A recent search at RE was looking for “preference for manufacturing in US under Bayh-Dole.” There’s a series of articles here on 35 USC 204. There’s also discussion of the related march-in provision at 35 USC 203(a)(4) and the broader objective at 35 USC 200.
But let’s save you all some time.
There is no preference for US manufacturing under Bayh-Dole.
Section 204 is titled “Preference for United States industry.” Sounds nice. 204 says it takes precedence over any thing else in Bayh-Dole, and Bayh-Dole says it takes precedence over anything that has gone before but for Stevenson-Wydler. Still sounds nice. But it’s an empty gesture.
Section 204 is more accurately a preference for United States manufacture in limited exclusive licensing to use or to sell in the United States. If a federal contractor owning a subject invention does not license, or licenses non-exclusively, there’s no preference requirement. A federal contractor can make product based on a subject invention anywhere in the world and and be the only source to sell it in the United States, as far as section 204 is concerned. If a federal contractor assigns a subject invention rather than licensing it, there’s no preference requirement on the assignee–only on the assignee’s exclusive licensee to use or to sell–but an assignee may have no need to license anything to anyone. Just make product or have it made, and sell it. No United States manufacturing obligation.
Furthermore, even if, in the rare instance, a federal contractor would grant an exclusive license to use or to sell, section 204 gives the federal agency the right to waive the US manufacturing preference if the contractor tried and could not get an exclusive licensee to accept US manufacturing or didn’t bother to try because it didn’t seem that there would be US manufacturing. And on top of that, Bayh-Dole is structured to require section 204 to go into a patent rights clause that becomes part of each federal contract for research or development. Nothing in Bayh-Dole requires any federal agency to enforce any part of that patent rights clause. So a federal agency doesn’t even have to waive the US manufacturing requirement in those rare licensing cases–it can just ignore the whole thing and choose not to march-in if it ever came to that.
So section 204 is an empty gesture. It’s not even a “preference” for US manufacturing. It’s a gesture at a preference. If you feel the need to repeal Bayh-Dole, here is a great starting point–the central premise for the law as introduced by Senator Bayh, that Bayh-Dole would stimulate US manufacturing and return the country to global technology leadership. Eliminate section 204 and Bayh-Dole reduces to superseding Congressional conditions on specially allocated research and development funding and executive branch research contracting policy that made nonprofits and contract research organizations make a persuasive case that their holding exclusionary rights in federally supported inventions was better for the public than open access–and in particular, and most significantly, and expressly with regard to public health inventions.
Section 204 is beyond empty, really. The only federal contractors for which exclusive licensing might matter are the nonprofits and contract research organizations–organizations without manufacturing capabilities for the most part and not in the business of selling technology products. But a requirement based on granting an exclusive right to use or sell anticipates a company that can manufacture but which desires another company to do the using or selling. Again: a federal contractor that’s a manufacturer acquires an invention made under contract. The manufacturer grants an exclusive license to sell product to another company. Per 204, the manufacturer requires the exclusive licensee to acquire its product from the manufacturer, and the manufacturer, per 204, must make that product in the United States (“manufacture substantially”–meaning either most products made in the United States and some made in other countries or most of each product made in the United States and some parts of each product may be made elsewhere). Or, the manufacturer doesn’t care to manufacture and grants non-exclusive licenses to make. Now the exclusive licensee to sell may acquire product from any of those manufacturers, but mostly from manufacturers making in the United States.
These are arcane possibilities. They just don’t happen, though they are theoretically possible. The moment things go non-exclusive, then there’s no United States preference required. And if things are non-exclusive, then Bayh-Dole reduces to allowing federal contractors to create a bureaucratic patent paywall between federal funding and public access–essentially cutting profit-takers in on what’s discovered in research pitched as being in the public interest. If things are full-on exclusive–granting all substantial rights in an invention–to use, to sell, and to make–then the transaction is not merely an exclusive license–it’s an assignment of the subject invention and a conveyance of the right to enforce a patent on that invention. The United States manufacturing preference is specific to limited exclusive licensing–to use or to sell. The preference does not operate with assignments unless the assignee has acquired the invention only to turn around and engage in limited exclusive licensing.
An owner of a subject invention can easily avoid the section 204 requirement if it must license anything. Just grant two licenses, not one. Or even, just grant a single non-exclusive license. Just the possibility that there could be other non-exclusive licenses is sufficient to neuter section 204’s requirements. Not an exclusive license. No 204 issues.
