The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
FEDERAL RESERVE BANK OF RICHMOND MONTHLY REVIEW The Role o f the M oney Supply in the Conduct o f M onetary Policy Evolution o f the Concept of the Demand fo r M oney Volume 59 Number 12 DECEMBER 1973 The Role of the Money Supply in the Conduct of Monetary Policy The following letter, dated N ovem ber 6, 1973, by Arthur F. Burns, Chairman o f the Board o f Governors o f the Federal Reserve System, was written in response to inquiries by Senator P roxm ire regarding criticisms of m onetary policy during the past year. T he H onorable W illiam P roxm ire M any observers have blamed these difficulties on United States Senate W ashington, D. C. the management o f public econom ic policies. C er tainly, the Federal budget— despite vigorou s efforts to hold expenditures dow n— continued in substantial Dear Senator P r o x m ir e : deficit. I am w riting in further response to your letter of September 17, 1973, which requested comm ents on certain criticisms of monetary policy over the past year. T here has also been an enorm ous grow th in the activities of Federally-sponsored agencies which, although technically outside the budget, must still be financed. T he results of efforts to control wages and prices during the past year have been disappointing. A s stated in your letter, the criticisms a r e : ( 1 ) Partial decontrol in early 1973 and the subsequent “ that there was too much variation from time to time freeze failed to bring the results that were hoped for. in the rate of increase in the money supply, that M onetary policy has been criticized on somewhat monetary policy was too erratic, too much character contradictory counts— for being inflationary, or for ized by stops and starts’ ’ ; and (2 ) “ that the money permitting too high a level of interest rates, or for supply had increased much too much last year, in failing to bring the econom y back to full employment, fact that the increase would have been too much even if we had been in the depths o f a recession instead of or for permitting excessive short-term variations in enjoyin g a fairly vigorous econom ic expansion.” These criticisms involve basic issues with regard to the role of money in the econom y, and the role that the m oney supply should play in the form ula tion and execution of monetary policy. These issues, along with the specific points you raise, require careful examination. the grow th o f the money supply, and so on. One indication o f dissatisfaction with our public policies was provided by a report, to which you refer in your letter, on a questionnaire survey conducted by the National A ssociation o f Business Econom ists. O f the respondents, 38 per cent rated fiscal policy “ over the past year” as “ p o o r ” ; 41 per cent rated monetary policy “ over the past year” as “ p o o r” ; only 14 per cent felt that the w age-price controls under Criticism of Our Public Policies Phase I V were “ about right.” D uring the past tw o years the A m erican econom y If this sampling is at all indicative, the public policies on which we have has experienced a substantial measure o f prosperity. relied are being widely questioned. Real output has increased sharply, job s have been o f the above group, in fact, went on record for a created for millions of additional workers, and total significant change in fiscal policy. personal income-—both in dollars and in terms of question whether they favored a variable investment real purchasing pow er— has risen to the highest levels tax credit, 46.5 per cent said “ yes,” 40 per cent said ever reached. Y et the prosperity has been a troubled one. “ n o,” and 13.5 per cent expressed “ no opin ion .” increases have been large and widespread. M any members In response to a Price Let me turn now to the questions raised in your F or a letter and in some other recent discussions about time, the unemployment rate remained unduly high. monetary policy. Interest rates have risen sharply since the spring of role of money supply in the conduct o f monetary 1972. M ortgage money has recently becom e difficult to obtain in many comm unities. A n d confidence in p o lic y ; the extent and significance of variability in the dollar at home and abroad has at times wavered. havior of the money supply during 1972-73. 2 I shall discuss, in particular, the the grow th o f the m oney su p p ly ; and the actual be M ONTHLY REVIEW, DECEMBER 1973 Role of M oney Supply F or many years economists have debated the role o f the m oney supply in the perform ance of econom ic systems. O ne school of thought, often termed “ m one tarist,” claims that changes in the money supply in fluence very importantly, perhaps even decisively, the pace o f econom ic activity and the level of prices. M onetarists contend that the m onetary authorities should pay principal attention to the money supply, rather than to other financial variables such as inter est rates, in the conduct of monetary policy. T hey also contend that fiscal policy has only a small inde pendent impact on the econom y. econom ic disturbances tend to be self-correcting, and they therefore argue that a constant or nearly co n stant rate o f grow th of the m oney supply w ould result in reasonably satisfactory econom ic performance. But neither historical evidence, nor the thrust of explorations in business-cycle theory over a long century, give support to the notion that our econom y is inherently stable. O n the contrary, experience has demonstrated repeatedly that blind reliance on the self-correcting properties of our econom ic system can lead to serious trouble. D iscretionary econom ic p oli cy, while it has at times led to mistakes, has m ore often proved reasonably successful. T he disappear ance of business depressions, which in earlier times A nother school of thought places less emphasis on spelled mass unem ployment for w orkers and mass the m oney supply and assigns m ore im portance to bankruptcies fo r businessmen, is largely attributable the expenditure and tax policies of the Federal G ov ernment as factors influencing real econom ic activity to the stabilization policies of the last thirty years. and the level of prices. T his school emphasizes the need for monetary policy to be concerned with interest econom y tend of themselves to generate business rates and with conditions in the m oney and capital markets. Som e econom ic activities, particularly resi dential building and State and local governm ent co n struction, depend heavily on borrow ed funds, and are therefore influenced greatly by changes in the cost and availability of credit. In other categories o f spending— such as business investment in fixed capi tal and inventories, and consum er purchases o f dur able good s— credit conditions play a less decisive role, but they are nonetheless important. M onetarists recognize that monetary policy affects private spending in part through its impact on interest T he fact is that the internal w orkings of a market fluctuations, and m ost m odern economists recognize this. F or example, im proved prospects fo r profits often spur unsustainable bursts o f investment spend ing. T he flow of personal incom e in an age of afflu ence allows ample latitude fo r changes in discretion ary expenditures and in savings rates. D uring a business-cycle expansion various imbalances tend to develop within the econom y— between aggregate in ventories and sales, or between aggregate business investment in fixed capital and consum er outlays, or between average unit costs of production and prices. Such imbalances give rise to cyclical m ovements in the econom y. F lexible fiscal and monetary policies, rates and other credit terms. But they believe that prim ary attention to the grow th o f the m oney supply therefore, are often needed to cope with undesirable will result in a m ore appropriate monetary policy than w ould attention to conditions in the credit ished by the fact that our available tools o f econom ic stabilization leave something to be desired. markets. T here is general agreement am ong econom ists that, as a rule, the effects o f stabilization policies occur Needless to say, monetary policy is— and has long been— a controversial subject. Even the monetarists econom ic developments, and this need is not dim in do not speak with one voice on monetary policy. gradually over time, and that econom ic forecasts are an essential tool o f policy making. H ow ever, no Som e influential monetarists believe that monetary policy should aim strictly at maintaining a constant econom ist— or school o f econom ics— has a m onopoly on accurate forecasting. A t times, forecasts based rate of grow th of the m oney supply. H ow ever, what largely on the m oney supply have turned out to be that constant should be, or how broadly the money satisfactory. A t other times, such forecasts have been supply should be defined, are matters on which m one quite poor, mainly because o f unanticipated changes tarists still differ. in the intensity with which the existing m oney stock A n d there are also monetarists w ho w ould allow some— but infrequent— changes in the rate of grow th of the m oney supply, in accordance with changing econom ic conditions. is used by business firm s and consumers. Changes in the rate of turnover o f m oney have historically played a large role in econom ic fluctu It seems self-evident that adherence to a rigid ations, and they continue to do so. F or example, the grow th rate rule, or even one that is changed infre narrow ly-defined m oney stock— that is, demand d e quently, w ould practically prevent m onetary policy posits plus currency in public circulation— grew by from playing an active role in econom ic stabilization. 5.7 per cent between the fourth quarter o f 1969 and M onetarists recognize this. the fourth quarter o f 1970. T hey believe that most FEDERAL RESERVE BA N K OF RIC H M O N D But the turnover o f 3 m oney declined during that year, and the dollar value of G N P rose only 4.5 per cent. In the follow in g year, the grow th rate o f the m oney supply increased I recognize that one advantage o f maintaining a relatively stable grow th rate of the m oney supply is that a partial offset is thereby provided to unexpected to 6.9 per cent, but the turnover of m oney picked up briskly and the dollar value of G N P accelerated to and undesired shifts in the aggregate demand for 9.3 per cent. T h e m ovem ent out of recession in 1970 as to the em erging strength o f aggregate demand. into recovery in 1971 was thus closely related to the m oney grow th is maintained at a rather stable rate, greater intensity in the use o f money. good s and services. T here is always some uncertainty If O ccurrences and aggregate demand turns out to be weaker than such as this are very com m on because the willingness is consistent with the nation’s econom ic objectives, to use the existing stock of money, expressed in its interest rates will tend to decline and the easing o f rate of turnover, is a highly dynamic force in e co credit markets should help to m oderate the undesired weakness in demand. Similarly, if the demand fo r nom ic life. F or this as well as other reasons, the Federal Reserve uses a blend of forecasting techniques. T he goods and services threatens to outrun productive behavior of the m oney supply and other financial provide a restraining influence on the supply o f credit and thus tend to restrain excessive spending. variables is accorded careful attention. So also are the results o f the m ost recent surveys on plant and equipment spending, consumer attitudes, and inven capacity, a rather stable rate of m onetary grow th will H ow ever, it w ould be unwise for m onetary