Circumventing Bayh-Dole, Fitt the Second

Circumventing Bayh-Dole is easy. University administrators have been circumventing Bayh-Dole since the law became effective in 1981. Consider next

2) circumvention by exploit–to use Bayh-Dole’s limitations to end up with the best deal under Bayh-Dole and avoid public covenant requirements.

Circumventing Bayh-Dole without Noncompliance 

Our next circumvention is distinctive because it does not involve a failure to comply with Bayh-Dole. Instead it exploits the architecture of the Bayh-Dole apparatus. Under Bayh-Dole, universities have no obligation to take assignment of inventions made with federal support. Even if a university implements the (f)(2) written agreement requirement in the standard patent rights clause–which makes most any invention otherwise within scope of a federal funding agreement also a subject invention–the university does not have to own the invention.

If the university does not own a subject invention, then a different invention rights clause comes into play, that of 37 CFR 401.9. That invention rights clause instructs federal agencies to treat inventors as small business contractors but with even fewer restrictions than normal small business contractors. Under 401.9, inventors have no restrictions on assignment or use of income from licensing, don’t have to prefer small businesses, and even might not have to grant the government a royalty-free license. The inventor patent rights clause is the most advantageous of all the patent rights clauses that have been created under Bayh-Dole’s authority.

Now consider the logisitics. An inventor having made the university-required (f)(2) written agreement to protect the federal government’s interest in inventions, reports a subject invention to the university that hosts the research. The university in turn does not require assignment of the invention but instead takes an option to acquire title based on what happens next. The university then reports the invention to the federal agency on behalf of the inventor, but does not indicate whether it will elect to retain title (and, for that matter, it cannot elect to “retain” title because it does not have any title to retain). The relationship between the inventor and the federal agency is then governed by the (f)(2) written agreement and the inventor patent rights clause at 401.9. The federal agency has a choice: it can allow the inventor to retain title under 401.9 or it can request title under the (f)(2) written agreement (if the agency has a regulatory framework that incorporates such a requirement into its funding agreements–Bayh-Dole doesn’t provide that framework). Either way, the agency is set up for the circumvention.

If the federal agency allows the inventor to retain title–which under Bayh-Dole it must, since the inventor is by virtue of the (f)(2) agreement also a contractor–then as soon as that arrangement is official, the university exercises its option on title to the invention and thus obtains ownership of a subject invention with only the inventor’s patent rights clause obligations–401.9–not the standard patent rights clause that would otherwise apply to the university–401.14(a). The only requirements, besides filing a patent application and such, is to report utilization, require substantial manufacture in the U.S. for an exclusive U.S. license to use or sell, and to accept march-in procedures. There’s not even a requirement that the inventor grant to the federal agency a non-exclusive license. That’s sweet. Why would a university want to manage an invention under the standard nonprofit patent clause at 37 CFR 401.14(a) when it could have an even better situation under 401.9?

If, however, the federal agency requests assignment of title from the inventor, then the inventor goes ahead and assigns title to the university instead. Now the invention is under the university’s standard patent rights clause (401.14(a)) rather than the inventor’s patent rights clause, and we are back to dull, ordinary Bayh-Dole land.

Of course, the university doesn’t need to take an option to acquire subject inventions. It could just let inventors go their merry way and discuss financials only if the inventor requests university involvement–help with patenting, access to research and development resources not otherwise available, licensing work, and the like. In such a case, Bayh-Dole reduces to (i) patenting, (ii) U.S. mfg for exclusive licenses in U.S., and (iii) possible march-in (which never happens)–and even enforcing these requirements in the standard patent rights clause is at an agency’s discretion. Bayh-Dole, you see, requires only that the standard patent rights clause includes these requirements. Bayh-Dole does not require a federal agency to enforce these requirements once they are placed in a federal funding agreement (or any other requirements, for that matter).

A second compliant circumvention of Bayh-Dole involves focusing on the development of a subject invention rather than the subject invention itself. If an invention requires “development” to be useful, then there’s got to be much uncertainty about how one goes about the task. That’s the difference between Feynman’s “invention” of atomic rocket-propelled airplanes (much development left to make such a thing work) and the “invention” of a method to better focus one’s telescope in hot weather (which one has already got working and the issue for others is re-implementing). Thus, we can make a working assumption that if development is necessary, then there will be more inventing to come. That is, the line of work from first invention to commercial product will involve multiple additional inventions.

In engineering and chemistry, there might not be at first even a “line” of development. It will be more like a random walk through a potentially huge set of variations. When the compound in sweet clover hay that was causing cattle to bleed to death was finally isolated by Karl Paul Link and his students, they went on to create a hundred variations on the compound and worked to characterize these compounds. One of these compounds ended up becoming warfarin, a rat poison. Another was targeted as a blood thinner, but it didn’t work so well. Similarly, when Sawyer and Jung collaborated on compounds effective at treating prostate cancer, they produced over 150 variations that UCLA licensed to Medivation, a company looking to invest in new drug candidates. Xtandi was compound 169′. What UCLA didn’t tell Medivation was that they had a second series of compounds that did roughly the same thing, but were not within the scope of the licensed patents. So UCLA started another company to compete with their first company. Yeah, there were lawsuits.

