[This article comments on an article at Emory University’s technology transfer office’s web site. The article has finally been removed, so I have in turn removed some identifying elements. I’m keeping this article up, however, for the discussion of Bayh-Dole and copyright, since the lesson is useful, however difficult the learning might be.]
Confusion is like spilled milk. It’s only a mess because it has been spilled, and the effort to clean it up is much greater than the effort to pour a glass of milk to drink. Thus we have two articles rather than two sentences.
A little over a week ago, a university transfer office, posted “Bayh-Dole Reporting for Copyrights” on their technology transfer office blog. At first I thought it must be a joke title to draw in the curiously knowledgeable reader. But alas, no. I pointed out the problems with the article to the folks at the university via Twitter, since they had announced the article there, and all I got back from them was a look at my LinkedIn profile. The article is still up, uncorrected. [Update–the article has now been disappeared. Pity they didn’t correct it and use its catchy title to draw in the curiously knowledgeable reader.] So let’s break it down and see if we can tease out the confusion and learn something helpful in the process.
Let’s get one thing out of the way from the start. Bayh-Dole has nothing to do with copyrights. Bayh-Dole is part of federal patent law. Its scope is restricted to inventions that “are or may be patentable” (or covered by plant variety certificates) when owned by a party to a federal funding agreement. Nothing about copyrights. Nothing.
But it’s worse that simply that. Observe.
With the passage of the Bayh-Dole Act in 1980, researchers are generally allowed to establish rights to their data, even when their research was supported by federally funded grants.
Rights in data also has nothing to do with Bayh-Dole. Rights in data has nothing to do with copyright, for that matter. Rights in data is the subject of a different contracting and regulatory regime. The right in data runs toward trade secret–or more generally the right (under whatever theory) to withhold data from disclosure or use. In a funding agreement, the right in data has to do primarily with whether data is to be delivered or withheld from the commissioning party, and if delivered, the scope of the commissioning party’s freedom to disclose or use the data (and restrict even the contracted party from disclosing or using the data). Bayh-Dole does not require federal agencies to establish rights to research data, not for researchers and not for universities, and not for the agencies themselves.
Here’s what Bayh-Dole does do–it makes a show of excluding from FOIA reports pertaining to the use of inventions and it restricts for limited times federal publication of invention reports and patent applications. These restrictions on federal publication do not establish anyone’s “rights” in the data. They merely prevent the federal government from disclosing reported “data” outside of the federal government. But that data is generally not research data. It is the performance data pertaining to whether the patent system has been used to promote the utilization of inventions, while also promoting free competition and enterprise. That’s generally not the data that researchers think of as “their” data.
For intellectual property that is protectable by copyright, the federal reporting responsibilities under the Bayh-Dole Act are complex, time-dependent, and highly detailed.
More nonsense. Copyright is a form of intellectual property. Copyright vests in works of authorship that meet the federal definition–original, fixed in a tangible medium of expression. Such works are not “protectable” by copyright. Copyright happens. To “protect” a work by means of copyright means to assert the copyright. That is a matter of will, not law. It as if the article writer thinks that patent and copyright are interchangeable. We can revise the sentence to be a much clearer statement:
For works subject to copyright, the federal reporting responsibilities under the Bayh-Dole Act are [… uh… I can’t finish it the way it’s written… ahem,] non-existent. There.
There are no reporting responsibilities under Bayh-Dole for “intellectual property that is protectable by copyright.” It is nonsense, then, that these responsibilities are “complex, time-dependent, and highly detailed.” This is nonsense fantasy.
Now watch as the article shifts abruptly to NIH regulations:
All copyrightable works, including publications and data, may be copyrighted by the grantee (University) without NIH approval, unless stated otherwise in the terms and conditions of the award.
