美聯(lián)儲(chǔ)副主席 Jefferson：歷史回顧中的美國(guó)當(dāng)前貨幣政策周期 Is This Time Different Recent Monetary Policy Cycles in Retrospect

Forreleaseondelivery

10:00a.m.EST

February22,2024

IsThisTimeDifferent?RecentMonetaryPolicyCyclesinRetrospect

Remarksby

PhilipN.Jefferson

ViceChair

BoardofGovernorsoftheFederalReserveSystem

atthe

PetersonInstituteforInternationalEconomics

Washington,D.C.

February22,2024

Thankyou,Adam,andthankyoutoPetersonfortheopportunitytospeaktoyou

today.

BeforeIbegin,letmeremindyouthattheviewsIwillexpresstodayaremyown

andarenotnecessarilythoseofmycolleaguesintheFederalReserveSystem.

IwilltakethisopportunitytosharewithyoumyoutlookontheU.S.economy

andsomeupsideanddownsideriskstowhichIampayingspecialattention.Also,Iwillreviewpastmonetarypolicycyclesanddiscusswhatlessonswemaylearnfromthem.

Withthat,letmeturntomyoutlookfortheU.S.economy.

AggregateEconomicActivity

Growthinrealgrossdomesticproductin2023cameinmuchhigherthan

expectedbymostprofessionalforecasters,buoyedbystrengthinconsumerspending.

Towardtheendof2023,however,householdbalancesheetsbegantoweaken,as

indicatedbyhigherdelinquencyratesandafurtherdeclineinsavings.These

developmentsleadmetoexpectslowergrowthinspendingandoutputin2024.Evenso,

withoutaclearunderstandingofwhyconsumerspendinghasbeensoresilient,Isee

continuingstrengthinspendingasanimportantupsiderisktomyforecast.Onepossible

explanationisthatconsumersdonotwanttogiveuppreviouslevelsofconsumption,

perhapsbecauseofhabitformationasdescribedbyRobertPollak(1970)andan

optimisticviewoffutureincomeprospects.Anotherpossibilityistheoneraisedin

pioneeringworkbyJamesDuesenberry(1949)75yearsagoandlaterdevelopedinthe

contextofmodernmacroeconomicsbyJordiGali(1994).Sociallymotivated

consumption—or“keepingupwiththeJoneses”—couldcauseindividualstoconsume

morethanwhatispredictedbymodelsthatonlyconsiderhouseholdwealthandincome.

-2-

TheLaborMarket

Theimbalancebetweenlabordemandandlaborsupplyhasnarrowed,aslabor

demandhascooledwhilelaborsupplyhasimproved.Thereisevidenceofcoolinglabordemand,suchasthedeclineinjobopeningsby3millionfromtheirpeakinMarch2022.

Nevertheless,thelabormarketremainstightandjobopeningsremainabout20percent

abovetheirpre-pandemiclevel,asshowninfigure1.Atthesametime,layoffshave

remainedverylow,andthepaceofpayrollemploymentgainsremainsstrong,with

nonfarmpayrollmonthlyjobgainsinthepastthreemonthsaveraging289,000.The

unemploymentrateinJanuarywas3.7percent,alevelthatisstillnearhistoricallows.

ThefactthattheunemploymentrateandlayoffshaveremainedlowintheU.S.economyoverthepastyearamiddisinflationsuggeststhatthereisapathtorestoringpricestability

withoutthekindofsubstantialincreaseinunemploymentthathasoftenaccompanied

significanttighteningcycles.

