Speech

Communicating the Fed’s Policy Stance HMTreasury/GESConferenceIsThereaNewConsensusinMacroeconomics? London,England November30,2005 My short answer to the question viewedastotallylackinginrationalelements.1 posed in the conference title is Thatis,marketsdoreflecteffortsofprivateagents “yes.” The fundamental issues tolookahead,howeverimperfectlytheymaybe that created an enormous gulf abletodoso.2And“lookingahead”certainly between macroeconomists in the 1960s have includesformingexpectationsastowhatpolicy- been resolved. Of course, there are still things makerswilldo. todiscuss,becauseagreementonthemostimpor- Myplanistodiscusssomeoftheevolution tantfundamentalsdoesnoteliminatecontroversy thathasledtopolicyconcernovercentralbank about many important details. communication.Myperspectiveisprimarilyfrom IntheU.Scontext,themostimportantsingle myownFederalReserveexperience.Iknowthat issuewasthatin1965,say,economistsconducted thetopichasbeenextensivelydebatedwithinand modelingandpolicyexercisesinacontrol- withoutothercentralbanks,butIdonothave theoreticframework.Achangedviewofexpec- enoughsystematicknowledgeofthesedebates tationsledtoappreciationoftheimportanceof tocommentonissuesoutsidetheU.S.context. I’lldiscusstwoaspectsofcentralbankcommuni- thedistinctionbetweenrealandnominalinterest cation.Oneaspectis“body-language”communi- ratesandtheviewthatinthelongrunthePhillips cationthroughincreasedregularityofpolicy curvewasvertical.Somewhatlaterbutcertainly actionsandthesecondiswrittenandoralcom- by1985,say,almosteveryonebelievedthatexpec- municationthroughpolicystatements,speeches tationsofprivateagentsaboutwhatpolicymakers andtestimony. woulddohadtobeincorporatedinmodelsand Beforeproceeding,Iwanttoemphasizethat policyanalyses. theviewsIexpressherearemineanddonot In1965,expectationswerealmostuniformly necessarilyreflectofficialpositionsoftheFederal modeledinabackward-lookingway.Asthe ReserveSystem.Ithankmycolleaguesatthe rationalexpectationsanalysistookhold,the FederalReserveBankofSt.Louisfortheircom- argumentconcernedtheextentofrationalityin ments,butIretainfullresponsibilityforerrors. formationofexpectations.Wereexpectations rationalinthesenseofMuth(1961)orwerethey basedonbackward-lookingand/orrules-of-thumb CENTRAL BANK CREDIBILITY calculations?Iwouldnotclaimthatthereisa consensustodayonhowtomodelexpectations, Thereisalargeliteratureoncentralbank butwouldclaimthatallseriousmacroeconomists credibility.Theliteraturearosequitenaturallyin believethatexpectationscannotbeadequately thecontextofthetaskofbringinginflationdown 1 AnearlyimportantpaperinthisliteraturewasSargentandWallace(1975). 2 Ifinformationislimitedtothepropertiesofatimeseries,thentherationalexpectationistheappropriateextrapolationofthepasthistoryof thevariable.Thedistinctionisbetweenarationalexpectation,inthesenseofthebestguessgivenallavailableinformation,andacustomary, orhabit-driven,extrapolationofhistory. 1 MONETARYPOLICYANDINFLATION fromtheexcessiveratesprevailinginmostcoun- wouldreduceinflationwhichinturnwould triesinthe1970s.Centralbankshadalwaysbeen graduallyreduceinflationexpectations,setting concernedaboutcredibility,buttheimportance thebaseforrenewedeconomicgrowthatalow ofthatconcernwaselevatedbytherationalexpec- rateofinflation. tationsrevolutioninmacroeconomicsinthe1970s Credibilityandforward-lookingexpectations andtheunacceptablyhighrateofinflation.3 promisedtoreducethepain.Nevertheless,the Analysisofcredibilityissuesinvolvedthe