Invention is not a thing, 14

Bayh-Dole’s public protection apparatus, even unused as it is, makes it clear that the federal invention economic system is intended to be different from that of private exploitation of patents for financial gain. In the federal economic system, patents are to be used to promote use, not suppress use, of all instances of any given invention. This is dramatically different than the idea of simply expanding private speculation on patents to include inventions that have been made in publicly directed projects, as if patent speculators don’t have enough opportunities to speculate and need a federal welfare program in the form of research grants and contracts to help them in their struggle for huge, lottery-style upsides.

For governments that hold patents on inventions, whether federally supported or not, these issues are magnified. State governments, unlike private parties, have obligations to serve the public welfare. Even Bayh-Dole makes this distinction (though like much of Bayh-Dole largely unenforced). Contractors have one public protection apparatus, in the form of 35 USC 202-204; 206 and  37 CFR 401 requirements. Federal agencies have a different public protection apparatus, in the form of 35 USC 207-209 and 37 CFR 404. The striking difference concerns exclusive licensing.

On the contractor side, contractors may license exclusively with few controls–an easily circumvented and regularly waived requirement for a narrow set of exclusive licenses to require U.S. manufacturing. If an exclusive license functions as an assignment of the invention, then nonprofits are required to pass through with the license the nonprofit’s patent rights clause to the assignee. And the federal government may march-in and void an exclusive license for nonuse, failure to meet demand, or failure to comply with the empty U.S. manufacturing requirement.

On the federal side, however, there is a vast apparatus that purports to control exclusive licensing. The government is authorized to grant exclusive licenses, even to the point of assignment of the invention (35 USC 207, finally addressing the statutory gap made clear by the Supreme Court way back in 1933 with its Dubilier decision). 35 USC 209 then lays out criteria under which an exclusive license may be granted.

It’s an involved list of criteria. We will use a broad brush.

The granting of the exclusive license must be a “reasonable and necessary incentive,” there must be a plan submitted by the candidate licensee, the proposed scope of exclusivity must be “not greater than reasonably necessary,” the licensee must commit to achieve practical application within a reasonable time, and will not substantially lessen competition or violate antitrust laws.

All of these conditions involve qualifications, exceptions, and waivers that provide for pathways entirely at odds with this broad brush account. The “reasonable and necessary incentive” is one that will “call forth the investment capital . . .  needed to bring the invention to practical application.” It is not that the exclusivity is needed, but that an incentive is necessary. If there’s no need for an incentive, then there’s no authority for the exclusive license. But what constitutes the *need for an incentive*? Is it that the candidate company insists that it will not spend to develop an invention unless it has an exclusive license? Even if one company insists on exclusivity, what about other companies? It would appear that the case for exclusivity must be made by the federal government before any offer of a license to any particular company is even considered. That is, the federal agency must determine that without the incentive of an exclusive license offered to one company, no companies (or any other organizations) will develop any instance of the invention. This is very strange, as it leaves out the possibility that instances of the invention do not require any development or investment capital to achieve practical application.

There’s more–even if there is no need to call forth investment capital, the federal government can still grant an exclusive license if the license is a “reasonable and necessary incentive” to “otherwise promote the invention’s utilization by the public.” That’s a really strange otherwise. The logical structure is that the “otherwise” part is the more general statement–the federal government can grant an exclusive license when it is a reasonable and necessary incentive for a non-federal entity to promote the invention’s use by the public. If that is the case, then any where a non-profit organization is willing to take a license–already having a mandate baked into its articles of incorporation to promote the public interest–then it cannot receive an exclusive license because it already has sufficient incentive to promote the invention’s use. The upshot: the federal government cannot grant an exclusive license to any nonprofit that has a mission consistent with the public use of the invention. Strangeness.

But if the prospective licensee is a for-profit that would not otherwise invest in the invention if available to others, and all other prospective licensees meet this same standard, then the federal government may grant an exclusive license or even assign the invention–even if there are companies that would use the invention as it is (that is, do not need any incentive to spend on development, or development is not necessary). You see how screwed up this all this? The criteria for exclusive licensing authorize the government to deal with those companies that would refuse to invest in development unless they have an exclusive license and to exclude from consideration of a license all those companies that would use the invention or develop the invention without an exclusive license or without requiring the exclusive license as the necessary incentive to invest in development.

