Speech

Inflation Signals and Inflation Noise UniversityofArkansasatLittleRock PeabodyHotel LittleRock,Arkansas April6,2004 The employment report for March, thisreference,soI’llhavetoexplainthatsome released last Friday and showing a jobs analystsarguedthatthedisappearanceof gain of 308,000 over February, was anchoviesfromthecoastofPeruin1972had certainly good news. Everyone hopes somethingtodowithrisinginflationin1973. that monthly reports over the rest of this year Today,however,economistsuniversallyaccept and for years to come will also bring good news thepropositionthatsustainedinflationordeflaon the employment front. tionis,inthewordsofMiltonFriedman,“every- Assumingthatemploymentgainscontinue, whereandalwaysamonetaryphenomenon.” marketfocuswillnaturallyshiftfromemployment TheexperienceoftheGreatInflationbrought toconcernsoverinflationrisks.Indeed,some hometothepublicandpolicymakersalikethe marketcommentaryalreadyhasshiftedinthat burdenthatinflationimposesonaneconomy. direction.Mypurposetodayistoprovidemy Intheabsenceofinstitutionsadaptedtoaninflaperspectiveontheproblempolicymakersfacein tionaryenvironment,theefficiencyofmarket determiningwheninflationrisksarerising. pricestosignaltherelativescarcityofgoodsand Beforeproceeding,Iwanttoemphasizethat servicesisimpaired.IncomesandwealthareredistheviewsIexpressherearemineanddonot tributedcapriciously,evenwhentheinflationis necessarilyreflectofficialpositionsoftheFederal partiallyforeseen,inataxsystemthatisnotfully ReserveSystem.Ithankmycolleaguesatthe indexed.Regulationsonnominalinterestrates, FederalReserveBankofSt.Louisfortheircomsuchasdepositceilingsandusurylaws,interact ments;RobertH.Rasche,seniorvicepresident withtheinflationratetodistortdemandincertain anddirectorofResearch,providedspecialassissectorssuchashousing.Somefinancialinstitutance.However,Iretainfullresponsibilityfor tions,suchasthethriftsthatspecializedinhouserrors. ingfinance,weredriventowardsextinction.Itis possible,thoughtheevidenceisnotconclusive, thatevenmodestsustainedinflationnegatively BACKGROUND impactstherateofgrowthoflaborproductivity. Inthefourdecadessincethebeginningofthe Centralbankersworld-widehavetakenthese GreatInflationofthe1960sand70s,economists lessonstoheart.Sincetheinitialinflationtargetandcentralbankershaveacquiredamuchbetter ingexperimentbytheReserveBankofNew understandingofthesourceandconsequences Zealandin1990,thoughtatthetimetoberadical, ofinflation.WhentheGreatInflationbegan,it inflationtargetinghasbecomethestatedobjective wascommontociteoneoranotheridiosyncratic ofatleast11centralbanks,includingthoseof eventsasthedrivingforcebehindtheobserved developedanddevelopingcountries.Thesecenchangeinprices:OPEC,steelprices,anchovies tralbankspublisheitheranumericvalue,ora andforth.Anchovies?Fewtodaywillunderstand rangeofvalues,towhichtheycommitasaninfla- 1

MONETARYPOLICYANDINFLATION tionpolicyobjectiveoversometimehorizon.For expectationsinthemarket,theeconomyexperimostofthesecentralbanks,achievingtheinfla- encedasevererecessionin1981-82. tiontargetis,undernormalmarketconditions, Measuringinflationby12-monthchangesin thesinglepolicyobjective.Forthesepolicymak- theconsumerpriceindex(CPI),inflationfellfrom ers,theidea,embodiedintheoriginalPhillips apeakof13.3percentin1979to3.8percentin curveanalyses,thatitispossibletopermanently 1982,andremainedintheneighborhoodof4 trade-offalowerunemploymentrateorahigher percentintothemid1980s.Inflationroseabitin levelofrealoutputforahigherrateofinflation