I continue to be amazed at the persistence of the faux Bayh-Dole crowd. Like something out of The Road Warrior, they keep coming back to wreak havoc. Despite the text of the law, the Supreme Court ruling in Stanford v Roche, the ubiquitous university practice of obtaining written assignments from inventors, and the remarkable questions about how helpful state control of faculty scholarship has actually been for American innovation, the faux version of Bayh-Dole keeps coming back for more. Certainly there are huge amounts of money at stake in this discussion. I guess there is no reason to expect fair play, reason, or a broader perspective, let alone an admission from the faux folks that they got things a bit wrong and are sorry for that.
Two recent papers (“Best Practices for Drafting University Technology Assignment Agreements After FilmTec, Stanford v. Roche, and Patent Reform” and “Renewing the Bayh-Dole Act as a Default Rule in the Wake of Stanford v. Roche“) by Parker Miles Tresemer appear to be the latest effort to revive the faux Bayh-Dole with its vision of institutionalizing control of faculty scholarship when in the form of an invention as a really keen thing. I expect that what Tresemer has written will become the script for the next round of the effort to strip faculty of their rights and reduce them to corporate employees. If you are a faculty member who cares about academic freedom, has a deep distrust of state control of scholarship (if you work in a public university), and believe that innovation starts with individual initiative and is largely suppressed by compulsory institutional accumulation, then take notice. Now would be a good time to get at it and secure your freedom, before it is taken from you for good.
Tresemer has assembled substantial heft for his argument. However, on inspection, one finds much of the argument and evidence don’t hold up. As I get time, I will point out the problems, but it is like picking fleas from a dog. There are lot, and it may work better just to douse the pooch in a medicated oatmeal bath, which is what I propose to do here. In parts of the record that I know well–university licensing practice, the history of university involvement in technology licensing, the particulars of the Stanford v Roche case, the effect of institutional ownership on innovation–it appears that Tresemer suppresses, ignores, or is ignorant of, a great deal. Not that anything I could point out would likely get him to change his position. His is a case made as a lawyer would by selecting the bits and pieces that make for a coherent story, and not offering to explain the problem parts. It would help his case (if one wanted administrators to dominate faculty research) for him to get his facts right. I can’t help him with the reasoning, though. A man hears what he wants to hear.
Tresemer’s argument, however defective, is the most substantial effort so far to argue that Bayh-Dole was a vesting statute, that the Supreme Court ruined it, and now universities *have* to impose present assignments to get what Bayh-Dole intended for them to have all along. His argument is a scholarlied-up version of the standard myth–that Bayh-Dole was a magical law that sparked American innovation, vested title with universities so there wasn’t any “title uncertainty”, and that institutional control of faculty scholarship by federal fiat is just the keenest thing that anyone can imagine for innovation and economic vitality.
The subtext for Tresemer is that to get things put right, Bayh-Dole should be amended to make clear its intention to strip independent faculty investigators of their title to inventions, so that university administrators can sequester inventions for the benefit of speculators. Somehow, this combination of institutional and state control, combined with good buddy deals with speculators, is the very thing we need for innovation and “American competitiveness.” Well, that faux magic hasn’t happened much in thirty years. The current discussions are about open innovation and networks, non-institutional science, why there is such stagnation in formalized university research, badly conducted science, and, and why university tech transfer is in such a wretched condition, despised by many companies and investors alike. Isn’t it odd that in thirty years, universities have never reported publicly their Bayh-Dole metrics of invention utilization? Never. Doesn’t exist. The faux version of Bayh-Dole can’t tolerate it.
Tresemer makes a confident argument about a rhetorical vision of patent practice that flourished thirty years ago and has been shown to be ineffectual. He fails to grasp the core of Stanford v Roche, and continues the line of reasoning that somehow the case teaches the need for present assignments. The desire to retain privilege encourages white knuckles, and there is no indication that the idea of using the power of the state to enfranchise bureaucrats and their buddies is going to go away. So let’s have it.
