[Editor’s Note: this article is adapted from a 2003 essay in the Quarterly Journal of Austrian Economics entitled “New Keynesian Monetary Views: A Comment.” As economists abandon theory in favor of makeshift plans to flood the economy with stimulus, Hülsmann here provides some helpful reminders of the fundamental problems behind the current economic consensus on money.]
The essential fallacy of John Maynard Keynes and his early disciples was to cultivate the monetary equivalent of alchemy.
They believed that paper money was a suitable means to alleviate the fundamental economic problem of scarcity. The printing press was, at any rate, under certain plausible conditions of duress, a substitute for hard work, savings, and cutting prices (Hazlitt 1959, 1960).
The self-styled new Keynesians have not at all abandoned this fallacy, and they therefore do not differ in any essential respect from the old Keynesians, in spite of the pains they take to distinguish themselves from the latter. The new Keynesian recommendation for monetary policy is to “stabilize the growth of aggregate demand.” In plain language this means that the monetary authorities should never stop flooding the economy with paper money. This is recognizably the core tenet of the old Keynesian monetary program, which in itself had been nothing but even older fallacies clothed in the new language of aggregate analysis.
In many respects, new Keynesian views on monetary theory and policy seem to be even more fallacious than those of their predecessors. Whereas Keynes and his immediate followers were still trained in the old-fashioned art of economic reasoning, the new Keynesians are macroeconomic purebreds.1 Their expertise lies more or less exclusively in the field of modeling. As with the macroeconomics profession in general, they are devoted to a positivistic methodology, putting all their energies into modeling quantitative relationships among things that are the result of human action, rather than into the analysis of human action itself. Not surprisingly, therefore, their “science” of the economy resembles a hotchpotch of educated guesswork, conventions, and fictions, all designed to make the problems under consideration amenable to mathematical treatment. At one point in his exposition, Zimmermann briefly concludes that the optimal inflation rate should be zero. But then he goes on: “There is a consensus, however, that an inflation target between zero and two percent is optimal, allowing for measurement errors” (Zimmermann 2003, p. 7). Those who are bewildered by such turns of argument might also wonder how one can believe that policy recommendations derived from a mere consensus of academic economists have anything to do with science. Similar problems crop up with the frequent jumps between statements of facts and value judgments, and to new Keynesian uses of expressions such as “principles,” “first principles,” and so on. In the writings of Austrian economists these words have a completely different meaning.
Honest new Keynesians acknowledge that their models are not a realistic representation of parts of the observed economy. They admit that their reasoning is based on stipulations and conventions of a more or less fictional and therefore arbitrary character. But they believe that economic science cannot do any better than that, whereas Austrians think they know a better approach. As a consequence, from an Austrian perspective, new Keynesian policy prescriptions do not exactly stand on the firm ground of science. For example, is it really true that central bankers are supposed to be well-meaning economic tsars? And is it really true that central bankers seek to minimize deviations of output from potential output? More importantly still, even if we assume for the sake of argument that both of these questions can be answered in the affirmative, we still have still not even touched the fundamental problems of monetary policy analysis—problems that require sober examination before one can venture to make any “scientific” recommendations to policymakers? Here is one for a starter: How can central bankers, or for that matter anybody else, know the potential output of an economy at any point in time? And also: Can monetary policy close the gap between actual and potential output at all? Again the sober observer cannot fail to notice that new Keynesians have no genuine answers to these questions—they answer them (in the affirmative) by stipulation and by convention.
More to the point of monetary theory and policy, Austrian economists certainly agree with the commonplace [notion] that money is not neutral in the short run. But they would add that money is never neutral, neither in the short nor in the long run, because of the wealth effects that go in hand with any change in the money supply. Monetary policy therefore always has an impact on the economy, irrespective of whether the policy has been anticipated.2 Its impact on prices is not necessarily proportional, but in any case it permanently alters the distribution of wealth among the members of society and thereby affects the prices paid for the various consumer’s goods and producer’s goods. It makes some ventures profitable, and it makes other ventures unprofitable. It creates fortunes and elevated social positions on the one hand, and it destroys wealth and income on the other.
The fundamental question we have to confront in the theory of monetary policy is therefore not whether money affects the real economy—yes, it does, both in the short run and in the long run—but whether changes of the money supply can make society better off in the aggregate. Austrian economists who follow the approach of Mises and Rothbard believe that it cannot. By contrast, the intellectual edifice of Keynesianism—both old and new—rests squarely on the notion that money does alleviate the problem of scarcity for society as a whole. The entire case for monetary policy is based on the idea that “a decrease of inflation is followed by temporary output losses” (Zimmermann 2003, p. 6). At least in the short run, there is a tradeoff between inflation and unemployment.
But why should we believe that such a tradeoff is more than an accident of history—that is, why should we not believe that a decrease in inflation could with equal probability lead to temporary output gains ? At this point the new Keynesians essentially repeat the arguments of their predecessors, referring to sticky prices, the Phillips curve, and discontinuous price setting. But the Phillips curve at the very best is the statistical representation of contingent historical data (Phillips 1958; Fisher 1973). And even if we grant for the sake of argument that it faithfully represents certain historically contingent data, the fact remains that stickiness of prices is no fact of nature such as rainy weather or sunshine, but a variable that depends on political and cultural factors. On a free market, price stickiness is always as low as it can humanly be, and no monetary policy can reduce it further—as long as labor unions can hire economists trained in (new and old) Keynesian economics, at any rate (Mises 1931, 1958; Friedman 1968).
Similarly, there is no reason to assume that staggered nominal price setting per se induces disequilibria—”fluctuations in output” and “macroeconomic inefficiencies.” Contracts are always made for certain periods of time. In more or less many cases, one of the contracting parties may discover after the fact that the price stipulated in the contract was not adequate from his point of view. But how can this be said to be a general social problem? After all, the other party usually gains in such a situation, and thus there is no general problem. Most importantly, however, it is not at all clear how such problems can possibly be avoided or reduced through manipulations of the money supply. Is it really necessary to repeat all the insights about expectations and the dangers of inflation-induced moral hazard—insights that seem to have made it into large parts of the mainstream these past thirty years?
A final word on the Taylor principle: it does not rule out inflation spirals. Under certain conditions it is not enough to increase interest rates overproportionally, because even such an overproportional increase might not be sufficient to offset the other factors that cause an increase in the price level. These conditions seem to hold in many cases (Orphanides 2000). Moreover, and more importantly, the entire focus on inflation targeting and interest rate targeting in particular is misleading, as Austrians have long been arguing, because bubbles can build up even at stable prices and interest rates.3
Monetary theorists are well advised to regain the habit of discussing policy questions at a fundamental institutional level. It is relevant and useful to raise the question of whether we need central banks at all, and why. And it is relevant and useful to compare free market money with fiat money. Eclipsing such questions from the very outset might be convenient for those who derive income from the present institutions. But it is a disserve to science, and to society as a whole.
Our Christmas in July sale is now live! Get up to 60% off on our hottest products today!