Ten years ago Sara Boettiger and Alan Bennett, a couple of University of California licensing officers, published an article on Bayh-Dole in Nature Biotechnology, “Bayh-Dole: if we knew then what we know now.” Boettiger and Bennett paint a picture of the Bayh-Dole Act and university patent licensing practice, and then argue that there needs to be exceptions to that practice to support research and humanitarian interests. Their article makes for a complicated discussion, not because their writing is complicated–or even their points–but because they base their arguments on a faux version of Bayh-Dole–the standard faux version of the law–and then argue against this version without quite coming out and holding the law responsible for the patent practices at universities that the authors object to.
Here’s the question Boettiger and Bennett ask:
If we were to write similar legislation today, what issues would be addressed differently, given our experience with the Bayh-Dole Act over the past quarter century?
The answer cries out–nearly every issue would be addressed differently, from the conduct of science and public welfare research to institutional involvement in patenting to federal agency discretion in the requirements on inventions made with federal support to the use of the patent system itself. But our authors accept that Bayh-Dole is somehow “successful”–their burden is to try to carve out from that success a safe haven for the crumbs of success, to offer to the research community and to the poor.
In some way, Boettiger and Bennett’s article is not about Bayh-Dole, but about the university response to Bayh-Dole. Let’s work through the article. There some things not to like, and some excellent points that have waited ten years to be requited and will wait a lot longer if nothing is done with Bayh-Dole or with university patent practice.
Shifting the Incentive Structure
Boettiger and Bennett start by describing what they call the basics of Bayh-Dole. This description no doubt reflects their understanding and that of many of their colleagues in university technology transfer. They assert that Bayh-Dole “shifted the incentive structure” for federally funded research. This is an abstraction, of course, and implies that there was an incentive structure to shift.
Prior to Bayh-Dole, some federal agencies allowed contractors to patent inventions made with federal funds; others did not. What mattered, according to the Kennedy patent policy (and its variations after), was whether the public interest would be served by a given disposition of inventions. Funding purposes varied. Contractor capability and interest varied. The Kennedy patent policy stressed flexibility within uniform defaults. Thus, if one couldn’t figure out something better, go with the defaults. The defaults were simple. If a contractor was in the business of doing something, and invented something more, then let the contractor own patent rights, but with some restrictions on term of exclusivity to encourage rapid implementation and to meet any future government needs. Otherwise, the government should own the inventions and dedicate them to the public or license them non-exclusively so they are generally available.
The Kennedy policy distinguished a number of situations in this second category. Some areas of government contracting are dominated by the government–atomic power, for instance. In these areas, the market is with the government. Contractors taking out patents in federally supported work simply mess with each other in competing for more government work. That in turn turns decisions about any given research contract into the politics of picking contracting favorites, because if Firm A wins the contract to do key basic research and development, and invents core technology, then Firms B and C are locked out of government bidding to use or develop that core technology for nearly two decades. The idea is–if a field is new (as in new scientific knowledge) or controlled by the government, then the government should create a commons before letting any particular organization attempt to dominate that field.
A second situation involved public welfare. If the federal government’s purpose in conducting research was to address a matter of public welfare, then the public should be free to practice any new thing discovered or invented. Agricultural extension programs are a good example. Extension biologists might, working to improve the soil on farms in their area, develop a new amendment. The expectation then would be: the farmers in the area would be free to use this new amendment. That would be the point of the development work. That would be “success.” If companies wanted to supply the amendment at a price or quantity or quality better than farmers could do for themselves, fine. That’s even more success.
But if the “incentive structure” is “shifted” it would appear that such use cannot happen. Instead, the extension program should patent the new amendment and unless a single company agrees to produce it, no-one should be permitted to use it. And the company that gets the exclusive license gets to charge monopoly pricing–whatever the market will bear, whatever the company can extract as a share of whatever “value” the new soil amendment offers to the farmer. So, price so there’s just enough adoption to give some farmers with the ability to pay an advantage over the others, and then those farmers can drive the others out of the market (or force the others into the market for the soil amendment at the monopoly price or risk going bankrupt). That’s a nice scenario, if money is one’s object. Why a university would participate in such a thing evades me at the moment, but perhaps you think it’s a good thing, survival of the fittest and all, or at the very least it’s ambiguous what’s best and so should be debated by academics for a few decades while Bayh-Dole does its thing.
