ARehabilitationof MonetaryPolicy inthe 1950’s

By CHRISTINA D. ROMER AND DAVID H. ROMER*

Americanmonetary policy in the 1950’ s has FederalReserve of the 1990’ s. In particular, typicallynot been judged favorably. Monetar- theFederal Reserve in the early postwar era istssuch as Milton Friedman (1960), Karl showedthe same overarching concern about Brunnerand Allan H. Meltzer(1964), and ination that is the hallmark of post- CharlesW. Calomirisand David C. Wheelock monetary-policyorthodoxy. We alsoŽ ndthat (1998)criticize the Federal Reserve for stop- theFederal Reserve of the 1950’ s was not and-gopolicies and for amistakenfocus on free weddedto faulty indicators in its implementa- reserves.Keynesians such as Alan S. Blinder tionof policy.Finally, empirical analysis of the andStephen M. Goldfeld(1976) argue that the behaviorof the federal funds rate shows that FederalReserve targeted output below the nat- policymakersin the 1950’ s respondedmuch uralrate and therefore unnecessarily restrained moreaggressively to expectedin ation than did outputgrowth. policymakersin the 1960’ s and1970’ s. Theseunfavorable judgments seem strangely atodds with economic performance in this I.NarrativeEvidence decade.In ation, measured using the GDP de- ator,averaged under 2.0 percent per year be- Giventhat the time period is short, it ishard tween1952 and 1960, and it never went above totest statistically whether the Federal Reserve 3.3percent in a singleyear. Real GDP overthe ofthe 1950’ s was blessedwith good sense or sameperiod grew atan average rate of 2.9 goodluck. For thisreason, it is most useful to percentper year, and the unemployment rate analyzenarrative evidence. The records of the averaged4.7 percent. While there were two FederalReserve, speciŽ cally the Minutesof the recessionsduring this decade, that in 1954 was FederalOpen Market Committee (Board of exceedinglymild, and that in 1958 was sharp Governorsof theFederal Reserve System, var- butvery brief. Although this unquestionably iousyears) and the Congressi onaltesti- goodeconomic performance is not proof that monyof Federal Reserve Chairman William monetarypolicy was similarlygood in the McChesneyMartin,can reveal both the mo- 1950’s, it is certainly suggestive. At thevery tivationbehind policy actions and the pre- least,it implies that those who would criticize vailingframework usedto understa ndthe monetarypolicy in this decade are left with a macroeconomy. mystery:Why was performanceso good if mon- etarypolicy was pooror inept? A. AnOverarching Concern about In ation Thispaper suggests an alternative view of monetarypolicy in the 1950’ s, and hence a Themost obvious and signiŽ cant belief re- possiblesolution to themystery of thatdecade’ s vealedby the Minutes isa fundamentalabhor- outstandingeconomic performance. We show renceof ination by virtuallyall members of the thatpolicy in the 1950’ s was actuallyquite FederalOpen Market Committee (FOMC). In- sophisticated.Narrative evidence on the moti- deed,in reading the Minutes, oneperiodically vationof policymakersand their understanding hasto double-check the data. The discussion oftheeconomy shows that the Federal Reserve was oftenso ferventand the predictions so dire ofthe 1950’ s was remarkablysimilar to the thatit is hardto believethat in ation was actu- allyvery low. Theoverarching concern about in ation is revealedmost clearly in the statements the *Department ofEconomics, University of , membersmade and the actions they endorsed Berkeley,CA 94720-3880.We are gratefulto theNational Science Foundationfor Ž nancialsupport, to Patrick McCabe duringthe times when in ation began to accel- forresearch assistance, andto Anna Schwartz forcomments erate,if only modestly, in the mid and late andsuggestions. 