Bayh-Dole government-wide licenses

In a Twitter thread in which Jennifer Gunter argues that “Americans should not be bankrolling the pharmaceutical industry,” Steven Martin tweets

Let’s work this through. There is no need to revise Bayh-Dole for state and municipal governments to authorize the production of medicines under Bayh-Dole. They already have the right to do so under Bayh-Dole’s “government” license. Bayh-Dole requires a patent rights clause in every federal funding agreement that includes a provision that as a condition of a contractor’s right to elect to retain title to a subject invention, it must grant to the United States a license to practice and have practiced. See 35 USC 202(c)(4).

In this license, “practice” means to make, use, and sell–the substantial rights in any invention. So to practice and have practiced means that the United States has the right–without owing any compensation to an owner of a subject invention–to make, use, and sell and authorize others to make, use, and sell. Furthermore, “United States” means the federal government, state governments (yes, the states, united), and municipal governments.

To show you that this is true, look at the usage of practice in the Kennedy and Nixon executive branch patent policies.


Governmental purpose–means the right of the Government of the United States (including any agency thereof, state, or domestic municipal government) to practice and have practiced (made or have made, used or have used, sold or have sold) throughout the world by or on behalf of the Government of the United States.


(replaces the definition with the scope directly in the license grant requirement)

The Government shall normally acquire . . . at least a nonexclusive, nontransferable, paid-up license to make, use, and sell the invention throughout the world by or on behalf of the Government of the United States (including any Government agency) and States and domestic municipal governments, unless the agency head determines that it would not be in the public interest to acquire the license for the States and domestic municipal governments

Norman Latker, the NIH patent counsel who drafted Bayh-Dole, claimed that Bayh-Dole was a codification of the Institutional Patent Agreement program that Latker revived in 1968, under the Kennedy patent policy. The IPA master agreement uses the Kennedy definition of practice and scope of license that includes state and municipal governments.

Bayh-Dole adopts these usages without notice that anything is supposed to be different. Latker claimed Bayh-Dole was based on that past usage. Bayh-Dole is intended (by Congress, by Latker, by Bayh and Dole) to follow the usage of the executive branch patent policies on which Bayh-Dole is based.

Practice is not just federal government “use” for research purposes, as some antagonists to Bayh-Dole have argued. Count NIST among the federal agencies most opposed to Bayh-Dole. The Trump administration could do us all a favor and remove the delegation to NIST of Bayh-Dole regulatory administration.

“United States” in Bayh-Dole’s license scope is not merely the “Federal government.” If it were, then Bayh-Dole would read “the Federal agency shall have a license to practice for or on behalf of the Federal government.” But it doesn’t read that way. The law reads “for or on behalf of the United States”–and that scope necessarily includes state and municipal governments, just as it expressly does in the Kennedy and Nixon executive branch patent policies that Bayh-Dole is built upon. It would take a determination of exceptional circumstances for any given federal funding agreement for a federal agency to limit the scope of the license in 35 USC 202(c)(4).

The scope of the license to the United States is not merely broad with regard to what United States governments may practice and have practiced subject inventions. The scope of the license also pushes back toward what inventions are subject to the licensing requirement. Bayh-Dole’s definition of subject invention, funding agreement, and contractor are really, really broad. See 35 USC 201 (b), (e), and (c). These definitions in turn make the government license apply to a really, really broad range of patentable assets.

A subject invention is a patentable invention (“is or may be patentable,” or a darned plant variety that isn’t patentable) that is owned by a party to a federal funding agreement and arises from federally supported research or development work. Bayh-Dole calls “contractor” any party to a funding agreement.

If an invention is owned by a contractor, and “arises from” (35 USC 200) federally supported work, then it’s a subject invention. All United States governments are required to have a license under 35 USC 202(c)(4) to make, use, and sell anything within scope of that invention, claimed or otherwise, and to have that invention made, used, and sold on their behalf. That is, they have a license not only to product that’s offered for sale, but to anything claimed by a patent on a subject invention–all the variations in compounds, methods, applications, regardless of whether any of those variations have been offered as commercial products.

