Here’s a diagram that might help discussions of patent rights made with federal funding. A general monopoly A operates without any constraints of law, and in particular without regard for anti-trust law. Thus, such a monopolist may exercise ranging power, including power through the courts to enforce contracts and property rights without limitation for public policy, market power, or scope of claim.
A patent monopoly P, by contrast, is constrained by the Constitution, by the provisions of federal patent law, and by anti-trust law. A patent monopoly concerns only certain subject matter, excludes other subject matter (laws of nature, surgery methods, perpetual motion machines, stuff that’s not new, useful, and non-obvious), and runs for a limited time (presently, 20 years from application date). A patent monopoly is further limited by various matters of public policy–restrictions on tying and price controls, for instance–and by anti-trust concerns having to do with competition and effects on markets.
A patent monopoly on a subject invention S–one made under a federal research contract and owned by a contractor–is more limited still. Bayh-Dole restricts assignment and licensing, and requires extra procedures to go through to secure ownership clear of government claims to title. And the government must have a non-exclusive right to practice the invention. So the government is exempt from any private monopoly practice. But there’s more to a subject invention monopoly. It is constrained by the statement of Congressional patent policy placed in federal patent law (35 USC 200). Patents have the “attributes of personal property.” They aren’t personal property, however, and those property rights are expressly “subject to the provisions” of federal patent law (35 USC 261). The statement of Congressional policy is the provision specific to patents on subject inventions. Any property rights in patents on subject inventions is subject to the provisions of the Congressional statement of patent policy.
Let’s single out two key requirements. First, the patent system must be used to promote the use of subject inventions. That is, the patent system is expressly not available to be used to exclude the use of subject inventions. That is, a patent monopoly on a subject invention cannot be used to prevent all use–even though for an ordinary patent this may be a perfectly legitimate use. Second, the patent system must be used to promote free competition and enterprise. If a patent on a subject invention can be held as a monopoly for the full term of the patent, and for all areas of use, then there would be no point in stating a requirement regarding free competition and enterprise. A patent held as a monopoly for the benefit of any single user would fail this requirement, as it would be equivalent in effect to any ordinary patent.
Thus, a patent monopoly on a subject invention must break up rights among competitors, must separate use from sale, must run as an exclusive right for less than the entire term of the patent, or must prevent the application of other monopoly rights that would prevent free competition and enterprise with regard to the invention covered by the patent on the subject invention. While the owner of a patent on a subject invention enjoys a monopoly, the monopoly is different in character from that of an ordinary patent and much more restricted in its property right.
Finally, we see in the diagram a monopoly G owned by the federal government on a patent covering a subject invention. This monopoly is restricted not only by Bayh-Dole but also by federal agency regulations and rules. In particular, federal agencies operate without a profit motive. Any income from patent licensing goes to the U.S. Treasury, subject to Congressional authority to make budget allocations, not to the federal agency that administrates the patent. Further, federal agencies operate with a mandate to manage inventions in the public interest, not for any particular private interest, and certainly not a for-profit venture. That is, any for-profit venture aiming to acquire an exclusive interest in a government-owned patent on a subject invention has to subordinate its own goals to those of the federal agency. The for-profit venture must administrate the patent on behalf of the federal agency, doing for the public what the federal agency delegates to the venture.
A federal agency has no authority to expand the monopoly available to such a venture beyond what the federal government’s own patent property rights are in a subject invention. What a federal agency gains by licensing its patent rights exclusively to a for-profit venture is access to development and marketing resources that would be otherwise unavailable to the federal agency or available but not as well positioned for effective development and dissemination of the invention. The federal agency gains capability from a private partner, but the private partner does not gain expanded patent rights beyond the restricted monopoly provided by Bayh-Dole, as further limited by the grant of an exclusive license. That grant cannot be for the entire or substantially all rights available under the patent or the exclusive license becomes an assignment. The federal agency has to continue to control some commercial rights–fields of use, sublicensing, term less than the full term, right to enforce claims of infringement, right to settle claims of infringement–and without an obligation to enforce or settle on terms favorable to any exclusive licensee.
