tinaning

NBERWORKINGPAPERSERIESRULESVERSUSDISCRETIONINMONETARYPOLICYStanleyFischerWorkingPaperNo.
2518NATIONALBUREAUOFECONOMICRESEARCH1050MassachusettsAvenueCambridge,MA02138February1988TheresearchreportedhereispartoftheNBERsresearchprograminEconomicFluctuations.
AnyopinionsexpressedarethoseoftheauthorandnotthoseoftheNationalBureauofEconomicResearch,SupportfromTheLyndeandHarryBradleyFoundationisgratefullyacknowledged.
NBERWorkingPaper#2518February1988RulesVersusDiscretioninMonetaryPolicyABSTRACTThispaperexaminesthecaseforrulesratherthandiscretionintneconductofmonetarypolicy,frombothhistoricalandanalyticperspectives.
Thepaperstartswiththerulesofthegameunoerthegoldstandard.
Iheserules,ereill—definedandntadheredto;activediscretionarypolicywaspursuedtodefendthegoldstandard—butthegoldstandardcaineclosertoaregimeofrulesthanthecurrentsystem.
TheargumentsforrulesingenerldevelopedbyMiltonFriedmanaredescribedmoappraised;alternativerulesincludingtneconstantmoneygrowthratirule,interestraterules,nominalGNPtargeting,andpricelevelrulesareanalyzed.
Until1977thegeneralargumentformonetaryrulessufferedfromtheapparentdominanceofdiscretion:ifaparticularmonetarypolicywasdesirable,itcouldalways09adoptedbydiscretion.
Theintroductionofthenotionofoynamicinconsistencymadeastrongercaseforrules,thefinalsectionsanalyzetinecaseforrulesrathertnandiscretioninthelightofrecentgametineoreticapproachestopolicyanalysis.
StanleyFischerS-9035TheWorldBankISISHStreetWashington,DC2i)433Januaryi9BE.
RULESjERSUSDISCRETIONINMONETARYPOLICi,StanleyFiscnerLDiscretionarymonetarypolicyhaslongbeenananomalytoliberaleconomists,HenrySimonsaddressedtheissue2inhisclassicpaper'RulesversusAuthoritiesinMonetaryPolicy"(1948,originalini936;iThemonetaryproblemstandsouttodayasthegreatintellectualchallengetotheliberalfaith.
.
.
Theliberalcreeddemandstheorganizationofoureconomiclifelargelytnroughindividualparticipationinagamexutndefiniterules.
.
.
.
[Diefinite,stable,legislativerulesoftnegameastomonejareofparamountimportancetothesurvivalofasystemoasedonfreedomofenterprise.
(pp.
160—162,emphasisinoriginal;Simonsposedtheissueasoneofrulesversusauthorityor,rulesversusziscretion.
Thatdichotomyshouldratherbeseenasacontinuum,inwhimtheevtentofdiscretionlefttothemonetaryauthorityisdeterminedbytnespecificityoftheobjectivesitisgiven,andtheimmediacyofthelinkoetweenitsactionsandtheattainmentofthoseomiectives.
AtoneextremeofSiscretion,amonetaryauthoritycouldhavethefullpowersoftheFed(toouyandsellsecurities,settnediscountrate,reserverequirements,Andotherregulations(andbegiventheobjectiveofpromotingeconomicwell—being.
'WorldBankandMIT,andResearchAssociate,NBER.
ThispaperwaspreparedasachapterfortheforthcomingHandbookofMonetaryEconomicseditedoyBenjaminFriedmananoFrankHahn.
IamgratefultoOlivierBlanchard,BenFriedmanandMiltonFriedmanforhelpfulcomments,andtheNationalScienceFoundationforresearchsupport.
21thadbeendiscussedacenturyearlier,inthecontext0fthedisputebetweenthecurrencyandbankingschools(finer,1953,ChapterVi.
Fetter'1965,p.
174;quotestheChancelloroftheExchequersayinginl839idenytheapplicabilityoftnegeneralprincipleofthefreedomoftradetothequestionofmayingmoney'.
Bothtnecbective,economicwI—being,andthelinkbetweenthemonetaryauthoritysactionsandtheattainmentoftheobjectivearevague.
ttneothermzreme,thecentralbarkcouldbedirectedtoexpanoitsicingsongovernmentsecuritiesweekly,atanannualrateof4peryearIntermediatearrangementsenistinseveralcountries,inwricntnecentralbankisgivengeneralobjectivesbutrequiredtoreportregularlyndjustifyitplanofactionforareasonableperiodaheadItsplansnaybesummarizedBymoneyandinterestratetargets.
Itnayalberepuireztoreportontheexecutionandoutcomeofpactpolicies.
ThereportsmaybemadetothelegislaturetotheTreasuryortothepublic.
Intrelastcase,publiccrtticimandtheinherentthreatthatthelegislaturewillintervCne,constitutethelimitsonthepowersoftheinstitution.
Bisonsconscoeredalternativemonetaryrules,includingconstancycithequantityofmoneybut,notingthedangerofsharpchangesonthevelocitysidelp.
ld4,concludedthattheoptimalruleatleastforaninterimperiodwasforthecentralbanktostabilizethepricelevel.
Mints11950)likewisesupportedaprice—levelruIfortheUnitedStatesMiltonFriedman1)948)afterechoingBisonspointofprinciple,developedaframeworkinwhichfiscalandmonetarypolicywouldoperateautomatically.
Itincluded10O.
reservemoney,essentiallyconstantgovernmentspendingandtaxrates,and,throughfullmonetaryfinancingofdeficits,countercyclicalmonetarygrowth.
Friedean11948)raisedthequestionofwhetherlongandvariablelagsintheoperationofpolicymightcausetheactive(thouphautomatic)countercyclicalpolicyoftheframeworktbedestabilizing.
Infp.
.
rsm+cr(19b0)heargued,onthebasisofresearchlaterpuoliehedinFriedmanandSchwartz(l93),thatthetheFedha;frequentlyteenasourceofeconomicinstability.
GiventherecordandthetheoreticalimplicationSoflongandvariablelags,Friedmanadvocatedtherulethatthegrowthrateofmoneybeheldionstant.
