Speech

Inflation Targeting JuniorAchievementofArkansas,Inc. LittleRock,Arkansas February16,2006 PublishedintheFederalReserveBankofSt.LouisReview,May/June2006,88(3),pp.155-63 Iam delighted to speak with Junior Ifindcomputersabitmysterious,andIknow Achievement of Arkansas. I cannot report thatmanythinkthatmonetarypolicyiseven astoryabouthowJuniorAchievement(JA) moremysterious.FederalReserveofficialsused got me off to a good start, but I do have a todelightinaddingtothemystery,buttoday personal story—from my oldest son, Will. When advancesinmacroeconomictheoryhavemade I accepted this speaking invitation, I asked Will cleartheimportanceofcentralbanktransparency to reflect on his JA experience, and here is the toaneffectivemonetarypolicy. paragraph he sent me. SincecomingtotheSt.LouisFedin1998,I havespokenoftenonthesubjectofthepredict- IwasinvolvedinJuniorAchievementwhenI abilityofFederalReservepolicy,emphasizing wasin8thgrade.Mostentrepreneurial-minded kidsIknewgainedtheirbusinessexperiences thatpredictabilityenhancestheeffectivenessof onpaperroutes,paintinghousesorthelike. policy.1Predictabilityhasmanydimensions,but ButIwasdrawntoJA’sconceptofteachingthe oneiscertainlythatthemarketcannotpredict basicsofbusinessandfiguringouthowto whattheFedisgoingtodowithoutadeepundermass-producesomething.LittledidIknow standingofwhattheFedistryingtodo. whereitwouldleadme.InmyJAgroup,we TheFedhasstatedformanyyearsthatakey assembledwoodencoatpegsonboardsand monetarypolicyobjectiveislowandstable paintedthemupnicely.Iquicklylearnedthat inflation.Ibelievethataddingformalitytothat buildingasinglecoat-rackwidgetisnotso objectivecanclarifywhattheFeddoesandwhy. hard,butleadingahandfulofpeopletomake 50,withquality,ismuchharder.Andthatget- Thatismytopictoday. tingallofthemsoldforaprofitisevenharder Beforeproceeding,Iwanttoemphasizethat yet.Ican’tsayexactlywhichoftheskillsI theviewsIexpressherearemineanddonot learnedatJAhelpedmeenduprunningthe necessarilyreflectofficialpositionsoftheFederal WindowsbusinessatMicrosoft.Iwasabig ReserveSystem.Ithankmycolleaguesatthe dreamerbackthen,butevenIwouldnothave FederalReserveBankofSt.LouisfortheircomdreamtthatIwouldsomedaybeleadingateam ments;DanielL.Thornton,vicepresidentinthe of3,000professionalsthatcreatesoftwarethat ResearchDivision,providedspecialassistance. isusedin169countriesaroundtheworldand Itakefullresponsibilityforerrors. powers200,000,000newPCssoldeveryyear. JA,thanksforthejump-start! WillisaseniorvicepresidentforMicrosoft’s THE FRAMEWORK Windowsclientbusiness.Needlesstosay,Iam immenselyproudofhim.Idon’tknowthelist, TheFederalOpenMarketCommittee(FOMC) butwillbetthatnumerousotherJAalumniare hastheresponsibilitytodeterminemonetary inveryresponsiblepositionstoday. policy.TheCommitteeimplementspolicyby 1 SeePoole(1999)forthefirstofaseriesofspeechesonthistopic. 1

MONETARYPOLICYANDINFLATION settingatargetforthefederalfundsrate.Policy ofthetargetrange.I,givenmypreferredtarget predictabilitydoesnotmeanthatthepublicor range,wouldargueforasomewhatmorerestricthemarketscansuccessfullyforecastthetarget tivestance,tomovetheinflationratedowntoward federalfundsratenextweek,nextmonth,ornext thecenterofmypreferredrange.Thedifference year.Thetargetrateisbasedonpolicymakers’ betweenthesetwotargetrangesissmall,andyet currentinformationandbestestimateoffuture thatdifferencemightbeenoughtocallforasomeeconomicevents;thekeyobservationisthat whatdifferentpolicystance. incominginformationmaydepartfromthebest Obviously,theFedcannotsimultaneously estimateandindicatethatthetargetfundsrate pursuetwodifferentinflationgoals,andtherefore needstobechangedtoachievepolicyobjectives. thereiseveryreasonfortheCommitteetoagree