The debate about Bayh-Dole has to do with whether the above diagram is correct. In particular, is there any difference between the monopoly of an ordinary patent and the monopoly available for a patent on a subject invention? Clearly, there are limitations on licensing a patent on a subject invention–but these are mostly a matter of administrative bother. The preference for small businesses is made so silly by the implementing regulations as to be useless in practice. The government is given the right to fuss and bother, but cannot contest any given past licensing practice. The requirement for substantial manufacture in the U.S. for any license exclusive in the U.S. also is readily circumvented–certainly for patented methods it hardly applies, as methods are not manufactured. Otherwise, the requirement is simply one of protectionism, preventing imports of patented subject matter to the U.S., even when such imports might provide a higher quality product at a better price. The prohibition for nonprofits on assignment of a subject invention except with federal agency approval (except to a patent broker) limits exclusive licenses that operate as assignments, but is generally ignored and there is no means to enforce the provision anyway. It’s like forbidding speeding, but not stating any sanction for speeding, such as a fine or arrest or confiscation of the offending car. Finally, nonprofits are limited in what they can do with royalty income. A nonprofit can recover only costs of managing subject inventions–not other inventions–and any profit over costs must go to scientific research or education–not to other nonprofit endeavors, not to be invested in stocks, not to be used as administrative slush money, not to expand one’s medical services empire.
Beyond these restrictions–which aren’t enforced as it is–one argument is that a patent on a subject invention is like a patent on any other invention. This argument asserts that the only difference in a patent as a result of Bayh-Dole is the administrative pain of complying, as if Bayh-Dole adds an apparatus of pain as a reflexive effort to create red tape for an otherwise simple release by the government of its claim to ownership of subject inventions. In this reading, a patent is a patent, and a patent on subject invention is a patent with minimal governmental oversight. The only real limitation is anti-trust law. Bayh-Dole goes so far as to make that limitation express (35 USC 211):
Nothing in this chapter shall be deemed to convey to any person immunity from civil or criminal liability, or to create any defenses to actions, under any antitrust law.
No one can argue, then, that what they did they were authorized or required to do under Bayh-Dole and therefore can’t possibly be anti-trust or anti-competitive. Big whoop.
Another argument, however, is that the statement of Congressional patent policy for subject inventions, placed in federal patent law, has actual effect on the scope of the property rights available for a patent on a subject invention. The Congressional patent policy does not merely justify what is to follow, by providing color commentary, as it were. That would be the case if the Congressional statement of purpose was left in the legislative history, as arguments in favor of the law. Rather, by being placed as part of federal patent law, the Congressional patent policy necessarily changes federal patent law. It is not “surplusage” and there’s no authority to read the Congressional statement of policy and objective as if it has no effect, “for information purposes only.” A basic canon of statutory interpretation is that if a provision in a statute can be read to establish a reasonable action, that interpretation is to be assumed over one that sets the provision aside as extra or optional or to no effect.
So the debate can be reduced to whether the diagram at the start of this discussion is properly drawn. Is the monopoly in a patent on a subject invention the same as that for an ordinary invention, except for some extra administrative pain that is mostly of no consequence to patent practice? Or does the Congressional statement of patent policy and objective place meaningful limitations on the scope of property rights in patents on subject inventions? If a patent on a subject invention is just like other patents, but with pain attached, then the statement of Congressional policy means nothing for patent law. It’s a wish list for the law. Surplusage. If a patent on a subject invention is just like other patents, then the apparatus supporting subject inventions makes no sense. When federal interest in patents made with federal support was a matter of executive branch policy, “subject invention” was a definition regarding a deliverable–a scoping statement. There, it defined a deliverable of interest and the obligations under contract surrounding that deliverable.