The article starts by talking about researchers “establishing rights” in their data. Now the “their” is turned into an institutional claim. We have, without even a transition to mark the change in perspective, the idea that the NIH allows universities to “copyright” publications and data, at least as a default. The controlling regulation is 2 CFR 200.315(b):
The non-Federal entity may copyright any work that is subject to copyright and was developed, or for which ownership was acquired, under a Federal award. The Federal awarding agency reserves a royalty-free, nonexclusive and irrevocable right to reproduce, publish, or otherwise use the work for Federal purposes, and to authorize others to do so.
Previously, this stuff was in Circular A-110 (__.36), then 2 CFR 215.36. This regulation forms the funding agreement between federal agencies and grant recipients (the “Grantee” and now the much more compelling “non-Federal entity”). The regulation itself is screwed up with regard to copyright–what’s meant is that the grantee may retain ownership of copyrights in works created “under a Federal award.” The regulation leaves open how a grantee comes to own such works–it could be work made for hire or it could be subject to an assignment requirement or it could be freely negotiated for. Similarly, the government license fails to recite various rights of the copyright owner–for instance, to publicly display, or to create derivative works, or reproduce the work in copies, or transmit–but apparently covers all this with the mildly indifferent “or otherwise use the work.”
In copyright licensing, “use” is a wildcat term that opens the door to all sorts of strangeness. Use is not a right controlled by the copyright owner. A copyright owner cannot exclude “use” in general, as a patent owner might do with an invention. A copyright owner might gain some promise with regard to “use” in exchange for permission in a right the copyright owner does have–so, “You promise to pay me a royalty of 15% of adjusted gross receipts after your usual clever deductions AND not to use pages of my book when you run out of toilet paper, and in exchange I permit you to publish my book in a hardcover edition.”
But it’s not the case that the federal regulations establishing a grant funding agreement provide any mandate for a grantee such as a university to take copyright ownership of anything. Whatever compliance issues there may be, they don’t require taking ownership of copyright works or data. The copyright and data clauses (and Bayh-Dole’s standard patent rights clause) in the regulation are subordinate to a general statement pertaining to intangible property (2 CFR 200.315(a)):
Title to intangible property (see § 200.59 Intangible property) acquired under a Federal award vests upon acquisition in the non-Federal entity.
Intangible property is defined to mean “property having no physical existence.” Instances cited are “trademarks, copyrights, patents and patent applications.” I have highlighted the key distinction–between vesting of title and acquisition. The question that the distinction settles is whether an entity acquires intangible property on behalf of the federal government–in which case the entity passes title to the government–or on its own behalf. The regulation makes clear that title “vests” with the entity when the entity acquires a thing.
This is parallel with the requirements for equipment, which is where, in the dark history of federal contracting, these ideas were thrashed out. Early on, the idea was that federal money was held in trust by grantees, and so when they purchased equipment of any substance, they were in fact purchasing that equipment on behalf of the federal government, which permitted the use of the equipment in the immediate grant and then might expect to have the equipment, too, delivered to the government, since the government had done the paying and therefore the purchasing. This thinking was gradually unwound. And once unwound, for better or worse, was applied as well to intangible property.
There is no obligation in the regulation, however, for the entity to acquire intangible property. If it does, then title vests too. But there’s more, and this is the important point that the article skips over. As a condition of the vesting, the entity is restricted in the use that may be made of the acquired intangible property:
The non-Federal entity must use that property for the originally-authorized purpose, and must not encumber the property without approval of the Federal awarding agency.
Two prongs. First, the property must be used for the purpose established in the grant proposal–again, parallel with equipment. And then the fun part: “and must not encumber the property” without federal agency approval. What the heck? This wording is not in Circular A-110 at this point. It occurs with regard to real property at __.32. If real property is encumbered, then someone else has a claim on it–financial, say, or a right to take all the trees from a property or a right of access or use (such as an easement). The stipulation for real property is that a grantee is not allowed to mortgage or lease the property without federal approval. The reason is that the regulations anticipate that after the grantee has used the property for its original purpose, and for any related purpose after the grant for other federal grants, as approved by the funding agency, the funding agency has the right to direct the disposition of the real property (keep the property but compensate the government; sell the property and return a share of proceeds to the government; or transfer title to the property to the government or to whomever the government designates).