Onthegeneralissueofrulesversusdiscretion,hesuggestedthatdiscretionhadpermitteddestabilizingshiftsinmonetarypolicyamthecentralbankhaddeenwaydbypublicopinionaridpoliticalpressures)arthat,zecausethecriteriaforjudgingtheperformanceofthemonetaryauthorityareSOimprecise,theFedsdiscretionarypowershadenabledittoescapeseriouspublicscrutiny.
newsetofargumentsintherulesversusdiscretiondebatedevelopedfrorithedynamicinconsistencyliteraturebroughttomacroeconomicsbycydlandandPrescottl977)Attheformallevel,Friedmansanalysisuffermdfromthelogicalweaknessthatdiscretionseemedtodominateruleerifaparticularrulewouldstabilizetrimeconomy,thendiscretionarypolicymakermcouldalwaysbehavethatway——endretaintheflexibilitytochangetheruleasrieeded.
Thedynamicinconsistencyliteraturesnowedthatprecommitmentoymonetaryauthoritiescouldimprovethebehavioroftheeconomy.
InthischapterIfirstdiscussthegoldstandardasaquasi—automaticmonetarypolicyregime,thenturntotheissuesraisedbytheChicagoschool.
AlternativemonetaryrulesareesaminedinSectionIII.
Themodernliterature,vFriedmanil9O)esplainstheswitchfromhisearlier4rameworasresultingfromtheempiricalevidence,Cogentcriticismsofthe1O''.
moneyproposalandthelinkingofmonetaryandfiscalpolicybyClarkEarburton19db,ChapterId,originalin1952)mayalsohaveplayedarole.
Warburton(lbb,Chapter17,originalin1952)cameclosetoadvocatingaconstantmoneygrowthrule,suggestinganannualgrowthrateaveraging4—5l,withvariationsofnotmuchmorethan1—2/.
perannumtoallowforchangesinvelocity.
Seiden19d2)reviewsthedevelopmmntofsuppOrtfortheconStantgrowthraterule,givingsubstantialcredittoWarburton.
HoweverFriedmanidconfronttheissueofwhyaformalrulemightompreferabletoadiscretionarypolicy,makingananalogytothedillofRights(l9d2,pp.
29—24l.
4tenteredaroundtheconceptofdynamicinconsistency,anditsrelevancetotherulesversusdiscretiondebateisthefocusofecticnsitnroughfltTheGobStaniard.
puregoldstandardisafullyautomaticmonetarysystem.
Thespecie—flowmechanisminhiihthemoneystackadjuststhroughthebalance0fpaymentsrevealstheequilibratingtandence;inherentinthesystem.
WithallgoodstradedandtheirpricesequalizedworIwide,adustnentcomespurelythroughwealtheffectsastheoutflowofgold)specie4rmacountrywithacurrentaccountdeficitreducestheflowofspendinginthatcountry.
Withnon—tradedgoocsineachcountry,adjustmentsintherelativepricefhomegoodsthatshiftdomesticspendingbetweenhomeandtradedgoodsprovdsanadditionalstabilizingmechanism,asinHumesanalysis.
althoughfarfromfullyautomatic,thegoldstandardinitsneydaywasasclosetoamonetarysystemoperatedbyruleastherehasbeen.
Ithereforebrieflyreviewboththeoperation0ftheSystemandnineteenthcenturyanalysesfit,primarilyintheUnitedXin;ooewherethetheoryandpracticecfcentralbankingdeveloped.
Thecontextwasoneinwhichcommercialbankingwe;developingrapidlyandthequestionoftheeffects0fchangesinthequantityofbankdepositsontheeconomywasbeingdiscussed.
ThedebatebetweenthecurrencyandbankingschoolsprecedingPeelsBankActin1B44thatdeterminedtheformalstructureoftheBankofEnglandexplicitlyaddree;edtherulesversusdiscretionissue.
ThecurrencyschoolarguedthatthequantityofcurrencyshouldvarypreciselyasitwouldifallViner(1955,ChapterVI)present;aFullanalysisofgoldstandardmechanices.
5moneyweregold,meaningthatthebalanceofpaymentssnoulddeterminechangesinthequantityofcurrency;theydidnotviewbankdepositsasmoney;inctheyfavoredtneuseofmonetaryrulesratherthandiscretion.
Theoankingschooldisagreedontheseissues.
BotnschoolsoelievedthatthecurrencyshouldbeKeptconvertioleintogold6.
Thetemperofthetimes,soontoproducetnaCornLaws,favoredtneuseofmarketsandnotdiscretionaryauthority.
TheBankAct,reflectingcurrencyschoolviews',putastrictlimitontheBanksissueoffiduciarymoneyandrequiredallotherBankofEnglandnotestobebacked1007.
bygold".
TheBankofEnglandhadclaimedsince1832tobedeterminingtnequantityofcurrencybyfollowingcurrencyacnoolprinciples',buttnscurrencyschoolwasnonethelessseverelycriticaloftheBankamisuseofitsdiscretionarypowers.
''BankingscnoolopponentsoftheBillarguedbothtnatitwasamistaketotreatonlycurrencyasmoney,andthatthegobstancardrulefordeterminingthestockofmoneywasinanycaseunwise,sincetheappropriateoehaviorofthemoneystockdependedonwnethermovementsintneBankofEnglandsreservewerecausedbydomesticorforeigndisturbances,andon"GooddescriptionsoftneviewsofthetwoschoolsarecontainedinMints(1945,ChapterVI),andFetter(1965,ChapterVt,.
'IncludedinthisschoolwereLordOverstone,G.
W.
NormangrendfatnerofMontaguNorman),anoRobertTorrens.
Peelhimselfrecognizedthatinacrisis,discretionaryauthoritymighthavetobeevercised.
(Kindleberger,1978,p173).
"Silverwaspermittedtoconstituteuptoone—fifthofreserves,outtheBankdidnotdealinsilverafter1850.
'Thiswastneso—calledPalmerrule,describedin1832sy.
1.
HorsleyTamer,governoroftheBankofEngland,asguidingtheBanksoperations,Petter,1965,ppl32—l33(.
finerml955,p.
254(stronglysupportstheircriticisms'.
.