UnitedStatesandtheUnitedKingdombothsuf- natureofappointmentstogoverningboardsof feredsignificantrecessionsintheearly1980s.I centralbanksandtheincentivescentralbanks rememberacommentImadein1981tooneofmy face.Asforincentives,therehaslongbeenrecog- morehard-linerationalexpectationscolleagues: nitionoftheimportanceofpoliticalindepend- GiventheFederalReserve’sbehavioroverthe ence,lestmonetarypolicybeusedtobolsterthe previous15years,itwasnotrationaltoassigna electoralprospectsofthepartyinpower.The highprobabilitytotheFed’spromises,evenunder FederalReserveActin1913wasdesignedtocreate thenewChairman,PaulVolcker,tobringinfla- considerablepoliticalindependence.Theargu- tiondown.Intheevent,itwastheFed’swilling- mentontheotherside,ofcourse,wasthatina ness,supportedbyPresidentReagandespitea democracypolicydecisionsshouldbemadeby severerecession,tosustainapolicytoreduce electedofficialsorofficialsdirectlyaccountable inflationthatcreatedcredibility.TheFed’spolicy toelectedofficials;thatviewdeterminedthe morethananythingitsaidbuiltcredibility. institutionaldesigninmanycountries,including Centralbankcredibilityisanaspectofthe theUnitedKingdom. broaderissueoftrust.Credibilityandtrust,once Theissuewaseventuallydecidedinfavorof lost,canbeextremelyexpensivetoregain.I politicalindependencelargely,Ithink,onthe believethatmostpolicymakersrecognizethis basisofexperience.Therelativelygreatersuccess fact,andtherecognitionhasmuchtodowith ofmoreindependentcentralbanks—Deutsche effortstoenhancetransparencytobuildtrust. Bundesbank,SwissNationalBankandFederal ReserveSystem—convincedmanyobserversthat substantialpoliticalindependencewasthebetter POLICY RULES institutionaldesignfordemocraticcountries. Inanyevent,thegrowingconvictionof Thereisalargeliteratureonrulesversusdis- macroeconomiststhattherationalexpectations cretioninmonetarypolicy,datingfromthesem- modelneededtobetakenseriouslyprovideda inalpaperbyHenrySimons(1936).Therational powerfulanalyticalbasefordiscussingcentral expectationsrevolutionreinforcedthecasefor bankcredibility.Also,economistsrecognized rules.Inthemodelsspawnedbythisrevolution, thatthecostofbringinginflationdownmight monetarypolicywasgenerallyspecifiedbya dependimportantlyontheextentofcredibility. moneygrowthrule.Thisapproachworkedwell Inmodelsinwhichinflationexpectationswere inthemodels,butleftavoidformonetarypolicy- merelyextrapolationsfrompastinflation,reduc- makersbecausecentralbanksdidnotimplement inginflationwouldrequireapredictableperiod policythroughmoneygrowthtargets.Thereis ofpain.Tobringactualinflationdown,thecentral controversyovertheextentofmoneygrowthtar- bankhadtocreatearecession.Therecession getingbytheBundesbank,butfortheUnited 3 CredibilitywasacriticallyimportantissuefortheFederalReservein1979asitdecidedtochangepolicydramaticallytobringdowninflation. SeeFederalReserveBankofSt.Louis(2005). 4 TheFOMCbegansettingatargetforborrowedreservesin1982.Myviewisthatthistargetwasafundsratetargetatoneremove,because bankborrowingatthediscountwindowdependsonmoneymarketinterestrates.Also,pursuitoftheborrowingtargetwastemperedbythe behaviorofthefundsrate.Putanotherway,openmarketoperationsweredesignedinparttopreventthefederalfundsratefromdeparting toowidelyfromexpectation.Forallpracticalpurposes,Ibelieve,theFOMCwasoperatingonafederalfundstargetafterAugust1982. 