Another situation involved government contracting to develop a product for public use. Not necessarily a commercial product, not necessarily a commodity product, but still one that would benefit the public. Perhaps today it is unheard of to develop a product for public use that isn’t itself a commercial product, sold in markets, for a profit. But that’s the digital computer and the internet for you–to name just two. In the early 60s, too, the idea that chemistry might lead to better medicines was in full play–just as Vannevar Bush had anticipated. So finding new compounds, screening them for bioactivity, and messing with those that looked promising made a lot of sense. It’s just that there are more chemical compounds possible than there are atoms in the universe, so finding just the right ones takes a bit of luck and a bit of science. The Kennedy patent policy did not cede all products that might be produced to private, monopoly-based commercial producers. A new vaccine, for instance, could be produced and sold competitively by multiple manufacturers–if in doing so they could produce quality product more efficiently than could the government itself, using contractors. (Medivision, a holding company prospecting for drug candidates, did much the same thing to produce Xtandi–there was no particular reason why UCLA could not have done so itself, or the federal government could not have done so).
Between commercial contractor and areas dominated by government use and without an independent commercial market lay an ambiguous area, where contractors lacked capability but still might want patent rights. For this area, the Kennedy patent policy let agencies decide based on what appeared to be in the public interest. If a contractor provided a good argument for a patent position–that it would then invest resources to develop a product and make that product available quickly and on reasonable terms, then that would be a good thing. But if the argument wasn’t so well put or certain, then don’t bother. The burden was on the contractor to make a public case for a monopoly position, given that the contractor could expect to have royalty-free access if the government acquired ownership of the invention made by the contractor’s personnel.
For over two decades, ending in 1978, federal agencies (NIH, NSF) had used an “Institutional Patent Agreement” program, off and on, to allow universities and nonprofits to own patent rights to inventions made with federal funding. The deal was: if you decide to seek a patent and agree to timely development of an invention for use with public benefits, then you can obtain assignment to the invention, provided you do what you have committed to doing. The IPA program basically said:
do better than the federal patent commons or put things into the commons.
The IPA program, then, aimed to do better than the federal government in achieving the same goals that the federal government had. If there was a shift in the “incentive structure,” it was in the institution that managed the patent rights. The institution that managed could adopt the same goals that the federal government had–make things available as quickly as possible, drawing on private investment to supplement government support–or could focus on institutional profit, giving lip service to social goals. Everything that was possible for the federal government to do was possible for the university patent owner.
The condition under an IPA for the university was to get things done quickly, and if via a monopoly grant, then that grant could be for only a limited time. If a company paid for the monopoly privilege, fine. There was no mandate that the IPA program was to make money for universities, though there was the claim that a university financial incentive would motivate universities to get involved in patent licensing matters.
That incentive could be made to look benign–if a university owned patent rights, then it could recover its costs of making inventions available for development, and the prospect of cost recovery would motivate administrators to do what they wouldn’t do on behalf of the public with their own money. That’s a backhanded way of reasoning about university administrators, but perhaps it was apt. But the incentive wasn’t limited to cost recoveries. The incentive included “upside”–the prospect of making a lot of money. Research Corporation and WARF had done so. Then so could others. Chasing big money is different from making new research available for private use and development. There are different people involved, with different mindsets, with different swagger. As one dean of engineering at a big private university told me (I paraphrase), “I don’t want my faculty talking to administrative no-names. I want them dealing with the A-list celebrities of venture capital.” “I get paid when I get things out” becomes “Things won’t get out until I get paid.”
One might argue that unregulated institutional financial “incentives” are the worse possible “incentives” for university involvement in technology transfer. In any event, Bayh-Dole did not “shift” an incentive structure so much as to “add to” an existing structure already well developed at universities. Bayh-Dole, in effect, restored the IPA program, made it apply arbitrarily across all federal government research programs involving universities (and nonprofits, and small businesses). Bayh-Dole added federally supported inventions to the university-side patent management programs–from the federal agencies that resisted the claim that private monopolies were uniformly a better approach to benefiting the public, and in particular, HEW and DOE. But Bayh-Dole also came with fewer restrictions and less federal oversight than the IPA program.
In many ways, Bayh-Dole is a law to prevent public oversight of patents obtained on research that was funded to benefit the public. Unlike the IPA program, Bayh-Dole requires an arbitrary patent rights clause to be used without regard for the purposes of the public funding, the qualifications of the contractor, the contractor’s patent policies, or the contractor’s performance record in managing inventions in the public interest. Bayh-Dole also was much worse than the private invention management system, since it appeared to be mandatory (or was made to appear mandatory), and thus served as a basis to force (fool) university administrators into changing their private invention management policies to match the one that was presented as required by Bayh-Dole.