Research Enterprise

I’m going to work through the May 2010 amicus brief filed by AUTM WARF and others in Stanford v. Roche. The aim of the brief is to protect Bayh-Dole relative to a finding by an appeals court that a university can indeed screw up its Bayh-Dole rights. I’m not saying that’s what happened, but that’s the effect of the appeals court ruling. WARF, AUTM, and others want to set things right. Let’s check it out paragraph by paragraph, shall we? Warning: I’m writing this as I go. It covers the ground. I’ll need to go back and check some things, so there may be some interlinear revisions. I’ll try to show those. Here we go….

The opening pitch–failure to change the decision will “cloud” universities’ title. Hmmm. Only if universities fail to protect the government’s interest in inventions. Bayh-Dole is *not* about universities getting title–it is about creating a uniform policy for agency interest in research-originated patents supported with federal funding. Universities have the right to elect to retain title–but there is no obligation to do so, nor obligation in electing to retain title to holding that title. A university could, for instance, designate an affiliated research foundation to hold title. Or, a university could assign its Bayh-Dole interest to any organization having a primary function to manage inventions–and that organization could then elect to retain title, without the university ever having even seen an invention disclosure. So, we are not taken with the opening alarmist premise.

At stake: ownership. Let’s see about the second paragraph. Federal funding enables lots of things. Vitamin D, blood thinner, and MRI (all WARF based inventions). Well, Vitamin D and the blood thinner at least were both *inventor owned* though federally funded. In the case of the blood thinner, Link is said to have signed over his patent rights to WARF (not to the University of Wisconsin) because “I’m a loyal Wisconsin man” and “I wanted to do something for the university.” Neat. I like that. It wasn’t “because I had to or be accused on unethical behavior” or “I didn’t have the rights in first place, they were stripped away the moment federal dollars touched my work.” These are all inventions that are pre-Bayh-Dole. It appeared to be a heady time of great inventions, patent licensing, and money-making led by publicly spirited faculty. Not like today.

Now the pitch about owning inventions. This is tied to “enjoying returns”. But that’s fallacy. One doesn’t have to own an invention to enjoy a return on its commercialization. One could enjoy that return through more taxes, jobs for students, better health care, lower costs for goods–any number of things. One could even dispose of ownership claims in advance in return for “returns”. Further, there’s nothing to show that royalties from federally supported inventions have “contributed significantly” to the universities’ research and educational missions. More so, one can show that the licensing income, in all but a couple of handfuls of instances over 100 years, has been only a minor thing on university budgets for research, is largely regarded as pork funding and appropriated to administrative delights like new buildings and bridge funding for faculty who haven’t got a replacement grant.

Here’s the first real wrong thing. Bayh-Dole did not grant universities “initial title” to inventions. It granted them the right to elect to retain title, which then could be disposed of through assignment of Bayh-Dole obligations, subcontracting, or assignment, always with a lien on any transaction for government rights and possible government intervention in the case on non-use or unreasonable use.

Now, on to the Federal Circuit’s problems. The result, claims the amicus brief, is a return to pre-Bayh-Dole. Hardly. There doesn’t appear to be a lot of uncertainty over ownership pre-Bayh-Dole. It was the US government’s per agency funding policy absent a patent administration agreement, and if there was one of those, then that spelled out initial claims on title. That was sort of like divvying up the spoils, since initial title is with the inventors under US patent law, and the patent administration agreement was along the lines of who got to take that title away from the inventors. Bayh-Dole regularized this, so it is along the lines of: if you do it, then you have to also do these other things, but you don’t have to do it if you don’t want to. The amicus brief cannot imagine a university not wanting to, even while the brief is led by a foundation that’s not a university at all! Woah.

Now into the meat of it. In Stanford v. Roche, we have competing obligations to assign and a chronology of events. First there’s a promise to assign to Stanford everything Stanford wants. That’s a generic promise to assign, directed at Stanford’s interests. But Bayh-Dole is about protecting the government’s interest. The problem for Stanford, and for all universities that do something similar, is that it’s not true that protecting the university’s interests must also somehow ensure compliance with Bayh-Dole. Not so. But that’s been the assumption. But wait, there’s more. The brief argues that if inventors own federally funded inventions initially, then Bayh-Dole is undone. Think about it. Stanford has the policy that requires inventors to assign to Stanford, which it relied upon to comply with Bayh-Dole, but the amicus brief argues that Stanford didn’t have to have the promise to assign, because Bayh-Dole itself places title with university administrators not with the inventors, until the administrators decide they don’t want title. But what Stanford didn’t have was the agreement that was specified under Bayh-Dole–namely, a written agreement signed by each research investigator that protects the *government’s* interest. Why would Bayh-Dole require such a written agreement if the law could be interpreted so that inventors had no rights to their inventions? Why wouldn’t Bayh-Dole require university administrators to sign the agreement to protect the government’s interest, if they held title outright upon invention? Bizarre logic. It tumbles away in pieces. There is no recovery. But we continue.