TheInflationOutlook

Inflationmadeclearprogressover2023towardtheFederalOpenMarket

Committee’s(FOMC)2percentinflationobjective.Ibelievethatthisprogressreflectsboththeunwindingofpandemic-relatedsupplyanddemanddistortionsintheeconomy

aswellasrestrictivemonetarypolicy,whichhascooledstrongdemandandgiventhe

supplysideoftheeconomytimetocatchup.Asshowninfigure2,overthe12months

endedinJanuary,theFederalReserve’sstaffestimatesthattotalpersonalconsumption

expenditures(PCE)pricesrose2.4percent,downfrom5.5percentoverthepreceding12

months.CorePCEprices,whichexcludesenergyandfoodprices,rose2.8percent,

downfrom4.9percent.ThefiguresforJanuaryareestimatesthatincorporatethe

-3-

somewhatlargerconsumerpriceindex(CPI)increasewesawlastmonth.That

disappointingCPIreadinghighlightsthatthedisinflationprocessislikelytobebumpy.

TheJanuarydatanotwithstanding,theslowingincoreinflationhasbeenespecially

pronouncedinrecentmonths,asthe3-and6-monthchangesincorePCEpricesthroughJanuary,at2.5percentand2.4percent,respectively,clearlyremainbelowthe12-month

changeshowninfigure2.Themoststrikingmoderationhasbeenincoregoodsprices,

asshowninfigure3,whichhavedeclinedoutrightoverthepastyear.Inflationincore

services,bothinitshousingcomponentandnonhousingservices,hasalsoslowed,butnot

asmuch.Ibelievethatasthelabormarketcontinuestocool,coreservicesprice

increaseswillcontinuetomoderate.Ofcourse,Iremainattentivetootherpossibilities.

ALonger-TermPerspectiveonMonetaryPolicyCycles

Next,Iwouldliketohighlightkeyaspectsofpastmonetarypolicycyclesand

whatlessonswemaylearnfromthesepastexperiences.Forexpositionpurposes,my

reviewfocusesoneasingcyclesandtheirprecedingpeak-rateepisodesthatextendback

to1989;however,Ialsowillmakesomecommentsonanimportantepisodepriorto

1989.1AsofourlastmeetinginJanuary,mycolleaguesontheFOMCandIbelievethat

ourpolicyrateislikelyatitspeakforthistighteningcycleandthat,iftheeconomy

evolvesbroadlyasexpected,itwilllikelybeappropriatetobegindialingbackpolicyrestraintatsomepointthisyear.Iwillthereforestartwithadescriptionofeconomic

conditionsduringpastpeak-rateepisodes,givenwherewearetoday.

1Thestartdateof1989ismotivatedbythefactthatthiswastheearliestcycleinwhichtheFOMCwas

viewedasconsideringmonetarypolicyactionsintermsofdiscrete25basispoint,50basispoint,etc.,ratemovesinthefederalfundsratetargetand,assuch,ismorecomparabletotoday.SeeStigumandCrescenzi(2007)foradiscussionoftheFOMC’sincreasedfocusonfederalfundsratetargetinginthelate1980sas

wellasLindsey(2003),whodescribestheFOMC’sfurthershifttowardtargetingthefederalfundsratein1989bydiscontinuingthepracticeoftargetingborrowedreserves.

-4-

Thefirstcolumnoftable1liststhedatesofthelatestsixpeak-rateepisodes

precedingeasingcycles,definedasasequenceofratecutswithoutratehikesinbetween.

Thesecondandthirdcolumnssummarizeeconomicconditionsasdescribedinofficial

accountsofmonetarypolicyreleasedatthetime,suchasFOMCstatements,minutesof

Committeemeetings,andtheFederalReserveBoard’ssemiannualMonetaryPolicy

Report.Thetablealsorecordsinflationatthetimeofthepeakrateasmeasuredbythe

12-monthpercentchangeinheadlineandcorePCEpriceindexes.2

Mymaininferenceisthatmostofthetime,infiveoutofthesixepisodes,the

peakrateisreachedonceinflationiscontained,albeitinsomecaseswithrisksstill

present.Thereisonlyoneexceptioninthissample,theMarch1989toMay1989peak-rateepisode,wheninflationwaselevated—notedinthefirstrowofthetable.Theeasing

cyclefollowingthisparticularpeak-rateepisodebeganascoreandheadlinePCE

inflationwerestartingtocomedownfromanelevatedlevel,asillustratedinfigure4.3

Thisfigureshowsheadlineandcoreinflationtwoyearsbeforeandafterthestartdateofeachofthesixeasingcycles.Theredline,whichcorrespondstotheMarch1989peak-

rateepisode,standsout,withcorePCEinflationat4percentatthebeginningofthe

easingcycle,whilealltheothereasingcyclesshowcorePCEinflationatabout2percentatcommencement.Oursituationtodayisclosertothenormduringtheseepisodesthan

totheexception,asPCEinflationiscloserto2percentthanto4percent.