1950’s. For example,in mid-1955the economy 121 122 AEAPAPERS AND PROCEEDINGS MAY 2002 was quitewell recovered from therecession of havein ation and ... theremust be serious 1953–1954, and there were fears thatprices shocktreatment” (p. 27). were aboutto rise. Many members of the Theconcern over in ation and the desire FOMCspoke about the need to actdecisively to for tightpolicy continued for mostof 1959. preventin ation. In August, Chairman Martin InFebruary, H. G.Leedysummarized his view saidin one of his rare preparedstatements to the oftherole of monetarypolicy: “ TheSystem, of FOMC:“ Ination is a thiefin the night and if course,wanted growth as well as stability, but if we don’t actpromptly and decisively we will temporarilythere had to be a choicebetween alwaysbe behind” ( Minutes, 2August1955, growthand arresting in ationary psychology he p.13).In November, Governor J. L.Robertson wouldfavor the latter course” ( Minutes, 10 Feb- said,“ Ifeelthat there are in ationary pressures ruary1959, p. 22). In late May, Vice Chairman presentwhich should be checked now by a Hayesannounced that “ Inthe light of these Žrmer monetarypolicy— one Ž rm enoughto threatsto oureconomy, I amconvincedthat the curtailspending and thus dampen price pres- timehas come for adecisivesignal of theFed- sures” (Minutes, 16November1955, p. 20; em- eralReserve System’ s determinationto do its phasisin the original). In response to these partto check in ationary trends” ( Minutes, 26 concerns,the discount rate was raisedby afull May1959, p. 17). 1 percentagepoint between April and November, andother contractionary measures were taken. B. Modelof the Economy Thedislike of ination and the desire to Ž ght itwere evenmore obvious in 1958. Almost as Thenarrative record also provides crucial soonas the trough of the 1957– 1958 recession evidenceabout why monetary policymakers in was reachedin the spring of 1958, the FOMC the1950’ s dislikedin ation so. Their model beganto worry aboutin ation. The members ofhow the macroeconomy operated contained feltthat they had not reacted soon enough in botha remarkablymodern view of the causes of 1955,and they were willingto risk another ination and a Žrm beliefthat the output costs slowdownand Congressional anger to keep in- ofin ation were largeand imminent. As are- ationfrom risingagain. Chairman Martin said sult,they Ž rmlybelieved that in Ž ghtingin- “hedid not think that the System had faced in ationthey were encouragingboth short-run recentyears anything like the present problem, stabilityand long-run growth. whetherit be calledan inationary psychosis or Akeyfeature of the model of many FOMC inationary psychology. He didnot know how memberswas asensibleview of capacityor full todeal with the speciŽ cs of theproblem except employment.Most policymakers believed that bymoving in the right direction within the ination began to rise when there was stillsig- System” (Minutes, 19August 1958, p. 59). In niŽcant unemployment. For example,in July doingso, however, the System would have “ to 1955,when unemployment was 4.0percent, havecourage to assume the risks that were ViceChairman said that the econ- involved”(p. 58). By September, interest rates omywas “nearerthan we havebeen since early hadrisen back to their 1957 peak level, and 1953to fullutilization of plant, equipment, and ViceChairman Alfred Hayesexpressed concern manpower;prices which have been stable, in thatfurther action “ couldlead to interest-rate theaggregate, for twoyears may be aboutto get levelsso high as tobe harmful to the economy apushon theup-side due to pressure from costs andso high as to place the System in political andfrom anticipationof pricerises by business- jeopardy” (Minutes, 9September1958, p. 12). men,purchasing agents, and consumers” ( Min- His concern,however, was notshared by most utes, 12July 1955, pp. 26 –27).At thenext othermembers. Chairman Martin responded that“ If theSystem should lose its independence inthe process of Žghtingfor soundmoney, that 1 Monetarypolicymakers in the 1950’ s alsoexpressed wouldindeed be a greatfeather in its cap and concernover unemployment and output growth on many ultimatelyits success would be great” (p. 