NIST does not understand the government license in Bayh-Dole

NIST has issued a draft green paper that consolidates all the fake history, pseudo data as fact, misrepresentations of Bayh-Dole, and misconceived proposals all in one convenient place. I can’t hope to catch everything, but let’s take a look at one thing–the government license. NIST proposes to define that license–a nonsense move, and I will explain why.

NIST discusses what it calls a “government use license”:

The government use license refers to the “nonexclusive, nontransferable, irrevocable, paidup license to practice the invention or have the invention practiced throughout the world by or on behalf of the Government,” that applies to any federally funded invention

  • There is no such thing as a “government use” license.
  • The government’s license does not apply to any federally funded invention.
  • The entity benefiting is the “the United States” not “the Government.”

The term “government use license” does not appear in Bayh-Dole. Bayh-Dole specifies that the standard patent rights clause to be used in any federal funding agreement unless an exception applies or a statute overrides shall include a license to the government (35 USC 202(c):

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

(4) With respect to any invention in which the contractor  elects rights, the Federal agency shall have a nonexclusive, nontransferrable, irrevocable, paid-up license to practice or have practiced for or on behalf of the United States any  subject invention throughout the world . . .

Nothing here about “government use.” Not even “Government.” The term “government use” is used with the intent to limit the scope of the license by creating a misleading impression. In university administrative practice, “government use license” is just typical sloppy usage. The license established by statute is “to practice and have practiced on or on behalf of the United States any subject invention . . . .”

A subject invention is an invention that a contractor–a party to a federal funding agreement–owns and has elected to retain that ownership. Thus, only inventions that are subject inventions carry the requirement for a government license.

Now, let’s deal with “practice and have practiced.” University representatives responding to NIST’s call for public comment on federal technology transfer bitched and moaned that they didn’t know what the scope of the government license is. Funny–it would be the federal government that ought to be worried about the scope of the license. But no. University administrator ignorance isn’t public policy, but apparently NIST thinks their ignorance is important.

Norman Latker, who drafted Bayh-Dole, wrote that Bayh-Dole was essentially a codification of the Institutional Patent Agreement. He wasn’t being entirely truthful–since Bayh-Dole was rather an upsetting of the Federal Procurement Regulations (that he also was involved in creating)–making a special case of an extraordinary circumstance the new arbitrary default, and putting private exploitation of patent rights on inventions made in work supported in the public interest ahead of statutory protections of the public interest. But Latker did pull “practice and have practiced” from the FPRs (1975), which in turn implemented the Nixon executive branch patent policy (1971), which was a revision of the Kennedy executive branch patent policy (1963). In each of these documents, as well as the IPA master agreement, “practice and have practiced” is clearly defined. One would think that the variant “practice or have practiced” would be equally clear. Look:


In any case the government shall acquire at least a non-exclusive royalty free license throughout the world for government purposes.

Notice, this is not “government use”–it is government purpose–any purpose for which the government has authority to act. The policy then defines “government purpose”:

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.

The government purpose has nothing whatsoever to do with limiting the right granted to “use.” The scope of the grant is as broad as any action government is authorized to undertake.

Government–not merely the federal government, but also state governments and municipal governments. No government is to be sued for infringement when the federal government provides funding for work supporting an invention where the principal rights are with the contractor.

Practice–defined as “make, use, and sell”

Have practiced: defined as “have made, have used, have sold.”

It is clear that the scope of the license extends to anyone authorized to practice “on behalf of” government. Thus, “have made, have used, have sold” are captured by the “on behalf of” government.

There’s nothing ambiguous here. Pity university administrators are so clueless. Deeply troubling that NIST administrators are also uninformed about the law that they are responsible for.

The Institutional Patent Agreement program was revived by Norman Latker at the NIH in 1968 and operated under (and circumvented) the Kennedy patent policy and later the Nixon patent policy. The IPA master agreement requires a government license and includes a template license to be used to grant that license.

