Monetarism in Historical Perspective

Cagan, Phillip Article Monetarism in Historical Perspective Kredit und Kapital Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Cagan, Phillip (1976) : Monetarism in Historical Perspective, Kredit und Kapital, ISSN 0023-4591, Duncker & Humblot, Berlin, Vol. 9, Iss. 2, pp. 154-163, https://doi.org/10.3790/ccm.9.2.154 This Version is available at: https://hdl.handle.net/10419/292761 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Monetarism in Historical Perspective By Phillip Cagan, New York/N. Y.* The Bullionists versus the Anti-Bullionists, the Currency School versus the Banking School, and now the Monetarists versus the Fiscalists. What jolly good sport to participate in a famous debate. It encourages a flowering of thought and economic research, not to mention conferences and grants. Participants in the former debates were immortalized in numerous Ph. D. dissertations and scholarly histories of thought [9, 16]. Alas, no one cares about the history of economic thought any more. Hence we must provide our own account. Hence this collection of commentaries. But does any one other than the participants care much about this debate? Perhaps not. Yet the issues have far-reaching implications for the conduct of policy and are a serious matter. In his careful review of the issues, Thomas Mayer largely bypasses the historical antecedents to present-day monetarism.** I believe the antecedents merit attention. I. Professional Opinion in the 1940s and Early 1950s No one who was not in touch with the economics profession in the 1940s and early 1950s can quite imagine the state of thinking then in the profession at large on monetary theory and policy. The quantity of money was not considered important, indeed was hardly worth mentioning, for questions of aggregate demand, unemployment, and even inflation. Peruse the journals of that period! The analyses may have contained an "LM" curve, following Hick's [4] famous interpretation of Keynes, but textual discussions of theory and policy in article after article hardly mentioned the quantity of money at all. Textbooks in basic economics and even in money and banking mentioned the quantity theory of money, if at all, only to hold it up to ridicule. Those textbooks produced an entire cohort of professional economists who became the * Comments from Anna J. Schwartz on an earlier draft were most helpful. ** Vol. 8 (1975) pp. 191 and pp. 293. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15 Monetarism in Historical Perspective 155 teachers of hordes of economics students. There were, of course, many exceptions, most notably at the University of Chicago and among some teachers of monetary economics whose intellectual heritage reached back into the 1920s. But if you traveled among the profession at large, mention of the quantity of money elicited puzzled glances of disbelief or sly smiles of condescension. Monetarism is a reaction to that earlier inhospitable environment. Indeed, if there had not been a time when most of the profession said "money does not matter',' it would never have occurred to anyone to say "money does matter" [7]. Much of the monetarist message at first was intended simply to reacquaint the profession with the principles of money known in the 1920s, which had first been denied and then widely forgotten during the 1930s and 1940s. That purpose has now been accomplished, thanks no doubt more to several decades of inflation than to the preachings of the monetarists. Nevertheless, now that monetarists have come out of the wilderness, few would deny them self-satisfaction for the renewed interest in money and monetary policy and its favorable recognition again in textbooks. II. The Relative Importance of Money Indeed, monetarism has gained such attention that its opponents have feared that it might sweep away everything else. They have raised the spectre that monetarists claim that "only money matters". That was never claimed. What the opponents mistook for such a claim, perhaps, was the lesser one implied by some monetarists' writings that, among the list of influences on aggregate demand, money should stand at the top. The supporting argument is that changes in the money stock inevitably affect aggregate demand, albeit often slowly. The other influences have uncertain effects. They depend, unlike money, upon redistributive effects, money illusion, or cooperative changes in the money stock, in which the net results can differ materially from case to case. This issue was argued at the Brown University Conference on Monetarism in 1974 [15]. No one there disputed the importance of money, and most of the proceedings were devoted to the importance if any to attach to fiscal policy. The fiscalists pointed out that the major econometric models show a large effect from changes in government expenditures or tax rates. But it is not clear how much the effect depends upon the Keynesian structure of the models and the cooperation of an OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15 156 Phillip Cagan expanding money stock. On a theoretical level fiscal effects (with the