Liberty Matters

Reply to George Selgin on Austrian Business Cycle Theory

I appreciate George’s clarifications, which help push our conversation into more interesting realms. I am a great admirer of his work on monetary economics, believing he has made major contributions. I certainly did not intend to misrepresent George’s views. Nor do I have any fully developed business cycle theory of my own to offer as an alternative to his. I consider this to be the great unresolved issue in macoeconomics. While I believe, as I said before, that Austrian theory offers some penetrating insights, I do not find any of its variants to be entirely satisfactory.
George denies any intention to “revise the Mises-Hayek theory,” and yet he then proceeds to discuss how he “clarified” (his word) one aspect of the theory. Categorizing the varied theories of economists is inherently imprecise, especially when the differences are subtle and nuanced. I did concede that my four suggested variants of Austrian business cycle theory had elements in common, and that the goal of stabilizing MV came from Hayek (although not from Mises). Whether the writings of George (along with Larry and Steve Horwitz) deserve to be classified as a separate variant or merely an elaboration on Mises and Hayek is a judgment call. George is free to minimize his originality; I think it merits greater recognition.
Moreover, there is another respect in which George has extended Austrian business cycle theory. As mentioned in my previous comment, he has incorporated from the Monetarist analysis of Leland Yeager (which in turn goes back to at least the work of Clark Warburton) a story about the inherent disequlibrating effects of both excess supplies and demands for money. But unlike Yeager (who rejects Austrian business cycle theory), George replaced price stability as the equilibrating optimum with his productivity norm and its (usual) secular deflation.
I did misinterpret George’s explicit views about the empirical importance of velocity shocks, in part because a major portion of his 1988 book is devoted to explaining how an unregulated banking system would offset such shocks. That said, I still think George’s acceptance of velocity shocks as a potential source of depressions introduces a Keynesian element. I should hastily add that this is not intended as a criticism. Unlike many other Mises-influenced economists, I do not believe that Keynesianism is utterly bereft of value, despite my adamantly rejecting its policy implications. The essential element of Keynesian business cycle theory is that autonomous falls in velocity (what Keynesians traditionally refer to as a falls in autonomous expenditures, which are equivalent to increases in money demand) cause economic downturns. George is right that Monetarists also accept this as a theoretical possibility. But it is also a depression scenario that requires no previous malinvestment boom, giving it a decidedly un-Austrian flavor.
Moreover, one can make too much of the distinction between monetary and velocity shocks. Again as emphasized in my previous comment, how a shock is classified depends on what is included in the money stock. An extreme case is the Great Depression. Outside of Rothbardians, most proponents of Austrian business cycle theory accept that the Great Depression was made great by what Hayek came to call the “secondary deflation.” This entailed an enormous collapse of the broader measures of the money stock from 1929 to 1933, driven mainly by a series of banking panics, as well documented by Milton Friedman and Anna Jacobsen Schwartz (1963). Yet over the same four-year period the monetary base, directly controlled by the Fed, ultimately rose. Thus, what was a huge negative monetary shock from the perspective of the broader monetary measures was a negative velocity shock to the monetary base.
In response to my query (and apparently Ludwig Lachmann’s as well) about how looking at MV permits one to “precisely identify periods that represent artificial booms generating malinvestments,” George quite reasonably introduces expectations, viz.:
…once people come to anticipate a pace of spending growth such as might otherwise have set a cycle in motion, the cyclical effects of the growth are muted or neutralized, and remain so until expectations are again exceeded [emphasis mine].
This sounds to me suspiciously like pure Monetarism, in which downward turning points in MV growth become the primary cause of depressions. I’m unclear whether George is making a comprehensive claim; i.e., that any growth rate of MV, if constant and fully anticipated, will not generate a self-reversing malinvestment boom, until the growth rate changes. Or to put the question in concrete terms, is he saving that if MV growth in the United States had continued to remain constant after the expansion of 1922 to 1929, there would have been no garden-variety depression in 1929, much less a Great Depression? I may be missing something, but if there is another way to interpret George’s answer, I hope he will elaborate. To my question about how to determine if the central bank or something else causes a change in MV growth, George gives a straightforward answer. As long as there is a central bank that can target MV, then the central bank is responsible. But this is not really an answer to my question; it is a policy prescription. Indeed, it is a policy prescription that grants the Fed more ability to stabilize the macroeconomy, even if only in theory, than I think is warranted.
Again, consider the Great Depression. I concede that Friedman and Schwartz were right that the Fed, with a sufficient expansion of the monetary base, could have totally offset the deflation and even cut the bank panics short. But that still leaves open what triggered the panics. Friedman and Schwartz blamed a change in Fed policy, arguing that most of the bank failures were liquidity failures. Yet there are alternative explanations, including that of Charles Calomiris and others, who conclude that the banks were already suffering serious solvency problems. Surely this is an important economic question in and of itself, irrespective of whether the Fed, with a better policy, could have averted the subsequent deflation.
Perhaps because I was trained as a historian, I am at least as curious about the causes of depressions as about the proper policy. After all, depressions in the United States long predated the creation of the Fed. George and Larry have been in the forefront of economists revealing how numerous legal restrictions made the U.S. banking system peculiarly vulnerable to shocks and panics. Yet that still leaves unexplained the timing of the downturns. Were the initiating factors mainly domestic, or can we point the finger at the international factors, particularly the policies of the bank of England? Was it some combination, or do different cases require different stories?
Which brings us to the housing boom and the subsequent financial crisis. George and I agree that exceptionally low interest rates were a major cause, but we disagree as to why rates were so low. In a recent Econlib article (2013), I challenged the common but exaggerated notion of almost total Fed control over interest rates. Since many factors can affect interest rates, one must avoid the following circularity: “Why were interest rates so low? Because of Greenspan’s expansionary monetary policy. How do you know Greenspan’s policy was expansionary? Because interest rates were so low.” In other words, to blame Greenspan, some independent variable must be invoked. If I understand George’s reasoning (2008; but he can correct me if I am wrong), he relies (as do most other economists who blame the Fed) on some kind of Taylor Rule that specifies what the interest rate should have been to prevent the housing boom. But the Taylor Rule in all its versions makes the astonishing assumption that the underlying real interest rate (what Mises called the originary rate and Wicksell the natural rate) remains perfectly constant over long periods. I find this a particularly peculiar assumption for Austrian economists to accept.
The standard way the Fed is thought to affect interest rates is by changing its balance sheet, buying or selling securities and thus altering the monetary base. Yet the increase in the base during the housing boom was overwhelmingly dwarfed in size by the net inflow of savings from abroad. In one year alone, 2006, that annual inflow was reaching nearly $800 billion, far exceeding the mere $200 billion increase in the base for the entire half decade from 2001 to 2006. Furthermore, it is widely recognized that the savings inflow was not entirely market driven by was heavily influenced by the policies of the Chinese government, which had coupled a pegged exchange rate with significant internal capital and exchange controls. I have never encountered a plausible explanation of how the Fed can significantly affect interest rates other than by manipulating its balance sheet, nor a plausible mechanism that would magnify any resulting impact on interest rates way beyond the impact of other changes of like monetary magnitude in the demand for or supply of securities. But I am open to being persuaded.
P.S. I must apologize to the readers and my fellow participants in this discussion for the delay in writing this reply. I hope that doesn’t seriously inhibit possible responses.