彼得森經(jīng)濟(jì)研究所-是什么導(dǎo)致了美國疫情時(shí)期的通貨膨脹？（英）

WORKING

PAPER23-4

What

caused

the

US

pandemic-erainflation?BenBernankeandOlivierBlanchardJune2023ABSTRACTWe

answerthequestionposedby

thetitleby

specifyingandestimatingasimpledynamicmodelofprices,wages,andshort-runandlong-runinflationexpectations.Theestimatedmodelallowsusto

analyzethedirectandindirecteffectsofproduct-marketandlabor-marketshocksonpricesandnominalwagesandto

quantifythesourcesofUSpandemic-erainflationandwage

growth.

We

findthat,contraryto

earlyconcernsthatinflationwouldbespurredby

overheatedlabormarkets,mostoftheinflationsurgethatbeganin2021wastheresultofshocksto

pricesgivenwages.Theseshocksincludedsharpincreasesincommodityprices,reflectingstrongaggregatedemand,andsectoralpricespikes,resultingfromchangesinthelevelandsectoralcompositionofdemandtogetherwithconstraintsonsectoralsupply.

However,

althoughtightlabormarketshave

thusfarnotbeentheprimarydriverofinflation,we

findthattheeffectsofoverheatedlabormarketsonnominalwage

growthandinflationare

morepersistentthantheeffectsofproduct-marketshocks.Controllinginflationwillthusultimatelyrequireachievingabetterbalancebetweenlabordemandandlaborsupply.BenBernankeisDistinguishedSeniorFellowattheHutchinsCenteronFiscalandMonetaryPolicyattheBrookingsInstitution.OlivierBlanchardistheC.FredBergstenSeniorFellowatthePetersonInstituteforInternationalEconomics.JELCodes:

E31,

E37,

E52,

E6,E24Keywords:

inflation,monetarypolicy,

aggregate

demand,

Beveridge

curve,commodity

prices,

shortages,

inflation

expectationsThe

BrookingsInstitutionisfinancedthrough

thesupportofa

diverse

array

offoundations,corporations,

governments,individuals,aswell

asanendowment.A

listofdonorscanbefoundintheirannualreports

publishedonlinehere.The

findings,interpretations,andconclusionsinthisreportare

solelythoseofitsauthor(s)andare

notinfluencedbyany

donation.Note:Thispaperwaspreparedforaconference

on

the

recent

inflationheldattheHutchinsCenteronFiscalandMonetaryPolicyattheBrookingsInstitutiononMay23,2023.IthasbeenpublishedasHutchinsCenterWorkingPaper#86here.TheauthorsthankDavidReifschneiderandElaineBuckbergforhelpfulinformation,andPeterDiamond,JasonFurman,JosephGagnon,PatrickHonohan,EgorGornostay,

DanielLeigh,MauriceObstfeld,LawrenceH.Summers,RichardClarida,ángelUbide,DavidVines,andDavidWilcox

fordiscussions.Excellentresearchassistancewasprovidedby

JamesLeeandAthianaTettaravou.

TheauthorsthanktheHutchinsCenter,

theBrookingsInstitution,andthePetersonInstituteforInternationalEconomics(PIIE)forsupport.Thispaperhasbeensubjectedto

aPIIEpre-publicationdatareplicabilityreview.?

2023

The

Brookings

Institution.Allrightsreserved.IntroductionCentralbankersand

mostoutsideeconomistsfailedto

predictthesharprisein

inflationthatbegan

in2021,andpolicymakers,bothintheUnitedStatesandinotheradvancedeconomies,wereaccordinglyslowtoreact.Actionwas

furtherdelayedbytheviewthattheinflationburstwouldbetemporary.Misjudgmentsoftheeconomiceffectsofcovid-erafiscalprogramsareonepotentialexplanationofthefailuretoforecastthesubsequentinflation.FederalOpenMarketCommittee(FOMC)forecastsofinflationchangedlittlebetweentheSeptember2020andJune2021iterationsoftheSurvey

ofEconomicProjections,andtheCommitteecontinuedtoexpectthatinflationwouldreturnclosetotheFed’s2percenttargetby2023,despitecongressionalpassageofabillinDecember2020thatincluded$900billionforcovidreliefandthe$1.9trillionAmericanRescuePlansignedbyPresidentBidenin

