Kevin E. Noonan, a biotech patent attorney, made an interesting assertion in a LinkedIn comment on the fourth article in this series. Maybe he was being flippant, but let’s consider:
People against Bayh-Dole just support private industry (much of it abroad) being able to steal innovation paid for by US taxpayers
Behind Dr. Noonan’s assertion is the idea that Bayh-Dole is intended to serve as a form of nationalist protection. Federally supported research, so the argument goes, is published openly. Thus, foreign companies can easily obtain those results, create products, and sell them back to U.S. taxpayers, and not owe anyone anything–those companies don’t necessarily pay U.S. taxes, and don’t pay inventors a royalty, and don’t pay the contractors who hosted the federal research anything either. Somehow, by getting nonprofits to take inventions away from inventors (and from research teams, and collaborators and introducing a bureaucratic licensing step that might never happen, especially if a university holds out for an exclusive licensee, this would correct the situation and now the U.S. would be again the world leader because those patents would (if universities sued) prevent foreign companies from “stealing” those otherwise open access research results.
Of course, Bayh-Dole does not authorize universities to take title to inventions, or provide a mandate to do so, or give any special privilege such as a right of first refusal to do so, Bayh-Dole doesn’t get nonprofits to take inventions away from inventors–even though that is what some “advocates” for Bayh-Dole have claimed. To be against those “advocates” presumably would not throw one into Noonan’s pack of enablers of company “stealing.” No–the real oompf in Bayh-Dole is 35 USC 209, which (i) authorizes federal agencies to grant exclusive licenses and (ii) requires any company seeking any license, exclusive or non-exclusive, to submit a plan for development or marketing of any federally owned invention–even if all the company wants to do is to make and use the invention. Again, it is not obvious how federal agencies granting exclusive licenses–and no preference for granting those licenses to U.S.-based companies–does a better job getting the U.S. to the forefront of global technology than does open access.
These federal-side licensing regulations then set up a weird strangeness for open access on the contractor side. If a contractor takes ownership of an invention (made in work with federal support) and file a patent application, then the federal agency (with a big budget for this sort of thing) can take the invention, file a patent application, and license the invention (if it can, if ever) exclusively. What’s the effect? The federal agency by holding a patent for exclusive licensing can cut a university research team from companies that it might otherwise work with on the research results. Bayh-Dole authorizes federal agencies to use exclusive patent licensing to threaten the very research teams that it has funded with a huge obstacle to technology transfer. A research team, then, might need its university host to take ownership and file a patent application, to save the team’s connections (present, anticipated) with industry from the potential predatory patent practices of the federal agency that has funded the team’s work.
You follow? Inventors may feel the need to ask universities to take title and file patent applications to prevent federal agencies from cutting the inventors off from their industry connections. Universities would have to patent to preserve open access. But it costs $10K or $15K or more to get a patent, using a patent attorney. An invention would have to be pretty darned important for a university to pay $10K just to ensure that the inventor can make the invention open access for all. The university will want someone to reimburse the patenting costs. The obvious deal that does this is the exclusive license (though that is not the most effective deal for reimbursing one’s costs). So university licensing offices default to seeking exclusive licenses. While they wait for an exclusive licensee, their patent position prevents anyone from practicing the invention, especially in the U.S. with its strong IP law enforcement and weak invention working requirements.
Somehow–back Noonan’s assertion–universities holding out for exclusive licenses because that seems to be the easiest way to recover their patenting costs will result in global American technology leadership. It’s not a particularly credible assertion, but perhaps there’s an explanation someone can give for how such a thing operates. All the better if one can show from 40 years of evidence that this such a thing actually operates. No post hoc nonsense, now, mistaking coincidence for cause.
Sen Bayh did premise the need for Bayh-Dole with an argument about U.S. loss of global technology leadership–Japan appeared to be the big competitor, and electronics was the area of concern–for which universities patenting more federally supported work would be the answer. Bayh did not bother to point out that the proposed law didn’t prevent universities from licensing to those same Japanese companies, and didn’t require anyone to enforce their patent positions. The provision in Bayh-Dole that insists on U.S. manufacturing–35 USC 204–is limited in ways that make it silly, for all of Sen Bayh’s posturing.