2 CommunicatingtheFed’sPolicyStance Statesthepracticewasclear.Exceptfortheperiod genuinelynewinformationtowhichtheFed fromOctober1979toSeptember1982,theFederal shouldanddoesrespond.6 Reservedidnotpursueamonetaryaggregates policybutinsteadimplementedpolicythrough adjustmentsinitstargetforthefederalfundsrate.4 INFLATION TARGETING Ofcourse,observedmoneygrowthdidaffectthe TheTaylorruleassumesaspecifiedinflation fedfundsratetargetsfromtimetotime,butso target.Ambiguityoverthetargethasbeenreduced alsodidmanyothervariables. byadoptionofaformalinflationtargetbymany Foratime,therefore,macroeconomistswere centralbanks.Asubstantialliteratureoninflation intheuneasypositionofconstructingmodels targetingnowexists.7Theliteraturecoversnot closedbyamoneygrowthrulewhentheyknew onlytheissueofwhattargettosetandtheadvan- thattheFeddidnotpursuepolicythisway.In tagesofhavingatargetbutalsopolicyunderan 1993JohnTaylorshowedthatitispossibleto inflationtarget.Theconsensusisthattheformal modelFedbehaviorbyusingasimpleequation inflationtargethelpstoinformthepublicabout thatmakesthetargetfedfundsrateafunctionof thecentralbank’sobjectivebutthatpolicyshould therealinterestrate,thegapbetweentheactual notbeaslavetothetarget.Particularlyinthe anddesiredinflationrate,andthegapbetween UnitedStateswheretheFederalReserveAct actualandpotentialrealGDP. containsadualmandatespecifyinggoalsof TheTaylorrulefitexperienceprettywell, maximumemploymentaswellaspricestability, andcontinuedtodosoafter1993.Someversion policymakerscannotpursuethepricestability oftherulebecamethestandardwayofclosing goaltotheneglectofeverythingelse. macroeconomicmodels.Nevertheless,because IntheUnitedStates,andIbelieveelsewhere theFederalReservehasnotfollowedtheTaylor aswell,opinionisprettysettledthatemployment ruleclosely,oranyvariantofitorotherrule, thereremainedandremainstodayagapbetween andpricestabilitygoalsarenotcompetitivebut whattheFederalReserveactuallydoesandthe fundamentallycomplementary.Successinmain- notionofapolicyruleembeddedintheabstract taininglowandstableinflationcontributesto models. economicstability,maximizeseconomicgrowth FederalReservepolicyismuchmorerule-like andcreatesconditionsthatpermitthelabor thancommonlyappreciated.Ihavediscussed markettoclearatthehighestpossiblelevelof thisissueatsomelengthintworecentspeeches. employment.Ofcourse,thatlevelisdetermined Inthefirst,IarguedinsomedetailthatFedpolicy importantlybynonmonetaryconditionsthatfall hasbecomehighlypredictable,asmeasuredby outsidethecentralbank’sdirectareaofresponsi- theaccuracyofpredictionsinthefederalfunds bility.Alsoimportant,however,isthatlowinfla- futuresmarket.5Inthesecond,Iarguedthatpre- tionandhighcredibilitypermitthecentralbank dictabilityisevidenceofrule-likepolicy.Idetailed torespondconstructivelytorealandfinancial someofthecharacteristicsofpolicybehavior disturbanceswithoutraisinginflationfears.With thatareindeedhighlypredictable.Oneexample highcredibility,thecentralbankcanreducethe istheFed’sanalysisofthesignificanceofstatis- varianceoftherealeconomy. ticaldatabasedondetailedinformationthathelps AlthoughtheFederalReservehasnot todistinguishtransitorydisturbances,towhich announcedanumericalinflationtarget,Ben theFedoughtnotanddoesnotrespond,from Bernanke,thenomineetobecomeFedchairman 5 Poole(2005a). 6 Poole(2005b). 7 Twoconvenientsourceare:BernankeandWoodford(2005)andFederalReserveBankofSt.Louis(2004). 