53). occasions.Similarly, the FOMC expressed substantial con- cern aboutmaintaining stability in the bond market and GovernorJames Vardaman also expressed the soughtto avoid tightening around times oflarge Treasury viewthat Ž ghtingin ation was ofparamount reŽnancing operations. However, these concernswere importance.He said,“ thecountry was goingto clearly dominatedby the concern over in ation. VOL. 92 NO. 2 MONETARY-POLICYRULES IN PRACTICE 123 meeting,Malcolm Bryan said that “ theapparent ination were thoughtto occur quite quickly. presenttrends in the economy simply extend Indeed,in ation could actually cause a reces- themselvesto over-reach comfortable capacity sion.Martin expressed this view very clearly andthat, accordingly, an in ation is inevitable” inCongressional testimony in 1959. He stated, (Minutes, 2August1955, p. 23).Watrous Irons “Ihappento believe, Mr. Patman, that the subscribedto thesame view a few monthslater, 1957–58 recessionwas adirectresult of letting saying,“ Theeconomy was movingnearer ca- ination get substantially ahead of us”(Martin, pacityin many respects, and as this point ap- 1959bp. 1285). Thomas, the chief economist, proachedless efŽ cient means of production expresseda similarview. In , he wouldbe utilized and prices would tend to rise” said,“ Increasingdemands after mid-1955 re- (Minutes, 4October1955, p. 8). Again in 1959 sultedin relativelysmall increases in outputbut whenunemployment was 5.0percent, Woodlief markedadvances in prices ... .Distortionssuch Thomas,the chief economist, said, “ Theecon- asundue inventory accumulation, too hasty cap- omyis approaching the limits of resourceutili- italexpansion in some areas, too rapid a risein zation” (Minutes, 16June 1959, p. 6). debtburden, and consumer resistance to price Themembers of the FOMC and the Board increasesundermined the prevailing high activ- staff were certainlyaware thatthere was a ityand led to the recession of 1957– 58” ( Min- short-runtrade-off between in ation and output. utes, 22September 1959, p. 8). However,they were unitedin believing ada- Thebelief in theabsence of a long-run(pos- mantlythat there was nota positivelong-run itive)trade-off is certainly much more modern trade-off.Indeed, by far themost common view thanthe simplistic Keynesian model that held was thatif excessive demand resulted in in a- swayin the 1960’ s and1970’ s. Indeed, many of tion,output would actually fall in the long run. thestatements made by FOMCmembersin the Thisview is similar to those of many current 1950’s couldbe inserted into the narrative monetarypolicymakers, such as Alan Green- recordfor the1980’ s and1990’ s withoutnotice. span(see e.g.,Greenspan, 1997). Thatthe Federal Reserve had this model in the Thiswas clearlyChairman Martin’ s view. 1950’s suggeststhat the passionate statements Martinsaid in 1958: “ If ination should begin aboutthe dangers of in ation were notmere todevelop again, it mightbe thatthe number of windowdressing. Rather, they were partof a unemployedwould be temporarily reduced to coherentview that placed predominant empha- fourmillion [from thecurrent level of Žvemil- sison keeping in ation in check. lion],or some Ž gurein that range, but there wouldbe alargeramount of unemploymentfor C. Implementationof Policy alongtime to come. If ination should really geta headof steam up, unemployment might Brunnerand Meltzer (1964), Calomiris and riseto ten million or Ž fteenmillion” ( Minutes, Wheelock(1998), and others argue that an im- 19August 1958, p. 57). Martin repeated this portantsource of policymistakes in the 1950’ s viewin Congressional testimony in 1959, say- and1960’ s was afocuson free reserves(total ing:“ If totaldemands tend to run ahead of the reservesless required reserves less borrowed outputpotential, the general price level will reserves).And, there is no doubt that free re- beginto rise and this, in turn, will have an servesplayed an important role in policy in the adverseimpact both on