2ThePCEinflationnumbersintable1arereviseddata,whichmeansthedatathatpolicymakerswere

reviewingatthetime—theso-calledreal-timedata—couldhavebeendifferent.Inaddition,forsomeof

thesecycles,policymakersfocusedonCPIinflationmorethanPCEinflation.Also,notethatwhilethe

tablefrequentlyreferencesinflationrelativetoa2percentrate,itwasonlyforthelastpeak-rateepisodeforwhichtheFOMChadestablisheda2percentrateofinflationtobemostconsistentoverthelongerrunwithitsprice-stabilitygoal,peritsfirst“StatementonLonger-runGoalsandMonetaryPolicyStrategy”adoptedinJanuary2012.

3Thedatashowninfigure4arereviseddata.

-5-

Table2summarizesthereasonsgiveninFederalReservedocumentsatthetime

toexplaintherationaleforeasingpolicy.Whenstudyingmonetarypolicycycles,itis

importanttorecognizetheoften-multistagenatureofcyclesand,becauseofthis,table2

distinguishesbetweenthereasonsunderpinningthestartoftheeasingcycle,listedin

column2,andthoseunderpinningsubsequenteasings,listedincolumn3.Oneclear

exampleofacyclewithdifferentphasesandinvolvingmorethanonereasonforeasingis

themostrecenteasingcycle—theonelistedinthelastrow—whichtookplacebetween

August2019andMarch2020.Theinitial75basispointsofeasinginthiscyclewerearesultofdownsideriskstotheU.S.economyduetoweakerglobalgrowthandhightrade

uncertainties.Thesubsequenteasingsinthiscyclewereduetothedisruptionstothe

economyresultingfromtheCOVID-19pandemic.

Lookingattable2,twofactsstandout.First,mosteasingcyclesstartbecauseofconcernaboutslowingeconomicgrowth.Intable2,theoneexceptionistheeasingcycle

thatstartedinJuly1995andisassociatedwithwhatAlanBlinder(2023)haslabeleda

“perfectsoftlanding”example.4Thatparticulareasingcyclestartedpredominantly

becauseofreducedinflationconcerns.Alltheothereasingcyclesstartedbecauseeithertherewasaconcernaboutslowingeconomicgrowth,or,inonecase,becausetherewasa

concernaboutslowingeconomicgrowthandtherewerereducedinflationconcerns.

4Blinder(2023)labelstheJuly1995episodeasthe“perfectsoftlanding”andidentifiesother“softish”

landings(seehisTable1,page119).Blinder(2023)definesasoftishlandingasanoutcomeinwhichrealGDPdeclinesbylessthan1percentorthereisnoNBERrecessionforatleastayearafteranFOMC

tighteningcycle.

-6-

Thesecondfactthatstandsoutisthathistoryisrepletewitheventsthat

complicatemonetarypolicydecisions.Thethirdcolumnintable2,whichnotesthe

reasonsforsubsequenteasings,demonstratesthispoint.Itshowsthatfouroutofthesixeasingcycleshadmultiple“easingphases,”withlaterphasestriggeredbyeventslikethe1991GulfWar,the9/11terroristattack,theGlobalFinancialCrisis,andthepandemic.Theseeventsrequiredpolicymakerstotakeadifferentcourseofpolicyeasingfromthecoursetheymayhaveanticipatedearlierinthecycle.Specifically,becausetheseeventscontributedtothecontractionofeconomicactivity,policymakersmayhaveacceleratedpolicyeasing.ThemainmessagesthatIseeemergingfromthisreviewoftherecordarethatpolicymakersneedtoremainvigilantandnimble,incaseofadverseshockshitting

theeconomy,andthatpolicymakersneedsomegoodluck.

