Speech

How Should the Fed Communicate? “TheFutureoftheFederalReserve” CenterforEconomicPolicyStudies(CEPS) PrincetonUniversity Princeton,NewJersey April2,2005 Central bank communication is a topic butthatconceptdoesnottakeusveryfar.Forthe of increasing interest. There is a bur- FederalReserve,somemightputtheissuethis geoning literature in both the United way:Atthetimeofeachpolicydecision,theFed States and Europe, but I’ll confine needstoexplainwhatitdidandwhy.If“thewhat” my comments to the U.S. context.1 referstothetargetfederalfundsrate,thenclear Perhapsironically,onereasonfortheheight- disclosureofthepolicyactionhasbeeninplace enedinterestinFederalReservecommunication since1994.If“thewhat”referstostatementsor policyistheFed’ssuccessinmakingitspolicy hintsaboutfuturepolicyactions,thenitseems decisionsonthetargetfederalfundsratehighly thattheFeddoesnothaveasettledpolicyonsuch predictable.Ashasbeendocumentedinanumber disclosure.Moreover,disclosingorexplaining ofpapers,thefederalfundsfuturesmarketquote “thewhy”ofpolicyactionsisobviouslycomplex. thedaybeforeanFOMCmeetinghas,withonly I’llconfinemyanalysistotheeconomicsof afewexceptionsoverthepastdecade,predicted Fedcommunicationpolicy.Disclosureissuesin accuratelytheCommittee’sdecisiononthetarget thecontextofpoliticalaccountabilityoverlap withtheeconomicsissues,butarenotexactly federalfundsrate.Whatcapturesthemarket’s thesame.Theeconomicsissueisthis:Howcan attentionnowistheCommittee’sstatementand Fedcommunicationsmakemonetarypolicymore notitsactiononthetargetfundsrate. effective?Thatquestionrequiresaviewabout BecauseI’vedevelopedsomeprettydefinite howtheeconomyworks. viewsonthesubjectofFederalReservecommu- Mystartingpointistheconceptofafull nicationpolicy,itisespeciallyimportantthatI rationalexpectationsmacroeconomicequilibrium. issuetheusualdisclaimerbeforeproceedingfur- Thebasicideaisthattheprivatesectormakes ther.TheviewsIexpressherearemineanddonot decisionsinthecontextofitsexpectationsabout necessarilyreflectofficialpositionsoftheFederal howthegovernmentwillsetpoliciesofallkinds— ReserveSystem.Ithankmycolleaguesatthe fiscalpoliciesdeterminingtaxesandspending, FederalReserveBankofSt.Louisfortheircom- regulatorypoliciesandmonetarypolicies.I’ll ments,butretainfullresponsibilityforerrors. confinemydiscussiontomonetarypolicies. Thebaselineassumptionisthattheprivate economyandthecentralbankhavethesame A FRAMEWORK information.I’vediscussedelsewherewhatmight Theword“transparency”isoftenusedinthe happenwheninformationisasymmetric,but contextofanagency’sorfirm’scommunications, willnottakeupthattopictoday. 1 SeeespeciallyGenevaReportsontheWorldEconomy3,“HowDoCentralBanksTalk?”I’vedevelopedmyownthinkinginaseriesof speechesandpapers;seePooleandRasche(2000)andPoole(2003). 1 MONETARYPOLICYANDINFLATION Acriticalfeatureoftherationalexpectations federalfundsrate.Theprivatesectormusthave frameworkisthatthecentralbankneedstobe thesameview,forotherwiseitwillmakeforecasts clearaboutitsobjectives.Moreover,thefunda- oftheeffectsofpolicyactionsthatdifferfrom mentalobjectivesofmonetarypolicyshouldbe thecentralbank’sforecasts.Astable,efficient stableovertime,toprovideaconsistentbasisfor equilibriumisunlikelytoprevailwhenthepri- efficientplanningintheprivatesector.Forexam- vatesectorandthecentralbankhavesubstantially ple,aninflationobjectiverandomlychosentobe