Given the apparent intention of certain advocates for Bayh-Dole that the purpose of the law should be that universities come to own faculty inventions made with federal support, thereby effectively cutting off Research Corporation and other independent invention management agents from having direct access to faculty inventors, so as to give the advantage to the single-university-dedicated licensing offices, why did the Department of Commerce feel the need to insert the (f)(2) requirement in the Standard Patent Rights Clause?
Many university administrators and legal counsels rallied to amicus briefs fronted by organizations they populate, such as AAU, APLU, and AUTM, and argued that Bayh-Dole is a vesting statute, that it changes a fundamental of US patent law, in defiance of the US Constitution, to make title to inventions made with federal support vest with the university employer rather than with the faculty inventor. Short of the vesting argument, there are a number of variations: that Bayh-Dole requires inventors to assign, or prevents inventors from assigning to anyone else, or gives university employers “first right” or “right of first refusal” to such assignments. None of these arguments hold up on a reading of the law. It’s all really so much bombast, but it is presented as the suave pronouncements of authorities on the subject. Or, put another way, the arguments express a deep-seated projection that whatever Bayh-Dole was about, it should have been–and therefore really was–about university administrators gaining expedited ownership of faculty inventions, helped by a glorious new federal law to make administrative acquisition of inventions unilateral, expeditious, and certain.
The situation has been made all the worse because a number of “experts” among university administrators and legal advisors have asserted that Bayh-Dole vests/requires assignment of invention title to university employers. It’s now in university policies, in “educational” documents, in research contract templates, in popular articles, academic journals, most everywhere. Fortunately, the Supreme Court flatly rejected their assertion. Now they have no where to go. They certainly are not resigning in embarrassment. We cannot even expect them to say, “Gosh, we got that fundamental thing completely wrong! What were we thinking? Advocating for state / bureaucratic ownership of faculty scholarship in a country founded on personal freedom and limited government? We must have been high on something! We are very, very sorry and won’t make that mistake again!” That would have been a brave, credibility-saving thing to do. But it doesn’t appear that anyone is ready to make that sort of public confession. There is, apparently, no repenting in technology transfer. These folks appear to be determined–or desperate–to show that compulsory institutional ownership is “successful” even though that is not the general model that was in place prior to Bayh-Dole, though there were some instances of it at UC, Stanford, WARF, and MIT–some of the very same universities that led the vesting argument to the Supreme Court and which are now out in front of the present assignment craziness.
Compulsory institutional ownership carries advantages for those who dislike the idea of inventor choice, do not want a diverse network of invention management resources, and have no interest in seeing financial benefits arising from the use of inventions go anywhere other than back to administrative budgets at one’s own university. Compulsory institutional ownership:
1) Prevents independent competition from external invention management agents
2) Precludes such agents from showing successes that an institutional office cannot show
3) Limits the use of invention deployment models at odds with institutional choices
4) Ensures all licensing revenue is returned to the institution
5) Eliminates negotiation with faculty inventors over patenting, agent, licensing, or money
In short, holding a monopoly on invention management dissuades folks from questioning the monopoly, backed as it is by authority, claims of success, and a dearth of evidence that any other approach might be workable. What better, then, than to have the federal government requiring institutional ownership? That, in a nutshell, is what university administrative support for Bayh-Dole has been about. Administrators may be slow, but there is no question that they have found compulsory institutional ownership highly attractive. There have been no great or sustained outcries from administrators that the version of Bayh-Dole presented as a vesting statute is an abuse of federal powers, or is antithetical to an open research and instructional environment, runs against deeply held values of academic freedom relative to institutional (and state) control of scholarship, or that compulsory institutional ownership of creative work is strongly correlated with anti-innovation, whether in companies, governments, or universities. That is, no senior university administrators, to my knowledge, have spoken out publicly against the idea of compulsory institutional ownership of faculty inventions. Compulsory ownership was taken to be the law of the land, it was attractive, and it was a federal blessing to make the work of rainbow-chasing patent licenses for money that much easier.
Not only are the arguments for compulsory university ownership of faculty inventions not particularly substantive–the advocates of compulsory ownership were rebuked by the Supreme Court, their understanding Bayh-Dole is deeply flawed, and their views on innovation are not much better–but also they appear to have a totalitarian bent. One might say they are confused, as most of them I expect would claim a liberal orientation, but they happily suspend their views on diversity and freedom when it comes to seeking windfall profits from the corporate and capitalist worlds, in which a patent on exclusive offer represents a hopeful treat for wealthy consumption, in exchange for yet more money for research and research administration. But confused as it might be, the desire for compulsory ownership of faculty inventions is also well organized, rationalized, and repeated so often that it is held as truth, any internal contradictions and conflicting values notwithstanding.
