Speech

The Monetary Policy Model NationalAssociationofBusinessEconomics(NABE)AnnualMeeting Boston,Massachusetts September11,2006 Iam sure that all of us find it sobering to Ourmostfundamentalchallengesarebasically meet on the fifth anniversary of the tragic thesameonesSmithnoted.Uncertaintyoverthe eventsofSeptember11,2001.Manyof valueofmoneyisone;anotheristheincentive you were present at the NABE annual ofsovereignstatestodebasethevalueofmoney. meeting at the World Trade Center that day and Myaiminthislectureistoprovideaskeletal were fortunate to escape before the great towers pictureofhowIviewtheFederalReserve’stask. came down. Beforeproceeding,Iwanttoemphasizethat Itisadistincthonortobeheretodaytoaccept theviewsIexpressherearemineanddonot theAdamSmithAward.I’llbetalkingaboutthe necessarilyreflectofficialpositionsoftheFederal subjectIknowbest—monetarypolicy.Iwould ReserveSystem.Ithankmycolleaguesatthe haveusedthetitle,“MyMonetaryPolicyModel” FederalReserveBankofSt.LouisfortheircomexceptforthefactthatthemodelIuseisnotmine. ments,butIretainfullresponsibilityforerrors. Ourcurrentunderstandingofmonetaryeconomicshasbeenbuiltoncontributionsovermany hundredsofyears.Inpreparingthislecture,I RATIONAL EXPECTATIONS remindedmyselfofjusthowclearlyAdamSmith MACROECONOMIC EQUILIBRIUM understoodmonetaryissuesbygoingbackto rereadsectionsoftheWealthofNations.Smith Mostmacroeconomiststodayadheretoa beginsthebookwiththreeshortchaptersonthe modelbasedontheideaofarationalexpectations divisionoflabor.Chapter4isentitled,“Ofthe equilibrium.Allaspectsofgovernmentpolicy OriginandUseofMoney.”Smithexplainsthat entertherationalexpectationsmodel—the divisionoflaborrequiresexchange,andexchange “REmodel”forshort—butI’llconfinemyremarks requiresmoney.Hediscussesthedifficultyof tomonetarypolicy.Policymakersareassumedto conductingexchangeefficientlybecausemetallic haveasetofgoalsandaconceptionofhowthe moneyneedstobeweighedandassayed.Smith economyworks.Theprivatesectorunderstands, discussesthepracticeofprincesandsovereign totheextentpossible,policymakers’views.An statesindebasingthecurrency.Henotesthat equilibriumischaracterizedbyasituationin “[s]uchoperations,therefore,havealwaysproved whichtheprivatesectorhasaclearunderstanding favourabletothedebtorandruinoustothecred- ofpolicygoalsandthepolicymakers’modelof itor,andhavesometimesproducedagreaterand theeconomy,andthepolicymodeloftheeconmoreuniversalrevolutioninthefortunesofpri- omyisasaccurateaspossible.Ifthepolicymakers vatepersons,thancouldhavebeenoccasioned andprivatemarketparticipantsdonothaveviews byaverygreatpubliccalamity.” thatconverge,nostableequilibriumispossible Monetarypolicymakersareacutelyawareof becauseexpectationsastothebehaviorofothers thepotentialformonetarypolicytocreate“a willbeconstantlychanging. verygreatpubliccalamity”andtheyfeeldeeply Inthissetting,marketbehaviordependscentheirresponsibilitynottopermitsuchanoutcome. trallyonexpectationsconcerningmonetarypolicy 1

