We are working through an op/ed published in April 2021 by Niels Reimers, one of the recognized university TLO leaders from the 1970s on. We are working through it now because the Bayh-Dole Coalition is using quotes from it to try to influence federal policy makers to avoid using the public protections built into Bayh-Dole–and because Reimers just published a second op/ed, and we may as well take on both of them at once.
In his first op/ed Reimers has just made the claim that public-private collaboration (apparently imbedded in Bayh-Dole practice) has “spawned thousands of startups.” And I’ve pointed out that Reimers hasn’t made the connection between those startups and Bayh-Dole. He just switched metrics in the middle of his discussion, as if no one would notice, or everyone would assume that because there were startups after Bayh-Dole, they must have been caused, or “helped” or something by Bayh-Dole, and would not have happened, at any rate, without Bayh-Dole.
When a company takes a license to an “early-stage” university invention, development is exactly what the company will have to do, either on its own (exclusive license) or with others (non-exclusive license, shared platform, common standards, pre-competitive collaboration). Bayh-Dole then points to “maximum” participation in development. Bayh-Dole stipulates, then, non-exclusive licensing that favors small companies–a maximum number of small companies. Exclusive licensing to one startup runs against Bayh-Dole, as it excludes all the other small companies that otherwise would have an opportunity to participate in development. Same for “free competition and enterprise.” Free competition doesn’t happen, with regard to any new invention, because the rights get licensed exclusively to just one company. The competition that necessarily comes about, then, is between the single licensee and the rest of the industry, which has to design around or undermine or ignore.
Here’s a way to think about it. University administrators were determined to take ownership of all inventions made by university personnel. They misrepresented Bayh-Dole to give them a federal mandate to take all inventions made in federally funded work. The Supreme Court ruled they were wrong, but they have kept at it. So they got all these patents–50,000 and counting–citing federal support and tried to license them exclusively, and had a miserable time of it. Mostly, they licensed next to nothing. So they got the idea that rather than trying to license exclusively to existing companies, they would start their own companies, license to those companies, and then seek private investment to fund “development.”
There were a few successes this way, but mostly the startups lived off phase I and phase II SBIR grants until they wasted away. So the administrators had to find other forms of funding, and got states to set up “seed” investment funds. The state of Washington, for instance, used tobacco liability lawsuit settlement funds to set up a seed fund for biotech companies in the state. Much of that funding ended up with university startups. In the case of both the University of Utah and the University of Washington, administrators started creating paper companies without real operations, using university addresses and web sites, but pitching these companies as a proxy for the potential for economic growth. Look at that innovation, will ya! But it was all a sham, innovation theatre to get more money from the state that ought to have gone to support existing small companies but instead went to universities. Rather than using new technology to stimulate existing small companies, universities have abused Bayh-Dole to start their own, do their licensing deals without arm’s length interests, and those deals (mostly licensing rights back to inventors) work to cut out access by small companies–just the opposite of what Bayh-Dole states as its objective.
With one technology we had–not Bayh-Dole, thank god–we licensed non-exclusively and at a price within the discretionary spending limit of a typical group lead in a biotech company. Anyone could get the license, and most of the money we received went back to the lab to provide support for the licensees. They saw it as a great deal–pay $5,000 and get back access to $500,000 worth of programming, a common platform, and direct access to the programmers for any questions. At one point, the CEO of a small company thanked us for the program. “If you had done this the usual way, as an exclusive license to some company,” (I paraphrase), “my company would never have been able to get access at all, and couldn’t have distributed our own version.” We maximized participation by small companies. Unlike most TLO practice under the influence (but not compliant with) Bayh-Dole.
Prior to 1980, America was losing its economic edge, especially within ideas-driven industries like the life sciences.
Maybe, but since 1968, the NIH had been running an Institutional Patent Agreement program–a scam, really–under which nonprofits could circumvent executive branch patent policy and take out patents on biomedical inventions made in work with NIH support. If America was losing its “economic edge,” it wasn’t because inventions made in NIH work weren’t getting the benefit of private patent exclusion. Norman Latker, patent counsel at the NIH, implemented the IPA program (which was also adopted by the NSF in 1974), and later it was Latker who drafted Bayh-Dole, which he claimed was an implementation of the IPA program. The IPA programs were shut down in 1978 as ineffective–and against public policy, which they were on both counts. Of the 96 inventions disclosed to the NIH, only 4 produced any product. The success rate under the IPA program was about 1/4 that reported for non-federal inventions managed by university-affiliated licensing organizations–and also about 1/4 the rate for federally owned biomedical inventions. The IPA program was a total failure, unless a few universities making money from less than a handful of inventions while suppressing over 90 others is a technology transfer success. Which it isn’t.
If the U.S. was losing its edge in the life sciences, and federal patent policy had anything to do with it, then the very thing Bayh-Dole purports to do was already the thing that the NIH and NSF were already doing on the sly–and it wasn’t working. Nothing like politicized institutional bureaucrats to see that something isn’t working and work up a sweaty campaign to do more of that same thing, and make it government-wide.
America had some of the world’s greatest research universities. But there was no easy way to translate federally funded academic research into commercial technologies.
This is just fantasy history. True, the IPA programs were not an “easy way.” But the IPA programs were proto-Bayh-Dole, so again Reimers is being deceptive. And the big question–how to transfer federal research results into anything useful–had more to do with the problem of the relevance of federally chosen research projects to the work undertaken in companies than it had to do with particular patent positions taken on a subset of that work. The pharma industry did not demand exclusive rights to federally supported research results–and they have a track record of using all sorts of research results in their drug development efforts.