.
aurmngtneperiodfromabout1800toabout1860theBankofEnglandalmostcontinuouslydisplayedaninevcusaoledegreeofincompetenceorunwillingnesstofulfi,,therequirementswhichcouldreasonablybedemandedofacentralbank.
whethertnedisturbanceswerepermanentortemporary)Viner,1Z,p.
2ml.
VitnouontheoankingschoolsrealtillsdoctrineappearedtosuggestaruleforBankofEnglandoperations,theschooldidnotproposeanalternativelegislativerule,Viners)i95,p.
281)summaryoftheirpositiondescrbesthegeneralviewofproponentsofdiscretion:Peiiancemustbehadontnsgoodsenseandthecompetenceofthosewhonadchargeofthecreditoperationsofthebankingsyeten.
Despitethe1844BankAct!
BritishmonetarypolicyduringtneperiodfromtheBankActuntil1914wasactivelymanagec.
Theneedformanagementarose!
asthebankingscncolhadanticipated,fromthepresenceokfractionalreservebanking.
ClAimsontheBankofEnglandwerethroughouttheperiodasgoodasgold.
VirtuallytheentiregoldreserveofthecountrywasneldbytheBankofEngland.
Inseveralcrises,theBankinlendingfreelytomeetlargelyinternaldrainsofcurrencyexceededthelegallimitonitsuncoverediiabilties.
ItwastypicallylaterindemnifiedbyActofParliament.
Thecentralroletobeplayedbybankrateintheoperationofthegoldstandardwasnotanticipatedin1844.
Internationalcapitalflowsrespondingtointerestratedifferentialsmovedgoldfarmorerapidlythantheclassicalspecie—flowmechanism,Duringtheheydayoftheinternationalgoldstandard,from1880toWorldWar1,aninformalsetofrulesofthe'tFetterll9&,ChapterVI)describestheviewsofthecriticsoftheBanvAct,includingTooke,Fullarton,andJohnStuartMill.
'2ThiswastheeituationdecriedbyBagehot11906,originalin1873),whoarguedthatitwouldhavebeenbetterforthecommercialbankstoholdtneirowngoldreserves.
Butheeawnowayofmovingfroethecurrentsituationtothepreferredalternative,insteadvigorouslydevelopingtheviewthatinacrisistheBankshouldlendfreelyagainsthallgoodbankingsacuritiestLmQntheoperationofthegoldstandard,seeBordoandSchwartc)l984iparticularlytheintroductoryessaybyBordo,andEichengreenil9BZ.
TheEichengreenvolumecontainsseveralclassicsources;amongothers,W,hl.
Scameell'Theworkingofthegoldstandardrepaysreading.
7gameissaidtohavedevelopedtodescribethediscretionaryactionsthatmonetaryauthoritiesweretotakeinsupportofthesystem.
Acountrysuxferin;acurrentaccountdeficitwassupposedtotightendOmesticcredit,therebyprotectingthegoldreservemdgoldconvertibilitybyreducingoathcapitaloutflowsanddomesticdemand.
Tierulesofthegamewerenotformallyagreedto,notwelldefined,andnayriothavebeenimplemented.
TheMacmillanCommittee(19313paragraphs4b—47iunabletodefineoperatingrulesofthegoldstandard3setoutprinciplestowhich:ntralbanksinagoldstandardsystemwouldsubscribe3includingthestabilityofpricelevelsandexchangeratesasgoaLsofpolicy3ariatheavoidanceofnon-cooperativebehavior.
oodhart(19701definedthemainruletobeprotectionofthegoldreservetnrougnoanxrate,aridfoundthattheBanxfollowedtherule,Bloomfield(l99(,examiningdataforelevencountrieSovertheperiod1880l9l4,showedthatcentralbankspredominantlyviolatedtherulethatdomesticpolicyactionsshouldreinforcethespecie-flowmechanism.
'Thisissuggestiveofsterilization3whichwas"Moggridge(inBordoandSchwartz31984,p.
l9(saysthatthephrasewascoinedmocameintouseonlyin1930.
TheCuriliffeCommissionsFirstInterimBeportonthereturnoftheu.
(c,togoldattheendofWorldWarIcontainedaclearstatementoftrrcommonunderstanding:WheneverbeforethewartheBanksreserveswerebeingdepleted3therateofdiscountwasraised,This.
.
.
byreactingupontheratesvormoneygenerally,actedasacheckwhichoperatedintwoways,Ontheonehand,raisedmoneyratestendeddirectlytoattractgoldtothiscountryortokeepheregoldthateighthaveleft.
Ontieotherhand,bylesseningtiedemandsforleansforbusinesspurposes,theytendedtocheckexpenditureandsotolowerpricesintniscountry,withtheresultthatimportseradiscouragedarideuportsencouragedandtheevchangestherebyturnedinour'avour,Unlessthistw—f1dcheckiskeptinworkingorderthewnolcurrencysystemwillbeimperilled.
Paragraphl8"EconometricresultspresentedbySutton(l984iconfirmtheviewtriattheBankwasnotobeyingtherulesofthegame,thoughhebroadenstherulectoprecludetheBankreactingcountercyclicallytodomesticcontract1on.
SeealsopapersonGermany,Sweden,andItaly,by,respectivelf,McGouldrick,Jonung,andFratianniandSpinelli,inBordoandSchwartz1l984,suggestingtheruleswerenotfollowed.
acertainlyconnoninthegold—exchangestandardafterWorldWarI',notablyintnefaureoftheUnitedStatesandFrancstoexpandtneirhoneySuIi25inresponsetogoldinflows.
Inereisnoquestiontnattnegoldstandardmonetarysystemcidnotoperateoyrule,Inthefirstinstance,countrtsssometimesleftan:sonetinenreturnedtotnegoldstandard,oydiscretion,Andeverwnentheyaereontnegoldstandard,thecentralbank(intheUnitedStates,sometimestheTreasury)tookanactive,ifnotalwayssucceseful,roleinmanagingthesystemtomantanconvertibility,TheBankofEnglandmanipulatedtankrateactively,bydiscretion,mainlytoprotectthegoldreeerve.
