Brad DeLong has posted a link to an older paper of his, " 'Liquidation Cycles' and the Great Depression," in which he develops a reasonably coherent model of an economy in which austerity can be a rational response to a downturn. He even suggests that a couple of the business cycles in the U.S. in the 19th century might have been rational liquidations of railroad positions. It's well worth reading. And here's the comment I posted on DeLong's blog:
"This is a perfectly persuasive rational reconstruction, I think, of the liquidationist thesis. What strikes me is how much of the model I see echoed in (for example) Tyler Cowen's zero-marginal-product arguments (although he focuses on ZMP of labor, rather than ZMP of capital).
"What also strikes me, though, is that the liquidationists, who are (alas!) still with us are likely to argue that the liquidation was not allowed to procede, that the election of Roosevelt in the US and the abandonment of the gold standard in the US and England (and elsewhere) prevented the adjustments from occurring, that, far from having demonstrated that the collapse was deep enough to prove the liquidationists wrong, you [DeLong] have simply proved that they were right--standing in the way of liquidation only prolonged the agony.
"Of course, that argument also means that WWII was simply a barrier to the necessary adjustments as well. But in Germany and in Japan (indeed in Europe as a whole) the redundant capital was destroyed by the war, thus clearing the way for the adjustment to procede beginning in 1945.
"I'm actually serious--I expect this response, because, at its core, the liguidationist argument is not (I think) about economics, but about a morality that contends that 'no good will come of this...' "