Bayh-Dole’s definition of funding agreement makes clear that federal funding may support only part of the work in which invention and development of any given invention takes place. Wrap your mind around conceptually how Bayh-Dole represents the situation. First, establish an overall “work” to be done: for instance, find and develop new medicines for public use. Second, determine that there has been federal support for any part of that overall “work.” It does not matter whether the federal support is for a research part, or a development part, or a testing part (first actual reduction to practice may involve testing to demonstrate that the invention works as claimed, and may be essential where there is a history of failed claims, even if data from first actual reduction to practice is not necessary to obtain a patent–see In re Eddie L. King).

As the implementing regulations for Bayh-Dole make clear (37 CFR 401.1 Scope), once an invention is owned by a federal contractor, neither separate accounting nor chronology are determinative of whether a given invention is within scope of Bayh-Dole. The federal support could be for early research or could be for some phase of development, or could be to support clinical trials of a compound selected from the claimed invention to be prepared as a medicine. It does not matter whether federal money “paid” for the invention. Private money could “pay” for the invention. What matters is that the invention is owned by a contractor and the invention arises from the federally supported work. That is, the invention is within the scope of the federal funding agreement, and that scope is directed at the overall “work” for which the federal government provides at least some funding.

Thus, “federally supported invention” is a misleading shorthand, way too narrow. It’s really “invention arising from work in part federally supported.” These, conceptually, are lightyears apart. Development directed at first actual reduction to practice–such as a clinical trial–qualifies as part of the overall work.

The purpose of federal funding–other than perhaps NSF funding–is mission-directed. The NIH does not fund medicinal chemistry research merely to advance general scientific knowledge. It funds such research as a starting point (perhaps misguided, but that’s another thing) for the development of beneficial medicines. It’s just that the NIH has a longstanding policy (also perhaps misguided) of leaving the development of identified compounds to anyone other than the NIH.

Bayh-Dole makes up for the NIH’s refusal to do its own development of medicines in the public interest. The NIH intends its funding to result in medicines and the like. The nonprofit contractors that receive NIH funding announce in their formal policies their determination to own and commercialize every invention that gets made in their NIH funding. The NIH and its contractors agree that the overall work is to develop medicines. That’s the “work” established by Bayh-Dole’s definition of “funding agreement” at 35 USC 201(b). It’s not merely the “work” specified in a research proposal submitted to the NIH by a faculty researcher and its not merely the “work” that’s actually performed by contractor personnel in the “course” of doing the proposed work. The “work” in play in Bayh-Dole is the overall “work,” the work that the NIH as a matter of policy funds only a part of–a research part or a clinical trial part. Any invention from *any* part of this overall work, however, if owned by a federal contractor–any party to the funding agreement–is a subject invention, subject to Bayh-Dole, and subject to the government-wide license whenever a contractor elects to retain title to a subject invention.

This is a really, really, really broad claim. The claim is at least as broad as Mine Safety Appliances, a case from the 1960s in which inventors at the University of Southern California working on a Navy contract with an overall goal to improve pilot safety worked on a side project to develop their own helmet. Didn’t matter that they shifted their helmet work to a side budget and that their part in the overall work was to do some centrifuge studies on physiology. Even though the federal government did not have an ownership claim on their helmet invention, the government had an implied license because the contractor developed stuff that was within the overall project that the contractor had agreed to work on and received federal support for.

Bayh-Dole’s definition of funding agreement is also broader than a single contractor. The definition’s second sentence provides that a contractor creates new contractors by “any assignment, substitution of parties, or subcontract of any type.” Thus, a contractor adds a party to the funding agreement by any assignment of a subject invention. The new assignee becomes a contractor. The invention that’s assigned remains a subject invention. But more so, any inventions acquired by this new assignee and made within the general scope of the overall “work”–regardless of any additional federal funding–are also subject inventions. That is, any and all the inventions made in development by a for-profit assignee of an invention made under a federal grant are also subject inventions, also required to be timely disclosed, also subject to a requirement to elect to retain title, and also subject to the government-wide license. If the assignee-developer fails to disclose timely, then at any time (per the new NIST regulation change) the federal government can require assignment of the subject invention to the government, and thus there would be no need in such a case for a government-wide license.

When an exclusive license grants all substantial rights (to make, use, and sell) in an invention, the license acts as an assignment. It’s within the scope of “any assignment.” Courts looking squarely at subject inventions have ruled such exclusive licenses to act as assignments, even with reservations of rights for the government-wide license and for educational uses. There can be exclusive licenses that do not assign the invention, but those licenses must be for less than all substantial rights–an exclusive license to sell, for instance–and the licensee can have no right to enforce the patent. Most university exclusive patent licenses are also assignments of the claimed invention. Bayh-Dole’s provisions are specific to subject inventions, not to whatever patents might issue on those inventions. Assigning a subject invention–even by exclusive licensing–adds a new party to the funding agreement’s overall work, a new contractor with obligations same as the old contractor.