policy Recent trends in key producing and to aim at all times at a constant or nearly constant rate of grow th of m oney balances. T he m oney grow th spending sectors are analyzed. T he opinions o f busi rate that can contribute m ost to national objectives nessmen and outside econom ic analysts are canvassed, in part through the nationwide contacts o f Federal will vary with econom ic conditions. Reserve Banks. A n d an assessment is made of the probable course of fiscal policy, also o f labor-market usually weak, or if the demand fo r liquidity is unusu and agricultural policies, and their effects on the well above the desirable long-term econom y. needed for a time. tory plans. E vidence from all these sources is weighed. E fforts F o r exam ple, if the aggregate demand fo r good s and services is un ally strong, a rate of increase in the m oney supply trend m ay be A gain, when the econom y is experiencing severe cost-push inflation, a m onetary developments grow th rate that is relatively high by a historical through the use of large-scale econom etric models. A n eclectic approach is thus taken by the Federal yardstick may have to be tolerated fo r a time. If m oney grow th were severely constrained in order to R eserve, in recognition of the fact that the state of combat the element o f inflation resulting from such a econom ic know ledge does not justify reliance on any single forecasting technique. A s econom ic research cause, it might well have seriously adverse effects on production and employment. In short, what grow th rate of the m oney supply is appropriate at any given time cannot be determined simply by extrapolating are also made to assess econom ic has cumulated, it has becom e increasingly clear that money does indeed matter. But other financial vari ables also matter. In recent years, the Federal Reserve has placed somewhat m ore emphasis on achieving desired grow th rates of the monetary aggregates, including past trends or by som e preconceived arithmetical standard. M oreover, for purposes o f conducting monetary policy, it is never safe to rely on just one concept of the narrow ly-defined m oney supply, in its conduct m oney— even if that concept happens to be fashion o f monetary policy. able. But we have continued to give A variety o f plausible concepts merit careful careful attention to other financial indicators, am ong attention, because a number o f financial assets serve them the level o f interest rates on m ortgages and as a convenient, safe, and liquid store o f purchasing other loans and the liquidity position o f financial institutions and the general public. T his is neces sary because the econom ic implications of any given m onetary grow th rate depend on the state of liquidity, the attitudes of businessmen, investors, and co n power. T he Federal R eserve publishes data corresponding to three definitions of m oney, and takes all o f them into account in determining policy. T h e three m ea sures a re: ( a ) the narrow ly-defined m oney stock ( M i ) , which encompasses currency and demand de sumers tow ard liquidity, the cost and availability o f posits held by the nonbank p u b lic; borrow ed funds, and other factors. A lso , as the broadly-defined money stock ( M 2) , which also in nation’s central bank, the Federal R eserve can never cludes time and savings deposits at com m ercial banks (b ) a m ore lose sight of its role as a lender of last resort, so that (other than large negotiable time certificates o f de financial crises and panics will be averted. posits) ; ( c ) a still broader definition ( M 3) , which 4 M ONTHLY REVIEW, DECEMBER 1973 includes savings deposits at mutual savings banks and savings and loan associations. A definition em brac ing other liquid assets could also be justified— for example, one that would include large-denomination negotiable time certificates o f deposit, U . S. savings bonds and Treasury bills, com m ercial paper, and other short-term m oney market instruments. requirements set by the Federal Reserve. A s a result there is some slippage in monetary control. F urther m ore, since deposits at nonm em ber banks have been reported for only tw o to four days in a year, in co n trast to daily statistics for member banks, the data on the money supply— which we regularly present on a weekly, monthly, and quarterly basis— are esti There are many assets closely related to cash, and the public can switch readily am ong these assets. mates rather than precise measurements. W h en the infrequent reports from nonm ember banks becom e H ow ever m oney may be defined, the task of deter available, they often necessitate considerable revisions of the money supply figures. In the past tw o years, mining the amount of m oney needed to maintain high employment and reasonable stability o f the general price level is com plicated by shifting preferences o f the public for cash and other financial assets. the revisions were upward, and this may happen again this year. Som e indication of the extent of short-term vari ations in the recorded money supply is provided be Variability of M oney Supply Growth low. Table I shows the average and m axim um devi In the short-run, the rate o f change in the observed ations (w ithout regard to sign) o f M i from its aver m oney supply is quite erratic, and cannot be trusted as an indicator of the course of monetary policy. age annual grow th rate over a three and a half year This w ould be so even if there were no errors o f measurement. ation diminishes as the time unit lengthens; it is much larger for monthly than for quarterly data, and is also larger for quarterly than for semi-annual data. The record of hearings held by the Joint E conom ic Committee on June 27, 1973 includes a memorandum period. A s w ould be expected, the degree o f vari which I submitted on problem s encountered in co n Table I trolling the m oney supply. A s indicated there, weekto-week, m onth-to-m onth, and even quarter-to- DEVIATIONS IN M, FROM ITS AVERAGE RATE quarter fluctuations in the rate of change o f money OF GROWTH, 1970 THRU MID-1973 balances are frequently influenced by international Annual Rates of Change in per cent flow s o f funds, changes in the level of U. S. G overn A verage Deviation M axim um Deviation M onthly 3.8 8.8 Q uarterly 2.4 5.5 Because the demands of the public for money are subject to rather wide short-term variations, efforts Sem i-annual 1.8 4.1 by the Federal R eserve to maintain a constant grow th In our judgm ent, there need be little reason for concern about the short-run variations that occur in the rate of change in the m oney stock. Such vari ment deposits, and sudden changes in the public’s attitude towards liquidity. Som e of these variations appear to be essentially random— a product o f the enorm ous ebb and flow o f funds in our modern econ omy. rate of the money supply could lead to sharp shortrun swings in interest rates and risk damage to finan cial markets and the econom y. Uncertainties about financing costs could reduce the fluidity of markets Form of Data ations have minimal effects on the real econom y. F or one thing, the outstanding supply of m oney is very and increase the costs of financing to borow ers. In addition, wide and erratic movements o f interest large. It is also quite stable, even when the short-run rates and financial conditions could have undesirable outstanding supply of M i, seasonally adjusted, was effects on business and consum er spending. about $264 billion. These rate o f change is unstable. This O ctober the average O n this base, a monthly rise or adverse effects may not be o f m ajor dimensions, but fall in the money stock of even %2y2 billion would it is better to avoid them. amount to only a 1 per cent change. But when such In any event, for a variety o f reasons explained in a tem porary change is expressed as an annual rate, the mem orandum for the Joint E conom ic Committee, as is now com m only done, it com es out as about 12 to which I have previously referred, the Federal R e per cent and attracts attention far beyond its real serve does not have precise control over the money significance. supply. T o give one example, a significant part of the money supply consists o f deposits lodged in carefully the econom ic implications of variability in nonm ember banks that are not subject to the reserve M i grow th. T he Federal R eserve research staff has investigated T he experience of the past tw o decades FEDERAL RESERVE BANK OF RIC H M O N D 5 suggests that even an abnormally large or abnormally small rate of growth of the money stock over a period up to six months or so has a negligible influence on the course of the econom y— provided it is subse quently offset. Such short-run variations in the rate of change in the money supply may not at all reflect o f 1972. A lso, the grow th of M i, which on a m onthend basis appears very erratic in the first three quar ters of 1973, is much m ore stable on a quarterly average basis. F or example, while the level o f M i did not expand significantly between June and Sep tember, the quarterly average figures indicate further Federal Reserve policy, and they do not justify the attention they often receive from financial analysts. sizable grow th in the third quarter. T he thrust of monetary policy and its probable effects on econom ic activity can only be determined that the m oney available for use wras appreciably by observing the course o f the money supply and of other monetary aggregates over periods lasting six months or so. Even then, care must be taken to measure the grow th of money balances in ways that temper the influence of short-term variations. F or F or purposes o f econom ic analysis, it is an advantage to recognize larger in the third quarter than in the second quarter. Experience of 1972-73 D uring 1972, it was the responsibility o f the F ed eral Reserve to encourage a rate of econom ic expan sion adequate to reduce unemployment to acceptable example, the grow th of m oney balances over a quarter levels. can be measured from the amount outstanding in the last month of the preceding quarter to the last month of the curent quarter, or from the average amount outstanding during the preceding quarter to the aver effects of the w age-price control program , inflationary age in the current quarter. The first measure cap These objectives were to some extent conflicting, and tures the latest tendencies in the m oney supply, but may be distorted by random changes that have no lasting significance. T he second measure tends to monetary policy alone could not be expected to cope with both problem s. Continuation of an effective average out tem porary fluctuations and is comparable straint were urgently needed. The narrow ly-defined money stock increased 7.4 to the data provided on a wide range o f non-m onetarv econom ic variables, such as the gross national product and related measures. A com parison o f these tw o ways of measuring the A t the same time, despite the dampening pressures were gathering. M onetary policy, therefore, had to balance the twin objectives of containing infla tionary pressures and encouraging econom ic growth. w age-price program and a firm er policy o f fiscal re per cent during 1972 (m easured from the fourth quarter of 1971 to the fourth quarter of 1972). B e tween the third quarter of 1972 and the third quarter rate of grow th in M i is shown in Table II for succes sive quarters in 1972 and 1973. T he first colum n, of 1973, the grow th rate was 6.1 per cent. labeled M , shows annual rates calculated from endrnonths of quarters; the second colum n, labeled O . shows annual rates calculated from quarterly aver clined to 5.8 per cent, and a further slow ing occurred in the third quarter. ages. rates would require full analysis of the econom ic and financial objectives, conditions, and policies during the past tw o years, if not longer. Such an analysis Table II B y the first half of 1973, the annual grow th rate had d e Evaluation of the appropriateness of these growth GROWTH RATES OF MONEY SUPPLY cannot be undertaken here. Som e perspective on m onetary developments during 1972-73 may be O N TWO BASES gained, however, from com parisons with the exp eri Annual Rate of Change, in per cent M 1 9 .2 5 .3 1 1 1972 Q 6.1 8 .4 III 8 .2 8 .0 IV 8 .6 7 .1 ence of other industrial countries, and by recalling briefly how domestic econom ic conditions evolved during this period. Table III com pares the grow th of M i in the United States with that of other industrial countries in 1972 and the first half of 1973. T he definitions of M i differ somewhat from country to country, but are as 1 1 .7 4 .7 nearly com parable as statistical sources permit. I I 1 0 .3 6 .9 goes without saying that each country faced its own III 1973 0 .3 5.1 set of econom ic conditions and problems. It Y et it is useful to note that m onetary grow th in the United A s may be seen, the quarterly averages disclose States was much low er than in other m ajor industrial much m ore clearly the developing trend of monetary countries, and that it also was steadier than in the restraint— which, in fact, began in the second quarter other countries. 