I have been looking at the impact of the “export” of the faux Bayh-Dole Act from the US to other countries. By “faux” I mean the interpretation of Bayh-Dole that claims that the Act vested, mandated, and/or assured university ownership of federally supported inventions. That interpretation was set aside as erroneous by the US Supreme Court in Stanford v Roche over a year ago, but it still persists unchanged in many university documents and in scads of articles describing federal research policy. It does not matter whether someone is a proponent of Bayh-Dole or a critic–nearly all publications prior to the June 2011 decision have it wrong.
As I have worked through the problem, and in preparation for a talk in France next week, it seems to come down to whether people have a deep commitment to the idea that institutions should have the right to own the creative work of those associated with them, or whether the issue ought to be resolved by a reasoned discussion. If the matter is a reasoned discussion, then laying out reasons and facts ought to count for something. If the matter is just one of comparing urges and alliances, then there’s little hope for reason. One changes “sides” only by a conversion experience or as an act of betrayal or coercion, not through persuasion. It then becomes something rather farcical to attempt to reason with folks who have made a commitment to institutional ownership. What they present as “reasons” are more in the form of “rationalizations”–they aim to explain the position and quell dissent, but do not reason from history, from logic, from judgment. For the “science of science policy” folks, one would think making this sort of distinction would be fundamental.
Of course, this can all descend into claims and counterclaims that folks who prefer freedom over institutional compulsion when it comes to inventive work are just as deeply committed to their views and there’s nothing in the world but rationalizations, so it’s only a matter of which rationalizations happen to dominate the others. At that point, we are back to Thrasymachus’s argument in The Republic, and the question of whether there is a truth grasped by reason or whether any sort of bullshit rhetoric is fine so long as it wins out. Only the fool would choose a weak rhetoric, or so it goes. And it does appear to be the case that folks like to side with power, so it stands to reason that administrators within big institutional bureaucracies will tend toward taking and keeping power, delight in order and process and respect for authority, even if it is the authority of underlings with more power than their positions otherwise warrant when it comes to decisions about creative scholarship.
As a result of the faux interpretation of Bayh-Dole, universities across America have changed their policies to require assignment of inventions to the institution. At first this was represented, mostly, as a compliance step with Bayh-Dole. A university could not comply with Bayh-Dole (so the argument went) unless it had ownership. Therefore each university (so it was claimed) had to have agreements with its employees (and others) to take ownership of anything invented with federal funding. Later, this morphed into an assertion that this ownership was a condition of employment, and rationalized as the best thing possible for university-based invention, innovation, and economic development.
I am sure that some folks deeply believe this is true. Yet there are reasons that argue it does not hold up in fact. David Mowrey has a good talk on the matter, showing that universities were in IP practice before Bayh-Dole, and that the activity subsequent to Bayh-Dole is not entirely the result of Bayh-Dole, and furthermore, some of the activity is not all that keen, such as the friction between university administrators and companies over the disposition of inventions made with company support. (One refinement of a Mowrey point–Bayh-Dole not only reduced federal agency influence over university *licensing* policies, but also over university *ownership* policies.)
One could develop a variety of arguments, such as that creativity has rarely been fostered in compulsory bureaucratic systems. Another, that the metrics of commercial first use or sale do not support university contentions that Bayh-Dole is a roaring success. Another, that the university role in licensing prior to Bayh-Dole was largely voluntary and negotiated, and as a result of faux Bayh-Dole, universities have largely destroyed the very environment that provided the foundation for Bayh-Dole’s passage. Another that university licensing offices are generally poorly suited and resourced for patent licensing work. Yet another, that central planning, no matter how sincere, can never keep up with the opportunities and motivations of individuals at the cutting edge. One can’t very well plan for the innovation one desires, unless that innovation is incremental improvements and efficiencies that preserve if not entrench the power positions of the status quo. For all of the public rhetoric of “transformative innovation” it turns out university administrators have rigged up a system that has destroyed the diversity of opportunity and replaced it with a “more of the same, so long as we get some” approach. What is good for the bureaucrat is good for America.
We can point out another interesting feature of the discussion tied to the claims universities make to own faculty inventions. In US law, inventions are owned by their inventors. There is no “hired to invent” statute, no automatic vesting of inventions in employers. An employer must take an added step to secure an invention agreement with employees, and in some states there are restrictions on what an employer can claim of an employee in such agreements. For the agreement to hold up, it also must carry with it consideration. In a company, that consideration can be tied to the offer of employment, where the scope of interest in inventive work is within the scope of that employment.
Further, where the institution is an instrument of government, as with public research universities, the Bill of Rights in the US Constitution requires that private property taken for a public purpose must follow due process and the private owners must receive just compensation. When these public universities aim to take faculty inventions, they have to navigate these requirements. The challenge is showing that under the faux Bayh-Dole practices there has been both due process (including a true and accurate statement of the basis for the taking) or just compensation (many inventors–most–have never been paid anything).
Now, if one doesn’t care that inventors are not paid anything, and really likes the idea of institutional administrators taking what they want when they want it, then the effort is simply to throw up rhetoric to keep the discussion for moving into threatening territory. If we step around such gobbets, however, we encounter the question of just what is required of university inventors. In the pre-faux Bayh-Dole days, a university typically had a patent policy that set out a reporting process for inventions and a royalty sharing schedule if the inventors chose to assign their invention to the university or an affiliated research foundation for management. The policy statements assumed a voluntary choice. Clearly, if inventors choose the university for management of their invention, on the condition that the university will share royalties after costs but may never license the invention and may give up trying within a couple years, then that’s the deal.