Whatwemustmeanbyperfectlypredictableis onacommonobjective.Anagreed-uponcommon thatthepublicandthemarketsarenotsurprised objectiveismuchmoreimportantthanthesmall bytheFed’sresponsetothelatesteconomic differencebetweenmyownpreferredobjective information,understandingthattheinformation andtherangeofobjectivesIbelievearefavored itselfisnotpredictable. byothers. Althoughnewinformationcreatesasteady IftheFOMCweretodecideonacommon streamofmostlyminorsurprises,theFOMCought objective,thentheCommitteecouldcommunicate tobeclearaboutwhatitistryingtoaccomplish. ittothegeneralpublic.Discussionoftheformal, Atpresent,mostmembersoftheCommittee numericobjectiveandwhatitmeanswouldhelp wouldprobablybeprettyclosetogetheronhow marketstobetterunderstandmonetarypolicyand tostatetheinflationgoal.Abenefitofgreater wouldmakepolicymorepredictable.However, formalityindefiningtheinflationgoalisthat manydetailsmatterandaninflationtargetwill individualFOMCmemberswouldhaveaclearer notbeasourceofincreasedclarityunlessthe ideaastowhattheinflationobjectiveis. detailsarespecifiedappropriately.So,let’stalk Toillustratethispoint,Ihaveoftensaidthat aboutthoseimportantdetails.Tosimplifythe mypreferredtargetrateofinflationis“zero, language,I’llrefertoapubliclyannounced,speproperlymeasured.”Thatis,allowingasbestwe cificnumericaltargetrangeforinflationasa canformeasurementbias,whichmightbeinthe “formal”inflationtargetorobjective. neighborhoodofhalfapercentperyearforbroad measuresofconsumerprices,Ifavorliterallyzero inflation.Givenmeasurementbiasinprice WHAT IS INFLATION? indices,Imightstatemygoalasinflationbetween 0.5and1.5percentasmeasuredbythepriceindex IftheFOMCisgoingtoadoptaformalinflation forpersonalconsumptionexpenditures(thePCE objective,weneedtoagreeonwhat“inflation”is. priceindex).Otherspreferasomewhathigher Howeverinflationismeasured,itisimportantto rateofinflation,perhapsintherangeof1to2 distinguishbetween“highfrequency”inflation, percentasmeasuredbythePCEpriceindex.Still whichcentralbankshavelittlecontrolover,and othersmightfavoradifferenttargetrange,witha “lowfrequency”inflation,whichcentralbanks differentmidpointand/orawiderornarrower cancontrol.High-frequencyinflationistherate range.IftheFOMCdecidestodiscussinflation ofchangeinthepriceleveloverrelativelyshort targeting,alldimensionsofspecifyingatarget timeperiods—months,quarters,orperhapseven willbeconsideredcarefully. ayear.Low-frequencyinflationisaneconomy- Whydoesprecisiononatargetrangematter? wide,systemicprocessthatisaffectedbypast, Considerasituationinwhichtheactualrateof present,andexpectedfutureeconomicevents. inflationis1.5percent.Thosefavoringatarget Centralbanksacceptresponsibilityforlowrangeof1to2percentwouldsaythatthepolicy frequencyinflationbecausesuchinflation stanceisjustright;inflationisintheexactcenter dependscriticallyonpastand,especially, 2

InflationTargeting expectedfuturemonetarypolicy.WhenIadvocate fallingandothersarerising,thentheaverageof thattheFedestablishaformalinflationobjective, allprices,orthepricelevel,canremainconstant. Iamspeakingofthelow-frequencyinflationrate. Nevertheless,definingapriceindexwhenprices Asapracticalmatter,low-frequencyinflationcan arechangingatdifferentratesinvolvesmeasurebethoughtofastheaverageinflationrateovera mentissuesthatarecomplicatedatbothconcepperiodofafewyears. tualandpracticallevels.Foravarietyoftechnical reasonsthatIwon’tdiscuss,thebestwecando istoapproximatethetheoreticalconstructofthe SETTING THE TARGET RATE OF pricelevel.Expertsbelievethatpriceindices,such astheconsumerpriceindex(CPI)andthePCE INFLATION priceindex,haveanupwardbias.Thatis,ifthe TheEmploymentActof1946setsobjectives