money stock constant) depend upon the interest elasticity of the demand for money balances (because the effect of a government deficit on interest rates which induces a more active use of money balances is needed to support an increase in expenditures) and uponthe extent to which future tax liabilities of government bond interest are discounted (because government borrowing can be offset by an increase in taxpayers' saving) [5 a]. As I pointed out in my published comments at the conference [15], all this makes the magnitude of the long-run (say over a year or more) effect of fiscal policy uncertain.1 On the presumption that the initial impact of a change in expenditures from a fiscal action will take place before the subsequent offsests in private spending associated with the financing of the action, the effect is likely to be much greater in the short run. Even so, the short-run effect of a tax cut can also be small if the public views it as temporary, and the size of multiplier effects of fiscal expenditures are uncertain in the short run because of the variable size and speed of associated offsets (not to mention administrative delays). Monetary policy presents almost the opposite timing pattern. Its effects are delayed and, in the short run, variable from case to case and uncertain. But it builds up to a single long-run effect, aside from usually minor differences in the manner of changing the money stock (open market operations, change in reserve requirements, etc.). A sustained change in the money stock tends to produce a proportional change in aggregate expenditures in the long run. There are no important qualifications or offsets to the effect of monetary changes on aggregate demand in the long run. A change in the demand for money balances due to accompanying changes in interest rates is an offset, but it cannot last forever and is largely a very short run offset. The only qualification to the proposition that changes in money will have inevitable long-run effects on aggregate demand arises when the monetary change is a policy response to actual or anticipated changes in aggregate demand, for in some cases policy may accommodate the money stock to demand changes which have already originated in other short-run influences. In those cases the causal connection between money and aggregate demand becomes muddled, because money is then an endogenous 1 If consideration is given to various wealth effects of fiscal actions, even the direction of the effect becomes uncertain [8]. The main monetarist propositions do not, however, depend upon wealth effects. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15 Monetarism in Historical Perspective 157 variable and not an independent source of disturbances to the economy. But the possible endogeneity of money is not denied by monetarism. All that monetarism holds is that money should be properly controlled and that, if it is, it can control aggregate demand. Whether or not money belongs at the top of the list of influences on aggregate demand is not a theoretical question to argue about, so long as money is not buried in neglect at the bottom as it once was. The question is an empirical one and will be settled by the weight of evidence. It was the importance of gathering the evidence which motivated Milton Friedman and Anna Schwartz to undertake their monumental study of U.S. monetary history [3]. They found support for their view that money had been the most important source of disturbances to the U. S. economy over that period. It is hard to imagine taking strong exception to their finding, though many do. If we take another step and conclude that money will continue to be a serious potential source of disturbances unless properly controlled, we enter into monetarist territory. III. A Fixed Monetary Rule The preceding empirical proposition about the importance of money supports a strong prima facie case for a policy of constant growth in the money stock, the bête noire of the opponents of monetarism. It is not entirely a new proposal. The immediate antecedent was Henry Simons' early arguments in favor of a constant money stock [14]2. This has the benefit of simplicity of policy management and would avoid the horrendous mistakes of policy that produced devastating fluctuations in the money stock of the past. It has the disavantage, however, of not stabilizing the price level. Lloyd Mints proposed that policy control the money stock with a view instead of stabilizing the price level [10]. His proposal has since lost its appeal, because the price level is not a sensitive indicator for monetary policy, an objection strengthened by experience in recent years. We have learned that money affects prices with a long lag and that movements in prices and activity are not closely correlated in the short run. Consequently, a policy of trying to stabilize the price level would not be successful in the short run and would entail substantial fluctuations in monetary growth. 2 See also Robertson's discussion [11] of policy views in the 1920s. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15 158 Phillip Cagan It is natural, therefore, to turn back to the money stock itself as the best indicator of policy, and to propose a constant rate of growth consistent with long-run price stability. Since the long-run trend of monetary velocity changes from time to time, moderate periodic changes could be made in the fixed rate of monetary growth to achieve approximate long-run constancy of the price level. An argument frequently made against such a proposal is that a discretionary policy must always dominate a fixed rule. If the authorities seek to stabilize the growth of aggregate demand, constant monetary growth will by chance and only occasionally be the optimal policy and therefore should not be specified beforehand. Let the authorities choose the best policy as they go along; if constant monetary growth is the best, it will be chosen. This is a silly argument and misses the point. The monetarist position is based on the fact of our ignorance and the mistakes that are bound to result, whereas the opposing argument assumes optimization of a known model of the economy. Once stochastic terms are added to represent uncertainty, the optimal policy collapses toward virtual constancy of the policy instrument in the short run as the degree of uncertainty increases. As Friedman pointed out [1], policy makers must have forecasts and knowledge of the effects of their actions that are very good indeed before they can hope to reduce rather than add to the instability in aggregate demand. Maybe they will someday, but who can be optimistic? Nor is that all. Besides the effect of stochastic terms in models to represent uncertainty, the structure of models entails an unknown degree of misspecification which can aggravate policy errors well beyond the two standard deviations of normal stochastic terms. In addition, the kind of policy pursued by the authorities will itself affect the response of the economy. Mints [10, p. 9 and elsewhere] held the view that public expectations of a constant-price-level policy would help to keep the price level stable. Uncertain, discretionary policies undermine stabilizing behavior by the public. Thus Benjamin Klein [5] concludes that the demise of the gold standard in the period between the two World Wars and absence of any clear commitment to stabilizing prices has made the price level more volatile than it used to be. In technical terms, he finds that changes in the price level used to be regressive but now follow a random walk. The same point is developed further with the new idea of "rational expectations" [6, 12, 13]. If the public bases economic decisions on exOPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15 Monetarism in Historical Perspective 159 pectations formed rationally, that is, with all the available information and economic knowledge, they will take into account what is announced or known about government and central bank policy. The consequences are remarkable and intriguing. The standard "Phillips Curve" is vertical (that is, no tradeoff exists between inflation and unemployment), and monetary policy has no effect on output and employment but only affects prices. Such models based on rational expectations should not be taken as accurate descriptions of economic behavior, but it is an empirical question whether they are worse approximations to reality especially today than are the various econometric models on which discretionary policy proposals are based. These econometric models derive "optimal" discretionary policies to trade off inflation and unemployment; they are able to do so because they are based on lags and ad hoc mechanical formulations of expectations by the public which are distinctly irrational in terms of the models themselves.3 But the recent work on rational expectations, though consistent with the monetarist tradition, goes beyond it. Monetarism is not based on rational expectations and does not require them. Indeed, a monetarist in good standing need not oppose all discretionary monetary policy. It might be beneficial in particular circumstances. The monetarist view is rather that discretionary monetary policy has not been beneficial over-all in the past, that its performance in the future cannot by assumption be taken as beneficial without question, and that in any event discretion should be quite limited and undertaken with caution. (Keeping monetary growth reasonably constant gives the authorities plenty to do, so fixed monetary growth is far from a "passive" policy.) Monetarists and their opponents disagree on the degree of economic stability to be achieved with a policy of constant or near constant monetary growth. The monetarists say the resulting stability — no longer disturbed by domestic monetary influences — will probably be the best attainable and "good enough", whereas the opponents say "not good enough" and think they know how to do better. This issue goes to the heart of the debate and does not appear capable of being settled to 3 As one who helped to popularize mechanical expectations in a study of hyperinflations done long ago before expectations entered empirical models at all, I welcome work on rational expectations as an appropriate antidote. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15 160 Phillip Cagan everyone's satisfaction by the kind of evidence so far available. We need experience with constant monetary growth. German policy has moved in that direction and, under prodding from Congress, so has the U. S. Federal Reserve. Perhaps the future will bring the needed evidence, but it should be understood that reasonably constant monetary growth, starting in the midst of rapid inflation and world economic disorganization, needs more than a few years to demonstrate its beneficial effects. Debates tend to accentuate extreme views. It should be emphasized, therefore, that monetarists do not claim that reasonably constant monetary growth (and its necessary corollary, freely floating exchange rates) will produce a millenium free of disturbances to the economy and of fluctuations in economic activity. They claim only that economic instability will be much less than in the past. Monetarists may differ among themselves whether some discretionary monetary policy and short-run variations in monetary growth could, under good management, be beneficial and worth trying. Given the lags in monetary effects, however, discretionary monetary policy will outperform a policy of constant growth at best moderately, and worse results are a real danger. IV. Is a Discretionary Fiscal Policy Needed? Monetarists are accused of claiming that fiscal policy has no effect. Not so. The defense of those effects thrown up by fiscalists is in response to a different monetarist argument, namely that the preference generally given to fiscal over monetary policy as a means of moderating economic fluctuations is a mistake, and that the order of priority should be reversed. Fiscal measures, particularly expenditures such as unemployment compensation, obviously do have short-run effects, and nothing in monetarist thought opposes their use for purposes of short-run stabilization, especially in view of the long lags of monetary policy. Friedman for one has pointed to the beneficial automatic effects of the government budget for short-run stabilization and proposed that it be nondiscretionary [2]. One might also seriously contemplate an active discretionary fiscal policy. I see no basic conflict of such a view with monetarism on a theoretical level. The catch is the fiscal performance in practice. Here monetarists are skeptical of a record which fiscalists glorify. As monetarists sump up our experience with fiscal policy, decisions to change the level of governOPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15 Monetarism in Historical Perspective 161 ment expenditures have administrative lags just as long as the "outside" lag for monetary policy, while cyclical changes in tax rates, on which fiscal stabilization in the U. S. has been forced to concentrate, have uncertain effects on consumption expenditures. Because of forecasting errors, the over-all contribution to the stabilization of mild business cycles is open to question. In severe, prolonged recessions the contribution is more likely to be beneficial and is widely accepted. Tax changes for stabilization purposes, however, are difficult to control in a democracy because of their effects on the distribution of income. The consequence is the well known preference of elected governments for cuts rather than increases and the persistence of deficits when they are not appropriate. Perhaps the management problems of fiscal policy are not insuperable. Yet they are formidable. As a first step, the monetarist proposals for monetary policy should be inoffensive to a wide range of economists including neo-Keynesians who might nevertheless doubt the degree of success to be achieved with stable monetary growth alone. Whether a discretionary fiscal policy, despite the risks, is also desirable depends upon how unsuccessful a monetarist policy would be. The monetarist position is that, after several years of reasonably stable monetary growth, discretionary fiscal changes would not be needed as a stabilization instrument. V. Concluding Remark As is evident from the above discussion, monetarism in my view is more a set of propositions about policy and about the empirical research to support them than a particular theoretical model of the economy. The core of the monetarist view is that money is very important and that reasonably stable monetary growth should be the centerpiece of stabilization policies.4 References 1. Milton Friedman, The Effect of a Full-Employment Policy on Economic Stability: A Formal Analysis, in his Essays in Positive Economics, Chicago, 1953. — 2. Milton Friedman, A Monetary and Fiscal Framework for Economic Stability, American Economic Review, June 1948, 245—64, reprinted in his 4 To avoid misunderstanding I should add that this is a policy for an economy already on its desired path. If we start far off the desired path, as with the inflation of 1975—76, we can only approach the desired path gradually. 11 Kredit und Kapital 2/1976 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/ccm.9.2.154 | Generated on 2023-01-16 13:33:15