Tresemer’s work has four thrusts.
First, that the intentions of those pushing for Bayh-Dole must be considered ahead of the actual text and operation of the law.
Second, that Bayh-Dole in its perfect form vested title in inventions with universities (and others) despite federal patent law, patent office practice, and actual university practice.
Third, that because of the Stanford v Roche decision universities have to impose present assignments to be assured of getting title to federally supported inventions.
Fourth, that Bayh-Dole should be revised to make vesting of invention title in federal contractors perfectly clear.
In working through all this Trasemer relies on the faux history of university licensing and federal management of patents. In that he cites Stevens, an AUTM insider, rather than Mowery, who has done the actual historical legwork. Yes, the faux history shows up as political rhetoric in the legislative history of Bayh-Dole and in a bunch of AUTM documents, but the repeated assertions of the politics don’t make it so, except in an Orwellian sort of world. Mowery, in particular, has documented the rise of university-based patenting in the 1970s and has shown how Bayh-Dole was not the spark, but rather a political symptom, of such activity. See here and here and here: “the ‘transformations’ wrought by the 1980 Act followed trends that were well established by the late 1970s.”
Were universities not involved in seeking patent management agreements with federal agencies, there would not have been a pressing need to make uniform the federal agency invention management practice. Substituting a political history for any other narrative of history is convenient, but not persuasive–unless perhaps no one else knows any history. A similar thing happens when we turn to the effect of Bayh-Dole, where we find “success stories” but little to show that those “successes” came about as a result of Bayh-Dole. Indeed, universities don’t report their Bayh-Dole activities. There is a reason for that. The faux history was constructed to get at a political privilege. It will be repeated as fact so long as there is any hope in keeping the privilege.
I will deal with each in turn.
1) The legislative debates around Bayh-Dole don’t matter.
Yes, it’s clear some folks wanted Bayh-Dole to be a radical departure from US patent law. They wanted not merely “invent for hire” but rather federal law to push title from inventor to employer even when the employer wasn’t doing any employing–that is, when the federal government was paying for work proposed by faculty, and also paying the universities for releasing faculty to do that work. Tresemer is spot on about the legislative record on this point, as far as he takes it. It’s just that what those folks wanted is not in the law. It is not in the objectives of the law, not in the implementing regulations, not in the Patent Office regulations and guidance, not in university invention assignment practice. No one goes to the PTO and says, I don’t need to have a written assignment for this invention because I’m the employer and federal law vested title with me. That never happened, even before before Stanford v Roche. It is an artificial account of practice, constructed for political purposes, without regard to the evidence.
We have written laws and contracts not to stand as icons for whatever someone in power wants, but to set forth what is expected in plain writing, as a record to all. It is the writing, not an intention for benefit by some one group, that takes precedence. The legislative history of debates on Bayh-Dole is not relevant to the law except where the law cannot be reasonably read. The Supreme Court was able to read the law, and reasonably. So should we all, law professors especially.
Let’s emphasize this point. Bayh-Dole authorizes the Department of Commerce to create standard patent rights clauses for federal funding agreements used by federal agencies. The agencies are permitted to modify those clauses under certain circumstances. The actual requirements for any particular federal funding agreement therefore are not found in Bayh-Dole, but in each individual federal funding agreement, following a pathway such as 37 CFR 401.14(a) to 2 CFR 215.36(b) [now, due to changes, 37 CFR 401.14 and 2 CFR 200.315(c)] to the actual language of the patent clause inserted in a particular federal funding agreement with, say, a university, and on down to the subcontracts, the assignments, the substitutions of parties, and the written agreements that are to be required of universities of their employees, to protect the government’s interests. The disposition of invention rights is *contractual* not *statutory* with regard to universities and with regard to research personnel employed by universities. It is a matter of agreements. And any account of patent rights has to follow the agreements. That’s IP contracting 101. This is fundamental stuff. And here we have a law article that proposes to disregard all this and says all one has to do is to read the law colored by the intentions of advocates for a particular outcome and then one has the renewed the belief in the faux version of Bayh-Dole.