So we have Stanford’s promise to assign (to Stanford), and then the investigator signs an agreement with Cetus to learn about PCR that includes a present assignment of future inventions that are made arising out of the PCR information he receives. How far does that scope reach? Everything forever? That’s the big question left uncertain, but it’s not the tree AUTM and the amicus briefers are barking up. Instead, they are barking up the ownership tree under Bayh-Dole. Apparently if there’s looting to be had, then make sure you get in on the looting big-time. Why worry about a single invention with a quarter billion on the line when you can go after everyone’s inventions?

Let’s point out that Stanford administration apparently knew about the consulting. It wasn’t clandestine.

Okay, so the investigator at Stanford with federal funding co-invents. At that point, there are two competing assignment docs, Stanford’s and Cetus’s. One is a promise to assign future inventions within the scope of Stanford’s interests and the other is a later present assignment of future inventions within the scope of Cetus’s interests. Both are laid out in writing. Which one takes precedence? That’s the question. What a puzzle. But there’s a complication, and that has to do with timing. The Bayh-Dole order of things is: to obtain the benefit of the law, an institution hosting federally funded research has to obtain a disclosure, report it to the agency, and notify the agency of election to retain title. That is, under the law, the host does not have the benefit of standing in for the federal agency to dispose of title until it has notified the agency of its election to do so. It’s not that it gets to decide in its administrative head–I’ve got title–and then at convenience notify the government. It doesn’t have any standing to require assignment of title under Bayh-Dole until it has notified the agency of its election.

Another thought. If Bayh-Dole cancels Cetus’s claim (Cetus’s PCR stuff acquired by Roche), then it also cancels Stanford’s promise to assign claims. That is, the written agreement required by Bayh-Dole is not directed at university ownership but at protecting the government’s interest. What the university makes its employees sign but for protecting government interests, isn’t in Bayh-Dole. If the university doesn’t elect title, then that decision advances to the federal agency. If the federal agency doesn’t take title, then it can consider leaving the title with the inventors, or it can force title to the public domain. If the university does not elect title, and the agency does, it gets assignment *from the inventors* not from the university. If the agency decides not to take title, then Bayh-Dole allows the inventors to take title (subject to agency approval, which includes consulting with the university–but doesn’t include a university veto). The important point is, Bayh-Dole does not have the agencies assign anything to the inventors–they hold the rights under US patent law. After all the reviews, the rights are not requested, and they are where they always were until assigned: with the inventors under US patent law. The idea that Bayh-Dole requires a bunch of assignments from universities to agencies and agencies to inventors is unfounded. It hasn’t happened in 30 years. This is not an OMG moment where the university tech transfer community discovers it has been doing things all wrong. No, this is a WTF moment when certain visible parts of the university tech transfer community prove how badly they can reason about a fundamental of its activity.

Now for the chronology. As best I can make it out, it goes like this. The inventors disclose their invention to Stanford, which then prepares and files a patent application. It then notifies the government of its election of title, and then goes and gets an assignment from the inventors. Later, it knocks on Roche’s door, proposing a license. When this offer is turned down, Stanford sues. The pivotal act for the appeals court is the filing of the patent application. It is at that moment that everything comes into play formalizing the existence of an invention and freezing the camera on title. The court has to look at this. A promise to assign, a later present assignment of a future invention, federal funding requiring protection of the government’s interest in the invention, and a suit claiming that Stanford owns the patent, and a defense that it doesn’t.

