We have looked at an article by Boettiger and Bennett reviewing the Bayh-Dole Act after 25 years. We have picked over the description of the law and pointed out how our authors mischaracterize the law to their own disadvantage. Bayh-Dole doesn’t “shift the incentive structure”–it suppresses an approach without an “incentive structure” and lets an approach with an incentive structure to dominate. Bayh-Dole also doesn’t require a change in university policies–it implements a federal policy through a standard patent rights clause that supersedes any university policy or agreement on the same subject matter. Bayh-Dole doesn’t require patenting (a university must choose to patent), doesn’t require anyone to “encourage development” (a university must promote use). This stuff isn’t subtle unless one doesn’t know the law or has been conditioned to university-speak for years.
Having characterized Bayh-Dole, our authors make an insightful observation:
Many of the issues that are identified today as negative consequences of Bayh-Dole can be traced to the institutional policies structured to optimize institutional benefits and income, rather than to the Act itself.
There is something to this statement, but again, the cognitive dissonance is intense. First, we’ve been led through a characterization of the Act itself that does not track the Act itself. Second, the Act itself hasn’t been enforced, so it’s difficult to see how the Act itself can be adequately evaluated. What’s happened has happened because federal agencies are asleep at the wheel and university administrators have been eager to exploit whatever they can. Somewhere, not in the Act itself, university administrators were told (or began telling each other) that Bayh-Dole mandated that they seek money from patent positions.
There’s more, of course, to the cognitive dissonance. Whatever university patent policies have come to be, they surely are not “structured to optimize institutional benefits and income.” That’s about the last thing these new policies do. They optimize nothing–if “optimize” means to “make the best” of an opportunity. These policies expand ownership claims without selectivity, defocusing university licensing efforts. These policies provide no guidance on the purpose of university patent ownership–they waffle around about public benefits but lose their way in abstractions, so that most anything is possible, including trolling industry.
These policies result in practices that cost the university more money than the university makes by licensing. For state universities, breaking even or losing some money is apparently worth it if it allows the administration to launder state funds (for research administration) into royalty funds (free of state restrictions). Anything better than break-even is gravy, in this laundering mindset. But the worst effects of these policies shows up in practice, the own-everything and license-exclusively approach, the “dual monopoly” approach, performs worse than most anything else. Every ten to thirty years, it may result in a big financial gain–half a billion dollars, maybe more. But in the interim, most of the time it runs at a loss, or at break-even. It’s a gambler’s mentality. There’s nothing much one can do to “optimize” gambling but card counting and cheating.
If we adjust our authors’ wording a bit, we might say that universities have implemented policies that demand ownership of inventions and authorize the licensing of inventions for profit, giving preference to licensing arrangements that involve monopolies. It’s not clear at all whether such preferences result in more income or optimize a licensing program’s activities. It appears that the argument for monopolies came from the pharmaceutical industry, where it was considered a good. That argument then was adapted to the universities–if they obtained patent rights from the government, then the pharmaceutical industry would take exclusive licenses from the universities, for a royalty of 0.5% or maybe 1% or maybe 2%. Better to pay a little and gain a monopoly than to refuse altogether when there are billions to be made.
The expansion from pharma to all other domains, however, reversed the reasoning. Now, rather than “in our industry, a monopoly is good,” it became, “because monopolies are good, they should be imposed by federal policy on all other activity, industries, and markets.” In the bozonet idea of innovation, paying monopolies are the driver of public benefit. “Exclusive licensing”–often actually assignment of patent rights–comes to be seen as a public good in itself. Monopoly is the only good thing that a patent is good for, and Bayh-Dole’s genius (in this rendition) is to give universities a financial share of each monopoly. As unreflective university patent licensing folk are fond of repeating, “a non-exclusive license is just a tax.” Of course, were that true, then universities would not be out a-trolling their patents once the exclusive licensing route has failed. Trolling produces non-exclusive licenses. These being just “taxes” serve no promotional or encouragemental purpose. There would be no reason to troll a patent. Of course, universities do troll their patents, because their patent policies are about making money, and trolling patents aims to teach industry a lesson that it must not ignore university demands for exclusive licensing.
