4

Thefollowing article is reprinted from the July 1968 issue of the Federal Reserve Bank of St Louis Review.

Karl Brunner

The Role of Money and Monetary Policy

HE DEVELOPMENT of monetary analysis in A response to the criticisms of existing mone- the past decade has intensified the debate con- tary policy methods was naturally to be ex- cerning the role of money and monetary policy. pected and is welcomed. Four articles which de- Extensive research fostered critical examinations fend present policy procedures have appeared of the Federal Reserve’s traditional descriptions during the past few years in various Federal of policy and of the arrangements governing Reserve publications.1 ‘I’hese articles comprise a policymaking. Some academic economists and countercritique which argues that monetary im- others attribute the cyclical fluctuations of pulses are neither properly measured nor ac- monetary growth and the persistent problem tually transmitted by the money stock. The concerning the proper interpretation of authors reject the Monetarist thesis that mone- monetary policy to the established procedures tary impulses are a chief factor determining of monetary policy and the conceptions tradi- variations in economic activity, and they con- tionally guiding policymakers. tend that cyclical fluctuations of monetary growth cannot be attributed to the behavior of The critique of established policy procedures, the Federal Reserve authorities. These fluctua- which evolved from this research into questions tion are claimed to result primarily from the concerning the monetary mechanism, is de- behavior of commercial banks and the public. rived from a body of monetary theory referred to in this paper as the Monetarist position. ‘I’he ideas and arguments put forth in these Three major conclusions have emerged from articles deserve close attention. The controversy the hypotheses put forth. First, monetary im- defined by the critique of policy in professional pulses are a major factor accounting for varia- studies and the countercritique appearing in tions in output, employment and prices. Second, Federal Reserve publications bears on issues of movements in the money stock are the most fundamental importance to public policy. Under- reliable measure of the thrust of monetary im- lying all the fashionable words and phrases is pulses. Third, the behavior of the monetary the fundamental question: What is the role of authorities dominates movements in the money monetary policy and what are the requirements stock over business cycles. of rational policymaking?

‘Lyle Gramley and Samuel Chase, “Time Deposits in Alternative Approaches to the Analysis of the Financial Monetary Analysis,” Federal Reserve Bulletin, October Structure,” Monthly Review, Federal Reserve Bank of Kan- 1965. John H. Kareken, ‘Commercial Banks and the sas City, March 1968. Richard 0. Davis, “The Role of the Supply of Money: A Market Determined Demand Deposit Money Supply in Business Cycles,” Monthly Review, Rate,” Federal Reserve Bulletin, October 1967. J. A. Cacy. Federal Reserve Bank of New York, April 1968.

FEDERAL RESERVE BANK OF ST. LOWS 5

The following sections discuss the major as- role of the public’s and the banks’ behavior in pects of the countercritique. These rejoinders the determination of the money stock, bank may contribute to a better understanding of the credit and interest rates. issues, and the resulting clarification may re- Kareken’s paper supplements the Gramley- move some unnecessary disputes- Even though Chase arguments. He finds “the received money the central contentions of the controversy will supply theory” quite inadequate. His paper is remain, the continuous articulation of opposing designed to improve monetary analysis by con- points of view plays a vital role in the search structing a theory of an individual bank as a for greater understanding of the monetary firm. This theory is offered as an explanation of process. a bank’s desired balance sheet position. It also appears to form the basis of a model describing A SUMMARY OF THE the interaction of the public’s and the banks’ COUNTERCRITIO,UE behavior in the joint determination of the money stock, bank credit and interest rates. The four articles relied on two radically dif- The whole development emphasizes somewhat ferent groups of arguments- Gramley-Chase, suggestively the importance of the public’s and Kareken and Cacy exploit the juxtaposition banks’ behavior in explanations of monetary “New View versus Traditional View” as the cen- growth. It is also designed to undermine the tral idea guiding their countercritique. The ana- empirical hypotheses advanced by the Monetar- lytical framework developed by the critique is ist position. This is achieved by means of ex- naturally subsumed for this purpose under the plicit references to specific and “obviously “Traditional View” label. On the other hand, desirable” features of the model presented. Davis uses the analytical framework developed Cacy’s article develops neither an explicit by the critique in order to organize his framework nor a direct critique of the basic arguments. propositions advanced by the Monetarist thesis. Gramley-Chase describe their general argu- However, he provides a useful summary of the ment in the following words: general position of the countercritique. The Monetarist analysis is conveniently subsumed by “(New) developments have reaffirmed the bankers’ Cacy under a “Traditional View” which is jux- point of view that deposits are attracted, not taposed to a “New View” of monetary mecha- created, as textbooks suggest. In this new environ- nisms: “The new approach argues... that there ment, growth rates of deposits have become more suspect than ever as indicators of the conduct of is no essential difference between the manner monetary policy A framework of analysis [is in which the liabilities of banks and nonbank required] from which the significance of time financial institutions are determined. Both types deposits and of changing time deposits can be of institutions are subject in the same way to deduced. Traditional methods of monetary analysis the portfolio decisions of the public.” The new are not well suited to this task. The ‘New View’ in approach is contrasted with the Traditional monetary economics provides a nlote useful View, which “obscures the important role analytical framework. In the new view, banks— played by the public and overstates the role like other financial institutions—are considered as played by the central bank in the determination suppliers of financial claims for the public to hold, of the volume of money balances.”~The general and the public is given a significant role in deter- comparison developed by Cacy suggests quite mining the total amount of bank liabilities. -- clearly to the reader that the Traditional View Traditional analysis. --fails to recognize that substitution between time deposits and securities allegedly espoused by the Monetarist position may be an important source of pro-cyclical varia- cannot match the “realistic sense” of the New tions in the stock of money even in the face of View advocated by the countercritique. countercyclical central bank policy.” In the context of the framework developed by This general argument guided the construction the critique, Davis questions some basic proposi- of an explicit model designed to emphasize the tions of the Monetarist position:

2 Gramley-Chase, pp. 1380, 1381, 1393. ~lbid.,p- 7. ‘Cacy, pp. 5 & 7.

SEPTEMBER/OCTOBER 1989 6

“In the past five to ten years, however, there pirical conjectures are logical implications of the has come into increasing prominence a group of general program. The explicit separation of the economists who would like to go considerably three aspects is crucial for a proper assessment beyond the simple assertion that the behavior of of the New View. money is a significant factor influencing the

behavior of the economy. ---In order to bring a Section A examines some general character- few of the issues into sharper focus, this article istics of the countercritique’s reliance on the will take a look at some evidence for the ‘money New View. It shows the New View to consist of supply’ view. - a program acceptable to all economists, a re- It confines itself to examining the historical rela- search strategy rejected by the Monetarist posi- tionship between monetary cycles and cycles in tion, and an array of specific conjectures ad- general business. The article concludes that the vanced without analytical or empirical substanti- relationship between these two kinds of cycles ation. Also, not a single paper of the counter- does not, in fact, provide any real support for the critique developed a relevant assessment of the view that the behavior of money is the predomi- Monetarist’s empirical theories or central prop- nant determinant of fluctuations in business activi- ositions. ty. Moreover, the historical relationship between cycles in money and in business cannot be used to In sections B and C detailed examinations of demonstrate that monetary policy is, in its effects, specific conjectures centered on rival explana- so long delayed and so uncertain as to be an un- tions of cyclical fluctuations of monetary satisfactory countercyclical weapon.” growth are presented. The direct assault on the Monetarist position by Davis is discussed in AN EXAMINATION OF THE ISSUES some detail in section D. This section also states the crucial propositions of the Monetarist thesis A careful survey of the countercritique yield- in order to clarify some aspects of this position. ed the following results. The Gramley-Chase, This reformulation reveals that the reservations Kareken, and Cacy papers parade the New View assembled by Davis are quite innocuous. They in order to question the status of empirical provide no analytical or empirical case against theories used by the Monetarist critique in its the Monetarist thesis. Conjectures associated examination of monetary policy. The Davis with the interpretation of monetary policy (the paper questions quite directly, on the other “indicator problem”) are presented in section E. hand, the existence and relevance of the evi- dence in support of the Monetarist position, and A. The New View constitutes a direct assault on the Monetarist The countercritique has apparently been critique. The others constitute an indirect decisively influenced by programmatic elabora- assault which attempts to devalue the critique’s tions originally published by Gurley-Shaw and analysis, and thus to destroy its central proposi- James Tobin.’ The program is most faithfully tions concerning the role of money and mone- reproduced by Cacy, and it also shaped the tary policy. arguments guiding the model construction by The indirect assault on the Monetarist position Kareken and Gramley-Chase. The New View, as by Gramley-Chase, Kareken and Cacy requires a a program, is a sensible response to a highly clarification concerning the nature of the New unsatisfactory state of monetary analysis in- View. A program of analysis must be clearly herited in the late 1950’s. A money and banking distinguished from a research strategy and an syndrome perpetuated by textbooks obstructed array of specific conjectures.6 All three aspects the application of economic analysis to the are usually mixed together in a general descrip- financial sector. At most, this inherited litera- tion. It is important to understand, however, ture contained only suggestive pieces of an- that neither research strategy nor specific em- alysis. It lacked a meaningful theory capable of

