Remarks by Roger W. Garrison
November 24, 2001

Alan Ebenstein has posed a half-dozen questions in conjunction with his response to three reviewers of his Friedrich Hayek: A Biography (New York: St. Martin's Press, 2001).

The full text of Ebenstein's response is posted to the Hayek-L list archives.

The three reviews under discussion are:

Richard Ebeling, "F. A. Hayek: A Biography," Lugwig von Mises web site, posted March 26, 2001.

Bruce Caldwell, "Friedrich Hayek: A Biography, by Alan Ebenstein," The Independent Review, vol. 6, no. 2 (Fall), 2001, pp. 263-67.

Peter Klein, "Friedrich Hayek: A Biography, by Alan Ebenstein," Chronicles Magazine, January, 2002.

Ebenstein's Questions:
1. Did Hayek hold the view that in countries such as the United States and United Kingdom that aggregate prices did not rise during the 1920s?

2. Is this view factually correct?

3. Are points A through F [below] a correct portrayal of Hayek's views?

A. The Great Depression was caused by the expansion of money supply by central banks in the United States, United Kingdom, and elsewhere during the 1920s.

B. This expansion led to a misshaping of the structure of production during the '20s.

C. The Great Depression was not caused by deliberate deflationary action on the part of central banks.

D. The Great Depression was caused by the misshapen structure of production.

E. The deflation was primarily a result not of monetary, but of real, factors (i.e., maladjustments of industry).

F. A stimulative monetary policy would be an inappropriate policy response to the Great Depression.

4. Are points A through F factually accurate—did they really occur; or, in the case of F), was it optimal policy advice?

5. Is it a correct reading of Hayek that he thought a policy of significant credit expansion was followed by the United States federal reserve between 1929 and 1932?

6. Did the United States federal reserve follow a policy of significant credit expansion between 1929 and 1932?

There are many issues entangled among the half-dozen questions posed—issues involving historical fact, macroeconomic theory, and policy prescription. It is difficult to choose a starting point for a response. But let me start with the historical record of price-level changes during the 1920s. Here, there is no basis for disagreement between the Monetarists and the Austrians. Hayek recognized and Friedman documented that, by and large, the price level showed little movement during the 1920s, rising slightly only at the end of that decade. (A quick check of the CPI data provided by the Minneapolis Federal Reserve Bank shows no increase at all: The price index stood at 17.1 in 1923 and 1924, deviated little during the decade and again stood at 17.1 in 1928 and 1929.)

The Monetarist/Austrian difference comes in the supposed implications of the near-constancy of the price level. Friedman takes the absence of price inflation to be the hallmark of macroeconomic health. According to him, there was no money-induced boom during the 1920s—as evidenced by the virtual absence of price inflation. We can note that the Monetarist account of boom and bust focuses on the short-run/long-run relationship between inflation and employment. Periods of price inflation can cause employment to rise above its full-employment level. The dynamics hinge on misperceptions of the real wage rate during an inflation. When the perceptions finally get straightened out, employment reverts to its initial level and, due to complicating factors, possibly lower. The analytics have the economy moving up a downward-sloping short-run Phillips curve, after which the curve itself shifts rightward—to reflect changed expectations about inflation. Abstracting from complications, the initial and final equilibrium points both lie on a vertical long-run Phillips curve. This is the Monetarist story of boom and bust. The absence of price inflation during the 1920s means that there is no Phillips curve story to tell. (I realize that, on some occasions, Friedman employs his so-called "plucking model" to argue that there are no such boom-bust episodes; there are only lapses from trend-line growth followed by recovery.)

The Austrians and especially Hayek have always argued that the economy is put through a boom-bust cycle by relative-price movements and not by movements in the overall price level. Much of Hayek's Prices and Production and of his Monetary Theory and the Trade Cycle is devoted to this issue. He is critical of Wicksell and others for their undue attention to movements in the general level of prices and for their taking monetary theory to be almost exclusively concerned with the relationship between money and the price level—a relationship suggested by the equation of exchange: MV=PQ.

Real economic growth during the 1920s put downward pressure on prices in general. Think in terms of the equation of exchange where there are no changes in either M or V. If output Q is rising, then each unit of output must sell for a lower price. Growth is accompanied by price deflation, as actually occurred, for instance, during the 1890s. But, of course, there were changes in M during the 1920s. M was increasing. (The precise rate of increase as guaged by Friedman or as guaged by Rothbard need not concern us here.) The increase in M put upward pressure on prices in general. As it turned out, the consequences of a rising Q and a rising M were largely offsetting: Prices in general neither rose nor fell during the 1920s. Friedman would have to applaud the Federal Reserve for its adherence, in effect, to the monetary rule. Hayek, by contrast, referred to this outcome as "artificial price stability": The central bank had intervened to prevent the natural falling of prices during a period of otherwise healthy economic growth.