When a contractor assigns a subject invention. the issue that remains is what patent rights clause applies to the assignee. Bayh-Dole provides that the patent rights clause appropriate to the status of the new contractor applies to any such assignment *except* if the subject invention is owned by a nonprofit contractor. Then the nonprofit patent rights clause applies to the assignee (see 35 USC 202(c)(7)(A))–that’s the standard patent rights clause with the addition of all of  the nonprofit-specific provisions of 35 USC 202(c)(7).

The nonprofit patent rights clause is a big deal. Under the nonprofit patent rights clause, a contractor must account for all income earned with respect to any given subject invention–not only royalties (consideration for the grant of a license) but also any exploitation of the invention itself, such as, say, selling product using the invention or produced using the invention. The nonprofit patent rights clause requires contractors to use all income earned with respect to any given subject invention, after allowable deductions for expenses incidental to the administration of subject inventions (including royalty payments to inventors) to support scientific research or education.

When a nonprofit exclusively licenses (i.e., assigns) a subject invention to a for-profit company, that for-profit company necessarily, as a matter of federal law, must comply with the nonprofit’s patent rights clause. But more than that, becomes as well a contractor, too, and becomes a part of the overall work. In the case of developing inventions made in medicinal chemistry research, that overall work is the preparation of medicines. Any new inventions made by the for-profit assignee-contractor are also subject inventions whenever the new assignee-contractor acquires them. It doesn’t matter that the federal government doesn’t provide support for the new assignee-contractor’s development work. Bayh-Dole’s definitions don’t require such federal support. All that matters, as far as Bayh-Dole is concerned, is that the new assignee-contractor is now working on the overall work, and dedicating its own resources to this public effort rather than asking for additional federal support.

It has to be this way. Otherwise, the USC attempted exploit in Mine Safety Appliances would be an easy way to circumvent any government interest in any meaningful invention. Take federal money, but just when something interesting comes up, move the inventive and development work to a non-federal account and come after the government later for compensation. Licensing out development work is just another version of the attempt to shift development work (and any associated additional inventing) outside the overall work. Bayh-Dole makes it clear that such shenanigans is not permitted.

This outcome also serves Bayh-Dole’s overall policy to promote free competition and enterprise. If a contractor licenses non-exclusively for development, the licensees do not become parties to the funding agreement, aren’t contractors, and therefore any inventions they may make in the course of their development work fall outside Bayh-Dole’s scope. But if a contractor assigns a subject invention, then the assignee becomes a contractor, and whatever new that gets invented, when acquired by the new assignee, is a subject invention. When the assignor is a nonprofit and the assignee is a for-profit, the for-profit also agrees to give up its profit-seeking interests, or at least agrees that whatever profit-seeking it does with respect to any subject invention, whether licensed or newly made, is done to add more money to the support of scientific research or education.

Here is a special helpful note to companies that are not so taken with the idea of contributing to a public effort to create medicines, where doing so they agree to  give up all profits on those medicines for charitable purposes:

Refuse take exclusive licenses that amount to assignments from nonprofits.

Deal only in non-exclusive licenses or exclusive licenses to only the right to use or to sell.

Never accept the responsibility to enforce a nonprofit’s patent on a subject invention or settle cases of infringement.

If you take assignment, then you must choose to dedicate profits to the charitable causes that Bayh-Dole specifies. Be happy, in that case. You are doing a good thing, even if it seems to be against your nature. For all that, it is then up to you whether the public will benefit more from a lower price (you, foregoing profits) or from a higher price (you, using that extra money to support scientific research or education). Just keep in mind that scientific research or education isn’t product development. Isn’t marketing. Isn’t a new HQ. Isn’t investing in the stock market. Isn’t paying out dividends. Isn’t adding more staff or raising executive compensation. Hope this note helps.

This then is Bayh-Dole’s implicit statement of public policy: as a default, it is in the public interest that nonprofit contractors, and for-profit contractors agreeing to work on public projects as if they were nonprofits, get to decide the balance between low price medicines, and paying inventors, and making more money to support scientific research or education.