March2021.Thesetwoprograms

cameontop

ofthe$2.2

trillionCARESAct,passedinMarch2020andsignedbyPresidentTrump,whichhadalreadystrengthenedfirms’andhouseholds’balancesheetsand

increasedfutureabilitytospend.Overall,asashareofGDP,theheadlinecostsofthesethreecovid-erafiscalpackageswereabout4-1/2timesthesizeof

theAmericanRecoveryandReinvestmentAct(ARRA),enactedinresponseto

the2008financialcrisisandtheensuingrecession.1Accordingtoconventional

economictheory,expansionaryfiscalpoliciescanstokeinflationiftheycauselabormarketstobecomeoverheatedandoutputtoexceedtheeconomy’s

potential.

Indeed,someearlyanalysesoftheAmericanRescuePlanusingstandardfiscalmultipliersconcludedthattheadditionalfederalspendingwouldindeedoverheattheeconomy,possiblyleadingtohigherinflation(Blanchard,2021;

EdelbergandSheiner,2021).Thesestudies

acknowledgeduncertaintyaboutmagnitudes,asthefiscalimpactsonemployment,output,and

inflationwoulddependonmanyfactors,includingtheeffectsofthepandemiconpotentialoutput,thesizeoftheinitialoutputgap,households’propensitiestoconsumeoutoftheone-timechecksauthorizedbythebill,

andtherelationbetweenlabormarkettightness

andinflation.Inflationoptimists(ReifschneiderandWilcox

2022)arguedthat,evenifthenewfiscalspendingweretodrivedownunemploymentbymorethan

expected,a

largeburstofinflationwas

still

unlikely,asmanystudieshadconcludedthat

thePhillipscurveis

quiteflat(thatis,

inflationisrelativelyinsensitivetolabormarkettightness),and—inlightofmorethanthreedecadesoflowinflationin

theUnitedStates—inflationexpectations

werelikelytoremain

wellanchored.Incontrast,someeconomists(includingoneoftheauthors)arguedthatwageinflation,andconsequently

price

inflation,couldrise

muchmorethanpredictedbyconventionalcalculationspredicatedonaflatPhillipscurve(Summers,2021;

Blanchard,2021).Theirconcernswerethattheincreaseinaggregatedemandlikelytoresultfromtheunprecedentedlylargefiscaltransfers,togetherwiththecumulativeeffectsoftheeasing

ofmonetary

policybeguninMarch2020,could

causemoreoverheatingofanalready-tightlabormarketthan

theoptimistsexpected.AnextremelylowunemploymentratemightinturncausethePhillipscurvetosteepen.Moreover,higherand,consequently,morepsychologically...1.

Theheadlinecost

of

theCARESActandthe

December

supplementalappropriationspackagetogethertotaledabout15percent

ofgross

domestic

product(GDP)in2020,andtheAmerican

RescuePlanwas

roughly

8percent

of2021GDP.Bycomparison,theARRAof

2009wasslightlymorethan5percentofthatyear’sGDP.________________________________________________________________________What

Caused

the

U.S.

Pandemic

-Era

Inflation?1HUTCHINS

CENTER

ON

FISCAL

&

MONETARY

POLICYsalientlevelsofinflationmightleadinflationexpectationstode-anchor,raisingthe

potentialforawage-pricespiral.Thecritics’forecastsofhigherinflationwouldprovetobecorrect—indeed,eventoooptimistic—but,insubstantialpart,thesourcesoftheinflation,atleastinitsearlystages,wouldprovetobedifferentfromthosethey

warnedabout.Thelabormarketdidtightensignificantlyin

2021and

2022,asreflectedinseveralindicators,includingunsustainablyhigh

ratesofjobcreation,an

increasingratioofjob

openings2tounemployed

workers,and

lowlevelsofquits.