The court sees that it can reduce its problem to the two claims on assignment. One is a future obligation, the later one, a present obligation. Well, if nothing intervenes, then the court decides that the later, present assignment controls, especially since Stanford was aware of the consulting relationship. In a sense, the court finds that the later consulting carves out, with Stanford’s concurrence (at least by not opposing), the company’s interest. So if a patent application is filed, it becomes Cetus’s. If someone didn’t want that result, they had to do something–like, obtain an assignment *later* than the present assignment or reach agreement with Cetus as to the scope of that present assignment. Stanford didn’t do this. So the court looks at it and goes, geez (not really, courts don’t use that kind of language), it sure looks like Stanford thought it had this, but really, when it goes back to get an assignment from the inventors that pretty much tips its hand. 1) it didn’t have title until then, by its own actions; and thus, 2) it couldn’t possibly have obtained title from the inventor who had committed his future inventions to Cetus, because the moment the application was filed, it became Cetus’s because clearly that was the controlling instrument. Any later assignment was void. There was nothing left to assign. Yes, it appears Stanford messed up. Well now. Is it worth a quarter billion bucks? It looks like the court avoids Bayh-Dole altogether, which is as much as what it said it did in its opinion.

What would Bayh-Dole add to this situation? The amicus brief wants it to impose an ownership claim at the start of the federally funded research that comes after the signing of the present assignment of future inventions to Cetus. That is, the brief wants the federal funding to create a pre-emption of the private assignment of rights rather than to curtail the scope of that private assignment by imposing a new scope of interest. That pre-emption, of course, would also have to reach to anything in Stanford’s own promise to assign instrument that was within the scope of Bayh-Dole. That is, the only thing left to argue is invent for hire–that the university has the patent rights and doesn’t even know it–nor does any other university until this WTF moment when after 30 years folks realize they didn’t have to get assignments from their inventors.

This argument, however, is, let’s be blunt, a heap of parrot droppings. The better angle of attack would be on the competing scopes of interest in the present assignment and the later federally funded research, not in competing claims on title. The upshot of the amicus brief argument is that it is impossible for an inventor with private obligations to co-invent with university inventors receiving federal funds. For instance, a university researcher cannot go, gosh, we’re having this problem that we can’t solve, let’s call our friend at Company A and see what he says, or even, let’s call researcher friend at University B and see what she says. Either way, if the interaction results in an invention, it would be a joint invention, and title would be undivided among each of the joint owners. It really doesn’t matter if the other joint owner is a company or a university–the title is shared. Bayh-Dole doesn’t care about any of this. It doesn’t prohibit private participation in research, and doesn’t even extend the obligation to protect the government interest beyond the host institution’s *employees* and even then excluding clerical and non-technical employees. Something is amiss in the focus on title.

Here is what is going on. The purpose of Bayh-Dole is set out in the law. It is to use the patent system to promote the use of federally supported inventions. That goal may be realized through a variety of things, such as better university-industry relationships, supporting small companies, creating American jobs, protecting the public from non-use, commercialization, and lowering administrative costs to manage inventions. Nothing in that list of objectives says Bayh-Dole was a clever ploy to help universities make money from patents. While there is provision for what universities (and only universities and other non-profits) have to do with their patent money if they make it, there is nothing in the law that requires university money making or sets it out as an objective. There is nothing, for that matter, that prevents universities from sharing *all* patent royalties after costs with the inventors. Just none of them do this. Funny.

It is brazen nonsense, this amicus brief, setting out the idea that title is critical to university money-making under Bayh-Dole. It’s actually an admission that universities are not on the same page with the government on objectives. They are coming out arguing that Bayh-Dole is a covert financial aid package for universities allowing them to knock up industry for patent fees. For this to work, for lack of any better argument, they are arguing for an invent-for-hire reading of Bayh-Dole. Perhaps it is just sloppy thinking–or perhaps these folks really want a law that hands them patent rights with a mandate to beat up industry. Either way, it’s bad reasoning on the law, and a bad idea in practice. On the latter: the position is anti-inventor, as it purports to remove personal invention rights and place them with a bureaucracy. What about this points to creative class? Yes! A bureaucracy! That’s the ticket! Does it make any sense? Is that what the Economist thought was so inspired? That the federal government said, “why, let’s move this stuff from a federal bureaucracy to a bunch of risk-averse, under-resourced university bureaucracies! That will horse up this innovation thing.” No, sadly, Bayh-Dole did not arbitrate a competition between two bureaucracies over money-making interest in research patents. But that’s the thrust of the amicus brief.

So we get back to business. The court found two competing assignment obligations, saw that the latter controlled, looked at the chronology of events following and saw that Stanford had acted too late to prevent the present assignment to Cetus to operate and found that this operation was not pre-empted by later contracting conditions accepted by Stanford. So, title to Cetus, end of patent licensing dispute. Sorry, guys. Further, we can observe, that the dispute involved a product on the market apparently within the scope of patent claims (but did it benefit from the federal research as well?). That is, as far as federal invention policy goes, this would be a success–the public has the benefit of the invention on reasonable terms. Stanford has rights through two co-inventors, and the government has its non-exclusive license through the administration of those rights. Where is the government’s interest in the invention not protected?