Things have changed, however. Under faux Bayh-Dole, university inventors were required to assign inventions. For most there never has been “just compensation.” I have argued elsewhere that university faculty working on federally supported projects are not “employees” of the university in that work, in the sense that they are not assigned to do such work, but released to do it, and that they, not the university, decides what work is to be done, and how, and with what resources. Further, faculty are not doing the work for the benefit of the university. The work lies outside of the university’s scope of interest–except that sponsored projects offices feed off of the indirect cost charges and therefore have come to view sponsored research as a university directed activity, and technology licensing offices have come to feed off of licensing royalties and therefore see all research activity as necessarily to create inventions to be patented and licensed to industry for institutional profit. From their perspective, the business of the university is to make money on whatever inventions happen to be made in faculty research. It is therefore only right that everything is assigned to the university for their profit-seeking.
It turns out that in taking everything, a university sets up a portfolio model to income. Rather than serving each lab with support for its inventions, so they are properly placed, a technology licensing office retreats to a sales model in which hundreds to thousands of inventions are taken so that only a handful produce income and fewer still commercially available new products. In fact, just one lucrative patent license per decade is sufficient to call a program a “success.” The rest can rot. While this may be a nifty way to financial survival, it is a devastating approach for the vast majority of university research findings. Each is subject to a claim of institutional ownership. Most won’t be licensed. Most won’t become commercial products. And therefore most cannot even be used outside commercial development endeavors. Even if a university shared its one lucrative deal with *all of its inventors* that would still not compensate for the damage done to the laboratory relationships that are suppressed by holding ownership of the unlicensed inventions. The basic university position is: if speculators do not take these inventions, then no one can have them. Protecting the operating model of the licensing office becomes more important than the relationships that a given lab and its inventors might make with other organizations. Preventing “a race to the bottom” is the administrative code for this way of thinking.
In France, as in a number of European countries, however, there are three categories of invention. A first category is “inventions under mission”–in which the employer owns the invention but is obligated to provide the employee with “additional remuneration” for the inventive work. A second category is “assignable inventions beyond mission”–these inventions are assignable to the employer upon demand (or the employer may assert various forms of partial interest in the invention), but the employer must pay the employee a “fair price” for the assignment or other interest. The third category is “non-assignable inventions beyond mission.” These inventions are the employee’s and none of the employer’s durn business.
The question has arisen just what is owed by an employer in taking an employee’s invention in either of the first two categories. Just what is the remuneration or “fair price” that is required? An article by Thomas Bouvet in discussing a relatively recent court decision in France puts the issue this way (emphasis added):
The provisions regarding employee inventions apply to patentable inventions, whether actually patented or not. In other words the employer is obliged to pay additional remuneration (or a fair price) as soon as the employee creates an invention. The employer is under no obligation to file a patent; it can decide to keep the invention secret. Similarly, the employer is under no obligation to exploit the patent. However, when the remuneration of the employee, or the fair price, is dependant [sic] upon the exploitation of the patent, the employer may be obliged to exploit the patent or to pay the employee as if it had exploited the patent.
If the “just compensation” for US faculty inventors depends on a share of licensing revenues, then it would appear that US universities have an obligation to exploit the patent to obtain those revenues. If I take your house on the condition that I resell it and share proceeds with you, then I have to resell it. If I take the house and don’t resell it, then I owe the market value of the house regardless. As Cornell’s discussion has it:
The U.S. Supreme Court has defined fair market value as the most probable price that a willing but unpressured buyer, fully knowledgeable of both the property’s good and bad attributes, would pay.
For a patent that has a 20 year run, that’s a net present value calculation. And the sum involved can be much higher than what a university administrator might settle for, since the emphasis is on what a knowledgeable buyer would pay, not what a seller might take. Whether the argument is built around a law in France or one in the US, the principle is common: the exchange by which the employer comes to own an employee’s invention requires compensation above that paid for employment, especially where the invention falls outside of the assigned and directed scope of employment. Even if the requirement to assign is made a condition of employment, for faculty working on federally supported projects, the consideration for that assignment is *not* what has been paid as compensation for employment.
It would appear that where universities have taken ownership of inventions under a faux Bayh-Dole claim, they owe the inventors the value of the invention *as if they had exploited it*–that is, as if it had been licensed. Put another way, they owe the inventors a share of the net present value of the invention at the time it was assigned–that’s not the value at the time of assignment, but the value over time based on taking a patent position and licensing that position for a reasonable royalty. Whether or not that royalty actually comes about is a separate issue. Obtaining that royalty is a risk borne by the institution making the claim of ownership, not by the inventor who is often required in public universities not to participate in contracting involving the state. If this line of reasoning is followed out, one could imagine that universities have been exposed to a substantial liability by impulsive, rationalizing patent licensing officers.
If such assignments of inventions to universities had been strictly voluntary, without the faux Bayh-Dole claims and the back-filling assertions about conditions of employment, often made unilaterally and well after the fact of the establishment of the employment relationship, the universities would not have this exposure. They would have whatever obligations were mutually agreed to in the negotiations on which any given assignment was made.
Of course, it is worse than this under the standard patent rights clause authorized by Bayh-Dole. Under that clause, whatever agent obtains and retains ownership of a subject invention is required to make efforts to promote practical application–typically for universities by licensing it to industry. Failure to do that is a failure to comply with the federal funding agreement. It appears, however, that in the thirty odd years of Bayh-Dole, no agency has ever enforced Bayh-Dole diligence clauses. One can argue this is because universities have done such a great job of “commercialization”, but that’s not going to hold up under scrutiny. They haven’t done the job. It is only a matter of time until university inventors recognize, after Stanford v Roche, that they have been cheated and demand the just compensation they are due. If reasoning doesn’t alter the position of administrators happy with compulsory claims to ownership of faculty inventions made with federal funding, then perhaps fear of financial liability eventually will.