Rajat directed me to a post by Miles Kimball, entitled "Pro Gauti Eggertsson". He discusses Eggertsson's views on the role of inflation expectations in the recovery from the Great Depression. Like Miles, I'm a big fan of Eggertsson's work on the role of dollar depreciation, and particularly the expectations channel. Unlike Miles, I strongly disagree with the view that higher inflation expectations from adverse supply shocks can be expansionary, even at the zero bound. Here's Miles discussing Eggertsson's work:
In "Was the New Deal Contractionary?" (to which his answer is "No") Gauti argues that the much derided National Industrial Recovery Act was useful as a way to raise inflationary expectations. And he argues that its explicitly temporary nature was a recognition that ordinarily raising inflation expectations is a bad thing, but under the extraordinary circumstance of the Great Depression when the economy was hard up against the zero lower bound on nominal interest rates, higher inflation expectations could be helpful. Together with "Great Expectations and the End of the Great Depression," this paper constitutes a major bit of revisionist economic history, countering the generally dim view that many economists have taken of the details of FDR's policies as anything more than sleight-of-hand confidence-building measures. (Here there is some danger I am projecting onto others my own view before reading Gauti's papers, but I do not think I was alone in those views.)
In macroeconomics, there are very few propositions that we can be relatively certain are true. One of those is the claim that Paul Volcker's tight money policy of 1981 triggered a severe recession. Another is that the NIRA slowed the recovery from the Great Depression. In my view this claim is not even debatable. There is a large empirical literature on this, by many researchers, using a variety of techniques. My recent 500-page book on the Great Depression devotes about 25% of its pages to showing that FDR's 5 wage shocks (the first of which occurred under the NIRA in July 1933) each sharply slowed the recovery in monthly industrial production.
If the empirical evidence is clear, what about the theory? Why aren't higher inflation expectations expansionary? The simple answer is that looking at inflation expectations is reasoning from a price change. Thus if inflation rises due to a positive AD shock (more NGDP) then both prices and output will rise in the short run. But if NGDP is unchanged, then higher inflation leads to lower output. That's why supply shocks are contractionary.
The New Keynesian models are simply wrong. The role played by inflation expectations in NK models should be played by NGDP growth expectations. Switching to NGDP allows us to avoid reasoning from a price change, and avoiding mistakes like the claim that adverse supply shocks can be expansionary.
Companies do not care about inflation expectations; they care about total revenue expectations---which are tied in to NGDP growth. If you tell steel companies that they must raise their hourly wage rates by 20% in two months (as FDR did to steel and other manufacturers in July 1933) then the steel companies may expect higher steel prices, but they will also expect lower profits, as steel prices would probably rise by less than wages. They would react by sharply cutting production, which is exactly what manufacturers did after July 1933 (until the NIRA was declared unconstitutional in May 1935, after which industrial output soared.) In contrast, if you tell steel companies that you are going to sharply devalue the dollar, as FDR did in April 1933, then they will expect higher revenue from selling steel. This will make both steel prices and steel output rise rapidly. Indeed industrial production soared 57% between March and July 1933, after which FDR's NIRA aborted the recovery for 2 years.
It's about as perfect a set of natural experiments as I've ever seen, at least in the field of macroeconomics. The following table shows some data for 1933, in case you are interested.