We have a few things to clean up. First, Stanford can’t exclude Roche, so can’t collect royalties. That may be a downer for Stanford and even for the inventors. But that’s a money thing, not a public policy issue.

Second, we have the problem of competing scopes. Is it possible that the later scope of “planned and committed” activities with federal funding excludes the prior, broad scope of interest in the present assignment? That is, would accepting the federal scope of interest and its requirements mean that an invention within the scope of the federal funding agreement come later and displace the prior scope? If so, does Cetus/Roche have any cause of action against its consultant-come-inventor? Especially if Cetus/Roche knew its consultant with his assignment obligations was participating in federally funded research? My thought is that if the later assignment obligation takes precedence, then so too ought the later agreement on scope, on the same arguments (such as, Cetus/Roche did not object to the extent they knew, but in any event evidencing that the investigator had no intent to honor a broad interpretation of the prior present assignment and his intent would be reasonable under the circumstances). Otherwise, the investigator simply could never take other employment in this an area of expertise for receiving some training. That would appear to go against California labor law, and would be generally against public policy, as any generally drafted assignment agreement would always be interpreted against the consultant. That outcome, I think, should interest the courts in making clear the limits of the scope of present assignments in subsequent employment situations.

Third, we have the outcome that Cetus/Roche may be co-owner of a patent to a federally funded invention. If so, then does Roche have the same obligations that the university does? That is, should Roche have been required to disclose the invention and timely notify the government of its election to retain title? Is it possible to construe that the effect of the present assignment was to make Cetus/Roche a co-sponsor (in-kind) of the federal research by allowing its consultant (holding the thinnest of relationship in the continuing commitment to assign) to participate in the federally funded research? If so, what happens if Cetus/Roche has failed in its federal obligations? Could the government take title to its undivided interest? Could it be that the same federal preemption of private obligations holds–that regardless of the prior present assignment, Cetus/Roche actually has no position on the federally supported invention until it has elected to retain title? If it has failed to do so, and gets sued, does the government have to allow it to retain that title later? Don’t think so. So it would seem that there would be good appeals on the operation of Bayh-Dole without inventing a super silly WTF scenario in which university inventors for 30 years are stripped of their personal rights in inventions under US patent law. Surprise, surprise, surprise. No, we don’t need to go there.

1) scope of a later obligation to invention rights
2) obligations under federal funding to Roche and failure to comply
3) consequences of a broad present assignment for later employment interpreted against the inventor

Now back to the amicus brief. Dire effects if the Court doesn’t act, say the friends in their brief. This is throwing sticks in the air. That’s because it’s just an assertion, without a clear argument to compels this conclusion. The actual consequence of not acting is that universities will have to get their act together and learn what Bayh-Dole is about and how it operates. After 30 years, you think this would be something folks got. But they don’t. Clearly, there is self-interest everywhere. The Supreme Court shouldn’t spend time on this case for the reasons given in the amicus brief. The three issues above might be of some merit, if the dispute reaching the Court were about any of these. But if it is about title under Bayh-Dole, then it doesn’t appear there is anything gained from spending the Court’s time in explaining to university counsel that Bayh-Dole isn’t about what they claim, and they ought to do their homework.

Oh, and did I mention that universities ignore the fact that Bayh-Dole is embedded in federal research policy via OMB Circular A-110 at __.36 (b). Turn your attention to __.37 and you will see that it is federal policy that intangible assets acquired or improved under federal funding, including IP, including patent rights under Bayh-Dole, are to be managed in trust for the beneficiaries of the research. So, not a clever ploy to allow universities to make money by threatening to shut down products on the market under reasonable terms.

Whew, that’s just the summary of the argument! Do we need to go on? Yes, let’s hear these folks out.

Let’s look at the argument. The brief claims the appeals court “fundamentally misunderstands” Bayh-Dole. Might be better to say the court interpreted Bayh-Dole differently than prevailing practice, or in an unreasonable way. But there’s good indication that the court did understand Bayh-Dole, at least enough to see that it didn’t matter to decide the dispute it was presented with.

Title to federally funded inventions does not vest in the research institution. The research institution only has a right to elect to retain title. The election to retain title is a decision who has standing to obtain title from the inventors, as between the agency, which historically claimed that right in its research contracting, and the university, which in a few instances had negotiated exceptions to this claim on an agency by agency basis. In those instances, we can note, the institution went and got the assignments from its inventors–there was no invent-for-hire. Think “retain” as in attorney, not “retain” as in water.

