Political science, even at its best, has few, if any, redeeming aesthetic qualities. We do not offer beautiful theories of how the cosmos came to be; our prose is at best serviceable; if our diagrams convey the meaning they are supposed to and no more, then they have done their job. That means that political science has to justify itself on the pragmatic grounds of its usefulness. But much of political science is not only ugly, but not especially useful. It doesn’t say anything that non-political scientists might possibly care about knowing.
The implication is that political scientists need not only to think about how best to convey what they do to the public; they need to think about doing whats that ought to be so conveyed. This is not to say that political science needs to be in the business of pleasing the crowd; many of the truths that political science might want to convey might indeed be somewhat unpleasant to the sensibilities and prejudices thereof. But it needs to see itself as making a useful contribution to public discourse. Jeff (if I understand him rightly) is suggesting that political science needs to be engaged with public debates in some quite profound ways. This would require a major reshaping of how political scientists understand themselves.
I completely agree: political science should have implications for politics, and political scientists should be engaged in the public sphere. This doesn't just mean doing op-eds and getting cozy "advisor" positions in government; it also means doing work that is at least somewhat relevant to political discussion, and conveying that work in clear, understandable ways.
Sen. Coburn's proposal caused quite a lot of discussion on my campus, and I quite often found myself in the minority when I said that I couldn't really justify NSF funding for most political science work. (Since then, Elinor Ostrom was awarded the Nobel Prize for Econ, and her work has often benefited from NSF grants, but hers is clearly an exceptional case.) The thing is, tho, neither could anyone else justify it when put on the spot except with vagaries like "knowledge is good for its own sake". I don't disagree, but not all pursuits of knowledge are worthy of public subsidy, are they? Why is political science more worthy than anything else?
It's not that I think political science is undeserving relative to other NSF grant winners, or even government spending more generally. My argument isn't negative, it's positive: I don't see how political science is especially deserving, except in cases where extensive field work is required, or expensive randomized trials must be conducted (and I am very much in favor of this sort of research, so long as it is carefully designed and executed). An example of that would be Dr. Ostrom's work, or many of the field experiments that political economists like Chris Blattman are doing. But for average, run-of-the-mill research... well, if you want to justify it by saying that there are more egregious wastes of public expenditure, then fine. But that's not a very strong argument in my mind.
Now don't get me wrong: I'll soon be applying for these things, and if I win one I'll gladly accept it. And it's only $9mn/year, so it's not as if political scientists are breaking the bank. But as a matter of principle...