We may posit that for any “concept” that we can pull out of the air, there’s an implied “cost” to make that concept into a “commercial product.” We could imagine a w
alking cat-bus, say. We could work out the robotics, the software controls, the tiger-like furrish finish, the headlamp eyes; also conditioning the market for accepting walking public vehicles, training drivers, and the like. We could do all these things and produce a walking cat-bus, but it might not turn out to be profitable. We might find that the public won’t even subsidize it. Thus, just because a walking cat-bus might be patentable (well, it’s not patentable as a concept, obviously, because it has been anticipated, if not by Studio Ghibli, then by Chinese engineers) doesn’t mean that there must be some effort to create a commercial product because only then might the public benefit from the concept. The patent doesn’t come first in the logic. It comes later, and as an option. If we are concerned only with concepts that the public must have, and to have must first be commercial products, and to be commercial products requires great expense, and that great expense must be borne by a single source, and that source is not the federal government or a state government or a nonprofit institution, and the source of funding expects a financial return on that expense (i.e., some form of profit), then a patent might be the thing.
Bayh-Dole, in encouraging this sort of thought, makes it appear that every invention that might be made at a university, or at least in federally supported research, must be subject to this sort of calculus, but in the reverse order: first, obtain a patent; then, determine to create a commercial product (rather than offer a tool or enable industrial use); then exclude all other forms of support but for a single investor (so, no commons development, no consortia, no standards, no foundation or government support); then have that investor spend whatever is necessary to create a product (and so drive up the investment required and the need for substantial return on that investment); then use the patent’s monopoly position to receive the maximum return on that investment. Call that return on investment the inducement for investors to participate in this scheme. But point out, too, that the problem has been transformed into finding wealthy, speculative investors willing to share huge potential profits, but only if they have a monopoly for the entire run of the patent. Wealthy, risk-taking but sharing speculators. Now, that’s a rare breed of speculator. Talk about narrowing one’s market.
It is almost exactly the reverse of what university research produces that has any prospect for practical application for public benefit. Bayh-Dole, so the thinking goes, imagines that every invention made in university research must be handled as if it could be a profitable commercial product produced at great expense by a single private payer. The more obscure the invention, the more important it is that a monopoly in the patent right be preserved so that an investor will come forward and make the attempt. The argument goes then that universities must manage all inventions this way by default and must insist that a single private payer be identified or don’t license the invention at all.
This is not how things are depicted to the public, of course, but it is the practice. Here’s Jones, giving an account of the effort at MIT in trying to find licensees for its patents:
Again, it all sounds very good and reasonable. And as testimony it is designed to have the attributes of good and reasonable. But the underlying practice creates the problem. If a patent owner approaches companies with an offer to license, it amounts to a threat. The companies then are on notice that anything they do in the same area will be subject to a threat of infringement litigation with treble damages and attorney’s fees. If the companies have been sought out because they are already known to be working in the area, it’s even worse, as the invention now presents a potential barrier to further development. A company will have to decide whether to work around the patent or deal with it in some way. How important is the invention?
But the problem is, the company knows the university’s response if no one takes its offer of a royalty-bearing non-exclusive license with a demand that the company turn the invention into a commercial product–the university will offer someone an exclusive license and if it’s not that company, then the company will be locked out of the invention and a competitor will have exclusive rights. Or–there is an alternative–no one will license the university’s patent and in a couple of decades the invention will enter the public domain, where no one will likely care about it. The university’s offer of a non-exclusive license is set up to create the conditions for moving to an exclusive license. And the exclusive license is set up to reward the university for granting all rights in the invention–make, use, and sell–for the entire term of the patent.
Imagine that the university granted a non-exclusive, royalty-free license outright in any new invention. Would companies turn down such a license? Generally, no. But their concern would shift to developments and improvements and work-arounds, starting with the university lab that generated the initial invention. What about follow-on work? Will the university handle additional inventions in the same was as the original invention? Has the university already got a second invention that does things a bit differently that it isn’t talking about (what UCLA did to Medivation, say)? Will other companies rush to develop improvements and applications that if patented will block any further development by other companies, creating gridlock or an eventual monopoly for whoever gets to the patent office with those developments first? Even a bare naked upfront non-exclusive license may not be attractive to companies, then. Instead, one has to show that there will be a pre-competitive commons from which, at an appropriate time, competitive variations can emerge that provide advantages for their developers. This is the approach of “crossing the commons” to arrive at competitive positions that depend on the non-essential features of inventions contributed to a standard–faster implementations, with better quality and safety, at lower costs of manufacture or assembly.
