A Taxonomy of Business Cycle Theories
As part of our research project, my colleague and I are doing a taxonomy of business cycle theories. Haberler famously differentiated between endogenous and exogenous theories of the cycle. But we can add at least two more contrasts, I believe:
True cycles versus illusory cycles: Real business cycle theory does not contain any cycles. The illusion of cycles if produced by random shocks with hysteresis. (This is not a criticism of RBC, any more than pointing out the amphibians have no fur is a knock against amphibians!) Minsky's cycle theory, on the other hand, contains a true cycle, since he offers an explanation (true or false is not the point here!) as to why conditions (are likely to) return to a state similar to the starting state of the cycle and then proceed roughly as they did before.
Note that "true versus illusory cycle" is orthogonal to the question of "endogenous versus exogenous." A theory of the business cycle based on sunspots is a true cycle theory, because sunspots are cyclical, but clearly about as exogenous as such a theory can be!
Regular cycles versus cycles of arbitrary length: Austrian cycle theory is, I think, a true cycle theory, but the length of the cycle can be quite variable, depending on the actions of the central bank. Kondratiev waves, on the other hand, or supposed to be fairly regular in duration.
How else might we classify cycle theories? Any suggestions?
UPDATE II: I left off an obvious one:
Does Say's Law always hold or not? In Austrian theory, Say's Law holds and the problem is structural, not general. In Keynes's theory Say's Law is only a long-run tendency, and there are general failures of aggregate demand.
UPDATE I: OK, I've been fighting NYC traffic for 45 minutes and I'm in no mood, so listen up!
If I am classifying theories of evolution, and I mention that "Creationism is not, of course, a theory of evolution, since it denies that (macro) evolution occurs," that remark should be pretty uncontroversial, right? You would think a creationist would say, "Damned straight it's not a theory of evolution! That (macro) evolution business is nonsense!"
Similarly, when the proponents of a macroeconomic theory explicitly boast that they have advanced macroeconomics by moving from the fallacious idea that "each cyclical phase of the economy carries within it the seed that generates the next cyclical phase" to the idea that all we are really seeing is a series of random shocks, you might think it would be uncontroversial to note that this is not a cycle theory, but a theory as to why sequences of random shocks can produce what looks like cyclical behavior.
One commenter was actually so outraged by my "attack" on RBC (which was, of course, no attack at all!) that he said I was committing the "no true Scotsman fallacy." Well, if I say of someone who is 40th generation Scottish that he is not a true Scotsman because he votes conservative, that is the no true Scotsman fallacy. But if I point out that someone not from Scotland and with no Scottish ancestry is not a Scotsman, that is not an example of that fallacy!
So, when I mention that a theory explicitly denying that cyclical phenomena play any major part in macroeconomic fluctuations is not a cycle theory, you'd think the main complaint would be I'm just stating the obvious!
True cycles versus illusory cycles: Real business cycle theory does not contain any cycles. The illusion of cycles if produced by random shocks with hysteresis. (This is not a criticism of RBC, any more than pointing out the amphibians have no fur is a knock against amphibians!) Minsky's cycle theory, on the other hand, contains a true cycle, since he offers an explanation (true or false is not the point here!) as to why conditions (are likely to) return to a state similar to the starting state of the cycle and then proceed roughly as they did before.
Note that "true versus illusory cycle" is orthogonal to the question of "endogenous versus exogenous." A theory of the business cycle based on sunspots is a true cycle theory, because sunspots are cyclical, but clearly about as exogenous as such a theory can be!
Regular cycles versus cycles of arbitrary length: Austrian cycle theory is, I think, a true cycle theory, but the length of the cycle can be quite variable, depending on the actions of the central bank. Kondratiev waves, on the other hand, or supposed to be fairly regular in duration.
How else might we classify cycle theories? Any suggestions?
UPDATE II: I left off an obvious one:
Does Say's Law always hold or not? In Austrian theory, Say's Law holds and the problem is structural, not general. In Keynes's theory Say's Law is only a long-run tendency, and there are general failures of aggregate demand.