thelate1980s,andreached6.1percentin1990, hasbeenrelegatedtothetextbooksonthehistory inthefaceoftheoilshockthataccompanied ofeconomicthought. Iraq’sinvasionofKuwait. Othercentralbanks,includingtheFederal Intheearly1990s,theconsensuswasthat Reserve,havedeclinedtoquantifytheirinflation theeconomyhadnotreturnedtopricestability. objective.Nevertheless,most,ifnotallofthese Anysubstantialriskthatinflationmightrisewas institutionsacknowledgetheirprimaryrespon- clearlyundesirableandnecessitatedapolicy sibilitytoproducealowandstableinflation response.Butoverthecourseofthe1990s,inflaenvironment.InthecaseoftheFederalReserve, tionwasflattograduallyfalling,andwithfurther theFOMChasstatedasitspolicyobjectiveto declinesintheearlypartofthisdecademost achievepricestabilityinordertoachievemaxiobserversconcludedthatthebattleforpricestamumsustainableeconomicgrowth,therebyfulbilityhadbeenwon. fillingitsduallegislatedmandateunderthe Policydiscussionstodaymustconsidertwo- Humphrey-HawkinsAct. sidedoutcomes—theriskstopricestabilityare TheFOMChasneverdefined“pricestability” symmetric.Asignificantbreakoutofinflation numerically,butitscommitmenttopricestability aboveourcurrentsituationiscertainlynotdesirisnotindoubt.Ihaveindicatedonseveraloccaable—noonewantstothrowawaythehard-earned sionsthatmyowninflationtargetiszero,properly accomplishmentsofthepast25years.However, measured.Becauseallourmeasuresofinflation asignificantdeclineininflationfromcurrentrates havesomeupwardbias,mydefinitionofprice maynotbedesirableeither—noonewantsto stabilityisconsistentwithasmallpositivemeasreplicatethedeflationaryexperiencesofthe1930s uredrateofinflation.OurrecentinflationexpeorJapan’sexperienceinthe1990s.Theconsensus rienceis,Ibelieve,agoodapproximationofprice oftheFOMC,asreflectedinthemostrecentpress stability. release,isthattheupsideanddownsiderisksto ThepolicyproblemfacedbytheFOMC,and inflationarealmostbalanced. manyothercentralbankers,todayissignificantly Whentheprimarybattleagainstinflation differentfromthat25yearsago.Bythelate1970s, startedin1979,therewasastrongcaseforpaying theinflationrateintheUnitedStateshadbecome greatattentiontotherateofmoneygrowthasa unacceptablyhigh—bothfortheFOMCandfor measureofthethrustofmonetarypolicy.Money thepublicingeneral.Therewasdisagreementin theeconomicsprofessionastowhetherthecosts growthisnotirrelevanttoassessinginflationrisks ofdisinflationarymonetarypolicywereworth today,buttheemphasishaschanged.Foravariety bearing,buttheFOMCconcludedthatamore ofreasons,andespeciallybecauseexpectations disciplinedmonetarypolicywasnecessary.The oflowinflationaresoentrenchedinthemarkets, decisivedatewasOctober6,1979,whenthe short-runmoneygrowthisaninadequateindica- FOMCdecidedtoadoptthe“NewOperating torformonetarypolicypurposes.Whatweneed Procedures.”Subsequently,theFOMCpermitted todoinsteadistoextractasbestwecanevidence thefederalfundsratetoexceed20percent.With ofpossibleinflationarypressuresfromavariety thetightermonetarypolicyandchanginginflation ofothersourcesofinformation. 2

InflationSignalsandInflationNoise TRANSITORY AND PERMANENT ofClevelandpublishessuchanindex.1The medianCPI,byconstruction,willexcludeany CHANGES IN PRICES CPIcomponentpriceindexthatishighlyvolatile Howcanwebestreadthe“news”intheprice intheshortrun,asthiscomponentinflationrate datathatbecomeavailableeverymonth?The willtypicallyappearinoneofthetailsofthe problemistouncoverinformationthatmight cross-sectiondispersionoftheinflationratesof indicatethatahigherrateofsustainedinflation theCPIcomponents. couldbeathand.TheFOMCshouldrespondto