A mere non-exclusive license does not manage competition to block others from access. That’s why in open source licenses, there is often a share-alike requirement in the form of a trade of license for commitment not to sue for patent infringement for technology that has been contributed to the open source platform (such as the Apache 2.0 license). Such commons then provide a neutral ground to develop a shared platform of technology, any combination of which might be developed into a commercial product where there’s a specific application without preventing others from also competing. What’s the advantage of this approach? First, many companies may want the technology for their internal use and not to create a product. They don’t care if there’s a commercial version of the invention. They care that the invention is developed for commercial use. For that, they may be happy to contribute their share of the expense and technology, so that they gain access to the benefits of everyone’s contributions. This is the genesis of industry research consortia and standards. It’s better to have a platform maintained by the industry for its use than to have a single monopoly investor–especially if not already an industry leader–seeking to maximize profit return at the expense of everyone in the industry.
While there are times when a single monopoly investor may be the only way to develop a commercial product from a university-hosted invention, those times are remarkably rare. Bayh-Dole encourages university patent brokers to make it appear that those times are frequent, are the general case, just as Thomas Jones argued in 1978. The subtle change then is that Bayh-Dole gives university patent brokers the cover to break from the idea of widespread dissemination, offer a non-exclusive licensing program that’s unattractive, and then move over to monopoly methods with all their complexity but with the prospect for more lucrative deals, even if only one a decade.
Think of it another way. Bayh-Dole encourages university officials to pack all of the possible financial return in university research deployment into patent licensing instruments. If an invention is licensed royalty free, so that there is open and happy use of the invention immediately, then no money flows through the license and thus what is the point (so the argument goes) of the patent? (Well, publication in the literature, creating a higher bar for improvements and applications to create an initial commons, and protections against imports that unfairly compete with domestic development). But if there’s use by industry, then there is often opportunities to provide support for that use–research contracts, consortium formation and management, training workers, training students as future experts, consulting services (by individuals), supply of data sets, of test standards, of specifications. If a university lab becomes central in a developing network of use, then it will see new areas for research, problems that if solved will have general application, and the like. These, too, become valuable assets for the lab, guiding its work.
From a tax perspective, however, such support has to be managed so that it does not become “unrelated” to the university’s exempt purposes or the income a matter of “unrelated business” operations. One way to preclude such findings is to involve students in all aspects of the work, rather than employing only research professionals. Training students, you know, a basic exempt purpose. Another is to focus on teaching industry how an invention (and related technology works)–that, too, is standard exempt purpose fare–people pay tuition for instruction; so can companies. If the university’s work is focused on developing the public commons (library of tools, research platform, ad hoc standard) rather than working on the design of any particular commercial product in competition with others in industry, the commons itself becomes the protection against a finding of unrelated business income–widespread dissemination in the public interest, for which the university receives support from a number of sources for its public service.
Jones then moves through a discussion of the value of a profit motive, without which nothing gets done, and the patent system as a Constitutional expression of the profit motive. Jones does not acknowledge that there’s no profit motive set forth in patent law, and that the Constitutional premise for the progress of useful knowledge is based on the publication of inventions in exchange for exclusive rights. There may be a profit motive in those exclusive rights, but there’s nothing in the law that depends on a profit motive, and there’s nothing to show that the profit motive in patents is better than the profit motive in trade secrets or the profit motive in open publication of a new discovery. It all depends on one’s perspective on where one’s profit is likely to come–from a patent monopoly, from knowing more and better than others, or from getting acolades and better paying research positions.
Jones provides a list of reasons why the federal government is poor at commercializing research inventions and therefore should not own them. Jones ignores the idea that inventors own their inventions or that they might assign inventions directly to companies for development and moves instead to the idea that universities should own inventions and “transfer” them to industry–that is, to companies with a profit motive, supported by an exclusive position reserved for them by universities acquiring patents on research inventions.
Also omitted from Jones’s development is the problem of fragmentation. It’s fine if MIT invents magnetic memory for computers. But if hundreds of universities are working on carbon nanotubes, and each patents the heck out of its discoveries, then before any company can work on any nanotube, it needs scores of patent licenses, each from a different university on different terms, all demanding a royalty and that their technology is the most important one of the bunch–for the first and last licensors in, this may be in some way true. But overall, the lack of uniform treatment of licenses by universities is far, far worse than the lack of uniform treatment of patent ownership was by the federal agencies outside the IPA program and pre-Bayh-Dole. Why make federal agency patent practice “uniform” (i.e., arbitrary) and let universities present a fragmented, wildly varying front to industry? Seems like a job have done.
Jones picks up the idea that patent licensing is difficult, but presents the core arguments that persist to this day.