But,as

this

paperwillshow,thebehaviorof

nominalwages

in

2021and

2022wasbroadlyconsistentwiththerelationstheFedhad

reliedon,andmedium-terminflationexpectations

rosemodestly

bysomemeasuresbutdidnotde-anchor.Labormarketoverheatingwouldindeedultimatelyprovetobeasourceofpersistentinflation,as

thecritics

hadexpected,butthetraditionalwagePhillipscurvemechanismwasnotthemainevent,atleastnotuntilrecently.Inretrospect,thefailuretoforecasttheinflationburst

reflectedinlargepartthefactthat,infocusingonthelabormarket,boththeFedand

itscriticsunderestimatedtheinflationarypotentialofdevelopmentsin

goodsmarkets,thatis,

fromincreasesinprices

givenwages.Theshockstopricescamefromseveralsources.First,likeotherforecasters(includingparticipantsin

commodityfuturesmarkets),mosteconomistsdid

notanticipateeitherthemagnitudeorthedurationof

theriseincommodity

pricesthatbeganin

2021.Beyond

themore-familiar(butunexpectedlylargeandsustained)shocks

tofoodandenergyprices,thepandemicitselfledtounusualdistortionsinkey

productmarkets,suchasthosefornewandusedvehicles.Strongaggregatedemandandshifts

inthecompositionofconsumerspending(e.g.,fromservicestodurablegoods)duringthepandemiccombinedwithconstraintson

supplyinsomesectorstocreateshortages

andsectoralprice

increasesnotoffsetbydecreasesin

othersectors

(Guerrierietal.,2022;diGiovanniet

al.,2023).Thesesectoralmismatchesbetweendemand

andsupply

provedmoreintractableandlonger-lastingthanmany

hadexpected.Together,theseshockstopricesgivenwageswouldprove

tobethecriticaltriggersoftherise

in

inflation.This

isthestorywedevelopinthispaper.Westartwithasimpleanalyticalmodel,focusingonthebehaviorofwages,prices,

andshortandlong-runinflationexpectations,takinglabormarketslackandshockstopricesas

given.We

showtheverydifferentdynamiceffectson

wageandpriceinflationof

overheatinginthelabormarketand

price

shocksingoodsmarkets.Appreciatingthesedifferencesisessentialforunderstandingthe

changingrelativeimportanceofthetwosourcesofinflation,andbyimplication,theevolutionofinflationovertime.Wethenestimateanempiricalversionofthemodel,allowingforagenerouslag

structuretoaccommodateflexibledynamicsandestimatingitonthepre-covidperiod.Weexaminethestabilityofthewage,price,andshortandlong-runinflationexpectationsrelationsbetweenthepre-covidandcovidperiods,takingintoaccount

therolesoflabormarkettightness,energyandfoodpriceshocks,andsectoralshortages.Weusetheestimatedmodeltodecomposethesourcesofthepandemic-erainflationintotheeffectsofthevariousshocks.Ourdecompositioncapturesgeneral

equilibriumeffects,lags,andlinkagesamongalternativesourcesofinflationoperating

throughinflationexpectationsandotherchannels.Weconcludethat

labormarkettightnessmadeatmostamodestcontributiontoinflationearlyon,resolvingthepuzzleof

howinflationcouldrise

so

muchdespiteaflat

wagePhillipscurve.Instead,mostoftheearlyactionin

inflationcamefromthegoodsmarket,in

theformofsharp

increases

insome...2.

Bycontrast,theunemployment

rate,

whichremainedclosetoFOMCprojections,showedfewersigns

of

labor

markettightening.________________________________________________________________________What

Caused

the

U.S.

