Caveats

Brad DeLong has been on a bit of a run recently. I particularly liked this:

I think that modern neoclassical economics is in fine shape as long as it is understood as the ideological and substantive legitimating doctrine of the political theory of possessive individualism. As long as we have relatively-self-interested liberal individuals who have relatively-strong beliefs that things are theirs, the competitive market in equilibrium is an absolutely wonderful mechanism for achieving truly extraordinary degree of societal coordination and productivity. We need to understand that. We need to value that. And that is what neoclassical economics does, and does well.

Of course, there are all the caveats to Arrow-Debreu-Mackenzie:

The market must be in equilibrium.
The market must be competitive.
The goods traded must be excludable.
The goods traded must be non-rival.
The quality of goods traded and of effort delivered must be known, or at least bonded, for adverse selection and moral hazard are poison.
Externalities must be corrected by successful Pigovian taxes or successful Coaseian carving of property rights at the joints.
People must be able to accurately calculate their own interests.
People must not be sadistic--the market does not work well if participating agents are either the envious or the spiteful.
The distribution of wealth must correspond to the societal consensus of need and desert.
The structure of debt and credit must be sound, or if it is not sound we need a central bank or a social-credit agency to make it sound and so make Say's Law true in practice even though we have no reason to believe Say's Law is true in theory.