But in Bayh-Dole, there’s no executive branch patent policy. Bayh-Dole is not a regulation regarding the disposition of deliverables. It is a law regarding a new category of patentable invention–the subject invention. It applies regardless of who owns the invention. It dictates what federal agencies cannot do with regard to claims of inventions as deliverables. But Bayh-Dole also limits the property right available in patents on subject inventions. The issue is not how much bigger to make the monopoly available for subject inventions–it is how much smaller it actually should be than I’ve drawn.
If the monopoly available for a patent on a subject invention is to act as a trustee of the invention, a curator, an administrator, not with a profit motive but expecting only reimbursement for its direct costs (if even that), then a patent on a subject invention is nothing like an ordinary patent. Consider: with an ordinary patent, the patent owner can exclude use by others, even if the patent owner is not using the claimed invention itself. But Congressional policy limits the use of the patent system to promoting use, and does not authorize excluding use. A puzzle? Let me help solve it. Imagine a patent owner of a subject invention. The owner does not develop or use the invention. Someone else does develop and use the invention. The patent owner has no property right to exclude that use. That use does not infringe the patent owner’s property right in a subject invention. It is not infringement, it is success. If the patent owner is not using the claimed invention–or not using the invention in all its claims–then the patent owner has no property right to exclude others who do use the invention under any unused claims.
This is, I know, utterly ridiculous from the perspective of an ordinary patent. But patents on subject inventions are not ordinary patents. The monopoly is no ordinary patent monopoly. It is a trust monopoly. It it requires practical application to be established for the protection of that practical application to be available as a property right. Thus, to secure limited rights in a monopoly on a subject invention, the patent owner must use or license to one or more persons (companies, whatever) that use the invention, with benefits available to the public on reasonable terms. Fail any part of that definition of practical application, and there’s no patent right to exclude others from using. Once there is practical application, then there is a limited monopoly right in a patent on subject inventions, because then, arguably, exclusion of others promotes the use by those that are using–freeing them from competition that has not adopted early (when there was risk and development costs and costs to train a market) and not invested early (when the risk of investment and likely expenditure was greater).
There, does that help? Patents on subject inventions may create limited monopolies, but only once there is use that meets the definition of practical application in Bayh-Dole. The best way to achieve use is to license for it immediately, before ever there’s a commercial product. If there’s no way to achieve use short of a commercial product, then build in incentives for any exclusive licensee to achieve practical application before any other non-licensee does so. Then there is a rationale that excluding other users promotes the practical application achieved by the exclusive licensee. But the excluded users have to be ones practicing under the same claims that the exclusive licensee is using. If the patent identifies five possible ways to use a subject invention, and an exclusive licensee achieves practical application under one of those ways, then there’s no monopoly available to exclude others that practice under another of those ways. Otherwise, the patent is used to promote nonuse, not use–something expressly forbidden by Congressional patent policy. How could it be that a policy requiring the patent system be used to promote use ends up allowing the patent system to be used to prevent use? Can’t be. Doesn’t.
That there is no mechanism in the law for federal agencies to object to such patent behavior does not change the fact that the monopoly in patents on subject inventions does not extend to such behaviors. In fact, it would be entirely inappropriate for federal agencies to be granted the authority to enforce patent law–that is a matter for federal prosecutors, and a matter of defense for anyone wrongly accused of infringing a patent on a subject invention. Instead, Bayh-Dole restricts the property right available–the monopoly available–in patents on subject inventions. If the patent system has not been used to promote use–to the point of practical application–then there’s no monopoly formed. “Use or nothing” might be the public policy stated by Bayh-Dole. If that’s not an incentive to get to use with the assistance of the inventors, practicing every part of the patent that’s meaningful, before anyone else, then I don’t know what is. The monopoly right under Bayh-Dole for subject inventions kicks in with use. Until then, it is prospective, like a pending patent that still might pend after it issues, until there’s practical application. The monopoly in a subject invention does not start with the approval of a patent examiner regarding patentability.