differentviewsastohowtheeconomyworks.The 5percentoneyearand1percentanotheryear strategyofmonetarypolicydependscritically willnotyieldagoodequilibriumbecausethe onthecentralbank’sviewastohowtheecon- privatesectorwillmakemistakesinplanningits omyworks;thus,thecentralbankneedstocon- activities,andwillexpendresourcesunnecessar- veyitsviewsinthisregardtothemaximum ilyintryingtoinsureitselfagainsttheuncertain possibleextent.Ofcourse,reducinguncertainty inflationoutcome.Ifandwhencircumstances isnottheonlyimportantfeatureofmonetary ariserequiringfundamentalpolicyobjectivesbe policydesign—thesystematicpartofpolicyalso changed,itisimperativethatthenewobjectives needstobewelldesignedtoachievepolicy becommunicatedtothemarket.Althoughthere objectivesinthemostefficientpossibleway. hasbeenanongoingdiscussionintheFedabout Givenalltheuncertaintiesandcontroversies theadvantagesanddisadvantagesofaformal, inmacroeconomics,itmayseemthatanyattempt numericalinflationobjective,Ithinktheambi- toconstructacommunicationframeworkaround guitywithrespecttotheFed’sinflationand theassumptionofacommonunderstandingof employmentobjectivesisnotlargeandisnotthe howtheeconomyworksmustfail.Infact,the mainproblemtheFedfaceswithitscommunica- situationismuchmorehopeful.Ibelievethereis tionpolicies. acommonunderstandingaboutmanyfeaturesof Intheworldofarationalexpectationsmacro- macroeconomicbehaviorandthatwecansweep economicequilibrium,claritywithrespecttoboth areasofuncertaintyintoananalyticaluncertainty goalsandstrategyinpursuinggoalsisavirtue. basket.Thereisagreementaboutmostmacro- Thisisanimportantpointbecausetheolder economicrelationshipsataqualitativeanddirec- monetarypolicyliterature,fromatraditionthat tionallevel,atleastinthecontextofmonetary predatestherationalexpectationsrevolutionin policyconductedforstabilizationpurposes. macroeconomics,arguedattimesthatmonetary Appreciatingtheuncertaintythatmacroecono- policyeffectivenessdependsontakingmarkets mistsshareisanessentialelementofcraftinga bysurpriseandcreatinguncertainty.Some goodcommunicationstrategy.Thatis,thecentral observersstillmakethisargumenttoday.Ireject bankandtheprivatesectorcanhaveacommon thisview;Iknowofnoconvincingmodelswhere understandingofthenatureoftheuncertainty policy-induceduncertaintyyieldssuperioreco- basketandunderstandthesignificanceofthat nomicoutcomes. factfortheconductofmonetarypolicy. Onerequirement,then,foranefficient Thethirdessentialelementoftheframework rationalexpectationsmacroeconomicequilibrium isthateconomicchangeandeconomicoutcomes isthatthecentralbankhaveclearpolicygoals. thatdifferfromforecastsareaconsequenceof Anotherrequirementisthattheprivatesector thearrivalofnewinformationthatisinherently andthecentralbankhaveacommon,andcorrect, unforecastable,atleastatthecurrentlevelof understandingoftheeconomy’sstructure—of economicscience.Asnewinformationarrives— howtheeconomyworks.Thecentralbankmust oftencalled“shocks”toemphasizetheunfore- haveaviewastohowtheeconomyworksto castableorsurprisenaturethatisinherentinnew knowhowtoforecasttheeffectsofpolicyactions, information—theprivatesectorandthecentral whichI’llmeasurebyadjustmentsinthetarget bankhaveacommonunderstandingoftheimpli- 2 HowShouldtheFedCommunicate? cationsofthenewinformation,especiallythe “measuredpace”phrasethatindicatesthatthe implicationsforadjustmentsinthemonetary