6 MONTHLY REVIEW, DECEMBER 1973 Table III ANNUAL PER CENT RATES OF GROWTH IN M O NEY SUPPLY 4tn Q uarter 1971 to 4th Q uarter 1972 United States 4th Quarter 1972 to 2nd Q uarter 1973 7.4 14.1 United K ingdom G e rm an y France Japan until Novem ber that the unem ployment rate dropped below Sl 2 per cent. F o r the year as a whole, the / unemployment rate averaged 5.6 per cent. It may be o f interest to recall that unem ployment averaged 5.5 per cent in 1954 and 1960, w hich are com m only regarded as recession years. 5.8 14.3 15.4 23.1 Since the expansion of M i in 1972 was low rela 10.0 4.2 8.7 28.2 tive to the demands for m oney and credit, it was accompanied by rising short-term interest rates. L ong-term interest rates showed little net change last year, as credit demands were satisfied mainly in the short-term markets. T he next table shows, in summary fashion, the rates of change in the money supply of the United States, in its total production, and in the consum er price level during 1972 and 1973. The table is based In 1973, the grow th o f M x moderated while the transactions demands for cash and the turnover of money accelerated. G N P in current dollars rose at a It may be noted, in passing, that, 12 per cent annual rate as prices rose m ore rapidly. In credit markets, short-term interest rates rose according to data available as late as January 1973, sharply further, while long-term interest rates also the rate of grow th of M i during 1972 was 7 .2 % , m oved up, though by substantially less than short term rates. on the latest data. not 7 .4 % ; and that the rate of increase in real G N P was 7 .7 % , not 7 .0 % . In other words, on the basis of the data available during 1972, the rate of growth of M i was below the rate of grow th o f the physical volume of over-all production. T he extraordinary upsurge of the price level this year reflects a variety o f special influences. First, there has been a w orld-w ide econom ic boom super im posed on the boom in the United States. Second, we have encountered critical shortages o f basic m a terials. T he expansion in industrial capacity needed Table IV MONEY SUPPLY, GNP, AND PRICES IN THE U. S. (Per cent change at annual rates) to produce these materials had not been put in place earlier because of the abnormally low level of profits between 1966 and 1971 and also because of num erous impediments to new investment on ecological grounds. 4th Quarter 1971 to 4th Q uarter 1972 M o n e y sup ply (M i) 7.4 4th Quarter 1972 to 2nd Quarter of 1973 3rd Quarter of 1973 5.8 5.6 11.7 4.8 12.1 5.4 Prices Consum er price index (CPI) CPI excluding food result of crop failures in many countries last year. Fourth, fuel prices spurted upward, reflecting the developing shortages in the energy field. A n d fifth, the depreciation of the dollar in foreign exchange G ross N atio n al Product Current dollars 10.6 Constant dollars 7.0 T hird, farm product prices escalated sharply as a markets has served to boost prices o f im ported goods and to add to the demands pressing on our prod u c tive resources. In view of these pow erful special factors, and the cyclical expansion of our econom y, a sharp advance in 3.4 7.1 7.8 our price level w ould have been practically inevitable in 1973. The upsurge o f the price level this year 3.0 4.0 4.1 hardly represents either the basic trend of prices or the response of prices to previous monetary or fiscal T he table indicates that grow th in M i during 1972 policies— whatever their shortcom ings may have been. and 1973 approxim ately matched the grow th o f real In particular, as the above table shows, the explosion output, but was far below the expansion in the dollar o f food prices that occurred this year is in large part value of the nation’s output. responsible for the accelerated rise in the over-all A lthough monetary policy limited the availability of money relative to consumer price level. the grow th of transactions demands, it still encour T he severe rate of inflation that we have exp eri aged a substantial expansion in econom ic a ctiv ity ; enced in 1973 cannot responsibly be attributed to real output rose by about 7 per cent in 1972. monetary management or to public policies m ore gen so, unemployment remained unsatisfactorily throughout the greater part of the year. Even high It was not erally. In retrospect, it may well be that monetary policy should have been a little less expansive in FEDERAL RESERVE BA N K OF RIC H M O N D 7 1972. But a markedly more restrictive policy w ould have led to a still sharper rise in interest rates and risked a premature ending of the business expansion, demand deposits at com m ercial banks on a uniform without limiting to any significant degree this year’s It is important for the Congress to put an end to upsurge of the price level. basis from the standpoint o f reserve requirements. Im provem ents in our fiscal policies are also needed. fragmented consideration of expenditures, to place a firm ceiling on total Federal expenditures, and to Concluding Observations relate these expenditures to prospective revenues and T he present inflation is the most serious econom ic problem facing our country, and it poses great d iffi culties for econom ic stabilization policies. W e must the nation’s econom ic needs. recognize, I believe, that it will take some time for the forces of inflation, which now engulf our econom y and others around the w orld, to burn themselves out. In today’s environment, controls on wages and prices cannot be expected to yield the benefits they did in 1971 and 1972, when econom ic conditions were much different. Prim ary reliance in dealing with inflation — both in the near future and over the longer t e r m wili have to be placed on fiscal and monetary policies. T he prospects for regaining price stability would be enhanced by improvements in our m onetary and Fortunately, there is now widespread recognition by members of Congress of the need to reform budgetary procedures along these broad lines. It also is high time for fiscal policy to becom e a more versatile tool of econom ic stabilization. P a r ticularly appropriate would be fiscal instruments that could be adapted quickly, under special legislative rules, to changing econom ic conditions— such as a variable tax credit for business investment in fixed capital. O nce again I w ould urge the Congress to give serious consideration to this urgently needed reform . W e must strive also fo r better understanding o f the fiscal instruments. T h e conduct of monetary policy could be im proved if steps were taken to increase effects of econom ic stabilization policies on econom ic activity and prices. O u r knowledge in this area is the precision with which the money supply can be greater now than it was five or ten years ago, thanks to extensive research undertaken by econom ists in controlled by the Federal Reserve. Part o f the present control problem stems from statistical inade quacies— chiefly the paucity of data on deposits at nonm ember banks. A lso, however, control over the academic institutions, at the Federal Reserve, and elsewhere. T he keen interest of the Joint E conom ic m oney supply and other monetary aggregates is less Committee in im proving econom ic stabilization p oli cies has, I believe, been an influence of great im por precise than it can or should be because nonm ember tance in stimulating this widespread research effort. banks are not subject to the same reserve require ments as are Federal R eserve members. I hope that the Congress will support efforts to Committee in an effort to achieve the kind o f e co nom ic perform ance our citizens expect and deserve. I look forw ard to continued cooperation with the rectify these deficiencies. F or its part, the Federal R eserve Board is even now carrying on discussions with the Federal Deposit Insurance Sincerely yours, Corporation about the need for better statistics on the nation’ s m oney supply. T h e B oard also expects shortly to recom m end to the Congress legislation that will put 8 M ONTHLY REVIEW, DECEMBER 1973 A rth ur F . Burns Evolution of the Concept of the Demand for Money T he concept of the demand fo r m oney is one o f the m ore fundamental elements of supply. An erratically shifting demand function contem porary might offset the effect o f a controlled shift in the m acroeconom ic analysis. T his concept refers to the functional relationship, often expressed as a mathe m oney supply at one time yet accentuate it at another. H aving no firm grasp on the demand function, p olicy matical equation, between the quantity of m oney that people demand to hold and the variables (e.g., inter makers could not hope to assess correctly the m agni est rates, income, wealth, etc.) on which that quantity O ver the past tw o decades the demand fo r m oney has been the subject of many ingenious and com plex depends. D em and-for-m oney equations are key co m ponents of current theoretical and empirical models of the aggregate econom y. F or example, such equa tude or direction of the effects o f their policy actions. theoretical and empirical studies. thus, however. It was not always O w in g to the strategic im portance o f tions are used to represent the behavior o f the d e the m oney demand function, one might assume that mand side o f the so-called “ m oney-market sector” in there had been an early emergence o f a sophisticated large-scale econom etric models em ployed in simulat ing the influence of policy actions and other changes on the econom ic system. In these models, m oney de analysis of it. It seems reasonable to expect that the techniques and standards applied in dem and-for- m oney constructions at least w ould have matched mand equations help to determine the solution (eq u i those em ployed in ordinary com m odity demand anal librium ) values of national income, interest rates, the ysis. price level, and other measures of aggregate econom ic the methods o f monetary demand analysis differed from those o f traditional demand theory. activity. Such was not the case, however, and fo r years T h e chief reason for econom ists’ interest in the B y the turn of the century, com m odity demand dem and-for-m oney relationship, however, is its prac analysis was firm ly