18Theruiesofthegame1werefarfrombeingwell—defined,''Bloosfisldsexaminationofthepre—WorldWarIdatafollowstheformatofNurkees(1944)researcnoninter—Wsrmonetarypolicy,18Sayers(1958,Chapter2)providesacomcieeaccountoftheperiodafteri873,motingthattheBanksconcerntoprotectitsownincomeinterferedwiththedevelopmentofitscentralbankingactivities.
lfawtrey(l9a2coversalongerperiodandputsmoreemphasisonthemechanismsywhichbankrateaffectedrealactivity:hesawthelinkasbeingmainlythroughinventorydemans,andforthatreasonemphasicedthatchanoejintnerate,whichladtoinventoryaccumulationordecumulation,weremoresignificantthantheleveloftherate.
Nonetheless,thegoldstandardcameclosertoaregimeofrulestnanthecurrentsystem.
ThekeydifferenceisthatmonetarypolicyhadIclearlydefinedonjectivethatwasformostoftheperiodwithinthepowerofthemonetaryauthoritytoachievetomaintainconvertibilityofthecurrencyintogoldatafivedprice.
WhetherornottheruleennancedeconomicstaSilityrelativetoalternativefeasiblepoliciesisanothermatter,II.
TheChicagSchoolandRules_versusDiscretion.
TheFederalReserveSystemwasoriginallyevpectedtooperatewithiagold—5tandardsetting,althoughinspecifyingthatthemoneystockwastobeelastic,theFederalReserveActprovidedacontradictoryguidetononetarpolicy.
BecauseitstartedoperatingsuringWorldWarIwnenmanyofthebelligerentswereeffectivelynotonthegoldstandard,tneSystembeganoperationswitnnoeffectivelegislativecriterionfordeterminingthetotalstockofmoney.
Thediscretionaryludgmentofagroupofmenwasinevitablysubstitutedforthequasi—automaticdiscipline0fthegoldstandard(FrieSmanandSchwartz,1963,p.
193).
thedLscretionaryoperationofmonetarypolicycreatedlittlecontroversysolongasitwassuccessfulduringthe1920a.
ProposalsformonetaryreformintheUnitedStatesproliferatesafterthedebacieofmonetarypolicyintheBreatDepression,Impressedbybankfailures,agroupfChicagoeconomistsin1933advocatedasysteminwhichbankswouldhoidlOOP.
reservesagainstcheckabledeposits.
'WithlOOP.
reservesDankrunsHart(1952,1935inoriginal)providesanaccountandcritiqueofthe'ChicagoPlan",whichhesayswasindependentlydevelopedbyL.
aucruinCurrieatHarvard.
coulonotreoucethemoneystock,astheyhadintheGreatueprssson,ProponentsincludedIrvingFisher(1945origznalin935)2whosesimplestplanrequirecthemoneystocktobeheldconstantalteracurrencycommissionhedboughtsufficientgoverneentsecuritiestobringreservesupto100%.
kmongthealternativesproposedbyFienerwereconstancyofthepercapitanoninalmoneystockandapricestabilizationrule.
Fisheralsoimpliedtr,atinterestmightbepaidonreserves,MlthoughtheChicagoargumentsforruleswereoriginallydevelopedintheoontevtofthe100%reeerveplan,theyarenotinherentlyrelated.
SimonsandFishercouldeasilyhaveadvocatedstabilizingtheprizelevelinafractionalreeervesystem.
Friedmancouldlateradvocatetheconstantgrowthraterulewithoutrequiringtheeconomyfirsttomoveto100%money.
changeto100%reserveswouldhaveremovedonepotentialweaknessinanautomaticsystem——thedangerthataioesofconfidencewouldleadtorunsonbanks,Butitshouldhavebeenanticipatedthen,andisclearnowfromthefateofregulationintheseventies,thatitwouldhaveBeenimpossibletomaintaintherestrictionagainsttheuseofallotherfinancialintermediaryliazilitiesasmeansofpayeent.
ttThesystemwouldhaverequiredfrequentlegislativeordiscretionaryrulechangestokeepupwithprivatesectorattemptstocircumventtheartificialbarrier,andincreasedmonetaryuncertainty.
Ratherthanpreventsharpchangesinvelocitythe100%2'Inthepreface,FisherthankssembersoftheChicagogroupandacknowledgestheirmeeorandumonthe100%planastheoriginalsourceofmanyoftheideesinthebook.
2tSimons(1948,p.
l72)wasawareofthedifficulties.
Hethoughtthatawayofhandlingthemcouldbefound'whenweconceivetheproblembroadlyasthatofachievingafinancialstructureinwhichthevolumeofshort—termborrowingwouldbeminieized,andinwhichonlythegovernmentwouldbeabletocreate(anddestroy)eithereffectivecirculatingmediaorobligationsgenerallyacceptableashoardsmedia.
"11reserveproposalbymandatingtheuseofinefficientmethodsofbanking,wouldnaveensuredcontinuingshiftsinvelocity.
Monetaryrulesarediscusseointneremaincerofthischapterwithoutreferencetol)reserves.
itisusefultodistinguishtwolevelsofargumentintierulesversusdiscretiondebate.
Thefirstisgeneral,examiningtnecaseinprincipieforrulesratherthandiscretion,withoutnecessarilyspacsfyingtheruleordetailsoftheoiscretionarysystem.
Thesecondisspecific,discussingwnetnerthemoneysupplyshouldvaryonewayortheother,orwhetnerncminaxinterestratesshouldoefixed,andmayberelevanttobotnciscretionaryandrulessystems.
Friedmansgeneralargumentsforrulesratnerthandiscretionaretnataruleenaoiesthenonetaryauthoritytowitnstandpoliticalpressures,providescriteriaforjudgingitsperformance,andensurescertaintyaoouteconomicpolicyforprivateagents,Hisspecificargumentforaconstantgrowthraterulewasentirelypragmatic,thatthiswouldreduceeconomicinstaiiiity.
IbrieflydiscussFriedman'sgeneralcaseforrulesandtheninthenevtsectionexaminealternativeoperatingrules.
insulatingtnecentralbankfrompoliticalpressuresis,Ibelieve,aworthypurpose,thoughitaccordsillwithageneralpreferencefordemocraticdecisionmaking.
Theothertwoargumentsareunpersuasive.
--Thepaymentofinterestonreserveswouldreducetheincentivetoinventnon—bankdepcsitoryinstitutions,thoughitisunlikelyevensothatattemptstodifferentiatesharplybetweenianxsandotherintermediariesinordertocontrolthequantityofmoneywouldbesuccessful.