growthof demandsand 1950’s. For example,most FOMC meetings onmeans of Ž nancingincreased and improved endedwith some discussion of atargetfor free capacity.It will also have adverse effects on the reserves. efŽciency with which resources are utilized” However,we Žndnoevidencethat this focus (Martin,1959a p. 118). onfree reserveswas predominantor led to Twofeatures of this framework arenotewor- persistentmistakes. The narrative record shows thy.The Ž rst isthat the level of in ation at thatthe FOMC also paid close attention to in- whichMartin and others felt negative conse- terestrates, and goals for keyinterest rates were quenceswere likelywas verylow. No onewas oftenused as a supplementto instructionsabout contemplatingin ation of more than 5 percent free reserves.A verycommon instruction was whenmaking the dire predictions of long-run thatthe Account Manager should pay close consequences.Second, the negative effects of attentionto the “ color,feel, and tone of the 124 AEAPAPERS AND PROCEEDINGS MAY 2002 market” (Minutes, 30September 1958, p. 46). Theseconsiderations suggest that while tar- Toa largedegree, this instruction meant that he getsfor free reserveswere importantin the was towatchshort-term interest rates. Often the short-runimplementation of policy, nominal in- roleof interest rates was moreexplicit. For terestrates were predominantover longer hori- example,when Vice Chairman Sproul gave a zons.And since in ation varied little in the detailedsummary of whatvarious terms such as 1950’s, nominal interest rates provided a good “activeease” or “restraint”meant, the behavior indicationof tightness in credit markets. Fur- ofinterestrates was central( Minutes, 11 Janu- thermore,many FOMC membersshowed a ary1955, pp. 10 –12).Indeed, the FOMC often clearunderstanding of the distinction between chosethe target for free reservesto try to attain realand nominal interest rates. For example,in aparticularinterest-rate outcome. In ,Karl Boppsaid, “ Onereason for the 1955,for example,Martin asked the Account presentlevel of interestrates is the anticipation Managerwhat “ operations... mightbe followed offurtherin ation” ( Minutes, 13October1959, for theSystem account to provide a minimum p. 15). disturbanceto themarket during the immediate TheFOMC was alsoacutely aware ofthe future[that is, to keep interest rates steady]” lagsassociated with monetary policy. The (Minutes, 25January 1955, p. 9). The Account membersoften worried that in ation, while cur- Managerresponded by suggesting a rangefor rentlylow, was aboutto take off. For example, free reserves,and the FOMC adopted a target inSeptember 1958, Leedy said that “ theSystem withinthat range. And when the Committee shouldnot postpone the matter of lookingat the expecteda shiftin therelationship between free possibilityof in ation ahead of it. There were reservesand interest rates, it typically changed signsof recovery on every hand, and if the thereserves target. In , for example, Systemshould wait until there was recovery theFOMC expected that without open-market beyondany shadow of a doubtit seemed to operations,there would be alargefall in free himthat the System would have lost its oppor- reserveswith only slight upward pressure on tunityto do the kind of a jobthat it was sup- rates.Since the Committe efeltthat some rise posedto be doing”( Minutes, 9September1958, inrates was desirable,it decided to allow p.32).Similarly, in September1959, Governor thelarge decline ( Minutes, 29March 1955, Robertson“ expressedthe view that the System pp. 5–9). oughtto adopt an afŽrmative position of restric- Anexamination of the data on free reserves tivenessin orderto keepon top of thepotential andinterest rates conŽ rms thekey role of inter- inationary situation ahead. Otherwise, the Sys- estrates in policy-making. For mostperiods, temwould get behind the game and might never free reservesand the federal funds rate move catchup— repeating the mistakes of afew years togetherclosely, but