3 MONETARYPOLICYANDINFLATION inFebruary,isonrecordfavoringsuchatarget. monetarypolicy,enablingmarketparticipantsto AlthoughIamalsoonrecordfavoringaformal makemoreefficientdecisions.Theissueisnot numericaltarget,Ibelievetheissuehasnotbeen withtheprincipleofbettercommunication,but alargeoneintheUnitedStates.Manyobservers howtobemoreeffective.Ihavediscussedsome believethattheFedhasbeenpursuinganinfla- ofthedifficultiesinaspeechIentitled“Fed tiontargetrangeof1to2percentannualrateof Transparency:How,NotWhether”(Poole,2003). increaseofthecorePCEpriceindex.IftheFed Aseverycentralbankerknowsandhasmost doesadoptaformaltargetinthefuture,Idoubt likelyexperienced,communicationisdifficult verymuchthatstatisticaltestsforaregimebreak becauseitissoeasytobemisunderstood.Mis- wouldbeabletofindoneinaneconomicseries communicationaddsuncertaintyandcreates suchasaninflationindex,employmentorreal marketvolatility.Inaformal,rationalexpectations GDP. modelsuchuncertaintycanbemodeledbyadding arandomtermtothemarket’sunderstandingof thepolicyrule.Iknowofnomodelinwhich COMMUNICATION THROUGH addingstochasticdisturbancestothepolicyrule improvesoutcomesforinflation,employment POLICY STATEMENTS, SPEECHES, andgrowth. ETC.8 Increasedattentiontocommunicationhasa Arationalexpectationsequilibriumrequires benefitthatisfrequentlyoverlooked—animprove- thatthemarkethaveinformationabouttheFed’s mentintheclarityofinternaldeliberations.Ina policyrule.Themoreaccurateisthatinformation, committeecontext,explicitunderstandingof themoreefficientwillbeeconomicoutcomes. policygoalsandagreementonpolicydirection Themarketlearnsabouttheruleaboveallfrom mustprecedepubliccommunication.Weneed whattheFeddoes.Actualmonetarypolicyhas toknowwhatwewanttosaybeforewetrytosay beenhighlysuccessful,andthatiswhathas it.IhadavividlessoninthisregardwhenIfirst builtcredibility.Regularitiesinpursuingpolicy cametomycurrentposition.In1998,theFOMC havemadepolicymorepredictable,inthesense wasagreeingonapolicy“tilt”or“bias”andIdis- thatconditionalonnewinformationthemarket coveredthatdifferentmembersoftheCommittee hasagoodideaoftheFed’sresponse,ifany,to haddifferentinterpretationsofwhatthebias thenewinformation. meant. Althoughpredictablepolicy—thebodylan- Themostimportantcommunicationsissue guage—isthemostimportantfeatureofthecur- facingtheFOMCcurrentlyiswhetherandhowto rentsituation,improvedpolicycommunication continuetoprovideforwardguidanceonpolicy hasalsoplayedasignificantrole.Perhapsthe decisions.Startinginmid-2003,thepolicystate- mostimportantsteptheFOMChastakento mentattheconclusionoftheFOMCmeeting improvepolicycommunicationswastherelease statedthat“...policyaccommodationcanbemain- ofthepolicydecisionimmediatelyfollowing tainedforaconsiderableperiod.”Later,theCom- eachFOMCmeeting,startinginFebruary1994. mitteesaidthatitcouldbe“patientinremoving Othersteps,suchasmoretimelyreleaseofmin- itspolicyaccommodation.”Stilllater,theCom- utesofFOMCmeetings,havebeenhelpful.9 mitteesaidthatitwouldremoveaccommodation Ibelievethatthereisaconsensusthatbetter ata“measuredpace.” communicationfurthersthegoalofinforming Evenwhenthefederalfundsratewasat1 themarketsmorecompletelyaboutthecourseof percent,unpredictableeventscouldhaveoccurred 8 Woodford(2005)providesanexcellentrecentdiscussionofcommunicationsissues. 9 Forafulllistofsuchsteps,seePoole(2005a). 