Hoisted from Comments: Robert Waldmann on price flexibility and deflationary expectations:
Grasping Reality with Both Hands: Brad DeLong's Semi-Daily Journal: And Apropos of Keynes, ch. 19...: I am impressed by Krugman's brilliance (as always) but not convinced by his argument (is this the first time for me ?). I don't know exactly what to do. My problem is this. Krugman discusses the price level and notes it didn't matter given the liquidity trap with nominal interest rates as low as they can go. However, this was due to deflation. The price level may not have mattered in the usual way, but the deflation rate mattered a lot.
I can present a story in which the key issue was to get the deflation rate up to 0 and price flexibility was needed to do this. That is a more rapid deflation would be a briefer deflation. The story follows:
Banks fail and the money supply falls partly because their reserves become cash in circulation and the money multiplier thus falls and partly because fearing more failures consumers take cash out of banks (more vault cash to cash in circulation) and banks feel they need higher reserves. Finally some of the cash in circulation is really cash stored under mattresses so it is not working as money at all. The reduction of the money supply causes an increase in the relative price of money (a deflation) so 1% nominal is [a] huge real [interest rate] and investment collapses. Only when the deflation ends can firms afford to invest again.
The deflation ends when the real value of money is so high that there is enough so that money actually circulating fits MV = PY with a normal V. For this to happen, it is necessary that people have built up the desired amount of cash under matresses. There is a new lower equiilibrium price level, and deflation continues until it is reached.
Now, if Krugman bothers to respond to my argument, he will say "what about the Phillips curve? The price level affects nominal interest rates (unless there is a liquidity trap) and they affect aggregate demand, but the change in the inflation rate depends on the unemployment rate via the Phillips curve. The only way to end deflation was to get the unemployment rate down using fiscal policy and/or/including the WPA and public works projects.
To which I say: hmmm the old Phillips curve applies in normal times and not when the key issue is how much money people have stuffed under their matresses. There is no empirical evidence, because we have no information on the shape of the Phillips curve at unemployment over 15% do we?...
Maybe with more price flexibility there would have been a quicker, briefer deflation. Real interest rates would have been even huger during this deflation but investment can be less than zero. I we assume that the bankruptcy of most industrial corporations due to debt deflation would have been no big deal, we can conclude that the NIRA may have prolonged the depression by a few months (by slowing and extending deflation until it was a dead letter)....
Ah I see that it was all there in the General Theory (of course). My argument is the one Keynes deals with immediately before making the argument also made by Krugman
The contingency, which is favourable to an increase in the marginal efficiency of capital, is that in which money-wages are believed to have touched bottom, so that further changes are expected to be in the upward direction. The most unfavourable contingency is that in which money-wages are slowly sagging downwards and each reduction in wages serves to diminish confidence in the prospective maintenance of wages. ["my" argument]
It is, therefore, on the effect of a falling wage- and price-level on the demand for money that those who believe in the self-adjusting quality of the economic system must rest the weight of their argument; though I am not aware that they have done so.
That Keynes acknowledges the argument doesn't means he deals with it. If recent past deflation does indeed create expectations of future inflation and so lowers money demand, then price and wage flexibility is indeed a stabilizing force.