pricelevelweretrulyunchanged,thepriceindex formonetarypolicy—indeed,objectivesforall wouldshowalowrateofinflation. economicpolicy.2TheActdeclaresthatitisthe WhenaskedduringtheJuly1996FOMCmeet- “responsibilityoftheFederalGovernment...to ingwhatlevelofinflationdoesnotcausedistorpromotemaximumemployment,production,and tionstoeconomicdecisionmaking,Chairman purchasingpower.”Theseobjectivesarereflected Greenspanresponded,“zero,ifinflationispropintheFOMC’stwinobjectivesof“pricestability” erlymeasured.”4Greenspan’sviewthatthetheoand“maximumsustainableeconomicgrowth.” reticallycorrectdefinitionofpricestabilityiszero Althoughuseful,thesephrasesaresomewhat inflationstemsfromhisbeliefthateconomic vague.Forexample,inthelate1970sandearly growthismaximizedwhenthepricelevelis 1980s,theFedpursuedthegoalofpricestability unchangedonaverageovertime.5WhileIbelieve byreducinginflationfromdouble-digitrates;from thatthereisavirtualconsensusthattheeconomy themid-1980sintotheearly1990s,thegoalwas functionsbestwhenthetheoreticallycorrectmeastobringinflationdownfromthe4percentneigh- ureofinflationis“low,”noteveryoneagreeswith borhood.Overthepastdecadeorso,thegoalhas Greenspanthattruepricestability—azerorate cometomeankeepingtheinflationratelow. ofinflationproperlymeasured—isthebesttarget Butwhatinflationrateconstitutespricesta- fortheFed.Foravarietyofreasons,someeconobility?Ratherthananumericaldefinition,former mistsbelievethattheeconomyfunctionsbest ChairmanGreenspanpreferredaconceptual wheninflationcorrectlymeasuredis“low”but definition,suggestingthat“pricestabilityisbest notzero. thoughtofasanenvironmentinwhichinflation Whilethegoalofpricestabilityisspecificin issolowandstableovertimethatitdoesnot boththeFederalReserveActandtheEmployment materiallyenterintothedecisionsofhouseholds Actof1946,somesuggestthattheFOMClacks andfirms.”3ButdoesGreenspan’sdefinition theauthoritytoestablishanumericalinflation requirezeroinflation? objective.TheyclaimthatonlyCongresshasthis Becausemeasuringthepricelevelisadaunting authority.ThatCongresshasthepowertoestabtask,zerotrueinflationandzeromeasuredinfla- lishthegoalsofeconomicpolicyisindisputable; tionmaydiffer.Pricesofindividualgoodsand however,itdoesnotfollowthattheFOMCdoes serviceschangeovertime,butifsomepricesare nothavetheauthoritytoadoptaformalinflation 2 SeeSantoni(1986)foradiscussionofthecreationoftheAct. 3 Greenspan(2002,p.6). 4 TranscriptoftheFOMCmeetingheldonJuly2-3,1996,p.51. 5 Forcompleteness,InotethatFriedman(1969)arguedthattheoptimalrateofinflationwasnegative.Specifically,hesuggestedthateconomic welfarewasmaximizedwhenthenominalinterestratewaszero.Thisrequiresthattheinflationrateisequaltonegativeoftherealinterestrate. 3

MONETARYPOLICYANDINFLATION objectiveaspartofimplementingitsbroadcon- Adoptingaformalinflationobjective,and gressionalmandate.Itiscommonpracticefor successinachievingthatobjective,willalso Congresstoestablishobjectivesandguidelines enhancepolicymakers’abilitytopursueother andleaveituptotheagencyresponsiblefor policyobjectives,suchasconductingcountermeetingthoseobjectivestofillinthedetails. cyclicalmonetarypolicy.Isuspectthatsomeof Therealquestionisthis:ShouldtheFOMC thosewhoopposeaspecificinflationobjectiveare announcewhatitsinflationobjectiveis?Answer- concernedthatdoingsowillcausepolicymakers ingthisquestionissimpleinprinciple.Ifannounc- tobecomewhatMervynKing,Governorofthe ingaspecific,numericalinflationobjective BankofEngland,hascolorfullytermed“inflation enhancestheefficacyofmonetarypolicy,thenthe nutters.”King(1997)isreferringtopolicymakers answerisyes.Ifdoingsoreducestheefficacyof whoaimtostabilizeinflation,whateverthecosts. monetarypolicy,theanswerisno.Ibelievethe