The usual narrative regarding Bayh-Dole ignores all this, of course. And the usual narrative is wrong at almost every turn about the history of Bayh-Dole, what it does, how it does it, what matters for compliance, whether it has been successful, and the effect the law has had on research, innovation, and public benefit. That so many people rely on such a faulty–deliberately, indifferently, bullshitty, rah-rah-to-save-patent-broker-jobs–narrative does not change the underlying facts.
Look at the Stanford v Roche case. Scores of patent brokers and their mercenary attorneys representing hundreds of universities and organizations of universities and patent administrators recited in their amicus briefs the usual narrative–Bayh-Dole was inspired, has been successful, has produced wondrous innovation where there was none before. If Bayh-Dole were changed, inventions would sit on selves gathering dust, children would weep in the streets, the world would fall apart. Even Senator Bayh pitched the hope that by disenfranchising inventors from their inventions, Bayh-Dole gave hope to patent brokers to exploit patent rights (“in the public interest”) by creating private monopolies, giving a profit motive to universities where otherwise it wasn’t clear there should be such a thing.
The Supreme Court looked at all these versions of the usual narrative, across many amicus briefs, and rejected the narrative. It didn’t matter if Bayh-Dole was “successful.” It didn’t matter that important people representing important organizations claimed they knew how the law worked (vesting! no, first right! no, second right! no, inventors can’t assign to anyone else! no, er, well one of those ways surely). It didn’t matter if children would weep in the streets. The usual narrative was wrong. At points, the justices appear to mock the usual narrative:
It would be noteworthy enough for Congress to supplant one of the fundamental precepts of patent law and deprive inventors of rights in their own inventions. To do so under such unusual terms would be truly surprising. Had Congress intended such a sea change in intellectual property rights it would have said so clearly—not obliquely through an ambiguous definition of “subject invention” and an idiosyncratic use of the word “retain.”
Sweep away the usual narrative regarding Bayh-Dole. It is more than just wrong. It is deceptively wrong, even fraudulently wrong. Where it is recited by people in positions of public trust, in state and federal government, in universities, in research hospitals, the usual narrative betrays the public trust. What’s left? Time to read Bayh-Dole as it is written, in context, with a clear understanding of what has happened: 0.5% commercialization rate, 99.5% of inventions withheld from the public for the entire term of their patents, disruption of private initiatives to make inventions available and to develop them, suppression of commons, platforms, and standards. Suppression of research uses in industry. These are not the metrics of success.
The stark reality is that most research discoveries are not used. Ever. Patentable ones included. In fact, there’s a good argument that patenting a research discovery with the intent to assert ordinary patent rights in it works against rapid uptake of the discovery. It makes long odds even longer. It gives people incentives to design around, to avoid, to file blocking patents on improvements, to not use the discovery. Or to say screw the patent holder and use the discovery anyway. But no one has examined this point. Certainly no one in the habit of retelling the usual narrative.
What’s left, then, is space to construct a narrative that’s true to Bayh-Dole, and aware rather than blinded to the nature of research discovery and innovation that might come from access to research discovery. Patents have a role to play, no doubt. But the monopoly interest in such patents will vary according to the purpose of the funding and what might best bring new practices and products into use. Less than full term monopolies, less exclusivity than by assignment, and no right of exclusion until there’s practical application. Strange ideas, for the ordinary patent broker. But then, Bayh-Dole is strange. Federal funding in the public interest is strange. Research discovery is strange. Innovation is strange. Little of this thrives under institutional control. Little of it needs to call forth the hardened consciences of university patent administrators. Little of it requires the arbitrary (er, uniform) creation of private monopolies of the form of any ordinary patent.
If the stated purpose of Bayh-Dole was
arbitrarily to allow speculators to bet on the profitable exploitation of public suffering,
who would say Bayh-Dole is inspired? No, the law would never have passed in the first place. But that is Bayh-Dole’s unstated purpose. It is the effect in practice. University patent administrators do not have to operate this way, but most of them do. And a policy’s effect, unchallenged, is its intent.