Here’s a part of Bayh-Dole that’s odd. It is a requirement for a provision in the standard patent rights clause specific to nonprofits, 35 USC 202(c)(7)(C):
(7) In the case of a nonprofit organization …
(C) … a requirement that the balance of any royalties or income earned by the contractor with respect to subject inventions, after payment of expenses (including payments to inventors) incidental to the administration of subject inventions, be utilized for the support of scientific research or education;
There are two things, actually, going on worth being odd about. The first is the question of why there should be a restriction on nonprofit use of licensing royalties from patents on subject inventions (let’s call them “POSI.” There is no comparable restriction for small businesses, nor for large companies (with Reagan’s extension of Bayh-Dole by executive order). So companies, receiving federal support for research, can use money from licensing patent rights (and litigation, and strategic partnerships, and product sales, and whatever else) any way they like, but nonprofits are restricted to “support of scientific research or education.” Why?
Why wouldn’t it be the other way around? Nonprofits, having public charters that qualify them for tax-free status, should have no restrictions (since the restrictions are already built into their charters and into their tax waivers). For-profit companies, enjoying the benefits of federal subsidies, should have some sort of public covenant that limits how income from POSI can be exploited (that is, under what circumstances can a POSI be litigated for infringement, for instance) and restricts how money can be used that arises from that exploitation. Perhaps not all income would go to scientific research or education, but perhaps there’d be a 5% for science kind of thing, in lieu of paying a royalty to the government, say.
But it’s not this way. It’s bass-ackwards of this way. Nonprofits have the restrictions and for-profits can run like the wind. This is one of the ways in which nonprofits have the most restrictive standard patent rights clause in the Bayh-Dole apparatus.
But if we look at the history of the discussion around Bayh-Dole, things start to become clearer–not clear as in “clear water” but clear as in “it’s clear I just drank muddy water.” At least we see what we are up against.
We find some clues regarding the issue of nonprofit treatment of patent income in the run-up to Bayh-Dole. Two sources make helpful contributions. One is Dr. Thomas F. Jones, MIT’s Vice President for Research in 1978. The other is “Tax Aspects of University Patent Policy,” an article from the Journal of College and University Law by Norman Sugarman and Douglas M. Mancino.
Let’s start with Sugarman and Mancino. Norman Sugarman was the expert on nonprofit tax matters. He wrote the law on it, quite actually. Douglas Mancino, at the time of the article, was a recently graduated attorney who now “enjoys a national reputation as one of the country’s top health care and tax-exempt organization lawyers” (according to his bio at Seyarth Shaw LLP). Senator Nelson enters the text of the “Tax Aspects” article into the Congressional Record in the debates over a government-wide IPA program–what would become, two years later, Bayh-Dole.
Sugarman and Mancino identify three areas in which university patent activities might give rise to tax issues: 1) nonprofit status; 2) unrelated business income; and 3) handling royalty payments to inventors. I can’t hope to grapple with all the complexity of tax law–or the intervening years in which things may have changed–but let’s look briefly at “Tax Aspects” to get a feel for why Bayh-Dole has the nonprofit licensing revenues the way it does.
Sugarman and Mancino first discuss university tax-exempt status under IRC 501(c)(3). They point out that tax regulations require that no income from the activities of a nonprofit should “inure to the benefit” of any individual with a “personal and private interest in the activities of the organization.” But what then about university inventors, who may receive a share of royalties based on the patent licensing activities of a nonprofit university licensing program? Here’s their argument:
If the university acquires invention rights from an inventor and agrees to pay “reasonable compensation” for those rights, then whatever the university pays to an inventor is “for the property” the inventor has conveyed to the university. There is no “benefit inuring” to inventors from the licensing activity because the royalty sharing is part of an exchange of patent rights.
Sugarman and Mancino’s argument here depends on a transaction in which an inventor negotiates a deal with the university and “reserves the right” to receive royalties as part of the deal. Thus, royalty payments to an inventor are, in this account, just what’s due the inventor. The university is merely a “conduit.”
Sugarman and Mencino depict the relationship between an inventor and university as one in which the university’s share of royalty income is based on an equitable return for the services it provides in managing the patent work:
The relationship contemplated is one in which the university takes a share of royalty income to account for its services. This is unlike most current university patent policy royalty sharing schedules, which assume university ownership, deny any agency or need to account for services, and allocate a share to inventors as an administrative act of generosity.