Thelessonthatpolicymakersneedtoremainvigilantandnimbleisfurther

illustratedinfigure5,whichshowstheunemploymentratearoundthestartofeach

easingcycle.Ascanbeseenfromthischart,insomeeasingcycles—forexample,the

easingcyclesthatbeganinJanuary2001,showninblue,andSeptember2007,showningreen—theunemploymentraterampsupquickly,shortlyaftertheeasingcyclebegan.In

boththesecases,theeconomyweakenedrapidly.

IntheeasingcyclethatbeganinJanuary2001,moderatinggrowthoverthe

secondhalfof2000gavewayabruptlytosluggishgrowtharoundtheendoftheyear.

Economicweaknessspreadandintensifiedoverthefirsthalfof2001and—asshownbytheblueline—ayearaftertheeasingcyclebegan,theunemploymentratehadincreased

justshortof2percentagepoints.

-7-

IntheeasingcyclethatbeganinSeptember2007,themacroeconomicdatawerenotshowingmuchweakeningatthetimeofthecycle’sfirstcoupleofratecuts,although

financialmarketswereexhibitingheightenedandbroad-basedvolatilityandshort-term

fundingmarketsweresignificantlyimpaired.ItwasonlyinDecember2007that

incomingdatastartedtoshowmoresignificantspilloversofthehousingdownturnto

otherpartsoftheeconomy,whileseveralfinancialfirmsalsobegantoreportlarger-than-expectedlosses.Asthegreenlineinthechartshows,theunemploymentratewasaround

4-1/2percentatthestartofthe2007–08easingcycle—havingremainedbroadlystable

aroundthatlevelin2006—butthenrapidlyroseto6percentwithinayearofthefirsteasing.Mymotivationfordiscussingthesetwoepisodesistohighlighthowquickly

economicactivitycanweaken.

Anotherreasonwhypolicymakersneedtowatchallavailableinformationandbe

nimbleintheirdecisionmakingisthatdevelopmentsconcerninginflationcanlikewise

changerapidly.ThiswashighlightedrecentlybyRussia’sinvasionofUkraineinMarch

2022.Theinvasioncompoundedtheeffectsofpost-pandemicsupplyconstraintson

inflation.Inaddition,wealwaysneedtokeepinmindthedangerofeasingtoomuchinresponsetoimprovementsintheinflationpicture.Excessiveeasingcanleadtoastalling

orreversalinprogressinrestoringpricestability.FormerFedChairPaulVolcker

stressedthisdangerina1981speech,whenhepointedto1967asayearwhenmonetary

policyeasedinresponsetoconcernsaboutslowingeconomicgrowthandreduced

inflationconcerns,yetinflationsubsequentlyturnedbackup.

Finally,anotherobservationfromreviewingpastepisodesisthatcarefuleasingin

theJuly1995easingcycleallowedtheFOMCtoassessincomingdataandother

-8-

informationtomakesurethatinflationwasundercontrol.AsInotedearlier,theJuly

1995easingcycleisassociatedwiththeso-calledperfectsoftlanding.Inthisparticulareasingcycle,theFOMCstartedtoeaseasitobservedalesseningininflationconcerns,

leftratesunchangedforthreemeetingsasitwaitedformoreinformation,andthen

continuedtoease.

LessonsforCurrentMonetaryPolicy

Withtheknowledgeofpastexperiencesinhand,letmesayafewwordsaboutthecurrentmonetarypolicycycleandtheextenttowhichfuturepolicyislikelytoresemble,

ifatall,pastexperiences.BetweenMarch2022andJuly2023,theFOMCraisedthe

targetrangeforthefederalfundsrate5-1/4percentagepoints.Ourstrongactionshave

movedourpolicyratewellintorestrictiveterritory,andourrestrictivestanceofmonetarypolicyisputtingdownwardpressureoneconomicactivityandinflation.Iftheeconomyevolvesbroadlyasexpected,itwilllikelybeappropriatetobegindialingbackourpolicy

restraintlaterthisyear.

