Paul Krugman asks a good question:
And now as then, the whole notion [TC: what I call sectoral shift theories] falls apart when you ask why, say, a housing boom – which requires shifting resources into housing – doesn’t produce the same kind of unemployment as a housing bust that shifts resources out of housing.
This of course was also Sraffa's 1932 critique of Hayek. I would cite a few points:
1. Bloom, Floetatto, and Jaimovich focus on factors of confidence and option value in a real business cycle model to help explain the asymmetry. Alex and I present related arguments in our Principles text.
2. Standard models of matching and job search generate cyclical employment behavior in the required manner when combined with varying real shocks; see for instance Mortensen and Pissarides (1994). The asymmetry in these models comes from the difference between job creation and job destruction, which is again related to option value. M&P even build some simulations to show the effects can be quite large and that they fit to some extent with real world data. This is one of the best macro papers of the last twenty years and it is commonly recognized as such.
3. Caballero and Hammour have done relevant work in this area. You may recall their modeled proposition, published in the QJE: "Government incentives to production may alleviate unemployment but exacerbate sclerosis. In contrast, creation incentives increase the pace of reallocation." (As you'll see in a later post, that's not even the main claim I wish to make.) See also this paper.
4. Greg Mankiw, with Ricardo Reis, has some excellent and still undervalued papers on sticky information and business cycles.
5. The asymmetry between the upward and downward phases of the cycle was very much a concern of Knut Wicksell in his 1898 Interest and Prices. Wicksell's account, as I understand it, emphasizes the interaction between real and monetary shocks. This is still a better book than most of contemporary macroeconomics.
6. I still believe Krugman is correct in leveling "the asymmetry charge" against Austrian business cycle theory, with its excessive emphasis on monetary tricks. That doesn't mean the asymmetry charge always defeats a sectoral shock theory.
7. I don't think that sectoral shifts explain most business cycles. I was convinced by Christina Romer's work that a lot of downturns are simply the result of contractionary monetary policy at the wrong time. I'm also convinced by the Brainard and Cutler paper that sectoral shifts didn't have much to do with the downturn of the early 1980s. Nor do I think that the recent crisis is all about real shocks or that sectoral shifts explain every aspect of the downturn — hardly. Still, the theory behind real shocks, sectoral shifts and the like has improved a good deal since Schumpeter.
Overall I'm puzzled by Krugman's post in response. Why cite a weak argument by Schumpeter when so much stronger work is available to discuss?
By the way, Schumpeter himself charged that the Keynesians were misunderstanding the real accounts of the business cycle, including Pigou (and I read him as referring to himself as well); see p.944 in his History of Economic Analysis. Schumpeter was a selective re-reader of the past, however, so perhaps Krugman's interpretation of him remains correct.