Posted on November 30, 2010 by Gerry Barnett

Folks should read Bayh-Dole carefully, closely.

This would be a good time to lay out how folks think the law operates, and back up each claim with citations, and show how the claims and citations all work together. Nothing ignored, nothing contradictory, nothing grossly stupid.

It is especially important to separate out assertions of want that do not have grounding in the text of the law. If Bayh-Dole is your reason for having a job, then you better get it right-at the level of what the law says, how that law is interpreted into useful actions, and what those actions say about you, personally, as a professional working with creative people within a national framework of research innovation.

Look at the claim that Bayh-Dole gives universities title to inventions. The law never says that.

More so, it allows for things other than that. Even more so, why would one not want to exploit those other things, or at least explore them? Yet even more than that, why would one *want* a federal law to take title from individuals and plant it with a university? Why would one *want* the government to impose this requirement? Why would one *want* to take this position of dominance with regard to research inventors regardless? Why would one *want* to have a national research innovation policy with such an attitude toward inventors and research investigators? What has this bit of profession come to? An administrative love of order? Sloppy thinking is okay if it sounds good?

There is a lot of arcane stuff in this biz. Bayh-Dole, however, deserves a much better reading than it is being given by some of its most vocal, most determined "defenders". This, too, should be cause for concern, if not alarm. If the most vocal defenders can't read the law, can't explain it, and are reduced to claiming it stands for something they really, really want and want so bad that they have all adopted practices in line with that wanting rather than the law, because everyone knows what the law was intended to do so pay the actual text no real matter-well then, I think the game is up. Time for a big reset.

For those that want to take a deep dive, what follows are some questions about the text of Bayh-Dole as implemented in the CFR. They are not idle questions. They are not trumped up. They are intended to open up challenges that any close reading regarding title ought to address. At the root of each there are real answers, and those answers make good sense of the law throughout its various parts, and they come to a class of readings in about how Bayh-Dole allocates prospective right to title to certain inventions as between a contractor and an agency. As I have argued elsewhere, the readings I find do not assure either agency or university of title; that inventors are not constrained by the law or the SPRC (the Standard Patent Rights Clause), but must be bound in personal agreements regarding inventions made specific to each funding agreements; and that any reading that Bayh-Dole is a vesting statute results also in a finding that Bayh-Dole is unconstitutional under the 5th Amendment for lack of due process and lack of just compensation, either on the face of it, or as established by inequitable practice.

For those that have the time for something at the root of US university invention management, read on. It will take you an extra few minutes. I'm not sorry about that. It would be good to get a copy of 37 CFR 401 handy, as I cite it often.

*****

37 CFR 401.14(a)(b) says contractors may elect to retain title under the heading of "Allocation of Principal Rights". What does "Principal Rights" mean? Why is it not "Vesting of Title". Could it refer to the allocation of rights as between the contractor and the agency-that is, among the principal parties to the agreement? Or does it mean, allocation of the most important rights-the principal ones-suggesting that there are other rights that are not so allocated here but might be elsewhere, and if so, what are those other rights?

What does "elect" mean and why is it made in reference to notice to the government under (c)(2) and yet not require notification of the inventors, say? Why would the law leave inventors in the dark as to the disposition of title to their inventions? Or does it?

Moving on. Why does (b) qualify subject invention with "subject to the provisions of this clause"? Is that wasted words, a drafting artifact repeating the obvious? Or does it suggest that not all subject inventions are subject to the allocation of principal rights or the SPRC generally. What subject inventions might not be subject? None? Never? Really?

Notice that in (b), the government requires a paid up non-exclusive license, but (b) does not specify who grants this license. Why is that? If the university has title under (b), then why doesn't (b) just come out and say, the contractor hereby grants to the government a non-exclusive license and be done with it? Is this just bad or obtuse drafting, playing with us, hinting at meanings because it is too banal to out and say things directly? Or is there a broader set of meanings intended than a limited interpretation is willing to allow?