Sugarman and Mencino’s description of the relationship between inventor and university is at the heart of the agent approach to invention management. The agent–here, the university–acquires inventions in deals with inventors, and then uses its resources to undertake an effort in the public interest to make the invention available for public use and benefit. In the agent approach, the nonprofit may act as the agent of the inventor or may act as the agent of the public. Typically, when the nonprofit agent was a research foundation, it performed both roles–it was an agent for the inventor acting on behalf of the inventor and an agent for the inventor’s university, acting in the public interest.
These days, most universities now have sworn off the agent model. In its place is an implicit portfolio model in which the university acquires everything it can and seeks to make a profit from a handful of things. Winners pay for losers–and in a portfolio model, there can be many, many losers. Only one big deal every two decades is enough for a portfolio model to appear financially successful. But more importantly, many universities have also sworn off the idea that there is any kind of negotiation between inventors and universities over invention rights.
A misrepresentation of Bayh-Dole had something to do with this move. Bayh-Dole, so it went, vested ownership of inventions in the universities that hosted the federally supported research. (Or, because there wasn’t any consistency in the story, Bayh-Dole gave universities first right to inventions, or right of first refusal, or prevented inventors from assigning rights to anyone else, or required universities to own–none of this was true).
If federal law allowed universities to own inventions made with federal support simply by “electing title,” then there was negotiation over rights and no “consideration” for the transfer of rights. Any payment to inventors was a matter of institutional generosity. Some universities depicted payments as an inducement to inventors to participate in the “technology transfer” program. Even this, however, staled as universities adopted policies that made it compulsory to participate in the program. In these settings, payments took the form of “lessening the pain.”
Bayh-Dole requires federal agencies to require in funding agreements that nonprofits to share royalties with inventors (35 USC 202(c)(7)(B):
a requirement that the contractor share royalties with the inventor;
This requirement is also strange, again, as something spelled out for nonprofits and not for for-profit companies. Perhaps we can now see the answer. The drafters of Bayh-Dole wanted to make it clear that nonprofits could share royalties with inventors and not run afoul of federal tax law. However, the presence of this requirement does not mean that any form of sharing will be acceptable under federal tax law. Sugarman and Mencino envison a negotiated deal in which a university acquires private property and the inventor reserves a financial interest, which the university pays as a cost for acquiring the property. Thus, in Bayh-Dole we find the treatment of payments to inventors treated as an expense “incidental to the administration of subject inventions”:
after payment of expenses (including payments to inventors) incidental to the administration of subject inventions
If we add these two requirements up–share royalties with inventors and deduct payments to inventors as expenses–we start to see an expectation that to do both, and remain a nonprofit, the nonprofit has to negotiate for the personal property of inventors–their invention rights. Assertions in policy that a university will own all inventions as a condition of employment or use of resources do not address, then, either the negotiation under which such ownership comes about or the financial interest that an inventor may reserve.
Most patent policies now (and the administrators interpreting them) make it appear that inventors are compelled to assign to the university as a condition of policy, often without the need for a formal patent agreement, as if a university could by policy create a special new form of common law under which the university, not inventors, owned their inventions. There’s no negotiation, no reservation of a financial interest.
Some state universities go so far as to make it appear to be a personal conflict of interest for an inventor-employee to attempt to negotiate any aspect of assignment of inventions to the university. Such a negotiation would represent an unethical attempt to influence the official actions of the university and reserve to the inventor financial rights that are not allowed. This idea runs exactly opposite to Sugarman and Mencino’s idea that the negotiation is what enables the argument that paying royalty income to an inventor does not create a finding that earnings of the nonprofit have “inured” to the benefit of individuals–the inventors.
Even Sen. Bayh, in his amicus brief to the Supreme Court in Stanford v Roche insisted that there must be some sort of negotiation under which an inventor could reserve a financial share of any licensing income received by the university. It’s just that if Bayh-Dole were to be a vesting statute, then there would be nothing at all to negotiate. Most universities refuse to negotiate strategy, diligence, or the royalty share. They have a royalty schedule and that’s what inventors get–if there is ever anything to get, and the getting is pretty darn sparse.
