speeches · September 28, 2006

Speech

William Poole · President

Data Dependence MiddleTennesseeStateUniversity AnnualEconomicOutlookConference Murfreesboro,Tennessee September29,2006 PublishedintheFederalReserveBankofSt.LouisReview,March/April2007,89(2),pp.77-83 Iam very pleased to participate in the SOME BACKGROUND AnnualOutlookConferencehereatMiddle Morethanthreeyearsagonow,inJune2003, Tennessee State University. However, per- theFederalOpenMarketCommittee(FOMC)set haps strangely, I’ll not say much about the itsfederalfundsratetargetata40-yearlowof1 outlook. Others are better qualified than I to dis- percent,completing,asitturnedout,aseriesof cussthatsubject.MytopicishowtheFedadjusts reductionsfromarateof6½percentin2000.The policy when the economy departs from the cen- policystatementaccompanyingthechangeinthe traltendencyoutlook.Ofcourse,forecasterscom- policytargetconcludedwithaconcernaboutan monly have somewhat different views, but each “unwelcomesubstantialfallininflation.”The forecaster’scentraltendency,orbaseline,forecast declineintheinflationratewasonlyoneofa provideshisorherbestguessastohowtheecon- stringofsurprisestowhichtheFOMCreactedas omy will evolve. However, forecasters also need itbroughtitstargetratedown.Themostshocking to be able to say something about probabilities ofthesurprises,ofcourse,wastheterroristattack of other outcomes. The probability distribution ontheUnitedStatesonSeptember11,2001.It of possible outcomes is substantially affected by wouldbetimeconsuming,butnotdifficult,to policy responses to deviations from the baseline recountthishistory,pointingtothedatareleases outlookifandwhenthosedeviationsoccur.And, andeventsthatledtheFOMCtoreduceitstarget although I say “if and when,” everyone in the ratebetweenearly2001andJune2003;suchan forecasting business knows that our knowledge accountwouldprovideaclearillustrationofwhat offorecasterrorsrequiresthatweputmuchmore ismeantby“datadependence.” weight on the “when” than the “if.” Theroughlytwo-yearperiodafterJune2003 TheviewsIexpressherearemineanddonot wasquitedifferentinthesensethatmonetary necessarilyreflectofficialpositionsoftheFederal policydoesnotappeartohavebeenverydata ReserveSystem.Ithankmycolleaguesatthe dependent.FollowingitsmeetingonAugust12, FederalReserveBankofSt.Louisfortheircom- 2003,theFOMCissuedastatementthatsaid, ments.BillGavin,vicepresidentintheResearch amongotherthings,that“theCommitteebelieves Division,providedspecialassistance. thatpolicyaccommodationcanbemaintained Letmealsonoteattheoutsetthatthisspeech foraconsiderableperiod.”Thefundsratetarget issomethingofacompaniontoanotherspeechI remainedat1percentforafullyear.Theeraofa gaverecently,“UnderstandingtheFed,”which 1percenttargetendedwhentheFOMCraisedthe waspublishedintheSt.LouisFed’sReviewand targetto1¼percentonJune30,2004,apolicy isavailableonourwebsite.1 adjustmenttheFOMChadsignaledatitsprevious 1 WilliamPoole,“UnderstandingtheFed,”FederalReserveBankofSt.LouisReview,January/February2007,89(1),pp.3-13; http://research.stlouisfed.org/publications/review/past/2007/. 1 MONETARYPOLICYANDINFLATION meetinginMay.Bythen,astheeconomy’srecov- outtobeagoodtimetopausetotakestockof erycontinued,therewasnodoubtthattheFOMC wheretheeconomystoodandthelikelycourse wouldhavetoraiseitspolicytargetbyasubstan- ofeventsgoingforward.WhethertheAugustdeci- tialamounttosupportitslong-terminflation siontoholdthetargetfundsrateunchangedwill objective. turnouttobeapauseintheprocessofraising Ineachofthenext16consecutivemeetings, rates,alonger-lastingstop,oreventhepeak,will theFOMCvotedtoraisethetargetforthefederal dependontheeconomy’sevolutionincoming fundsrateby25basispoints,finallypausingat5¼ months. percentinAugustof2006.Itappearedtosome thatpolicywasonautopilot,astheFOMCraised thetargetby25basispointsmeetingaftermeeting, THE MODEL apparentlyindependentofincominginformation. Tooperatemonetarypolicyeffectivelyand Thatview,Ibelieve,wasmistaken.Whenthe tounderstandhowpolicyactionsaffecttheecon- FOMCbegantheseriesofrateincreases,inJune omy,theFederalReservereliesheavilyoneco- 2004,thestatementincludedthissentence: nomictheorydevelopedoverthespanofmany “Nonetheless,theCommitteewillrespondto decades.Thetheoreticalframeworkiscompli- changesineconomicprospectsasneededtofulfill catedinitstechnicalformandimplementation itsobligationtomaintainpricestability.”Similar butquitestraightforwardinitsbare-bonesabstract languagehasappearedineverystatementsince, framework.Therealeconomyevolvesalonga andtheminutesofthemeetingshaveemphasized trendthatisbuffetedbyavarietyofeconomic thesamepoint.Whathappenedoverthe18 shocks.Inflationevolvesalongatrendthatis monthsafterJune2004was,basically,thatincom- determinedbymonetarypolicyandalsobuffeted ingdataindicatedthattheeconomywassoclose bythesesameeconomicshocks.Althoughthese tothetrackexpectedearlierthattherewasnorea- shocksdrivethebusinesscycleandmakethe sontodepartfromthe“measuredpace”ofrate near-termuncertain,expectationsaboutlonger- increasesof25basispointsateverymeeting. Mypurposetodayistodiscussinasystematic termtrendsinbothrealoutputgrowthandinfla- fashionthedependenceofpolicyonnewinfor- tionhavebecomequitestable. mation.Icangiveyouafeel,thoughnotaformula, Long-runoutputgrowthhasalmostalways forwhypolicydecisionsaresometimesmoredata beenfairlypredictablebecauseitstrendisdeter- dependentthanatothertimes.Whenthetarget minedbythetrendsinthegrowthofrealfactors ratewasat1percent,oronlymodestlyabove,it suchasthelaborforce,thecapitalstock,andthe wasclearthatrateshadtorise,butasufficiently leveloftechnologyinscience,industry,and largesurprisewouldhaveledtheFOMCtostop, management.Thesetrendsevolveslowly;since slow,oracceleratetheincrease.Intheevent,data WorldWarII,realgrowthhasfluctuatedaround surpriseswereminimalandtheFOMCraisedthe a3½percentaverageandforecastsoffuture targetby25basispoints17timesinarow.Increas- growthtendtobecenteredonthatnumberor ingly,though,astheFOMCraisedthetargetfunds perhapssomewhatlowerbecauselaborforce rate,policymakersbecamemoresensitivetothe growthisslowingasbabyboomersretire. possibilitythatdatasurprisescouldalterthe Inflation,ontheotherhand,hasnotalways policycourse.Asitturnedout,thedecisionto beensopredictable.Before1987,therewerewide stopraisingrateswasdetermined,inmymind, swingsintheinflationtrendand,unlikethecase lessbydatasurprisesthanbytheeconomy’sslow- forrealgrossdomesticproduct(GDP),long- ingmoreorlessashadbeenexpectedmany horizonforecastsofinflationwereactuallymore monthsbefore.TheAugustFOMCmeetingturned uncertainthanshort-horizonforecasts.2Today, 2 SeeStephenK.McNees,“HowAccurateAreMacroeconomicForecasts?”FederalReserveBankofBostonNewEnglandEconomicReview, July/August1988,pp.15-36. 2 DataDependence afteraquartercenturyofeffortbytheFedto suchadvanceisonthehorizon;butevenifitwere, activelycontaininflation,inflationhasalso itwouldnotbeanexceptiontotherulethatthe becomemorepredictableoverallhorizons;and policyobjectiveshouldbeindependentofincom- forecastsoverlongerhorizonsarenowmuchmore inginformationaboutthecurrentstateofthe accuratethanthoseovershorterhorizons.3Evi- economy.Thepolicyobjectivedeterminesthe dencethatlong-terminflationhasbecomemore long-runinflationtrendinourmodeland,more predictableisimportant,becauseitmeansthat importantly,thenominalanchorfortheeconomy. theFedhasfoundawaytoanchortheinflation Thereactionofpolicytoincomingnews