The meaning of “promote” in Bayh-Dole gets dealt with, though indirectly, in Public Citizen v. NIH (2002). In this case, Public Citizen, an advocacy group, sued the NIH in an effort to get the NIH to disclose certain particulars of an exclusive license the NIH had granted to Johnson & Johnson. Public Citizen wanted to know how much money the NIH had received under the license and the royalty rate the NIH had negotiated. The NIH claimed the information was protected by FOIA. The court ruled in favor of the NIH–that releasing such information would have a detrimental effect on the NIH’s licensing program.
The upshot–no public accountability for federal agencies when they deal in patents. Whatever public agencies require (or give up) in patent deals must be understood to be information coming from whomever the agency is dealing with, and to release that information–a deal whose terms are entirely in the federal agency’s control–would damage the federal agency’s ability to do more such deals. “If we cannot do secret deals, then our licensing program based on secret deals will be damaged.” Well, well, tautologies abound. Meanwhile, the public loses oversight over matters of public importance. If Bayh-Dole anticipates federal licensing in the public interest, apparently it also asserts that the public is too stupid and dangerous to be informed of what the federal government is doing in its interest. Fundamental rule of Bayh-Dole: Sucks to be the public.
For all that, let’s look at how a sucks-to-be-the-public judge bounces around Bayh-Dole to help the NIH apply its suckiness.
There is plenty of strangeness in the ruling. Bayh-Dole is not the only bit that comes in for odd treatment. For example, the Court rules that a royalty rate, though it is negotiated, necessarily “comes from a person”–that is, the government must be thought to obtain the royalty rate from a company, and so according to this logic, the royalty rate is FOIA protected. That is a heck of a mental gyration. Consider the implication: *any* terms of *any* deal the federal government might make with *any* non-governmental entity could then be placed under FOIA with the argument that whatever the federal government offers by way of terms is secret because once the deal is signed, those terms must be construed as coming from the other entity, and therefore must be kept from disclosure. Now, just on general principles, the terms that anyone doing business with the federal government this way would want to keep secret would be just those terms that might come in for criticism from the public.
It is the government that has the power to set the royalty rate. Any company taking a license to an invention necessarily must accept the government’s royalty rate. Even if a company makes a counter offer in negotiations, in the end it is the government’s decision as holder of the patent whether to grant the license or not, and on what terms.
The NIH argued that disclosing the requested information would result, according to the Court’s paraphrase, in “substantial competitive harm and the impairment of the effectiveness of its licensing program.” The Court remarks on the many letters submitted by the NIH from universities and companies that revealing royalty rates would cause competitive harm and disparages Public Citizen for offering no such “evidence” that revealing royalty rates would not cause harm. Yet the letters themselves do not provide any evidence of harm. They merely state the assertions of their writers regarding harm. But these letters’ assertions become facts, according to the Court: royalty rates are usually kept secret, so the NIH must keep its royalty rates secret, too, because the royalty it charges is, in effect, the royalty rate of the company, and the company wants to keep that rate secret.
The Court discusses then how a competitor might come to understand the value that a company places on a technology by the royalty rate the company is willing to pay:
The Court notes that if a competitor realized information about how much a licensee was paying for a royalty, it would immediately know the value the firm placed on that particular technology. This knowledge, in turn, would certainly help competitors know which technologies and which areas of the market the licensee found particularly fruitful.
In the context of an exclusive license, this is just silliness. The fact of the exclusive license is the thing that reveals a company’s valuation of the technology. Perhaps the structure of the license–that there is a royalty rate and not a fixed sum or equity–reveals more about the value of the licensed technology, or at least the risk involved. A running royalty on sales shifts the risk to the licensor. A cash payment upfront shifts more risk to the licensee, but also suggests the licensee has great confidence in making back the value of the payment, and much more. But all this is beside the point, since the company taking the exclusive license does not have any competitors with regard to the licensed class of compounds from which the drug is produced.