5 Davis, pp. 63-64. tuation of monetary growth and propositions about proper 6 These three aspects of the New View will subsequently be interpretation of policy. elaborated more fully. Their program of analysis refers to ‘John 0. Ourley and Edward F. Shaw, Money in a Theory the application of relative price theory to analysis of finan- of Finance, (Washington: Brookings Institute, 1960). James cial markets and financial institutions. Their research Tobin, ‘Commercial Banks as Creators of Money,” Bank- strategy refers to a decision to initiate analysis in the con- ing and Monetary Studies, ed. Deane Carson (R. D. Irwin, text of a most general framework. Their specific conjec- 1963). tures refer to propositions concerning the causes of fluc-

FEDERAL RESERVE BANK OF St LOUIS 7 explaining the responses of the monetary sys- termediaries, or between money and other fi- tem to policy actions or to influences emanating nancial assets. from the real sector. The New View proposed a systematic application of economic analysis, in The strong impressions conveyed by the New particular an application of relative price theory, View thus result from the relative emptiness of to the array of financial intermediaries, their the formulation which has been used to elabo- assets and liabihties. rate their position. In the context of the formal world of the New View, “almost everything is This program is most admirable and in- almost like everything else.” This undifferentia- contestable, but it cannot explain the conflict ted state of affairs is not, however, a property revealed by critique and countercritique. The of our observable world. It is only a property of Monetarist approach accepted the general prin- the highly formal discussion designed by the ciple of applying relative price theory to the New View to overcome the unsatisfactory state analysis of monetary processes. In addition, this of monetary analysis still prevailing in the late approach used the suggestions and analytical 1950’s or early 1960’s.8 pieces inherited from past efforts in order to develop some specific hypotheses which do ex- Two sources of the conflict have been recog- plain portions of our observable environment. nized thus far. The Monetarists’ research strat- The New Viewers’ obvious failure to recognize egy was concerned quite directly with the the limited content of their programmatic state- construction of empirical theories about the ments only contributes to maintenance of the monetary system, whereas the New View in- conflict. dulged, for a lengthy interval, in very general programmatic excursions. Moreover, the New A subtle difference appears, however, in the Viewers apparently misconstrued their program research strategy. The New View was introduc- as being a meaningful theory about our obser. ed essentially as a generalized approach, in- vable environment. This logical error con- cluding a quite formal exposition, but with little tributed to a third source of the persistent attempt at specific structuring and empirical conflict. content. The most impressive statements pro- pagated by the New View were crucially in- The latter source arises from the criticism ad- fluenced by the sheer formalism of its exposi- dressed by the New Viewers to the Monetarists’ tion. In the context of the New View’s almost theories of money supply processes. Three of empty form, little remains to differentiate one the papers exploit the logically dubious but object from another. For instance, in case one psychologically effective juxtaposition between a only admits the occurrence of marginal costs “New View” and a “Traditional View.” In doing and marginal yields associated with the actions this they fail to distinguish between the inheri- of every household, firm, and financial interme- ted state of monetary system analysis typically diary, one will necessarily conclude that banks reflected by the money and banking textbook and non-bank financial intermediaries are syndrome and the research output of restricted in size by the same economic forces economists advocating the Monetarist thesis. and circumstances. In such a context there is This distinction is quite fundamental. Some for- truly no essential difference between the deter- mal analogies misled the New Viewers and they mination of bank and non-bank intermediary did not recognize the logical difference between liabilities, or between banks and non-bank in- detailed formulations of empirical theories on

6 Adequate analysis of the medium of exchange function of various cost or yield functions associated with different money, or of the conditions under which inside money assets or positions is required by both problems. The becomes a component of wealth, was obstructed by the blandness of the New View’s standard program cannot programmatic state of the New View. The useful analysis cope with these issues. The reader may consult a of the medium-of-exchange function depends on a decisive preliminary approach to the analysis of the medium of ex- rejection of the assertion that “everything is almost like change function in the paper by Karl Brunner and Allan H. everything else.” This analysis requires proper recognition Meltzer, in the Journal of Finance, 1964, listed in footnote that the marginal cost of information concerning qualities 9. He should also consult for both issues the important and properties of assets differs substantially between book by Boris Pesek and Thomas Saving. Money, Wealth assets, and that the marginal cost of readjusting asset and Economic Theory, The Macmillan Company, New positions depends on the assets involved. The analysis of York, 1967, or the paper by Harry Johnson, ‘Inside the wealth position of inside money requires recognition of Money, Outside Money, Income, Wealth and Welfare in the marginal productivity of inside money to the holder. Monetary Theory,” to be published in the Journal of Adequate attention to the relevant differences between Money, Credit and Banking, December 1968.

SEPTEMBER/OCTOBER 1989 8 the one side and haphazard pieces of unfinished tion about the response of bank credit, money analysis on the other side.9 stock and time deposits to changes in ceiling rates, or to changes in the speed with which A related failure accompanies this logical er- banks adjust their deposit-supply conditions to ror. There is not the slightest attempt to assess evolving market situations. Every single aspect alternative hypotheses or theories by systematic of the banks’ or the public’s behavior emphasiz- exposure to observations from the real world. It ed by the countercritique has been analyzed by follows, therefore, that the countercritique the Monetarists hypotheses in terms which scarcely analyzed the empirical theories advanc- render the results empirically assessable. Little ed by the Monetarist critique and consequently remains, consequently, of the suggestive failed to understand the major implications of countercritique assembled in the papers by these theories. Gramley-Chase, Kareken and Cacy.11 For instance, they failed to recognize the role assigned by the Monetarist view to banks’ be- B, A Monetarist Examination of the havior and the public’s preferences in the New View’s Money Supply Theory monetary process. The objection raised by the New View that “the formula [expressing a basic Three sources of the conflict have been dis- framework used to formulate the hypothesis] cussed thus far. Two sources were revealed as obscures the important role played by the logical misconstruals, involving inadequate con- public” has neither analytical basis nor meaning. struction and assessment of empirical theories. In fact, the place of the public’s behavior was A third source pertains to legitimate differences discussed in the Monetarist hypotheses in some in research strategy. These three sources do not detail. Moreover, the same analysis discussed explain all major aspects of the conflict. Beyond the conditions under which the public’s the differences in research strategy and logical behavior dominates movements of the money misconceptions, genuinely substantive issues re- stock and bank credit.b0 It also yielded informa- main. Some comments of protagonists advocating °Asexamples of the empirical work performed by the posure to observations, I would have to withdraw my Monetarists, the reader should consult the following works: statements. A detailed analysis of the banks’ and the Milton Friedman and Anna Jacobson-Schwartz. A public’s role in the money supply, based on two different Monetary History of the United States, 1867-1960, hypotheses previously reported in our papers will be (Princeton: Princeton University Press, 1963). Philip developed in our forthcoming books. This analysis, by its Cagan, Determinants and Effects of Changes in the Stock very existence, falsifies some major objections made by of Money, (Columbia: Columbia University Press, 1965). Cacy or Gramley-Chase. Much of their criticism is either Karl Brunner and Allan H. Meltzer, “Some Further In- innocuous or fatuous. Gramley-Chase indulge, for in- vestigations of Demand and Supply Functions for Money,” stance, in modality statements, i.e. statements obtained Journal of Finance, Volume XIX, May 1964. Karl Brunner from other statements by prefixing a modality qualifier like and Allan H. Meltzer, “A Credit-Market Theory of the “maybe” or “possibly.” The result of qualifying an em- Money Supply and an Explanation of Two Puzzles in U.S. pirical statement always yields a statement which is Monetary Policy,” Essays in Honor of Marco Fanno, 1966, necessarily true but also quite uninformative. The modality Padova, Italy. Karl Brunner and Robert Crouch, “Money game thus yields logically pointless but psychologically ef- Supply Theory and British Monetary Experience,” Methods fective sentences. Cacy manages, on the other hand, of Operations Research Ill—Essays in Honor of Wilhelm some astonishing assertions. The New View is credited Krelle, ed. Rudolf Henn (Published in Meisenheim, Ger- with the discovery that excess reserves vary over time. He many, by Anton Ham, 1966). Karl Brunner, “A Schema for totally disregards the major contributions to the analysis of the Supply Theory of Money,” International Economic excess reserves emanating from the Monetarists’ Review, 1961. Karl Brunner and Allan H. Meltzer, “An research. A detailed analysis of excess reserves was Alternative Approach to the Monetary Mechanism,” Sub- developed by Milton Friedman and Anna Schwartz in the committee on Domestic Finance, Committee on Banking book mentioned in footnote 9. The reader should also note and Currency, House of Representatives, August 17, 1964. the work by George Morrison, Liquidity Preferences of 10 The reader will find this analysis in the following papers: Commercial Banks, (Chicago: University of Chicago Press, 1966), and the study by Peter Frost, “Banks’ Demand for Karl Brunner and Allan H. Meltzer, “Liquidity Traps for Excess Reserves,” an unpublished dissertation submitted Money, Bank Credit, and Interest Rates,” Journal of to the University of California at Los Angeles, 1966. The Political Economy, April 1968. Karl Brunner and Allan H. classic example of an innocuous achievement was sup- Meltzer, “A Credit-Market Theory of the Money Supply plied by Cacy with the assertion: “. .-the actual volume and an Explanation of Two Puzzles in U.S. Monetary of money balances determined by competitive market Policy,” Essays in Honor of Marco Fanno, Padova, Italy, forces may or may not be equal to the upper limit 1966. established by the central bank” (p. 8). Indeed, we knew “The reader is, of course, aware that these assertions re- this before the New View or Any View, just as we always quire analytic substantiation. Such substantiation cannot knew that “it may or may not rain tomorrow.” The reader be supplied within the confines of this article. But the should note that similar statements were produced by reader could check for himself, If he finds, in the context other authors with all the appearances of meaningful of the countercritique, an analysis of the monetarists’ ma- elaborations. jor hypotheses, an examination of implication, and ex-