All those records of intentions, even of the sponsors of Bayh-Dole, just don’t matter. It starts with the language of the law as we have it, and have it now. If the words failed to convey the intention of some pressure group, then that’s a problem in the drafting of the bill for that pressure group. The court in Shaw v. The Regents of the University of California made it clear that the university can’t demand that its intention dictate how its invention policy (for which it claimed the force of law) must be read. It goes the other way. Start with the text, and ask what the reasonable effect of the text must be:
To resolve the parties’ dispute concerning the meaning and effect of the patent agreement, we resort to the traditional rules governing interpretation of contracts, which “teach us that the overriding goal of interpretation is to give effect to the parties’ mutual intentions as of the time of contracting…. Where contract language is clear and explicit and does not lead to absurd results, we ascertain intent from the written terms and go no further.”
The true intent of a contracting party is irrelevant if it remains unexpressed.
Tresemer appears dismayed by such a thing. He writes: “The Court grounded its decision not in the Act’s legislative history, but instead in its language.” It is as if the Court is supposed to first decide what a pressure group wanted, and then make the law do that regardless of what it says. To be sure, that is one way of making sure the courts serve special interests, but it raises odd thoughts about justice and the rule of law. Everything I’ve seen points to a kind of fundamentalist belief among the faux Bayh-Dolers that comprehensive, compulsory institutional ownership of faculty scholarship, to be sold off to speculators for development, really is a bright idea that will lead to a glorious world, wealth, health, and jobs, just like Vannever Bush imagined. We are not debating the outcomes, or even the claimed intentions, but the methods and their consequences. But while we are at it, we can notice that the outcomes for this approach are so bad that universities simply don’t report them. Rather than being so sure that compulsory institutional ownership of inventions is so very good, folks would do well to entertain some doubt about it, and encourage a freedom to explore possibilities rather than require service to a doctrinal bureauklepticism about innovation that does little good at all, at some expense and damage to scholarship.
Tresemer’s discussion of the legislative history does show, however, how convoluted the drafting of Bayh-Dole had become by the time it reached final form, and especially the substitution of “contractor” for “subject inventor.” I hope to comment on that later.
2) Bayh-Dole never vested ownership of inventions with universities.
Bayh-Dole makes no express change to the invention ownership provisions of federal patent law. The PTO never created regulations to accommodate such vesting. There is nothing in the MPEP about it. Universities never relied on such vesting, but have always sought written assignments in practice. Even Stanford sought assignments in the Roche case. Why would Stanford do that?
No, folks have said one thing, but practiced another. The university line has been, “you have to assign as a condition of federal law” not “federal law transfers your ownership rights as an inventor to the university.” Further, when pushed, the universities cannot point to just where it is that federal law requires the assignment. It doesn’t. They then fall back on the argument that “elect to retain title” must mean, by implication, that the universities already magically have title. The Supreme Court corrected this argument to be: universities have title when they get it by assignment, and once they properly have got title, then they may elect to retain it.
If these folks would look at how the law is implemented, by means of contracts, they would find that the (f)(2) agreement is critical, and (f)(2) is not in Bayh-Dole. It is in the standard patent rights clause, that becomes part of each federal agreement. That (f)(2) agreement is to be required not by the patent rights clause, but by university. That’s what the (f)(2) clause says. Why not read it?
(2) The contractor agrees to require, by written agreement, its employees, other than clerical and nontechnical employees, to disclose promptly in writing to personnel identified as responsible for the administration of patent matters and in a format suggested by the contractor each subject invention made under contract in order that the contractor can comply with the disclosure provisions of paragraph (c), above, and to execute all papers necessary to file patent applications on subject inventions and to establish the government’s rights in the subject inventions….