Friedman1962)remarksontheinherentlyundemocraticnatureofcentralbanking,referringtotheimplicitviewofSmileMoreau,agovernoroftheBankofFranceietweentheWars,thathe,Norman,SchacntandStrongzouloruntneeconomiesoftneworldifonlytheywereleftalonetodoso.
Themorerecentviewisthatcentraloanksfollowtheprevailingpoliticalwinds.
Societyofcoursemakesarrangementstosnieldothertypesofoecisionmaking——forinstancelegal——fromimmediatepoliticalpressures.
whatgoodszttoevaluatetneperformanceofttecentralsankfitiengagedinanevercisethatisirrelevanttotheisnaviorofthemacroeconomicariablestnetmatterFarbettertirezuiretnscentralDanKeacyyear,orquartertoevplainitsactions,andtosubjecttneevpianaticntotneuncertainevaluationsofanimprecisesclence——aswedoatpresentCertaintyiDouteconomicpolicyisnotacompellingargumentforruleseither.
Economicagentswantcertaintyaboutpricesandaboutoutput;cansaszzetheyhavenoinherentinterestinthebehaviorofthestockoKmoney.
IfadiscretionarypolicythatproducedanunpredictaDlepatriformoneyensuredpricestabilityandfullemployment,theuncertaintyaboutmonetarypolicywouldbeofnoaccount.
Leavingasidethedynamicinconsistencyargumentthatwillbediscussedlater,itisdifficulttoevaluatetheargumentofprinciplemadebybisonsandmostproponentsofrules.
successfuldiscretionarymonetarypolicythatmaintainsreasonablepricestabilityandemploymentwilllikelydomoretmaintainthegeneraluseoffreemerketsandpersonalfreedomsthananunsuccessfulrulethatcausesdiscontentovertiebasicorganizationoftieeconomy.
Iwillthereforeanalyzealternativepolicieswithdifferingdegreesofdiscretioninterms4thelikelyeconomicoutcomes,leavingthereadertofactorinhisorherpreferenceforrulesasamatterofprinciple.
III.
lternativeRules.
InthissectionIccn;ideralternativemonetarypolicies:tnieFriedmancntantmoneygrowthrulethepossibilityoffeedbackmonetaryrules,includingthefrequentmodernproposalsforaneinalNPrule;the13Fisher-SimonsruleofapriceleveltargetandtheFisher"compensateddollar"proposal.
Idonotconsiderexchangeratebasedrules.
Lh_catnI_Qth_jQAkLvFriedman'sargumentfortheconstantgrowthraterule-'ratherthananactivefeedbackruleisthattherearelonganbvariablelagsintheeffectsofmonetarypolicy.
Thusanyactivepolicythatesponostocurrenteventsmayhaveitseffectsonlyataninappropriatetime.
Letioethelevelofatargetvariable,saynominalSNP,andntbethelevelofthemoneystock.
Theaimofpolicyistominisioethevarianceoft,,,conditionaloninfornationavailableuptoperiod(t—si.
SupposethatcI,Y=C1m,tYt—.
4t"i,tmt—i+Ctwhere.
andSi.
tarestochasticcoefficients,andStisadiaturoanceterm,whichcanbetakentoiewhitenoise.
Considernowtheroleoflags.
First,intieabsenceofsags,activepolicymaybeunnecessary:ifthem,intl:kwereidenticallycero,anCgi'enthatiswhitenoise,activemonetarypolicycouldnotreducethevarianceofnominalGNP.
Longlagsinthesystem,reflectedinthem.
,makepolicypotentiallymoreuseful.
Themechanismsthatproducesuchlagsarelikelyalsotocausemonetarypolicytoworkwithlonglags.
Longlagsintheeffectsofpolicybythemselvesarenotnecessarilyanimpedimenttotnesuccessfuluseofactivecountercyclicalpolicy.
lFecherendCooper,lT73).
Ifthecoefficientsin(1)werenotstochastic,thenoptimalmonetarypolicycouldevactiyoffsetthelaggedeffectsofearlierdisturbancesandmonetar,policybysetting"4kftersomediscussion,Friedman(1960)proposestherebenoseasonalvariationinthegrowthrateofmoney.
Thiswouldrestoretheseasonaltointerestrates.
2mt—C=3CEf—11Unefifficulcywiththepolicyin(2;isthatitcouldproduceinstrumentinstability,requiringeverlargerchangesinthemoneystocktooffsetitslaggedeffects(Holbrook,1972).
itwouldtemorelikelytodosothemoreslowlytheeffectsofmoneyonoutputbuildup;ifweresnail,therule;2)wouldcallforlergefluctuationsinthemoneystocu,Inthatsenselonglagsofpolicycouldmaketheactivepolicyin)2uncesirable,butallowingforcostsofinstrumentinstabilityintheobjectivefunction,optinalpolicywouldstillinthismodelbeactiveUncertaintyaboutthelagcoefficients,themeansthatactiveuseofmonetarypolicyaddsvariabilitytoincomeActivepolicycanstillbeused,cautiously,toreducethevarianceofoutput,butthegainmaynesmali.
2Thepresenceofvariablelags,then,makesoptimalpolicylessactive,andinthatsenseisanargumentinfavoroftheconstantgrowthraterule.
Totallyinactivepolicycouldbeoptimalifthemereuseofactivepolicyaddsuncertaintytothesystemindependentoftheparticularactivepolicyfollowed.
26FrieomansevidenceforlongandvariablelagsintheeffectsofmoneyisbasedonacomparisonofturningpointsinthegrowthrateofmoneyandcyclicalpeaksinactivityinNBERreferencecycles.
v'Reducedformevidence,forexampleoftheStLouisFedmodel(AndersenandJordan,l97(H2Thisistheeffectofmultiplieruncertainty,discussedbyBraimerd(1967).
ThThiecouldbethecaseifeconoeicagentsviewedtheFedaseitherbeingtotallyinactive,orelsepotentiallyasourceofinstability——asimplicitinFriedeanedevelopmentofthecaseforaconstantgrowthraterule.
27Friedmam(1969,originalin1961)discusseshisevidenceandcriticismsofit.