Back in the day, most university patent policies involved voluntary assignment. Only if someone was expressly hired to invent might the university claim an ownership interest. Otherwise, the concern focused on equitable interest, based on the support above and beyond ordinary that the university may have provided to enable the invention or its development. If a university published a royalty schedule, it was a “take it or leave it” thing–meaning, if one wanted to use university invention management services, then the policy specified the deal. An inventor could leave it. But when administrators moved to expand the scope of what the university claimed (or at least claimed to expand the scope) and made assignment compulsory, they didn’t bother to consider how their actions also changed the status of the royalty schedule. Instead of negotiating, they were “taking.” Anything that came back to inventors was not part of the “taking” deal. Share involved a separate policy of administrative generosity.
We might find that Bayh-Dole, then, in these two requirements–royalty sharing and treating royalty sharing as an expense–implicitly requires universities to adopt a voluntary invention assignment policy, so that there is a legitimate negotiation with the inventor and that any sharing of licensing income with the inventor is a matter of that negotiation and not of a policy handing out inuring treats to individuals.
One can see in the tax problem of how nonprofits share licensing income with inventors a reason why some universities refuse to share income with non-inventors working in the same laboratory–doing that might be seen as income inuring to the benefit of individuals. But such taboo is silly if the university shares as a matter of administrative generosity rather than because there’s consideration for an invention assignment. And of course, if an invention turns out not to be patentable after a licensing deal has been done and income received and shared, then the university is sharing income with inventors who aren’t actually inventors–just like the other lab personnel, and thus we are back to taboo all around.
Sugarman and Mencino then consider whether a patent licensing program is properly related to a university’s “exempt purposes.” They argue that patent licensing is related to research. Here’s the money passage (so to speak):
Sugarman and Mencino connect the dots: universities conduct research to discover and to train, so if a patent licensing program should “encourage and stimulate” research. Then it is “directly related” to the exempt purpose. There it is–royalties should be used for “scientific research or education.”
In the IPA program, we find a similar requirement:
That is, nonprofits should use royalties after costs for exempt purposes in order to avoid problems with their tax status and unrelated income taxes.
Sugarman and Mencion pack this down with a further argument related to how university patent programs should be conducted to meet the requirements of exempt purposes:
Here we have the roots of Rev Proc 2007-47 (and its prior ilk). “Scientific research is an exempt activity if patents resulting from such research are made available to the public on a nondiscriminatory basis.” There are various ways to do that–license non-exclusive, royalty-free. License for a non-discriminatory fee–that is, companies pay the same, whether sponsors of research or not. Or license exclusively based on some premise other than favoritism or prospect for payment, such as a better plan or a finding of better capability than others. Sugarman and Mencino then repeat the choice phrases–“greatest possible public benefit” and “widest possible exploitation” and “place the public benefit over private profit-making.” It’s just that these choice phrases have to be taken as prescriptive rather than emblematic or intentional. That is, a university patent program has to actually do these things, not merely make statements that aspire to these things and turn around and license exclusively to preserve monopolies or worse not license at all and wait until companies appear to be using an invention anyway and then suing them for infringement.
At least at one time, forty or so years ago, a university patent licensing program could run afoul of federal tax law and jeopardize the tax-free standing of a university by managing patents in a discriminatory manner, and not for the greatest possible public benefit or widest circulation. It was possible for a university to place private profit-making over public benefit–and it still is. It’s just that now university administrators equate private profit-making with university benefit, and by extension that university benefit is public benefit. One wonders whether present practice is afoul of federal tax law, or whether we are at the point that the argument goes “we have been out of compliance this way for so long you can’t begin to enforce the law now.”
Sugarman and Mencino then repeat these requirements on university invention management in taking up the issue of unrelated business income. I’ll show you the paragraph, so you can see the repeat:
First, the patent management activities must be bound to the university’s exempt purposes. Second, the income from those activities must be bound to the university’s purposes. For both, the patent management must be connected to research (to stimulate research, to train people in research) and the income from patent management must arise because the patent management seeks public benefit over institutional or personal inventor income. If this is the case, our tax expert authors conclude, “it should be fairly clear that the income derived from the university patent activities contributes importantly to the accomplishment of its exempt purposes other than the need for income.” And in Bayh-Dole, the goal is utilization of inventions arising from federally supported research or development.