anchored in the theory o f utility- tical policy implications. M acroeconom ic analysis suggests that certain properties of the m oney demand m axim izing behavior— part o f the general theory of rational choice. T he utility-m axim ization analysis function may critically influence the degree o f effec provided economists with a set of pow erful analytical techniques that were em ployed systematically in iden tiveness o f monetary policy. Especially im portant are ( 1 ) the interest rate elasticity of the demand for money, i.e., the responsiveness or sensitivity o f the quantity of money demanded to changes in market in terest rates and ( 2 ) the stability of the functional relationship between m oney balances demanded and tifying both the general form and the principal inde pendent variables o f com m odity demand functions. B y contrast, techniques o f m icroeconom ic value theory were not introduced into monetary analysis until the m id-1930’s, and even then their application the independent variables (interest rates, income, etc.) of the equation. F or example, if the amount of money demanded is extrem ely interest elastic, m one was incomplete. tary policy may be powerless to stimulate the econ ysis. One unfortunate consequence of this delay was that for years monetary theorists virtually ignored, om y because, in this case, the slightest fall in the rate N ot until the 1950’s was money demand analysis fully integrated into the rational choice fram ew ork em ployed in ordinary demand anal of interest, resulting from a policy engineered expan or at least did not exam ine systematically, the influ sion of the m oney stock, would simply induce cash ence o f cost and yield considerations on m oney- holders to absorb all the new m oney into idle hoards. holding decisions. Consequently, no increase in expenditure w ould en W h y were econom ists so slow in applying the sue. Even if the demand for money is not excessively methods and procedures responsive to interest rate changes, however, policy ysis to m oney? makers may still be confronted with the problem of cedures consist of and how did m oney demand theory an unstable demand relationship. Instability o f the m oney demand function w ould hinder monetary p oli depart from them ? cy by making it difficult for the authorities to predict counterparts? the impact of policy-induced changes in the money sition from the older approaches to the current ones, of traditional demand anal W h at did these methods and p ro H o w did the older approaches to the demand for m oney differ from their current FEDERAL RESERVE BA N K OF RIC H M O N D W h at were the main stages o f tran 9 and how did this development manifest itself in suc cessive form ulations o f the m oney demand fu n ction ? W h a t explanatory variables were stressed in each form ulation? M ost important, what were the chief these variables are the substitution and incom e effects, the tw o main forces influencing the amount o f any good demanded by an individual. T he substitution policy implications of the alternative v iew s? This article seeks to answer these questions by tracing against the backdrop o f orth od ox demand analysis follow ing a change in relative prices, which, by alter the substitution of relatively cheaper fo r relatively the main lines of development of the theory o f the demand for m oney from the early decades of the dearer good s in the consum er’s budget. T he incom e effect refers to the demand impact of a price-induced century to the present. T h e basic m ethod em ployed in ordinary demand analysis is outlined in the fo l low ing section, which then serves as a point o f refer change in the real purchasing pow er o f the individ ence for the remainder o f the article dealing with the of a single g oo d has the effect of increasing the co n effect refers to the shift in expenditure patterns ing the rates of exchange am ong com m odities, induces ual’s total budget, which alters his entire range o f alternatives. F o r example, the low ering o f the price study o f dem and-for-m oney relationships. sumer’s effective budget, thereby expanding his field Conventional Demand Analysis o f choice and thus influencing his demand fo r all com m odities including the particular g ood in question. C o m m o d ity d e mand analysis is founded on the theory o f rational choice, or optim izing (i.e., constrained utility-m axi m izing) behavior. T he analysis begins at the level of the individual consum er and seeks to deduce how he will allocate his limited m oney resources— the socalled budget constraint— am ong the various goods available to him at given market prices in such a way as to m axim ize his total satisfaction or utility. T he conclusion is that he will apportion his budget in such a w ay that the rate at which he can substitute one good for another via market exchanges is just equal to the rate at which he is willing to d o so (h is sub jective tra d e -o ff). T he next stage o f the analysis is to derive the market demand function by aggregating over all the individual demand functions. In general, relative prices and incom e will also be the chief arguments in the market demand functions. T h e im portant point, however, is that the analysis proceeds from the level o f the individual decision-m aking unit to the m a rk et; market demand functions follow from individual d e mand functions. Finally, in the last stage o f the analysis, the market demand and market supply equa tions are solved simultaneously to determine the market clearing levels of price and output. T o summarize, traditional demand analysis: ( 1 ) is In principle, the individual’ s demand function for any com m odity can be derived from this analysis of budget-constrained utility-maxim ization. T he result ing demand functions indicate the relationships be tween the quantities demanded o f the g ood s and the variables on which those quantities depend. W ith tastes and preferences given, the individual’ s demand for any com m odity will depend chiefly on tw o sets of variables: ( 1 ) relative prices (the price o f the good based on the principle of optimization or rational choice, ( 2 ) starts at the level o f the individual deci sion-m aking agent and then proceeds, via aggrega tion, to the market level, and ( 3 ) predicts that d e mands will be largely determined by relative prices and incom e constraints. The Demand for M oney— Transactions Velocity in question in com parison with the prices o f all other Approach T o the m od ern e co n o m ist, it seem s natural to construct dem and-for-m oney equations in goods that com pete for the consum er’ s expenditure) exactly the same way that com m odity demand equa and T he relative price variables represent the tions are formulated. T h e analysis begins at the level of the individual m oney holder, determines the appro terms in which one g ood can be substituted for priate budget constraint and the relevant opportunity (transform ed ex cost or relative price variables that enter his demand- T hese variables also measure the op p or for-m oney function, and derives via aggregation the tunity costs of obtaining one g ood in terms of the total market dem and-for-m oney function. T h e analy (2 ) straint). changes. the consum er’s incom e in to) other (his budget co n goods via market amounts of other good s w hose purchase must be sis then proceeds to investigate both the substitution sacrificed or foregone. T he remaining variable, in effects between m oney and com peting assets stem come, indicates the individual’s com m and over all m ing from changes in relative (com parative) rates o f goods and return and the incom e (o r w ealth) effects on the thus serves to fix an upper limit on the amount of demand fo r m oney stemming from changes in na the com m odity he can purchase. tional income. (the purchasing pow er of his budget) Relative price and incom e variables play stra tegic roles in demand analysis. 10 O perating through T his conventional approach is of relatively recent origin, how ever. M ONTHLY REVIEW, DECEMBER 1973 U p through the early decades of the twentieth century, econom ists as a rule did not examine the demand for m oney along the lines o f orth od ox demand theory. Rational choice principles, or utility-maxim ization analysis, it was argued, could not explain w hy any individual would want to hold money, because cash holdings as such were believed to produce no direct satisfaction. In short, money the product of the average stock o f m oney, M , and the average number o f times each unit o f m oney turns over in financing exchanges (v e lo c ity ), V — must equal the aggregate value of transactions, P T — the product of the total number of transactions, T , and the average price per transaction, P. was view ed as just a mechanical medium o f exch ange, Fisher argued that the transactions velocity o f m oney was determined by slow ly changing techno i.e., something used to facilitate market transactions and to circulate goods, but not in itself a utility- logical and institutional factors, e.g., state o f develop ment o f the banking system, frequency o f receipts and yielding asset. T he distinguishing feature of a medium o f e x disbursements, length of the payment period, degree o f synchronization o f cash inflows and outflows, change is that it is transferred or circulated, not that it is held. A ccordin gly, theories that focus exclu rapidity o f transportation and comm unication, etc. sively on the m edium -of-exchange function o f m oney tend to ignore the demand fo r m oney p er se and instead concentrate on how fast money circulates from hand to hand. T his interest in the circidation v elocity, or rate of use, o f m oney was particularly character istic of most nineteenth and early twentieth century monetary theorists. T hese analysts pointed out that the efficiency with Since these factors were subject to only gradual, evolutionary change, velocity could be considered a virtual constant in the equation o f exchange. T he constancy of velocity implies the com plete stability o f the demand fo r m oney. A n d with the latter absolutely stable, monetary policy could be expected to exert a pow erful, predictable influence on prices and nominal income. U sing the equation o f exchange, Fisher demonstrated the potential potency which m oney facilitates exchanges depends on how o f monetary policy in this stable demand case. rapidly it circulates in market transactions. velocity, V , a constant, and transactions, T , also assumed to be a constant determined by the full- In the limit, the turnover velocity of a perfectly efficient medium o f exchange w ould approach infinity; and the demand for cash balances correspondingly w ould approach zero. It was, of course, recognized that perfect efficiency in circulation is never achieved, that velocity is necessarily finite, and that people often' W ith capacity utilization o f the econ om y’s productive re sources, the equation could be expressed in a form , P = ( V / T ) M — (con stan t) M , showing a constant proportional relationship between average prices and the m oney stock. T his expression implied that a maintain inventories of idle cash fo r sustained inter given percentage change in the m oney stock w ould vals of time. T he existence of positive cash balances, cause the same percentage change in the price level. however, was attributed not to rational choices and utility-m axim izing decisions but rather to institu The Cambridge Cash Balance Approach tional “ frictions” in the econom ic system. In sum, initial step in m oving the theory of the demand for the rate of circulation of m oney— and by implication money in the direction o f ordinary demand analysis