Throughout 37 CFR 401.14(a) the language used is "retain title" not "own". Why is that? What does it mean to "elect to retain title" rather than, say, to "own title". Why would the law be drafted in such a flowery way if the intended purpose was to vest invention ownership in universities outright? Why in 37 CFR 401.9 do inventors also have the right to retain title? How can this be if they never had title under (b)? Why would the law not read: "the university at the agency's request will assign title to the inventors" if that is what is meant.

Further, look at the definitions. What makes something a "subject invention"? If it is a subject invention for one co-inventor, is it a subject invention for other co-inventors even if they were not working with federal funding? Or is only a subject invention insofar it is an invention "of the contractor"?

Why is the definition limited to "any invention of the contractor". What does "of the contractor" do by way of limiting the scope of subject invention? Does "of the contractor" mean "already owned by the contractor" or "claimable by the contractor by some means" or "made within the planned and committed activities of the grant" or "made by contractor employees other than clerical and non-technical employees in performance of work under this contract"? Is there a reason why "of the contractor" here is a general possessive and not more specific?

The definition of invention is limited to "is or may be patentable." This, in contrast to some university IP policies that define an invention as "whether or not patentable". Subject invention is further limited to conception or first actual reduction to practice. One might ask, is it possible that an invention can be a subject invention (first actually reduced to practice) but not subject to the SPRC (because not an invention of the contractor)?

We may ask, what makes an invention "subject"? It may not be a new invention-since first actual reduction to practice may trigger the designation. Is it possible that an invention, to be subject has to be claimable by the contractor under some means the contractor has established for this purpose rather than one that the contractor is entitled to by operation of law? That is, an invention becomes a subject invention within the context of the funding agreement when it meets all the elements of the definition, and no other inventions can be made subject inventions by private agreement if they do not meet the elements of the definition.

Look at the treatment of Contractor Actions to Protect the Government's Interest in (f). Why should the contractor be required to "have executed" documents if the contractor has title outright? Why is the contractor asked to "convey" title rather than assign it? When a contractor elects to retain title, what rights for the government does the contractor "establish" and what rights "confirm"? Why is it "establish or confirm" and not "establish and confirm"?

Why are written agreements with research personnel required in (f)(2)? Why are non-technical personnel excluded? Is it because they cannot invent? Or is it because they are not expected to invent? Or is it because the government does not intend to have an interest in their inventions? If the SPRC excludes such people, can university policy force them to participate anyway? Is their work an instance of a subject invention that is not subject? What about non-employees?

Why is (f)(2) restricted to employees? Is "employee" given a different meaning than that assigned by HR? That is, is an independent contractor to a university working on a federal grant an employee for (f)(2)? If so, can the university have an agreement with the independent contractor regarding inventions at odds with (f)(2)? If an independent contractor is not an employee, then is the contract with this person in the form of a subcontract (see (g)), or does the arrangement fall within the "substitution of parties" in the broader definition of funding agreement in 401.2(a)-that is, the independent contractor is subject to his/her own funding agreement with the government, outside that of the university's?

For that matter, why are nonprofits singled out at (k) and required to share royalties with inventors (k)(2)? Why not also small businesses? What is it about nonprofits that requires such special treatment? Similarly, why can small businesses assign title in subject inventions to anyone, but nonprofits are limited to those organizations that have a primary function in the management of inventions (unless they get agency approval)? And why is the limitation directed to the management of inventions and not the management of patents? Can someone manage patents and not inventions? Or are these all the same thing, really, and folks drafting the law really don't care about distinctions of this sort? Hmmm.

Onward. Why is the royalty sharing with inventors called out as an *expense* incidental to the administration of subject inventions in (k)(3)? Why are these expenses restricted support of subject inventions only? (Do you know of *any* university that keeps separate accounting for Bayh-Dole royalty income relative to its use by its technology transfer program? No? Surprised?) If the obligation to share is in the SPRC as an expense, then it is apparently not a function of the university's own IP policy or part of an employment agreement. Is the purpose of (k)(3) merely to repeat that royalties should be shared per (k)(2)? Why say something twice? Why call it out in this way? Or does (k)(3) require a particular kind of accounting, so that royalty sharing is an expense of the university's subject invention management program, and not something that happens later, after the university has gobbled up (or not) administrative expenses?

