Two reactions to this piece by Paul Krugman:
1. If Krugman believes that disinflation is "incredibly expensive," then he should be more wary of exploiting the Phillips curve tradeoff that he imagines exists.
2. Krugman should do some research on the debate about disinflation that occurred circa 1980. At the time, various "gradualists" thought that the sacrifice that would have to be made in reducing inflation would be very much larger than it actually turned out to be. People like Tom Sargent (see "The Ends of Four Big Inflations" and this paper) had a better grip on what was going on than the adaptive expectationistas.