was taken simultaneously by several Continental and the dem and-for-m oney balances— was thought to be determined by technological and institutional factors associated with the aggregate payments mechanism rather than by the subjective processes of individual decision-making. Consequently, analysis tended to center on statistical measures of the aggregate trans actions velocity o f m oney, rather than on cost and yield considerations affecting the m oney-holding choices of individual optimizers. The British economists. A m on g the m ore influential o f these analysts was the small grou p associated with Cam bridge U niversity in England. In a series o f writings spanning the period 1917-1930, econom ists o f the Cam bridge school contributed at least four innovations to m onetary analysis, thereby advancing it beyond F isher’s transactions velocity form ulation. T h e first innovation was to concentrate on m oney T he leading exponent of the transactions velocity in relation to final output, or national income, rather approach was the A m erican economist, Irving Fisher, than on the much broader and m ore inclusive aggre^ whose principal w riting in the field of monetary gate, total transactions. theory, T he Purchasing P o w e r of M o n ey , appeared attention to the properties that make m oney a desir T his innovation directed in 1911. Fisher did not write out an explicit demand- able ob ject to hold as distinct from an object to spend. for-m oney function. Instead, he examined the be A s long as one associates m oney with gross trans havior o f velocity within the fram ework of his cele actions, one necesarily tends to think o f m oney e x brated equation of exchange, M V = PT. This fo r mula is an identity stating that total spending, M V — clusively as a means o f exchange. B y recasting the analysis in terms of income, rather than transactions, FEDERAL RESERVE B A N K OF R IC H M O N D 11 however, the Cam bridge school opened up the possi bility of interpreting m oney as something m ore than just a medium o f exchange. equal m oney demand, M s = M d, resulting in the cash balance equation, M = kP y. W ith emphasis shifting from transactions to in plicitly stated a rudimentary m oney demand function and drew the demand curve corresponding to it. com e, F isher’s equation of exchange was eventually restated as M V y = P y = Y , where M is the stock of m oney in circulation, Y is nominal national income, y is real income or the national product valued at constant prices, P is the price level of the national product, and V y is the incom e velocity o f money, i.e., A s a fourth innovation, Cam bridge econom ists e x F rom the assumption that the com m unity w ould wish to hold a constant quantity o f real (p rice deflated) cash balances at the full-capacity level o f real output, an expression was derived showing the the ratio of nominal national incom e to the money quantity of nominal balances demanded, M , as a function of the exchange value of the monetary stock or, alternatively, the rate of turnover o f money as it circulates against the national product. A few unit (i.e., the reciprocal o f the price level, 1 / P ) . A dm ittedly an artificial construct, this particular d e analysts, particularly in the United States, began to use this equation to investigate the behavior of incom e mand equation expressed the product of the tw o variables M and 1 /P as a fixed constant. Real in velocity. T he Cam bridge school, however, fo r the most part generally eschewed analysis of m oney’ s com e and the m on ey /in com e ratio— the other factors that conceivably could m fluence the demand for circulation velocity and instead focused on the famous m oney— w ere interpreted as fix ed parameters in the “ Cam bridge k ,” i.e., the desired ratio, k, o f money function. balances to income. graphed in the Cartesian plane, with nominal m oney balances demanded, M , measured along the horizontal In other w ords, the Cambridge school sought to explain the proportion of annual incom e that the com m unity of decision makers wished to hold in the form of m oney, not how rapidly m oney T he function was of a special fo r m : when axis and the exchange value o f m oney, 1/ P , along the vertical axis, the equation described a dow nw ard- turns over in buying the national product. sloping rectangular hyperbola. This focus on the cash balance ratio was the second novelty of the Cam bridge approach that ultimately coordinates of each point on this demand curve would led to a m ore conventional interpretation o f the de T he product o f the be the same and equal to the constant quantity o f real 1/V y. balances demanded, M /P . T his special demand curve was used by the Cam bridge school to demonstrate the validity o f the quantity theory o f m oney. T he quan But the k ratio implies a desired holding o f m oney tity theory asserts that, because people look to the balances, and is thus m ore suggestive o f conventional purchasing pow er rather than to the mere money demand theory than is V y, the rate of spending o f money. T he Cam bridge emphasis on cash holdings value of their cash balances, the price level w ould have to vary in direct proportion to the nominal suggested that m oney might be a utility-yielding as m oney supply to maintain real balances intact. set, and also that the demand for money, like the demand for any com m odity, is a matter o f individual bridge form ulation, however. mand for money. Form ally, k is just the reciprocal of the incom e velocity o f circulation, i.e., k = choices and decisions. O ne crucial element was m issing from the Cam T here was nothing in A s a third innovation, the Cambridge school re the m oney demand equations analogous to the rela tive price arguments that appear in ordinary demand form ulated the equation o f exchange in a manner functions. N o variables entered the Cambridge equa more consistent with orth od ox demand and supply tions to represent the opportunity costs of cash hold analysis. T his step involved replacing the symbol ings, i.e., the yields on alternative non-m onetary for the incom e velocity of money with its reciprocal, assets. the Cam bridge k, and then incorporating the latter for m oney to respond to changes in these costs or into the Cam bridge cash balance equation M = yields. kP y. Y et one norm ally w ould expect the demand F or example, if cash holders behaved ratio T h e cash balance equation was interpreted as the nally, a rise in interest rates w ould probably induce a equilibrium solution o f a three-equation dem an d / fall in k (the cash balance ratio) as people sought to supply system, rather than as a simple identity as had econom ize on cash holdings and to substitute interest been the equation of exchange. earning assets for m oney balances in their asset p ort Specifically, the Cam bridge form ulation implies ( 1 ) a dem and-for- folios. Similarly, falling interest rates, by low ering k P y with the incom e co n the opportunity cost o f holding m oney relative to the straint, P y = Y , appearing as an explicit independent brokerage costs o f converting it into and out o f bonds, variable; w ould m ost likely cause a rise in the cash balance m oney equation M d = (2 ) supply M g = an exogenously determined m oney M ; and ( 3 ) an equilibrium (m arket- clearing) condition stating that m oney supply must 12 ratio. Strangely, how ever, there was no explicit recognition of any yield or substitution effects in the MONTHLY REVIEW, DECEMBER 1973 Cambridge equations. Instead, the Cambridge k ratio was treated as a numerical constant rather than as a variable whose magnitude is functionally related to rates of return on non-cash assets, i.e., k = k (r ). A pparently the failure to include interest rates as a determinant of m oney demand stemmed from the C am bridge sch ool’s inability to see the full im plica tions of its analytical approach. A fter constructing a fram ework conducive to the study of factors influenc ing cash-holding decisions, Cambridge economists failed to exploit this innovation fully. T rue, one can K eynes’s Formulation A s p re v io u s ly m en tion ed , the main shortcom ing of the Cambridge cash balance analysis was its failure to incorporate yield or cost variables into the money demand function. This oversight was partially rectified in John M aynard K eynes’s analysis o f the speculative or liquidity preference demand for money, presented in his 1936 classic, T he General T h eory of E m ploym ent, Interest, and M on ey. Keynes separated the demand for money into tw o distinct p a rts : ( 1 ) a demand for transactions or find in the writings of the Cambridge school refer active balances to satisfy the transactions and pre cautionary motives for holding cash and ( 2 ) a d e ences to a representative individual striking a balance mand for idle or asset balances to satisfy a speculative between his holdings of cash and non-cash assets as motive. well as some mention of trade-offs between costs and M 2, respectively. Keynes labeled these tw o demands M i and returns (convenience, safety, etc.) on cash holdings. It was in conjunction with the speculative demand But such passages were infrequent, and the insights that he gave explicit consideration to the yields on they contain were never incorporated systematically assets that com pete with money in individuals’ p ort into the Cam bridge analysis. folios. F or the most part, Cambridge economists, when describing the determ i nants of k, referred to the same technological-institu tional factors that Fisher had cited in his discussion of velocity. T he constancy of k in the Cambridge analysis had the same policy implications as the invariability of velocity in F isher’s theory. Keynes argued that individuals make their cash-holding decision by com paring the interest in com e that w ould be sacrificed by holding money with the expected capital gain or loss on holding bonds. The latter depends on decision makers’ anticipations of future movements in bond prices and the degree o f certainty with which those expectations are held. A ccord in g to Keynes, these anticipations are form u Both implied stability of lated via com parisons of the current rate o f interest the dem and-for-m oney function and thus the powerful with some expected “ norm al” or permanently main influence of monetary policy on prices and nominal tainable rate. If the observed rate is above the normal incomes. rate, individuals will expect it to fall. Since bond prices vary inversely with bond yields, however, A ccord in g to the Cambridge analysis, the impact of policy-induced changes in the m oney supply w ould not be weakened or negated by perverse or unexpected shifts in the demand for m oney. T o the contrary, the constant desired k ratio was interpreted as a reliable strategic link in the transmission m ech anism connecting m oney to prices. Thus, any in anticipations of falling interest rates mean expecta tions of rising bond prices and thus capital gains. The higher the current rate of interest the greater the amount o f capital gains expected. W h y ? Because the larger the spread between the current and e x pected maintainable rates, the greater the likelihood crease in the m oney supply would, first, raise the actual m on ey/in com e ratio above the desired level ; actual k w ould be greater than desired k. Then, in dividuals, finding that they were holding more cash than they wanted in relation to their incomes, would increase spending. T he increased expenditure in a that the interest and the greater pected to fall. the m ore costly expected capital fully em ployed econom y w ould push up market prices, incom e foregone. thereby raising nominal income. This increased rate