MONETARYPOLICYANDINFLATION andtheeffectsofmonetarypolicyontheeconomy, Clarity of Goals includingeffectsoninflation,employment,and Fortheprivatesectortoformaccurateexpecfinancialstability.Astableequilibriumrequires tationsaboutfuturemonetarypolicyandoutcomes thatmarketsbehaveaspolicymakersexpectand ofkeyeconomicvariables,policymakersneedto thatpolicymakersbehaveasmarketsexpect. statetheirgoalsclearly.Theliteratureoninfla- Itiseasiesttodescribetherationalexpectationtargetingemphasizesthispoint,andIhave tionsequilibriuminacontextofcertainty.But, longbelievedthattheFOMCcouldimprovethe ofcourse,allthereallyinterestingquestionsarise claritywithwhichitconveysitsobjectivesto inacontextofuncertainty.Oneformofuncerthegeneralpublic.Inthepast,Ihavestatedmy taintyconcernsfuturestatesoftheworld.In ownpersonalinflationobjectiveas“zeroinflaprinciple,wecanthinkaboutamodelinwhich tion,properlymeasured”buthavealsosaidthat marketresponsesandpolicyresponsestonew FOMCagreementonaninflationobjective,which informationreflectmaximizingbehavior.Inthe somemightexpressasa“comfortzoneof1to2 privatesector,inresponsetonewinformation, percentinflation,”ismoreimportantthanwhich householdschangeconsumptionplanstomaxi- precisespecificationisselected.Therearepracmizeutilityandfirmsmakeoperatingandinvest- ticaldifficultiesthatcanandshouldbeaddressed, mentdecisionstomaximizeprofits.Similarly, suchaswhatpriceindextouse,overwhatperiod policymakersrevisethestanceofmonetarypolicy tomeasurepricechanges,andwhatdegreeof inanattempttopursuepolicygoalsasefficiently tolerancetoadoptifinflationrunsoutsidethe aspossible.Thecontinuousflowofnewinforma- range.Idonotbelievethatuncertaintyaboutthe tionincludeseverythingthathappens—weather Fed’sinflationobjectiveisalargeissueatpresent disturbances,technologicaldevelopments,routine butdobelievethatthereisanopportunityto economicdatareportsandthelike.Thus,we improveclarity. canthinkoftheeconomyasevolvingefficiently inresponsetostochasticdisturbancesofallsorts. Maintaining Credibility Ofcourse,inpractice,thereisalsouncertainty Tomaintaincredibility,themonetaryauthoraboutmarketandpolicyresponsestonewinforitiesmustdeliverwhattheysaidtheywould mation,andthereareimprovementsovertimein deliver.Credibilityisessentialtothestabilityof knowledgeabouthowtheeconomyworks.Never- longer-terminflationexpectations.Centralbanks theless,theREmodelprovidesthecoreinsights aroundtheworldemphasizetheimportanceof thatshapemyviewsastohowtomakeroutine achievinglowandstableinflation.IntheUnited policydecisionsandhowtodesignalonger-run States,theFederalReservelostconsiderable programtoimprovepolicy. credibilityinthe1970sbecausetheinflationrate rosetounacceptablelevels.Withimpairedcredibility,theFOMCunderPaulVolckerhadtopur- MONETARY POLICY sueasustainedanti-inflationarypolicyevenin thefaceofthemostsevererecessionsincethe IMPERATIVES IN THE MODEL GreatDepression.Thecostofrestoringcredibility Wecanthinkoftheactual,observedequilib- makesclearthereasonfornotlosingit. riumasafullrationalexpectationsequilibrium Credibilityisnot,however,one-dimensional. undercurrentpolicy.Theremaywellbeanalter- Sustainedlowinflationisdesiredforitsown nativepolicythatwouldinduceanewrational sakebutevenmoreforthecontributionitmakes expectationsequilibriumthatwouldhavemore tohighemploymentandeconomicgrowth.Thus, desirablepropertiesthanthecurrentequilibrium. whileinflationdamagescredibility,soalsocan Thinkingofthemodelthiswayprovidesanagenda highunemployment.Thereisafinebalancehere. forlong-runimprovementinmonetarypolicy. Weknowthatmonetarypolicycannotaffect 2