The patent rights clause requires the contractor to take an action. It is the contractor that then requires the written agreement. That written agreement is not an assignment agreement placing rights with the contractor. The (f)(2) written agreement includes a commitment to “establish the government’s rights”. The authorization to make this written agreement comes from the contractor, not the law, not the funding agreement, not the patent rights clause. This is just the kind of agreement that anyone seeking to file a patent application would want with any inventor. Here, the effect of (f)(2) is to get the contractor out of the way. Clearly, if the contractor fails to obtain title for a given invention, it is (f)(2) that the government will rely upon to obtain title or a license to an invention. For the government, the greatest title uncertainty comes if the government does not have this direct relationship, as authorized by each contractor, with potential inventors. The (f)(2) arrangement has no need of the “establish the government’s rights” provision if inventors have no rights to convey to the government.
For all that, there is absolutely no reason for Bayh-Dole to require the Department of Commerce to create standard patent rights clauses, or for these to be incorporated into funding agreements, if as “critics of the Court’s decision” claim, Bayh-Dole was to perform a statutory excision of inventor title rather than shape the conditions for a contractual transfer of title without the need for time-consuming agency review of each case. Why have all the show of contracting, and clauses, and exceptional circumstances, and agreements, if the statute supersedes all this apparatus and vests title anyway? Where is the “title certainty” if title flits magically from inventor to contractor, and then if the contract messes up or doesn’t want it, back to the inventors, or to the funding agency, or to an organization that the contractor has a deal with, all without the need for written documents of assignment, based on things like a notice to “retain title” or a failure to provide such a notice. Where is the “certainty of title” in all that mess? Just what invention is subject to such claims?
The intention to own via vesting has haunted Bayh-Dole throughout its history. It haunts especially university tech transfer office restatements of Bayh-Dole. However, even the most official documents never come out and say, Bayh-Dole vests title. They are very careful to use “retain title” and never once mention how radical a departure this idea about invention ownership is for any American inventors, let alone university faculty inventors. What did university faculty do to deserve such abuse at the hands of patent adminstrators and legislators? There never was vesting. Vesting is just a strong-arm tactic for university administrators to take by a claim of force what previously they had accepted with goodwill and mutual agreement.
3) Stanford v Roche does not require present assignments.
In the Stanford v Roche case, Stanford used an agreement under which employees promised to assign to Stanford inventions in which Stanford had an interest. This was by no means all inventions that they might make. Indeed, Stanford’s policy said that inventors would own their inventions “whenever possible.” That is, Stanford would only have an interest in employee inventions when there was some requirement incumbent on Stanford or the employee that prevented employee ownership of the invention and required Stanford to rely on the promise to assign.
The Stanford policy is not remarkable in this regard but for the “whenever possible.” The University of Washington and University of California policies have the same structure–that employees will assign to the university what the university has an interest in. All the attention has been on making stronger the demand for assignment–as a condition of employment, as a condition of use of resources, as a condition of receiving federal support for research, now by present assignment rather than promise to assign–and no attention is placed on what the policy sets forth for what the university actually may have an interest in.
At one time, that “interest” was, apparently, what the university was required to have an interest in by contract, whether with inventors directly, as in the case of commissioned work or work specially supported with university funds, or with sponsors of research that designated the university as the beneficiary of invention ownership as a condition of their funding. Stanford’s policy made clear that the deal was: you assign what we have to require you to assign, and you keep everything else. Later, universities got the idea that if they inserted a requirement that they must own inventions into contracts with sponsors, and the sponsors agreed to that university demand, then the university was “required” by the contract to own inventions, and could turn around and demand ownership from inventors.
You can see how this would work wonderfully well with Bayh-Dole when it came along. Bayh-Dole requires university ownership, so despite any qualms we might have about that, there’s nothing we can do except require inventors to assign. The assignment then is made out to be necessary for the university to comply with federal law (or, more accurately, with its funding agreements with the government, which in turn are established by federal law). That’s the actual shape of the vesting argument as it was established in university practice. Tresemer doesn’t bother to point this out. Probably he doesn’t know it, and doesn’t much care. It is, after all, a history that the faux political history of Bayh-Dole doesn’t want told.
