In 2013, David Schwartz wrote an article posted at Tech Transfer Central’s “Technology Transfer University Reporter” about the “substantially manufactured” requirement in Bayh-Dole (35 USC 204). Schwartz reports advice offered by a workshop panel on handling NIH waiver requests, but his sources don’t do so well with Bayh-Dole. Here we will attempt to straighten things out.
The “substantially manufactured” requirement is part of what, by its own assertion of precedence, is the most important provision in federal patent policy. Thus, we might expect commentaries on the provision to be directed at how to build up American manufacturing capabilities when licensing exclusively, but this is not the case with the discussion presented in Schwartz’s article. Instead, the focus is on how to get that waiver and so avoid the obligation by engaging the NIH bureaucracy.
We have discussed section 204 recently. While set out as important, it turns out that the actual provision is designed to be hopelessly weak, to the point of being useless in developing American industry. At exactly those points where it might be strong, it wilts. It makes big fuss over an implausibly narrow demand on licensing behavior, and within that demand it sets out an ambiguous standard, and then authorizes federal agencies to waive the requirement at will, with necessary fussing over paperwork, while placing enforcement in the unworkable march-in provisions, where the logic of compelling non-exclusive licenses defeats the premise of the provision.
Of course, march-in has never been successfully used for any reason, let alone to enforce the substantially manufactured provision, so in the big scheme of things, complying with section 204 simply doesn’t matter. What matters is not making a show of not complying. That’s the essence of Bayh-Dole practice. Beyond disclosing subject inventions, notifying the government with regard to electing to retain title, placing a government rights notice in patent applications, and responding to invention utilization requests, there’s next to nothing that federal agencies care enough about to review, let alone enforce.
Let’s work through Schartz’s article for what it teaches us about the state of advice on section 204. It is worth pointing out that in the quoted material below the requirement at 35 USC 204 is “Preference for United States industry”–not “Preference to [sic] U.S. manufacturing.” Schwartz doesn’t get the heading accurately. That’s not the best indication of care. Here’s our opening “answer”:
When federal funding is involved, the “preference to U.S. manufacturing” is a requirement (See 35 U.S.C. 204 of the Bayh-Dole Act below).
The “preference” becomes a requirement only when a small business or nonprofit obtains ownership of a subject invention. If a small business or nonprofit does not obtain ownership, the requirement does not operate. Federal funding is not the threshold, then–obtaining title to an invention made in the performance of work under a funding agreement by a small company or nonprofit is. The distinction matters. The invention might not have been “funded” with federal money but still may have been made in the performance of work under the funding agreement. As the implementing regulations put it at 37 CFR 401.1, “Separate accounting for the two funds used to support the project . . . is not a determining factor.”
The commentary that follows Schwartz’s statement of the requirement lays out the argument for why the requirement is silly:
Since manufacturing activities in industries such as electronics, medical devices and consumer products are mostly conducted outside of the U.S., obtaining a manufacturing waiver is an attractive route for exclusive licensees and can only be obtained if U.S. manufacturing is not commercially feasible or if the university is unable to find a licensee that can — or will — engage in U.S. manufacturing.
The purpose of section 204 is to provide specific guidance on the use the patent system “to promote the commercialization and public availability of inventions made in the United States by United States industry and labor” (35 USC 200). Thus, waiver would mean giving up on this fundamental policy and objective of Bayh-Dole. The policy objective of the law is rather to encourage manufacturing activities inside the United States. Stating as a premise that those activities are “mostly” not in the United States merely perpetuates the idea that one should not manage one’s use of patents on subject inventions to meet the objective of the law, but rather be used to justify avoiding the objective of the law. If one were attempting to meet the objective of Bayh-Dole, this is where the advice would be directed toward providing US manufacturers with the most favorable terms possible–such as, say, a royalty-free license. I know, I can almost see your shocked faces.
Section 204 concerns only a specific sort of exclusive license–an exclusive license to use or to sell. By this enumeration, what section 204 is also saying is that an owner of a patent on a subject invention must grant at least a non-exclusive license for someone to make the invention in the United States, unless that requirement is waived by the federal agency. Without that license to make, or a waiver, a patent owner cannot grant anyone an exclusive license to use or to sell in the United States. Thus, the challenge to the patent owner–as an objective of the law, and the most important objective at that–is to find or create United States manufacturing capability for the subject invention. Schwartz, however, recommends a different approach, which we will get to shortly.
Schwartz represents the waiver approach as an “attractive route for exclusive licensees”–but later we discover that it is the patent owner that must seek the waiver. This becomes a muddle that we can sort out as well.
Here’s the first question and part of the answer:
Who should apply for a manufacturing waiver?
Let’s be as clear as possible. The waiver provided in section 204 is to obtaining the exclusive licensee’s agreement to substantially manufacture in the United States. Here’s the bit from 204:
However, in individual cases, the requirement for such an agreement may be waived by the Federal agency under whose funding agreement the invention was made . . . .