We are at story time. Before 1980. The federal government owned default rights to inventions. Not so, at least not accurate. The brief wants there to be invent for hire prior to Bayh-Dole, so Bayh-Dole doesn’t have to change the law. But before Bayh-Dole, the federal government contracted in each disposition of funding for research so that it would receive assignment of any inventions it wanted to own. The default for title was in the research contract, not in patent law. (There are only a very few instances, for short durations, where title is not initially with the inventors as a personal right under US patent law–one instance was in inventions involving atomic/nuclear weapons in the 1950s). Note, for what it’s worth: federal educational funds–such as scholarships and fellowships–carry no such claims on invention. We then get the conventional issues about agencies not doing a good job licensing the patents they did obtain. Then we get an assertion that Bayh-Dole clearly vested ownership in the contracting institutions. No chapter and verse yet, but we are getting there. I’ve previously done my own chapter and versing in this blog, so won’t bore you with the truth.

Rights are allocated to the contracting institution if the institution elects to retain title. What isn’t said is that the *rights* are *rights under contract* not *rights under patent law*. That is, the contracting institution has the ability under Bayh-Dole to stand in ahead of the federal agency in deciding the disposition of patent rights in federally funded inventions. That is, electing to retain title means that the authority under the federal contract is transferred, with conditions, to the institution. It has not got title to the invention, only the right to retain the title relative to any claim that might be made under the funding arrangement by the agency (except in the case of exceptional circumstances or other modification of the standard patent rights clause–another thing, for later). To get actual title, the institution has to go get an assignment–an act by which an inventor’s personal rights are transferred legally to another. Again, Bayh-Dole does not require this assignment be to the contracting institution. It could be to any organization that has as one of its primary functions the management of inventions. The right to retain title and the benefits of holding title are entirely distinct things, but easily confused with sloppy reasoning.

Section 202(a) no where states or suggests that inventors own initial title because it does not have to. They do under patent law. They did under prior federal contracting practice. They still do when they are federal employees under Stevenson-Wydler. What would be required in 202(a) is just the opposite–that patent law as being changed as to ownership to place title outright with the institutions, presumably upon notice to the government. More so, 202(a) would have to be at pains to make clear that all other interest in federal inventions, such as by co-inventors who were not employees of the institution, is also made void by the law. The entire title, undivided, would be with the institution. Not there. No reason for it. The law didn’t change. The MPEP didn’t change. University patent practice didn’t change. For 30 years. Until this WTF moment. WTF.

The institution’s rights to obtain title ahead of the government are what are allocated by Bayh-Dole. The step of securing those rights remains. It’s not a matter for Bayh-Dole so long as the government’s interest is protected, including an interest in seeing federally supported inventions being used for the public benefit. Bayh-Dole harmonized agency rules on patents. Bayh-Dole allowed institutions to step in and claim rights on the same research contracts in the same manner as the agencies, with conditions imposed on the institutions’ actions. No requirements on how the institutions dealt with patent ownership. Wide open. Hence the choice discussion in places like HBJ, which AUTM in particular has tried to shout down, consistent with their efforts here. Is there a connection? Hmmm.

Now to the issues of inventors. They may claim patent rights if the agency’s interests in the patent rights are waived. Those interests include first that the research host not elect title, and second that the agency itself does not require assignment of title. What is truly interesting is that if an inventor is allowed to claim patent rights directly under Bayh-Dole, the obligations on the inventors are much less burdensome than those imposed on the universities! See 37 CFR 401.9. The inventors have to notify the agency if they cease to pursue patent protection, include a federal funding statement in issued patents, report on utilization, prefer US industry in exclusive licenses, and agree to march in provisions. No royalty sharing, no use of remaining funds for research or education, no limitations on assignment, none of that.