FEDERAL RESERVE BANK OF St LOUIS 9

the New View should probably be interpreted both bank credit and the money stock respond as conjectures about hypotheses to be expected positively to economic activity. The Gramley- from their research strategy. It should be clear- Chase model also implies that the responses of ly understood that such conjectures are not logi- both the money stock and bank credit to cal implications of the guiding framework. tn- monetary actions are independent of the stead, they are pragmatic responses to the gen- general scale of the public’s and the banks’ in- eral emphasis associated with this approach. terest elasticities. Uniformly large or small in- terest elasticities yield the same response in the A first conjecture suggests that the money money stock or bank credit to a change in the stock and bank credit are dominated by the monetary base. public’s and the Banks’ behavior. It is suggested, therefore, that cyclical fluctuations of monetary A detailed discussion of the implications growth result primarily from the responses of derivable from a meaningfully supplemented banks and the public to changing business con- Gramley-Chase model is not necessary at this ditions. A second conjecture naturally supple- point. We are foremost interested in the rela- ments the above assertions. It is contended that tion between this model and the propositions the money stock is a thoroughly “untrustworthy mentioned in the previous paragraph. The first guide to monetary policy.’ proposition can be interpreted in two different ways. According to one interpretation, it could Articles by Gramley-Chase and Kareken at- tempt to support these conjectures with the aid mean that the marginal multipliers g, and h of more explicit analytical formulations allegedly (i = 1, 2) are functions of the banks’ and the public’s response patterns expressing various expressing the general program of the New types of substitution relationships between dif- View. The paper contributed by Gramley-Chase has been critically examined in detail on ferent assets. This interpretation is, however, another occasion, and only some crucial aspects quite innocuous and yields no differentiation relative to the questioned hypotheses of the relevant for our present purposes will be con- sidered at this point.’2 Various aspects of the Monetarist position. first conjecture are examined in this and the A second interpretation suggests that the next section. The second conjecture is examined growth rate of the money stock is dominated by in sections D and E. the second component (changes in income) of the differential expression: A detailed analysis of the Gramley-Chase model demonstrates that it implies the following AM = g, AB’ + g AY reduced form equations: 3 This result is not actually implied by the M = g(B’, Y, c) 3 Gramley-Chase model, but it is certainly consis- E = h(B°,Y, c) h>o>h,, and h1 >g,’ tent with the model. However, in order to explaining the money stock (M) and bank credit derive the desired result, their model must be (E) in terms of the extended monetary base (B’), supplemented with special assumptions about the level of economic activity expressed by na- the relative magnitude of g, and g2, and also tional income at current prices (Y), and the ceil- about the comparative cyclical variability of AB’ ing rate on time deposits (c).’~ and AY. This information has not been provided by the authors. The Gramley-Chase model implies that mone- tary policy does affect the money stock and Most interesting is another aspect of the bank credit. It also implies that the money stock model which was not clarified by the authors. responds positively and bank credit negatively to Their model implies that policymakers could economic activity. The model thus differs from easily avoid procyclical movements in AM. This the Monetarist hypotheses which imply that model exemplifying the New View thus yields

2 13 ‘ The reader may consult my chapter “Federal Reserve 1n the Gramley-Chase model, 93 and h, are indeterminant. Policy and Monetary Analysis” in Indicators and Targets of ‘~Thisimplication was demonstrated in my paper listed in Monetary Policy, ed., by Karl Brunner, to be published by footnote 12. The monetary base is adjusted for the ac- Chandler House Publishing Co., San Francisco. This book cumulated sum of reserves liberated from or impounded in also contains the original article by Gramley-Chase. Fur- required reserves by changes in requirement ratios. ther contributions by Patric H. Hendershott and Robert Wemntraub survey critically the issues raised by the Gramley-Chase paper.

SEPTEMBER/OCTOBER 1989 10 little justification for the conjectures of its analysis of firm behavior developed on the first proponents. level of his investigation. A central property of the Gramley-Chase This disregard for the construction of an model must be considered in the light of the economic theory relevant for the real world is programmatic statements characterizing the carried into the second level of analysis where New View. Gramley-Chase do not differentiate the author formulates a system of relations between the public’s asset supply to banks and describing the joint determination of interest the public’s demand for money. This procedure rates, bank credit, and money stock. A remark- violates the basic program of the New View, able feature of the Karenken model is that it namely, to apply economic analysis to an array yields no implications whatsoever about the of financial assets and financial institutions. response of the monetary system to actions of Economic analysis implies that the public’s asset the Federal Reserve. It can say nothing, as it supply and money demand are distinct, and not stands, about either open market operations or identical behavior patterns. This difference in about discount rate and reserve requirement ac- behavior patterns is clearly revealed by dif- tions. This model literally implies, for instance, ferent responses of desired money balances and that the money stock and the banking system’s desired asset supply to specific stimuli in the en- deposit liabilities do not change as a result of vironment. For instance, an increase in the ex- any change in reserve requirements ratios. pected real yield on real capital raises the None of the conjectures advanced by the public’s asset supply but lowers the public’s countercritique concerning the behavior of the money demand. It follows thus that a central money stock and the role of monetary policy analytical feature of the Gramley-Chase model find analytical support in Karenken’s analysis. violates the basic and quite relevant program of To the extent that anything is implied, it would the New View. imply that monetary policy operating directly on Karenken’s construction shares this fundamen- bank reserves or a mysterious rate of return on tal analytical flaw with the Gramley~Chase reserves dominates the volume of deposits—a model, but this is not the only problem faced by practically subversive position for a follower of his analysis. The Karenken analysis proceeds on the New View.” two levels. First, he derives a representative bank’s desired balance sheet position. For this C. Alternative Explanations of purpose he postulates wealth maximization sub- Cyclical Fluctuations in Monetary ject to the bank’s balance sheet relation be- Growth tween assets and liabilities, and subject to reserve requirements on deposits. On closer ex- The examination thus far in this article has amination, this analysis is only applicable to a shown that even the most explicit formulation monopoly bank with no conversion of deposits (Gramley-Chase) of the countercritique, allegedly into currency or reserve flows to other banks. representing the New View with respect to In order to render the analysis relevant for a monetary system analysis, does assign a signifi- representative bank in the world of reality, ad- cant role to monetary policy. This examination ditional constraints would have to be introduced also argued that the general emphasis given by which modify the results quite substantially. It the New View to the public’s and the banks’ is also noteworthy that the structural properties behavior in determination of the money stock assigned by Karenken to the system of market and bank credit does not differentiate its pro- relations are logically inconsistent with the im- duct from analytical developments arising from plications one can derive from the author’s the Monetarist approach. It was also shown that