Consider then what happened in Stanford v Roche. Stanford has a faculty member serving on the advisory board of a company with hot new technology, polymer chain reaction (PCR). That faculty member sends over a post-doc in his lab to work at the company for nine months, to learn PCR and work with company scientists. Stanford signs a material transfer agreement with the company, agreeing that the company has a royalty-free license to anything Stanford might invent with the material. Stanford allows the post-doc to sign an agreement with the company that any inventions made as a result of his work at the company will be assigned to the company. Stanford is fine with all this. It is entirely well formed under Stanford policy. The inventor will own whenever possible, and working outside of Stanford and its resources is perfectly fine.
Now the post-doc returns to Stanford and lo, “invents.” The invention is in just the area he had been working on with the company. It is an invention clearly within the scope of his obligation to assign to the company. It is just this sort of situation that any company is most concerned about, and why such assignments are required. It would make no difference whether the company had a present assignment or a promise to assign–the deal with Stanford was that the post doc would assign to the company. It happens the company included both a promise to assign and a present assignment in its contract with the post-doc. If it hadn’t had the present assignment, then no doubt it would have been diligent in getting an assignment document executed. Meanwhile Stanford goes off, files patent applications, then much later does its federal reporting stuff out of order and apparently out of compliance, and then goes to get assignments from its inventors. Years more later, when the patents have issued, and the company has product on the market (Roche, now, having acquired PCR assets from Cetus), Stanford shakes down the company for a license, and when the company cites the assignment by the post-doc, Stanford sues for a quarter billion dollars or so. This is any company’s horror story–play nice with a university and have them turn on you and try to break the deal you have made.
The CAFC and Supreme Court analyses of the situation do not have worry all of this. The CAFC followed the valid assignments. The company had one first, and Stanford didn’t. Stanford’s assignment document comes after patent applications have been filed, after election to retain title (why does Stanford even request an assignment document? if Stanford has retained title, then what was conveyed in the assignment instrument?). The basis for the company’s assignment is the present assignment language. Everyone fixates on that. But it is simply the way that the assignment was conveyed. If Stanford had had a present assignment instead of promise to assign in the original agreement, it would not have mattered in the least: the invention was out of scope of Stanford’s claim. The courts did not have to discuss this because it is a hypothetical. They could get to a decision without it. What the CAFC found is that the company had an assignment, it was valid, and Stanford didn’t, and when it went to get assignment, there was nothing for the post-doc to assign, since he no longer held his title.
Stanford’s own promise to assign language did not operate, even though Stanford tried to make it operate by requiring the inventors to assign to Stanford. The assignment language fails not because it lacked a present assignment, but because the invention was outside the scope of what the policy identified as Stanford’s interest. This is where a whole lotta lawyers seem intent on getting it wrong. Maybe it is a kind of ghost story useful for scaring up new business from administrators who have the money to pay. Dunno.
Look at it another way. Even if Stanford had had a present assignment, and even if the invention had been in scope of the claim, Stanford had released its post-doc to make a commitment that might conflict with a future Stanford claim. Stanford accepts this. They don’t say, “sure, our post-doc will assign all that to you, unless he comes back to Stanford and we direct him to just this same work and then he invents, and then it will be ours, not yours, bruhaha.” The company relies on Stanford’s knowledge of the commitment, which functions as a restriction in the scope of Stanford’s future possible claims. Even under a 35 USC 261 analysis of competing assignments, it is the company’s assignment that wins out. It is the prior assignment, because at the time that it was made, Stanford clearly had no claim whatsoever in any inventions that the post-doc would make. They were not within the scope of employment at Stanford, and Stanford was under no obligation to require assignment. Further, Stanford has knowledge of this prior assignment to the company. Stanford could not succeed with a claim unless it (1) obscures its own assignment conditions and (2) finds some way to make the assignment required.
