Growth and financial instability: Schumpeter’s hypothesis

I have taken my own advice and read (most of) Schumpeter’s Business Cycles with some care. He has completely blown my mind — and I am left bewildered by how it is possible that this body of work has been all but forgotten.

All the elements of what is now known as the Kindleberger-Minsky model of financial crises were present in Chapter IV of Schumpeter’s Business Cycles, and indeed Minsky cites his advisor as an important source for the financial instability hypothesis.

There is a crucial aspect of Schumpeter’s analysis that is, however, typically omitted from discussions of “Minsky moments.” Schumpeter separated out the “displacement” and “boom” phases of crises as fundamentally productive phenomena: displacement is naturally caused when a transformative innovation is funded by credit creation through the financial system and a boom is the inevitable result. Thus, Schumpeter is careful to construct his argument so that there is no doubt that we need the financial system to create credit. Credit creation ensures that growth due to innovation is accompanied by growth in the money supply, and thus that innovation does not result in deflation.

In short, for Schumpeter displacements and booms are an essential part of the process by which innovation drives economic growth.

Unfortunately the same financial system that creates credit to fund innovation, also creates credit to fund many other activities, including the finance of inventories, expansion of existing businesses, and consumption, all of which appear initially to be justified by the dynamics of the boom, but which in the end cause the economy to overheat. Kindleberger calls this phenomenon “overtrading.” The counterpart to an overheated economy is bad debt. As this economy works through the bad debt, “abnormal” liquidation takes place and “destroys many things which could and would have survived without it. In particular, it often liquidates and weeds out firms which do not command adequate financial support, however sound their business may be.” 155. When the economic circumstances are particularly adverse, debt deflation can set in and cause a depression. This is the stage that Kindleberger named “revulsion.”

Overall, Schumpeter didn’t just describe the dynamics of bubble, he also argued that there was a close connection between these dynamics and the capacity of an economy to take advantage of innovation and to grow. In the process, he concluded that almost every economic “catastrophe” can be attributed to dysfunction in the banking sector – and in particular to a failure on the part of bank lenders and business borrowers to exert appropriate control on the use of credit.

Schumpeter’s error presumably was to acknowledge that it was beyond the scope of economic theory to determine how to discipline the banking and business communities in their use of credit. 156. Thus, the intellectual debates of the middle of the 20th were dominated by economic theorists who could offer simpler answers (spend money, increase the money supply) to extraordinarily complex problems.


Schumpeter’s Monetary Theory of the Macroeconomy

Brad DeLong needs to give Schumpeter’s 1939 work “Business Cycles” a read. DeLong writes that Schumpeter “genuinely seems to have no clue at all that the business cycle is a feature of a monetary economy,” that he did not understand Wicksell, and that he was a sound money ideologue. I have never studied the history of Schumpeter’s thought, so it is possible that there is an “early” Schumpeter of whom these accusations may be accurate, but by 1939 Schumpeter had not only grappled with Wicksell’s approach but written a very cogent critique of it, had worked out a monetary theory of the business cycle, and had aggressively criticized sound money advocates.

The problem is not that Schumpeter was not a monetary economist, but that he offered a fundamental critique of the way money is introduced into the Walrasian model upon which both modern Keynesianism and much of modern neo-monetarism are built. With Tobin and Friedman allied against his ideas, all that has been passed down into the modern canon of Schumpeterian macro is Minsky.

DeLong apparently has missed something very important about Schumpeter: His “entrepreneurs-disrupt-the-circular-flow-and-cause-structural-change-and-growth-theory of enterprise” is fundamentally a monetary theory. Schumpeter’s analysis is built on a whole theory of the monetary system that makes it possible for entrepreneurs to obtain the funds for their yet-to-be-realized projects without diverting too much in the form of real resources from the already established — and reasonably well functioning — economy. In Schumpeter’s theory at the core of a capitalist economy lies not just innovation, but more importantly the monetary finance of innovation.

This is what Schumpeter has to say about Wicksell (pp. 128 ff.):

The necessity of reconciling a nonmonetary theory with obvious facts of the sphere of money and credit is, in particular, responsible for the idea that there are two kinds if interest rates, a “natural” or “real” one which would also exist in a barter economy and which represents the essence of the phenomenon, a permanent net return from physical means of production, and a monetary one, which fundamentally is but the former’s reflex in the monetary sphere. The two may, nevertheless, differ of course or be made to differ by monetary policy or by an expansion or contraction of bank credit, but this constitutes a disturbance from which a definite string of consequences, among them the business cycle itself, has been deduced. The roots of this idea reach very far into the past and are clearly discernible in the English monetary discussions of the fourth and fifth decades of the nineteenth century. Its role in the thought of our own time is due to the teaching of Knut Wicksell and to the work of a brilliant group of Swedish and Austrian economists. For us, however, there is no such thing as a real rate of interest, except in the same sense in which we speak of real wages : translating both the interest and the capital items of any loan transaction into real terms by means of the expected variation in an index of prices, we may derive an expected and, by performing the same operation ex post, an actual rate of interest in terms of “command over commodities.” But nominal and real rates in this sense are only different measurements of the same thing or, if we prefer to speak of different things even in this case, it is the monetary rate which represents the fundamental phenomenon, and the real rate which represents the derived phenomenon. Hence, the money market with all that happens in it acquires for us a much deeper significance than can be attributed to it from the standpoint just glanced at [i.e. Wicksell]. It becomes the heart, although it never becomes the brain, of the capitalist organism.41

