Let’s look back in time, do a “thought experience”. That is, experience thoughts that might help clarify things that don’t appear all that clear for folks in the doing.
In your thoughts, imagine a university faculty member years ago, say, 1975, wants to do research with the US government under a standard grant in which the government claims title to any inventions. The faculty member writes a proposal, and the grant goes to the university. The faculty member makes an invention, reports it to the government, and the government requests title. The faculty member signs a document that conveys title to the government. It’s a standard Patent Act situation. The government claims title, but it does so by the normal means, by asking for it as an obligation of the federal contract or terms of award under which the funding is made available. There is no automagicality. The government when it requests title expects to get it, but it doesn’t pass a law that simply takes title outright, but rather goes through the effort to construe title as a deliverable under the contract. Think of it as a form of respect for due process, and for the social constraints of the transaction.
It might be more efficient to have a law that just strips title and saves everyone the formality. But it wouldn’t respect the transaction. And it would place way more power in the government. And it would appear to violate the US Constitution, which gives the government the right to reserve patent rights for inventors, not for the government. So the government gets its rights by an assignment from individuals, because that’s how the whole patent thing is set up. One can be efficient to the point of destroying the social structure on which everything is founded. Yes, you get a new social structure, but it’s more like fascism, and cedes way too much to a central power as all knowing so better able to manage people’s lives than the people themselves. Sounds sort of good, on paper, if you disregard where the expertise lies in research and innovation systems–it is often at the edges. When folks start talking efficiency I pay attention. It can be good, but they had better have a clear idea what they are saying is wasteful. If the expertise at the edges is the waste, then I’m going to be skeptical.
Back to our historical thought experience. Federal funding, an invention, a request, an assignment from the inventor. There’s no reason for the inventor to assign to the university which then assigns to the government. The university could introduce this step, but it wasn’t needed, Back Then.
Now let’s vary things a bit. The faculty member runs an open lab, typical of university research. That openness kicks up a discussion with a scientist a company and together the faculty member and company scientist invent something within the scope of the federal grant. The company scientist is not paid under the grant and has an obligation to assign patent rights to the company. It’s just a chat at a conference that hits on something. Imagine, then, that there are two co-inventors.
The faculty member reports the invention to the government and the government as usual asks for title. Here’s the rub: the faculty member assigns the rights the faculty member has—an undivided interest in the invention. But the faculty member does not have to assign the company scientist’s rights (or break the company’s interest in the scientist’s rights). That’s not part of the bargain. Nor is it part of the bargain that the faculty member not consult with anyone in industry or run a closed lab or force everyone to agree to the same government claims on title simply for talking with the lab about their work. None of that. Not there. The obligation is not to assign all right to title in the invention to the government, but only the right in title that one has. That’s the deliverable under the grant. That’s what the government requests.
It is a big step to go from the personal obligation of a university investigator to arguing that the investigator had to secure all rights to any invention, that the purpose of the government claim was to have exclusive right to the invention. It was not that way for most government grants to universities (some were different, where nuclear weapons or propulsion systems were involved and the labs were closed and disclosure was not permitted).
Perhaps it’s clear now—the right to title in a federal grant is a personal obligation arising from a personal patent interest, a special performance tied to the rights one has, not a demand that in its potential implications forces the lab to become secretive, bureaucratic, and anal retentive, just so all research inventions made have no external parties holding an interest. Nor is there a mandate for university administrators to step in and force the lab into compliance with such an inferred demand, by preventing publication, requiring subcontracting for every interaction, and demanding confidentiality agreements and commitments to assign patent rights for every visitor or communicant with the lab.
Things change when the university wants to take ownership of title, and the government is willing to allow this. What the university then bargains for, with the government, is the title that the government otherwise would receive. That is, the title that the faculty inventor is prepared to assign as an obligation of the award. That’s an undivided interest in the invention, but not necessarily the entire title to the invention, because the faculty inventor does not control the interests of any inventors external to the award. What the university is able to negotiate is the redirection of this assignment—rather than to the government, to the university, or to a research foundation working with the university and the inventor.