UPDATE I: OK, I've been fighting NYC traffic for 45 minutes and I'm in no mood, so listen up!
If I am classifying theories of evolution, and I mention that "Creationism is not, of course, a theory of evolution, since it denies that (macro) evolution occurs," that remark should be pretty uncontroversial, right? You would think a creationist would say, "Damned straight it's not a theory of evolution! That (macro) evolution business is nonsense!"
Similarly, when the proponents of a macroeconomic theory explicitly boast that they have advanced macroeconomics by moving from the fallacious idea that "each cyclical phase of the economy carries within it the seed that generates the next cyclical phase" to the idea that all we are really seeing is a series of random shocks, you might think it would be uncontroversial to note that this is not a cycle theory, but a theory as to why sequences of random shocks can produce what looks like cyclical behavior.
One commenter was actually so outraged by my "attack" on RBC (which was, of course, no attack at all!) that he said I was committing the "no true Scotsman fallacy." Well, if I say of someone who is 40th generation Scottish that he is not a true Scotsman because he votes conservative, that is the no true Scotsman fallacy. But if I point out that someone not from Scotland and with no Scottish ancestry is not a Scotsman, that is not an example of that fallacy!
So, when I mention that a theory explicitly denying that cyclical phenomena play any major part in macroeconomic fluctuations is not a cycle theory, you'd think the main complaint would be I'm just stating the obvious!
I'm not certain that RBC is an "illusory" cycle. For example, one shock that would produce a "real" business cycle in the agricultural sector is an abnormally dry year. If we assume there is some long-run mean amount of rainfall, it seems we do have a "cycle" since output will return to its "normal" level once the rains return.
ReplyDeleteNo, jeremy, that is a one-time shock and a recovery. To have a cycle, we would have to have a theory as to why the recovery would create a new drought in time.
DeleteBy this definition, it seems that Austrian Cycle Theory is not a cycle theory either. You have a one-time credit cycle, which eventually ends with a recovery. A new cycle requires a new monetary shock.
DeleteKeynesian theory is not cyclical either, it seems by this definition. Investors get pessimistic, AD declines, then somehow you get recovery once the pessimism ends.
Perhaps you could say that animal spirits is cyclical. But so is the weather!
Good question about ABCT. Some people would agree with you. But I think there is a real cycle theory there in the "hair of the dog" aspect: in response to the downturn, the central bank is tempted to, and often does, flood the market with liquidity again, leading to a new boom. *That* generates the cycle, and without that I agree it is a one-time shock theory.
DeleteAnd the sunspot theory is definitely cyclical, it is just an exogenously driven cycle.
Keynes: I think there is a cycle story in there. There *definitely* is in Minsky's version of Keynes.
But I agree, these things can be muddy.
"Does Say's Law always hold or not? In Austrian theory, Say's Law holds and the problem is structural, not general"
ReplyDeleteWhile most Austrians may believe that Say's law holds and that recessions caused by inadequate AD can never occur I'm not sure that is relevant to ABCT.
What part of ABCT is dependent upon Say's law holding ? If anything the bust phase of ABCT could probably be shown to be an example of Say's law not holding (though probably this would not be recognized as such by proponents of the theory)
Excellent point, rob. I should have said, instead, that ABCT does not call attention to what might happen if Say's Law fails to hold, but instead focuses on how, even if Say's Law holds, structural imbalances may lead to a downturn.
DeleteAnd, rob, let me say, that comment is illustrative of why i really appreciate my regular commenters, and bother blogging instead of just jotting these ideas down in a notebook.
DeleteIt was your posts on Say's law from a few weeks ago that made me realize that an implicit assumption I had picked up while attempting to learn economics (mostly from Austrian sources) was in fact erroneous.
ReplyDeleteFactoring that in to my thinking has definitely helped me to understand economic better.
So: Thanks for blogging !
Thank God we are posting these mutual love fest posts from different geographical locations, or the night might end in acts we would both regret later!
Delete:-)