Note 41: Moreover, profits in our sense display no tendency toward equalization. This and the essentially temporary character of profits in our sense should be sufficient to make it quite clear that both our distinction between profit and interest and the relation between them is not identical with an old distinction between normal business profits and contractual interest. However much the writer welcomes anything that will link his teaching to older doctrine, he must point out, first, that normal profits and interest are, according to this view, still the same thing— exactly as contractual and directly earned rent of natural agents is— which he thinks erroneous, and, second, that the analytic problem which he undertook to solve by his theory of interest was precisely to show how it is possible that a theoretically permanent income flows from essentially transient sources and that it should not disappear as a net return through a process of imputation.

One of the points that Schumpeter is making is that if you are going to bring up Walras’s Law then it is clear that you are talking about a barter and not a monetary economy, because the model in which Walras’s Law can be derived is fundamentally a barter economy. (See Jakab and Kumhof for a modern treatment of this issue.) Schumpeter doesn’t tack money onto a model of barter and call it “monetary economics.” Instead he studies the banking system to understand what it does, and then develops a model where the business cycle is driven by phenomena in the money market.

Schumpeter goes on to explain that he has a completely different notion of “capital” than that used by Wicksell. Roughly speaking Schumpeter views capital as financing rather than as a stock of productive goods. He writes (pp. 130 ff.):

Capital in this sense is not goods but balances, not a factor of production but a distinct agent which stands between the entrepreneur and the factors. It can be created by banks because balances can. Its increase and decrease are not the same as increase and decrease of commodities or any particular class of commodities. Its market is simply the money market, and there is no other capital market. No realistic meaning attaches to the statement that, in the latter, “capital” (= some kind or other of producers’ goods) is being ‘lent in the form of money.” But again as in the case of interest the introduction into our analysis of this concept of capital does not do away with the problems of what is traditionally referred to as real capital—on the contrary, they reappear though in a new garb—and results arrived at by means of a monetary theory of capital not always invalidate, but in many cases only reformulate, the proposition of “real” theories of capital. If our understanding of the processes of capitalist society hinges on realizing the fact that monetary capital is a distinct agent, it also hinges on realizing how it is related to the world of commodities.

In addition, Schumpeter explained very clearly that “catastrophes” in capitalist economies are usually due to defective regulation of the banking system. If I remember correctly, DeLong has acknowledged that one of his errors prior to the crisis was to underestimate dysfunction in the banking system. Perhaps if he were more familiar with Schumpeter, he would have been less likely to make that error. Schumpeter writes (p. 117):

Moreover, bankers may, at some times and in some countries, fail to be up to the mark corporatively : that is to say, tradition and standards may be absent to such a degree that practically anyone, however lacking in aptitude and training, can drift into the banking business, find customers, and deal with them according to his own ideas. In such countries or times, wildcat banking develops. This in itself is sufficient to turn the history of capitalist evolution into a history of catastrophes. One of the results of our historical sketch will, in fact, be that the failure of the banking community to function in the way required by the structure of the capitalist machine accounts for most of the events which the majority of observers would call “catastrophes.” [my emphasis]

Thus, Schumpeter’s error is not to fail to write monetary economics, but to have such a profound understanding of the relationship between money and finance that his work was not understood by the founders of modern macroeconomics.

Furthermore, when it comes to “sound money” Schumpeter writes of the 1830s experience with “reckless banking” in the U.S. (pp. 236 ff):

Whatever our opinion might be if we placed ourselves on other possible standpoints, however strongly we may feel it our duty to condemn both the misconduct involved and the public opinion that not only condoned but fostered it, the fact still remains that we have before us the clearest historical instance by which to illustrate the function of credit creation. It was the financing of innovation by credit creation— the only method available, as we have seen in the course of our theoretical argument, in the absence of sufficient results of previous evolution—which is at the bottom of that “reckless banking.” This undoubtedly sheds a different light upon it. Those banks filled their function sometimes dishonestly and even criminally, but they filled a function which can be distinguished from their dishonesty or criminality. Sound money men of all times, hence, threw and still throw away the baby with the bath by condemning the principles of that practice, however understandable their clamor for policing and controlling the practice itself may have been. The people felt this. So did some of the advocates of inflation, though they were unable to formulate their case correctly.

Thus, Prof. DeLong is making very unfair accusations against Schumpeter when he claims that Schumpeter “denies that anything other than budget surpluses and ‘sound money’ can ever be appropriate economic policies.” Schumpeter indicates that the problem with people like DeLong is that they are unable to formulate their case correctly, because they don’t have a good understanding of the relationship between money and finance.

I think that Schumpeter would agree with my view that those who claim that monetary policy can solve the problem of “a shortage of safe assets” are confused. We have a shortage of safe assets, because we have a dysfunctional banking system that is currently incapable of creating safe assets. If this is correct, then holding interest rates at zero will not solve the problem and get the capitalist engine running again. That is like pouring gas into a broken engine and expecting it to run, when of course what is needed is that you make the effort to learn how to fix the engine.