The rest of the apparatus is just as ghastly–and even more so once the implementing regulations at 37 CFR 404 have got done with them (for instance, expanding the plan requirement to all licenses, not just exclusive and “partially” exclusive licenses. For our purposes, however, the remarkable bit comes at the end of the requirements, having to do with antitrust law:

granting the license will not tend to substantially lessen competition or create or maintain a violation of the Federal antitrust laws;

Again, it is a mealy mouthed criterion: combining “granting” with “will not tend” and with “to substantially lessen” leaves plenty of operating room for anti-competitive licensing practices by federal agencies. The criterion requires a federal agency to determine in advance of the exclusive license that it won’t “tend” to “lessen” competition. How could this determination possibly take place? How could an exclusive license do anything but “tend to lessen” competition? Perhaps the argument is that there cannot be any competition for an invention that is unanticipated and therefore no exclusive license could possibly lessen that non-existent competition. But that argument makes nonsense of the statute–there’s no reason to state something that is tautologically true. No, the competition to be lessened is the competition that would take place if the invention was made available non-exclusively. There, the determination must be that no company will take a non-exclusive license or would use or develop any instance of the invention were it in the public domain. For that determination to take place, one would have to actually offer the invention for non-exclusive licensing for a reasonable period of time–such as from the time a patent application is filed to a year after a patent has issued.

No one in their right mind (other than speculators and folks with an agenda to capture a market) takes a license until they see claims allowed on a patent application and those claims are in line with what they need to practice. Without claims allowed on a patent, a federal agency cannot reasonably make a determination that an exclusive license is necessary. Without offering the invention for non-exclusive licensing, a federal agency cannot make a determination that an exclusive license is necessary. And if a company does take a non-exclusive license, then it is clear that an exclusive license is neither reasonable or necessary to promote the utilization of the invention by the public, even if a company would be willing to buy up all the rights and so obtain an exclusive license. In that case, the logic would go–the federal government may offer an exclusive license when doing so would provide an incentive for a company to exclude all others from working an invention they otherwise are prepared to work. Again, that would be very strange.

It is not sufficient that a federal agency announces that they are considering an exclusive license to a given company, to see if anyone objects. No. The federal agency has to first offer a non-exclusive license on reasonable terms that are made public in advance of any negotiation with any single company. That offer has to stand for at least some time after a patent issues to provide potential users of the invention to see exactly what instances of the invention are claimed by the patent. There really is no other way to do it and meet Bayh-Dole’s requirements for the exclusive licensing of federally owned inventions.

The NIH’s IPA program, restarted in 1968, has a similar apparatus for exclusive licenses for contractors–something Bayh-Dole pretty much drops. In the IPA program, a grantee must first determine “that nonexclusive licensing will not be effective in bringing such inventions to the commercial market in a satisfactory manner.” There is no guidance at what “effective means” nor what “satisfactory manner” means nor why the use of such inventions outside any “commercial market”–research use, DIY use, internal company use–is excluded from consideration.

There’s a gesture toward requiring non-exclusive licensing first:

Exclusive licenses should be issued only after reasonable efforts have been made to license on a nonexclusive basis , . . .

To make a reasonable attempt to license on a non-exclusive basis, a grantee would necessarily have to make the terms of such a license available. It is not “reasonable” to say–this invention is available for licensing non-exclusively and not identify the terms–cost, restrictions, structure of payment, risk requirements, and the like. But the IPA master agreement immediately takes back the requirement:

. . . or where the grantee has determined that an exclusive license is necessary as an incentive for development of the invention or where market conditions are such as to require licensing on an exclusive basis.

So, try non-exclusive licensing first except you don’t have to if you decide that an exclusive license is necessary as an incentive or because “market conditions” require exclusive licensing. Read slowly this is all nonsense with a purpose. If no one takes a non-exclusive license, that should be a good basis for determining that one may have to try an exclusive license. There’s subtlety here, of course: a company (one can imagine) might take a non-exclusive license to prevent other companies from taking the exclusive license they need and so disable all investment in developing the invention. So, in this bizarre logic, if one can imagine a company taking a non-exclusive license to thwart exclusive licenses, then one must start with exclusive licensing to thwart those that would thwart development by taking a non-exclusive license. Sigh. Just make non-performing non-exclusive licenses revocable. That’s what Bayh-Dole does with march-in. But apparently this is all too difficult for university patent administrators and federal policy makers.

Essentially, the IPA exclusive licensing provisions boil down to you can grant exclusive licenses any time you think that’s the way to go. And in practice, that’s what nonprofits operating under the IPA program did–grant exclusive licenses and attempt to grant exclusive licenses. All that apparatus about non-exclusive licensing and criteria that must be met to deal in exclusivity was just cover for a clear pathway of exclusive licensing that was intended all along–it was just that there needed to be the appearance of an apparatus that made it look like any exclusive license was supported by a determination that exclusivity was the only way to benefit the public. Pay no attention to the fact there’s no enforcement mechanism for the IPA requirements–just as there isn’t for Bayh-Dole’s federal licensing requirements. At least for contractor licensing, federal agencies have a march-in right (such as it is, designed not to work). But there’s no oversight at all for federal agency licensing of government-owned inventions. Federal agencies can just decide what the words mean–even to mean something quite opposite from what a normal person might take them to mean–and go their merry way.