If this discussion were about logic and reasoning that held up under scrutiny, then university administrators would never have accepted a vesting version of Bayh-Dole, would have complied with the Standard Patent Rights Clause from the outset, and would have preserved faculty choice of invention management agent. But they were snookered, seduced, and bullied into a position that over time they have come to adopt as their own. There is no easy way to back down from a professionally avowed truth, even when it is clearly not the truth. It’s like the great tragic heroes, running blind and mad at the end, finally persuaded of the horror of their situation, except that here there aren’t any heroes, and it is more a matter of bathos, of lost opportunities, of spoiling of a good thing by ineptitude and weakness made out to be a grand thing.
The Supreme Court decision in Stanford v Roche offered a way out: university folks could have gone, “we stand corrected”–but they didn’t. Instead, corny advice from legal commentators claimed that the decision turned on the failure of Stanford’s agreement to assign losing out to a later assignment. Stanford’s policy provided that inventors would own “if possible”; Stanford encouraged its employee to work for months at the company as a way to gain access to proprietary technology; Stanford knew of the assignment obligation and accepted it. Stanford had no claim on inventions made in that work, not under policy, not in equity, not legally, not under Bayh-Dole. The failure was Stanford’s later attempt to obtain assignment, despite all this, and long after Stanford had filed a patent application. The later Stanford assignment was void, if for no other reason than, as the CAFC ruled, the inventor had no rights to assign to Stanford by the time they got around to asking for an assignment.
Thus, vesting under Bayh-Dole was the straw to grasp at, and making it appear that Stanford’s invention policy all along intended for such vesting to happen, if not by federal law then by private agreement, only to be thwarted by the lack of pseudo-magical language, using “promise to assign” rather than “hereby assign.” Of course, this is all hack argument. Anyone being honest would fill it out: “promise to assign all inventions in which Stanford is required to take assignment” and “hereby assign these and only these inventions”. Neither of these forms of agreement would have done a lick for the facts in the Stanford v Roche situation. Stanford was not required to take assignment, had abandoned its future interest in such assignment, hadn’t even sought assignment when Stanford might have had an argument for it (such as before filing a patent application), and at any rate was precluded from the assignment by 35 USC 261, given that Stanford had notice of the prior assignment. The advice that as a result of Stanford v Roche universities must be more diligent in getting assignments upfront in the form of present assignments is total bunk. The strategy doesn’t work for the logic of the case, it is damaging to collaborative relationships, it creates huge uncertainties, it does not operate as claimed, it hammers academic freedom, it is inventor loathing, and it is antagonistic to innovation and economic benefits from discovery and invention made at universities. Other than all that, well, geez.
The University of California and the University of Washington, in particular, stand out for being the leaders of this bunk, imposing present assignment obligations on faculty while claiming that doing so is not a change in policy and is required by the outcome of Stanford v Roche. Administrators at these universities have doubled down to make sure that compulsory institutional ownership of inventions persists, even if federal law never required it and the only reason why faculty accepted it was that they were told straight-faced by senior administrators that federal law did require it.
Back, then, to the question of the (f)(2) requirement. If the purpose of Bayh-Dole was to secure compulsory institutional ownership of federally supported inventions, and the law was not going to do this by means of secret, automagical vesting, then what would have Congress needed to see in the law? Clearly, Bayh-Dole would have required universities to have a policy and paperwork that clearly established the obligation of inventors using federal funds to assign their inventions to the university. More particularly, the universities would have had to require assignment from their employees paid from federal funds managed by the university. Think about that for a moment: the employees are being paid with federal not university money, are working on a scope of work that they have proposed and is outside of their usual work assignments, and the university is additionally compensated for the cost of managing this work and providing the facilities for it, but as a condition of shifting some of their salary from university accounts to federal accounts, the federal government requires the university to require them to assign any inventions they make to the university. Where is the equity in that? What has the university provided that it has not been paid for? Nothing.