FOMCanticipatesfurtherincreasesinthetarget policyinstrument.Thereisanextensiveliterature rateatfuturemeetings. showinghowinterestratesrespondtosurprises Thesituationprevailingforaboutayearnow, intheroutineflowofeconomicdataonemploy- inwhichithasbeenreasonabletoexpectfuture ment,inflation,housingstarts,industrialproduc- increasesinthetargetfederalfundsrate,ishighly tionandsoforth. unusual.Inmyexperience,mostofthetimeitis Theresponsetonewinformationinthefed- notespeciallyclearwhatthefuturepaceofpolicy eralfundsfuturesmarketisadirectbetonhow adjustmentswillbe.Thisfactflowsfromthe theFederalReservewillrespondtothenewinfor- observationIofferedearlier,thatchangeinthe mationinsettingthetargetfederalfundsrate. economyandpolicyadjustmentsinresponseto Giventheaccuracyofthefedfundsfuturesmarket changeflowfromthearrivalofnewinformation. inpredictingFeddecisionsonthetargetfedfunds Ordinarily,policyadjustmentscannotbepre- rate,thisliteraturesupportstheviewthatthe dictedwithmuchaccuracybecauseshocksthat rationalexpectationsframeworkisusefulfor createtheneedforpolicyadjustmentsarenot understandingactualmarketbehaviorasitrelates forecastable. tomonetarypolicydecisions. Thispointisextremelyimportant.Isome- timeshearpleasfrommarketparticipantsthat theFedshouldmakeclearinadvancewhatitis APPROACHING NIRVANA goingtodo,toreduceuncertaintyandmakeplan- ningeasier.Alittlethoughtshouldleadmarket Theabstractionofthefullrationalexpecta- participantstostopthinkingthisway.Suppose tionsmacroeconomicequilibriumprovidesa theFedhadadvertisedanexpectationofan frameworkforacommunicationagenda.How unchangedtargetfederalfundsrateinearly canFedcommunicationshelpmovetheeconomy September2001.ShouldtheFednothave closertothenirvanaofthefullandefficient respondedtothe9/11shock?Obviously,theFed rationalexpectationsequilibrium? shouldrespondtosuchashock;moregenerally, Toprovideaveryspecificfocusformy theFedneedsalwaystobeopentorespondingto remarks,I’llconcentrateontheshortstatement newinformationwhenitisofsuchamagnitude releasedattheconclusionofeveryFOMCmeet- ing.Thatstatementisthefirststepofanexten- tocallforapolicyresponse,orwhensmaller siveprocessofexplainingpolicythroughminutes individualpiecesofinformationaccumulateto ofFOMCmeetings,speechesandtestimony. suchapoint. Animportantfeatureofthepolicystatement Letmeformalizethispoint.Considerboth isanexplanationofthereasonsfortheFOMC’s thetargetfedfundsratecomingoutofanFOMC decisiononthetargetfederalfundsrate.The meetingandtheratecomingoutofameetingsix mostrecentstatement,issuedtheafternoonof monthslater.Howmuchofthechangeinthe March22,isagoodexample.Thestatement targetrateoverthesix-monthintervalisdeter- emphasizedtheFOMC’sincreasedconcernover minedbyinformationavailableatthefirstmeet- inflation,asevidenceisaccumulatingthatfirms ingandhowmuchbyunforecastableinformation perceiveanincreaseintheirpricingpower.From arrivingbetweenthetwomeetings?Ihavenot myperspective,themarketreactiontothatstate- attemptedaformalstatisticalinvestigationof mentmadealotofsenseandreflectedmyown thisquestion,butamwillingtoassertthatmost assessmentofachanginginflationenvironment. ofthetimethechangeinthetargetrateovera Thechangedlanguageinthestatementnotonly six-monthintervalisdrivenbynewinformation explainedthe25-basis-pointsincreaseinthetar- overthesixmonthsandnotmuchbyinformation getfederalfundsratebutalsotherenewalofthe inhandatthebeginningoftheinterval. 