Ibelievethatjusttheoppositehashappened. answerisyesforavarietyofreasons. Thedebateisfundamentallyabouttherelationshipbetweenthelow-inflationobjectiveand thehigh-employmentobjective.Evenbefore THE CASE FOR AN INFLATION BritisheconomistA.W.Phillipspublished TARGET researchin1958thatgaverisetowhatquickly cametobecalledthePhillipscurve,manyecon- Ihavealreadypointedoutthataformal omistsbelievedthattherewasanegativerelationinflationgoalshouldimprovethecoherenceof shipbetweeninflationandunemployment—i.e., internalFeddeliberationsbyfocusingattention lowerinflationresultedinhigherunemployment. onhowtoachieveanagreedgoalratherthanon Somepreferredtothinkofcausationasgoingthe thegoalitself.Adoptingandachievingaformal otherwayaround—thathigherunemployment inflationobjectiveshouldreducerisksforindiresultedinlowerinflation. vidualsandbusinesseswhenmakinglong-term Theinflation-unemploymenttrade-offwas decisions. thoughttobepermanent.Societycouldhavea Becausethebenefitsofpricestabilityare permanentlyloweraverageunemploymentrate indirectanddiffuse,theyaredifficulttoquantify. byacceptingahigheraveragerateofinflation.In Oneareawherethebenefitsofpricestabilityare thelate1960s,MiltonFriedman(1968)and mostapparentisthelong-termbondmarket.Itis EdmundPhelps(1967)challengedtheideaofa notsurprisingthatthe10-yearTreasurybond permanenttrade-offbymakingthetheoretical yieldhasgenerallydrifteddownwithactualand argumentthatthePhillipscurvemustbevertical expectedinflationsincethelate1970s.Thereducinthelongruninaworldwhereeconomicagents tioninlong-termbondyieldsreflectsmarket arerational.Subsequentevidenceconfirmedthe participants’expectationsoflowerinflationand Friedman-Phelpsview,andfeweconomiststoday theirincreasedconfidenceaboutthelong-term believethatthereisanylong-runtrade-off. inflationrate.Moreover,thevolatilityofthemar- Averticallong-runPhillipscurvedoesnot ket’sexpectedrateofinflation,measuredbythe implythatonelong-runinflationrateisasgoodas spreadbetweennominalandinflation-indexed anyother.Rather,thedynamicsoftheFriedman- 10-yearTreasurybondyields,hastrendeddown Phelpstheoryimplythatinflationwouldaccelsincethelate1990s,suggestinganincreased eratecontinuouslywerepolicymakerstopursue confidenceintheFed’sresolvetokeepinflation apolicyofkeepingtheunemploymentrateperlow.Ianticipatethattheadoptionofaformal manentlybelowitsnatural,orequilibrium,rate. inflationobjectivewouldresultinsome,probably Thisequilibriumratecametobecalledthe modest,furtherreductioninthelevelandvari- NAIRU—thenonacceleratinginflationrateof abilityofnominallong-termbondyields. unemployment. 4

InflationTargeting TheFriedman-Phelpstheorydemonstrates itnotforthefactthatthecredibilityestablished whyapolicyofkeepingtheunemploymentrate overtheyearssincePaulVolckerdramatically permanentlybelowitsnaturalrateisfutile.Itdoes alteredthecourseofmonetarypolicyinOctober nottellustheinflationratethatmaximizessocial 1979.7 welfare,whichIwillcalltheoptimalinflation Ibelievethathavingaformalinflationobjecrate.Economictheorydemonstrateswhyinflation tivewillfurtherenhancetheFed’scredibilityand, iscostly,andworldwideexperiencedemonstrates consequently,itsabilitytoengageincounterthat“high”inflationand“slow”economicgrowth cyclicalmonetarypolicy.Thereasonissimple. appeartobeinexorablylinked.Everyoneacknowl- ThemoreopenandprecisetheFedisaboutits edgesthat,beyondsomerate,inflationreduces long-runinflationobjective,themoreconfident economicgrowth.Thegoalsofpricestabilityand thepublicwillbethattheFedwillmeetthat maximumsustainableeconomicgrowtharenot objective.Theobjective,andtheaccompanying substitutes,asimpliedbytheoriginalPhillips obligationtoexplainsituationsinwhichthe