4 CommunicatingtheFed’sPolicyStance thatwouldhaveledtheCommitteetodepart decadeorso.Ibelievethatthereisaconsensus fromitsforwardguidance.Thesettingofpolicy thatimprovedpredictabilityhasimprovedthe mustbeconditionaloninformationathand,and effectivenessofmonetarypolicy.Thisconsensus wheninformationchangessufficientlythepolicy inturnisfundamentallyaconsequenceofthe settingmustalsochange.Historically,theFederal consensusintheeconomicsprofessionthatthe Reservehasnotprovidedforwardguidancefor rightwaytothinkaboutthemacroeconomyis fearthatitwouldlockitselfintoapolicystance inthecontextofarationalexpectationsmacro thatmight,undernewinformation,nolongerbe model.Expectationsmaynotbefullyrational,but appropriate.Inprinciple,thereisnoreasonwhy thatisstilltherightstartingpointforanalyzing theFedcannotexplainthenatureofthecondi- theeconomyandmonetarypolicy. tionalityandconveytheviewthatpolicyguidance dependsoninformationavailableatthetime guidanceisoffered. REFERENCES Puttingasidetheunusualcircumstancesthat Bernanke,BenS.andWoodford,Michael,eds. tookthefederalfundsrateto1percent,forme “TheInflation-TargetingDebate,”NBERStudiesin theissueiswhetherundernormalandroutine BusinessCycles,32.ChicagoandLondon: circumstancesforwardguidancewillconvey UniversityofChicagoPress,2005. informationorwhetheritwillcreateadditional uncertainty.Ifconditionalityofpolicyisunder- FederalReserveBankofSt.Louis.Inflation stood,theneventsthatleadtheFOMCtodepart Targeting:ProspectsandProblems.Federal frompreviouslystatedforwardguidanceshould ReserveBankofSt.LouisReview,July/August notcausedifficulty.Themarketwillunderstand 2004,86. thatguidanceisnotapromisethatmustbekept toretaincredibilitybutinsteadawayofsumma- FederalReserveBankofSt.Louis.Reflectionson rizingtheCommittee’sviewoftheprobabledirec- MonetaryPolicy25YearsAfterOctober1979. tionofpolicy.Then,whenunexpectedevents FederalReserveBankofSt.LouisReview,March/ movepolicyadifferentway,marketswillcome April2005,87(Part2). tothesameconclusionaboutthepolicysignifi- canceofanunexpectedeventasdoestheFOMC. Muth,JohnF.“RationalExpectationsandtheTheory ofPriceMovements.”Econometrica,July1961,29, pp.315-35. SUMMARY Poole,William.“FedTransparency:How,Not Inthecontextoftoday’sunderstandingof Whether.”FederalReserveBankofSt.Louis howtheeconomyworks—anunderstanding Review,November/December2003,85,pp.1-8. basedonrationalexpectationsmodels—policy communicationisanessentialaspectofmonetary Poole,William.“HowPredictableIsFedPolicy?” policy.Privateagentsmakemoreefficientdeci- FederalReserveBankofSt.LouisReview, sionswhentheyunderstandmonetarypolicy.Of November/December2005a,87,pp.659-68 course,thesameargumentholdsforotherpolicies, suchasfiscalandregulatoryones.Clarityrequires Poole,William.“TheFed’sMonetaryPolicyRule.” thatpolicyhaverule-likecharacteristics.Indeed, CatoInstitute,Washington,DC,October14,2005b. aresearchagendaimprovingthespecificityof themonetarypolicyruleishighlydesirable. Sargent,ThomasJ.andWallace.Neil.“‘Rational’ FederalReservepolicypracticeandpolicy Expectations,theOptimalMonetaryInstrument, discussionhasdonemuchtoimprovethepre- andtheOptimalMoneySupplyRule.”TheJournal dictabilityofFedpolicydecisionsoverthepast ofPoliticalEconomy,April1975,83,pp.241-54. 5 MONETARYPOLICYANDINFLATION Simons,HenryC.“RulesVersusAuthoritiesin MonetaryPolicy.”JournalofPoliticalEconomy, February1936,44,pp.1-30. Taylor,JohnB.“DiscretionversusPolicyRulesin Practice.”Carnegie-RochesterConferenceSerieson PublicPolicy,1993,39,pp.195-214. Woodford,Michael.“CentralBankCommunication andPolicyEffectiveness.”PresentedattheFederal ReserveBankofKansasCitysymposiumThe GreenspanEra:LessonsfortheFuture,Jackson Hole,WY,August25-27,2005. 6

William Poole (2005, November 29). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_20051130_poole

@misc{wtfs_speech_20051130_poole,
  author = {William Poole},
  title = {Speech},
  year = {2005},
  month = {Nov},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_20051130_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}