JEP2312.docJournalofEconomicPerspectives,forthcomingCommentswelcomeEscapingfromaLiquidityTrapandDeflation:TheFoolproofWayandOthersLarsE.O.SvenssonFirstdraft:January2003Thisversion:December2003AbstractExistingproposalstoescapefromaliquiditytrapanddeflation,includingmy“FoolproofWay,”arediscussedinthelightoftheoptimalwaytoescape.Theoptimalwayinvolvesthreeelements:(1)anexplicitcentral-bankcommitmenttoahigherfuturepricelevel;(2)aconcreteactionthatdemonstratesthecentralbank’scommitment,inducesexpectationsofahigherfuturepricelevelandjump-startstheeconomy;and(3)anexitstrategythatspecifieswhenandhowtogetbacktonormal.Acurrencydepreciationisadirectconsequenceofexpectationsofahigherfuturepricelevelandhenceanexcellentindicatorofthoseexpectations.Furthermore,anintentionalcurrencydepreciationandacrawlingpeg,asintheFoolproofWay,canimplementtheoptimalwayand,inparticular,inducethedesiredexpectationsofahigherfuturepricelevel.IconcludethattheFoolproofWayislikelytoworkwellforJapan,whichisinaliquiditytrapnow,aswellasfortheeuroareaandtheUnitedStates,incaseeitherwouldfallintoaliquiditytrapinthefuture.JELClassification:E52,F31,F33,F41Keywords:Zerobound,nominalinterestrates,optimalescapeLarsE.O.SvenssonisProfessorofEconomicsatPrincetonUniversity,Princeton,NewJersey.HeisalsoResearchFellow,CentreforEconomicPolicyResearch,London,UnitedKingdomandResearchAssociate,NationalBureauofEconomicResearch,Cambridge,Massachusetts.Hishomepageis~svensson.1IntroductionInthelastdecadeortwo,centralbanksallovertheworldhavebeenquitesuccessfulinachievinglowandstableinflation.Averageannualinflationintheindustrializedcountrieshasfallentoalevelbelow2percent.Averageinflationhasnotbeensolowsincethe1950s,asshowninTable1.Inemergingcountries,inflationisnowthelowestsincethe1960s(InternationalMonetaryFund(2003b)).Table1.Averageannualinflationinindustrialcountries(consumerpriceindex,percent)1950-592.81960-693.21970-798.21980-895.61990-992.72000-031.8(Source:InternationalMonetaryFund(2003b).)Thesegainsagainstinflationaregoodnews.Theyhavebroughtsubstantialbenefitsintermsofreduceddistortions,lessuncertainty,andimprovedresourceallocation.Buttheyalsoraisenewrisks.Unanticipatednegativeshockstodemandorsupplycanalwayscauserecessionsandlowerinflation−and,startingfromalowinflationlevel,evendeflation.Insuchasituation,theappropriateresponsebycentralbanksistolowerinterestratesandthiswaystimulatetheeconomyoutofrecessionandtoolowinflation.Butwithlowinflationorevendeflation,anegativeinterestratemayberequiredtoprovidesufficientstimulustotheeconomy,whereasnominalinterestratescannotfallbelowzero.Theeconomymightthenbecomecaughtinaliquiditytrapandaprolongedrecessionanddeflation.Thispaperbeginswithadiscussionofthecausesandconsequencesofaliquiditytrapanddeflation,withsomeemphasistoJapan’sexperiencesincethe1990s.Itthendiscussespolicyoptionsforpreventingaliquiditytrapanddeflationfromoccurringandforescapingfromaliquiditytrapanddeflationiftheyhavealreadyoccurred.Whereaspolicyforavoidingaliquiditytrapanddeflationislesscontroversial,thereisafairamountofcontroversyabouttherangeofpoliciestoescapefromaliquiditytrapanddeflation,includingmyownproposal,the“FoolproofWay”(Svensson(2001,2002)).CausesandConsequencesofaLiquidityTrapandDeflationHowcanaliquiditytrapanddeflationarise,andwhyaretheyaproblem?Anincreasingnumberofcentralbanksaimbothtostabilizeinflationaroundalowlevelandtokeepoutputclosetoitspotentiallevel.Butmonetarypolicyoperatesunderconsiderableandunavoidableuncertaintyaboutthestateoftheeconomyandthesizeandlagoftheeconomy’sresponsetomonetary-policyactions.Unanticipatedshockstodemandandsupplyareunavoidable.Becauseofthelagsintheeffectofmonetary-policyactions,goodcentralbanksareforward-looking,useavailable2informationabouttheeconomyandanticipatedshockstoconstructforecastsofinflationandoutput,andrespondtotheseforecastsasbestastheycan.Asubstantialrealizedoranticipatednegativeshocktoaggregatedemand,becauseof,forinstance,theburstingofanasset-pricebubble,acorrectionofoveroptimisticgrowthandproductivityexpectations,increaseddoubtsaboutfuturepensionsandbenefitsduetodemographicdevelopmentsand/orrecklessfiscalpolicy,increaseduncertaintyforgeopoliticalorotherreasons,willlowerbothactualinflationandoutputaswellasforecastsoffutureinflationandoutput.Ifinitialinflationislow,thismaybeallthatisneededfornotonlyatemporaryrecessionbutatemporarydeflation.Whencentral-bankforecastsindicaterecessionandtoolowinflationorevendeflation,theappropriateresponseistolowerthecentralbank’s“instrumentrate,”theshort(-maturity)nominalinterestrateitusestoimplementmonetarypolicy–thefederalfundsrateintheUnitedStates.Alowershortnominalinterestrate,combinedwithsluggishprivate-sectorinflationexpectations,willlowerstheshortrealinterestrate–thenominalratelessexpectedinflation.Expectationsoflowerfutureshortrealratesthenlowerlonger(-maturity)realrates,theratesthatmatterforconsumptionandinvestmentdecisionsandtherebyaggregatedemand.Thelowerrealrates,withsomelag,stimulateaggregatedemandandoutputandbringtheeconomyoutofrecession.I