curve,butcomplements.Monetarypolicymakers objectiveisnotachieved,shouldincreasethe canmaketheirgreatestcontributiontoachieving Fed’scredibility. maximumsustainableeconomicgrowthbyachiev- Becauseitwillbemucheasierforthepublic ingandmaintaininglowandstableinflation. todeterminewhethertheFOMCispursuingits Thatinflationandeconomicgrowthare inflationobjectiveifthatobjectiveisknownwith complementsdoesnotimplythatpolicymakers precision,adoptingaformalobjectiveforinflashouldnotengageincountercyclicalmonetary tionalsowillenhancetheFed’saccountability. policywhencircumstanceswarrant.Forexample, HavingaformalobjectivemakestheCongress’s withinflationwellcontainedattheendofthe andthepublic’sjobeasier,therebyenhancing long1990sexpansion,theFOMCbeganreducing accountability.IftheFOMCmissesitsinflation itstargetforthefederalfundsrateinJanuary2001, objective,itwillhavetosaywhytheobjective somewhatinadvanceoftheonsetofthe2001 wasmissed.Bythesametoken,theFOMCwill recession.Thefundsratetargetwasreducedstill havetoexplainwhyitfailedtorespondtoaparfurtherin2002and2003asincomingdatarevealed ticulareventwheninflationappearedtobewellthattheeconomywasrespondingsomewhatmore containedwithintheobjective.Inessence,having slowlythanexpectedandthatactualandexpected aspecificinflationobjectivewillhelpthepublic inflationremainedwellcontained.Thefundsrate betterunderstandwhatIhaveelsewherecalled targetwaseventuallyreducedto1percentand “theFed’smonetarypolicyrule.”8 remainedthereforslightlymorethanayear. Thosewhosuggestthatadoptingaformal inflationobjectivewillcausepolicymakersto SPECIFYING THE TARGET becomeinflationnuttersand,somehow,limitthe Fed’sabilitytopursueotherpolicyobjectives Thattherearedifferencesofopinionabout shouldexamineactualexperience.Notonlydid theoptimalinflationrateisnotareasonforhaving theFed’scommitmenttopricestabilitynotprevent afuzzyobjective.Ifthereareimportantdifferences itfromengagingincountercyclicalmonetary ofopinionwithintheFOMContheappropriate policy—itfacilitatedit.6Suchanaggressivecoun- target,whichIdoubt,theCommitteeoughtto tercyclicalmonetarypolicyaspursuedstarting resolvethosedifferencesandnotpermitthemto inearly2001wouldhavebeenunthinkablewere beasourceofuncertainty. 6 Forevidenceonhowinflationinterferedwithcountercyclicalpolicyinthepast,seePoole(2002). 7 Forthoseinterestedinunderstandingtheissuesthatleduptoandsucceededthisevent,seeFederalReserveBankofSt.Louis(2005). 8 Poole(2006). 5

MONETARYPOLICYANDINFLATION Figure 1A CPI and Core CPI 3-Year Moving Averages 12 10 8 6 4 2 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 –2 CPIfor AllUrban Consumers:AllItems CPIfor All Urban Consumers:AllItems Less Food & Energy Difference Becausethetargetshouldapplytolow- reducemeasuredprices.Suchataxchangewould frequencyinflation,thetargetneedstobestated beaone-timeeffect—thepricelevelwouldchange intermsofeitherarangeorapointtargetwith whenthenewtaxlawtookeffectbuttherewould anunderstoodrangeoffluctuationaroundthe notbecontinuingpressureovertimetendingto pointtarget.Thechoiceismoreamatterofthe lowerprices.Supposetheone-shotpricelevel mosteffectivewayofcommunicatingthetarget changetookmeasuredinflationoutsidethetarget andwhatitmeansthanamatterofsubstance. range.Withaformalinflationtarget,theFOMC Aspecifictargetrange,suchas1to2percent wouldhavetheresponsibilityofexplainingwhy annualchangeinaparticularpriceindex,hasthe amonetarypolicyresponsetothistargetmiss advantageoffocusingattentiononlow-frequency wouldbeunnecessaryandperhapsharmful. inflation.Evenhere,therecouldbespecialcircum- Aformalinflationtargetneedstorefertoa stances,whichtheFedshouldexplainshould particularpriceindex.Thatthereisnopriceindex theyoccur,thatwouldjustifydeparturefromthe