The specific nature of the problem that is caused by this artificial price stability derive from the fact that new money enters the economy through credit markets. (This, by the way, is a historical/institutional fact that the Monetarists systemmatically ignore.) While the price level remains largely constant during the 1920s, the interest rate, whose job it is to keep investment in line with saving, is artificially low. As I have come to express the matter: Padding the supply of loanable funds with newly created money drives a wedge between saving and investment. At a rate of interest below its natural level, savers save less but investors invest more. Hayek gave emphasis to the particular pattern of investment: Low interest rates favor long-term projects. Resources are allocated disproportionally to the early stages of production.

For a time the relative prices of the varous factors of production are inconsistent with the preferences of consumers. Those prices reflect instead the policy-induced changes in the rate of interest. Markets that allocate resources intertemporally have been thrown out of equilibrium. These are the considerations that led Hayek to the conclusion that such policy-induced booms are fundamentally unsustainable. The Federal Reserve fostered a little more growth during the 1920s than savers were willing to finance.

During a credit expansion, the capital structure becomes excessively future oriented. The conflict between production plans and consumer preferences eventually precipitates a bust. Of course, if we could "assume" homogenous capital, there need be no actual bust, just a swift and costless reorientation of the capital structure. Capital in the form of cement mixers, for instance, could be used instead to mix Pina Coladas. The raw fact that capital is actually heterogenrous—that it can be reoriented only within limits and only at some cost and over some period of time—means that the economy suffers a downturn. That is, there is a bust.

We see from this line of argument that the Austrian theory of the business cycle is a theory of the unsustainable boom. It is not a theory of depression per se. The theory explains why and how a policy-induced boom will do itself in. It does not explain what complications might aggravate the subsequent bust.

One such aggravating factor is a collapse of the money supply. Another is a rise in the demand for money (liquidity) beginning with the downturn. That is, both M and V can fall, putting strong downward pressure on PQ. And to the extent that P doesn't fall in lockstep with the fall in MV (and how could it?), then Q falls, sending the economy into deep depression. (Here, I'm framing the issues in a way that will not win the approval of all Austrians. Rothbard, for instance, counted as "inflationary" any actions of the central bank that served to prevent or delay the money supply from collapsing to its gold base.)

The question of whether or not the central bank should resist price deflation by injecting more money to offset a falling V is a question that divides the Austrians. And not suprisingly so. This is a secondary issue with them. The primary issues are the origins of the boom and the inevitability of the bust. The Austrians would not say that the depression that began in 1929 was "caused" by deflationary pressures any more than they would say that the destruction of San Francisco in 1906 was "caused" by fire. They would instead identify the causes as the policy-induced artificial boom in the first case and, in the second case, the earthquake. Of course the quaking of the earth can quickly and dramatically get compounded by the outbreak of fires; and the quaking of the capital structure can quickly and dramatically get compounded by deflationary pressures. Alternatively stated: an economywide intertemporal disequilibirum caused by too low an interest rate can set the stage for monumental blundering by the Federal Reserve. (In Friedman's account of events, there is only inherent ineptness and exogenous blundering .)

In the 1930s, when Hayek took for granted the appropriateness of a central bank, he made several arguments against reflating. First, he argued in the broadest terms against undue attention to the price level and against the belief that if only the price level can be stabilized, the economy will be set straight. Second, he argued that the process of increasing the money supply impinges directly on the interest rate, exacerbating the very kind of misallocations from which the economy is already suffering. And third, he argued that a deflationary pressures may cause prices and wages to become more flexible. I take this third argument to evidence some exasperation on Hayek's part. Hayek was constantly confronting the Keynesian position according to which expansionary policies were justified on the grounds of a supposed wage and price inflexibility. All of Hayek's arguments add up, in my view, to serious doubts on his part that the institution responsible for igniting and fueling the artificial boom by expanding credit can then undo the damage by, well, expanding credit. (Here my earthquake/fire metaphor breaks down. San Franscisco's Fire Department was not responsible for triggering the quake. We should not be reluctant, then, to rely on the firefighters to fight the fires.) All Federal Reserve efforts in the direction of expansion during the 1929-1933 period (and there were ongoing debates within the Federal Reserve about just what it should be doing) were judged to be ill-advised by Hayek and judged to be wholly inadequate by Friedman. We know from Friedman's monetary history that the net effect of the actions of the Federal Reserve, the commercial banks, and the depostors was a dramatic monetary contraction.