Gettingbacktothetitleofmytalktoday:Willthistimebedifferent?Myanswer

is,ofcourseitwill.Everytimeisdifferent.Butwecanstilllearnfrompastepisodes.

Wecannotknowiftherewillbeunanticipatedexogenousshocksthatrequireapolicy

responsedifferentfromwhatwillbeenvisagedatthebeginningoftheeasingcycle.Allwecandoisassesstherisksasbestwecan,giventheavailableinformationandourbest

forecasts.Intheabsenceofunanticipatedexogenousshocks,policymakerscanweigh

multiplefactors,includingkeepingpolicyrestrictiveenoughtotampdownapossible

resurgenceofinflationduetothestrengthofaggregatedemandoreasingsoonertoavoid

anundueincreaseinunemployment.Unfortunately,thehistorythatIhavereviewed

-9-

todaysuggeststhatweshouldnotbesurprisedifsomekindofunanticipatedshock

occurs.Giventhatwemustacceptthatuncertaintyispresent,weconsidertherisksthat

canaffectouroutlookandforecasts.

Lookingahead,Iseeatleastthreekeyrisks.First,asImentionedatthe

beginning,consumerspendingcouldbeevenmoreresilientthanIcurrentlyexpectittobe,whichcouldcauseprogressoninflationtostall.Second,employmentcouldweakenasthefactorssupportingeconomicgrowthfade.Third,geopoliticalriskscouldremainelevated,andawideningoftheconflictintheMiddleEastcouldhavegreatereffectson

commodityprices,suchasoil,andonglobalfinancialmarkets.

Iremaincautiouslyoptimisticaboutourprogressoninflation,andIwillbe

reviewingthetotalityofincomingdatainassessingtheeconomicoutlookandtherisks

surroundingtheoutlookandinjudgingtheappropriatefuturecourseofmonetarypolicy.

-10-

References

Blinder,AlanS.(2023).“Landings,SoftandHard:TheFederalReserve,1965–2022,”JournalofEconomicPerspectives,vol.37(Winter),pp.101–20.

Duesenberry,JamesS.(1949).Income,Saving,andtheTheoryofConsumptionBehavior.Cambridge,Mass.:HarvardUniversityPress.

Galí,Jordi(1994).“KeepingupwiththeJoneses:ConsumptionExternalities,PortfolioChoice,andAssetPrices,”JournalofMoney,CreditandBanking,vol.26

(February),pp.1–8.

Lindsey,David(2003).“AModernHistoryofFOMCCommunication:1975–2002,”memorandum,BoardofGovernorsoftheFederalReserveSystem,Divisionof

MonetaryAffairs,June24,

/monetarypolicy/files/FOMC20030624memo01.p

df.

Pollak,Robert,A.(1970).“HabitFormationandDynamicDemandFunctions,”JournalofPoliticalEconomy,vol.78(4,part1),pp.745–63.

Stigum,Marcia,andAnthonyCrescenzi(2007).Stigum'sMoneyMarket,4thed.McGraw-Hill.

Volcker,Paul(1981).“DealingwithInflation:ObstaclesandOpportunities,”

speechdeliveredattheAlfredM.LandonLectureSeriesonPublicIssues,KansasState

University,Manhattan,Kansas,April15,

/title/451/item/8236.

IsThisTimeDifferent?

RecentMonetaryPolicy

CyclesinRetrospect

PhilipN.Jefferson

ViceChair,FederalReserveBoard

PetersonInstituteforInternationalEconomics,February22,2024

Disclaimer:TheviewsIwillexpresstodayaremyownandnotnecessarilythoseoftheFederalOpenMarketCommittee(FOMC)ortheFederalReserveSystem.

