An alternative national science foundation 11/17
How would you design it?
Tyler Cowen points to a report entitled Increasing Politicization and Homogeneity in Scientific Funding: An Analysis of NSF Grants, 1990-2020. I am sure that Tyler would agree that the time is ripe for some institutional innovation in science funding.
The National Science Foundation disburses funds for programs and projects intended to promote scientific research. It solicits funds from Congress, which has a pool of resources obtained coercively through taxation.
An Alternative National Science Foundation (ANSF) would disburse funds for programs and projects intended to promote scientific research. It would solicit funds from donors on a voluntary basis.
Why do we need an ANSF? I think that the NSF is too oriented toward incumbents, and incumbents are standing in the way of progress. It is hard for young minds and heterodox thinkers to get grants. On top of that, we are getting the craze for Diversity, Equity, and Inclusion. Before science falls apart completely, we need new institutions.
How would you design the ANSF so that it provides a value proposition to donors? Think of the ANSF as an intermediary acting on behalf of donors in disbursing funds to promote science. If donors could disburse funds more efficiently by themselves without an intermediary, then they do not need an ANSF. If the most efficient intermediary is the legacy NSF, then donors should just try to give to the NSF somehow, perhaps by providing supplemental matching grants of some sort.
Imagine writing a white paper describing the ANSF. Think of the white paper as containing a proposal to take to donors. Some of the questions that the white paper should answer:
How will the ANSF set priorities? What role will donors play? Will the ANSF collect and disburse funds, or will it only play an advisory role?
How will the ANSF evaluate the results of its process, and how will it go about improving its process?
What will be the organizational structure of the ANSF? What roles will be filled by permanent staff? By consultants? How should different branches of the organization be tailored to different scientific needs? To what extent should it be a hierarchical organization or instead a network of disparate organizations?
How could one prevent or overcome a tendency for a single ANSF to become corrupt and stale as time passes? (That is what I think has happened to the NSF.)
What can be learned from existing mechanisms for promoting science? venture capital, in biotech or alternative energy, for example; grants funneled through individual polymaths, like Scott Alexander or Tyler Cowen; special-purpose research foundations, such as the research arm of the Crohn’s and Colitis Foundation; other charitable intermediaries, such as GoFundMe or GiveWell or Open Philanthropy
How could you start the ANSF with a prototype, and what would be the expansion plan?
I am thinking of running a contest in which I publish white paper proposals here, with the winners getting cash prizes. If you are interested in trying to write a white paper, let me know at arnoldsk@us.net
“I think that the NSF is too oriented toward incumbents, and incumbents are standing in the way of progress. It is hard for young minds and heterodox thinkers to get grants.”
Indeed. According to the Browser, the NIH awards 7 times more funding to scientists >65, than to those 35 and younger.
And here is a new organization for building institutions of basic science I learned about today, also affiliated with Tyler Cowen:
https://newscience.org/
A good NSF - and even a good ANSF - would not fund research activities per se. Instead, it would select from among published, peer-reviewed reports deemed significant in the national interest (by specific agencies or other gov't authorities) and use a lottery system to assign them to one or more registered independent verification and validation laboratories for fully funded replication. It would make the results public.