When, in the 1970s, Hayek considered the possibility of competitive, decentralized banking, he had long since left the business-cycle issues behind. But subsequent developments in this area have suggested that (1) the mechanism of decentralized banking—precisely because of the decentralization—do not allow for the kind of economywide credit expansion that ignites and fuels an artificial boom, and (2) increased demands for liquidity—whatever the reason for the increase—will be accommodated by an increased money supply. In retrospect, it might be better to judge the performance of the central bank in its handling of a crisis (even a crisis the central bank itself had caused) on the basis of what would happen automatically in a decentralized banking system. By that standard, we could hope that the central bank would accommodate the demands for liquidity—providing just enough and not too much. But this hope is largely a vain one because the central bank lacks the timely market guidance as to just how much liquidity to provide. It tends, as Friedman never tires of pointing out, of overdoing it in both directions.

It is true that in his earliest writings on the subject of business cycles, Hayek argued that artificial booms can be initiated by almost anything. Cycles are virtually inevitable, he claimed, because of the elasticity of the supply of credit in the banking system. He suggested, for instance, that unusually high profit opportunities in a partucular industry can trigger a boom. The idea here is that, because of elastically supplied credit, banks can extend credit to the high-profit industry without withholding credit from industries that compete for the same resources. This notion of inherent cyclical tendencies is closely akin to Knut Wicksell's rockinghorse theory. The particular nature of the disrupting force is irrelevant. It is the very structure of the system that matters. Hit it, kick it, throw money at it, or whatever. If subjeceted to any disturbance, the rockinghorse rocks and the economy cycles.

Soon enough, however, Hayek came to consider credit expansions by the central bank as having a special claim on our attention. These are the historically most relevant instances of cyclical patterns. In his Constitution of Liberty (1960), Hayek emphasized the political incentivies faced by the monetary authority, significantly anticipating Public Choice theory and the political business cycle and adding to the historical and contempory relevance of policy-induced booms.

It is not at all unusual for a thinker such as Hayek to change his mind or his emphasis over time—although the various distinctions between a Hayek I and a Hayek II can easily be overdrawn. We should remember that Mises once endorsed Wicksell's view that the boom was initiated by technological changes that caused the natural rate of interest to rise. He only later came to associate credit expansions with inflationist ideology. And I suppose we might also remember that Milton Friedman was once very Keynesian in both his thinking and his policy preferences—as he himself reported in his and Rose's Two Lucky People (1998).

Let me end with a summary of Hayek's achievements and few questions of my own.

Hayek was very much alive to the microeconomic foundations of the macroeconomic phenomenon that we call business cycles. The price system guides the allocation of resources. The interest rate allocates resources intertemporally. An interest rate that "tells the truth" will cause producers to allocate resources among the various stages of production in accordances with consumers' intertemporal preferencs. It will maintain the right balance between current consumption and economic growth. An interest rate distorted by central-bank policy will set the economy onto an unsustainable growth path. The market process set in motion by credit expansion will not play itself out as economic growth but rather will do itself in. The conflict between consumers and producers, which will eventually—but only eventually—be won by consumers, takes the form of a boom-bust cycle. Hayek gave us the analytical apparatus for understanding the nature of this market process.

Hayek borrowed the term "natural rate on interest" from Wicksell to signify the rate that entails no money-induced distortions. Friedman deliberately chose a parallel term, the "natural rate of unemployment," to signify a rate that, similarly, entails no money-induced distortions. Now newly injected money does come through credit markets and hence impinges, in the first instance, on interest rates. This aspect of money creation would seem to give Hayek's theory a firm historical and institutional link to reality. By contrast, the wage rate and the unemployment rate are affected only after the new money has led to price inflation and only as a result of workers misperceiving the real wage rate. They are willing to work more, perceiving that the real wage rate has risen, when, in fact, the real wage rate has fallen. Eventually, however, workers straighten out their perceptions of the real wage and the unemployment rate reverts back to its natural level.

Now for some questions:

1. Is my summary paragraph true to the the labor-market dynamics that we call short-run/long-run Phillips curve analysis?

2. Is this the Monetarists' view (and is it Freidman's view) of boom and bust?

3. In particular, do the Monetarists (and Friedman) believe that real wages actually fall during the boom?

4. Is it true that real wages fall during the boom?

5. In setting out a monetary theory of boom and bust, how do the Monetarists (and Friedman) justify focusing on the wage rate and labor markets when, in fact, money impinges most directly on the interest rate and loanable-funds markets—and hence on capital markets?

6. And finally, why do the Monetarists tend to look with contempt and disdain upon throrists who do focus on interest rates, loanable-funds markets, and capital markets when analysing the effects of credit expansion?

My Time and Money: The Macroeconomics of Capital Structure (London: Routledge: 2001) provides discussion and works toward a resolution on these issues. But I and possibly others would benefit from answers offered by economists who are more Monetarist than Austrian in their thinking.