Learning from Latker’s 1984 “Federal Initiatives for Innovation” Talk, 3

Norman Latker, formerly patent counsel at the NIH and chief architect of Bayh-Dole and its extension by Presidential memorandum to all federal contracting, argues that if federal inventions are not privately owned and exploited for their exclusionary and financial value, then the results of federally supported research are wasted, and the public will not benefit from federal research because companies will not adopt or invest in anything that they cannot own exclusively–or, more particularly, given that American companies at least won’t adopt much of anything made in federally supported work, that patent speculators ought to be given “incentives” of patent exclusion to put up the money to make commercial products that, when sold at whatever price the market will bear and in whatever form might be most profitable, will then serve the public interest.

Latker proposes a plausible fantasy. Fom a practice point of view, it’s nonsense. Companies adopt new technology all the time that does not come with a gift of exclusionary rights. There’s no additional uncertainty if the federal government takes ownership of inventions made in its operations and releases those inventions, uniformly, without conditions or payment requirements. Such open access amounts to a more robust public commons–a sort of federal open innovation. If the federal government had the right to deal exclusive rights in any such invention, then a company might hesitate–why should it contribute to research that could, when any invention arises, be diverted to a federal agency that could play favorites in the market, cutting out everyone else with an interest.

And that’s just what Bayh-Dole has done–by act of Congress authorized federal agencies to deal in exclusive patent rights. 35 USC 207(a)(2):

(2) grant nonexclusive, exclusive, or partially exclusive licenses under federally owned inventions, royalty-free or for royalties or other consideration, and on such terms and conditions, including the grant to the licensee of the right of enforcement pursuant to the provisions of chapter 29 as determined appropriate in the public interest;

This provision of Bayh-Dole runs directly against the recommendations of the 1947 Attorney General report, ignores the arguments laid out there for why it is bad public policy for the federal government to use its power to sue or threaten to sue citizens for infringement and so subordinate government purposes to the private interests of a company favored by a federal agency with an exclusive patent position. To grant the right of enforcement means that the federal agency in effect assigns ownership of the invention to a private party. What’s left utterly unexplained is when it is “determined appropriate in the public interest.” You might think that 35 USC 209(a) takes up federal exclusive licensing, but if you read the provisions carefully, you will find that it reduces to “granting the license is a reasonable and necessary inventive to . . . promote the invention’s utilization by the public” and that the “scope of exclusivity is not greater than reasonably necessary to provide the incentive.” The operative verb is “promote”–the exclusivity promotes, rather than results in, public use. In this usage, it is all but tautological that an incentive intended to promote something can be determined to promote that something: “If a federal agency determines that an exclusive license will be an incentive to promote public use, then of course the federal agency is authorized to grant that exclusive license and assign the invention to the private company of its choosing.”

And “reasonably necessary” incentives in exclusive licensing come down to what the exclusive licensee holds out for. One might imagine a reverse auction in which federal agencies start out offering all rights to a federal invention and no public protections, and then interested companies bid against one another, proposing restrictions on the scope of their exclusivity until finally one company emerges as having proposed the greatest restrictions possible on their exclusivity, and so wins the grant of the exclusive license. Such stuff never happens. But that’s the sort of thinking that Latker set up in Bayh-Dole. And worse, there’s not even a requirement to set up a bidding process. Instead, federal agencies just do a mind-experiment–we are talking administrative minds, here, ones like Latker’s–and if they imagine what the most restrictive bid will be, and offer an exclusive license on those terms. But if their favored company turns down the offer and demands less restrictive terms, or–whatever terms the federal agency wants so long as the federal agency agrees not to enforce them, or builds into the deal that there’s no consequence for breaching the agreement–then well, it would appear that “reasonably necessary” will be whatever the company favored with a federal offer decides is “reasonably necessary.”

There’s no standard evoked by “reasonably necessary” and the “reasonably” does not add public strength to the statute–just the opposite, it weakens the federal position–what would a “reasonable” person do in the same situation? But the federal government is not just some other “reasonable” person who might act at times “unreasonably” and thus have to be held to the standard of what some other “reasonable” person would do. Nothing about “reasonably necessary” works in practice or as public policy. It’s a shambles. It’s nonsense.

The Kennedy patent policy called for incentives to “draw forth private initiatives.” Latker translates that into “private risk capital.” Private initiatives and private risk capital are two potentially very different ideas. Private initiatives involve any action taken to use or develop an invention. The federal government worked on encouraging such private initiatives by offering technology transfer services–instruction in how new federal stuff works, for instance. But here, in Kennedy’s patent policy, the incentive will be a patent incentive–a contractor will be allowed to own a federal invention, to be exploited in non-governmental markets, provided the contractor has an established position in such markets and the technical capability to work the invention for those markets, and provided that the contractor actually does so, within three years of a patent issuing on such invention.