2

RoadmapofTalk

·OutlookontheU.S.economyandrisks·Reviewofpastmonetarypolicycycles

·Lessonsforcurrentmonetarypolicy

3

Figure1:JobOpeningCountsIndex

2019=100

Weekly

20192020202120222023

Note:JOLTSjobopeningcountsattheendofeachweekdividedbytheaveragejobopeningcountsintheyear2019.ThegrayshadedbarindicatesaperiodofbusinessrecessionasdefinedbytheNationalBureauofEconomicResearch.TheshadedrecessionperiodextendsfromFebruary2020throughApril2020.

Source:BureauofLaborStatistics,JobOpeningsandLaborTurnoverSurvey(JOLTS).

4

Figure2:PCEandCorePCEInflation

10

Note:Percentchangeinthepersonalconsumptionexpenditures(PCE)priceindexfrom12monthsago.Corereferstothepriceindexexcludingfoodandenergy.ThefigureplotsdatafromtheBureauofEconomicAnalysis;dataforJanuary2024areFederalReserveBoardstaffestimatesbasedonavailableconsumerpriceindexandproducerpriceindexdata.ThegrayshadedbarindicatesaperiodofbusinessrecessionasdefinedbytheNationalBureauofEconomicResearch.TheshadedperiodextendsfromFebruary2020throughApril2020.

Source:BureauofEconomicAnalysis,PCEpriceindex;BoardofGovernorsoftheFederalReserveSystem.

5

Figure3:SelectComponentsofCorePCEInflation

9

8

7

6

5

4

3

2

1

0

-1

-2

Note:Percentchangeinselectcategoriesofthepersonalconsumptionexpenditures(PCE)priceindexfrom12monthsago.ThefigureplotsdatafromtheBureauofEconomicAnalysis;dataforJanuary2024areFederalReserveBoardstaffestimatesbasedonavailableconsumerpriceindexandproducerpriceindexdata.ThegrayshadedbarindicatesaperiodofbusinessrecessionasdefinedbytheNationalBureauofEconomic

Research.TheshadedperiodextendsfromFebruary2020throughApril2020.

Source:BureauofEconomicAnalysis,PCEpriceindex;BoardofGovernorsoftheFederalReserveSystem.

Table1.Summaryofpastsixpolicyratepeakepisodes

Startandenddateofpolicyratepeaks

precedingeasingcycle

EconomicGrowthBackdrop

InflationBackdrop

Mar.1989-May1989

(3months)

Moderatingandlaterslowinggrowth

Elevatedinflationandtightlabor

markets:

PCEinflation4.7%

CorePCEinflation4.6%

Mar.1995-June1995

(4months)

Moderatinggrowth

Inflationcloseto2%andtightlabor

markets:

PCEinflation2.17%

CorePCEinflation2.20%

Apr.1997-Aug.1998

(17months)

Strongdomesticgrowth,butrisksfromstrainsinemergingmarketeconomiesandassociated

financialturmoil

Inflationcloseto2%andtightlabor

markets:

PCEinflation1.9%

CorePCEinflation2%

June2000-Dec.2000

(7months)

Moderatinggrowth

Inflationcloseto2.5%andtightlabor

markets:

PCEinflation2.7%

CorePCEinflation1.7%

July2006-Aug.2007

(14months)

Moderategrowthdespitehousingsectorweakness

Inflationmoderatingandtightlabor

markets:

PCEinflation3.4%

CorePCEinflation2.5%

Jan.2019-July2019

(7months)

Solidgrowth,butrisingdownsiderisksfromweakerglobalgrowthandtradeuncertainty

Inflationmuted:

PCEinflation1.5%

CorePCEinflation1.8%

Note:Thesecondandthirdcolumnssummarizeeconomicconditionsasdescribedinofficialaccountsofmonetarypolicyreleasedatthetime,such

asFOMCstatements,minutesofCommitteemeetings,andtheFederalReserveBoard'ssemiannualMonetaryPolicyReport.ThePCEinflation6

numbersarereviseddata.

7

1989-92…1998··2007-08

1995-962001-03--2019-20

Figure4.InflationaroundtheStartofEasingCycles

CorePCEInflation(12monthper