The waiver is not for US manufacturing, but for having an agreement regarding United States manufacturing before granting an exclusive license to use or sell in the United States:
no small business firm or nonprofit organization . . . shall grant to any person the exclusive right to use or sell any subject invention in the United States unless such person agrees . . .
Agreement to use United States manufactured products is a precondition of granting the exclusive license to use or sell, unless agreement is waived.
So what’s the difference? It is the patent owner who must grant at least a non-exclusive license to at least one US manufacturer (which may be the exclusive licensee for the right to use or sell). The waiver is directed at not having to obtain an agreement that the licensee of exclusive right will source American-made product. What the exclusive licensee actually does is immaterial–section 204 is about obtaining an agreement or a waiver, not about requiring, as a matter of patent law, that federally supported inventions in the hands of small companies and nonprofits as an asset to license exclusively must be produced by at least one manufacturer located in the United States. Instead, section 204 is about paperwork–agreement paperwork or waiver paperwork. The thrust of Schwartz’s discussion is that waiver paperwork can be more attractive. There’s some sense to that–but there’s also a degree of risk that Schwartz does not acknowledge, though we should add that all risk with regard to section 204 is idle bureaucratic fantasy except when one flaunts non-compliance.
Here’s what Schwartz writes:
The licensee, since they need it for their business purposes, right? Wrong. 35 USC section 204 of the Bayh- Dole Act is clear that the tech transfer office should be the one to submit the application since it relates to the licensing and historical efforts related to that.
Section 204 has absolutely nothing to say about who submits requests for a waiver. All 204 does is stipulate that an exclusive licensee must agree and that a federal agency may waive upon a “showing” by the patent owner that they tried and failed or they didn’t try at all. Even the NIH form is silent about who should click the “submit” button. The instructions say “grantee organizations and contractors” should “provide information and justification.” But 204 includes as well “assignees”–and so clearly even the NIH hasn’t got things right.
Now here’s the thing. If the owner of a subject invention grants an exclusive license to all the substantial rights in the invention–to make, use, and sell–then the exclusive license also operates as an assignment of the invention. The licensee is also the assignee. This is a tough nut for university folks to swallow (hint: shell and chew the nut first). They mistake an exclusive patent license for an exclusive license to rights in the invention. Thus, they insist that they have not assigned formal title to the patent, which they haven’t, but they have assigned the underlying invention. And universities do this all the time, in defiance or ignorance of the standard patent rights clause (37 CFR 401.14(a)(k)(1)).
Assignees need a waiver only if they in turn intend to grant an exclusive license to use or sell in the United States. Thus, from the outset is critical to point out that if the exclusive license is to make, use, and sell in the United States, no agreement is required by 204, and no waiver of that not-required agreement should then appear attractive. If the assignee of the invention in the United States wants to manufacture in another country, the issue is whether the owner of the invention also has a patent in that other country. If so, then the assignee needs a license for that country. Nothing in 204 precludes the patent owner from granting that license. Of course, if the original owner of the invention conveys worldwide rights, then we are done–the assignee can do what it wants, subject to whatever other patents may be out there that would block manufacturing in one country or another. You can see how narrow 204 is. A pebble in the stream.
When should one apply?
Only apply for a waiver at the request of a licensee who wants an exclusive license to your intellectual property that was federally funded and who wants to manufacture substantially outside of the U.S.
Section 204 is clear–a subject invention owner cannot grant an exclusive license to use or sell without obtaining an agreement to use substantially American-made product or a waiver for this agreement. Again, this situation only arises when the exclusive license on offer is to use or sell (or, both). If the exclusive license is to make, use, and sell, then the offer is to assign, and assignees are not required to agree to anything as a condition of the transaction. For nonprofits, the standard patent rights clause requires federal agency approval for assignments of subject inventions other than to an organization that has as a primary function the management of inventions. It’s entirely up in the air whether a company that manages its own patent portfolio meets this standard. Arguably it does. Universities certainly conduct their practice as if they do not have to obtain agency approval for such assignments.
We might, then, refine Schwartz’s text: Only apply for a waiver to the agreement to manufacture substantially in the United States before granting an exclusive license to a subject invention to use or to sell (or both) only when the potential licensee declines to make that agreement and you have no better mutual alternative.
It is not a matter of a licensee requesting anything–it is a matter of a potential licensee refusing something specific. There are alternatives–assign the invention instead of exclusively licensing only the sell or use rights. Use the exclusive license instrument to mask the assignment. No one will worry it. Or grant only a non-exclusive license to use or sell, and don’t grant more such licenses for a time, perhaps a long time. That’s a sole license, not an exclusive license (the owner of the invention has not given up its own right to use or sell), so again one is entirely outside the scope of 204. Or for appearances grant a second non-exclusive license to a company that is unlikely to manufacture in any volume–say, one of those paper shell startup companies that universities create to game economic development statistics and fool state legislatures to allocate more money for university use. The company doesn’t have to operate–it just holds a non-exclusive license. Make the license non-assignable. No one will worry it, even if it smacks of the usual university scamming.