15orBarro1978,orPlishkin(1q83)typicallyfindsareasonablyclose:cnnectionbetweenmoneygrowthandthesubsequentbehaviorofnominalGNF.
'intheIt,LouiSmodelactivemonetarypolicycanbestabilizingevenwnenlagsaretreateoasstochastic.
(CooperandFischer,1974Intneaboveexample,azero(logaritnmofthe)moneystockminimizesthenoiseaodedtotnesystembyactivemonetarypolicy.
Tiequestionarises0fwhatisthecorrespondingmonetarypolicyinpractice(Diamond,l98.
Fredmanregardsapolicyofmaintainingaconstantgrowthrateofmoneyasinactive,tnougheveninthiscasethemonetaryrulehastospecifywhetherpastmissesinattainingthegivengrowthratearetobeignoredorcorrected.
Alsotobeconsideredarethequestionsofwhichmonetaryvariabletotarget,andwnetheranalternativepolicysuchasattemptingtofivthenininalinterestratemigntproducealowervarianceofthetargetaria0a.
-ExceptiftheargumentforaruleisbasedontheprincipleofminimizingtheFedsdiscretion,thesequestionscannotbeansweredwitnoutuvingananalyticandultimatelyaneepiricalmodel,TneFedadiscretcnxminimizedbygivingitataskthatitcanaccomplishexactly,theamplestofwhichistorequireittoincreaseitsportfolioeachweekatagivenanruumrate,say4'!
.
orOX.
3oStatisticalinferencewouldplacethemininun3ThevariabilityoflagsfoundbyFriedmancouldresultfromtheonievionofocnerfactorsthatmovethecycle,suchasfiscalpolicy.
.
solongasabaselevelofanominalvariablesuchasthemoneyatockiaspecified,aninterestratepeggingruleneednotproduceindeterninacyitthemoneystock.
See8lanchardandFischer(1988,Chapter.
0anoMcCai,m(1981)ItisoecausehebelievestheFed-sautnorityshouldbeminimizedtnatFriedmanhasrecentlymovedfromhisformerproposalthatthemoneygrzwtnruletargetMlor12totheviewthatthemonetarybaseshouldOsnabconstant,uncertaintyabouttheoutcomeofpolicyatthehistoricalaverageieveiohthemonetaryvariableintheregressionrelatingtnebehaviorofthetargetvariable;reaioutput,ornominalGNP(totheinstrumentvariable.
Thestrongestargumentagainstaconstantgroathrateruleformoneyisthatthevelocityofallmoneystockshaevariedsubstantiallyendwithcomeshort—runpredictability.
Itcanbearguedtnatthesevarationsarethemseiveeinducedbyunstablemonetarypolicy,butitishardtooelievethatshockstothedemandformoneythatcauseinterestratemovements,andtechnicalprogressinthepaymentssystemwillnotcausecontinuingfuturechangesinvelocity.
Becausethebehaviorofthestockofmoneyperseisnottheultimategoalofpolicy,thereisnoreasonotherthanthefearthatanyactivepolicywilldegeneratefornottakingsuchchangesintoaccountinsettingmonetarytargets.
InterestRateversusMg.
fl!
jTargets:Inafasousarticle,Poole(1970)analyoedinthecontextofafiaedpriceIS—UImodeltheguestionofwhetheroutputwouldoemorestableifmonetarypolicyheldtheinterestrateorthequantityofmoneyfixedinthefaceofshockstotheIScurve(shocksfromimvestment,consumption,orgovernmentdemandforgood;)andtheLMcurve(moneydemandshocks).
Thewell—knownansweristhatinterestratepeggingstabilioesoutputifshocksareprimarilyfrosmoneydesand,andthatmoneystockfixingispreferableifshocksareprimarilyfroetheIScurve.
Ifthepricelevelisallowedtovary,andwithrationalexpectations,monetarypolicycannotaffectthebehaviorofoutputunlessthemonetaryauthorityhasaninformationaladvantage,orequivalently,ifsomepricesarefixedbeforemonetarypolicydecisionsaremade3t.
Assumeas'IabstractherefromMundell—Tobinandothereffectsthroughwhichchangesinthegrowthrateofmoneyaffectrealvariablesevenwhenallmarketsclear.
17seemsrealisticthatthemonetaryauthoritycanreactafterwagesareset.
namodelwithISandLMcurves,plusanaggregatesupplycurveinwnicnoutputisanincreasingfunctionofthepricelevelrelativetotheexpectedpricelevel,resultssimilartoPoolesareoctainedwithrespecttoISandUIshocxs.
Therelativeimpactofasupplyshockunoermoneyanninterestraterulesoependsontneparametersinthenodeiwhanthelevelofoutptisrekvt:.
littleaffectedivpricelevelmovements,mona7stcctargxtin;xtxoiiicvsoutputrevtivetointerastritetargeting.
r4onx5NPTarjp_;AmongmonetaryrLestnatallowfor:nan;ingeiocic,anominalCNPrulehasreceivedconeidaranleattention*herulewouldspecifyatargetpathfornominalGNP,forinstanceonethatgrowsatU.
.
IfthetargetpathfornominalSMPispre—spetifiedforalltime,policyacceptsalinearandone—for—onetradeoffbetweenchangesinthepricesevelandoutput.
misimpliesforinstanceacceptanceoftheneedforarecessionof57.
ofrealSNPinthefaceofasupplysnocxtnatraisesthepricelevel5%.
ItisunliKelytnatsuchatradeoffwouldbeaccepcedftnecnoicewereputexplicitly.
rfonethelsss,supposethatatargetpstnfornominalSNPnasoxenspecified.
lonetarypolicywouldthenbechoseneachperiodtobringGNPascloseaspossibletotnespecifiedpath.
Givenlagsintneoperationomoney,andifcurrentmoneygrowthhassmalleffectsoncurrentnominaiSNP,suchapolicyislikelytoproduceinstrumentinstability,everincreasingfluctuationsinthemoneystock.
34ThenominalGNPrulewoulotnennaveno33TheseresultsaredevelopeoindetailinBlancnardandFischer1985,Chapter1))°3Forinstance,Bean(1983),Tayloril985;'lnstrumentinstabilitywouldnotbeaproblemifwehadfullconficsncsinourmodels,butwecanbesuretnattnemodelswillnotcontinuetodescroocrealityifthemoneysupplyfluctuatesmorethaniteverhashietorcaliy,becalculatecimposingsomecostsonvarationsinmoneygrowth,TaylorlB5evanineseconomicperformancewithnominalONEtargeting,empnasioingtheciffcuitiescausedDytnelaggeoresponsesohoutputandpricestopreviouspolicy.