Now we get to “sharing royalties with the inventor”. Here we find that universities are making a case for conversion. The royalty sharing language is specific to non-profits at 37 CFR 401.14(k)(2). The royalty sharing shows up as part of the *expenses* incidental to to the administration of subject inventions. It is not largesse of universities. It is expected that universities, just like industry, will pay inventors royalties in exchange for obtaining title or a license. There is no reference to a standard royalty sharing schedule. It is almost as if the inventors and university would decide on a appropriate sharing, on a case-by-case basis, like in the old days, when a university inventor might work it out with a research foundation. The idea is that the university negotiates its royalties with inventors, who are funded by the government not the university, doing work they proposed not assigned by the university and for which the university acts as a contracting steward not as the organization controlling the work. In short, these are not corporate inventors, and universities are not told to pay a token amount. They are to share royalties and not to dilute the royalties by cutting side deals (as we learn in Singer vs. The Regents of the University of California). There is no support in the law for the claim that “the basic right to royalties belongs to the institutions”. Further, we have Circular A-110 that says it is federal policy that the institutions act as trustees with regard to inventions (and patents) acquired or improved with federal funding. Bayh-Dole says: inventors are also beneficiaries. Circular A-110 says: institutions are trustees for the beneficiaries. The overall impact is clear: universities hold patent rights on behalf of inventors, and for the benefit of the public. It’s not about the university as corporate owner. If there’s one thing federal research policy is *not* about it is that universities are in it for themselves! That would be to breach their ethical duties as trustees, and would make them indistinguishable from corporate contractors, thereby erasing the need for special treatment under Bayh-Dole!

For federally funded inventions, the university is *never* a corporate owner, acting free and clear without obligations to inventors, the government, and the public. That’s federal policy. What desire to transmorgrify university research into a heaven for bureaucrats managing patent rights could lead to this brief’s argument? Regardless of the motivations or capacity of the drafters of the brief, with all due respect, they haven’t got it right.

I suspect that they are deliberately not getting it right. They are not in it to have an honest discussion about things. They are in it to manipulate the language of the law. I’m thinking they think the Supreme Court is going to play dumb on this, and won’t bother to read Bayh-Dole, or if so will want to re-interpret the law to give universities a leg up on litigation. Let’s read on. It’s just too good to be sloppy, and it’s too clear to be an OMG mistake.

The brief argues that the law requires (“Congress”–well, yes) that funding agencies approve all assignments of rights in federally funded inventions. But the exception noted “to invention management organizations” is not accurate. Look at 37 CFR 401.14(k)(1). There must be approval for assignment except if to an organization which “has as one of its primary functions the management of inventions” + flow down of Bayh-Dole. That can be any organization with a “primary function” in managing inventions. Arguable, companies with a robust patent shop fit the bill. The apparent concern is that assignments might be made to organizations unqualified to manage the obligations under Bayh-Dole that also must flow down, hence agency approval in those cases. It’s totally narrow to say only “patent management organizations”. But it is consistent with mounting a reading that Bayh-Dole took invention rights away from inventors, handed them outright to universities, and made the money making a university right. The alternative reading is that Bayh-Dole changed nothing about where inventions vest initially, gave universities the opportunity to step in on uniform terms, and required that they respect their inventors as well as promote the objectives stated in the Act, which don’t include making a ton of money from licensing. Though that possibility is anticipated, it is not an objective. It is not federal policy. It is not in the patent law. And it is bad public policy. But we find universities unabashedly arguing for it. Is it greed? Is it arrogance? Is it desperation? What drives these folks to be so given to attack their own inventors and think the Supreme Court will not see through it?

The amicus brief then argues that if inventors held personal rights then there would be no reason for the law to require limits on assignments. That’s too funny. Isn’t it? Back at 37 CFR 401.9 we see that there are no default limitations on assignments by inventors of their personal rights, should everyone waive ahead of them in the claiming line. The limitations are specific to the non-profits! It is the non-profits the law has questions about, not the inventors. It is the non-profits the law doesn’t trust with the public interest outright, or relationships with industry, and therefore imposes limitations on assignment to qualified organizations. Yeah, make it out that inventors are the problem, when they are the asset.

The amicus then repeats language from the Senate report that sets up that the research institutions have the right to elect to retain title. This, via assertion, means, somehow, that they have title under patent law, not that the framework for the disposition of rights in federal research contracting is that institutions have a contract right (federal contracting) ahead of agencies. We have to go to a report summarizing the findings of the Congressional Budget Office to get to something as close as “automatically grant small business and non-profits title to inventions”. The concern here is not that patent law is changed to dispossess inventors of their personal rights, but that federal contracting administration is changed to eliminate bureaucratic dithering on the agency side. That’s what has become automatic. Not invent-for-hire. But invent-into-consistent-federal-framework-with-conditions-on-institution-management.