ImustconfessthatIamuncomfortablewith thesignalandnottothenoiseinthedata.Separatarbitrarilydefinedfilters.Filteringoftransitory inginflationsignalsfrominflationnoiseisa shockswheretheshocksareunderstoodandcan seriouschallenge. beidentifiedisclearlyappropriate.Anexample Throughoutmostofthepost-WorldWarII istheincreaseintobaccopricesfollowingthe historyintheUnitedStates,inflationhasbeena legalsettlementwiththemajortobaccocompahighlypersistentprocess.Neverthelessthereisa niesseveralyearsago.Itwaswellunderstood lotofrandommonth-to-monthvariationinthe thatthiswasaone-timeincreaseintobaccoprices measuredpriceindexes.Itisthisnon-systematic thatwouldfinancethefundsestablishedaspart variationthatwecall“noise.”Somehow,policyofthesettlement.Itwaspossibletoestimatethe makershavetolookthroughthe“noise”todiscern impactofthosepricechangesontheoverallCPI thesignalabouttheunderlyingtrendininflation andmosteconomistsfilteredthisimpactoutin inordertoformulateappropriatepolicyactions. ordertoassesstheinflationsignalatthetime. Thissignalextractionproblemistherationale Absentspecificinformationonwhichto forthedevelopmentofanumberofsupplemenbaseestimatesof“inflationnoise,”caremustbe tarymeasuresofinflation,beyondthebroad-based takentoassurethatthetechniquesusedtofilter priceindexescollectedbythegovernmentstatistransitoryinflationarerobust.ThelateKarl ticalagencies.Inparticular,inadditiontotheCPI Brunnerusedtocriticizethe“uppertailtheory andpersonalconsumptionpriceindex,“core” ofinflation”—thatinflationiscausedbythe measuresofinflationthatexcludefoodandenergy increasesinthepricesofthoseitemsthathappricesareoftenhighlighted. penedtoberisingthefastestatanypointintime. InmonetarypolicyreportstoCongress,con- Ifthese“uppertail”ratesofinflationarefiltered gressionaltestimony,andpublicspeeches, outoftheinflationmeasure,thentherecanbe ChairmanGreenspanandothermembersofthe noinflation!Clearlyarbitraryfilteringcanbeused FOMChavefocusedonthecoremeasureofthe personalconsumptionpriceindex.Therationale todefineawaysubstantiveproblems.Ibelieve fortheconstructionofthecoremeasuresofinfla- thatadditionalresearchintodefiningand tionisnotthatfoodandenergyareunimportant extractinginflationtrendshasthepotentialto itemsofhouseholdconsumption.Thecoremeas- providevaluableinsightstomonetarypolicyurescametoprominenceinthe1970swhenfood makers. andenergypriceswereextremelyvolatile.Under ThoughtheevidencesuggeststhatU.S.inflathoseconditionsthecoreinflationmeasureslikely tioninthesecondhalfofthetwentiethcentury providedhelpfulfiltersthroughwhichtodiscern wasahighlypersistentprocess,itispossiblethat asignaloftheinflationtrend.Removingthese theobservedpersistencemayresultfromtheway componentsremovesasourceofshort-runnoise thatmonetarypolicywasimplemented.Unlike thatcanobscureunderlyingpricedevelopments. organizedassetmarkets,whereeconomictheories Anotherapproachtoidentifyingthesignalis suggestthatpricechangesshouldapproximate themedianCPIindex.TheFederalReserveBank arandomwalk,thereisnostrongtheoretical 1 Eachmonth’sreleasecanbefoundathttp://www.clevelandfed.org/research/data/mcpipr.htm. 3