Pandemic

-Era

Inflation?2HUTCHINS

CENTER

ON

FISCAL

&

MONETARY

POLICYrelativeprices,includingcommoditypricesandpricesin

sectorsinwhichstrongdemandconfrontedlimitsonsupply.Totheextentthatcommodity

pricesstabilizeandsectoralshortagesmoderate—processesthatbothseemwelladvancedasofthiswriting(June2023)—thegoods

marketcomponentofinflationis

likelytodecreasein

importance,

andthelabormarketcomponenttobecomemoredominant.Lookingforward,withlabormarket

slackstillbelowsustainablelevelsand

inflationexpectationsmodestlyhigher,weconcludethattheFed

hastoslowtheeconomytoreturninflationtotarget.Theextentofthatslowingwilldepend

howeveron

theevolutionofcertainstructuralfeatures

ofthelabormarket,notablytheefficiencyoftheprocessofmatchingworkerswithjobs.1.

ABriefLiteratureReviewMuchrecent

workhas

providedempiricalanalysesofthepandemic-erainflation.Balletal(2022)istheclosesttoours,andwelearnedfromit.They

proposea

decompositionofinflationintoweightedmedianinflation(insteadoftheconventionalcoremeasureofinflation)andaresidual,equaltoheadlineinflationlessmedian

inflation.Their

useofmedianinflationratherthemoreconventional

measureofcoreinflationis

intendedtoremovelargepricespikes,duenotonlytothepricesofenergyandfood,whichcoreinflationdoes,butalsothoseduetosectoralshortages,whichhavemadethis

inflationepisodedifferentfromearlierones.They

findthattheincreaseinoverall

inflationresultedprimarilyfromshockstotheresidual(headlinelessmedian

inflation)butthatlabormarkettightnesshas

beenanimportantandincreasingcontributortomedianinflation.Theseconclusions

arebroadlyconsistentwithours,butdifferencesinsampleperiodandmodelspecification(e.g.,theirspecificationimpliesthattheeffectsoflabormarkettightness

arefeltonlygradually)leadthemtobegenerallymorepessimisticthanwe

areaboutthecostsofreversingtheinflationsurge.Cecchettietal(2023)studytherecentinflationandtheprospectsfordisinflationinahistoricalcontext.Aswedo,

they

provideastripped-downmodeloftheinflationprocess.However,incontrasttothelargelyunconstraineddynamicsthat

weallowintheempiricalmodelpresentedbelow,they

imposeatightNewKeynesianstructure.Theyarguethatnonlinearities

inthePhillipscurveareimportantforexplaininginflation(seealsoBenignoand

Eggertsson2023,

GagnonandSarsenbayev2022).Incontrast,wefindthatalinearrelationbetweenlabormarkettightnessand

wages,estimatedonpre-coviddata,fitsthecovid-eradatareasonablywellandisadequatetoexplaintherecentbehaviorofwages.Cecchettiet

alalsoconcludethat,basedonthehistoricalrecord,disinflationislikelytobecostly,andthey

faulttheFedforfailingtotightenpolicy

pre-emptively.GagliardoneandGertler(2023)focusonthemonetary-policyoriginsoftheinflation.UsingacalibratedDSGEmodeltostructuretheiranalysis,theyarguethattheFed’scommitmenttohighemploymentinthefaceofinflationaryoil

priceshocks

explainsthesharp

rise

inthe

paceof

priceincreases.GagliardoneandGertler’sattentiontooilpriceshocks

parallelsourownfocusonfactorsdirectlyaffectingpricesgivenwages.Their

main

contributionsconcerntheroleof

theFed’sreactionfunctionandpolicyframework

in

theover-expansionof

aggregatedemandduringtherecoveryfromthepandemic—animportanttopicbutnotthefocusofourownpaper.KochandNoureldin(2023)oftheInternational

MonetaryFundconductapost-mortemoftheirinstitution’sfailuretoforecastthe(global)inflation.Likeotherpapers

intheliterature,theirwork

findsanimportantrolefor

increasedaggregatedemandand,

especially,pandemic-inducedsupplyconstraintsinsettingofftheinflation.Theyalso

pointtoover-estimationofoutputgapsasa

factorleadingtoexcessivestimulus.However,they

paylimitedattentiontodynamics,andin

particulartheyplaceless________________________________________________________________________What

Caused

the

U.S.