in oppositedirections. The ago” (Minutes, 1September1959, p. 21). mainexception occurs in 1956, when both se- riesrise considerably. Throughout the year the II.StatisticalEvidence FederalReserve expressed an intentto increase thedegree of restraint. The “ Recordof Policy Toseeif policymakersbacked up their words Actions”for the27 March meeting states that withactions, one needs to suppleme ntthe theCommittee felt that “ theSystem would be narrativeanalysis with statistical evidence. To derelictin its duty if it did not exercise addi- thisend, we lookat how the federal funds tionalrestraint.” On 17 April, the Committee rateresponded to developments in the macro- “agreedthat there should be no relaxation of economyin the 1950’ s andcompare those re- pressures”; atthe 7 Augustmeeting, it moved sponseswith the responses in other periods. 2 “tostrengthen credit restraint” ; andon 21 Au- Becausethe 1950’ s sampleperiod is inherently gust,“ TheCommittee felt that credit policy limitedand the variation in in ation in this shouldbe made somewhat more restrictive” decadeis small,this empirical analysis must be (Boardof Governors, 1956 pp. 26, 28, 36, 37). Therise in free reserves,therefore, was evidently 2 JohnB. Taylor(1999) shows that the response of the anaccommodative move taken to achieve a de- federal fundsrate toeconomic variables provides a sensible siredbehavior of interest rates in the face of shifts descriptionof policy even in eras whenthe Federal Reserve inthe normal behavior of reserves. was more directlytargeting some othervariable. VOL. 92 NO. 2 MONETARY-POLICYRULES IN PRACTICE 125 viewedas a suggestivecheck on the narrative TABLE 1—ESTIMATED FORWARD-LOOKING analysisrather than as a conclusivetest. MONETARY-POLICY RULE

Sample In ationa Outputb Constant A. SpeciŽcation 1952:1–1958:4 1.178 (0.876) 20.040(0.295) 20.562(1.874) 1964:1–1979:3 0.891 (0.090) 0.269 (0.112) 1.410 (0.517) Theparticular speciŽ cation that we consider 1979:4–1987:3 1.263 (0.187) 20.056(0.287) 4.614 (0.992) isa forward-lookingTaylor rule (see e.g., Richard 1987:4–2000:4 1.390 (0.305) 0.672 (0.315) 2.311 (0.760) Claridaet al.,2000). In its simplest, descriptive Note: Numbers inparentheses are standarderrors. form, aTaylorrule shows how the Federal a One quarterahead. Reservechooses the federal funds rate in re- b Deviationof outputfrom trend, one quarter ahead. sponseto in ation and departures of output from trend.A forward-lookingTaylor rule takes We estimatethe rule over four samples. The intoaccount the fact that the monetary authority 1950’s sampleis 1952:1– 1958:4. We starttwo typicallyresponds to expectationsof thesevari- yearsinto the decade because the Federal Reserve ables.As discussedabove, this forward-looking was unableto pursue independent monetary pol- behaviorwas animportant feature of policy- icyuntil the Treasury– Federal Reserve Accord of makingin the 1950’ s. 1951.We stopat the end of 1958 for reasons Theforward-looking Taylor rule that we con- discussedbelow. The second sample corresponds sideris simply roughlyto the late 1960’ s andthe 1970’ s; it runs from 1964:1to 1979:3. The third and fourth samplesare the Volcker and Greenspan eras: # (1) it 5 a 1 bEtpt 1 1 1 gEt ~Y 2 Y!t11 1979:4–1987:3and 1987:4 –2000:4,respectively. where i isthe federal funds rate, p isin ation, B. Results and Y 2 Y# isthedeviation of outputfrom trend. Timeis measured in quarters. To implement ThecoefŽ cient estimates are given in Ta- thisspeciŽ cation, we regressthe federal funds ble 1.4 Themost important result is that the rateon the leads of actual in ation and the weighton expectedin ation in the policyrule in deviationof output from trend,instrumenting the1950’ s isquite similar to that in the Volcker withinformation known at time t.For instru- andGreenspan eras, and noticeably larger than ments,we use(in addition to the constant) the thatfor the1960’ s and1970’ s. In both the contemporaneousand two lagged values of in- 1950’s andthe