Note that “Subject Invention” is defined differently in the IPA master agreement than in Bayh-Dole. Bayh-Dole’s definition requires a contractor to own the invention before the law applies, while the IPA’s definition requires only that the invention was made under a federal contract.


The Grantee shall grant to the Government of the United States a nonexclusive, irrevocable, royalty-free license for governmental purposes and on behalf of any foreign government, pursuant to any existing or future treaty or agreement with the United States under each U.S. or foreign patent application it elects to file on a subject invention. The form of the license to be granted shall be as set forth in Exhibit “A”, attached hereto, and by this reference made a part hereof.  Any license issued by Grantee shall be made expressly subject to the license to the Government of the United States.

Exhibit A then has this grant of license, including a definition of “governmental purpose”:

The Licensor, in consideration of the premises and other good and valuable consideration, hereby grants and conveys to the United States Government a royalty-free, nonexclusive and irrevocable license for governmental purposes on behalf of any foreign government pursuant to any existing or future treaty or agreement with the United States under the aforesaid patent application and any and all divisions or continuations, and in any and all patents or reissues which may be granted thereon during the full term or terms thereof. As used herein, “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.


Whenever the principal or exclusive rights in an invention remain in the contractor, the Government shall normally acquire, in addition to the rights set forth in Sections 1(e) [reports], 1(f) [march-in for nonuse], and 1(g) [march-in for public purposes],

(1) 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 and domestic municipal governments, unless the agency head determines it would not be in the public interest to acquire the license for the States and domestic municipal governments; . . .

The Nixon revision does away with “practice and have practiced” and replaces “practice” with “make, use, and sell.” The Nixon revision retains the scope to include any U.S. government. To practice an invention “on behalf of” any given government amounts to “have made, have used, have sold.”

The Nixon policy revision requires codification of the policy. This codification took the form of the Federal Procurement Regulations, which recites the scope of the license in the Nixon revision verbatim. The FPRs included this grant in its long form standard patent rights clause (1-9.107-5):


Minimum rights acquired by the Government. With respect to each Subject Invention to which the Contractor retains principal or exclusive rights, the Contractor:

(1) Hereby grants to the Government a nonexclusive, nontransferable, paid-up license to make, use, and sell each Subject 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 . . .

The Nixon patent policy was even embedded in federal statute at 15 USC 2218(d) (for fire prevention and control research). Even though the FPR was later replaced by the Federal Acquisition Regulations, federal agencies retain by federal statute their authorization to act under the authority of the Nixon patent policy for all inventions not preempted by Bayh-Dole; that is, for patentable inventions made in work receiving federal support and not owned by a party to the federal funding agreement.

It’s clear that throughout federal policy from 1963 to 1981, the scope of the government license included states and local governments and the government in all these forms have the right to make, use, and sell inventions and authorize others to do so on their behalf. In this usage “practice and have practiced” meant “make, use, and sell, and have made, have used, and have sold.”

There never was a “government use” license for federal contracting. That is sloppy talk now making its way all dressed up into NIST documents. The government license was for governmental purposes. Not government “use.” That license included all substantial rights in any invention covered by the policy–to make, to use, and to sell.

NIST continues with its sketchy commentary:

The primary benefit of the government use license is that the government can use research that it funded for its mission-driven purposes without a threat of legal challenges for patent infringement.

This, too, is much nonsense.

  • The issue is not the government use of research, but of patented inventions.
  • The scope is not limited to “mission driven purposes” but rather extended to the broadest authority of the government to act.

Whoever drafted the NIST green paper is either grossly uniformed or deliberately misrepresenting the case. I wouldn’t be surprised to see AUTM officials involved, as the NIST paper’s style, references, and usage all reflect AUTM’s on-going, significant misrepresentation of Bayh-Dole.

The federal government has for a long time had the benefit of 28 USC 1498:

Whenever an invention described in and covered by a patent of the United States is used or manufactured by or for the United  States without license of the owner thereof or lawful right to use or manufacture the same, the owner’s remedy shall be by action against the United States in the United States Court of Federal Claims for the recovery of his reasonable and entire compensation for such use and manufacture.