Oncetheimpactsofpastshocks,includingtnelaggedeffectsotmonetarypolicy,arerecognized,analternativemodeohONEtargetingmayccemployedinwhichthemonetaryauthorityannouncespoliciescnataresvpsctsctobringthesconomybscvtoatargetpathgraduallyCrtheaimmhtbetoproduceagivengrowthrateohONEsacnyear,withpestdeviationsfromtargetforgiven.
ThesimplestmethocohcalculatingtherequiredgrowthrateofmoneyistouseaforecastingequationforvelocityccderivetnegrcwtnrateofmoneyimpliedbytheintermediatetargetlevelsofONE,ThiswasthepolicyhollowedcytheSundesbankfromiE7btolEfT,witntargetmcnaygrowthsetannuallyPischer,1968)NrelatedinterpretationisthatundernominalONEtargeting,tnemonetaryautnorityannounceseachperiodanominalONEtargetratherthanatargetgrowthrateofmoney.
WiththenominalONEtargetjustifiedinpublic,thisisapolicythatgivesthemonetaryauthoritytheciscreticntoadjustmoneygrowthinresponsetochangesinvelocitywithintheperioD.
Fromthecontroltheoryviewpointthischangemakesitpossibleinprincipletocomeclosertoachievingtargets.
Whetherthatwouldactuallyhappenwoulddependamthemonetaryauthorityssuccessatpredictingvelocityandtheententtowhichthegreaterdiscretionwouldenableittopursuectnerobjectives.
19Onthisinterpretation,inthecontextofasimplethree—equationmacroeconomicmodel——consistingof19andL.
McurvesplustheLucamsupplyfun:ticn——apolicythatfivesnominalStOPwithinaperiodcompletelyoffsetstheeffectsofdemandshockscsnocksenteringtheISandLhcurvesonoutputandprices.
Thusiftheultimategoalsofpolicyaretoveepoutputandthepricelevelatsometergetlevel,monetarypolicytnatsuccessfullytargetsnominalStOPisappropriatewhentheeconomyisaffectedcydemanosnocfsHowever,ooviously,anadversesupplyshockwouldraisetnepricelevelaboveitstargetlevelandreduceoutputbelowitstargetlevel.
)SiancnardanoFischer,1988,Chapter10),PriceuevelRules;AsinthecameofnominalStOPtargeting,apricelevelrulmcanbmviewedeitherasthespecificationoftheobjectiveofmonetarypolicyorthebasisforanoperatingrule.
TheFisner—SimonapricelevelruleisinSimonsmerelythespecificationofatargetforeonetarypoiicy.
Fisher1945,p.
25navestnespecificproposalthattheFedexpandthemoneymtocvwhenthepricelevelfallsbelowtargetandcontractwhenthepricelevelrisesabovetarget.
Thedynamicpropertiesofmuchapolicycannotimmvah.
latedwithoutanexplicitmodel;lagsintheoperationofmonetarypolcyraisethepossibilitytnepolicywouldactuallybedestabilizing,particularlybecausemonetarypolicyseemstooperatemoreslowlyonpricesthanonrealSNP,ThissuggeststhataFianer—SimonmrulecouldcausesignificantfluctuationsinrealStOPintryingtostasilizethepricelevel.
Fishers"compensatedsollar"proposal(1920,recentlyrevivedby'fall&l984),proposesthatthedollarieexchangeableintogold,butthattnevalueofgoldthatisexcnangedforadollarbefixedinrealterms,definedsaybytneOH.
Thenotionisthatthegoldstandardcheckonexcessissueofcurrencywcuiostillbeaveiaole,thesimplegoldstancarocrtericnofnairtiningconvertliiitywouldshillbethere,huhhoeseculareffectsocnxngeeintherelativepriceofgoldcntheaggregatepricelevelwoulddisappearTheproposalgiveetheFedasimpleruletofollow——maintainccnvertibiiity——ancappearshopromisepriceleveletaoiiityTnedifficultywitntheacnemeishostitsdynamicsarenotunoerstood,particularlywrebnertherewouldoedeshabil-icingspeculationagainstthestandard.
Fornstence,fthepriceofgoldwereadjustedmonthlyinaccordancewiththechangeirtheCR1,anticipatedcnengeeintheCPIwouldallowndvidualetospeculatebycuyingorsellinggoldinadvanceofthecnengenparity.
Frobaclytheimpositionofsuffcientiylargetransactionscostswouldreducetheevbentofmuonspeculation,butthedesirabilityandconsequencesofsuchtostsOCvCnotbeenevpiored.
Further,itisentirelyunclearnowtheschemeinduceethewagefievibilitythatmustbeneedediftheaggregatepricelevelishobeshade.
LQgJgggL_ELpij±gJNoneofthesealternativepolicyproposalsconnandswideassentwithintheprofesedon.
SinceLucasspolicyevaluationoritiqueyl97i,thereheebeennoacceptedwayofevaluatingdetailedpolicyproposals,Therehaebeenatendencytoevaluateproposalseitherinverysimpleeconometricmodels,whicharesetupasmuchfortractabilityasforrealism,orinverysimpletheorsticaleodele,aieomarkedmainlybyv5lntheintroductiontothe3,5.
editionofWebb(1R34,Jamesh.
RandJr.
claimsthisispreciselythepolicyfollowedinBritainafterts1Q31departurefroethegoldstandard.
36Thiaincludestheconstantgrowthraterule,whichrepresentsasignificantchangefroepreviouspracticeendthereforemaynducschangesineconomicstructureinunexpectedways.
21tractaoiiity.
Oneauchmodeitnatprovidesconsioerableinsightistnetnrse—equationmacronodalconsistingcrfISandhcurvesplusanaggregatesupplycurveHoweverfewappeartobeconvincedbysuchexercises——andtneyareright,tornowapolicywouldoperateinpracticeoependstoaconsiderableextentonthelagswithwhichpolicyaffectstherelevanttarget'ariat1asYetpolicyevaluationandFedpolicymakingcontinue.