5 ‘ Two direct objections made to the Brunner-Meltzer because an “analytical underpinning” has been provided. analysis by Kareken should be noted. He finds that the Karenken also finds fault with our use of the term “money questioned hypotheses do not contain “a genuine supply supply function.” Whether or not one agrees with this ter- function” of deposits. Accepting Karenken’s terminology, minological preferences surely does not affect the relation this is true, but neither does the Gramley-Chase model between observations and statements supplied by the contain such a supply function. But the objection has no hypothesis. And it should be clear that the status of evidential value anyway. If a hypothesis were judged un- hypothesis depends only on this relation, and not on satisfactory because some aspects are omitted, all names attached to statements. hypotheses are “unsatisfactory.” Moreover, the cognitive status of a empirical hypothesis does not improve simply

FEDERAL RESERVE BANK OF St LOUIS 11 the only explicit formulation advanced by the The various conjectures advanced by Gramley- New Viewers does not provide a sufficient basis Chase, Cacy, and Davis in regard to causes of for their central conjectures. It is impossible to movements in money and bank credit can be derive the proposition from the Gramley-Chase classified into two groups. One set of conjec- model that the behavior of the public and tures traces the mechanism generating cyclical banks, rather than Federal Reserve actions, fluctuations of monetary growth to the re- dominated movements in the money supply. But sponses of banks and the public; the behavior the declaration of innocence by the countercriti- of monetary authorities is assigned a com- que on behalf of the monetary authorities with paratively minor role. The other group of con- respect to cyclical fluctuations of monetary jectures recognizes the predominant role of the growth still requires further assessment. behavior of monetary authorities. The detailed arguments advanced to explain In the following analysis the framework pro- the observed cyclical fluctuations of monetary vided by the Monetarist view will be used to growth differ substantially among the con- assess these conflicting conjectures. The em- tributors to the countercritique. Gramley-Chase phasis concerning the nature of the causal maintain that changing business conditions mechanisms may differ between the various modify relative interest rates, and thus induce conjectures regarding sources of variations in countercyclical movements in the time deposit money, but the following examination will be ratio. These movements in demand and time applied to an aspect common to all conjectures deposits generate cyclical fluctuations in emphasizing the role of public and bank monetary growth. On the other hand, Cacy behavior. develops an argument used many years ago by Wicksell and Keynes, but attributes it to the In the context of the Monetarist framework, New View. He recognizes a pronounced sen- the money stock (M) is exhibited as a product of sitivity of the money stock to variations in the a multiplier (m) and the monetary base (B), public’s money demand or asset supply. These (such that M = mB). This framework, without variations induce changes in credit market con- the supplementary set of hypotheses and theo- ditions. Banks, in turn, respond with suitable ries bearing on the proximate determinants of adjustments in the reserve and borrowing money summarized by the multiplier and the ratios. The money stock and bank credit conse- base, is completely neutral with respect to the quently change in response to this mechanism. rival conjectures; it is compatible with any set of observations. This neutrality assures us that Davis actually advances two radically different its use does not prejudge the issue under con- conjectures about causes of cyclical fluctuations sideration. The Monetarist framework operates of monetary growth. The first conjecture at- in the manner of a language system, able to ex- tributes fluctuations of monetary growth to the press the implications of the competing conjec- public’s and banks’ responses. Changing busi- tures in a uniform manner. ness conditions modify the currency ratio, the banks’ borrowing ratio, and the reserve ratio. The first group of conjectures advanced by The resulting changes generate the observed the countercritique (behavior of the public and movements in money. His other conjecture at- banks dominates movements in money) implies tributes fluctuations in monetary growth to that variations in monetary growth between Federal Reserve actions: “the state of business upswings and downswings in business activity influences decisions by the monetary authorities are dominated by the variations in the mone- to supply reserves and to take other actions tary multiplier. The second group (behavior of likely to affect the money supply.”” monetary authorities dominates movements in

6 ‘ Davis, p. 66. One argument about monetary policy in the sense correct. But the quote could easily be misinter- same paper requires clarification. Davis asserts on p. 68 preted due to the ambiguity of the term “policy.” This that the money supply need not be the objective of policy, term is frequently used to designate a strategy guiding the and “given this fact, the behavior of the rate of growth of adjustment of policy variables. Is is also frequently used to the money supply during the period cannot be assumed to refer to the behavior of the policy variables or directly to be simply and directly the result of monetary policy deci- the variables as such. The quote is quite acceptable in the sions alone.” This quote asserts that the money supply is first sense of ‘policy,” but thoroughly unacceptable in the “simply and directly the result of policy alone” whenever second sense. policy uses the money supply as a target. This is in a

SEPTEMBER/OCTOBER 1929 12 money) implies that, in periods with unchanged reserve requirement ratios and ceiling rates on time deposits, variations in the monetary base dominate cyclical changes in monetary growth. The movements of the monetary multiplier which are strictly attributable to the changing of requirement ratios can be separated from the total contribution of the multiplier and com- bined with the monetary base. With this adjust- ment, the second group of conjectures implies that the monetary base, supplemented by the contribution of reserve requirement changes to the multiplier, dominates variations in the money stock. In this examination of contrasting explanations of monetary fluctuations, values of the money stock (M), the multiplier (m), and the monetary base adjusted for member bank borrowing (B) are measured at the initial and terminal month of each half business cycle (i.e., expansions and contractions) located by the National Bureau of Economic Research. We form the ratios of these values and write: whole sample period, whereas ~ (p. a) is only mB .084. The half-cycle from 1929 to 1933 was ~ ~~2;or l~= a/3 M m B omitted in the computations, because move- 0 0 0 ments in the money stock and the multiplier The subscript I refers to values of the terminal were dominated by forces which do not dis- month and the subscript 0 to values of the in- criminate between the rival conjectures under itial month. ‘I’hese ratios were measured for consideration. The sample period, including 1929 to 1933, still yields a substantially larger each half cycle in the period March 1919 to De- cember 1966. They were computed for two def- value for Q(p, /3). The same pattern also holds initions of the money stock, inclusive and for subperiods. In particular, computations bas- exclusive of time deposits, with corresponding ed on observations for 1949 to 1966 confirm monetary multipliers. the pattern observed for the whole sample period. The results thus support the second Kendall’s rank correlation coefficients between group of conjectures but not the first group. the money stock ratios (j4 and the multiplier These results also suggest, however, that forces ratios (a), and between (p) and the monetary operating through the multiplier are not quite base ratio (/3) were computed. We denote these negligible. The surprisingly small correlation Q(p, correlation coefficients with ~ (hi, a) and ~ (i-i, /3). a) does not adequately reveal the operation of The implications of the two rival conjectures these forces. Their effective operation is revealed can now be restated in terms of the two coeffi- by the correlation Q(p, /3), which is far from cients. The first group of conjectures implies perfect, even in subperiods with constant re- that Q(p, a) > Q(J1, /3); while the second group serve requirement ratios. This circumstance implies that in periods of unchanged reserve re- suggests that the behavior of the public and quirements ratios and ceiling rates on time banks contributes to the cyclical movements of deposits, the coefficient q(pt~,/3) exceeds the coef- monetary growth. The main result at this stage, ficient Q(p, a). The second group implies nothing however, the clear discrimination between the about the relation of the two coefficients in two groups of conjectures. The results are quite periods of changing reserve requirements and unambiguous on this score. ceiling rates on time deposits. It follows, therefore, that observations yielding the ine- Additional information is supplied by table I. For each postwar cycle beginning with the quality Q(~1,/3 > Q(p, a) disconfirm the first group and confirm the second group. downswing of 1948-49, the average annual growth rate of the money stock was computed.