last two decades of the 20th ationand the contemporaneous deviation of centurythe point estimate is greater than 1, outputfrom trend.We usemultiple lags of indicatingthat in response to a risein in ation ination because the quarterly series tends to theFederal Reserve raised the nominal funds uctuatesubstantially. The deviation of output rateby enough to also raise the real funds rate. from trend,in contrast, is quite smooth, so the Inthe late 1960’ s and1970’ s thecoefŽ cient is contemporaneousvalue is anexcellent predictor below1, indicating that the Federal Reserve ofnextperiod’ s value. 3 reducedthe real funds rate when in ation rose. Theweight on expectedin ation is estimated lessprecisely in the 1950’ s thanin other de- cades.However, the point estimate and the nar- 3 Data onthequarterly average ofthefederal fundsrate rativeevidence presented in Section I tella very for1954:1 to 2000:4 are takenfrom Citibase. We extend similarstory. The Federal Reserve of the 1950’ s thisseries backto 1950:1 using data fromEdward J. was deeplyconcerned about in ation and acted Martens(1958). (The data inMartens [1958] are reported aggressivelyto control it on several occasions. onlyin graphical form. After deducingthe numbers from thegraph, we checkedand calibrated our deductions in a Thiscan be seen in Figure 1, which shows the periodof overlap between theseries inMartens and that fromCitibase.) We measure ination as thequarter-to- quarterchange in the log of the GDP deator (at anannual 4 We alsorun the regressions using the three-month rate). Thedeviationof outputfrom trend is calculated as the Treasurybill rate as theindicator of policy stance andthe difference between thelog of real GDP anda logtrend. The deviationof quarterly industrial production from trend as trendseries is estimated usingthe Hodrick-Prescott Ž lter themeasure ofthe output gap. Neither ofthese changes appliedto the period 1952:4 – 2000:4. affects theresults appreciably. 126 AEAPAPERS AND PROCEEDINGS MAY 2002

FIGURE 1. FEDERAL FUNDS RATE FIGURE 2. FEDERAL FUNDS RATE AND EXPECTED INFLATION AND EXPECTED OUTPUT DEVIATION federalfunds rate and expected in ation (mea- suredas theŽ ttedvalues of the regression of the ofconcernabout in ation, even if thatconcern leadof in ation on the instruments) during the isnot captured by our regression estimates. 1950’s. This Ž gureshows that there is a close ThecoefŽ cient estimates reported in Ta- andstrong relationship between the two series ble1 showthat the weight put on the expected for muchof thedecade. outputgap in the 1950’ s was small.The coef- Partof the imprecision of the estimation is Žcientis essentially zero and very imprecisely theresult of theFederal Reserve being particu- estimated.Figure 2 graphsthe expected out- larlyconcerned about expected in ation in the putgap (measured as the Ž ttedvalues from late1950’ s. Figure 1 showsthat, while expected theregression of the output gap at t 1 1 on the ination derived from theŽ rst-stageregression instruments) andthe federal funds rate in roseslightly in 1958,its rise was smallrelative the1950’ s. The obvious positive correlation tothe tightening by the Federal Reserve. As a betweenthe two series does not show up inthe result,this looks like a timewhen the Federal multipleregression because of correlation be- Reservewas notresponding to expected in a- tweenexpected in ation and the output gap. tion.