Thus, the last hope was to rebuild Bayh-Dole as a vesting statute. This is where the old advocates come in. They argue it was always their intent that Bayh-Dole would vest title with universities, in a radical departure from US patent law. While there is plenty of evidence, as Trasamer reports it, that this is what such advocates may have wanted, it is an argument of convenience. It is odd, because even Allen and Bremer wrote this after the Stanford v Roche decision: “So Bayh-Dole stands, based on its traditional premise that agreements are in place with researchers assigning intellectual property rights to their universities for inventions made in whole or in part with federal money.” Of course, even that is only correct insofar as “traditional premise” means the “faux premise” of Bayh-Dole, since that premise is not stated in the Act, and that means anyone can state a premise. The actual Bayh-Dole Act does not care one bit whether an inventor using federal funds assigns to his or her university. The funding agreements established by the agencies, however, do care, and that is the point of procurement of inventions as distinct from the process by which such procurement is administrated. The agencies decide how to shape the patent rights clauses, and the standard one at 37 CFR 401.14(a) provides for a delivery clause in the form of the (f)(2) agreement. It is that (f)(2) agreement that the agencies care about. Once that is in place, inventors can assign or not and the government still has its rights. Of course, I don’t know of a single university that has decided to comply with (f)(2). Instead, all the emphasis is on getting ownership, rather than complying with funding agreements or protecting government rights. Isn’t that odd? I sure wish ONR or DOE would do a few audits!
The faux Bayh-Dole is not so much a hoax as a clever scheme that takes up old expended political arguments, rewrites history and practice, and sees whether it can make Stanford another quarter billion dollars while inducing the Supreme Court to turn Bayh-Dole into that grail of a vesting statute. If Bayh-Dole vests title, or forces title to be assigned to the university, then Stanford’s promise to assign must somehow take precedence, the assignment to the company when the invention is made is voided, and title is restored to Stanford, so it can continue its litigation for infringement. The whole thing fails not because Stanford didn’t have clever language of a present assignment in its original agreement, but because Bayh-Dole is no vesting statute and therefore the invention was never within scope of the Stanford policy–and contractual–claim.
We are dealing with folks who lost in a grand scheme to claim the personal property rights of tens of thousands of faculty inventors, for their financial advantage, justified by a fluffy but undocumented claim of economic vitality. They equate autocracy with “title certainty.” Autocracy is not the horse I want to hitch the innovation wagon to. But these folks are never going to admit they have made up just enough of it to make the story sound plausible, especially to folks who haven’t ever been involved in university IP policy and practice. I do not expect they will ever have their “come to Oprah” moment. Instead, they will bully, obfuscate, and distort to keep up a game face on it all. Is there an exit for them by which they save face? Can university tech licensing folk ever be able to speak openly about their practice? Or will they all hide behind the shield of their faux history and metrics, figuring that no one has the energy to get at them, finally, and show that it just isn’t working the way they claim?
Tresemer raises a number of interesting points regarding assignment of inventions. Again, it is worth having the discussion on these matters. But he ignores the status of faculty at a university with regard to their “employment,” disregards matters such as eminent domain taking of private property–especially in light of the Stanford v Roche decision–when the university is an instrument of the state. More broadly, it is clear that Tresemer cannot imagine how patents could be used for anything other than making money from risk-adverse speculators willing to pay for the chance at big wealth. It is as if there is no role for patents in quality assurance, setting standards, providing access to shared technology through cross-licensing, or protecting lines of future research–all of which involve forms of non-exclusive licensing, or a separation of claims into essential and non-essential, where the essential claims are the ones that are available non-exclusively. None of this exists in the mythic account of Bayh-Dole and the use of the patent system to promote the use of federally supported inventions. Yes, there can be and are exclusive licenses. But industry didn’t need exclusive licenses for Cohen-Boyer. Or for Axel. Or for Hall, or a whole lot of the foundational patents that launched the biotech industry. It’s a myth with just enough truth in it to sound good. The purpose of the myth is to aggrandize the selling off of faculty scholarship to speculators as a public good. Out of context, it sounds as good as it can get. In context, it looks like a shill for bureaucrats more interested in money than scholarship or innovation.