MONETARYPOLICYANDINFLATION basisforhighlypersistentinflationrates.Recent theirrisingsteelcosts.Butmostothermanufacresearchsuggeststhatthepersistenceofinflation turersaretryingtopushsteel-pricejumpsofup intheU.S.hasdiminishedsincethemid1980s.2 to30percentto50percenttoothercompanies Economichistorianswhostudythegoldstandard alongthesupplychain,creatingtensionbetween periodconcludethatinflationinvariouscountries steelproducers,theirbiggestcustomersand wasmuchlesspersistentduringthatperiodthan numeroussmallersuppliersbetweenthem.”3 inthetwentiethcentury.Thelessonforthose Similarconcernsareoftenexpressedaboutthe tryingtoseparateinflationsignalsfrominflation depreciationofthedollaragainstthecurrencies noiseisthatthefiltersmaynotberobustto ofourmajortradingpartnersasasourceofCPI changesinmonetarypolicyregimes. inflationdirectlythroughthepriceofimported goodsorindirectlythroughinducedpricechanges ondomesticallyproducedgoodsthatcompete FORECASTING INFLATION withimports. Acriticalquestionforallcentralbankers,and Withoutquestion,individualfirmsoftendo theFOMCinparticularatthepresenttime,is passalongincreasesinpricesoftheirinputs. howinflationwillevolveinthenearfuture.Are Theissueiswhetherthisphenomenonisgeneral near-terminflationary(ordeflationary)pressures enoughtoexplainoverallinflation.Undoubtedly reallyquiescent?Theanswertothisquestion therearetimesduringwhichinflationshocks requiresaforecastingmodel,oralternativelya aretransmittedthroughsuchmechanisms.Howsetofleadingindicatorsofinflation. eversuchforcesarenotuniversal.Automobile manufacturers,facedwithweakdemandinrecent Supply Chain Theories of Inflation yearshaveeffectivelycutretailpricesthrough incentiveprograms,discounts,rebatesandzero Oneviewthatcurrentlyisreceivingconsidinterestfinancing.Inthisenvironment,theyhave erableattentioniswhatIwilllabelthe“supply demandedandobtainedsubstantialpriceconceschaintheoryofinflation.”Proponentsofthisview sionsfromtheirsuppliers.Inthiscase,downward characterizeinflationshocksasoriginatingin pricepressureshavebeentransmittedbackward rawmaterialsmarketsandsubsequentlytransthroughthesupplychainstartingfromretailmarmittedthroughintermediateproductstofinished ketsratherthanforwardfromcommoditymarkets. productsandfinallytoconsumerprices.From Backwardandforwardpricepressurescan thisperspective,inflationofcommodityprices isaleadingindicatorofPPIinflationwhichin existatthesametime.Theforwardpressures turnisaleadingindicatorofconsumerprice havebeenvisiblerecentlyinsomeindustries, inflation. butbackwardpressureshavebeenevidentfor Toproponentsofthesupplychainview,rapid severalyearsandseemlessnewsworthyatpresent. inflationinscrapsteelpricesandotherbasic Competitiveforcescanfrustrateeffortstopush commoditypricesinthepastsixmonthsisa increasedinputcostsupthesupplychain.The causeofsignificantconcern.Lastmonth,aWall WallStreetJournal,alsoinanarticlelastmonth, StreetJournalarticlereportedthatsomefirmsare reportedthatairlines,facingsubstantialincreases passingalongmaterialspriceincreases.“Indeed injetfuelprices“havemadeatleast12attempts ahandfulofcompanies—amongthemmakersof toboostairfaresinthepast2½monthsalone. mattressesandgymequipment—alreadyhasor Butmostoftheeffortshavefailedtostick,and arepreparingtoaskshopperstopaymoretocover increasinglythespoilerhasbeenoneormore 2 SeeT.CogleyandT.Sargent,“EvolvingPost-WorldWarIIU.S.InflationDynamics,”NBERMacroeconomicsAnnual,2001,pp.331-72;and A.LevinandJ.Piger,“IsInflationPersistenceIntrinsicinIndustrialCountries?”FederalReserveBankofSt.LouisWorkingPaper2002-023. 3 “CompaniesFightRisingSteelPrices,”WallStreetJournal,March8,2003,p.A3 4