(Furthermore, because expected in ation derivedfrom theŽ rst-stageregression falls in III.Conclusion 1959,if one continues the estimation through 1959the estimated coefŽ cient on in ation falls Likecentral bankers of the1990’ s, monetary considerablyand is measuredeven more impre- policymakersof the 1950’ s hada deep-seated cisely.)However, as described in SectionI, the dislikeof in ation and acted to control it. Their mainreason for thetightening by the Federal dislikeof ination was rootedin amodelof the Reserveat the end of the 1950’ s was itscon- economythat emphasized the costs of ination victionthat in ation was aboutto rise. In this andthe absence of apositivelong-run trade-off context,it is useful to note that the Federal betweenoutput and in ation. Reservewas notalone in fearingin ation at the TheseŽ ndingsprovide important insights endof the 1950’ s. The Livingston survey of intothe performance of the economy in the expectationsfor theCPI sixmonths ahead rose 1950’s. One key reason that in ation was low steadilyfrom mid-1958through the end of andsteady was almostsurely that the Federal 1959.5 Thus,the Federal Reserve was actingout Reservewas workingto achieve those goals. Andone likely reason that recessions were brief andmild is that in ation never got seriously out 5 TheLivingston survey data are fromthe Federal Re- ofhand.As aresult,the Federal Reserve never serve Bankof Philadelphia web site: http://www.phil. hadto undertake a disination of themagnitude frb.org/econ/liv . ofthoseof the 1970’ s and1980’ s. VOL. 92 NO. 2 MONETARY-POLICYRULES IN PRACTICE 127

Our Žndingsmay also provide insight into the for AmericanMonetary Policy?” in Michael policymistakes in the late 1960’ s and1970’ s. If D.Bordo,Claudia Goldin, and Eugene N. monetarypolicymakers in the 1950’ s hadŽ gured White,eds., ThedeŽ ning moment: The Great outthe essence of sensible policy, the mistakes of Depressionand the American economy in the the1960’ s and1970’ s cannotjust have been the twentiethcentury. :University of resultof continuing ineptitude or misunderstand- ChicagoPress, 1998,pp. 23– 65. ing.Rather, something must have changed. One Clarida,Richard; Gal ´õ ,Jordiand Gertler,Mark. obviouscandidate is the model of the economy. “MonetaryPolicy Rules and Macroeconomic ThomasMayer (1998) and Taylor (1999) suggest Stability:Evidence and Some Theory.” Quar- thata naiveKeynesia nmodelwith an exploit- terlyJournal of , February2000, abletrade-off between output and in ation and, 115(1),pp. 147– 80. later,a natural-ratehypothesis with an unreal- Friedman,Milton. Aprogramfor monetary sta- isticallylow estimate of the natural rate were bility. :Fordham University Press, thekey sources of the in ation of the 1960’ s 1960. andthe 1970’ s. Our Žndingthat these models Greenspan,Alan. “CurrentMonetary Policy.” areso different from thatin the low-in ation Speechat theStern School of Business, New 1950’s andpost-Volcker 1980’ s and1990’ s YorkUniversity, 8 May1997. addscredence to this view. Martens,Edward J. “FederalFunds: A Money MarketDevice.” Unpublis hedmanuscrip t, REFERENCES PaciŽc CoastBanking School, April 1958. Martin,William McChesney. “A Year of Blinder,Alan S. and Goldfeld,Stephen M. “New RecessionandRecovery .”Statementbe- Measuresof Fiscal and Monetary Policy, fore theJoint Economic Committe e,6Feb- 1958–73.” AmericanEconomic Review, De- ruary1959; reprinted in the Federal cember1976, 66(5),pp. 780 –96. ReserveBulletin , February1959a, 45(2), Boardof Governors of the Federal Reserve Sys- pp. 110 –19. tem. 43rdAnnual Report. Washington,DC: . Testimony,in Hearingsbefore the Boardof Governors of the Federal Reserve JointEconomic Committee, 86thCongress, System,1956. 1stSession, 27 . Washington, DC: . Minutesof the Federal Open Market U.S. GovernmentPrinting OfŽ ce, 1959b, pp. Committee. Washington,DC: Boardof Gov- 1231–1504. ernorsof the Federal Reserve System, vari- Mayer,Thomas. Monetarypolicy and the Great ous years. Ination in the : The Federal Brunner, Karland Meltzer,Allan H. The Fed- Reserveand the failure of macroeconomic eralReserve’ s attachmentto the free re- policy,1965– 79. Cheltenham,UK: Elgar, serveconcept. HouseCommittee on Banking 1998. andCurrency, 88th Congress, 2nd Session. Taylor,John B. “AHistoricalAnalysisof Washington,DC: U.S. GovernmentPrinting MonetaryPolicy Rules,” in JohnB. Taylor, OfŽce, 1964, pp. 1– 64. ed., Monetarypolicy rules. Chicago:Uni- Calomiris,Charles W. and Wheelock,David C. versityof Chicago Press, 1999,pp. 319 – “Was theGreat DepressionaWatershed 41.