TheMonetaryPolicyModel employmentinthelongrun,butwealsoknow pleteknowledgeonthepartoftheprivatesector thatmonetarypolicymistakescancreateunem- ofcentralbankthinkingandanalysis,Federal ploymentoveranuncomfortablylongshortrun. Reserveresearchfindingsshouldbedistributed Whenunemploymentrises,policymakersneed widely.Moregenerally,themodelpointstothe tobeabletoexplainincrediblefashionwhythe importanceofreducingasymmetriesofknowlproblemisnotaconsequenceofamonetarypolicy edgethroughcentralbanktransparency. mistake,forthatperceptionisalwayspresent Afindingoftheoptimalcontrolliteratureis thatwhenapolicyauthorityusesallavailable amongsomeobserversinsuchcircumstances. informationasefficientlyaspossibleinpursuing Thereis,afterall,somehistoricaljustificationfor itsgoals,simplecorrelationsbetweenobservable suchaperceptiongiventhatalmostalleconomists variablesandgoalvariablesmaygotozero. agreethatmonetarypolicymistakescontributed AlthoughtheexactresultwillbemodeldependtotheseverityoftheGreatDepression.Giventhe ent,weshouldnotexpectcorrelationssuchas importanceofhighemployment,aperiodofsusbetweenmoneygrowthandnominalGDPgrowth tainedexcessiveunemploymentmaycreate orbetweenunitlaborcostsandinflationtobe doubtsaboutfuturepolicy,andthisuncertainty independentofthepolicyregime.Thisisafamilisamanifestationofimpairedcredibility. iarpropositioninthesecuritiesmarkets,where TheFederalReserve’scredibilityisalso itisstatedastheefficientmarketshypothesis. affectedbyothermatters.Soundperformancein Observableinformationshouldbequickly dealingwithfinancialcrisesandregulatory reflectedinsecuritiesprices,leavingnoriskresponsibilitiesenhancescredibility.Reputational adjustedprofitopportunitiesfromtradingon riskappliestotheinstitutionasawhole.Poor publiclyavailableinformation.Thereis,ofcourse, performanceofresponsibilitiesoutsidethemonanopportunitytoexploitnonpublicinformation. etarypolicyrealmcanaffectthepublic’sconfi- Inthemonetarypolicycontext,researchsugdenceinthecentralbank’sleadership. geststhatinflation-forecastingmodelshavenot Monetarypolicysuccessdependsonhigh workedverywellinrecentyears.Thereason,I credibility,butitisimportanttorecognizethat believe,isthattheFederalReservehasbeenpretty credibilityadherestotheentireinstitutionthat successfulinexploitingallavailablepublicinforconductsmonetarypolicyandnotjusttothe mationinitsmonetarypolicydecisionsaimedat departmentwithintheinstitutionresponsible maintaininglowandstableinflation.However, formonetarypolicy. thereisanopportunitytoexploitnon-public information.TheFederalReservehasanextensiveprocessofgatheringanecdotalinformation SOME IMPLICATIONS OF THE frombusinesscontacts.Muchofthisinformation ispublishedintheBeigeBooktwoweeksbefore MODEL everyFOMCmeeting.ThepolicymodelI’m ThemodelI’vesketchedprovidesawayof sketchingcertainlyleavesroomforgreaterand thinkingaboutcurrentpolicydecisions.Atleast, moresystematicefforttogatherandexploitanecthemodelprovidesawayforyoutothinkabout dotalinformation. mythinkingaboutcurrentpolicydecisions.The Whatdoesthemodelimplyaboutthevalue modelalsoprovidesaframeworkforanalyzing offinancial-marketinformationtopolicymakers? potentialpolicyimprovementsovertime. Considerthefedfundsfuturesmarket.Doesthat Themodelclearlycallsforimprovedclarityof market’sresponseto,say,theemploymentreport centralbankgoals,atopicI’vealreadydiscussed. helppolicymakerstoassessthesignificanceof Clearly,also,themodelsupportscontinuing thenewdata?IntheREmodeltheclearanswer macroeconomicsresearchtoimproveunderstand- is“no.”Thebehaviorofthefedfundsfutures ingofhowtheeconomyworks.Becauseafull markettellsuswhatthemarketbelieveswillbe rationalexpectationsequilibriumrequirescom- theFed’sresponsetothenews.Thatisuseful 3