trend. dependsonthestateoftheeconomyrelativeto Thus,ourbasicmodelisofaneconomyin thetrends.Theprivatesectorneedstoknowthe whichbothrealgrowthandtheinflationrateare FederalReserve’sinflationobjectivesothatit buffetedbyeconomicshocksintheshortrunbut knowshowtoviewfluctuationsaroundthetrend. thentendtoreturntopredictablelong-termtrends. Recently,severalindividualFOMCmembers Thefluctuationsofbothoutputandinflation havecharacterizedthelong-runinflationgoalas aroundtrendshavemoderatedagreatdealover a“comfortzoneof1-2percentinflation”asmeas- thepast25years,partlyandimportantlybecause uredbyinflationinthechainpriceindexforper- ofbettermonetarypolicy.Thisbetterpolicyis sonalconsumptionexpenditures.Althoughthe duetotheFedconcentratingonitsobjectivefor FOMCitselfhasnotadoptedaformal,quantita- long-runpricestabilitythroughamoresystematic tiveinflationobjective,severalmembers,includ- reactiontoincominginformationabouttheeco- ingme,havesaidthattheybelievethatgreater nomicshocks. clarityaboutthelong-runobjectivewouldhelp Atonetime,manyeconomistsbelievedthat boththeCommitteeandthemarketstomakemore therewasaninherenttensionbetweenstabilizing informeddecisions. inflationandstabilizingtherealeconomy.Over Itismucheasiertoagreeonalong-runinfla- thepast25years,wehavelearnedthatacondi- tionobjectivethanonshort-runpolicyactions tionforstabilizingtherealeconomyisstabilizing consistentwiththeobjective.Thereisagreement long-runinflationexpectations.Thus,oneofthe ontwoconflictingprinciples.First,itisalltoo mostimportantthingstounderstandaboutthe easytooverreacttoshort-rundevelopments. dependenceofmonetarypolicyactionsonarriving Agreementonthatprincipleisreflectedinthe informationisthattheFederalReservehasadeep FOMC’semphasisoncoreinflation—inflation commitmenttoachievingalong-runoutcomefor measuresexcludingvolatilefoodandenergy inflationthatisinaccordwithitspricestability prices—asaguidetoshort-runpolicy.Moreover, objective.Putanotherway,short-runpolicyis above-trendinflationmaybeacceptableunder stronglymotivatedbylong-runconsiderations. somecircumstances,providedweareconfident thatpastpolicyactionshavebeensufficientto slowinflationinthefuture.Nevertheless,there MONETARY POLICY isalsoagreementonasecondprinciple:Itisall Afundamentalcomponentofmonetarypolicy tooeasytoallowwishfulthinkingoninflation isthedecisionaboutthelong-runpolicyobjective todelaynecessarytoughpolicydecisions.The forinflation.Thisaspectofpolicyshouldnotbe FOMCdoesitsbesttomaketherightchoices datadependent.Itispossiblethatanadvancein when,asisoftenthecase,“alltooeasytooverre- economicknowledgewillteachthatweshould act”collideswith“alltooeasytoallowwishful haveadifferentlong-runinflationobjective.No thinkingoninflation.” 3 Seeevidenceonforecasterrorsover3-,12-,and24-monthintervalsfrom1997through2006inWilliamT.GavinandKevinL.Kliesen, “ForecastingInflationandOutput:ComparingData-RichModelswithSimpleRules,”FederalReserveBankofSt.LouisWorkingPaper 2006-054A,September2006. 3 MONETARYPOLICYANDINFLATION Inonesense,long-runpolicyistheaccumu- Reserveisgoingtododependsonwhatthemarket lationofindividualshort-runpolicydecisions. andtheeconomyareanticipatedtodo.Thefull However,ifindividualdecisionsreflectonlyreac- rationalexpectationsmacroeconomicequilibrium tionstoshort-rundevelopmentsintheeconomy, occurswhenthemarketbehavesastheFederal thenthereisnotellingwherelong-runpolicy ReserveexpectsandtheFederalReservebehaves willgo.TherightwayfortheFedtothinkabout asthemarketexpects.Inbothcasesweassume short-runpolicydecisionsisthattheyhavetobe thattheexpectationsarefullyrational,bywhich partof,orfitinto,acoherentlong-termplan.The wemeanthattheexpectationsarefullyinformed market’sunderstandingofthisplaniscentralto onthebasisofallavailableinformation.The thedeterminationoflong-terminterestrates.In abstractionofafullrationalexpectationsmacro- general,therateonanybonddependsonexpected economicequilibriumprovidesapowerfulstart- shortratesoverthehorizonofthebond.Thus,the ingpointforanalysisofadata-dependentpolicy. 