Latker’s direction is different. His incentive is offered not to contractors with capability but to other contractors, the non-profits and contract research organizations that have no non-governmental markets and generally lack the capacity to work any invention in non-governmental markets, at least not on any scale beyond that of providing research tools to other organizations. It’s here that Latker’s emphasis on “private risk capital” hits home. The implied pathway is that the nonprofits and CROs will grant licenses to the inventions that they are allowed to keep, having demonstrated that keeping exclusive control over such inventions will better serve the public than would federal open access. Latker argues that the only license worth granting is an exclusive license and the target for such a grant is an organization dealing in “private risk capital.” If nonprofits granted royalty-free non-exclusive licenses–well, that would be just what the federal government was doing, but without the bother of long pages of technical licensing prose to do it. If open access to invention rights was the thing, then federal agencies were doing it way better than nonprofits could hope to do.

Latker, though, focused on exclusivity–the power of a patent to exclude all other practice of an invention. That power may exclude, then, companies that would readily take up practice of invention without exclusive control themselves. The space for “private risk capital” then is even greater where companies would adopt without any exclusive position. A patent that can be used to exclude others that would practice an invention–and develop it–without exclusionary controls will have greater value than a patent claiming an invention no one cares to practice at all and so will exclude no one, ever. Latker’s design of Bayh-Dole is to introduce a patent barrier between federal inventions and public use. That patent barrier consists of a nonprofit or CRO federal contractor assigning federal inventions to companies formed by private risk capital, to be “developed” as commercial products–until sold off to an established company or to other investors. This is, essentially, what the venture-capital-backed biotech industry has done. But the federal invention contribution to this activity is not one of necessity–it is one of opportunity. Bayh-Dole allows this patent barrier to be created and exploited without regard to whether private initiatives would be undertaken without the need for patent exclusivity incentives.

Again, we are not talking about federal contractors chosen for their technical capability and activity in non-governmental markets. We are talking nonprofits, and especially universities and their research foundations. We are talking about these nonprofits dealing preferentially with speculative capital, using patent exclusivity even if such exclusivity would not be indicated if universities were dealing with companies that already had capability and ready markets. Essentially, Latker proposes forestalliing as federal policy. Universities take up patent rights, pass these to speculators (“private risk capital”) rather than offer the inventions openly to “the market,” and these speculators then sell into that market on whatever terms that market will bear–necessarily a mark up from “open access.” As Latker’s policy heirs in this debate argue, those terms must be solely what the speculators decide–commercially reasonable terms–or the “incentive” of the patent system will not be sufficient to get them involved at all with federal inventions prior to those inventions being practiced by the public or industry.

The alternative–that if speculators cannot be attracted to forestall federal inventions, using nonprofits as a front, then the public won’t benefit from federal research–then shows up as a glass almost entirely empty. The public might not benefit anyway. Inventions are easy and can be irrelevant to public interest and welfare. And for inventions that have compelling uses, companies with capability have good reason–and good track record–in adopting what they need, without any demand that one of them hold exclusive rights to block the rest of them. Latker was clever, but he was also bonkers.

Latker’s approach may sound seductive–inventions need owners, owners need the full incentives of patent exclusion, their financial return is the engine that attracts them and others back for more–but it doesn’t work in practice and doesn’t scale to deal with multiple federal agencies granting hundreds to thousands of contracts in overlapping areas of research and development. Consider: hundreds of contractors–a mix of for-profit and nonprofit contractors–in a given area of technological interest, each taking out patents on their little bits of invention in that area, and the nonprofits insisting that (as their incentive to participate) they will deal only exclusively to find speculative risk capital to commercialize independently their little bits. Patent exclusion combined with a demand for exclusive licensing makes the development of cumulative new technology impossible–platforms, impossible. Standards, impossible. Stuff gets delayed for 20 years until all those petty patents on bits of technology expire rather than used more rapidly and in better form than if it were all open access.

That Latker’s approach does not work may be considered from the evidence. Universities hold 50,000+ patents on federal inventions in the Bayh-Dole era, but cannot account for more than a handful of meaningful inventions “commercialized.” Maybe a hundred new drugs and new applications for existing drugs. And even there, it’s not clear that Bayh-Dole catalyzed something that would not have happened otherwise or was a parasite sucking energy from activity that would have taken place more readily without Bayh-Dole and its welter of nonprofit patent demands.

But Latker’s approach doesn’t work in a broader way–wherever a given “technology” requires multiple inventive bits in order to work, or work well, then many patent owners all seeking exclusive deals destroy the opportunity for the development and use of that technology. One or two patent owners have to defeat the rest and then may “clean up” financially on the potential market, if they can manage the development work to provide product for sale in that market. This situation is great for speculative investors. There’s nothing like the uncertainty of a market to create a betting parlor that promises great rewards for a few winners at the expense of everyone else.