No, the pharma industry was concerned about what it called “contamination.” If a pharma company worked with a federal agency or with a federal contractor in “developing” and invention in which the federal government had rights, did the federal government’s claims extend to any inventions that the company also had, directed to the same area of disease research and development? If so, then pharma companies would decline to “participate” in the federal efforts. The bureaucratic goal, then, was to enable “participation” so that NIH research efforts could be connected to the leading pharma companies. Think of it as a federal need for reputational associations. Or, as a showing of “public-private” partnerships, but now based on a limitation on the scope of federal claims on inventions–“you companies can help us develop new medicines based on federally sourced inventions, but if you have your own inventions directed to this same effort, you don’t have to make those available as well–we just want to be associated with you to validate the importance of the federal research. It’s not really about working together to solve a public health issue.” Blunt. But don’t you get tired of bureaucratic abstract wording that hides it all?
Of course, if universities were going to drop exclusive licenses in their laps, pharma companies and biotech firms were not going to say no. The result has been that university TLOs insisted (and continue to insist) that the companies they try to license to won’t take licenses if they aren’t exclusive. The TLOs don’t bother to point out all the research results these same companies use when there are no patents, no licenses, no royalties, no audits, no exclusivity, no restrictions on sublicensing, and the like. I feel a John Lennon anthem coming on here, but let’s not.
We can take Reimer’s worry about “no easy way” and “commercial technologies” another route, too. The stated objective of Bayh-Dole is “utilization” not “commercialization.” On its face, Bayh-Dole does not address the problem of “translating” research results into commercial products. Instead, it concerns itself with use. And for that, Bayh-Dole states a restriction–use the patent system (only) to promote the utilization of inventions arising in federally supported work. 35 USC 200, again. I’ve supplied the “only,” which is necessarily implied–otherwise, the law states a truism that doesn’t mean anything or add anything to the law: “use the patent system to promote (or use it not to promote, we don’t care) utilization of inventions.” You see, that reading doesn’t work. So there must be an implied “only.” In the way Bayh-Dole lays things out, open access–free competition–for utilization may lead to commercial opportunities. By using the patent system to promote use by, say, non-exclusive licensing favoring small business access, a university de-risks a new technology, spreads the development and testing costs throughout an industry, and creates cumulative technology, technology platforms, and working standards. From these assets might arise commercial opportunities.
Instead, university IP practitioners, aiming for high-money patent license deals, skip the utilization step that is front and center in Bayh-Dole and claim that only by exclusive licensing a patent monopoly might some research result be “commercialized.” This claim may be true to the extent that the IP practitioners demand that the only licenses they will grant will be exclusive and directed toward commercialization–mass production of a product, paying royalties–rather than toward utilization. Of course, if these are the only deals IP practitioners will accept, then that’s the only way anything is going to happen–or nothing happens and the IP practitioners then talk about how crunchingly hard it is to do deals the way they say Bayh-Dole has set things up. Bayh-Dole provides “no easy way.” Reimers has things backwards, but then that’s what politics can do. Federal government open access–whether there’s a patent or not–was an easier way than any practice under Bayh-Dole.
At the time, most university advances backed by federal dollars never left the laboratory. The patents simply reverted to the government, which generally did little with them.
The patents did not revert to the government. The government took control of the inventions and made them available to the public, mostly with no licensing requirement. In the few instances in which licensing was required, the licenses were royalty free and were used to screen “applicants” for capability. Everyone had access–the university (which did even worse with inventions) and inventors (who did not have to worry about university administrators demanding exclusive control), and entrepreneurs and companies (who could work with anything and establish proprietary positions on inventive improvements).
When stuff doesn’t leave a laboratory, it generally means that no one is leaving the laboratory, even to consult, or to graduate. No one is visiting the lab who would care. The transfer problem is often a people moving problem, and a patent monopoly that starts from the premise that no one else has access doesn’t much change that.
When lawmakers drafted Bayh-Dole, fewer than 5% of federally held patents had been licensed to private companies.
More nonsense. No, really, this is not just a plausible difference of opinion. This is nonsense. Most of the patents held by the federal government when Latker drafted Bayh-Dole for the lawmakers to front were defense-related. The Defense departments (and the Kennedy patent policy) allowed federal contractors to deal in patents on inventions made under contract. The inventions that the federal government obtained, for the most part, were inventions that the contractors did not want. The only area under the Kennedy (and Nixon) patent policies where company contractors generally had to accept federal open access was public health. If a contractor did want exclusive control over an invention that a federal statute directed was to be federally owned, they could request for a determination of rights, and with some delay, obtain a federal release. Except, generally, for public health inventions, where DHEW routinely refused to waive the government’s claim. Under the Kennedy patent policy, a federal agency could waive the government’s claim to exclusive rights (and so to open access) if a contractor could show that its exclusive control via patent of a given invention would better serve the public than open access.
Bayh-Dole comes along and says contractors no longer have to make a case that their dealing in patent monopolies on inventions in public health will better serve the public than open access. That’s the heart of it. Bayh-Dole’s great advance is that contractors don’t have to bother with what will best serve the public. If they can make money from denying access, that’s all good under Bayh-Dole, or so folks like Reimers would have you believe.
So those federal patents were ones companies had already rejected. As Admiral Rickover explained in Congressional testimony, the defense departments obtained the patents as a kind of, er, “defensive” filing. Yes, it was his term! Anyone could use the inventions–but no one could get exclusionary control. No one needed licenses, so whether there was 5% licensed or not really doesn’t matter, other than for political spin to those not in the know, to mislead them. As it is, universities now hold more than 50,000 US utility patents citing federal funding, most not licensed, not utilized–and most definitely not available to just anyone, without a license or with a license that will be royalty-free.