10-yearTreasurybondratedependsonexpecta- tionsofshort-terminterestratesoverthe10-year horizon. CAN THE MARKET PREDICT Marketexpectationsaboutfutureinterestrates DATA DEPENDENCE? dependontheinteractionoftwointerrelated sourcesofinfluence.One,obviously,concerns The“Taylorrule”isastylizedviewofthe FederalReservedecisionsontheintendedfederal Fed’sreactiontoincominginformation.In1993, fundsrate.Alsoimportantareexpectationsasto StanfordeconomistJohnTaylorproposedasimple thedemandsforandsuppliesoffundsinthe formularelatingthefederalfundsrateto(i)a privatemarket.Forexample,withsimultaneous long-runinflationtargetand(ii)short-rundevia- investmentandhousingbooms,creditdemands tionsofinflationfromthattargetandshort-run willbehighandinterestrateswilltendtobebid deviationsofrealGDPfromameasureof“poten- up.Inpursuingitspolicygoals,theFOMCwill tialrealGDP.”4Taylorsuggestedthathissimple beadjustingthefederalfundsrateasneededto relationshipcharacterizedinbroadoutlinethe keeptheinflationratelowandstable.Thus,the actualbehaviorofthefederalfundsrateinthe marketformsexpectationsabouttheunderly- earlyyearsoftheGreenspanFOMC.Theessence ingstateoftheeconomythatwillbearonFed ofthisrelationshipisthatinthelong-runthe decisions. FOMCseekstokeepthefederalfundsrateroughly TheFederalReserveisconstantlyevaluating consistentwithalevelthatisbelievedtoproduce thesituationinthemarketsandtryingtoadjust atargetlevelofinflation.Taylorassumedatarget theintendedfederalfundsratetoproduceasatis- rateofinflationof2percentperyearmeasured factoryequilibriumintheeconomy.Whenweput bythetotalconsumerpriceindex(CPI).Inthe theFederalReserve’sandthemarket’sdecisions shortrun,therelationshipimpliesthattheFOMC andexpectationstogether,wehaveamacro- adjuststhetargetfederalfundsrateupaseither economicequilibrium. theobservedinflationrateexceedsitstargetor TheinteractionbetweentheFederalReserve realGDPexceedspotentialrealGDP.Conversely, andthemarketsmaybeconfusingatfirstsight, undertheTaylorrule,theFOMCreducesthetar- andindeedwasconfusingtoeconomistsforgen- getfederalfundsratewheninflationfallsbelow erationsuntilconceptualbreakthroughsinthe itstargetand/orrealGDPfallsshortofpotential 1960sand1970sclarifiedtheissue.Marketbehav- realGDP. iordependsonexpectationsastowhattheFederal TheTaylorrulereflectstheprimacyofalong- Reserveisgoingtodo,andwhattheFederal runinflationobjectivewhileincorporatingshort- 4 JohnB.Taylor,“DiscretionversusPolicyRulesinPractice,”Carnegie-RochesterConferenceSeriesonPublicPolicy,39,December1993, pp.195-214.Taylorcomparedthevaluesofhisformulaagainsttheobservedhistoryofthefundsratefrom1987through1992. 4 DataDependence runstabilizationefforts.Theruleprovidesafor- outlookshouldbeconsistentwithkeepingthe mulaforcomputingabaseline,orreference, economyclosetofullemployment,basedonthe interestratethatisconsistentwithpolicyachiev- CongressionalBudgetOfficeforecastofpotential ingtheFed’sobjectivesforbothoutputstabiliza- GDPgrowthof3.24percentin2007.Asforinfla- tionandpricestability.IdiscussedtheTaylorrule tion,thecentraltendencyforecastofFOMCpar- insomedetailinthespeechImentionedearlier, ticipantsfor2007is2to2¼percent.Thus, “UnderstandingtheFed,”andreferyoutoits inflationisexpectedtorecedeonlyveryslowly publishedversioninthisReviewifyouwantto fromitscurrentlevel. digintothesubjectmoredeeply. Therearetwocasesinwhichtheeconomic