direct use of the invention–and that the foundation can identify these companies more readily and provide better technical assistance than can the federal agency. The moment, however, that the foundation claims that it will secure its patent monopoly for the purpose of assigning the invention to a commercial concern, we are back to whether the government should be in the business of subsidizing patent monopolies that have the potential to exclude the public from the practice of the invention.

It is the delay in determining the public interest in allowing university research foundations to exploit patent rights in inventions made in university public-interest research (the box on the right in our diagram) that triggered the political moves that led to Bayh-Dole (after the IPA program failed to be expanded government-wide and was shut down for doing sweetheart monopoly patent deals with the pharmaceutical industry, and not being productive about it, either).

The third layer of the Kennedy patent policy regulatory apparatus was the imposition of a public covenant on inventions made with federal support, when the contractor was permitted to retain “principal rights.” The public covenant was developed, it appears, in response to objections from the Department of Justice and Health, Education, and Welfare regarding a proposal (from the president’s MIT-connected science advisor) that was circulating to permit federal contractors to own their inventions. Here’s some text from a Washington Post article from January, 1963:

Forme Secretary of Health Education and Welfare Abraham A. Ribicoff, in a letter to the Wiesner group, gave this reaction to its proposals: “If there is any single criticism of the draft statement that I would make it would be that it lacks a forthright statement of what the public interest is or requires on inventions deriving from Government support…”

“At the outset, it should be pointed out that the Statement of Policy Considerations seems to be unduly wighted in the direction of a consideration of the interest of industry in the entire matter. It would seem more appropriate for the emphasis in this sort of statement be cast more heavily upon the rights of the public to inventions derived from activities supported or paid for by public funds”

Assistant Attorney General Nicholas deB.Katzenbach also registered objections to the proposed patent revisions. He noted that some agencies, such as the Atomic Energy Commission, as a matter of law do not sign away exclusive rights to contractors at the time services are contracted for.

He warned the White House group against trying to reverse such policies “by Executive action.” Katzenbach added:”… this department is on record with COngress as opposed to any legislative solution of this problem which would grant exclusive rights to contractors when the contract is made. We have not been persuaded that this position is unsound.”

The public covenant required of inventions made in federal contracts appears to be a response to these criticisms. Contractors who retain principal rights to inventions have three years from the issuance of a patent to bring them to the point of practical application or grant non-exclusive licenses; are to provide written reports of their commercial use of inventions; and must accept government march-in for public purposes stipulated in the contract or by regulation or to address public health needs. The government, of course, is to receive its non-exclusive license to make, use, and sell (and have made, used, and sold) for or on behalf of the Government (including states and municipal governments). In short, the Kennedy patent policy introduces a working requirement for inventions made with federal support and held by a contractor and limits the patent property rights in those inventions–government license, march-in, three years of monopoly rights.

The idea of a public covenant then shapes a rather different insight: to enable private patent monopolies on publicly supported inventions, the necessary political piece is to announce protections for the public interest. Those protections do not have to be enforced, necessarily–they just have to be presented along with the proposal for contractor ownership of patents on inventions made in publicly funded work. And that’s the strategy that was used in Bayh-Dole to move university-hosted “basic research” to the “commercial” box, as if university research was conducted on behalf of future patent monopoly-based “commercialization partners”–“early stage” research, meaning so “early” in the “commercialization” process that no company is even aware that the invention has been made on its behalf.

Thus, instead of the “free play of free intellects,” as Vannevar Bush described it, Bayh-Dole made university faculty subservient to the future commercial interests of speculative investors and large corporations, all brokered by university administrations attracted to the idea of sharing the income from commercial success (including monopoly pricing) with those future commercial interests–if only they can be identified and persuaded to develop the invention under patent monopoly. This is the general case, then. The special case is that Bayh-Dole makes this general case to cover for the primary urge, which was to prevent university faculty in the area of medicinal chemistry publicly available and to also preclude HEW from insisting that federally supported inventions in medicinal chemistry be made available only on a non-exclusive basis, as a matter of public health.

In its gross outline, the argument becomes: “Why make something available to the public as an opportunity for everyone when the public could be made to pay a great deal more if only there was an intervening patent monopoly?” This in turn becomes the refrain “what is available to all is used by none.” Funny how the digital computer and those inventive internet communication protocols violate that refrain, among other obvious examples. A more subtle version of the argument is: “Why allow faculty inventors to develop their own inventions when the institution can profit from asserting the role of a broker to ensure inventions get into the hands of the pharmaceutical industry and so maintain their speculative value?”

There are development pathways by which research-originating inventions are used without any commercial product required–anyone who can practice the invention with ordinary skill may be able to benefit from the invention. Many research tools follow this pathway–any researcher can implement the tool. A commercial product version comes later, after the tool has proven its worth. The commercial version then must compete with the “do it oneself”–not only on price but also on technical specifications, availability, and support. Such competition between a technological user community and commercial product development is an essential feature of research innovation.

The idea behind Bayh-Dole is that such development cannot be allowed to take place for prescription drugs–that all such inventions must be routed through the pharmaceutical industry as regulated by the federal government. But that assumes that new discoveries must be dedicated–by law– to an institutional commercialization effort even before they have been demonstrated in practice. For inventions in general, such thinking is bonkers. More interesting is the possibility that even for therapeutic compounds, such thinking doesn’t hold up. We might think only of the markets for supplements and for “generic” drugs to see that a patent monopoly is not necessary. Even the argument that government regulation ensures that prescription drugs have a therapeutic effect that outweighs their potential harm falters as we consider how many prescription drugs have only a mild statistically significant effect, mostly making acute conditions chronic, often at a substantial price over the actual cost for either manufacturing or the development research.

Once Bayh-Dole removes university-hosted research from its own box centered on public access to inventions, it is easy to see how the product commercialization mindset arises. Universities do not have to have any particular competence in invention management or patent licensing, are not accountable for their actions, their patent policies and programs are not subject to federal inspection (as they were, at least, in the IPA program). All universities are asked to do is to step in and take faculty inventions on behalf of future patent monopoly commercialization partners. Instead of allowing contractors that have capability and an established commercial position to own and develop inventions made with federal support, Bayh-Dole allows university administrations to own inventions in the hope that they will strike an exclusive license (almost always an assignment of the invention) with a company willing to pay.

This activity is seen as a great public service, even virtuous. There is no mention of the value of cumulative technology, of the public domain, of standards, of libraries of technology tools, of immediate use in research and industry without an interposed monopoly for “commercial development.”

If we wanted to see how Vannevar Bush’s vision for university-hosted scientific research on the “frontiers” might work out, we would introduce a program that took the institutional controls out of the mix and restored the “U” box to its own invention protocols–that principal investigators had the responsibility to determine the approach to be taken with any given invention–publish, assign to the government for public access, license non-exclusively on FRAND principles, start a company, or assign to an existing company, so long as the federal government gets its broad non-exclusive license.


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