the quantity of cash demanded to satisfy the specu of spending w ould continue until the subsequent rise lative motive. Conversely, if the observed rate is below the e x in the price level and nominal income was sufficient rate will fall (b on d prices will rise ), the amount by w hich it can be e x Thus, the higher the current yield, are idle cash holdings in terms of gains sacrificed, as well as interest Consequently, the smaller will be to bring the actual cash /in com e ratio into equality pected normal rate, anticipations o f rising bond yields with the desired ratio. and declining bond prices render cash the preferred Since the desired ratio is co n stant, nominal income and the price level w ould have asset in individuals’ portfolios. to rise in exactly the same proportion as the money pecting the price o f bonds to fall at a rate that w ould A n individual e x stock. A t this point the com m unity w ould just be m ore than offset the interest earned on them w ould willing to hold the augmented stock of m oney and be motivated to hold zero-yield cash rather than the the adjustm ent process w ould be complete. overpriced bonds. FEDERAL RESERVE BA N K OF RIC H M O N D Generally, the low er the current 13 rate, the m ore unanimous will be the expectations that interest rates will subsequently rise, im posing capital losses on bond holders. Thus, the low er the current rate of interest the greater the num ber o f people w ho prefer to hold cash rather than bonds and positive rate of interest so low that if the current rate were actually at that level, no one would expect it to go any low er and everyone w ould expect it to rise. In other w ords, anticipations of falling bond prices therefore the greater the total quantity o f cash de w ould be unanimous. A t this point, anticipated cap ital losses would offset interest returns, and there manded. A ggregating over all individual portfolio would be no advantage to holding bonds. Cash w ould optimizers gives a smooth dow nw ard-sloping fu n c become a perfect substitute for bond holdings, and the demand for money w ould becom e insatiable, i.e., tion, M 2 — f ( r ) , relating the quantity o f speculative or asset balances demanded to the current interest rate. A s for the transactions balance com ponent o f the infinitely sensitive to the slightest change in the rate of interest— a pathological condition that Keynes called absolute liquidity preference. U nder these cir total demand for money, i.e., the portion held to cumstances, any increase in the m oney supply w ould finance day-to-day purchases and to provide a reserve for emergencies, Keynes agreed with his predeces be com pletely absorbed into idle cash balances with sors, Fisher and the Cam bridge school, that it would exhibit a simple, linear (p rop ortion a l) relation to bank acted to increase the money supply by purchas nominal income. In fact, K eynes’ s form ulation o f the transactions demand function is identical to the up of bond prices w ould simply induce individuals to C am bridge sch ool’s and may be expressed by the Cam bridge cash balance equation, M i = k Y . T he no reduction in interest rates. Thus, if the central ing bonds on the open market, the slightest biddingsell their bonds to the central bank and absorb the cash proceeds. Since at the floor rate of interest the reader will note that Keynes did not apply rational demand for cash is insatiable and the willingness to sell bonds is absolute, no amount of open market choice, optim izing considerations to the transactions com ponent of the demand for money. T his task re operations would overcom e absolute liquidity p refer ence and force interest rates to g o any low er. mained for later analysts, w h o showed that trans Both the instability and infinite elasticity properties actions balances also respond to cost and yield c o n of the m oney demand function, Keynesians pointed siderations. out, w ould have pessimistic policy implications. C om bining the tw o com ponents of demand gives the Keynesian total money demand function, M = stability of the money demand function w ould make M i -f- M 2 = k Y -f- f ( r ) . A ccordin g to this fu n c In accurate forecasting of the effects o f monetary policy impossible. Confronted with a volatile and unpre tion, the quantity of money demanded w ould vary in dictable direct proportion to income and inversely with the interest rate. It should be noted that the last term in this equation is im properly specified. A ccord in g to w ould never know whether shifts in demand would m agnify or nullify policy-induced shifts in the m oney supply. M oreover, even if the monetary authorities K eynes’s discussion, the demand fo r speculative bal ances depends on the current rate o f interest in rela tion to some expected normal rate. Thus the latter rate properly should be included as one o f the explan atory variables determining the quantity of money demanded. Keynes and his follow ers, how ever, chose m oney demand function, the authorities could predict the behavior o f m oney demand, m one tary policy still w ould be powerless if conditions of absolute liquidity preference prevailed. In the latter case, increases in the m oney supply w ould have no effect on nominal incom e or econom ic activity through the interest rate channel. Since no cash holder would to treat the expected rate as an exogenous factor be willing to bid for bonds, there w ould be no rise in contributing to erratic shifts in the functional rela bond prices and consequent fall in interest rates to tionship between the quantity of m oney demanded stimulate business investment spending. M oreover, none of the monetary injection w ould enter the spend and the current rate. K eynes and his follow ers also broke new ground ing stream. Instead, all of the new money w ould be in their discussion o f the behavior of the dem and-for- absorbed in idle cash balances. In short, the econom y money function. would be caught in a liquidity trap. First, unlike Fisher and the Cam bridge school, Keynesians argued that the money demand function is highly unstable, shifting errati deep depression, m oney stock changes w ould be ne cally under the impact of volatile market expectations. gated by offsetting changes in velocity or the Cam Second, Keynesians thought that in times o f deep depression the m oney demand function w ould becom e T o summarize, Keynesians argued that in times of bridge k. W ith variable velocity, ©r k, absorbing all the impact of m oney stock changes, none w ould be horizontal (infinitely elastic) at some floor rate o f transmitted to nominal income. interest. between money and econom ic activity postulated by T hey argued that there is some critical 14 M ONTHLY REVIEW, DECEMBER 1973 T he rigid linkage earlier economists w ould be severed. Thus, K e y nesians arrived at policy conclusions at variance with those reached by Fisher and the Cambridge school. J. R. H icks’s Analysis K e y n e s ’s ch ie f c o n tr ib u tion to the analysis of the demand for money was the introduction of a variable representing the cost of holding cash (the rate of interest) into the money demand function. transactions or spending that is closely related to the flow of income. H ick s’s use of the wealth constraint, by contrast, called attention to the stock of m oney as a store of wealth, i.e., a service or utility-yielding asset alternative to other asset stocks. In addition to his pioneering proposal that ordinary demand analysis be applied to m oney in its role as a balance sheet asset, H icks also took the initial step This innovation permitted exam i in extending the theory of choice or optim izing be nation of the substitution effects on the demand for havior to explain the demand for transactions (as distinct from asset) balances. P rior to H icks, no one money stemming from changes in relative rates of return. In giving explicit consideration to the yields had attempted this. Even Keynes had limited his on assets that com pete with money, Keynes became application of rational choice analysis to the asset one o f the founders of the portfolio balance approach com ponent of m oney demand. M oreover, no one previously had provided a convincing explanation of why individuals would be willing to hold trans to monetary analysis, i.e., the approach that inter prets the demand for money as part of the choice of an optimum portfolio of assets. A t least equal recog nition for originating the portfolio approach, however, actions balances when riskless, interest-bearing assets of virtually instantaneous redeemability (e.g., time should be given to British econom ist John R. Hicks, deposits) were available. who in 1935 first suggested that the demand for voluntarily sacrifice the option of holding interestyielding, speedily-convertible assets for the option o f holding cash ? Because the latter option may be less money be treated as a problem of balance sheet equi librium or asset choice to be analyzed along the lines of orth od ox com m odity demand theory. H icks pointed out that if money were to be analyzed as a capital asset and not just as a mechanical medium W h y w ould transactors costly, said H icks. In short, the existence of trans actions balances could be explained as the outcom e of rational, cost-m inim izing behavior. M ore specifically, of exchange, the dem and-for-m oney equation would the only reason for holding transactions balances, sug have to include as explanatory variables total wealth gested H icks, was the conversion costs (brokerage and expected rates of return on other assets. fees, effort and inconvenience, etc.) of transferring The wealth variable would represent the budget co n cash into earning assets and vice versa. straint on money holdings, since at the m axim um individuals could choose to hold their entire wealth pointed out that it would not pay to get out o f cash portfolios in the form o f cash. tw o-w ay conversion costs exceeded the interest in A nd the yield vari H icks into earning assets for short periods of time if the ables w ould represent both the opportunity costs of com e foregone by holding cash. holding money and the portfolio-substitution effects H ick s’s observation that the demand for trans actions balances stems from cost-m inim izing behav of changes in relative rates of return. Individual portfolio optimizers would com pare these yields with the imputed convenience and security yield on money balances in deciding whether to substitute other assets for cash in their balance sheets. H ick s’s specification of wealth as the constraint ior, together with his proposal that cash holdings be analyzed as a com ponent o f a portfolio of assets, served to eliminate much o f the remaining disparity between m oney demand theory and conventional de variable, it should be noted, was a significant depar mand theory. T he final steps, however, were taken in the 1950's by analysts w orking along the lines ture from the Cam bridge and Keynesian form ula tions, both of which used incom e in the m oney d e opened up by H icks. Chief am ong the contributions stemming from H ick s’s w ork were M ilton Friedm an’ s mand equation. elaboration o f the portfolio balance approach and This shift from income to wealth as the constraint variable underscored the shift from the B aum ol’s and T o b in ’s refinement o f the transactions transactions approach to the capital asset or p ort cost approach. These studies succeeded in integrating folio approach in H icks’s article. Incom e is a m agni money demand analysis into the optim izing behavior tude having the time dimension of a flow (an amount fram ework o f orth od ox demand analysis. that occurs over an interval of time, or so much per W ealth, 