The courts saw that the assignment to the company controlled. They did not see that Bayh-Dole vested title so as to void that assignment. They did not see that Stanford’s own policy requirement superseded the assignment to the company. They cited the present assignment to the company, because that was the form of the assignment. They did not have to construct hypotheticals and write a self-help book for attorneys in all the variations on the case. Present assignments have nothing to do with Bayh-Dole compliance, and would not have resolved Stanford v Roche in favor of Stanford.
4) Bayh-Dole needs to be revised, but not to make it a vesting statute.
Bayh-Dole as it is written creates a pattern of contracting with regard to agency rights to future inventions made by university personnel working with federal support. The deal is that if the inventors assign to an approved invention management agent, then the agency need not review the details of the assignment, so long as the assignee complies with the conditions of the patent rights clause in the funding agreement. That invention management agent can be any organization that has a primary function the management of inventions, or the inventor’s employer, even if that employer is clueless and incapable of managing inventions on its own. If the inventors want to assign to anyone else, then the agency has to approve, and we are back to the old, pre-Bayh-Dole regime of agency review and the setting of conditions. If the inventors don’t assign, then the deal is between the agency and the inventors, mediated by the (f)(2) agreement and within the guidance of 37 CFR 401.9.
In short, Bayh-Dole is a law of pre-approvals for assignments, not a vesting statute that removes the need for assignments. It is up to inventors and employers to determine what the nature of those assignments might be. For a small company receiving federal funds, it may be that the employees doing the inventing are working outside the scope of their employment, or inventing things that the company has no interest in. For that, the company need not have any sort of expectation of assignment, and need not make assignment a condition of employment. All the more so for universities with regard to faculty. What an absurdity that university administrators–or lawyers, even law professors–can imagine that “really” faculty are “just employees” and all the stuff about academic freedom, tenure, shared governance, and the importance of an independent faculty are just “fictions” and the administration really does assign, direct, review, and approve faculty scholarship. It’s nonsense, of course, but it’s dangerous nonsense.
The aim of the faux Bayh-Dole crowd is to use federal powers to strip faculty of inventions. The motivation is to build up their “technology transfer” programs–which is the point of AUTM licensing survey metrics. Those metrics track the growth of technology transfer activity, and the “success”of Bay-Dole is reported in terms of the “success”of technology transfer as a bureaucratic industry of policies, procedures, accumulation, and money. It’s not that technology transfer itself is bad, but that it willingly accepts this rhetorical cover for its operations. No one reports their actual metrics. If success is money for universities to be used in research, then other than a few big hits, tech transfer has been a huge bust. If success is in innovation–new products on the market–then again, there is evidence that tens of thousands of research assets have been sequestered to create a very few results. If success is in jobs, then universities have a dismal track record, and it is laughable that they propose to take their licensing income, multiply by 50, and divide by whatever they take to be a “good” salary, and arrive at a number of jobs “created” by the license.
Bayh-Dole, as a law that builds pre-approval into federal funding agreements, is pretty interesting stuff. It leads to thoughts about how breakthrough networks come about; it respects the flexibility that inventors in various employment circumstances might need to decide how best to proceed; and it leaves available the various ways by which the patent system might be used to promote use of inventions, without dictating products over standards, or proprietary investment over open innovation.
Bayh-Dole, however, must carry the blame for the damage it has caused to universities. Fire is easier to see at its onset than drought. And Bayh-Dole, in its faux vesting form, as advocated by Tresemer, among others, has been the cause of a great drought in university contributions to the public domain, to the commons (by means of non-exclusive licensing), and to industry (through collaboration without high overhead costs). While Bayh-Dole is made to stand for prosperity through basic research, it is in point of practice doing just the opposite–it is undermining basic research, starving the commons, corrupting university administrators, and stripping faculty of their freedom to teach and practice what they have discovered–all so that administrators can sell off assets to speculators. We live in an age that worships speculators–in mortgages, in energy, in student loans–whatever is easiest to place bets on. Bayh-Dole has turned into just this sort of play for university discoveries and inventions.