204 applies only to exclusive licenses to use or sell. So 204 doesn’t apply to non-exclusive licenses. A university could license to Japanese companies, to anyone, non-exclusively, and those companies don’t have to manufacture in the U.S. Once a university is granting non-exclusive licenses, then 204 doesn’t apply. A university could then refuse to license to companies that would manufacture in the U.S. A university could require any company to manufacture in the U.S. to pay higher royalties than companies manufacturing elsewhere (even U.S. companies manufacturing elsewhere). 204 still wouldn’t apply, so long as no license was exclusive to the U.S., to use or to sell. Nothing in Bayh-Dole requires non-discriminatory non-exclusive licensing if a nonprofit is going to grant non-exclusive licenses.
And that’s just the start of it. Even if a contractor grants an exclusive license to use or to sell in the U.S., 204 allows the federal agency to waive the requirement that any product based on that invention be manufactured substantially in the U.S. All that wind up about the need for national protection of inventions made by nonprofits and then when it gets down to it, federal agencies are delegated with the discretion to waive the requirement–as if federal agencies have any particular insight into why the requirement ought to be waived. The criteria for waiver are that a contractor tried but couldn’t get agreement for an exclusive licensee to source U.S. made product or that the contractor decided it wasn’t worth trying. Well, that’s some powerful (not) protectionism there.
Worse. The 204 waivable preference for U.S. industry applies only to limited exclusive license–to use or to sell. If a nonprofit grants an exclusive license to all substantial rights–to make, to use, and to sell–then the transaction is an assignment, regardless of the label (“Exclusive License”) on the document. The patent gets licensed, but the invention claimed by the patent gets assigned. A licensee that has a right to enforce the patent owns the invention. Courts have held repeatedly that conveying all substantial rights (even if rights are reserved for educational or nonprofit use) assigns the invention. Assignment then is not a transaction within scope of 204. A nonprofit, then, may assign an invention, even using an exclusive license to make, use, and sell, and avoid 204 requirements. Other parts of Bayh-Dole apply to nonprofit assignments–for which see 35 USC 202(c)(7)–but not the 204 provision. 204 would apply to the new assignee only if that assignee then set about offering an exclusive license of its own (which, given the naming convention of the original transaction, might be labeled a “sublicense”).
Now consider one further defect in Bayh-Dole. Bayh-Dole does not require a nonprofit that does acquire title to a given invention to file applications for foreign patents. A nonprofit can file in the U.S., and all Bayh-Dole requires is that a federal agency can then request to take the foreign rights. How often do you think that happens? Or if the federal agency does take the foreign rights that it files foreign patent applications, or if it obtains foreign patents that it goes off and sues companies operating in these foreign jurisdictions (even U.S. companies!) for infringement, so as to protect U.S. manufacturing interests? No, it just doesn’t happen, certainly not regularly, as would be needed to stop those thieving foreign companies.
For all that, a patent application is a notice of a future claim to a right to exclude all others, published 18 months after filing. What do those nefarious foreign companies do? They monitor U.S. patent applications and whenever one looks concerning, they get their engineers to work and file as many patent applications as they can on all the variations, applications, combinations, and improvements that they can. And they file as well in the U.S. Now come three or five years down the road when some U.S. university is out looking for its exclusive licensee to “develop” the invention into a “commercial product” has to deal with all those other encircling patents that aim to block that development until there’s a cross-license in play that defuses the effect of both the original patent and the company’s encircling development and improvement patents. It comes back to open access, but the hard way, and only for companies that can afford to play patent games.
What’s the upshot in university patent practice? The university locks up U.S. rights–no one can legally practice the invention in the U.S. until university bureaucrats give permission, and they by default uniformly hold out for an exclusive licensee–and the rest of the world, including those foreign companies pleased as punch to develop inventions into products without exclusive patent positions, has perfect freedom to go about its business, but for the U.S., where those products are now locked out by university fixation on granting an exclusive license and making money from it. And with more universities going proprietary on more inventions, there’s more motivation for foreign companies (and even U.S. companies) to patent around the university patents to block development, applications, and improvements until they get access in return. Bayh-Dole practice, then, delays, blocks, disrupts access to federally supported inventions primarily in the U.S. It’s bassackwards! To keep title, nonprofits should have to file for foreign patents, and delay, block, disrupt, and seek their money as a tax on foreign companies. And they would have to move fast to development and improvements to preclude foreign companies from obtaining patents to block those developments and improvements. But no, that is not how it is.
