Ms McArdle asks which federal agencies work.
She nominates the FDIC, but doesn’t say what she means by work. So, I’ll try to extrapolate from her nomination. The FDIC seems quite effective at containing the failures of small banks. If you go beyond this (very narrow) metric, I think the picture becomes much less clear. For example, it’s looking like investors that bought failed banks are doing pretty well – probably too well given the federal guarantees involved. If you broaden the scope to overall financial regulation, the FDIC is certainly less captured by big finance than the other banking regulators, but it’s still captured in its own way. Since the FDIC sits on pile of cash that decreases if banks fail, the FDIC has every incentive to be a risk averse regulator. If the goal is a regulatory structure that best benefits society (whatever that means), I think the case can be made the FDIC is (on its own) overly risk averse. However, I’ll be the first to admit that this sort of "capture" is likely to be less problematic that the capture evidenced at other financial agencies.
Here are some criteria to consider when determining if an agency is effective: 1) a technical focus (agencies that can hire any random person tend to be staffed by idiots – if some technical expertise is required, you solve this most important problem); 2) targeted at solving specific tasks; 3) consistent funding (nothing screws up an agency more than uncertainty about funding); 4) regulation of industries (or whatever) that are relatively stable; and 5) the appointment of the head of the agency is not a huge political fight. To the extent that these criteria are met, the agency will work just fine.