011 the other hand, has the M oney as a Capital A s s e t: Friedman’s Contribu time dimension of a stock (s o much existing at a tion F o llo w in g H ic k s ’s s u g g e s tio n , M ilto n F rie d given point in tim e). unit of tim e). T he rationale for the income man in 1955 em ployed the procedures of conven constraint in the Cam bridge and Keynesian form u tional com m odity demand theory to construct a de lations was that money is used to finance a flow of tailed dem and-for-m oney FEDERAL RESERVE BA N K OF RIC H M O N D function. Like H icks, 15 Friedm an interpreted the demand for m oney as a problem in balance sheet equilibrium or asset choice. H ow ever, unlike earlier analysts (especially K e y n e s), he did not seek to explain the demand fo r m oney in terms of special m otives that are satisfied by cash holdings. Instead, he treated money as a capital good, which, like any other capital good, yields a flow of specific substitute for a narrow range of financial assets led him to reject the Keynesian conclusions that m onetary policy may be relatively ineffective. Keynes had argued that if a monetary change does influence econom ic activity, it does so through an indirect interest rate mechanism rather than through services that makes it desirable to hold. T his ap proach is characteristic o f orth od ox demand analysis, the direct expenditure o f m oney for goods. S pecifi cally, in the Keynesian m odel, a m onetary increase initially upsets the preexisting optimum cash-bond which avoids considerations of psychological motives m ix in wealth portfolios. prom pting the purchase of goods. tempt to restore portfolio equilibrium by substituting Friedm an’s contribution consisted of a step-by-step demonstration of how a household’s m oney demand bonds for cash. T he increased demand for bonds pushes their prices up and their yields down. F all function could be derived from first principles of Equally important, h ow ing interest rates stimulate investment expenditure for new capital goods, and the increased investment ever, was his precise and complete specification o f the spending has a multiplied effect on national income. relevant constraint and opportunity cost variables entering the function. T he relation between these variables and the quantity of money demanded may But K eynes thought these indirect effects w ould be weak for tw o reasons. First, he thought cash and orth od ox demand analysis. be expressed as M d = f ( W , rb, re, rr, p ) , where W is w ea lth ; rb, re, rr are the expected real rates o f re turn on bonds, equities, and real assets, respectively; and p is the anticipated percentage change in the price level, a measure of the expected rate o f depreci ation in the purchasing pow er of m oney balances. Like H icks, Friedm an specified wealth as the ap W ealth holders then at bonds were such close substitutes that only slight reductions in bond yields w ould be necessary to in duce people to add the extra cash to their portfolios. Thus, increases in the m oney supply could generate only slight reductions in interest rates. Second, he thought that investment spending w ould be insensi tive to small changes in the interest rate. defined Friedman, however, argued that since m oney is a substitute for real good s and services as well as bonds, wealth somewhat unconventionally as the present changes in the quantity of m oney w ould spill over into value of expected future receipts from all sources, including human wealth (personal earning capacity) the market for consum ers’ and producers’ goods, as well as real property and financial capital. spending and not merely a weak indirect effect via the bond interest rate. That is, excess cash balances are at least as likely to be spent directly for good s as for long-term bonds. M oreover, Friedm an main propriate budget constraint, although he Unlike Keynes, w ho view ed bonds as the only asset co m peting with cash, Friedm an regarded all types of wealth as potential substitutes for cash holdings in individuals’ balance sheets. Thus, in sharp contrast to the single interest rate variable in the Keynesian liquidity preference equation, Friedm an’s list of rela tive yield variables entering the money demand func tion included the expected rates of return on bonds, thereby exerting a strong direct effect on private tained that since the substitution effects between m oney and other assets w ould be dispersed over such a wide spectrum o f assets, the particular substitution effect between m oney and any one class o f assets would tend to be small. Thus, contrary to Keynes, equities, and real assets. One additional novel feature Friedman believed that the quantity o f m oney d e was the inclusion of the expected rate of change o f the manded w ould be relatively insensitive to changes in price level as a measure of the rate o f return on (o r the yield on bonds. the depreciation cost o f) money holdings. T he ra relatively unresponsive to interest rate changes, then tionale for this particular variable is that cash hold the problem of hoarding anticipated in the Keynesian A n d if the demand for m oney is ers, recognizing that inflation erodes and deflation liquidity trap case will not arise. augments the purchasing pow er o f cash balances, believes, as did Fisher and the Cam bridge school, would take the expected rate of depreciation or ap preciation of money into account in form ulating their that the link between m oney and nominal incom e is cash holding decisions. m oney stock have a pow erful impact on econom ic N obod y prior to Friedman had thought to incorporate the anticipated rate of In sum, Friedman strong and relatively stable and that changes in the activity. inflation into the demand function as one of the rateThe Inventory Approach to the Transactions D e of-return or relative yield arguments. Friedm an’s interpretation of m oney as a general mand for M oney substitute for all form s o f wealth rather than as a K eynes-H icks-F riedm an 16 MONTHLY REVIEW, DECEMBER 1973 T h e ch ie f d e fic ie n c y o f the asset choice or portfolio form ulation was that it was not addressed specifi cally to the question of w hy anyone would hold trans actions (as distinct from asset) balances when they had the alternative of holding riskless, easily co n tion via his conjecture that transactions balances are consists not o f utility m axim ization but o f cost m ini mization. Specifically, the optim izing transactor will attempt to manage his cash inventory in order to minimize the sum o f tw o c o s t s : ( 1 ) the opportunity costs o f holding cash instead of bonds and ( 2 ) the transfer costs (brokerage fees, inconvenience, etc.) of converting bonds into cash. A s long as the costs held simply to avoid bond conversion costs. vertible, interest-yielding securities. A s previously mentioned, H icks had attempted to answer the ques But of frequent bond conversions exceed the interest in H ick s’s con jectu re was not supported by rigorous come foregone by holding cash, the rational trans proof until the m id-1950’s, when W illiam J. Baumol and James T obin applied inventory management actor will add to his average cash balances and reduce the number of conversions from bonds to cash. N o analysis to transactions balances, thereby integrating the latter into the rational choice optim izing fram e work of general demand theory. Baumol and T obin showed that the optimum in ventory o f transactions balances could be expressed in terms of the so-called lot-size or square-root fo r tice that nothing is said about the services that m oney yields. In the B aum ol-T obin analysis, transactions money is held fo r one reason only, to minimize the total costs o f transactions. In sum, the notion of the balancing o f tw o co m peting costs in order to minimize their total form s mula of inventory theory. A ccord in g to this form ula, the basis of the demonstration that an inventory o f the average cash holdings of an individual transactor would be directly proportional to the square root of transactions balances is kept, not fo r the services it provides, but because it is too costly to continually both the volum e o f payments (transactions) to be switch in and out of bonds. made and the brokerage cost per conversion from bonds to cash and inversely proportional to the The M oney Demand Equation in the S M P Model square root o f the carrying cost (interest incom e fore T his article has sketched the evolution over time of the theory of the demand fo r m oney. This lon g p r o gon e) of the cash inventory. Thus, given the level of transactions, the size of average cash holdings would depend on the relation between the yield on securities and the costs of buying and selling them. If the latter were greater than the form er, it would not pay to substitute earning assets for m on ey; co n sequently, cash holdings w ould be relatively large. But as yields rose relative to conversion costs, it would pay to econom ize on transactions balances. Hence, the size of average cash holdings w ould be negatively related to the rate of interest. If the v o l ume of transactions varied, however, the demand for money w ould vary in the same direction but less than proportionately. T his implies that the m ore prosper ous a person is, the smaller w ill be the increase in his cash balance necessary to cover a given increase in his transactions. cess o f theoretical developm ent has culminated in the money demand equations appearing in large-scale econom etric models currently in use. It is, therefore, fitting to close the article with a brief look at the dem and-for-m oney equations in one particular econ o metric model. F or this purpose, the S S R C -M I T P E N N ( S M P ) model was chosen. T he S M P model is em ployed by Federal R eserve economists in fore casting future levels o f econom ic activity and in simulating the effects of policy actions and other changes on the econom ic system. A s represented in the S M P model, the demand for money is broken dow n into its demand deposit and currency com ponents. T here is a separate equation for each com ponent. In each case, the basic form of the equation is M = k ( r ) Y , where M is the quantity H o w did the B aum ol-T obin inventory analysis co n of m oney demanded, k is the Cam bridge cash balance tribute to the closing of the gap between orth od ox or m on ey/in com e ratio (the reciprocal o f the income demand theory and m oney demand th eory? velocity o f circulation) expressed as the inverse func In the first place, it indicated that the quantity of trans tion of a vector of interest rates on short-term assets, actions cash demanded, like the quantity of a com r, and Y is the level of nominal national income. m odity demanded by a household, w ould be deter These basic equations are used to express the general mined to a significant extent by relative price or long-run equilibrium or desired level of m oney bal opportunity cost variables. Second, and m ore im por ances, M *, associated with the current interest rate tant, it demonstrated that rational choice, or opti vector and current income. Since time is required to mizing behavior, governs the demand for transactions adjust actual cash balances to this desired level, h ow balances just as it does the demand for asset balances ever, the actual demand for m oney in a given time or com m odities. T he only difference is that in the period may differ from the long-run equilibrium level. case of transactions balances optim izing behavior A ccordin gly, a stock adjustm ent process is in cor FEDERAL RESERVE B A N K OF R IC H M O N D 17 porated into the monetary sector of the S M P model to represent the dynamic sequence o f short-run ad