Section 204 also appears to try to damp down the tying of a patent license to acquisition of the licensor’s product. Normally, if a patent owner ties the patent license to a requirement that the licensee must also buy the patent owner’s product, then there’s a prospect for patent misuse, if not antitrust. Bayh-Dole stipulates that a patent owner must make such a tie or the patent owner must license the right to make to at least one United States manufacturer. To avoid making an assignment of the invention, the patent owner must then also grant the right to sell to at least one other company–not the manufacturer. In other words, if there’s to be a license, then the patent right must be broken up. If broken up into non-exclusive bits, then fine–there’s no reason for Bayh-Dole, really. If broken up into exclusive bits for the United States, then the manufacturing right has to include at least one United States manufacturer, unless the 204 requirement is waived, or the federal agency refuses to march-in, or the federal agency ignores enforcement of 204.
Bayh-Dole is directed at nonprofits. Small businesses are thrown in but there was little pressing demand for small businesses at the time–the policy need was different–to break up the economic power of the existing United States companies. Think about it. Bayh-Dole’s premise–on the Senate floor and in section 200–is to restore United States global leadership in technology manufacturing. But it does so by making it appear that small businesses are to get patent ownership preferences, especially in the area of public health, so that the United States leading companies may be excluded from access to such inventions made under federal contracts. Does that work for you? It may be that there are good reasons to spread federal contracting around so that some few big, effective, talented companies don’t gobble up all the opportunity and federal money and getting bigger, more effective, and more talented, but it would take a pretty subtle logic to argue that weakening big, effective, talented companies is the inspired genius way to restore United States global leadership in technology.
If a nonprofit acquires a subject invention, then it pretty much has to license to meet Bayh-Dole objectives. Without Bayh-Dole, the nonprofit would not need to license–the federal agency would acquire the invention and provide open access (no patent or royalty-free non-exclusive license, often with such indifference that no formal paper license was bothered with). The nonprofit would have access to the invention, as would researchers everywhere, as would companies everywhere. (And keep in mind–if there’s only a U.S. patent, then everyone everywhere else other than the United States has open access–only the United States is burdened with the overhead of dealing with a patent that excludes others from practicing the invention). If the nonprofit licenses non-exclusively, then what’s the point of Bayh-Dole? It’s just a profit-taking diversion of opportunity, so nonprofits get more money (if even that) for research. Getting more money for research might be a good thing. But–to ask a policy question–is it worth putting 50,000 inventions made in federally funded work behind patent paywalls in the United States? And is it worth doing that when the money for research that universities get from licensing patents turns out to be pretty lousy money for the most part, especially compared to the federal funding now available.
If the nonprofit licenses exclusively–all substantial rights–then that’s an assignment. No 204 issues. If the nonprofit licenses non-exclusively, then there’s no point to Bayh-Dole. The nonprofit just makes it harder and more expensive for anyone to practice what’s been invented by making everyone accept a 35-page contract that demands payment, audit rights, insurance, notification of change in insurance, and the right to delay assignment or sublicense of the invention. If the nonprofit licenses for limited exclusivity, then 204 applies, but can be waived, or not enforced, or ignored.
To make Bayh-Dole appear to be an effective policy tool, then, the dominant nonprofit licensing practice would be to break up patent rights with limited exclusive licensing. One company does the selling, and another company or companies does the United States manufacturing. If we wanted metrics to see how nonprofits were doing, we would ask:
For how many subject inventions have you successfully broken up patent rights with limited exclusive licenses?
That is, to do Bayh-Dole properly, to make Bayh-Dole seem even remotely relevant, a nonprofit would, for each subject invention, aim to grant at least two licenses–an exclusive one to sell or to use in the United States, and at least one license (sole, exclusive, or non-exclusive) to manufacture in the United States. That would be the package for each subject invention. If there are many such packages, then we might say that Bayh-Dole has been successful. But there aren’t such packages. Nonprofits do not break up patent rights in this way. Nonprofits for the most part do not grant exclusive licenses just to sell or to use. Instead, they assign subject inventions but label the deals “exclusive license” and then throw in the section 204 language–which doesn’t apply except if the assignee goes off and does an arcane deal in which it grants an exclusive license to use or to sell in the United States.
And as a measure of meeting Bayh-Dole’s express objective to use the patent system to promote free competition and enterprise:
For each subject invention, how many United States manufacturers have you licensed?
Those would be the metrics. They don’t exist. Partly because no one cares to gather them–not universities, not AUTM, not federal agencies. And partly because there’s nothing there. There aren’t any United States manufacturers licensed and connected to limited exclusive licenses to use or to sell. United States companies get assignments from nonprofits (trading on the value of patent rights–taking inventions from inventors and passing them to companies that didn’t invent, for a fee) or non-exclusive licenses (for which Bayh-Dole creates waste effort, for the most part).
So much for section 204 preferences for United States industry in Bayh-Dole. There isn’t any. Consider, instead, section 202’s implicit preference for United States industry. Try working that out.