thatadequatelyreflectstheeconomy’struerate target.Thewaytherangeisexpressedinteracts ofinflationisyetanotherreasongivenfornot withtheperiodoverwhichinflationisaveraged. adoptingaspecificinflationobjective.Myown Anarrowerrangewouldbeappropriatefora judgmentisthatthePCEpriceindexmeasures targetexpressedasathree-yearaveragethanfor consumerpricesreasonablywellandhassome ayear-over-yeartarget. advantages,whichcanbeexplained,overtheCPI. Tounderstandwhatsuchatargetmeans, Moreover,theFOMCcouldreasonablymaintain supposestatesweretoabolishsalestaxesand arateofincreaseinthisindexinarangeof,say raiseincometaxestooffsettherevenueloss.The 1to2percent,onatwo-yearmovingaveragebasis effectofthischangeintaxstructurewouldbeto undermostcircumstances. 6

InflationTargeting Figure 1B PCE and Core PCE 3-Year Moving Averages 10 8 6 4 2 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 –2 PCE:Chain-type Price Index PCELess Food and Energy: Chain-type Price Index Difference Overtime,refinementsinthepriceindexor categoriesofgoods.Rather,experienceindicates introductionofbetterindicesmayleadtosubsti- thatfoodandenergypricesaresubjecttolarge tutionofanotherindexforthePCEindexorjustify short-rundisturbancesthatarebeyondtheability achangeinthetargetrange.TheFOMCwould ofmonetarypolicytocontrolwithoutpolicy thenhavetoexplainwhyitwasadjustingthe responseshavingadverseconsequencesforgeneral objectiveorindexusedtoevaluatetheobjective. economicstability.Ifweexaminetotalandcore Theformaltargetprovidesavaluablevehiclefor priceinflationoverthreeyears,say,mostexperiexplaininganimportantissueintheconductof enceisthattheaveragesarequiteclose.Thatis, monetarypolicy.Experiencewithinflationtarget- foodandenergypricesdisplaysubstantialshortinginindustrialeconomiessuggeststhatissues runvariabilitythatyieldslargechangesinthe ofthissorthavenotbeenimportant.Themarkets short-runrateofinflationinoverallpriceindices arealreadywellinformedaboutsuchissues— withoutaffectinglonger-runinflation.(Seethe andarebecomingincreasinglyso.Conducting chartsinFigure1,whichtracktheCPIandPCE thisconversationwiththemarketswillimprove indicesfrom1960through2005.) theclarityofmonetarypolicyandthereforeits effectiveness. OverthepastdecadeorsotheFedhasgravi- HOW MUCH DIFFERENCE tatedtothepositionofplacingprimaryemphasis WOULD A FORMAL INFLATION onthecorerateofinflation,asmeasuredbythe TARGET MAKE? PCEpriceindexexcludingfoodandenergy.The reasonisnotthatfoodandenergyareunimpor- Thereisalargeandgrowingliteraturecomtant—theseareobviouslytwoveryimportant paringtheperformanceofinflation-targeting 7

MONETARYPOLICYANDINFLATION countrieswiththeirnon-inflation-targeting wouldneverhavehappened?Surelynot.Wouldit counterparts,especiallytheUnitedStates.This havehelped?Ibelievethattheanswerissurelyyes. literaturefindsfewstatisticallysignificantdifferencesbetweencountriesthathaveestablished inflationtargetsandthosethathavenot.This CONCLUDING REMARKS findinghasledsomeanalyststoargue,“ifitisn’t Inflationtargetingisanapproachtomonetary broke,don’tfixit.”Thereareanumberofreasons policyadoptedbymanycountries,inmostcases whysuchfindingsarenottoosurprising:The inthecontextofasocietalefforttoaddressundebenefitsfrompricestabilityarediffuseanddiffisirablyhighinflation.TheUnitedStates,fortuculttomeasure;theindustrializedeconomies nately,isnotdealingwithaninflationproblem arehighlyinterconnected,sothatsomeofthe atthistime.Thecaseforadoptinganinflation benefitstocountriesthathaveinflationtargets targetisthatitshouldhelptoavoidinflationin spillovertothosethatdonot;thegrowthrate thefutureandshouldincreasetheeffectiveness effectissmall,soitwilltakealongtimebefore