InflationSignalsandInflationNoise budget-priceairlines,whichseeachanceinthe historyofCPIinflationitself.Thus,wesimply currentsqueezetoextendtheirmarketshares.”4 donotobservePPIinflationbeingpassedalong Inthiscasecompetitioninconsumermarketsis overtimeintotheCPItoanysignificantdegree. limitingthetransmissionofpricepressures Anotherapproachtotestingthesupply-chain throughthesupplychain. theoryofinflationistoexaminetherelationship Norareexchangeratedepreciationsneces- betweenvariousstagesofprocessinginthePPI. sarilyleadingindicatorsofpricechangesatthe Itturnsoutthatthelead/lagrelationshipforthe retaillevel.First,theU.S.remainsarelatively PPIstageofprocessingisevenweakerthanthat closedeconomywithalargefractionofthegoods forthePPIandCPI.Considerthestatisticalequaandservicesconsumedhereproduceddomesti- tionexplainingPPIinflationforfinishedgoods cally.Second,researchindicatesthatthe“pass- usingthehistoryoffinishedgoodsinflationand through”ofexchangeratefluctuationsisnot thecurrentandprevious12monthsinflationof instantaneous,completeorconstant.5 thePPIindexforintermediateproducts.Theinter- Thereareseveralwaystotesttheideathat mediateproductsinflationaddslessthan1perinflationatanearlierstageofprocessingfeedsinto centtothepredictivevalueoftheequation.Using inflationatalatterstage.Considerastatistical thesameapproachtoexplaintheinflationrate equationinwhichwetrytoexplainmonthlyCPI forthePPIforintermediateproducts,wefindthat inflationfromtheproducerpriceindex.The inflationinthePPIindexforcrudematerials consumerpriceindexisanindexofretailprices. addsonly4percenttothepredictivevalueof ThePPI,whichusedtobecalledthe“wholesale theequation. priceindex,”isanindexofpricesatanearlier Tosummarizethisdiscussion,althoughitmay stageofproduction.Theequationemploystwo seemlogicalthatincreasesincrudematerials groupsofexplanatoryvariables.Thefirstgroup prices,suchaspetroleum,wouldfeedforward consistsoftheprevious12monthsofCPIinflation. intosemifinishedgoodsandthenforwardagain WeincludetheCPIhistorybecausewewantto tofinishedgoods,infacttheinflationprocess determinethecontributionofthePPIoverand doesnotworkthisway.Dependingonconditions abovethecontributionoftheCPIhistoryitself. inindividualmarkets,sometimesinflationdoes Thesecondgroupconsistsofthecontemporanefeedforward,butsometimesitfeedsback.Wejust ousandprevious12monthsinflationofthePPI cannotreliablyconcludethattoday’smaterials priceindexforfinishedgoods. pricesinflationwillbetomorrow’sfinishedgoods Mostofthepredictivevalueisintheconteminflation. poraneousPPItermandthelaggedCPIinflation terms.ThefactthatthecontemporaneousCPI “Gap” Theories of Inflation andPPIinflationratesmovetogetherisaconsequenceofinflationshocksthataffectpricesatall Anotherpopularframeworkinwhichto stagesofprocessingatthesametime.6Thetwelve analyzethetransmissionofinflationisthe“gap” laggedPPIvariablesaccountforonly11percent model.Variousimplementationsofthismodel ofthevarianceintheCPIinflationnotattributa- arerootedintheexpectations-augmentedPhillips bletothecontemporaneousPPIinflationandthe curve.Atypicalempiricalrepresentationofthis 4 “GrowingHeftPutsBudgetAirlinesinthePilot’sSeat,”WallStreetJournal,March29,2004,p.A1. 5 SeeP.K.GoldbergandM.M.Knetter,“GoodsPricesandExchangeRates:WhatHaveWeLearned?”JournalofEconomicLiterature,1997, 35(3),pp.1243-72;andP.S.PollardandC.C.Coughlin,“SizeMatters:AsymmetricExchangeRatePass-ThroughattheIndustryLevel,” FederalReserveBankofSt.LouisWorkingPaper2003-029C. 6 ThesimplecorrelationbetweenmonthlypercentagechangesofthePPIcrudematerialspriceindexandthePPIintermediateproductsprice indexis0.48;betweenpercentagechangesinthePPIintermediateproductspriceindexandthePPIfinishedproductspriceindexis0.74 andbetweenpercentagechangesinthePPIfinishedproductspriceindexandtheCPIis0.64overtheperiodMarch1948throughDecember 2003. 5