There is also a quote around, equally unreflective in the context of inventions, that things that aren’t owned aren’t valued. When we own things, we tend to value them more than the same things merely in our possession. And there are odd arguments that support the proposition: that if someone owns a forest, that someone will take better care of it than if the forest is open to anyone’s exploitation. There are arguments about wild animals in Africa, that if they are owned, then folks will look after them for profit better than if the animals are not owned by anyone. That’s an argument for the power of money over, say, ethics. Maybe we learn something there.
But in the context of inventions and patents, the ownership argument gets muddy as it expands from possession to exclusion. An invention in owned by common law. When you invent, you own. No statute needed. What does a patent do for you? It allows you to exclude others from doing what you have figured out how to do. Owning a patent is different from owning an invention, then. Two different things. Thus, to argue an invention that’s not owned won’t be valued as much as one that is owned only works if the invention is not all that useful. Otherwise, regardless of differential “value,” a useful invention will get used, if it happens to fall into the right hands at the right time, and an invention that isn’t so useful won’t be used–it doesn’t matter that the invention is in the public domain. Or, maybe it does matter–that someone might choose not to use an invention that’s owned by another if it is not clear what the intentions are of that owner, just as one might choose to stay off someone’s property, simply because they have a claim to it.
That is, in the context of inventions, it may well be that use is most promoted when there is no ownership claim. Perhaps, however, where there is the prospect of improvements, use is also promoted by a commons–that contributions will come from a number of participants, creating the likelihood of access to even better implementations in the future. From such improvements–developments, call them–one might see ways to standardize (to coordinate uses, to optimize training and support), and one might see ways to sell products to folks who would like the convenience (and efficiency) of having someone else design and make and configure a version of the invention. For that, folks are happy to pay, knowing that they have alternatives. Something that’s owned as a monopoly, in this context, amounts to a trade secret, and to deploy an invention as a trade secret, one has to extract promises not to disclose and to use only as directed and to pay or whatever for the privilege. One can deploy inventions without patents–still involves contracts, or involves the threat of retaliation if the promises are not kept.
What the ownership issue really drives after, however, is ownership of a patent rather than ownership of the underlying invention. What is the point of a patent to an owner, but to exclude others? To the federal government, the point of a patent is to publish what might otherwise be a trade secret. That is, the target of the patent system is the inventor who would keep an invention secret (using contracts to enforce the secret, guild style) but for the right offered by a patent. Anyone who wouldn’t keep such secrets has no particular need of the patent system’s offer–universities, say, in the era before the affiliated patent management organizations came to dominate thinking about new money for research. So it may well be true–and so true as to be a truism–that a patent that’s not owned by someone is worthless.
A patent that is owned, however, may be used in various ways. The university administrator argument is that the only, best, first, primary, optimal use of a patent is to create a monopoly around the possible development of a product. That’s the discussion that’s up in the air, that we might have if we could look at university patent licensing, invention by invention, and see which patents, in which industries, at what points in time in the development of those industries, were licensed at all, were licensed exclusively, and in being licensed exclusively resulted in commercial sale or use, and if so, did that commercial sale or use result in public benefits on reasonable terms. Then we’d have a basis for discussion. Otherwise, it’s all a matter of guesses based on the bits and pieces of information that trickle out. And those bits and pieces are not so good, whether ascribed to the Act itself or to university patent and licensing policies.
Whether Bayh-Dole has resulted in problems, or whether university administrators filled with big whoopee to profit from patent licenses has resulted in problems–or both have resulted in problems–is an odd discussion. Clearly, if the universities’ patent practices are not competent, not insightful, not consistent with university and public values on the matter, it doesn’t much matter what the Act itself does.
The problem is with the university administrators, not with the law. Despite Bayh-Dole placing restrictions on university administrators–more than on small business contractors, and more still than on inventors acting on their own–university administrators have largely ignored the law. They’ve claimed the law gave them ownership of inventions when it did not. They’ve assigned inventions (and patents) without federal approval instead of licensing, at best, exclusively. They’ve used royalty income for unauthorized purposes. And they’ve refused to implement the core provision of the standard patent rights clause, that flows down the standard patent rights clause to inventors themselves and makes those inventors parties to the funding agreement. What possible changes in a law that’s not enforced could possibly change the behaviors of university patent administrators? That’s the fundamental question, and one that calls into question whether Bayh-Dole should exist at all. Universities had a so-so history of dealing with federally supported inventions under the IPA program before Bayh-Dole. Nothing much appears to have changed.
Pingback: Ten Years After 25 Years After Bayh-Dole, Part 3 | Research Enterprise