The correlations obtained are quite unam- The expression M = mb was then used to com- biguous. The value of ~ (p. /3) is .537 for the pute the contribution to the average growth

FEDERAL RESERVE BANK OF St LOUIS 13

//~ ~ ~ /~ âIaMitasc ~ ~a*taaii~i: ~7 ~ ‘~aC~4~2 ~

;~~z ~ :: /~/ ~ ~:J~T:~d7 //4A~/ ~ / ~/ / ~//NN~ // ~~/ ~ / 7~ / ~ ;:~~> /7\ ~ / :~:

/ ~ / ~~~ ~ “ c~ç C K~4 ~~~

~/ //~ ~ ~/< c ~~\,‘ / ~%/~//~ ~ //\~ /fl ‘ ~ _r/~ ~ ~// ~/, //;~ ~ ~c ~ ~ ~N7~ ~r ~/~‘~/‘~~:~~~ ~ ~/4 /~~/\ ~~/ // ~‘ /~ /~~\/~~<~ ~/tc/~ ~ ~/> / C’~~/~//~c ~, ~/ t~/ % ~ // ~, ) r~ ~ ~ /~/ /// A~ ~ ‘/ ~ <~/,/~c ~/~/ ~ ~Ac/~ ~~7,// / ~ ~ \>/ // / ~ ~ / ,c~ ~ ~ / ~~ /~~v t ~ / ~\ /~ \ N // ~/ ///~/// ~ ~~ ~ c~/& ~ -\~ ~// ~/~/ /~7//~\/ ~ // ,/ ~ /\~~// ~ ~ ~/ ~/ -/ ~* 4 /: ~ // ~‘t ~ ~~~~~ _~/&/~/~\~ ~~/“~ ~~~ ~~ //~ /// /~/~// / /\ ~ \~/~ / // / g~ ,~ ~ ~//~A \\ ~ v ~~~~~/ ‘~~, -\ ~ ~ < / < / \ /~ / ‘~/~/~‘/ ~rk~/’ ~ ~ ~ \ ~* ;~t 4~ ~ ~ ~// ~ ~ ~ 4~/ t~~r ~~

,t ~ ~~/// ~ ~ ~ // // ~ / \,/ N

///~/ / ~~ ~/ / 4\‘ / ~ / 7 / ~/ /~/~ 7~/~\ ~ ~ —\ /~ ~~//\~/77~ ‘m~7~/, / /~ , /7/~~ N ?; 1~:~4~:T~~ N ~~/~/// ~~7/

rate of money from three distinct sources: (i) It appears that bank credit is comparatively less the behavior of monetary authorities (i.e., the exposed to the push of Federal Reserve actions monetary base and reserve requirement ratios), than was the money stock. On the other hand, and the public’s currency behavior, (ii) the time the money stock is less sensitive than bank deposits substitution process, and, (iii) the varia- credit to the time-deposit substitution tions in the excess reserve and borrowing ratios mechanism emphasized by Gramley-Chase, and of commercial banks (Wicksell-Keynes mecha- the Wicksell-Keynes mechanism suggested by nism). Cacy. Most astonishing, however, is the negative association between the average growth i-ate of The rank correlations between each contribu- bank credit and the Wicksell-Keynes mechanism tion, and the average growth rate of the money stock over all postwar half-cycles clearly sup- emphasized by Cacy. port the conclusion of the previous analysis that It should also be noted that the average cyclical movements in the money stock are growth rate of money conforms very clearly to dominated by Federal Reserve actions. the business cycle. Such conformity does not hold for bank credit over the postwar half- Table I also presents the results of a similar cycles. This blurring occurred particularly in examination bearing on causes of movements in periods when the ceiling rate on time deposits bank credit. The reader should note the radical was increased. These periods exhibit relatively difference in the observed patterns of correla- large contributions to the growth rate of bank tion coefficients. The behavior of monetary credit emanating from the time deposit substitu- authorities, supplemented by the public’s cur- tion mechanism. rency behavior, does not appear to dominate the behavior of bank credit. The three sources A regression analysis (table II) of the reduced contributing to the growth rate of money all cx- form equations derived from the Gramley-Chase erted influences of similar order on bank credit. model confirms the central role of the monetary

SEPTEMBER/OCTOBER 1989 14 base in the money supply process. Estimates of the regression coefficient relating money to in- come are highly unstable among different sam- Table Ill ple periods, relative to the coefficient relating money to the monetary base. Furthermore, estimates of regression coefficients relating money to income occur in some periods with signs which contradict the proposition of Gramley-Chase and Cacy, or exhibit a very small statistical significance. These diverse patterns of coefficients do not occur for the estimates of coefficients relating money and the monetary base. It is also noteworthy that the average .24 growth rate of the monetary base (adjusted for 61 changes in reserve requirement ratios), over the 71 upswings, exceeds without exception the 45 average growth rate of adjacent downswings. This observation is not compatible with the con- tention made by Gramley-Chase that policy is countercyclical. Additional information is supplied by table III, which presents some results of a spectral analy- sis bearing on the monetary base and its sources. Spectral analysis is a statistical procedure for decomposing a time series into seasonal, cyclical, and trend movements. After such an analysis was conducted on the monetary base and its sources, a form of correlation analysis was run between movements in the monetary base and movements in its various sources. The results of this procedure (table III) indicate that (n is lhe lCevnesrun conreplion (not Lii lu run- movements in Federal Reserve credit dominate I LNCd \\ tb Kt’_~IWS ~1fl\ I enshrined in standard seasonal and cyclical movements in the mone- horinuIatuon.~ot the inrome~expenditure(name— tary base. n ork. In We view, the interest rate h the main

In summary, preliminary investigations yield link heRs een nrone\ and economic acli’ it) - 1 he no support for the contention that the behavior tither Vii’n rejects the traditional neparatuon ol 111 of banks and the public dominates cyclical move- ei’iir~(ini~it[lie )I\ iliti.) ft 5 rthtilrthl income ments in the money stock. The conjectures ad- ,ir~iI~’.is(nNi(l—l) (ci)E~oi~iicSJ;iniii j)r’ite l111’liI\ vanced by Gramley-Chase or Cacy are thus (rtiitio (~l()nl)ll)j(—’,). \ccorcling li~Ihirs r.itlu’i ~ disconfirmed, whereas Davis’ second conjecture lntIll)lrl :inrij t’tril)Il)_~rn~’nit~iii’ {‘\hIIaIried h~a —~trit— that fluctuations in monetary growth may be at- able applirarinn of relati~ u~~’theory. With tributed to Federal Reserve actions seems sub- r-eLfar-l.l to clisc-ii,siniis 1)1 tirt inripart ol morie~ stantially more appropriate. However, further anici rrii,nretzir~aclior~sor~‘rl)Irl)riri(- arti~it~ this investigations are certainly useful. latter den ha’~been termed the \lunt’larisl po’itioui. Ilus ~05ilh0t1 rims he dWided into the EL Relevance of Money and n eak \lnneL i-RE thesis anti the ~Lrong Monetary Actions with Respect to \honetarist lhe~j~.In a ‘,ense, both the Nm \ ren and the \honetarkt extension ol he tradi— Economic Activity tional en are represenled in tire n ak At present, a broad consensus accepts the \honetarisl po~ilion). relevance of money and monetary policy with ibm lohlii~xiO~4 (ii~r1Is~i1,r~¼clix 1101) [lii’ xvuak respect to economic activity. But this consensus and We ~l rung \lrrnetarist thesk Fbi xx enk concerning the relevance of money emerges thi’’,i’, is cr)nrhiarecl xx ith ‘~rjn3ieaspecI~ol lire from two substantially different views about the illrlirnr(’.e\llenlirtnrre ~rpr)rI)arl~Ii’ th’ lintrrrrrir1a— nature of the transmission mechanism. One turn cit nratioiral ernn~omniraclidtx. tl1r ~trninrg