MONETARYPOLICYANDINFLATION frameworkpostulatesthattheinflationrateis inginflation.9Thefocusoftheiranalysiswas determinedbyinflationexpectations,current forecastsofinflationonatwelve-monthhorizon. andlaggedvaluesofa“gapmeasure,”current Theystartedtheiranalysiswith“conventional andlagged“supplyshocks”andundetermined specificationsofthePhillipscurve”thatrelated residualfactors.Intheseequations,acommon thechangeintheinflationratetopastvaluesof specificationofsupplyshocksincludeschanges anunemploymentgapmeasure,pastchangesof intherelativepricesoffoodandenergy,changes inflationandcurrentandpastvaluesofvarious inrelativeimportprices,and“productivity measuresof“supplyshocks”—thetypeofspecishocks.”“Gaps”typicallyaremeasuredasthe ficationdiscussedabove. differencebetweenrealGDPandameasureof StockandWatsonreachedseveral “potentialGDP”suchasthatconstructedbythe conclusions: CongressionalBudgetOffice,orbythedeviation 1.Inout-of-sampleforecaststhevarioussupoftheunemploymentratefromanestimated plyshockmeasuresdidnotimprovethe equilibriumunemploymentrate.7 forecastingperformanceoftheirmodels; Thereareseveraltrickypartsofsuchanalyses; 2.whiletheestimatedrelationshipbetween oneisuncertaintyoverthelevelofpotentialGDP changesintheinflationrateandcurrent oroftheequilibriumunemploymentrate.Another andpastchangesininflationandunemisobtainingareliablemeasure,orproxy,forthe ploymentfailstatisticaltestsforstability, expectedrateofinflation.Tomeasureexpected ineconomictermstherelationshipis inflation,oneoftwoapproachesisfrequently robust; applied.Thefirstistorepresentexpectedinflation 3.alternativemeasuresofrealeconomic asaweightedaverageofpastobservedratesof activitygeneratemoreaccurateforecasts inflation.Thealternativeapproachistoembed thandoequationswiththeunemployment thegapequationwithinamodeloftheentireeconrate; omyandtoequateexpectationsoffutureinflation 4.theadditionofinterestratesandinterest withthemodelbasedforecastsofinflation.8 ratespreadsfailstoimprovetheforecasting RecentresearchattheFederalReserveBank performanceoftheestimatedmodel; ofSt.Louisshowsthatwithseveraldifferent 5.commoditypricesdonotimproveinflation modelsofexpectedinflation,includingthelagged forecasts;and inflationproxy,neitherthe“gap”termnorthe “supplyshock”termsaccountforthemajormove- 6.forecastsusingtheunemploymentrateoutmentsintherateofinflation.Theexpectedinfla- performsimplemodelsusingonlylagged tiontermtrumpstheotherfactorsasthemajor changesininflation,butthegaininforemovingforceintheU.S.inflationhistory. castingaccuracyisrelativelysmall.Their analysisiswiderangingcoveringatotalof 168economicindicators.Anoverallassessmentoftheirresultsisthatourabilityto FORECASTING INFLATION— deliversignificantlymoreaccurateforecasts EMPIRICAL EVIDENCE ofinflationbeyondthosethatcanbegen- In1999,JamesStockandMarkWatsonpub- eratedfromthehistoryofinflationitselfis lishedanexhaustivestudyofmodelsforforecast- quitelimited. 7 RobertJ.Gordonhasauthorednumerousstudiesapplyingthisframework. 8 R.Clarida,J.;GaliandM.Gertler,“TheScienceofMonetaryPolicy:ANewKeynesianPerspective,”JournalofEconomicLiterature, December1999,37(4),pp.1661-704. 9 J.H.StockandM.W.Watson,“ForecastingInflation,”JournalofMonetaryEconomics,1999,44,pp.293-335. 6

