Is the NSF Plan Unconstitutional?

The 5th Amendment of the US Constitution includes a guarantee that the government cannot seize private property without making a due compensation. These protections were extended to the states with the 14th Amendment.

So, it’s a bit problematic to read two bullet points of the new National Science Foundation policy released last week:

That all peer-reviewed scholarly publications resulting from NSF-funded research will be made freely available and publicly accessible by default in the NSF Public Access Repository, or NSF-PAR, without embargo or delay.

That persistent identifiers, or PIDs, and other critical metadata associated with peer-reviewed publications and data resulting from NSF-funded research will be collected and made publicly available in NSF-PAR.     

The processes, expenses, and expertise that result in a peer-reviewed publication come from the private sector, so claiming all peer-reviewed scholarly publications is what lawyers would term “a taking,” which requires compensation. Add to this the seemingly new requirement that publishers also provide PIDs and “other critical medadata,” and the scope of the taking has grown — and may grow larger still.

The notion of a taking has been present in all OSTP policies, but with embargoes in place, publishers decided it made sense to cooperate. The most recent OSTP guidance hedges on this whether the taking — or any part of the guidance — is legal via the major qualification, “as appropriate and consistent with applicable law.” The NSF policy makes no such hedge, and presumes that its action is legal.

It’s worth recalling that the OSTP guidance was grilled by Congress in October 2022, and that there are major discontinuities between its saccharine public face and the more sober and considered language used in the actual document.

Maybe NSF policy will come to stand for “not so fast” in the future. After all, this new policy seems to fly in the face of basic protections against uncompensated government taking.