In engineering, variations in design can shift out from under one patent to another. Or work that starts in the public domain can quickly move within the scope of new patent claims as variations or improvements or alternative methods or added components or features or steps are necessary to create a working product from a working initial invention. This stuff happens all the time, in engineering a product might have fifty or three hundred distinct inventions, any one of which might be enough to prevent others from copying the product or making a functional equivalent. What’s the point? In many development environments, the first patent is not as important as the later patents. Z Corp (since acquired), for instance, used its exclusive patent license from MIT to file a bunch of additional patents on components of powder-based 3d printers, and those additional patents preserved Z Corp’s monopoly for much longer than the life of the original patent.

Bayh-Dole is built on the assumption that one invention, one patent, one exclusive license will create a product. This assumption is in general a rather poor one, resulting in FOIL technology as the primary outcome of federally supported research–fragmented ownership, institutionally licensed. Even a market-busting, anti-trust edgy monopoly sounds more likely to be successful. A better assumption is that there will be multiple inventions with multiple owners, multiple patents, and multiple licenses to develop any given product, and that some of these patents will be used to *prevent others* from developing versions of the product that one has already ruled out as being less profitable (harder to make, limited features, directed at smaller markets, directed at less important applications–save the life of your pet lizard?–hah, we are saving the lives of race horses!–race horses? hah-hah, we are saving humans!).

Here, then, is the circumvention of Bayh-Dole. Go ahead and turn that initial invention into a subject invention if you can’t work the system to induce the government to allow the inventors to retain rights and work a private deal with the university. But move further research out from federal funding, such as into a startup company that’s a shell company that “leases” space in the same lab that did the original inventive work. Now do the development work in this shell company, but without federal support, and file patent applications on the inventions that result. Aim to surround the original subject invention so that key pathways toward applications are blocked by these new patents. Once you have the new patents, release the core POSI to the winds–do anything but give up ownership.

When it comes time to license the package of patents to a commercial concern (and the license might take the form of acquiring the shell company and assigning the license to the new owner, perhaps with some changes), the POSI–the patent on the subject invention–is licensed for no royalty. There’s nothing to report to the federal government, no worries about exclusive licensing or small business preferences and the like. No fuss over whether the exclusive license is an assignment, and no bother with restrictions on royalties. Instead, the university licenses the additional patents–the variations, the improvements, the steps toward application–and these carry the value of the deal and whatever conditions are necessary. Since these other inventions are not subject inventions, Bayh-Dole does not apply. Even if the later inventions are closely related to the federal research, they aren’t subject, as 37 CFR 401.1 makes clear in discussing the proper scope of subject inventions.

There is nothing at all in Bayh-Dole that requires a patent owner to place value on a POSI or generate licensing income at all from a subject invention. (There might be such a requirement if the federal government expected a share of any licensing income.) Thus, it is entirely up to the university what elements in a commercialization relationship will carry the value of any upside back to the university. To circumvent Bayh-Dole and its public covenant and its fussy apparatus, all one has to do is to zero out the value of the patents on subject inventions and ride the improvement patents that are not subject to Bayh-Dole. There’s still the cost of the original patent work, yes, and there’s still reporting to be done, but the circumvention makes the reporting simple–the patent is available for all to use, royalty-free. Except no one can actually do anything with the invention because all the important next steps are covered by other patents, which have been chosen to carry the speculative and perhaps also commercial value. Nice, eh!

As an alternative, if the development of variations or of improvements or of applications moves outside the scope of the original subject invention, so that the subject invention is no longer part of the product design, then the university can simply drop the management of the subject invention and focus on the non-subject inventive developments. So what if the federal government patents the original subject invention? It’s moose turds at that point, and the commercial work has long departed the scene–by design. This alternative is not for all subject inventions, of course, but can work for those in which the research shows signs of bounding from first variation to others.

Getting the subsequent work immediately out of the scope of the federal funding is the first step in working the circumvention. So long as the federally supported research has not proposed, as planned and committed activities, to pursue the commercially relevant variations, then subsequent work is not encumbered by Bayh-Dole. Why not get inventions free and clear when one can? Why carry around in your shiny new car all the floor mats and back seat clutter from your old, federally funded car when you don’t have to? Circumvent–and all the while complying with the law, unlike most university administrative circumventions in play now. Isn’t it odd that university administrators likely will balk at using these compliant circumventions while defending their present, non-compliant practices? Too funny.

If Bayh-Dole is such a shallow law that university administrators feel the need to cheat on it to make it “even better” or “what it was intended to be,” then surely the best approach of all is to use methods to exploit the law to commercialize new things. What does it matter how the law is exploited–whether by circumventing its requirements through indifferent non-compliance or by complying in a way that avoids most of the administrative mess of the law?

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