InflationSignalsandInflationNoise ANCHORING INFLATION inassetmarketswhen,forreasonsthatarenot wellunderstood,“bubbles”develop.Reality EXPECTATIONS chasesexpectations,whichinturnchasereality. Howisamonetarypolicymakertointerpret Eventsaredeterminedby“inflationarypsycholotheaboveconclusions?Onepossiblereaction gies.”Pricesfollowexplosivepaths,eitherupward wouldbedespair—itisclosetoimpossibleto ordownward,foratime. discernthatnearfutureinflationwilldiffersys- Withinafewweeksofthedramaticchange tematicallyfromrecentpastinflationandsociety inpolicydirectionin1979,ChairmanVolcker willjusthavetolive,atleastforawhile,with testifiedbeforeCongressontheFOMC’snew recenthistory. operatingprocedures.Marketexpectationsplayed Ihaveanoptimisticalternativeinterpretation aprominentroleinhisthinking.“Theclearand oftheavailabledataandresearchresults.My presentdangerwasthatfailuretodealwithinflaconclusionisthattheunfoldinginflationary tionandinflationaryexpectationswouldintime experienceismoststronglyanchoredbyhowthe producemore—notless—economicinstability, publicandfinancialmarketparticipantsexpect ultimatelywithhigherpricesandgreateruneminflationtoevolve.Welldesignedandarticulated ployment.Inthatsetting,thepriorityforpolicy policyundersuchconditionscanproducegreat wasdecisiveactiontodealwithinflationary outcomes.However,badlydesignedpolicies pressuresandtodefusethedangerousexpectaunderthesameconditionscanproducedisasters! tionalforcesthatwerejeopardizingtheorderly Ifmycharacterizationoftheimportanceof functionoffinancialandcommoditymarkets.”10 inflationaryexpectationsasadeterminantof inflationiscorrect,thenthereareimportantlessonsformonetarypolicymakers.Inaneconomy CONCLUDING COMMENT thatworksthisway,itisessentialthatthecentral TheFOMCisunavoidablyinasituationof bankclearlyarticulateitsinflationobjectives,and havingtoapplyitsbestjudgmenttoavarietyof havethismessageregardedashighlycredible.If economicindicatorsofpossibleinflationarypresacentralbankiscommittedtoalow-inflation sure.I’venotdiscussedtodayarangeofinformaenvironment,andthatcommitmentiscredible, tiononinflationpressures,whichincludethe thenthegeneralpublicwillbelievefutureinfla- rateofproductivitygrowth,therateofincrease tionwillbelow.Underthesecircumstancesthe inunitlaborcosts,measuresofwageinflation economyhasastrongexternalnominalanchor. anecdotalreportsfrombusinessfirmsontheinfla- Thatnominalanchorgeneratesbehaviorbybuyers tionenvironmenttheysee.Instead,IhopeI’ve andsellersthatproduceslowandstablerealized convincedyouthatthereisnoregularandreliinflation.Inrecentvernacular,atanygiventime ablerelationshipbetweeninflationinmaterials fewfirmsintheeconomyhaveany“pricing pricesorgoodsatanearlystageofprocessing power.” andretailpriceinflation. Inthistypeofeconomy,ifthecentralbank Ofcriticalimportancetomaintaininglow failstoarticulateaninflationobjective,orifit andstableinflationistheFOMC’scommitment lackscredibilitywiththepublicthatthestated toactaggressivelywheninflationriskschange, objectivewillbepursued,thenthedownsiderisk eitherupordown.Thatcommitmentanchors totheeconomyisenormous.Environmentsin marketexpectationsoflong-runinflation,and whichexpectationsarenotanchoredexternally makestheeconomymorerobusttoshort-run areinherentlyunstable.Weobservethisbehavior inflationshocks.Short-rundisturbancesdonot 10StatementbyPaulA.VolckerbeforetheSubcommitteesonDomesticMonetaryPolicyandonInternationalTrade,InvestmentandMonetary PolicyoftheCommitteeonBanking,FinanceandUrbanAffairs,U.S.HouseofRepresentatives,Nov.13,1979,FederalReserveBulletin, December,1979,p.959. 7

MONETARYPOLICYANDINFLATION automaticallygetbuiltintoinflation,whichhelps todampentheimpactsontheeconomyofinflationshocks. ThisstableenvironmentalsohelpstheFOMC toavoidmistakes.Aboveall,wedonotwantto respondtoinflationnoise,whichwouldaddfurtherinstabilitytotheeconomy.Extractingthe inflationsignal,andrespondingtoit,iswhatwe trytodo. 8