Thus, for making and using patented inventions, the federal government is immune from infringement suits. At best, a patent owner can bring a claim for compensation. The governmental purpose license extends the government’s license in two ways. First, it is broader than just the federal government–states and local governments are expressly included up until Bayh-Dole. Second, the license includes the right to sell, which 28 USC 1498 does not address.

NIST then reveals the agenda driving this obfuscation of the language in Bayh-Dole “to practice or have practiced”:

An issue that has been noted is the use of the government use license to obtain discounts on products developed from federally funded R&D, primarily pharmaceuticals. A 2003 GAO report
concluded that the government use license does not bestow the broader right to purchase royaltyfree (i.e., discounted) products that happen to incorporate a federally funded invention if not produced under the government’s license.

Funny how GAO report conclusions somehow for NIST seem to have the effect of law. Here’s a passage from the start of that GAO report describing Bayh-Dole:

 The Bayh-Dole Act also provides federal agencies and their authorized funding recipients with a “nonexclusive, nontransferable, irrevocable, paid-up license” to practice these federally funded inventions for government purposes.

As we have seen, there is no such limitation “for government purposes” in Bayh-Dole. It’s just not there. Even the NIH criticized the GAO:

In commenting on a draft of this report, NIH stated that because we tie the exercise of the government’s license rights to the needs of the federal government, we give the impression that the government’s license rights are more limited than they actually are. While we agree with NIH that federal agencies and their funding recipients have unrestricted rights to use a federally funded invention for federal government purposes, it is
important to recognize that they can use these rights only to meet needs that are reasonably related to the requirements of federal programs.

The GAO’s assertion–persisting without any evidence to support it–apparently is good enough for NIST, as if the GAO is a fourth branch of government able to create new law by simply saying existing law should be made to mean whatever the GAO says it means. Some other time, we can look at how the GAO misrepresented Bayh-Dole in 2003, and why its report must be redone in light of the Supreme Court’s 2011 decision in Stanford v Roche.

Or, given that things like this don’t arise out of our collective nightmarish void unbeckoned, we might hypothesize that agents of innovation darkness have made it a point to get to NIST officials, GAO officials, and others with their fake history, fake data, and misrepresentations of Bayh-Dole. After all, Norman Latker worked on presidential patent policy, implementing regulations to that policy, programs to circumvent the implementing regulations, and when those programs were shut down as ineffective and contrary to the public interest (giving pharmaceutical companies monopoly control over inventions made with federal support, laundered through universities), new law to supersede presidential policy and federal statutes, and the implementing regulations for that new law. It’s like a beast in a horror movie showing up again and again. Now Latker is no longer with us, but others who worked with him still are, carrying on in his tradition of manipulating government policy and federal law in the service of the pharmaceutical industry.

The effort is to prevent the government from using its license to avoid paying whatever pharmaceutical companies wish to charge for prescription drugs based on inventions made with federal support. This is the heart of one of the big bubbles of corruption created by the misrepresentation of Bayh-Dole, contractor non-compliance with Bayh-Dole’s standard patent rights clause, and federal agency non-compliance with Bayh-Dole and non-enforcement of the standard patent rights clause.

There’s a convoluted set of arguments, misunderstandings, and misrepresentations that university administrators throw up with regard to the license. I use the image deliberately. They focus on the “royalty-free” aspect and insist that Bayh-Dole requires universities to include a requirement in their licensing agreements for subject inventions that licensees must sell to the government by reducing the price of any sales by the amount of the royalty due the university. This is pure nonsense, mistaking a grant of rights for a downstream sales transaction.

The government license is to practice and have practiced, not to create leverage to negotiate a lower price if the government chooses to purchase a product rather than making it or having it made under its own authority–under the government’s license.