Pandemic

-Era

Inflation?3HUTCHINS

CENTER

ON

FISCAL

&

MONETARY

POLICYweightthan

wewouldoncommodity

priceshocksandtheir

effectsonotherprices

andinflationexpectationsovertime.2.

ASimpleModelofWage-PriceDeterminationTohelp

interpretrecentinflation

developmentswithoutimposingtoomucha

prioristructure,weuseabare-bonesbutflexiblemodelofaggregatewage-pricedetermination.Ourapproachtoinflationisaggregative,or“topdown”;witha

partial

exceptiondiscussedinthenextsection,wedo

notconsider

themicroeconomicdeterminantsofsectoralpricechanges,e.g.,factorsaffectingthebehaviorofrentsorthedifferentialimpactofthepandemicondifferentindustries.Ourmodelreducestofourequations,whichjointlydeterminenominal

wages,prices,andshort-runandlong-run

inflationexpectations.Asalreadynoted,whenwetakethebare-bonesmodeltothedatainthenextsectionweincorporateamoreflexiblelagstructure,toallowforricherdynamics,and

includemeasuresofkeyshockstoproductandlabormarketstoaccountmoreexplicitlyfortheforcesaffectinginflation.ThewageequationWestartwithanaggregatewage

equation.We

makethestandardassumptionsthatnominalwagesdependontheexpected

pricelevelandthedegreeoflabormarketslack.Inadeparturefromthestandardspecification,wepursuethequestionof

whetherworkersacceptlossesin

purchasingpowerresultingfromunanticipatedpriceshocks,orwhetherinstead,intheirwagebargaining,theytryto“catch

up”

withpriorinflation.3Specifically,weassumethatthenominalwage

w

ineachquarterdependsontheexpectedpriceforthatquarter

??,anaspirationrealwage

??

(seebelow),and

anindicatorofthetightnessofthelabormarket

?:(Eq1)

?

=

??

+??

+??Wages,prices,andexpectedpricesareinlog-levels,so

differencesshouldbeinterpretedas

growthrates.Weuse?

todesignatetherealaspirationwageas

areminderthatthat

variableisin

realterms,indistinctiontothenominal

wage

w.Wediscuss

empiricalmeasuresoflabormarkettightness

?

inthenextsection,onlynotingherethesignconventionthat

weinterpret

highervaluesof

?

as

correspondingtoatighterlabormarket.We

suppressthetimesubscriptforcurrent-quartervariables.Weincludetheaspirationrealwage

??

inthemodelasa

devicefor

modelinga

possiblecatch-upeffect,inwhichworkersseek

tomakeup

forpastlossesofpurchasingpower,implyingadegreeofreal-wagerigidity.Theaspirationwagemaybethoughtofas

thereal

wagethatworkersbelieveisachievable—afocalpointinwagebargaining,perhaps—givenrecentdevelopmentsinwages,prices,andotherfactors.Alternatively,theaspirationrealwagemightbethoughtofasrepresentingthelong-runequilibriumrealwage,towardwhichtheactualrealwage

graduallyadjusts.Adoptingthenotation

?(??)

tostandforthevariable

?

lagged

?

quarters,

weassumethat

theaspirationrealwageineachquarterisaweightedaverageoftheprevious

quarter’saspirationwageand

realizedreal

wage,plusa

shock

term

?