We can summarize, then, contra the amicus brief: “In short, initial patent rights in federally funded research belong to the inventors unless and until a federal agency or a contracting research institution acting in conformity to the law and standing in for the federal agency to accomplish the government objectives set forth in the law, requires other disposition of title within the scope of its authority and consistent with any agreements or state law that may govern such transactions.” The principle claimed by the amicus brief is not at all clear. The conversion of a private right disposed by federal research contracting into an outright claim to title by university administrators is not supported in the law, in practice, or in the Patent Office. It’s a wild gambit. Someone must have said it’s worth a try and hired some mercenaries to make a go of it. But whatever. It’s wrong, it’s bad, it looks intentional, and it should be set aside.

Now to Part B. I’m assuming if you have got this far then it’s either very important or you are with me for some other reason. Thanks for coming along, in any event. Part B of the brief looks like chest thumping time. Now that we have shown we can misread the law, we can show how important that misreading is to us. And given that the law’s objectives were to benefit small business, American labor, university-industry relations, university inventors, and protect the public from non-use, bad patent managers, and administrative overhead, with commercialization thrown in without indicating whether the universities got to hold a payment receiving end of the whip at every turn–given all this, we can see how the law was really intended to benefit university technology transfer offices. What do you say? Is that what we expect?

Now we move to an argument that university-owned inventions have been important. But this is a different thing from subject inventions. Only 70% or so of university research funding comes from the federal government, and of that only so much actually goes toward research with an inventive profile. Some goes to support curriculum development and taking a bunch of photos of stars. And look at the numbers–university licensing at $100b over a decade is $10b a year–and 70% of that is say $7b. So let’s say across 100 universities we have $7b to $30b. That’s economic impact. When AUTM does this calculation, they take the licensing income reported and multiply by a factor reflecting an expected royalty rate–50 for 2%, for instance. Making a conservative estimate, imagine an average 5% royalty–that would mean multiply reported licensing revenues by 20. Since we have cited those robust estimates, we divide by 20 to get presumed licensing activity–or, $350m to $1.5b. These are blips. Divide by 100, say, and we get $3.5m licensing income, on up to maybe $15m, per year per institution. It’s not a lot, especially where those universities are receiving upwards of $100m each in federal funding each year. And we’re not even pointing out that most of the income reported comes from less than 5% of the patents under management–about the same licensing rate that the federal agencies had and which was so damnably poor to motivate the passing of Bayh-Dole. Durn if that were the case, no? Thus, no university I know of reports separately its federally supported patent work, or shows how many such inventions are actually licensed to practical application. Clearly, they do not want people to know, and they do not think it even important that people know. Funny, the research is supposed to be public, but not the innovation activity. Even Bayh-Dole protects under FOIA university-submitted utilization reports. So the universities would have to want to provide the information for anyone to get at it.

We get to the problem of uncertainty in title. That has always been with us. If an investigator phones a friend, or has a chat at a conference, or visits a company, there’s always a chance of joint invention, and with it comes shared title. Federal research contracting is not designed, at least for fundamental research, to isolate the contractor and prevent any other interaction. The clarity on title sought under Bayh-Dole has to do with the issues pertaining to agency contracting, not to the nature of inventing and collaboration in research. To argue that a university must have exclusive title or the whole system fails is just nuts. We work all the time with issues of title. Who claims to have invented? Are they an inventor of what’s claimed? What’s claimed when the patent issues? What organization are they with? Who do they owe their rights? Will they or their organization cooperate with us? All this happens, every time an invention report comes in. Bayh-Dole doesn’t change any of that. These folks want to isolate university investigators, prevent them from having any other collaborations with other universities or with companies, and while they are at it, strip any other inventors of their personal rights in inventions that carry some federal funding. Don’t think it’s a good idea, really.

In this, I think the appeals court got it exactly right, as summarized by the brief: “Under the court of appeals’ decision, that presumption of title has disappeared: title may belong to individual inventors or to whomever those inventors may have assigned it (advertently or inadvertently).” Yes! Exactly! Except the presumption of title *never existed* so it could not *disappear*. There is no *presumption* at all. There is a contractual deal, set out in federal regulations, over who has first dibs, if you wish, on the inventor’s rights. It is up to the university to conduct its affairs so if this is not desired, or is contrary to the government’s interests, that it does what is necessary to comply. The government may be damaged by the university’s actions as contractor when it fails to take corrective action. It is damaged if it does not have rights to what it funded (it only needs non-exclusive rights for its own purposes). It is damaged if an invention is not used or its benefits are not available on reasonable terms. If a university screws up, but it’s only because it doesn’t get any licensing money because it only holds an indivisible right to as joint invention owned by others as well, what is that to the government?