University patent administrators are determined not to allow anyone to lower the price charged for drugs made under license from universities, even when the drugs are based on inventions made in work receiving federal support. The higher the drug price, the higher the university royalties. If pharma is going to screw the public and the federal government, then university administrators want in on the action.  Apparently NIST is ready to go along with them–a clusterfucking of the public, apparently. Here’s what NIST proposes:

Intended Action 1. Define the scope of the “government use license” . . .

There is no “government use” license. Congress placed into the Bayh-Dole the language “practiced and have practiced.” That is not a matter of “government use” but of “government practice” of the licensed invention–make, use, and sell. NIST has no authority to define the scope of “practice and have practiced.” That is a matter for Congress, not NIST.

Other than for perpetually confused university administrators, “practiced and have practiced” must mean what it meant under the Kennedy patent policy and the IPA program–and what the Nixon policy and its codification in the FPR made express–to make, use, and sell. If Congress had wanted anything other, then Congress could easily have made that clear. Given that Norman Latker drafted the IPA master agreement, was involved in the Nixon patent policy and the FPRs, and drafted Bayh-Dole, if Latker had wanted to make clear there was to be a change, certainly he could have done so. But he didn’t.

NIST continues:

. . . for use directly by the government—or a government contractor in the performance of an agreement with the government—for a government purpose only, . . .

The “government purpose” here doesn’t exist in Bayh-Dole–there is no “government purpose” limitation in Bayh-Dole. The license is “to practice and have practiced by or on behalf of the United States.” NIST is off trying to define the scope of something that doesn’t exist but for the fantasy misrepresentations in the minds of university administrators, AUTM officials, a few vocal Bayh-Dole pundits, and apparently some wandering folks at NIST.

. . . including continued use in research and development by the government.

The idea that the government’s license is limited to a list of government purposes such as “use in research and development” amounts to a useless license. The effect is to expose the government to claims for compensation broader than that provided by 24 USC 1498–that somehow the government has to agree to take a narrower license than the government otherwise has–if only clever university administrators can persuade wallowing NIST officials to ignore the law and listen to their wormtongues.

. . . The scope of the government use license should not extend to goods and services made, sold, or otherwise distributed by third parties if the government—or a government contractor in the performance of an agreement with the government—does not
directly use or consume those goods and services.

We reach peak bozonet. The government license is to practice and have practiced. That scope of rights has nothing to do with what price the government chooses to pay for products it purchases. There is no scope of license to define, even if NIST did have the authority to define the scope of a license and thus alter the language of the federal statute without going through Congress.

However, if the federal government, or any state government, chose to make, use, and sell a drug that was based on an invention made in federally funded research or development (including when subject inventions are assigned by nonprofits to pharmaceutical companies), then the patent holder would have no claim for compensation as well as no standing to bring an action for infringement.

Thus, NIST proposes taking the law into its own hands:

Implement regulatory change under the Bayh-Dole Act to (i) update the definition of government use license and its use directly by the government—or a government contractor in the
performance of an agreement with the government—for government purpose only and not for the use of a third party, and (ii) clarify the appropriate processes and use of the government use
right based on a consistent interpretation of the definition restricting its scope of use.

Again, “government use” does not appear in Bayh-Dole nor in the implementing regulations or in the standard patent rights clause at 37 CFR 401.14(b). NIST has no authority to introduce a change to the statutory language by adding a definition that the statute does not use and does not need–and which is opposed to the scope of the license the government is, by statute, to be granted. The rest of the NIST agenda, here, amounts to farts. If there’s no basis for the definition of “government use,” then there is also no basis for clarifying that definition or restricting its scope to “government use.”

We should repeal Bayh-Dole if only to remove policy on inventions made with federal support from the massing bozonet of folks who don’t read the law, don’t think clearly about the law, habitually misrepresent the law, and for all that come up with really horrible ideas about how to encourage innovation. Their vision amounts to a private police state for inventions in which university bureaucrats serve the interests of speculators, trolls, and pharmaceutical companies–however they can make money. If that’s the innovation engine that federal law should adopt, then let’s out with putting it that way and see how it does.

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