Does the royalty sharing extend to all inventors of a subject invention or just those who have assigned title to the university? If there's vesting, of course, then how would one separate things out? (k)(2) does not appear to restrict inventors, and even includes expressly federal inventors when the agency approves and the invention is assigned per 37 CFR 401.10. And what do you make of the use of "transfers or re-assigns the right it has acquired in the subject invention from its employee to the contractor"? What is the difference between transfer and re-assign? Why distinguish these two? One might think, hey, if there's a federal employee involved as a co-inventor within scope of a funding agreement, then it is possible that the agency transfers its right to have the employee assign to the government to the contractor. But that would mean that the government does not expect to hold rights outright in such inventions all the time, so could transfer the right to require assignment-that is, no vesting here. This would be a reason for using transfer as well as re-assign. Or does transfer mean, really, only "license" rights as in 35 USC 202 (e)? Just another word for it. Synonyms you know. Drafting artifacts. Don't mean anything.

*****

At this point, you may go-that's a lot of detail and things are pretty complicated in there. Some folks go-"The law is apparently sloppy, but we are clear on our intention to read it everywhere to conform to our will, even if that means distortions, inconsistencies, and stuff that is superfluous. Never mind all that, let's get back to what we *want* out of the law, and keep it simple." Some folks go-"We will never know, until there is a court case and a judge tells us, so why bother, it is not up to us?" Some folks go, "We don't have time for all this detail talk, we will resolve this without being so close to the law. It's more like guidance-except when someone crosses us, then it's law the way we say it is and their resistance is unethical, selfish, and inept." I'm not persuaded these sorts of responses are in the best interests of the law or federal innovation policy generally. So let's find better ones.

Finally, we turn to the Constitutional basis of all this. This is tough. We don't spend time talking the Constitution, usually. We are a people of will and impulse, not of text and interpretation. The US Constitution gives Congress the right to manage patents as means of securing to inventors rights in their inventions. Invention rights are personal rights. When someone invents, it is first in their head, it is thought-it is conception. The inventor possesses it. Under US patent law, inventors hold title until they dispose of title in keeping with the law. The Fifth Amendment says private property cannot be taken for public purposes without due process and just compensation. If Bayh-Dole is cited as the authority for such a taking, then there's a constitutional issue with regard to both how that taking comes about, and what the compensation is for that taking.

Now let's go back to 37 CFR 401 and we might see the problem for implementation.

If the agreement is just between the researchers and the government, then the due process is the formation of the agreement-proposal, award of funding, SPRC, and do the work. The just compensation is the funding in support of the proposed work. Title to the government. How it was in the old days. It's all good. From that point, the government could give up title if it served public policy, or grant licenses.

But now we have Bayh-Dole, aiming to dish title in some way to other than the federal government in exchange for funding. We have state and private entities on the other side of this deal on title, not the investigators and inventors. How does a personal right of inventors come to be held by these other entities? This is the fundamental challenge of Bayh-Dole.

The money paid by the agency to support the research goes only to direct and indirect costs. The deliverable by way of rights is still paid for by the government, but the deliverable by way of rights to these other entities is *not*. The effect of (f)(2) is to implement the deal between the researchers and the government, outside any deal the researchers may have with anyone else. But it is the university's obligation to go get that deal. What is the consequence if the university fails to get an (f)(2) deal with its research employees and some of them invent? That is a central question.

But now, consider what the deal is for the university to obtain title to inventions from its research personnel.

Here are some options:

1) The university invokes Bayh-Dole as a vesting statute. It has the inventors' personal rights outright. Where is the due process as between the inventors and the university? What is the just compensation? It cannot be the funding paid by the government-that's already in the (f)(2) deal between the government and the investigators. It has to be something else. If it is only the promise to share royalties-well, that only works if there are royalties. Otherwise, it is a taking without compensation. It is not sufficient to say i) the property lacks commercial value at the time of the taking or ii) since we pay some few inventors, that counts as if we are paying all of them.