MONETARYPOLICYANDINFLATION informationfortheFed,butdoesnothelpin asonewhoroutinelytalkswiththepressafter understandingwhatthenewinformationitself everyspeech,Icantellyouthatimpartingthe meansfortheevolutionoftheeconomy.Thesame correctinterpretationofthepolicysignificance argumentholdsforotherfinancialprices,which ofrecentdataisnotasimplematter.Itisalltoo reflectamixtureofdirecteffectsfromnewinfor- easytocreateanunintendedmarketdisturbance. mationandindirecteffectsfromanticipatedpol- Thus,Ipersonallyreallydotakeseriouslythe icyresponsestothenewinformation.The practicalrelevanceoftheextrememodel. implicationforpolicymakersisthatweneedto IhavediscussedtheimplicationoftheRE provideourownanalysisoftheimplicationsof modelforcentralbankcommunicationbecause newinformationandcannotrelyonfinancial itseemstomethattherearemanyopportunities marketstodothatjobforus. forimprovingcommunication,andmanypossi- Themodelalsocontainsasurprisingimpliblecommunicationspitsintowhichitiseasyto cationforcentralbankcommunication.Consider fall.Centralbankcommunicationsarenotjust themostextremeversionofthemodel,inwhich “PR”issues,whichIputinquotestoindicate therearenoinformationasymmetries.Thecentral howsomemightdismisstheirimportance,but bankandtheprivatemarketshavethesameinforessentialtothemonetarypolicyprocess.Expecmation.Theeconomyevolvesasstochasticshocks tationsarecriticaltohowmarketsperformand occur.Whenshocksareobserved,thecentral expectationsareaffectedbothbywhatthecentral bankandthemarketsgetthenewinformationat banksaysandbywhatitdoes. thesametimeandbothunderstandtheimplicationsofthenewinformationforeachothers’ actions.Moreover,andthisisthesurprisingresult, CONCLUDING REMARKS thecentralbankdoesnotneedtosayanything Theeconomicsprofessionhasconvergedto aboutthepolicyimplicationofanyparticular abasicconceptionofhowtheeconomyworks shock.Themarketsalreadyknowtheimplicabasedonarationalexpectationsviewoftheworld. tionsbecausethereisfullknowledgeandno Thatcoreideaisattheheartofeverymodeltoday informationasymmetry. andattheheartofmonetarypolicypractice. Thepracticalimportofthisimplicationfor Nevertheless,thereisampleroomfordisagreecentralbankcommunicationpolicyisthatcommentandfurtherdevelopment.Some—andI municationsshouldfocusonpolicyfundamenincludemyselfinthisgroup—thinkaboutpolicy talsofgoalsandthemodelofhowtheeconomy inawaythatisprettyclosetoanaustereversion works.Theeconomyworksbestwhenpolicyofthemodelinwhichinformationasymmetries makersdisclosethesystematicpartofpolicy playaminimalrole,whileothersseeamuch andminimizetherandompart.Thatis,policy largerroleforimperfectionsinthemarket’sinforshouldnotitselfbeasourceofrandomdisturmationset.Inanyevent,theREmodelprovides bance.Intheextreme,austereversionofthe anorganizingframeworkandcommonlanguage modelIamnowdiscussing,centralbankcomforanalyzingmonetarypolicyissues.Adam municationaboutpolicyresponsestoindividual Smithposedfundamentalissueswell,butinthe shocksisunnecessaryandmorelikelytocreate contextofthegoldstandard.SinceSmith’sday, marketdisturbancesthanenlightenment.Only we’vecomealongway.Indeed,we’vecomea whenthecentralbankbelievesthatthemarketis longwaysincethe1960sand1970swhenfunmisinterpretingthepolicysignificanceofashock damentaldifferencesbetweenKeynesiansand iscommentontheparticularshockdesirable. monetarists,asthesplitwasthendescribed,left Wecanthinkofsuchcommunicationasbeing monetarypolicyinastatethatcouldonlybe designedtoreduceaninformationasymmetry. describedasincoherent.Fundamentaladvances Youmaybelievethatmyargumentfroman ineconomicsreallyhavemadeadifference. extreme,austeremodelisfar-fetched.However, 4