NowI’llturntosomecommentsonfuture newswillprettyclearlypredictachangeinthe Fedpolicy,butIwanttoremindyouthatIam Fed’spolicystance.Ifincomingeconomicindi- speakingformyself—otherFOMCparticipants catorsshowthatbothoutputandinflationare mayhavedifferentviewsabouthowfuturepolicy risingabovetheseforecasts,then,intheabsence adjustmentswilldependonarrivinginformation. ofanyotherinformation,wecanexpectthatthe Alleconomicindicatorsmayhaveimplications FOMCwillincreaseitstargetfederalfundsrate. fortheevolutionoftherealeconomyandinflation. Ontheotherhand,ifbothoutputandinflation Iemphasize“may”becausewehavetofilterout comeinweakerthanexpected,weareunlikely toseefurtherincreasesinthefederalfundstar- asbestwecanpossibledataerrorsandinconsis- get;indeed,ifeconomicweaknessispervasive tenciesacrossvariousindicators. enough,theFOMCwillatsomepointreducethe BeforeIdiscussfutureFedpolicyinanydetail, targetfundsrate. Ibeginwithawarning.Newinformationdrives Themostinteresting—nottomentioncontro- bothmarketadjustmentsandpolicychanges,but versialanddifficult—casesarethoseinwhichthe newinformationisinherentlyunpredictable.To outlookforinflationandoutputmoveinoppo- gainasenseoftheimpactofnewinformationon sitedirections.Insuchcases,theFOMChasto interestrates,I’veanalyzeddatafromtheeurodol- callonallitsexperienceandjudgmenttoreacha larfuturesmarketanddiscussedtheresultsin decision.Itisverydifficultformetobeprecise somedetailin“UnderstandingtheFed.”The aboutthejudgmentsIamlikelytoreachbased bottomlineofthatanalysisisthatforecastsembed- onincominginformationbecauseahostofcon- dedintheeurodollarfuturesmarketexplain42 siderations,someofwhichIcannotforesee,may percentofthevarianceoffluctuationsintheactual enterthecalculus.ButI’llmakeastabathow eurodollaryieldthreemonthsahead.Thus,unpre- thingscouldplayouttoillustratemythought dictableeventsevenoverathree-monthhorizon process. areresponsiblefor58percentofthevarianceof Acriticallyimportantconsiderationinmy theeurodollaryield.Overasix-monthhorizon, mindconcernstheinflationprocessandthe unpredictableeventsareresponsibleformorethan importanceoftheFed’scommitmenttolowand 70percentofthevariance.Thus,Icandiscuss stableinflation.Itismyconvictionthattemporiz- variousscenariosbuthavenowayofknowing ingonactionstocontrolinflationisaninvitation whichscenariowillcometopass. totrouble.Acceptinghigherinflation,orevena Let’sstartwiththeoutlookfortherestof continuationofthecurrentrateofinflation,in 2006.ForecastsmadebyFOMCmembersand anefforttosustaincurrentemploymentlevels transmittedtoCongressinJulywere3¼to3½ willonlyleadtomoregrieflater.Onceinflation percentgrowthforrealGDPandanincreasefor andinflationexpectationsrise,theeconomywill thecorepersonalconsumptionexpenditures becomelessstableandreducinginflationfrom (PCE)chainpriceindexof2¼to2½percent.As anelevatedratewillbemorecostlythantaking for2007,thecentraltendencyoftheFOMCmem- themedicinenow.Havingsaidthat,ifinflation bers’GDPforecastsis3to3½percent.Thisgrowth pressuresareeasing,evenifonlygradually,and 5 MONETARYPOLICYANDINFLATION thereisagenuineprospectthatinflationwill information.Still,Ibelieveforecastersshould returntothecomfortzone,thenIseenoreason assignarelativelylowprobabilitytodeepreces- toacceleratethedeclineininflationbymaintain- sionpreciselybecauseoftheFOMC’sdemon- ingarestrictivepolicyinthefaceofdeclining stratedwillingnesstoactaggressivelyasnecessary. employment.Policyneedstobeasdisciplined I’vegivenyoumytakeonwhatdatadepend- asnecessarytogetthejobdone,butnotmoreso. encemeansandtheattitudesthatunderliemy Thelong-runinflationgoalandtheattitude likelyresponses.I’vealsoemphasizedthatan I’veexpressedaboutwhatriskstotakesuggest