When the agency to become the Department of Commerce gets around to writing the implementing regulations and patent rights clauses for the Bayh-Dole Act, they confront this very problem. How does Bayh-Dole communicate its requirements? As law applied to federal agencies. How do those requirements get to universities? By patent rights clauses in funding agreements. What is the extent of agency power to require universities to take title to the personal property of inventors? It can be done, but it requires compliance with the Fifth Amendment’s restriction on taking of personal property for public benefit–in this case, for the benefit of the federal government, to better manage inventions made with its support.
The Intellectual Property Organization made an argument based on the Fifth Amendment in its amicus brief arguing that the Supreme Court should not hear the Stanford v Roche appeal. The IPO, after pointing out that vesting is not supported by the text of the law, and that vesting would have nasty effects on private ownership of inventions otherwise legitimately and fairly assigned, makes the case that if the federal government required university employers to take assignment, it would amount to a “taking” under the eminent domain clause of the Fifth Amendment:
A patent “confers upon the patentee an exclusive property in the patented inventino which cannot be appropriated or used by the government itself, without just compensatino, any more than it can appropriate or use without compensation land which has been patented to a private purchaser.” James v. Campbell, 104 U.S. 356, 358 (1881)
The IPO, citing case law, continues:
The Takings Clause may be implicated not only when the government takes property for its own use, but also when it mandates its transfer to a private entity.
Without due process and just compensation, the federal government is not permitted to take personal property for public use. And a patent in the US is personal property. The same goes for state governments, and hence, for public universities. If federal law requires universities to take ownership of inventions, then there’s a Constitutional standard to be met. That same standard is there if a public university makes substantially the same demand for title. And as the Supreme Court reaffirmed, employment alone is not sufficient to claim ownership of an employee’s invention. Thus, it cannot be a condition of employment, imposed later and by unilateral policy, that leads to a claim such as the University of California’s, that it all along intended that a promise to assign inventions in which the University established it had an interest meant that really inventors instantly assigned their inventions to the University whenever they made them, and the University would then assign back any inventions that it found it did not want, or that California law precluded it from owning.
Even if one wanted to pursue such forms of compulsory assignment, there would still have to be “just compensation.” For inventions, that surely cannot be merely a promise to share royalties with the inventor, even if there never are any royalties, or any effort to secure royalties, or any effort to secure royalties above expenses. Of course, an inventor could agree to such a deal, and that might be persuasive. But if the inventor is threatened (say, with being sacked, or disciplined) or if the deal is merely imposed, then there is no way that the compensation is “just.” In fact, there is no compensation for the taking, at the time of the taking! So how could it be “just” compensation other than for the voluntary concurrence of the inventor, at which point, it is a mutually negotiated, voluntary, private transaction, not a taking. Otherwise, the government, whether federal or state, owes inventors what their inventions are worth at the time of taking–and that would be, speaking justly, a reasonable share of the net present value of the invention as patented and practically applied. In a taking, that value is owed upfront. It is not a matter of “We taking over your house for public purposes, and we may pay you something in the future, whenever, if we get around to it, but otherwise we don’t have to pay you anything.” In a voluntary, private transaction, it may be negotiated lots of ways.
For the Department of Commerce, it must have been clear that the federal government was not going to vest title of inventions with universities merely for their role in hosting federally supported research. Further, there was nothing in Bayh-Dole that required universities to take ownership, or required inventors to assign ownership to the university, or to anyone else. There is not even a requirement in Bayh-Dole that inventors assign ownership to the federal government. The Act is entirely focused on creating uniform practices for federal government invention procurement in research contracts, and for limiting federal government claims on inventions made with federal support when an inventor did choose to assign such an invention to other than the government.
Thus, there was nothing compelling in either writing a patent rights clause that required university ownership of inventions or that required inventors to assign inventions to their employer. It could be done, of course, but it would be expensive, and would have involved more administration not less, and wasn’t at all necessary, given the stated objectives of the law and its language. Commerce had to ensure delivery of at least the right to practice inventions as part of the bargain. That’s the “procurement” part of contracting–that the commissioning party not get merely the performance of a duty (such as “to do research”) but also to obtain the benefit of the results (that is, “delivery” and “license” or “assignment”).
But there’s yet another problem for Commerce. The funding agreement is set up between the university and the federal agency providing the funds. If the federal government demands that the university deliver title to the invention, or a license to practice, then we are back to the compulsory assignment/taking problem, which Commerce chose to avoid. It is not sufficient, then, to say that the federal government pays the university for research and gets in return rights to inventions made with that funding. That is *not* the deal, because it will require eminent domain proceedings for the university to comply.
