[gajomi]

This is a very good point and I think I would like to refine my comment in response to it.

I think it is important to distinguish between individual vs collective accountability, as well as accountability within a field and the accountability of the field to larger society. My argument for supporting basic research in a field assumes that there is internal accountability within that field and that the research makes sense to those within it even if there is no direct application. In this view accountability of the individual to society is mediated by the accountability of the field at large to society. This is what allows for various sorts of bet hedging (if we think of a field as a firm) to improve the chances of hitting the rare event where a basic result has a big impact. If there is a deterioration of transparency accountability within a field, perhaps due to the mechanisms the OP describes in points (1-7) then (8) does become a problem. So I have perhaps a different understanding of (8) then the OP. I do not argue that some particular flavor of esoteric mathematics be supported because 'it will be useful someday'. Rather I argue that, so long as there is intellectual integrity within mathematics (or whatever other field) that one should expect some small fraction of the esoteric theoretical research will yield big returns, and that these returns will subsidize the failed efforts. This kind of mechanism allows for the possibility to support t=infinity timescale research with accountability checks of the collective at finite timescales.

Now, if we start to deviate from my assumption of integrity within a field and my scenario of field as idealized firms, there are all sort of interesting game theoretical questions that pop up. For example, what sort of internal rewards and punishments within a field are needed to keep things honest? I don't know, but I think it is interesting to think about.

[dnautics]

>assumes that there is internal accountability within that field

Yeah, that is a huge, and completely wrong assumption about the sciences.

I could rattle off a list of names just in my field (you should google them to find out exactly what they did): Dalibor Sames/Bengu Sezen. Leo Paquette. Homme Hellinga. Peter Schultz/Jonathan Zhang. Geoffrey Chang. Armando Cordova. To varying degrees of fraud, or negligence (e.g. Chang wasn't fraudulent, just almost criminally[0] negligent) And those are the people who got caught, and the ones I know about. Guess how many of them still have faculty positions?

[0]I feel safe saying this (with the "almost" qualification) because misappropriation of federal funds is a crime.

Irrespective of your assumption, you still need some sort of justfication for giving taxpayer money to science. Presumably, because it's good for society. Well fine. If it's so good for society, why can't you find people to give money to science as a free-will donation?