If any federal money supports any part of any drug invention or development or actual reduction to practice (i.e. clinical trials), then when any contractor–research host or assignee (even by exclusive license)–acquires an invention arising from any part of that work and elects to retain title:

(1) the United States has a government-wide license–for or on behalf of federal, state, and municipal governments

(2) to practice = make, use, and sell, and have practiced any subject invention

(3) for-profit assignees–i.e., contractors–must comply with the nonprofit’s patent rights clause.

And (3) is a Big Deal. The special non-profit patent rights clause provisions require a contractor to use all income earned with respect to any subject invention for scientific research or education, deducting only admin costs incidental to subject inventions. For-profit contractors working under a nonprofit exclusive license cum assignment cannot be involved for the money.

For Bayh-Dole to operate–to be actual law–the federal government must enforce it:

(1) Governments–federal, state, and municipal–must act on the United States license practice and have practiced (to make, use, and sell)

(2) Any government with the right to audit must audit for-profit contractors acting under the non-profit patent rights clause to ensure that all income less allowable expenses earned with respect to any subject invention goes to scientific research or education.

Anyone who argues that enforcing Bayh-Dole or encouraging governments to act on their licensed rights would “chill” biomedical innovation and prevent public benefit is an enemy of Bayh-Dole. And is talking nonsense.

Under Bayh-Dole, the “commercial” market for drugs in which a company can exploit patent exclusivity on subject inventions for maximum profit is limited to the “market” that no federal, state, or municipal government has the authority to act in, except federal agencies acting under Bayh-Dole’s march-in provisions. That is, the government-wide license is for any “market” within the government-wide right to make, use, and sell. There’s that “sell” word that indicates a market. The federal agency march-in, by contrast, is for any other market–any market outside the government-wide right to make, use, and sell.

If, for instance, one could imagine a health-related compound that no government had any authority to develop, acquire, or supply as a matter of the general welfare, then that would involve a “market” in which companies could exploit patent positions to maximize prices and profits. A cream to remove skin blemishes, perhaps, or a recreational drug to induce safe and pleasurable hallucinations. In such a market, a federal agency may use march-in if for some reason a contractor fails to achieve practical application, or the demand outstrips supply in an area of public health or safety or a regulation requires public use of the product and the contractor cannot supply it on reasonable terms. For the public interest market, however, governments do not need to rely on march-in–they have their license and can go to it without the threat of disruption for infringement or compensation.

Any state or county government that chooses to rely on Bayh-Dole needs to take these steps:

(1) Identify a subject invention on which a drug is based.

*there must be at least one contractor.

*there must be federal support for some part of the work.

*federal support may be to a contractor or direct federal expenditures for any part of the overall work from research to product testing (first actual reduction to practice).

(2) Determine that a contractor has granted the license to practice for or on behalf of the United States

*if not, the contractor is in breach of the election to retain title provision–demand that the federal government take title to the subject invention.

*if so, then the state or county government has the license it needs to practice and have practiced any subject inventions involved.

(3) Authorize companies to make and sell the drug or any variation within the claimed invention.

*authorization sets the payment (if any) for the production of the drug–nothing to negotiate on price.

*the state or county could pay for production and give the drug away.

If a patent monopolist contractor has a problem with this–gets chilled–then it can take its case to the Court of Federal Claims for reasonable and entire compensation. But it will have a hard time suing any state for infringement–sovereign immunity applies. But also, so do

Mine Safety Appliances

In re Eddie L. King

Bayh-Dole 35 USC 200, 201, 202(c)(4), 202(c)(7)(A).

The questions for states and counties–

Is public health worth standing up for the rights that governments have under Bayh-Dole?

If state and local governments are paying for medicines arising from federally supported research or development–from any contractor in the chain of exclusive patent rights–then they have to decide whether to grouse about it or act on the government-wide rights that Bayh-Dole demands that they have and can use to authorize making, using, and selling of subject inventions.

Is public health worth taking on the university and company looters of public interest?

If companies are making medicines arising from subject inventions available to the public at a cost lower than what any other companies or governments could reach, then let it go–the looters are doing a great job. However, if they gouge you $800 for a medicine that can be produced profitably for $80, then screw that. Time to act, not grouse. Stop listening to fretful attorneys worrying about infringement or demands for compensation. Bayh-Dole is there to be acted upon, not used as an excuse to screw over the public “in the public interest.”

We need our seven samurai.

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