justments to the desired long-run equilibrium p osi that rising interest rates induce the substitution o f relatively higher yield assets (b o n d s ) fo r relatively low er yield ones (m o n e y ) in asset portfolios and ( 2 ) tion described by the equation M * = The the B aum ol-T obin hypothesis that the ratio o f cash stock adjustm ent mechanism embodies the hypothesis k (r )Y . to transactions (o r in com e) is controlled by the in that in any given period money holders adjust their terest obtainable on bonds relative to the cost o f c o n verting cash into bonds. cash balances by a fixed proportion o f the discrepancy between the desired stock, M *, and the actual stock Still another feature is the stability o f the m oney of m oney in existence at the end o f the previous demand relation. Statistical tests of the S M P equa tion indicate that the parameters o f the equation are very stable, changing at best only slow ly over time. period, M _ i. If b represents the proportion o f the gap closed in a given time period, then the change in m oney balances, A M , may be written as A M = b (M * — M _ i). Since the long-run equilibrium demand for m oney is expressed as M * = k ( r ) Y and since the change in the demand for money by defi nition is A M = M — M _ i, the demand for money in the current period may be shown to be M = b k (r ) Y — (1 — b )M _ L . These findings conform to Friedm an’s hypothesis that the demand for money is a stable function o f a rela tively few variables. Perhaps the m ost controversial aspect of the S M P formulation is the use of current incom e rather than wealth as the constraint variable. Similar to other models that stress the transaction or cost-m inim izing T he S M P m oney demand equation is particularly approaches to the demand for money, the S M P model instructive because it embodies so many of the ele ments or features stressed in the various theories specifies incom e as the appropriate budget constraint on m oney holding. A s previously mentioned, h ow surveyed in this article. One such feature is the prominence of yield and income variables in the equa ever, analysts who adhere to the portfolio-balance or tion, which conform s to the post-K eynesian practice of incorporating relative price (y ield s) and budget constraint arguments into the demand function. the relevant budget constraint. Since incom e and wealth are closely related variables, however, the difference between the tw o approaches may not be as M oreover, the appearance of the transactions-related great as it appears at first glance. Specifically, wealth is the present value o f expected future income. A n d income variable in the equation links the S M P ver sion to earlier theories that emphasized the role of money as a transactions medium of exchange. In fact, the authors of the S M P money demand equa tions designate their form ulation “ the neo-Fisherian approach” and state that the demand for m oney is utility-m axim izing approach maintain that wealth is expected future income may be measured empirically by the weighted average of current and past levels of income. It follow s, therefore, that m oney demand equations using current incom e as the constraint are not incompatible with equations em ploying wealth as basically related to the flow of transactions and arises from a lack of synchronization between receipts and payments. T he same hypothesis, of course, underlies Irving F isher’s form ulation as well as the B aum ol- the constraint, as long as the form er equations also include lagged values of past income as supplementary variables. It can be demonstrated that recursive sub stitution for lagged values o f the money demand vari T ob in theory. able M _ ! in the equation M = b k ( r ) Y — ( 1 — b ) M _ i A nother noteworthy feature is the reemergence of the Cam bridge k in the S M P model. In fact, the causes the transactions variable, Y , to enter the equa tion in the form o f an exponentially weighted average S M P m oney demand equation may be viewed as an of current and past levels of income. Thus, the S M P im proved, augmented version of the old Cambridge stock adjustm ent form ulation reconciles the incom e cash balance equation. and wealth approaches to specifying the constraint on T he chief difference between the older and the newer version is the latter’s treat the demand fo r money. ment of k as an inverse function of the rate o f interest rather than as a numerical constant. T he original Summary T h is a rticle has tra ced th e seq u en ce Cam bridge form ulation implied that changes in bond o f steps that brought the theory o f the demand for yields w ould exert no substitution effects on cash money into the rational choice, optim izing fram ew ork holdings. o f conventional demand analysis. B y contrast, the S M P equation implies that, at a given level of income, a rise in bond yields T he integration process was accom plished in three stages. will induce a reduction in cash balances demanded. First was the development, in the 1920’s, of the The postulated inverse relation between k and the cash balance approach, which replaced the concept o f rate o f interest is consistent with both transactions velocity with the alternative concept o f K eynes-H icks-F riedm an 18 balance sheet (1 ) the hypothesis the desired m on ey/in com e ratio. M ONTHLY REVIEW, DECEMBER 1973 T he cash balance analysis tended to shift attention to factors affecting the holding, as distinct from the spending, o f money. It also led to the form ulation of a rudimentary de- SELECTED BIBLIOGRAPHY m and-for-m oney function containing the cash balance ratio as a parameter and nominal income as the co n straint variable. Baumol, William J. “ The Transactions Demand for Cash: An Inventory Theoretic Approach.” Quarterly Journal o f Economics, 66 (November 1952), 545-58 T he second step was taken in the 1930’s, when variables representing relative yields and costs were Friedman, Milton. “ The Quantity Theory of Money— A Restatement.” Studies in the Quantity Theory of Money. Ed. Milton Friedman, Chicago: University of Chicago Press, 1956, pp. 3-21. introduced into the money demand function, and the demand for m oney was interpreted as part o f the choice of a portfolio of assets optimally com posed with regard to alternative yields. Incorporation of in terest rate variables into the demand function per mitted systematic analysis of the substitution effects between m oney and com peting assets stemming from changes in relative rates of return. T he most recent stage occurred in the 1950’s when it was shown ( 1 ) that the demand function for asset money could be derived from the utility-maximization analysis of conventional demand theory and ( 2 ) that holdings of transactions balances are consistent with rational cost-m inim izing behavior. A ssociated with each of these stages of theoretical development were certain policy implications. The cash balance approach led to the belief in the potency o f monetary policy. T his belief stemmed from the assumed stability (con stan cy) o f the cash balance ratio linking money to nominal income. Coinciding ------------------. “ Money: Quantity Theory.” International Encyclopedia o f the Social Sciences. Vol. 10. Ed David L. Sills. New York: Macmillan and Free Press, 1968, 432-47. Hicks, John R. “ A Suggestion for Simplifying the Theory of Money.” Economica, 2 (February 1935). 1-19. Laidler, David. The Demand fo r M oney: Theories and Evidence. Scranton, P a.: International Textbook, 1970. Modigliani, Franco. “ Liquidity Preference.” Interna tional Encyclopedia o f the Social Sciences. Vol. 9. Ed. David L. Sills. New York: Macmillan and Free Press, 1968, 394-408. ----------------------, Robert Rasche, and J. Philip Cooper “ Central Bank Policy, the Money Supply, and the Short-Term Rate of Interest.” Journal of Money, Credit and Banking. 2 (May 1970), 166-218. with the introduction o f cost and yield considerations into money demand analysis in the 1930’s, however, was a reversal of the belief in the potency o f m one tary policy. Analysts began to emphasize such policydebilitating forces as the high interest elasticity and the extrem e instability of the dem and-for-m oney rela tionship. These views contributed to the doctrine, widespread in the 1930’s and 1940's, of the ineffec tiveness of monetary policy. Since the 1950’s, h ow ever, the consensus has tended to swing back tow ard belief in the potential pow er of monetary policy. Recent theoretical and empirical findings o f the sta Patinkin, Don. “ Keynesian Monetary Theory and the Cambridge School.” Banca Nazionale del Lavoro Quarterly Review. (June 1972), 3-23. Teigen, Ronald L. “ The Demand for and Supply of Money.” Readings in Money, National Income, and Stabilization Policy. Ed. W . L. Smith and R. L. Teigen, 2nd Ed. Homewood, Illinois: R. D. Irwin, Inc.. 1970, p p . 74-92. Tobin, James. “ The Interest-Elasticity of Transactions Demand for Cash.” The Review o f Economics and Statistics, 38 (August 1956), 241-47. bility and relative interest inelasticity of the demandfor-m oney function tend to support this latter view. Thomas M . H um phrey The M o n t h l y R e v i e w is produced by the Research Department o f the Federal Reserve Bank o f Richmond. Subscriptions are available to the public without charge. Address inquiries to Bank and Public Relations, Federal Reserve Bank of Richmond, P. O. Box 27622, Richmond, Virginia 23261. Articles may be reproduced if source is given. Please provide the Bank's Research Department with a copy o f any publication in which an article is used. FEDERAL RESERVE BA N K OF RIC H M O N D 19 MONTHLY REVIEW INDEX Volum e 59, 1973 FEDERAL RESERVE B A N K OF R IC H M O N D JANUARY Financial Highlights of 1972 Personal Income in the Fifth District in 1971 Employment and Unemployment Since 1969 FEBRUARY Philip H. Davidson and B. G ayle Burgess John W. Scott W illiam E. Cullison Econometric Models: The Monetarist and Non-M onetarist View s Com pared Joseph M. Crews Forecasts 1973 Financial Forecasts: 1973 W illiam E. Cullison and Carla R. Gregory Philip H. D avidson MARCH The Dismal Science Revisited Bank Affiliates and Their Regulation: Part I Thomas M. Humphrey W illiam F. Upshaw APRIL Bank Affiliates and Their Regulation: Part II i Time and Savin gs Deposits in the Fifth District The District Economy in Perspective: 1972 Spotlight on Agriculture W illiam F. Upshaw Susan P. Krug B. G ayle Burgess S ad a L. Clarke JUNE Bank Affiliates an d Their Regulation: Part III W illiam F. Upshaw International Agricultural Trade and the U.S. Balance of Payments MAY Thomas E. Snider Sources of Bank Expansion in the Fifth District: Internal and External G row th Labor Turnover: Another View of the Labor Market Clyde H. Farnsworth, Jr. Glenn Picou Rural Housing in the Fifth District Thomas E. Snider Thomas M. Humphrey B. G ayle Ennis Using the Futures M arket to Hedge Thomas E. Snider Recent Changes in Fifth District S M S A 's A U G U ST C h an gin g View s of the Phillips Curve The Fifth District Labor Force JULY Patricia G. Rhodes Susan P. Krug The Household W orker DECEMBER The Cyclical Behavior of Consum er Credit Glenn Picou Sharon M. Haley Economic Aspects of Health Care and Medical Insurance Program s W illiam E. Cullison Fifth District Bank Loans: 1965-1972 NO VEM BER Timothy Q. Cook D ou glas H. Lemmonds Behind the Unemployment Rate OCTOBER Some Factors Affecting Long-Term Yield Spreads in Recent Years Edge Corporations: A Microcosm of International Banking Trends SEPTEMBER Susan A. Whitlock The Role of the Money Supply in the Conduct of M onetary Policy Arthur F. Burns Evolution of the Concept of the Dem and for M oney Thomas M. Humphrey MONTHLY REVIEW, DECEMBER 1973