For intangible property, though, especially copyrights and patents, this requirement is strange. For patents, a grantee would not be permitted to assign the patent or even exclusively license rights (that would encumber the patent, so to speak) until the patent had been used for the purposes of the grant (which raises really interesting–and important–questions about how a patent might be used when the purpose of the grant is not to try to make money by transferring monopoly patent positions to company “partners,” nor to engage speculators in raising money to develop commercial products, nor to troll industry for infringement). In practice, university administrators selectively ignore this regulation.
For copyrights, there’s a similar analysis and similar ignoring of the regulation. Ignoring a regulation is rather easy, and one might think it decomplicates the situation. The regulation, however, includes its own disposition of the copyright. There are two parts, one here and one in the next section. Here, the federal government insists on a right to use the work subject to copyright in any “otherwise” way and authorize others to do so as well. That splits the copyright into a private right and a public right–just as Bayh-Dole does for inventions. The public right is outright–there’s no additional formalities. The regulation does not even require the grantee-entity to issue a paper license. The government simply “reserves” the right. Fun fantasy writing by government officials. Why couldn’t Bayh-Dole use this same approach–just make up the idea that the government “reserves” its license and therefore doesn’t have to bother with universities sending in paper notes confirming the license? Ah, because Bayh-Dole is not about innovation; it’s about bureaucracy! Almost none of the mechanisms of Bayh-Dole have anything to do with innovation. Those that did were amended out of the law starting in 1984.
Once the copyright has been split between private exploitation (controlled by the grantee-entity) and public exploitation (controlled by the federal government without further obligation to anyone), the regulation then imposes a public covenant on the use of the copyright on the private side–2 CFR 200.316:
Real property, equipment, and intangible property, that are acquired or improved with a Federal award must be held in trust by the non-Federal entity as trustee for the beneficiaries of the project or program under which the property was acquired or improved.
The grantee-entity must use the copyrights and patents not as owner but as trustee. The regulation serves to shift the presumption of benefit from the government (as sponsor expecting deliverables) to the public purpose of the government’s support–for the “beneficiaries of the project or program.” For inventions, those beneficiaries would be the people and organizations identified in the funding call, the grant proposal, in the communications between the investigators (and their host institution) and the funding agency, and/or the funding agreement. The beneficiaries, unless expressly called out, would not be the host institution, speculative investors, or companies looking to sell product. If a university acting as a trustee licensed a company to sell product based on a patent acquired in the course of a federally funded project, the university would have to ensure that the beneficiaries of the project obtained the primary benefit–not the university, not the inventors, not the company. But that would mean price controls, and university administrators ignore such things, just as they ignore 2 CFR 200.316.
More from our problematic article:
Under awards whose primary purpose is educational, these data rights extend to fellows, students, and trainees as well.
Bayh-Dole expressly excludes such awards from its scope (see 35 USC 212). So Bayh-Dole expressly has nothing to do with awards “whose primary purpose is educational.”
But now our article is talking about data rights, not copyrights, and certainly not inventions. Data rights are handled by 2 CFR 200.315(d):
The Federal Government has the right to:
(1) Obtain, reproduce, publish, or otherwise use the data produced under a Federal award; and
(2) Authorize others to receive, reproduce, publish, or otherwise use such data for Federal purposes.
There’s also a section on dealing with data requests under FOIA in 315(e). The basic gesture is that the government may do what it otherwise will with data subject to claims of trade secret or private confidentiality just as it may with works subject to copyright. The FOIA section limits what data the government must withhold from public disclosure. It would be interesting to see any federal educational grant include data right claims of any sort. Pell Grants, for instance, don’t have any data rights requirements. I can’t think of a federal educational grant program that does have a data rights clause.