FEDERAL RESERVE BANK OF ST. LOUIS 15 thesis supplements the weak thesis with special a Keynesian framework, in spite of substantial assumptions about our environment, in order to accumulation of outstanding government debt establish the role of monetary forces in the when a budget deficit continually occurs. And business cycle. lastly, the operation of interest rates on invest- 1, The Weak Monetansi Thesis ment decisions has usually been rationalized with the aid of considerations based on the ef- According to the %veak Monetarist thesis, fects of borrowing costs. monetary impulses are transmitted to the economy by a relative price process which These aspects of the income-expenditure ap- operates on money, financial assets (and liabili- proach may be evaluated within the framework ties), real assets, yields on assets and the produc- of the weak Monetarist thesis. The effects of tion of new assets, liabilities and consumables. fiscal actions are also transmitted by the relative The general nature of this process has been price mechanism. Fiscal impulses, i.e., govern- described on numerous occasions and may be ment spending, taxing, and borrowing, operate interpreted as evolving from ideas developed by just as “indirectly” as monetary impulses, and Knut, Wicksell, Irving Fisher and John Maynard there is no a priori reason for believing that Keynes.17 their speed of transmission is substantially The operation of relative prices between greater than that of monetary impulses. The money, financial assets and real assets may be relative price conception of the transmission equivalently interpreted as the working of an mechanism also implies that a constant budget interest rate mechanism (prices and yields of deficit exerts a continuous influences on eco- assets are inversely related). Monetary impulses nomic activity through persistent modifications are thus transmitted by the play of interest in relative prices of financial and real assets. rates over a vast array of assets. Variations in Lastly, the transmission of monetary impulses is interest rates change relative prices of existing not dominated by the relative importance of assets, relative to both yields and the supply borrowing costs. In the process, marginal costs prices of new production. Acceleration or of liability extension interact with marginal re- deceleration of monetary impulses are thus con- turns from acquisitions of financial and real verted by the variation of relative prices, or in- assets. But interest rates on financial assets not terest rates, into increased or reduced produc- only affect the marginal cost of liability exten- tion, and subsequent revisions in the supply sion, but also influence the substitution between prices of current output. financial and real assets. This substitution modi- This general conception of the transmission fies prices of real assets relative to their supply mechanism has important implications which prices and forms a crucial linkage of the mone- conflict sharply with the Keynesian interpreta- tary mechanisms; this hnkage is usually omitted tion of monetary mechanisms expressed by in standard income-expenditure analysis. standard income-expenditure formulations.’~In the context of standard income-expenditure The description of monetary mechanisms in analysis, fiscal actions are considered to have a Davis’ article approaches quite closely the no- “direct effect” on economic activity, whereas tion developed by the weak Monetarist thesis. monetary actions are considered to have only This approximation permits a useful clarification and “indirect effect.” Furthermore, a constant of pending issues. However, the criticisms and budget deficit has no effect on interest rates in objections advanced by Davis do not apply to the

“The reader may consult the following studies on this the Transmission Mechanism,” Proceedings of the aspect: Milton Friedman and David Meiselman, “The American Economic Association, May 1963. Karl Brunner Relative Stability of Monetary Velocity and the Investment ‘The Relative Price Theory of Money, Output, and Multiplier in the United States, 1897-1958,” in Stabilization Employment,” unpublished manuscript based on a paper Policies, prepared by the Commission on Money and presented at the Midwestern Economic Association Credit, Englewood Cliffs, 1963. The paper listed in foot- Meetings, April 1967. 18 note 21 by James Tobin should also be consulted. Harry The paper on “The Effect of Monetary Policy on Expen- Johnson, ‘Monetary Theory and Policy,” American ditures in Specific Sectors of the Economy,” presented by Economic Review, June 1962. Karl Brunner ‘The Report Dr. Sherman Maisel at the meetings organized by the of the Commission on Money and Credit,” The Journal of American Bankers Association in September 1967, ex- Political Economy, December 1961. Karl Brunner, “Some emplifies very clearly the inherited Keynesian position. The Major Problems of Monetary Theory,” Proceedings of the paper will be published in a special issue of the Journal of American Economic Association, May 1961. Karl Brunner Political Economy. and Allan H. Meltzer, ‘The Role of Financial Institutions in

SEPTEMBER/OCTOBER 1989 18 weak Monetarist position. They are addressed to served association between accelerations and another thesis, which might be usefully labeled the decelerations of monetary forces and economic strong Monetarist thesis. activity depends on the relative magnitude of monetary accelerations (or decelerations). The Z The Strong Monetarist Thesis same analysis also reveals the crucial role of If the theoretical framework of the weak changes in the rate of change (second differ- Monetarist thesis is supplemented with addi- ences) of the money stock in explanations of tional and special hypotheses, the strong Mone- fluctuations in output and employment. It im- tarist thesis is obtained. An outline of the strong plies that any pronounced deceleration, occurr- thesis may be formulated in terms of three sets ing at any rate of monetary growth, retards of forces operating simultaneously on the pace total spending. It is thus impossible to state of economic activity. For convenience, they may whether any particular monetary growth, say a be grouped into monetary forces, fiscal forces 10 percent annual rate, is expansionary with and other forces. The latter include respect to economic activity, until one knows technological and organizational innovation, the previous growth rate. The monetary dynam- revisions in supply prices induced by accruing ics of the Monetarist thesis also explains the information and expectation adjustments, capital simultaneous occurrence of permanent price- accumulation, population changes and other inflation and fluctuations in output and employ- related factors or processes. ment observable in some countries. All three sets of forces are acknowledged by The nature and the variability of the “Fried- the strong thesis to affect the pace of economic man lag” may also be analyzed within the activity via the relative price process previously framework of the Monetarist thesis. This lag outlined. Moreover, the strong Monetarist point measures the interval between a change in sign of view advances the crucial thesis that the vari- of the second difference in the money stock and ability of monetary forces (properly weighted the subsequent turning point located by the Na- with respect to their effect on economic activi- tional Bureau. In general, the lag at an upper ty) exceeds the variability of fiscal forces and turning point will be shorter, the greater the other forces (properly weighted). It is argued absorption speed of the economy, and the further that major variabilities occurring in a sharper the deceleration of monetary impulses subset of the other forces (e.g., expectations and relative to the movement of fiscal forces and revisions of supply prices induced by informa- other forces. Variability in the relative accelera- tion arrival) are conditioned by the observed tion or deceleration of monetary forces neces- variability of monetary forces. The conjecture sarily generates the variability observed in the thus involves a comparison of monetary vari- Friedman lag. ability with the variability of fiscal forces and What evidence may be cited on behalf of the independent “other forces.” According to the strong Monetarist thesis? Every major inflation thesis under consideration, the variability of monetary of impulses is also large relative to provides support for the thesis, particularly in cases of substantial variations of monetary the speed at which the economy absorbs the impact of environmental changes. This predomi- growth. The attempt at stabilization in the Con- nance of variability in monetary impulses im- federacy during the Civil War forms an im- pressive piece of evidence in this respect. The plies that pronounced accelerations in monetary forces are followed subsequently by accelera- association between monetary and economic ac- tions in the pace of economic activity, and that celerations or decelerations has also been pronounced decelerations in monetary forces observed by the Federal Reserve Bank of St. are followed later by retardations in economic Louis.hhl Observations from periods with diver- gent movements of monetary and fiscal forces activity. provide further evidence. For instance, such The analysis of the monetary dynamics, using periods occurred immediately after termination the relative price process, is accepted by both of World War II, from the end of 1947 to the the weak and the strong Monetarist theses. This fall of 1948, and again in the second half of analysis implies that the regularity of the ob- 1966. In all three cases, monetary forces pre-

‘~U.S.Financial Data, Federal Reserve Bank of St. Louis, and Time Deposits, 1914-1964” in the September 1964 week ending February 14, 1968. Also see “Money Supply issue of this Review.

FEDERAL REBERVE BANK OF St LOUIS 17 vailed over fiscal forces. The evidence adduced The same logical property applies to Davis’ here and on other occasions does not “prove” second argument (b). ‘the timing relation ex- the strong Monetarist thesis, but does establish pressed by the Friedman lag, in particular the its merit for serious consideration. chronological precedence of turning points in monetary growth over turning points in eco- Davis’ examination is therefore welcomed. His nomic activity, can probably be explained by objections are summarized by the following the influence of business conditions on the points: (a) observations of the persistent associa- money supply. Studies in money supply theory tion between money and income do not permit strongly suggest this thesis and yield evidence an inference of causal direction from money to on its behalf. ‘The cyclical pattern of the curren- income; (b) the timing relation between money cy ratio, and the strategy typically pursued by and economic activity expressed by the Fried- monetary policymakers explain this lead of man lag yields no evidence in support of the monetary growth. And again, such explanation contention that variations in monetary growth of the timing relation does not bear negatively cause fluctuations in economic activity; (c) the on the strong conjecture. correlation found in cycles of moderate ampli- tude between magnitudes of monetary and eco- The objection noted under Davis’ point (c) is nomic changes was quite unimpressive; (d) the similarly irrelevant. His observations actually length of the Friedman lag does not measure confirm the strong thesis. The latter implies the interval between emission of monetary im- that the correlation between amplitudes of pulse and its ultimate impact on economic ac- monetary and income changes is itself cor- tivity. Furthermore, the variability of this lag is related with the magnitude of monetary ac- due to the simultaneous operation and interac- celerations or decelerations. A poor correlation tion of monetary and non-monetary forces. in cycles of moderate amplitude, therefore, yields no discriminating evidence on the validity Davis’ first comment (a) is of course quite true discriminating evidence on the validity of the and well known in the logic of science. It is im- strong thesis. Moreover, observations describing possible to derive (logically) causal statements or occurrences are more appropriate relative to any general hypotheses from observations. But the formulated thesis than correlation measures. we can use such observations to confirm or For instance, observations tending to disconfirm disconfirm such statements and hypotheses. the strong Monetarist thesis would consist of oc- Davis particularly emphasizes that the persistent currences of pronounced monetary accelera- association between money and income could tions or decelerations which are not followed by be attributed to a causal influence running from accelerated or retarded movements of economic economic activity to money. activity.