ofmonetarypolicyinalow-inflationenvironment. onecandistinguishastatisticallysignificant Theincreaseinpolicyeffectivenessshould growth-rateeffect.Finally,manyofthecountries arisefromtwoconsequencesofaformalsystem thatadoptedaninflationtargethadahistoryof ofinflationtargeting.Thefirstconsequenceisthat inflation.Adoptingatargetwasamanifestation themarketwilllikelyholdinflationexpectations ofasocietalcommitmenttobringdownandkeep morefirmly.Thesecond,andprobablymore downtherateofinflation. important,consequenceisthattheinflation- GiventhattheUnitedStatespursuedasuc- targetingframeworkprovidesavehicle,orstruccessfulanti-inflationpolicyafter1979withouta ture,withinwhichtheFOMCcanbetterexplain formaltarget,andestablishedahighdegreeof itsmonetarypolicyactionsandthepolicyrisks monetarycredibility,thereisnoreasontoexpect itmustface.Inflationtargetingshouldincrease toobservemeasurableeffectsfromadoptinga accountabilitynotsomuchbykeepingscoreof targetnow.Nevertheless,Icannothelpreflecting targethitsandmissesbutratherbyencouraging onothercasesinwhichlowinflationprevailed amuchdeeperunderstandingofhowmonetary butdidnotlast.ConsiderU.S.policyerrorsof policydecisionsaremade.Thatunderstanding thetypethatoccurredinthemid-to-late1920s dependsoncontinuingFOMCcommunications andinJapaninthelate1980s.Inbothofthese withthemarketsandthepublicandFOMCwillinstances,policymakersfailedtorespondtodefla- ingnesstolistenaswellastalk. tion.Ibelievethataformalinflationtargetwould havefocusedattentiononthepolicymistake leadingtodeflationandwouldhaveincreased REFERENCES publicpressureonthecentralbankstorespond FederalReserveBankofSt.Louis.“Reflectionson moreforcefully. MonetaryPolicy25YearsAfterOctober1978: Similarly,theFedfailedtotightenpolicy ProceedingsofaSpecialConference.”Federal appropriatelyinthelate1960sasinflationbegan ReserveBankofSt.LouisReview,March/April itsascent.Intheearly1960s,astoday,theFed 2005,87(2,Part2). enjoyedahighdegreeofmarketconfidenceand inflationexpectationswerelow.Atthattime, Friedman,Milton“TheRoleofMonetaryPolicy.” onlyasmallminorityofeconomiststhoughtthat AmericanEconomicReview,March1968,58(1), monetarypolicywas“broken”inanyimportant pp.1-17. way,andthusthecasefor“fixingit”wasminimal. Wouldaformalinflationtargetin1960havebeen Friedman,Milton.“TheOptimumQuantityof anironcladguaranteethattheGreatInflation Money,”inTheOptimumQuantityofMoneyand 8

InflationTargeting OtherEssays.Chicago:AldinePublishing,1969, pp.1-50. Greenspan,Alan.Chairman’sRemarks.Federal ReserveBankofSt.LouisReview,July/August 2002,84(4),pp.5-6. King,Mervyn.“ChangesinU.K.MonetaryPolicy: RulesandDiscretioninPractice.”Journalof MonetaryEconomics,June1997,39(1),pp.81-97. Phelps,EdmundS.“PhillipsCurves,Expectationsof InflationandOptimalEmploymentoverTime.” Economica,August1967,34(3),pp.245-81. Phillips,A.W.“TheRelationbetweenUnemployment andtheRateofChangeofMoneyWageRatesin theUnitedKingdom,1861-1957.”Economica, November1958,25(100),pp.283-99. Poole,William.“Synching,NotSinking,theMarkets.” SpeechpreparedforthemeetingofthePhiladelphia CouncilforBusinessEconomics,FederalReserve BankofPhiladelphia,Philadelphia,August6,1999; www.stlouisfed.org/news/speeches/ 1999/08_06_99.html. Poole,William.“Inflation,RecessionandFedPolicy.” SpeechpreparedfortheMidwestEconomic EducationConference,St.Louis,April11,2002; www.stlouisfed.org/news/speeches/2002/ 04_11_02.html. Poole,William.“TheFed’sMonetaryPolicyRule.” FederalReserveBankofSt.LouisReview,January/ February2006,88(1),pp.1-11;originallypresented asaspeechattheCatoInstitute,Washington,DC, October14,2005. Santoni,G.J.“TheEmploymentActof1946:Some HistoryNotes.”FederalReserveBankofSt.Louis Review,November1986,68(9),pp.5-16. 9