The NIH and the company claimed that if one knew the royalty rate and the income received, one could determine how much money the company was making from sales of the licensed invention, and could surmise how important the invention was to the company’s business. It is easy to make it uncertain, if not impossible, to determine total sales from a royalty rate and income. A deal may involve upfront fees and license minimum payments and milestone payments or cash payments indexed to share value–any deal structured this way will make it difficult to determine just from a royalty rate and income received what total sales have been. Furthermore, the importance of a transaction to a company may be found in the scope of the license and how much money is guaranteed payment.
A running royalty on sales, while it might (rarely) result in substantial income, is a rather poor way of determining value of an invention to a company. There are times when a company may allow a high royalty rate because they never intend to pay anything based on that rate–they will pay another way, such as through a sublicensee (and the sublicensing royalty rate), or through a license buyout, or they will renegotiate the license later, or they will dedicate the invention to a standard or in cross-licensing and not earn anything from direct sales.
A low royalty rate, by contrast, may mean that the company does not value the invention all that much and has misled the federal agency to agree that the company’s valuation should be the value of the invention to the public. A low royalty rate, too, may mean that the company has low margins already and cannot take the deal otherwise. And it is odd that anyone thinks that there is only one royalty rate in play–there may be different rates for different products or product areas, for different jurisdictions, and for different forms of use or sale, and those rates may change over time or as various objectives are met.
A higher than average royalty rate (one would have to know the average royalty rate a federal agency obtains in its exclusive licenses) might mean that the company taking the license does not care about the royalty rate. It may be that the company mostly wants to prevent others from gaining access to the invention until the company’s own product position has resolved–then the company may terminate the license or sell off the division or subsidiary holding the license, and not pay a dime as a running royalty on sales. These sorts of things happen all the time with startups. It’s just nonsense to assume that a company agrees to a higher royalty rate because the invention is that much more valuable to the company. Many other factors come into play.
For that matter, only a very silly person would think that a patent license arrangement always takes the form of a running royalty on sales. You know–how AUTM calculates jobs by assuming that all reported income arises from a running royalty on sales, multiplies by the inverse of what they say is a typical royalty rate to create a figure AUTM calls “total sales,” and then divides by what they say is a typical salary and gets–er–a figure AUTM calls “jobs.” Then AUTM invites people to treat its goofball figures as facts. AUTM assumes its readers are all very silly persons–and some clearly are!
You get the idea. The NIH provides the judge with plenty of “information” by which the judge can wallow around in whatever fallacies she wants so long as they sound reasonable to her in rationalizing the idea that revealing royalty rates necessarily must cause competitive harm to a company taking an exclusive patent license from the federal government.
Exclusive patent licenses, if they convey all substantial rights in an invention, function as assignments of the invention. When a federal agency assigns an invention to a private person, the federal agency in effect re-issues the patent. A conventional patent issues with the conditions set forth in conventional (non-Bayh-Dole) patent law. An invention owned by the federal government, however, is subject as well to Bayh-Dole’s addition to federal patent law, including 35 USC 200 and the regulations with regard to licensing authorized by 35 USC 206 and 208.
A federal assignment of an invention along with an exclusive license in the patent amounts to a re-issue of the patent monopoly, but now under the terms not only of conventional patent law but also Bayh-Dole’s requirements at 35 USC 207 and 209. Those requirements then become the subject of the implementing regulations at 37 CFR 404, and those regulations then are interpreted for any given transaction by the federal agency. The public conditions that attend to the federal agency’s reissue of the patent then are contained in the license agreement. If the public does not have access to the terms of this agreement, Bayh-Dole has created a secret part of federal patent law, in which the public does not know the property rights under which a patent has been issued to a private party.
As the Supreme Court in Stanford v Roche commented with regard to the lack of third party rights in Bayh-Dole, saved only because Bayh-Dole dealt (for contractors) only with a priority of rights between the contractor and federal agency once a contractor had acquired rights: deeply troubling. One might find in the Supreme Court’s whisper an opening to find that Bayh-Dole is–if the federal government asserts this Public Citizen decision as a proper reading of law–unconstitutional in denying third parties access to the information and right of appeal that due process otherwise allows.