efficientrationalexpectationsequilibriumrequires thatIwillhaveabiasinthewayIinterpretincom- thatthemarketbehaveasthepolicymakersexpect inginformation.Ifdataontherealeconomycome andpolicymakersbehaveasthemarketexpects. inweakerthanexpected—ifitappearsthatthe Themarket’sevaluationoftheprospectsforpol- economyisfallingbelowthebaselineforecast icyisrevealedinthefuturesmarketsforfederal path—thenmybiaswillbeinthedirectionof fundsandeurodollardeposits.Currentfutures wantingtobesurethatthedatapaintaconsistent pricespredictthatthefederalfundstargetis picturebeforeI’lladvocateapolicyeasing.But expectedtobeginmovingdown.Becausethese ifthepictureisconsistent,andinflationriskis marketquoteschangedaybydayinresponseto receding,thenI’llnothesitatetoadvocatepolicy newinformation,Idonotwanttoattempttobe easing. particularlypreciseastothetiming—anythingI WhatIhopetheFOMCcanaccomplishisto writeasIdrafttheseremarksmaybeoutofdate retainfullmarketconfidencethatthelong-run bythetimeIdeliverthemorwithinafewweeks, rateofinflationwillremaininthecomfortzone. anyway.WhatIcansafelynoteisthatthemarket’s Ihopethatforecastersassignverylowprobabil- expectationoffuturepolicyeasinghasbeentaking itytoinflationoutcomesoverthemediumterm holdgraduallysincelateJune,say,inresponse of3to5yearsoutsidethecomfortzonenomatter todataontherealeconomysuggestingthatreal whattheincomingdatalooklike.AlthoughIam growthisslowingandinflationdatasuggesting talkingaboutinflationoverahorizonwellbeyond thattheworstmaybeoveronthatfront. theusualforecasthorizonof1to2years,thelong- Althoughexpectationsaboutfuturepolicy runinflationoutlookhasadirectbearingonthe actionsarerevealedtransparentlyinthefutures forecast.Thelongbondratetodaydependscriti- marketforshort-terminterestrates,Iwantto callyonexpectedinflationoverthematurityof underscoremyearlierpointaboutthelimited thebond.Thus,ratesthatenterimportantlyinto accuracyofthoseforecasts.Someoftheforecast anyeconomicforecast,suchasmortgageand misseshavebeenprettydramatic.Forexample, corporatebondrates,dependonthelong-run inDecember2000,thefuturesmarketforecasts inflationoutlook.Thisoutlookhasbeenquite wereforadeclineintheeurodollaryieldof35 stableinrecentyears,andthatfactisevidenceof basispointsoverthefollowingthreemonths amajormonetarypolicysuccess. andatotalof67basispointsoverthesix-month Withlong-runinflationcontained,theFOMC period.Instead,theFOMCactedaggressivelyto hasflexibilitytorespond,vigorouslyifnecessary, lowerthefundsratetargetstartinginJanuaryand toeconomicweaknessshoulditarise.TheFOMC continuingthroughMay2001byatotalof250 broughtthetargetfederalfundsratedownaggres- basispoints.TheFOMCactedaggressivelyas sivelyin2001inresponsetoincominginforma- incominginformationpointedtogrowingweak- tion.Aggressiveeasingkepttherecessionmild. nessineconomicactivity.BoththeFOMCand Iftheeconomycomesinbelowthebaselinefore- themarketsweresurprisedbyincominginforma- castincomingquarters,theFOMCwillhaveroom tionindicatingthattheeconomywasweakening toactasaggressivelyasrequired.Ihavenoidea quicklyandsignificantly. whatscaleofeasingmightbeappropriate,for AlthoughIcannotpredictunpredictablenew thatwilldependonthenatureoftheincoming information,I’vetriedtoprovideasenseofhow 6 DataDependence Imightrespondtonewinformationasitarrives. thecentralbankcaninprinciplespecifyapolicy Inote,however,thatitisrarethatasingledata rule,orresponsefunction,thatguidespolicy reportisdecisive.Theeconomicoutlookisdeter- adjustmentsinresponsetoincominginformation. minedbynumerouspiecesofinformation.Impor- Toachieveagoodresult,thegeneralpublicand tantdatasuchastheinflationandtheemployment marketparticipantsneedtounderstandtheobjec- reportsarecross-checkedagainstotherinforma- tivesandtheresponsefunctionsothattheprivate