3 MONETARYPOLICYANDINFLATION Ifmyclaimiscorrect,thenmostofthetime thenextsixmonthsorsoseemsclearlygreater theFOMCcannotprovideaccurateinformation thantheriskthatinflationwillfallbelowadesir- tothemarketastotheprobablecourseofthe ablerange.Theaimofmonetarypolicyshouldbe targetfedfundsrate,intermsofaspecificpath tocounterinflationpressureswithalessaccom- measuredinbasispoints.Thefuturepathwill modativepolicystance,sothathigheractual beconditionalonfutureinformationthatcannot inflationdoesnotextendbeyondunavoidable itselfbepredicted.Attemptstoprovidespecific transitoryeffects.Putanotherway,monetary forward-lookingguidancewillproveinaccurate policyshouldensurethatinflationpressuresdo andevenmisleadingtothemarket.Moreover,the notgetbuiltintoinflationexpectationsanda Fedcouldcreateacredibilityproblemforitself higheractualinflationrateoveramedium-term ifforwardguidanceistoospecific.Ifthemarket horizon. actsontheguidance,andtheFedsubsequently TheFed’scommunicationissueisthis:Can respondstonewinformationinawaythatdeparts thepolicystatementprovideanaccurateviewof fromtheguidance,thenthemarketwillnaturally theextenttowhichtheCommitteebelievesthe feelthatithasbeenmisled.ButiftheFedfailsto policyoutlookhasbecomeasymmetricwithout respondtonewinformationthatseemstodemand triggeringfirmmarketexpectationsofanadjust- aresponse,intheinterestofdoingwhatitsaidit mentofthefederalfundstargetatthenextmeet- wasgoingtodo,thenfailuretorespondmayalso ing?Asymmetrydoesnotnecessarilyimplya damagecredibility. highprobabilityofaratechangeinaparticular Inshort,theFedshouldonlyofferspecific directionatthenextFOMCmeeting,orthesev- guidanceforthefedfundstargetpathincircum- eralmeetingsafterthat.Whatasymmetrydoes stancesinwhichitishighlyprobablethatthe implyisthatifnewinformationmostlycomesin specifiedpathwillremainappropriateinthe oneway—positivesurprises,forexample—then faceofnewinformation.Generally,commitments apolicyadjustmentwillprobablybeappropriate. needtobecarefullyconditionedsothatthemar- Thecommunicationchallengeistoprovide ketunderstandsthatnewinformationmayrequire themarketacorrectsenseoftheprobabilities adifferentpolicycourse.TheFedneedstoexplain andofthesortofnewinformationthatmight asclearlyaspossibletheconditionalityofa justifypolicyaction.Idonothaveasettledview commitment,andhowthenatureofthecommit- onthisissue,butleantowardthepositionthat mentmustdependoncircumstances.Mostof tryingtoconveyanaccuratesenseofasymmetry thetime,themostlikelyfedfundstargetratesix ismorelikelytobemisleadingtothemarketthan monthsinthefuturewillbeclosetotherateat helpful.Theproblemisthatthemarketmayread thebeginningoftheperiod;theoutcomewill astatementofasymmetryasahintofexpected dependonshocksovertheinterval. policyactionatthenextFOMCmeeting,when Itisalsotruethatmostofthetimethepolicy nohintisintended.Myownpreferenceistodis- outlookisasymmetric.Asymmetryarisesbecause cusstheissueingeneral,andletthemarketsense itisordinarilythecasethattheeconomy’sper- adevelopingasymmetryfromobservingthesame formancepointstotheneedforthefederalfunds informationflowtheFOMCobserves.Then,when targetrateincomingmonthstoeither:a)staythe subsequentshockscreatethecaseforfuturepolicy sameorincrease;orb)staythesameordecrease. action,themarketwillrespondtothoseshocks Itisrareforthesituationtobesuchthattheodds