,standing?foralltheotherfactorsthataffectwage

determination:...3.

KaminandRoberts(2023)analyzehowalternativeforms

of

real-wagecatch-upmightaffectthe

courseoftheeconomy.________________________________________________________________________What

Caused

the

U.S.

Pandemic

-Era

Inflation?4HUTCHINS

CENTER

ON

FISCAL

&

MONETARY

POLICY(Eq2)

??

=

???(?1)+(1

??)(?(?1)??(?1))

+??When

estimatingthemodel,wealsoallowwagegrowthtodependonan

exogenous

trendinproductivitygrowth.Forsimplicityofexposition,weignorethattermhere.Puttingthetwo

equationsabovetogethertoeliminate

??

yields:(Eq3)

?

??(?1)

=

(??

??(?1))

+?(?(?1)???(?1))

+?(?

???(?1))

+??Examinationoftheseequationssuggeststwocasesofinterest,dependingon

thevalueofthe“catch-up”parameter

?.If?

=

0

then,from(Eq2),theaspirationrealwage

ineach

quarteris

justtherealwageintheprevious

quarter(adjustedfortrendproductivity

growth,intheempiricalversionofthemodellater)plusanerrorterm.Further,from(Eq

3),inthespecialcaseof

?

=

0,thechangeinthenominalwage

equalstheexpectedrateofinflationinthe

previousquarter,

??

??(?1),

plusatermthat

dependsonthedegreeoflabor-market

tightness

?—inshort,astandardexpectations-augmentedwagePhillipscurve.If

?

≠

0,ontheotherhand,then(Eq3)

impliesthatamongthefactorsdeterminingnominal

wagegrowthisthedifferenceinthepreviousperiod’s

pricelevelandthepricelevelthat

hadbeenexpectedforthatperiod,

?(?1)???(?1).Wecallthistermthecatch-up

term.Theexistenceofa

catch-up

termimpliesthat,intheirbargainingwithemployers,workersseektobecompensatedforlastperiod’sunexpected

inflation.Inotherwords,totheextentthat

workers’expectationsorfocalpoints(orthoseoftheir

institutionalrepresentatives)influence

thewagebargainingprocess,theeconomyexhibitsadegreeofreal-wagerigidity.Notealsofrom(Eq

3)that,if

?

≠

0,thetermrelatingtolabor-markettightnesscanberewrittenas

(1

??)?

+?(?

??(?1)),implyingthatwagespotentiallydependonboththelevelandchangeof

labor-markettightness.The

priceequationGivenwages,weassumethatthepricelevel

?

dependson

thelevelofnominal

wagesplusashock

term?

thatcapturestherelativecostsofnonlaborinputs,variationsinmarkups,andotherfactorsaffecting?price-setting:?

=

?

+??Or,in

firstdifferences:(Eq4)

?

??(?1)

=

(?

??(?1))+(??

?

?

(?1))?Intheempiricalmodel,wewillfleshouttheshock

termbyincludingvariablessuchascommodity

priceshocksandsupply-chainproblemsthatincreaseprices

givenwages.Sincepricesshoulddependon

unitlaborcostsratherthanwagesper

se,intheestimationwealsoincludetheproductivitytrendinthepriceequation.We

willnottreat

producermarkupsasanindependentvariablebutnotethatmarkups

willendogenously

increase

inourmodelwhenstrongsectoraldemandconfrontslimitedsupply

(seebelow).Inflationexpectations

equationsThe

modelisclosedbyequationsthatdescribethebehaviorofthepublic’sshort-runandlong-runinflationexpectations.Short-runinflationexpectationsareaweightedaverageof

long-runinflationexpectations,

??,

andlastperiod’sinflation:________________________________________________________________________What

Caused

the

U.S.

Pandemic

-Era

Inflation?5HUTCHINS

CENTER

ON

FISCAL

&

MONETARY

POLICY(Eq5)

??

??(?1)

=

???

