> michael wrote:
>
> >For the most part, you cannot refute any economic theory unless is is
> >logically inconsistent. ...
>
> That's what Krugman and the neoclassicals would like us to believe,
> isn't it? You certainly can refute a theory by showing it is not
> consistent with empirical evidence, can you not?
>
> Bill
Crudely speaking, there's internal critique and external critique. Internal critique can take two forms. One is logical consistency and the other is empirical testing. In the first, we accept the assumptions and the framework and examine the theory on its on grounds to see if the conclusions it claims are in fact legitimately derivable from its own assumptions. The Cambridge critique of neoclassical capital theory, marginal productivity theory (neoclassical theory of distribution), and price theory and the Sonnenschein-Mantel-Debreu impossibility theorem are examples of this kind of internal critique. The latter, empirical testing of a theory is a little different. Because it is tough to empirically test an entire *theory*, one can challenge particular results, specific claims. Of course, over time, you can build up a record and use this to argue for or against it. Of course it is still a problem, because of methodlogical problems of empiricism, testing, etc., and what constitutes "proof" and "refutation." (Also, I would distinguish between empiricism and using hisotry, but there are still complicated issues here).
External critique means you don't accept the assumptions, framework, methodology to begin with. So you can argue why there are better methodological approaches than methodological individualism or logical positivism or empiricism or why the data of a theory are the wrong data to take or you can get into the issues that Jeff raised about vision and ideology and so on, attack the starting point and offer alternatives.
My long winded two cents.
Mat