tion.TheFOMCisawareofthepossibilityofdata economycandetermineitsactivitieswithfull revisionsandshort-runanomalies.Sometimes knowledgeofhowthecentralbankwillact.Of dataoughttobediscountedbecauseofanomalous course,uncertaintyisaninherentcharacteristic behavior. oftheeconomicworld.Whatshouldbepredict- Anexamplewastheincreaseintobacco ablearethecentralbank’sresponsestothenever- pricesinlate1998.Tobaccopriceshadatransi- endingsequenceofsurprisesthatcharacterize toryimpactonmeasuredinflation,bothtotaland theeconomicenvironment. coreindices,duringDecember1998andJanuary Marketcommentaryoftenindicatesfrustra- 1999,butproducednolastingeffectontrend tionthattheFOMCdoesnotlayoutaclearerpath inflation.Similarly,informationaboutrealactivity forpolicy,arguingthattheFOMCisunpredict- sometimesarrivesthatindicatestransitoryshocks able.Thatview,Ibelieve,isoffbase.Typically toaggregateoutputandemployment.Anexample theFOMCcannotbepredictablewithregardto ofsuchatransitoryshockisthestrikeagainst thepathofthetargetfederalfundsratebecause GeneralMotorsinJuneandJuly1998.Similarly, newinformationdrivingpolicyadjustmentsis theSeptember2005employmentreportreflected notpredictable.Allofuswouldliketobeableto theimpactofHurricaneKatrina,whichwas predictthefuture.WeintheFeddothebestwe expectedtobe,andturnedouttobe,temporary can,butthemarketsshouldnotcomplainthat fromanationalperspective. theFOMClacksclairvoyance!WhattheFOMC Transitoryandanomalousshockstothedata strivestodoistorespondsystematicallytothe areordinarilyrathereasytoidentify.BothFedand newinformation.Thereisconsiderableevidence marketeconomistsdevelopestimatesofthese thatthemarketdoessuccessfullypredictFOMC aberrationsinthedatashortlyaftertheyoccur. responsestotheavailableinformationatthe Theprincipleoflookingthroughaberrationsis timeofregularlyscheduledmeetings.5 easytostatebutprobablyimpossibletoformalize withanyprecision.Weknowtheseshockswhen weseethem,butcouldneverconstructacom- CONCLUDING COMMENT pletelycomprehensivelistofsuchshocksexante. Tosaythatpolicyisdatadependentmeans Policymakerspiecetogetherapictureofthe thatpolicychangeswilldependontheincoming economyfromavarietyofdata,includinganec- newsaboutthestateoftheeconomy,bothreal dotalobservations.Whenthevariousobserva- growthandinflation.Thatthepolicysettingis tionsfittogethertoprovideacoherentpicture, datadependentisagoodsign.Itmeansthatpolicy theFedcanadjusttheintendedratewithsome isinarangethatcanbeconsideredneutral— confidence.Themarketgenerallyunderstands thatis,thoughttobeconsistentwiththeFed’s thisprocess,asitdrawssimilarconclusionsfrom longer-runpolicyobjectives.Itisimportantto thesamedata. rememberthatthelong-runinflationobjective So,givenpolicyobjectives,andgivenaview shouldnotbedatadependent.Iftheobjectiveis abouthowpolicydecisionsaffecttheeconomy, wellunderstood,peoplewillknowwhetherthe 5 See,forexample,WilliamPoole,“HowPredictableIsFedPolicy?”FederalReserveBankofSt.LouisReview,November/December2005, 87(6),pp.659-68;http://research.stlouisfed.org/publications/review/past/2005/. 7 MONETARYPOLICYANDINFLATION currentinflationrateisaboveorbelowthedesired trend.Theywillknowhowtointerpretincoming informationtogaugewhatitmeansforthepolicy stance.Ibelievethatisjustaboutexactlywhere wearetoday. 8

Cite this document

APA

William Poole (2006, September 28). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_20060929_poole

BibTeX

@misc{wtfs_speech_20060929_poole,
  author = {William Poole},
  title = {Speech},
  year = {2006},
  month = {Sep},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_20060929_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}