All this is on the federal side of Bayh-Dole’s exclusive licensing requirements. But when the contractor on the other side of Bayh-Dole’s public protection apparatus is an instrument of state government, such as a public university, then we might ask whether an apparatus similar to that on the federal side should apply. That is, non-exclusive licensing should be required unless the state entity can show that exclusive licensing is necessary–that all instances of the invention require development and that the financial support for such development cannot be obtained through collaboration, private donations, or government-provided funds. To show any such thing, one would have to reasonably attempt to develop instances of the invention with non-exclusive licensing–and continue those efforts until a reasonable time–one year, say–after a patent has issued.

This is not a matter of “commercial product” development but rather the migration of an invention from reduction to practice to practical application. The scale at which practical application is achieved for some instance of an invention is distinct from the use that that instance or any other instance of the invention might have in non-mass-produced settings. There is no reason why the effort to create a mass-produced version of an instance of a given patented invention must exclude the right of others to prepare craft versions of that same instance: build-it-yourself rather than wait for someone to solve the scale-up, quality assurance, feature selection, general compatibility, distribution, training, and post-sale support problems of mass production.

For state entities, exclusive licensing of inventions should not be conducted as if disposing of private property, whether the purpose is to make money for the state or to play favorites among companies or to give industries with substantial presence in the state some advantage over companies operating in other states. If there is to be exclusive licensing, it should follow at the minimum something like the Bayh-Dole requirements for federally owned inventions, but without the mealy mouthed approach.

Especially important is following the Bayh-Dole limitation on exclusivity (35 USC 209(a)(2):

the proposed scope of exclusivity is not greater than reasonably necessary to provide the incentive for bringing the invention to practical application . . . or otherwise to promote the invention’s utilization by the public.

That is, a state must use an exclusive licensing variation such as the ones I have outlined–restricting exclusively licensing for instances that are difficult, releasing for other licensing any instances of an invention not timely developed to practical application, and granting exclusivity for substantially less than the full term of the patent.

When a state acquires a patentable invention, that invention becomes a state property. Unlike real property or chattels, however, that invention (and any patents that the state obtains) can be used by multiple persons simultaneously. To dispose of the invention by exclusively licensing it to a single company (or, as is often the case, assigning it to a company) denies the public the opportunity to benefit from the invention in all its instances other than as dictated by the activities of the company licensee. This is true even if one makes the argument that only by an exclusive license will the chosen company attempt to create from a selected instance a “commercial” product that would then be made available for sale to the public.

Even if such a thing happens, there is no reason for other instances of an invention to remain suppressed, other than that it is in the financial interest of the exclusive licensee that other instances are suppressed. The argument then amounts to equating public interest with a company’s financial interest. That’s bad enough on its own–but if the state itself has a financial interest in the company’s financial interest, then it is not in a position to decide the public’s interest, any more than is a judge who has a business relationship with a plaintiff or defendant or attorney.

Suppressing the use of any instance of an invention runs against Bayh-Dole. Suppressing otherwise potentially beneficial instances of an invention directed at public health goes further and runs against public interest. And equating private profit-making with the public interest is necessarily antagonistic to public interest, no matter how wonderful profit-making might be to companies, even with Friedman’s argument that the public purpose of a corporation is to make a profit for its owners. The public purpose of profit-making is not equivalent to the public interest in having access to the benefits of using instances of an invention. The commercialization of instances of inventions directed at public health arises from a context of government-assured public access to those inventions made with public support, not from a government-assisted suppression of public access.

An invention is not a thing. It is a collection or set of things. An invention provides a doorway of opportunity for new practice. A patent provides an inventor with the ability to claim for a limited time the right to exclude all others for as much of that opportunity as the inventor can describe–including variations and applications. In U.S. patent law, an inventor has no general obligation to work any patented invention. Some instances of an invention may be easy and require no development to be used. Other instances may be difficult and require substantial effort (and money) to become useful. And yet other instances may be so difficult that it is impossible within the term of a patent to expect that the instances will be developed to the point of practical application.

However, Bayh-Dole does create a working requirement for federally supported inventions. Institutional contractors, at least, have an obligation under Bayh-Dole’s standard patent rights clause to work federally supported inventions that they acquire. And that means, to work each instance of such an invention and to release those instances that they do not work, and if they will not work instances or release them, then the government must march-in and require or enable such release. To argue otherwise is to argue that Bayh-Dole has it all wrong or should be ignored or can only be successful if it is turned on its head and made to say what it does not in fact say.

This is all the more the case with health-directed inventions made with government–federal or state–support. And it is all the even morer the case with inventions owned by state governments, which have an obligation not to be conflicted between their efforts to govern for the good of all (including all companies) and their financial or power-brokering interest in chosen, favorite companies and investors.

There’s no excuse for state universities licensing state-supported inventions exclusively without that exclusivity restricted to just those difficult instances that demonstrably do require private development to be worked available only through exclusive control by a speculative investor chosen by the state.

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