It may be possible to rehabilitate Bayh-Dole. If so, the starting point is to require each contractor to make public an annual report that identifies each invention that has been assigned, and identify the state of patenting, the date of first commercial use or sale of the invention, and a range for cumulative gross royalties (patent expenses; $0 to $100K, $10K to $1M, $1M to $100M, $100M+).
The second change is to make the inventors have the first discussion with the funding agency. That is, once the invention has been reported, the inventors decide whether to elect to retain title, or convey title to the agency. If the inventors chose to retain title, then they are subject to 37 CFR 401.9 requirements, unless they assign to a pre-approved agent, which is then subject to 37 CFR 401.14(a) requirements, as may be modified in the relevant funding agreement. Make this an condition of the federal funding, so that the employer, in the case of a federal grant–which covers most university activity–cannot establish a contractual claim to any invention other than by mutual agreement–not as a condition of employment or use of resources or involvement in federally funded research.
The short form is that the government is providing grants for the support of independent faculty, not for the financial exploitation of university administrators. The funding agreement by which money is conveyed to a university cannot give standing to a university to force assignment of inventions to itself. Instead, the (f)(2) agreement should be incorporated into an amended Bayh-Dole Act, and include a requirement for release by the employer of any prior employer claims to inventions that might be made under a federal grant (as distinct from a federal contract). That is, the federal agreement should take precedence. Universities may still be in the patent licensing business, if their faculties so choose, but they need to get out of the compulsory assignment business.
Tresemer spends a great deal of time discussing all the various ways that university administrators can nail down rights to faculty inventions. He ignores the particular issues of federal funding and inventions and expands the idea of university ownership to intellectual property generally, regardless of the source of funding. He shows how a university can cleverly beat out another university when hiring away a faculty member by not taking notice of any obligations that new hire might have to his or her former institution, and using 35 USC 261 cleverness to acquire rights to inventions made previously. Tresemer aims to lock down about every possible variation for ownership that he can think of. It is a script that I am sure faculty will start to see surface in proposed changes to policy, justified post Bayh-Dole. Or, if administrators follow the lead of the University of California and the University of Washington, they will claim that the new assignment language isn’t a change in policy at all, just a better reflection of what the administrators had wanted all along and just hadn’t, even with thirty years to do it, found the right words for it.
The problem, however, is not merely in creating a dictatorial, institution-demanding view of university research activity, down to demanding detailed note-taking on anything inventive, but rather in addressing the policies under which universities should come to make such demands of faculty. Such policy discussions are not the domain of lawyers, not even law professors, but rather of the university community generally. For that, there simply is no meaningful connection between the use of research-originated inventions and institutional ownership. I challenge readers who disagree to present the case. Inventions may be used and developed without any ownership whatsoever. Inventions may be owned by institutions as a result of voluntary assignment. Institutions may deploy inventions as public standards without seeking payment. The internet was built that way, you know.
However, the faux Bayh-Dolers don’t care about all this. They want outright ownership, first and foremost. They have no respect for inventors, even faculty inventors who may be at the forefront of science. It is all a matter of asset creation and control, to be sold off in a mill that produces about one big deal a decade at the luckiest institutions, and perhaps once in a hundred years at the rest.
As Tresemer puts it:
While inventors and patent purists may take issue with employment agreements designed specifically to divest inventors of their own creations, it is critical to view such agreements through the lens of the Bayh-Dole Act and the problems it was designed to address.
Inventor loathing at its most refined–and this, coming from a law professor. I guess freedom is nothing compared to the pleasure of serving the man. The lens of the Bayh-Dole Act is a poor place to consider how basic research contributes to innovation or economic vitality. Especially when the lens is that of the faux Bayh-Dole Act.