Thus, Commerce had to find a way to have a deal directly with the potential inventors, and it had to reach those potential inventors through the university. For that, Commerce was constrained, then, by what a university could be required to require of anyone with regard to inventions, within the bounds of employment agreements, invention agreements, state laws, and federal laws. Not to mention established practices, which at universities were all over the place with regard to inventions and patents. Furthermore, there was also the problem of small business firms, which were also included under Bayh-Dole, and small businesses may–or may not–have implemented requirements that employees assign inventions to the employer, and even if they have, those agreements might not be scoped to include federally funded research.
Whatever the university-side advocates of Bayh-Dole thought they were getting, when Commerce came to draft the Standard Patent Rights Clause, and got to the heart of how the government got its rights in inventions, regardless of who might also want title or rights, Commerce had to use agreements, not federal statutory power, and had to run the agreement through the employment relationship, without providing the employer with new leverage to take title of inventions. Anything in the SPRC that gave the employer leverage–a mandate, a requirement, a compliance step–could be construed as requiring the taking of personal property. What Commerce could require, however, is a voluntary decision on the part of anyone would who might invent within the federally funded project, to agree to deliver title or license rights to the government, as a condition of getting paid with federal funds. And why not throw in report the invention and assist in patenting?
Thus, the (f)(2) agreement. The (f)(2) agreement establishes a conditional invention deal with each potential inventor who is also an employee of the university. That deal is made with the concurrence of the employer, and indeed, the employer delegates performance of these core invention management activities to the potential inventors. The (f)(2) requirement acts as a release, relative to any employer’s claim on an employee’s inventions made as a condition of the federal funding.
In its way, (f)(2) is brilliant. It was no easy task to sort through all the conditions that might attend nonprofit, university, and small business dealings with research personnel. Even among universities, practices varied wildly, statements of principles varied wildly, and it was this wildness, apparently, that was integral to the general productivity of the university approach. In particular, the university approach appeared to succeed because it was (i) inventor-initiated; (ii) largely voluntary; (iii) highly selective; (iv) provided choices; and (v) was prepared to use exclusive licenses. All that the federal government needed, in all of this, was that the federal government was not denied the benefit of using what was invented for government purposes.
The core of federal research contracting with university faculty is a federal deal with the inventors, before they invent, not with their employers.
It is perfectly well formed that the federal government may cut a voluntary deal with potential inventors, that as a condition of receiving pay from federal funds held by the university to conduct the proposed research, these potential inventors agree to establish the government’s rights in their inventions. There are three possible outcomes. If the inventors assign their invention, then they require as well the obligations of the SPRC, and the government interest in their inventions moves to the assignee. If the inventors don’t assign to an approved agent, then they agree to assign to the federal government on request. If the federal government allows them to keep their invention, then they agree to grant to the federal government the nonexclusive license it desires. The university in all this is merely the steward of federal money, conveyer of contract requirements, and reporter of inventions. Only when an inventor chooses to assign to the university does it become an invention management agent, with a host of other requirements kicking in under the SPRC.
It is in this way that the SPRC navigates among the Taking Clause, employer leverage on inventive employees, and the diversity of practices between nonprofits and small businesses, and among universities, to reach a sound, appropriate, uniform, and manageable deal with potential inventors working with federal funds. As university administrators choose not to comply with the SPRC, and instead impose their own demands on employees, and even on non-employees, as they participate in federal funding, they may believe that what they are doing is rendering the (f)(2) requirement superfluous. What they are actually doing, however, is running headlong into the very problems that (f)(2) sought to avoid. For public universities, this includes the very real problem of just compensation for taking personal property.
There will be administrative handwaving about this, to be sure, but I am waiting for the first class action lawsuit against a state for failure of its public universities to pay “just” compensation for the taking of inventions under a claim–utterly and demonstrably false, especially after June 2011–that federal law or regulations require the state to take title or prevent inventors from assigning title to anyone but the state. I fear it will only be by litigation that university administrators will come to accept that their practice is either illegal or very expensive. Considering that the net present value of inventions taken by public universities over the past two years alone–we are talking thousands of inventions–could easily run to hundreds of millions of dollars, I would expect there is more than one law firm around that will take an interest. If university administrators cannot be persuaded to alter their policies and practices based on arguments drawn from diversity, freedom, and innovation, then perhaps they will pay attention when they see the significant financial liability into which they are leading their institutions. Do the thought experiment, my university administrator friends, and save yourselves the pain of defending what should not, and cannot be, defended.