Then there is the utter stupidity of folks in the U.S. trying to export Bayh-Dole (or their fake version of Bayh-Dole) to the rest of the world. Everyone else, they argue, should try this same protectionist scheme. And what does that scheme do? Ah, it denies access in the U.S. to research supported by non-U.S. governments. And there’s a huge asymmetry in play. In the U.S. to mess with the rest of the world, one has to file patent applications in the U.S. and in the rest of the world–or in five or ten or twenty countries (go look at the list of countries named by U.S. universities in any PCT application). But if you have a nonprofit in, say, Mexico–it will file patent applications in Mexico and in the U.S. and be done with it. For all that, the nonprofit may not even care about Mexico, given the state of markets there, and file only in the U.S., where there is a much bigger market and where there are companies that may be willing to pay to have access to that market. Surely you see how this works against U.S. protectionism–Mexico nonprofits, should they exploit a Bayh-Dole-like law in Mexico, will delay, block, disrupt, and attempt to make money from the U.S. market, creating even more problems for companies with operations in the U.S.
Having made this little trip around Bayh-Dole’s section 204, we can return to Noonan’s assertion. Anyone against Bayh-Dole is in favor of foreign companies “stealing” federally supported research. Well, 204 does a crappy job at promoting the development of U.S. manufacturing capabilities to “catch up” with the Japanese (or whomever else is doing things really well). 204 is poorly conceived, ineptly drafted, promoting practices that run against what it purports to be about–that central reason for the law, to help the U.S. regain technological leadership (what a ludicrous policy goal anyway for nonprofit basic research and public health research). According to Noonan, being “against” this crappy drafting in 204 somehow aligns one with foreign interests. That doesn’t wash.
We can finish up by considering Noonan’s use of “stealing.” Open access is not stealing. Open access provides a basis for effective technology transfer, and for collaboration and competition both. Other things are often needed for technology transfer–assistance, access to data, access to experimental setups and specification of instruments and settings, and the like–but open on rights is a good start. Without ownership positions, there’s no need for the bureaucratic step of undoing proprietary positions with a license embedded in a contract that demands payment and attempts to shift the risk of use entirely onto the company recipient.
If, however, one has taken an ownership position on results, does not publish those results, and refuses to give permission for access, might one say that a company using those results has “stolen” anything. Even then, in some areas, results are so expectable that companies can independently come up with functionally similar technology without ever having access to the proprietary results held back by a university. In information technology, for instance, independent development is common. The patent, then, operates to allow for the publication of proprietary results that no-one can use for twenty years–and no-one can use even if developed independently. Companies then have incentives to design around the federally supported work–avoid it, obsolesce it, undermine it, lock it out of standards. They won’t get permission (only one exclusive licensee, if ever, will get permission–and a free ride to come sue anyone caught “infringing”).
Bad enough. Now add in that federal agencies spread overlapping research around. The agencies overlap on research. Universities overlap on research, and the are often scores of overlapping projects in any given area of technology development. Each university takes out its own proprietary bit in the overlaps and sits like a spider in its web waiting for a bug to engage in an exclusive relationship. If a company needs access to more than one or two bits of these proprietary rights defined by patents, it’s all but impossible. If all the universities and federal agencies practiced open access, there would not be this problem. But Bayh-Dole enables exclusive positions for both contractors and federal agencies. Anyone defects on the commons turns the rest into patsies. Bureaucrats hate being patsies, so they all defect on each other. I get how this situation is perfect for patent attorneys–there’s compliance, there’s patent work, there’s potential for litigation, there’s advising on how to avoid litigation, there’s license negotiation, there’s ownership policies and invention assignment agreements–gawd, it’s heaven for patent attorneys. But how this
utter mess of practice heaven for patent attorneys enables either effective technology transfer or American global technology leadership is beyond me.