+(1

??)(?(?1)??(?2))Long-run

inflationexpectationsinturnevolveasa

weightedaverageoflastperiod’slong-runinflationexpectationsandactualinflation:(Eq6)

??

=

???(?1)+(1

??)(?(?1)??(?2))The

parameters

?

and

?

capturethedegreeofanchoring

ofshortandlong-runinflationexpectations.Ifboth?

and

?

arecloseto1,expectationsarewellanchored.

Both

?

and

?

playimportantthoughdifferentrolesindeterminingthedynamic

effectsofshocksoninflation.Theparameter

?

affectsshortrundynamics,andthedegreetowhichpriceinflationaffectswageinflation:

thelower

?,themorepersistentthedynamiceffectsofapriceshock.Theparameter

?

affectslongrundynamics,andinparticulartheeffectoftransitoryshockson

longruninflation.Inspectionoftheequations

aboveshowsthat,inthelong-runsteadystate,bothinflationexpectationsandtheinflationrateconsistentwithfullemployment

areindeterminate.Foranyvalueoflong-runexpected

inflation

??,assumingthat

?

=

0,themodelsolutionimpliesthatshort-runinflationexpectations,wageinflation,and

price

inflationall

equalthearbitrarilychosenvalueof

??.Putanotherway,thevalueof

??

atanypointintimeisdeterminedbythehistoryofinflation.Anepisodeof

higherinflationleadstohighersteady-stateinflation;thelongertheepisode,orthelower

?,

thestrongertheeffectonsteady-stateinflation.Onecanthinkofthe“de-anchoring”thatpessimistsworriedaboutatthestartoftheinflationepisodeasdecreases

ineither

?

or?.Otherthings

equal,suchde-anchoringwouldleadtostrongerand

more

persistentinflationfollowinganinflationaryshock.Whentheaspirationwageissubstitutedout,asabove,themodeldeterminestheevolutionoffourendogenous

variables:

??,??,

?

,and

?.Despiteitssimplicity,thebasicmodeldisplays

interestingdynamicproperties.Toillustrate,Figure1belowshows,foralternativeparameterassumptions,themodel-implieddynamicresponses

ofinflationtoaone-timeshock

totheprice

equation.Specifically,wesimulatethefullmodelundertheassumptionthattheprice

shock

?

rises

permanentlybyoneunitin

period1.

(Notethat?apermanentshock

tothepricelevelisaone-periodshock

topriceinflation).Weconsidertwocases:

Inthefirst,weassumealimitedcatch-up

effectandstableinflationexpectations(?

=

0.2,

?

=

0.9,

?

=

0.95).4Inthiscase,asFigure1shows,theshock

totheprice

levelleadstoasharp

increasein

inflationthat

isalmostcompletelyreversedafterafewperiods(seethecurvemarked“weak

feedback”).

Thelowpersistenceofinflationinthiscasereflectsthe

weak

catch-up

effect(workersareunabletorecouptheeffectsofunexpectedinflationontheirreal

wages)and

well-anchoredinflationexpectations.Becauseofthe(small)effectoftheinflationshockon

long-runexpectations,however,inflationendsup

permanentlyslightlyhigher,by0.06percentinthisexample.Withalimitedcatch-up

effectandwell-anchoredexpectations,theconventionalwisdomthatmonetary

policymakerscan“l(fā)ookthrough”temporary

supplyshocksisjustified.The

responseofinflationtoaone-timepriceshockisquitedifferentwhenthecatch-upeffectisstrongerand

inflationexpectationsarelesswellanchored(say,

?

=

0.6,

?

=

0.7,

?

=

0.9).

Inthissecondcasetheinflationresponse

toaone-timepriceshock

diesdownonlyslowly

(seethe

curvemarked“strong...4.

Wedonotneedto

assignavalueto?

as

doing

so

wouldnotaffectthesesimulations.______________________________________________________________________