Hahn contra fg? Isn't it? (amv)

In our ongoing discussion about the usefulness of rational expectations models with representative agents, I shamelessly quote Frank Hahn on my behalf:

The arguments that are offered in support of the view that we can treat the world as if it were in continuous Walrasian equilibrium are so appallingly bad and so much at variance with the Lucasian resolve to pursue serious and rigorous economics that they are barely worthy of consideration. One is told that not only will all Pareto-improving possibilities be exploited by markets but also they will be exploited the instant they arise. (No distinction is made between perceived and actual possibilities, and no attention is paid to the manner in which agents communicate.)

There is the even more disreputable argument that some three or four log-linear econometric equations not immediately at variance with the microeconomics of a single-agent and economy-wide equilibrium "work."

Then, of course, there is the "street light" defence, which says that because the Walrasian equilibrium theory is the only fully developed theory we have, we had better use it. But the drunk searching for his keys under a street light, when he lost them in a dark alley, will certainly not find them. Ptolemaic theory was the street light before Kepler.

(Hahn, p. 224, in: Frydman, R., Phelps, E. (eds), Individual forecasting and aggregate outcomes, 1983, CUP)