And Latker’s approach places too great an emphasis on commercial product and not enough on methods, professionally sourced product (such as just in time or custom manufacture), technology platforms, libraries of technology tools, and standards.

We also see that Latker’s approach hasn’t worked by the sequence of amendments and claims made about the true “intent” of Bayh-Dole. Bayh-Dole came into effect in 1981. By 1984, Latker had the law amended (along the lines proposed in S. 2171) to weaken public protections. In parallel, Latker drafted the implementing regulations and standard patent rights clauses to further weaken federal agency capability to enforce public protections in the law. And later, by then former senators Bayh and Dole published articles claiming that their memory of their personal intentions for the law never included the idea that federal agencies would use the law to protect the public from unreasonable pricing of products based on federal inventions.

The weakening of public protections in Bayh-Dole doesn’t indicate that the law is working great and even more greatness can be squeezed out of it by further attacking public interests by attempting to attract private risk capital demanding even fewer limitations–and attracting that risk capital ahead of industry generally, and ahead of platforms, libraries, and standards. Rather, the weakening of public protections indicates that the law isn’t working at all. NIST proposes even weaker public interest terms–march-in for unreasonable terms cannot involve unreasonable price; the government license to practice and have practiced cannot actually authorize practice but must be restricted to federal “use” for limited purposes; inventors cannot own their inventions, even if made parties to a funding agreement, and despite the Supreme Court’s ruling that Bayh-Dole gives no authority to federal agencies to require inventors to give up their inventions to contractors that host federal research, inventors have to do so anyway.

These sorts of changes indicate the law hasn’t worked as claimed. There would be no need for NIST to propose “unleashing innovation” by rewriting Bayh-Dole upside down by regulatory fiat if Bayh-Dole was “working.” Bayh-Dole is a failed Latker prophecy propped up by a community of patent speculators that survive on misrepresentations of the law, on deceptive metrics to suggest the law has been successful when it has not, and on just enough speculator money–one lucky big patent deal every 20 to 30 years for a handful of universities–to cover the costs of maintaining the scheme and keeping university administrators willing to gamble more resources for the chance to get the next big deal.

Most federal inventions made in work hosted by nonprofits now get withheld from open access, aren’t licensed, and even when licensed, mostly don’t result in commercial product, and even in those rare instances, there’s no assurance that Bayh-Dole’s public protections have operated, will be audited for compliance, or will be enforced. Actually–the assurance is just the opposite–there won’t be, and never has been, federal agency action to enforce public protections required by Bayh-Dole.

And why should there be? Latker drafted Bayh-Dole so that the law requires federal agencies to use a standard patent rights clause unless they can justify some other clause following a process that pretty much ensures they can’t, but Bayh-Dole does not require federal agencies to enforce that clause. They can not ask for reporting. They can waive obligations. Even if they determine that an owner of a Bayh-Dole invention is harming the public with nonuse or unreasonable use or failure to require U.S. manufacturing, the federal agency can just walk away. Yeah, here’s 37 CFR 401.6(h):

(h) An agency may, at any time, terminate a march-in proceeding if it is satisfied that it does not wish to exercise march-in rights.

March-in isn’t march-in if a federal agency decides–no standard given–that it doesn’t want to march-in. Another instance of a Bayh-Dole implementing regulation thwarting the law.

But there’s another thing going on with Bayh-Dole. Latker appears to have wanted to get federal agencies out of development work altogether. The Harbridge House report on government patents noted that in the two study years covered by the report, federal agencies that supported development of inventions had a near 100% success rate when they released those developed inventions for public use.

When federal agencies develop inventions made in the work they fund to the point of practical application, industry adopts those inventions. According to Latker, Japanese companies are happy to adopt American inventions and develop them to practical application, even without exclusive patent positions. That leaves the problem of American companies. Was Latker right that American companies would not use what Japanese companies were happy to use, and that the reason why was that American companies refuse to use anything that doesn’t have a patent owner willing to assign rights to them? Nothing about the claim makes any sense. What does make sense is the idea that patent speculators might indeed choose to fund development of only those inventions for which they could obtain exclusive patent rights.

Latker, in drafting Bayh-Dole, aimed to overturn federal policy that public health inventions made in federally supported work ought to be open access unless a contractor could justify exclusive control as better serving the public interest. In his approach, exclusive patent control always better serves the public interest, by default, not by the evidence. By the evidence, exclusive patent control provides speculators with control positions that, on rare occasions, they can exploit for substantial profit. But for most everything else, exclusive patent control destroys opportunity by fragmenting cumulative research technology and disrupting the formation of platforms and standards.

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