2) The university invokes its standing under (b) as secured by (c)(2) and makes requests under (f)(2) for title. That is, it stands in for the federal agency, and holds title not as outright owner but as steward for federal interests. In this option, the due process is the (f)(2) transaction, and the just compensation is the funding agreement, since the university is only a substituted party for the federal agency's interest.

3) The university invokes its own IP and research policy, and employment agreements, perhaps also with state laws. It ignores the apparatus of the SPRC and in its place requires title based on these conditions-use of facilities, in the course of extramural research, and the like. There is no vesting under Bayh-Dole. For public universities, there is still an issue of taking, insofar as the public university is an instrument of government. The claim is that the provision of facilities and salary are compensation for title, as well perhaps as efforts to secure licenses and share royalties (even those these cannot be consideration because under the SPRC they are independently required of the university and therefore cannot be consideration). The problem as to facilities and salary is that within the funding agreement, the government pays for these as direct and indirect costs. So one is left wondering exactly what, in the context of the SPRC, the university is offering to its inventors that it otherwise is not required to offer.

4) The university negotiates a deal with inventors, under which the inventors agree to compensation that can be shown to be just. The negotiation is the due process. Whatever the university's IP policies and the like, the negotiation and compensation have standing independent of these and instead take place within the context of the SPRC and the university's election of standing to have the negotiation.

For all the administrative convenience of saying Bayh-Dole vests title of subject inventions with universities, and for all the potential feel-good of that (with the concerns about why that feels good I've already mentioned), at the root of Bayh-Dole, one would do well to select readings that deal with due process and just compensation.

A vesting reading allows situations such as this: a university research employee makes an invention in consulting for a company but also in collaboration with other university research employees who have federal support. The company holds rights under a present assignment of future inventions. It is an entirely proper agreement under university policy and with regard to patent law. The university declares the invention a subject invention and claims title vests with the university. The *company* is the entity from which title to the consulting employee's share of the invention is taken. What is the due process for that? Where is the just compensation?

If Bayh-Dole is read as a vesting statute, it is plain wrong either in its intent or in its implementation. Even if it is read to mean that it retroactively changes the scope of what a research employee can obligate, it is still a taking without notice, without compensation. I don't see how anyone would try to read Bayh-Dole that way. I don't see why anyone would want Bayh-Dole to have this effect. It is wrong on so many levels.

Here is where it ends up, however. AUTM, 40+ universities, the AAU, and a number of other organizations have made a compelling case that they do read Bayh-Dole as a vesting statute. It's beyond showing all these folks that they don't know how to read the law. That would be disrespectful of the claim of practice. The fact established by AUTM et al. is that this reading is in practice and is broadly so. Thus, even if Bayh-Dole can be read other ways, such as via option (2) or (4) above, because such a wide and auspicious group of practioners insist that Bayh-Dole is a vesting statute without the requirements for due process and just compensation, the law *still is unconstitutional* for *precisely the reasons AUTM et al. advance*. The language is sufficient broad to permit an impermissible taking.

That's what S v R illustrates only too well. What S wants is title in H's invention to vest after the fact by federal authority in S, so that S can sue R for the value of that title! S wants to take from R under federal authority without just compensation or due process approximately what the title is worth at the time of the taking. I could see a court calling it even-if S gets title, then it owes to R about what it is suing R for. That's a wash. Then the outcome could very well be: yes, Bayh-Dole is a vesting statute, but if you don't do (2) or (4), then you owe inventors and third parties and any deals done in the past without due process or just compensation are void. At least then universities if they keep title are on the hook to pay up for each invention or waive that election to retain title and keep out of it.

AUTM is the kid who kicked the hornet's nest. It's a big nest, it turns out, and it's got hornets. Let's see if they are up for the consequences.