Oh, and then we get the “university inventors are stupid, gullible, and greedy, and university administrators aren’t” argument from the previous brief! We need not comment here, since we’ve done that elsewhere. But will they include the academic freedom argument? Yes! What audacity! Call the AAUP, ask what they think about it! Let’s get this clear: academic freedom is impinged because stupid, gullible, and greedy university inventors will dispose of their rights to prevent research from going forward, products won’t be built, and we will suffer a dark age. Yeah, like how university inventors built the internet. Or how, 50 years ago, they built the idea of using research patents to promote more research via organizations like Research Corporation, WARF, and a number of other university-affiliated research foundations, not run by bureaucrats, and often run in collaboration with industry, which helped to motivate the effort rather than being a target. And hey, that infrastructure formed the basis for the National Science Foundation, and from there it’s an easy step to see how Bayh-Dole offered a mid-course correction for federal agencies. The very model of technology transfer via research patents was built by university inventors. These people writing the brief do not care or do not know. Now they want to claim not that these inventors are not capable, though they also are responsible for drafting the proposals for research, choosing collaborators, hiring personnel, overseeing acquisitions of instruments and data, doing the analysis, writing reports, responding to criticism and inquiries–all this–and they are incompetent to develop their own thoughts on patent work, because, well, one guy at Stanford signs a present assignment and folks can’t sort out title. Bother.

The brief ends with more dire warning about cloudy title and doubly pernicious effects. Perhaps this will motivate the court, if the cloudy reasoning and pernicious intent that went before was not effective. Ominous, but in the form of an actual argument. Just a warning.

With regard to whether an inventor’s assignment may override the government’s rights under Bayh-Dole, this case doesn’t have standing. The government does have rights via the co-inventors. So the thing proposed is a hypothetical. A mystical beast, like a dragon or a unicorn. Let’s say we wring our hands in the presence of this beast. Change the situation. A university investigator consults for a company, signs a present assignment and then in the university joins a federally funded research program in the same area as the assignment, and then invents alone with no one else. What then? At this point, one might ask, why didn’t the university get a written agreement at the time the federal funding was awarded? Or why didn’t the university ask the investigator if there were any conflicting agreements? Aren’t these systemic problems–that is, with university administrators, who earlier we saw were a good alternative to the stupid, gullible, greedy inventors? Hmmm….

Under Bayh-Dole, it is the contractor–the university that is responsible for compliance. If it does not have standing to request title, then it did not obtain the written agreement it was obligated to obtain. It may have obtained another written agreement (such as a promise to assign, or even a present assignment of future inventions)–but it got the wrong one! What it needed was an agreement to protect the government interest *in that research* subject to the standard clause at 37 CFR 401.14. If it doesn’t get that agreement, then the personnel should not be doing the research. Second, even if there’s some agreement hung out there on assignment, the university should do things in the proper order: notify, elect to retain title, then get assignment, then file. If necessary, at assignment, and there’s something out there, go do business and deal with the problem before filing. Finally, it’s the government that grants the patent right, so it’s also the government that lets itself be sued for infringement. I would expect it could come up with a defense that the patent is invalid, or that the patent owner has no standing to sue for failure to follow Bayh-Dole (or other federal patent law) to get at the government for infringement. Thus, I don’t see how the government is shaking in its boots for rights shifting from an inventor to a company. What I would think the government would care about is whether the invention is used, and whether in that use the public has a benefit on reasonable terms, and whether research is adversely affected, and whether there are working relationships between universities and companies. Those things might require some adjustments. But not whether a university makes money, or whether in a few instances in thousands there’s a cloudy title.

There, that’s the scene. It would be great if the organizations involved would withdraw this brief. It would be great to see an argument based on any of the three theories I’ve identified above. I really do not want to see universities get into invent-for-hire. But if they do, then that would be a great reason to go after changes to Bayh-Dole, and reset things for the next 30 years. If Bayh-Dole was inspired legislation, now brought low by an aging, conservative, inflexible bureaucracy, it’s time for a new set of opportunities, looking globally, looking at international collaboration on grand challenges that benefit from research and IP from many places, and that level the playing field for small companies as well as communities of practice that don’t want to work through commercialization or products or corporate factory ethos at all. Maybe that’s the next phase. Maybe if this amicus brief wins out, we’ll get there sooner. Maybe I should just let it ride. But, no, we can push for extensions to Bayh-Dole anyway. We don’t need to make this a crisis of bad reasoning in order to have the opportunity not to waste it.

Here’s hoping to this particular line of amicus briefing, with all due respect to its drafters, and to the institutions they represent, and to the investigators and inventors I’m sure they never talked to, not going anywhere!