bychangingitspolicyexpectations,whichwill oftighteninginthenearfuturearethesameas berecordedinthefederalfundsfuturesmarket. theoddsofeasing.Thepolicysituationbecomes I’veemphasizedtheimportanceofclarityof asymmetricwhentheeconomicsituationbecomes policygoals.I’veemphasizedtheimportanceof asymmetric.AsIspeakrightnow,forexample, reducinguncertaintywithrespecttoFedpolicy theupwardthrusttotheeconomyappearsquite whereverpossible.I’llfinishwithoneothercom- substantialandtheriskofhigherinflationover ment.Policystatementsneedtobecraftedcare- 4 HowShouldtheFedCommunicate? fully,withspecialattentiontothepossibility CONCLUDING COMMENT thatlanguagewillbemisunderstood.Thistaskis Clearpolicycommunicationrequiresapolicy muchharderthanitmightseem,inpartbecause framework.Theabstractionofarationalexpec- economists’languagecontainswordswithiden- tationsmacroeconomicequilibriumprovides ticalspellingbutdifferentmeaningsthaninordi- suchaframework.Anessentialfeatureofsuch naryEnglish.ItisimportantthattheFedavoid anequilibriumisthattheeconomyevolvesin inadvertentlymisleadingthemarket.Forthis responsestoshocks—newinformationthatcan- reason,Ihavebeenanadvocateofcraftingthe notbepredictedinadvance.Appropriatemone- policystatementattheconclusionofeachFOMC tarypolicymustalsoevolveinresponsetothe meetingfromstockphrasesthathavebeengiven sameshocks.Ibelievethatthisframeworktakes clearmeaningthroughdiscussionovertimeand usalongwayinthinkingwithprecisionabout sustainedandrepeateduseovertime.Anyone communicationissues.Weneedtobeasclearas whohasreadFOMCmeetingstatementsknows wecanbeaboutwhatweknow,andemphasize thatcontinuityisanimportantfeatureofthem, thatmonetarypolicyactionsmustbeconditioned andthatthemarketappropriatelyfocusesonwhat onnewinformation.That’smymessage—Ihope hasandhasnotchangedfromonemeetingtothe I’vecommunicatedclearly! next. Althoughstatementsinrecentyearsreflect considerablecontinuity,changesusuallycomeas REFERENCES asurprisetothemarket,andtheinitialmeaning ofnewphraseshasnotalwaysbeenclear.For “HowDoCentralBanksTalk?”InAlanBlinder,etal., thatreason,IthinktheFOMCcouldimprove eds.,GenevaReportsontheWorldEconomy3. clarity,especiallywhenpolicydirectionchanges, Geneva:CentreforEconomicPolicyResearch. byagreeinginadvanceonstockphrasesto InternationalCenterforMonetaryandBanking describedifferentsituations.Thestockphrases Studies,2001. fromwhichpolicystatementswouldbeput togethercouldbeexplainedinadvanceofactual Poole,W.andRasche,R.H.“PerfectingtheMarket’s use.Somewillregardresorttostockphrasesas KnowledgeofMonetaryPolicy.”Journalof boilerplatethatistheveryoppositeoftrans- FinancialServicesResearch,2000,18(2/3),pp. parency.Myviewisthatahandfulofstandard 255-98. optionstodescribetheCommittee’ssummary Poole,W.“FedTransparency:How,NotWhether,” evaluationofthestateoftheeconomyandthe FederalReserveBankofSt.LouisReview,2003, stanceofpolicywouldpromoteclarityandthere- 85(6),pp.1-8. foretransparency.Thepurposeoftransparency isnotservedwhenthemarketgreetsanewstate- mentwithmultipleinterpretations. 5

William Poole (2005, April 1). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_20050402_poole

@misc{wtfs_speech_20050402_poole,
  author = {William Poole},
  title = {Speech},
  year = {2005},
  month = {Apr},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_20050402_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}