On the Stabilizing Effects of Fractional-Reserve Banking

     

George Selgin believes that I tend to confuse Mises’s opinions with my own. Clearly, all participants in this online discussion have been strongly influenced by Mises’s writings, though each of us has driven this inspiration into somewhat different directions. It is therefore tempting indeed to conflate the master’s opinions with one’s owns. My impression is that this holds true for Professor Selgin, too, no less than for anyone else.

Since the discussion turned to Mises’s business-cycle theory, I should like to comment on three specific issues:

(1) Professor Hummel thinks that Mises’s business-cycle theory boils down to an “empty tautology.” Mises would probably have agreed that it is a tautology. In Human Action (chap. II, sect. 3), he stated: “Aprioristic reasoning is purely conceptual and deductive. It cannot produce anything else but tautologies and analytic judgments. All its implications are logically derived from the premises and were already contained in them.” But he went on to point out that tautologies are not always empty. Quite to the contrary, they can very well contain important insights into (causal) features of the real world.

Moreover, in Human Action he further elaborated the argument, already presented in the Theory of Money and Credit, that the propositions of economic theory concern ordinal rather than cardinal relations. This implies, for example, that business-cycle theory does not allow us to specify the precise date of future crises, and it also does not allow us to specify ex ante their magnitude. Rather, such specifications must be given in the light of “historical” judgments pertaining to the contingent future circumstances of time and place. As Mises underscores, these “historical” judgments always contain an element of subjectivity.

It is also worthwhile to recall that economic science, like any science, only provides partial insights (therefore the usual ceteris paribus clause). Mises’s business-cycle theory highlights only one chain of causation, but there are many others that might play out simultaneously. Again it is historical judgment, not a priori theory, which might disentangle the relative weight of each factor that bears on a concrete situation.

(2) I agree with Professor Hummel that the definition of the components of the money supply is a critical issue of Mises’s business-cycle theory. It is also correct that on this question there is no general agreement. However, this difficulty is, again, not a problem for theoretical analysis, but for historical analysis. The very point of Mises’s typology of money is that the technical characteristics of various financial instruments are more than often irrelevant to deciding whether or not they belong to the money supply. What counts is whether they can be redeemed on demand at par; and whether this is the case must be determined for each concrete historical context.

This being said, and since we are at it, I should like to mention for the record that I disagree with Mises on the central role of the money supply. I do not think there is an a priori causal relation between (“artificial”) increases of the money supply and intertemporal disequilibria (see Hülsmann 1998). (3) Lawrence White claims that an increase in the money stock “that counteracts an incipient excess demand for money is ... equilibrating. It does not lower the market interest rate below the natural rate, but prevents the opposite discrepancy.” Similarly, George Selgin argues that in such cases, credit expansion serves “not to drive lending rates below their ‘natural’ levels, but to keep them from rising above those levels. Expressed in the simple terms of the equation of exchange, the argument amounts to saying that, holding reserves constant, expansion of M is stabilizing rather than destabilizing so long as it serves to offset a like decline in V.”

I completely disagree with this view. Let me begin by highlighting that Professor Selgin conflates two very different meanings of stabilization. At the end of the passage I just quoted, he addresses the stabilization of aggregate demand (M x V). The production of fiduciary media by fractional-reserve banks may indeed, under certain circumstances, entail a stabilization of this aggregate. But so what? Why is such a state of affairs more beneficial than a shrinking aggregate demand or a rising aggregate demand? I know fairly well how a Keynesian economist might respond to this question. With Jeffrey Hummel, I wonder whether George Selgin’s response is really any different in substance. But my point is that the stabilization of aggregate demand is not the same thing as intertemporal stabilization. It does not at all follow that the former implies the latter, or the other way round.

This brings me to the central question: Is it really the case that credit expansion, when it occurs simultaneously with an increase in the demand for money, does not drive interest rates below their natural levels, but prevents them from rising above those levels? The whole argument is premised on the notion that the increase in the demand for money, if unchecked by a corresponding increase in the money supply, would entail an intertemporal disequilibrium. But why should this be so? It is true that the increase in the demand for money would tend to entail a temporary increase of market interest rates (the latter would not necessarily be higher than before, but rather higher than they otherwise would have been). But why should we interpret this event as an increase of the interest rates above their natural level? Why is that temporarily higher level not itself the natural level? Why should the structure of production not be adjusted to interest-rate changes resulting from variations of the demand for money? As Dan Mahoney (2011) has recently pointed out, it is precisely when fractional-reserve banks prevent changes of the interest rates that they steer the structure of production away from the state in which it should be.