Indeed it could, but our present state of Point (d) still remains to be considered. Once knowledge rejects the notion that the observed again, his observation does not bear on the association is essentially due to a causal influ- strong Monetarist thesis. Davis properly cau- ence from income on money. Evidence refuting tions readers about the interpretation of the Friedman lag. The variability of this lag is pro- such a notion was presented in section C. The bably due to the interaction of monetary and existence of a mutual interaction over the non-monetary forces, or to changes from cycle shorter run between money and economic ac- to cycle in the relative variability of monetary tivity, however, must be fully acknowledged. growth. But again, this does not affect the Yet, this interaction results from the conception strong thesis. The proper interpretation of the guiding policymakers which induces them to ac- Friedman lag, as the interval between reversals celerate the monetary base whenever pressures in the rate of monetary impulses and their on interest rates mount, and to decelerate the prevalence over all other factors simultaneously monetary base when these pressures wane. Ad- operating on economic activity, usefully clarifies mission of a mutual interaction does not dispose a concept introduced into our discussions. This of the strong Monetarist thesis. This interaction, clarification provides, however, no relevant inherent in the weak thesis, is quite consistent evidence bearing on the questioned hypotheses. with the strong position and has no disconfirm- ing value. To the contrary, it offers an explana- In summary, the arguments developed by tion for the occurrence of the predominant Davis do not yield any substantive evidence variability of monetary forces. against the strong Monetarist thesis. Moreover,

SEPTEMBER/OCTOBER 1989 18 the discussion omits major portions of the aims, influences the monetary system and the 2 evidence assembled in support of this position. ° pace of economic activity. Thus, actual changes in the monetary base are quite meaningful and K Gountercvclical Polkv and the appropriate measures of actual behavior of Interpretation of Monetary Policy monetary authorities.~’ The usual assertion of the New View, at- The information presented in table IV sup- tributing fluctuations of monetary growth to the ports the conjecture that monetary policy- public’s and the banks’ behavior, assumed a makers’ interpretation of their own behavior strategic role in the countercritique. The coun- has no systematic positive association with their tercritique denied, furthermore, that monetary actual behavior. Table IV was constructed on actions have a major impact on economic activi- the basis of the scores assigned to changes in ty. With the crumbling of these two bastions, policies, according to the interpretation of the the monetary policymakers’ interpretation of Federal Open Market Committee.22 Positive their own behavior becomes quite vulnerable. scores were associated with each session of the In a previous section, the substantial contribu- FOMC which decided to make policy easier, tion of the monetary base to the fluctuations of more expansionary, less restrictive, less tight, monetary growth has been demonstrated. These etc., and negative scores indicate decisions to facts, combined with repeated assertions that follow a tighter) less expansionary, more restric- monetary policy has been largely counter- tive course. The scores varied between plus and cyclical, suggest the existence of a pronounced minus one, and expressed some broad ordering discrepancy between actual behavior of the of the revealed magnitude of the changes. monetary authorities and their interpretation of this behavior. An examination of the sequences of scores easily shows that the period covered can be A crucial question bearing on this issue per- naturally partitioned into subperiods exhibiting tains to the proper measure summarizing actual an overwhelming occurrence of scores with a behavior of the monetary authorities. Two ma- uniform sign. These subperiods are listed in the jor facts should be clearly recognized. First, the first column of table IV. The second column monetary base consists of “money” directly cumulated the scores over the subperiods listed issued by the authorities, and every issue of in order to yield a very rough ranking of the base money involves an action of the monetary policymakers’ posture according to their own authorities. This holds irrespective of their interpretation. knowledge about it, or their motivation and aims. Second, variations in the base, extended Table TV reveals that the FOMC interpreted by suitable adjustments to incorporate changing the subperiods from August 1957 to July 1958, reserve requirement ratios, are the single most and from July 1959 to December 1960 as important factor influencing the behavior of the among the most expansionary policy periods. money stock. And this second point applies ir- The period from November 1949 to May 1953 respective of whether Federal Reserve author- appears in this account as a phase of persistent- ities are aware of it or wish it to be, or ly tight or restrictive policy. The next two col- whatever their motivations or aims are. Their umns list the changes of two important vari- actual behavior, and not their motivations or ables during each subperiod. The third column 2 QMilton Friedman’s summary of the evidence in the Forty- as a chapter of a book edited by David Meiselman. 21 Fourth Annual Report of the National Bureau of Economic The reader may also be assured by the following Research is important in the respect. Davis overlooks in statement: particular the evidence accumulated in studies of the monetary policy refers particularly to determination of money supply mechanism which bears on the issue raised the supp1y of (the government’s) demand debt This by point (a) in the text. A persistent and uniform associa- demand debt coincides with the monetary base. The quote tion between money and economic activity, in spite of is by James Tobin, a leading architect of the New View, large changes in the structure of money supply processes, on p. 148 of his contribution to the Commission on Money yields evidence in support of the Monetarist theses. and Credit, “An Essay on Principles of Debt Manage- The reader should also consult Chapter 13 of the book by ment,” in Fiscal and Debt Management Policies, Prentice Milton Friedman and Anna Schwartz listed in footnote 9; Hall, Englewood Cliffs, 1963. 22 Studies in the Quantity Theory of Money, edited by Milton The scores were published as Appendix It to “An Alter- Friedman, University of Chicago Press, 1956; and a doc- native Approach to the Monetary Mechanism.” See foot- toral dissertation by Michael W. Keran, Monetary Policy note 9. and the Business Cycle in Postwar Japan,” Ph.D. thesis at the University of Minnesota, March 1966, to be published

FEDERAL RESERVE BANK OF St LOUIS 19

~

N \N ~N(~ N

tábIet\t\ N N / N The$soclMtOftbthleen P ky~nakers’1ntetpátatièrot~oiiçy, Chanies in ths MosSrBS and CftMtget~f’reaR~ánt. N

N C ScØt’eS NN N N cttanØsN1n’~bp

0 Po~cymakevt N NN N lntøptestt*n tP* ~ettoØIn tt~ø~PeSt N

PerIods tev~hePe8bd , htS~4I*O4tN NN NN NN N

eisattS4 4-US NNN~~N N + N 2184\CtM\ \\e~\N~ \~\ ~ 34~~ N ~ -+-tn ~Ssz

S5~ N~1N N’~44 N ~7NNN~

N iisi ~ N kS$GNN~ NN4-fl N N NN N~8~bs N aN N N NaN ~ N~ N4N N

15 ~ flN ~~N4fl ~X N ~N \~NN\N’N\/

N N NN NNN ~ N/~N N/ /~/ ~ N N describes changes in free reserves, and the the correlation between the policymakers’ de- fourth column notes changes in the monetary scriptions of their posture, and the movement base. A cursory examination of the columns im- of free reserves, is impressively close. This cor- mediately shows substantial differences in their relation confirms once again that the Federal broad association. The rank correlation between Reserve authorities have traditionally used the the various columns is most informative for our volume of free reserves as an indicator to gauge purposes. and interpret prevailing monetary policies. Yet little evidence has been developed which estab- These rank correlations are listed in table V. lishes a causal chain leading from changes in The results expose the absence of any positive association between the policymakers’ own in- free reserves to the pace of economic activity. terpretation or judgment of their stance and Another observation contained in table IV their actual behavior, as indicated by move- bears on the issue of policymakers interpreta- ments in the monetary base. The correlation tion of their own behavior. Changes in the coefficient between the monetary base and cumulated scores and free reserves between the cumulated scores has a negative value, sug- gesting that a systematic divergence between periods listed always move together and are perfect in terms of direction. By comparison, stated and actual policy (as measured by the the co-movement between cumulated scores and monetary base) is probable. On the other hand, changes in the monetary base is quite haphaz- ard; only three out of eight changes between periods move together. This degree of co-move- Table V ment between cumulated scores and the monetary base could have occurred by pure Rank Correlation Between Changes in chance with a probability greater than .2, the Monetary Base, Changes in Free whereas the probability of the perfect co- Reserves and the Cumulated Scores movement between cumulated scores and free of Policymakers’ Interpretations reserves occurring as a matter of pure chance is less than .004. The traditional selection of free reserves or money market conditions as an Cumulated scores and base -- 09 indicator to interpret prevailing monetary policy ~umL1latedscores and free reserves -- .70 Free reserves and base — .26 and to gauge the relative thrust applied by policy, forms the major reason for the negative 551 >,;1~4~7 \~// association (or at least random association) be- ;,//NNNN~~\~~N/N/, ~N~/N, / tween stated and actual policy.

SEPTEMBER/OCTOBER 1989 20

Attempts at rebuttal to the above analysis also be very misleading guides to monetary often emphasize that policymakers are neither policy. Thus, we can obtain a series of proposi- interested in the monetary base, nor do they at- tions about a vast array of entities, asserting tach any significance to it. This argument is ad- that each one can be a very misleading guide to vanced to support the claim that the behavior the interpretation of policy. We only reach a of the monetary base is irrelevant for a proper useless stalemate in this situation. examination of policymakers’ intended behavior. The ‘this argument disregards, however, the facts usual solution to the indicator problem at stated earlier, namely, movements in the mone- the present time is a decision based on mystical tary base are under the direct control and are insight supplemented by some impressionistic most frequently advanced argu- the sole responsibility of the monetary author- arguments. The ities. It also disregards the fact that actions may ments emphasize that central banks operate yield consequences which are independent of directly on credit markets where interest rates motivations shaping the actions. are formed, or that the interest mechanism forms the centerpiece of the transmission pro- These considerations are sufficient to cess. Accordingly, in both cases market interest acknowledge the relevance of the monetary rates should “obviously” emerge as the relevant base as a measure summarizing the actual indicator of monetary policy. behavior of monetary authorities. However, These arguments on behalf of market interest they alone are not sufficient to determine rates are mostly supplied by economists. ‘the whether the base is the most reliable indicator monetary authorities’ choice of money market of monetary policy. Other magnitudes such as conditions as an indicator evolved from a dif- interest rates, bank credit and free reserves ferent background. But in recent years a subtle have been advanced with plausible arguments change has occurred. One frequently encoun- to serve as indicators. A rational procedure ters arguments which essentially deny either must be designed to deter-mine which of the the existence of the indicator problem or its ra- possible entities frequently used for scaling tional solution. A favorite line asserts that “the policy yields the most reliable results. world is very complex” and consequently it is This indicator problem is still very poorly impossible or inadmissible to use a single scale understood, mainly because of ambiguous use to interpret policy. According to this view, one of economic language in most discussions of has to consider and weigh many things in order monetary policy. The term “indicator” occurs to obtain a “realistic” assessment in a compli- with a variety of meanings in discussions, and cated world. so do the terms “target” and “guide.” The in- ‘this position has little merit. The objection to dicator problem, understood in its technical a “single scale” misconstrues the very nature of sense, is the determination of an optimal scale the problem. Once we decide to discuss mone- justifying interpretations of the authorities’ ac- tary policy in term of comparative statements, tual behavior by means of comparative state- an ordinal scale is required in order to provide ments. A typical statement is that policy X is a logical basis for such statements. A multiplicity more expansionary than policy Y, or that cur- of scales effectively eliminates the use of coni- rent policy has become more (or less) expan- parative statements. Of course, a single scale sionary. Whenever we use a comparative con- may be a function of multiple arguments, but cept, we implicitly rely on an ordering scale. such multiplicity of arguments should not be confused with a multiplicity of scales. Policy- The indicator problem has not been given makers and economists should therefore realize adequate treatment in the literature, and the that one either provides a rational procedure recognition of its logical structure is often which justifies interpretations of monetary obstructed by inadequate analysis. It is, for in- policy by means of comparative statements, or stance, not sufficient to emphasize the proposi- that one abandons any pretense of meaningful tion that the money supply can be a “misleading or intellectually honest discussion of such guide to the proper interpretation of monetary policy. policy.” This proposition can be easily demon- strated for a wide variety of models and hy- Solution of the indicator problem in the potheses. However, it establishes very little. The technical sense appears obstructed on occasion same theories usually demonstrate that the rate by a prevalent confusion with an entirely dif- of interest, free reserves, or bank credit can ferent problem confronting the central banker—

FEDERAL RESERVE BANK OF St LOUIS 21 the target problem. This problem results from teract in the determination of bank credit, in- the prevaihng uncertainty concerning the na- terest rates, and the money stock, in response ture of the transmission mechanism and the to the behavior of monetary authorities. But the substantial lags in the dynamics of monetary recognition of such interaction implies nothing processes. with respect to the relative importance of the causal forces generating cyclical fluctuations of In the context of perfect information, the in- monetary growth. Neither does it bear on the dicator problem becomes trivial and the target quality of alternative empirical hypotheses, or problem vanishes. But perfect information is the the relative usefulness of various magnitudes or privilege of economists’ discourse on policy; cen- conditions which might be proposed as an in- tral bankers cannot afford this luxury. The im- dicator to judge the actual thrust applied by pact of their actions are both delayed and monetary policy to the pace of economic uncertain. Moreover, the ultimate goals of activity. monetary policy (targets in the Tinbergen-Theil sense) appear remote to the manager executing The Monetarist thesis has been put forth in general policy directives. Policymakers will be the form of well structured hypotheses which inclined under these circumstances to insert a are supported by empirical evidence. This ex- more immediate target between their ultimate tensive research in the area of monetary policy goals and their- actions. These targets should be has established that: (i) Federal Reserve actions reliably observable with a minimal lag. dominate the movement of the monetary base over time; (ii) movements of the monetary base It is quite understandable that central bankers dominate movements of the money supply over traditionally use various measures of money the business cycle; and, (iii) accelerations or market conditions, with somewhat shifting decelerations of the money supply are closely weights, as a target guiding the continuous ad- followed by accelerations or decelerations in justment of their policy variables. This response economic activity. Therefore, the Monetarist to the uncertainties and lags in the dynamics of thesis puts forth the proposition that actions of the monetary mechanism is very rational in- the Federal Reserve are transmitted to economic deed. However, once we recognize the rationali- activity via the resulting movements in the ty of such behavior, we should also consider monetary base and money supply, which initiate the rationality of using a particular target. The the adjustments in relative prices of assets, choice of a target still remains a problem, and liabilities, and the production of new assets. the very nature of this problem is inadequately understood at this state. The New View, as put forth by the counter- This is not the place to examine the indicator critique, has offered thus far neither analysis nor evidence pertaining relevantly to an ex- and target problem in detail. A possible solution to both problems has been developed on an- planation of variations in monetary growth. Moreover, the countercritique has not devel- other occasion.23 The solutions apply decision theoretic procedures and concepts from control oped, on acceptable logical grounds, a system- atic justification for the abundant supply of theory to the determination of an optimal choice of both indicator and target. Both pro- statements characterizing policy in terms of its blems are in principle solvable, in spite of the effects on the economy. Nor has it developed a “complexity of the world.” Consequently, there systematic justification for the choice of money is little excuse for failing to develop rational market conditions as an optimal target guiding monetary policy procedures. the execution of open market oper- ations. CONCLUSION But rational policy procedures require both a reliable interpretation and an adequate deter- A program for applying economic analysis to mination of the course of policy. The necessary financial markets and financial institutions is conditions for rational policy are certainly not certainly acceptable and worth pursuing. This satisfied if policies actually retarding economic program suggests that the public and banks in- activity are viewed to be expansionary, as in the

23 The reader may consult the chapter by Karl Brunner and ner. The book will be published by Chandler House Allan H. Meltzer on “Targets and Indicators of Monetary Publishing Co., Belmont, California. Policy,” in the book of the same title, edited by Karl Brun-

SEPTEMBER/OCTOBER 1989 22 case of the 1960-61 recession, or, if inflationary pretation of policy, and how do you actually ex- actions are viewed as being restrictive, as in the plain the fluctuations of monetary growth? The first half of 1968. major contentions of the academic critics of the past performance of monetary authorities could The major questions addressed to our mone- possibly be quite false, but this should be dem- tary policymakers, their advisors and consul- onstrated by appropriate analysis and relevant tants remain: How do you justify your inter- evidence.

FEOERAL RESERVE BANK OF St LOUIS