EconomistswithevperienceconfidentlypronounceontneerrorsoftneFedsways.
TneFedcontinuestomakediscretionarypolicy,recentlywithconsiderablesuccessintermsofitsoojecties.
Eachisusingimplicitandsometimessvplicitmodels,ofconsiderablesophistication.
Thenaturalvehiclesforstudyingalternativepolicyrulesaretnelarge—scaleeconometricmodels,someofwnichnavemetthemarkettestocommercialsuccess,Givenaneconometricmoosl,anobjectivefunction,arccomputingability,optimalfeedbackrulesformonetarypolicycanbecalculated.
However,giventhevarietyendinadequaciesofexistingmodels,itwouldbedifficulttojustifysnshrininganyoftheserulesinlegislation.
Untilandifeveranewgenerationofmodelsthatmeetstnedemandingstandardsoftneprofessionisoevelopeo,therewillbenogenerallyacceptedprofessionalbasisfordiscussingsitermativepolicyrules.
fihatmonetarypoliciesshouldbeadoptedinthsmeantime'Ataminimum,itiscleartnatmonetarypolicyshouldadJustforpredictablechangesinvelocity.
Itmightbepossibletofindsimplefeedbackruissthatperformwellinavarietyofmodels,7andtorecommendthemasabasisfor'CooperandFischer(1972)foundfeedbackrulestnatreactedtothebehaviorofinflationandunemploymentwhichstabilizedoutputinbotntneSt.
LouisandMPGmodels.
monetarypolicy,ineycouldserveinthefirstinstancea;anindicatorofwhatmonetarypointyshouldbe.
Prudencewouldsuggestyear;ofpuolicandprofessionalciscussionbeforeanattemptwasmaoetoputmuchrule;nntoisgimlation.
Italsosuggeststhattheruleincludeprocedurestcrit;ownamendment,Frnedeanin19dBofferedhoemonetaryandfiscaldrameaorkverymuonamatentativeproposalforprofessionaldiscussion.
23:,jjmiclnconsiste_ncy;theBasicExample.
Until177,itappearedthatdiscretiondominatedrules,sinceanygcoorulecouldbeadoptedbyoiscretion.
7Tneconceptofdynamicinconsistency,oroughttomacroeconomicsintherulesversusdiscretioncontextnydlandanoPrescott,1977;completelycnangedtnedebate.
lynanicinconsistencyoccurswnenafuturepolicydecisionthatformspartofanoptimalplanformulatedataninitialdateisnolongeroptimalfromtneviewpointofalaterdata,eventnoughnonewinformationhasappearedintnemeantime.
Anexampleofdynamicinconsistencyduetonrescott1l977,oeveiopedinFiscner1l9G,istnatofoptimaltaxationinasystemwitncapital.
Underrationalexpectationsthesolutiongivestaxratestnatareootimaiconditionalontheirbeingexpectedbyprivateagents,ButoncecapitaliSinplace,itssupplyisinelasticandagovernmentactingtomaxmicetnewelfareoftnerepresentatiieindividualwouldtaxcapitalmoreneavily.
Theproblemisthatifthepublicexpecteotnegovernmenttoiclateitsannouncement,economicwelfarewouldoelowertnanifthegovernmentcoulocommittseltofollowingtnrougnonitspromisedtaxrate.
8'meapplicationtotherulesversusdiscretiondebatecomesfromtneclaimtnatpolicywillbedynamicallyconsistentifdeterminedbyrules.
By9Friedman;1972;hadarguedtnatapolicyaoopteooyrulewouldstaoiiioeprivatesectorevpectationsrelativetotnesamepolicycarrieoout0,oiscretion,Dottnebasisforthatargumentwasnotclear.
4lnthissection1drawfreelyonmy19Gbsurveyarticle;seealsoCukierman119851.
41freciselythesameproblemoccurswitntheoptimalinflationtaxandmoneyholdingICalvo,1978);themonetaryauthoritycanalwaysimposealump—sumtaxbydiscretelyincreasingthemoneysupplyand,oncetneprixwtesectorhasformedexpectations,istemptedtodoso.
contrast1agovernmentorcentralbankwithdiscretionmayunderrationalexpectationsteexpectedtomakethesnort—runoptimaldecisioneverytimeitcan,thereforegainsnothingfromitsopportunism,andonaverageproducesaworseoutcomethanwouldagovernmentabletotieitshands.
IntnieeectonIpresentesxnplePhillips—curveexampleofdynamicinconsistency,anddiscueetherelevancetotherulesversusdiscretiondeoatwoftheexample.
InthenextsectionIpresentextensionsthattakereputationintoaccount4-ALLc_ihAPP.
LhJSupposethatthepolicymixerhasasingleperiodlossfunctoonquadraticintherateofinflationim)andinthedeviationofrealoutput(yfromatargetlevel:Li;ramw(y—xy*P;axO,k>l.
kierey'canccinterpretedasfullemploymentoutput.
Thetargetlevelcfoutputexceedsthenaturalrate.
Theassumptionk>1iscrucial.
Thecostplausiblejustificationisthattaxdistortionscausethenaturalrateofemploymenttobetoolow.
ThatjustificationallowsthelosefunctionI.
)Itobeconsistentwiththesingleperiodutilityfunctionofprivateagents.
Analternativeviewisthatthegovernmenthasdifferenttastesthantheprivatesector.
43Inanyevent,dynamicinconsistencymayoccurwhetherornottheprivatesectorandthegovernmenthavethesametastes,42ThisstructurewasintroducedbyKydlandancPrescott)1977ianddevelopedbyBarroandGordon(1983,1983a),BackusandDriffill(1985,1985a;Canconeri(1985),Rogoff(1985)andothers.
4mMoresophisticatedtheoriesthatrecognizeheterogeneityinprivatesectortastesandthatseektogroundthegovernmentsobjectivefunctionintheelectoralprocesscouldprocuceautilityfunctionforsomegovernmentsthatwouldseektoraiseoutputabovethenaturalrate,Cukierman)1T85containeanestendeddiscussionofthispoint.
Theintrtemporaiosaunction,aciaountebsumoxtheformNl#6Imaymoreplausiblydi#eroetweentheprivatesectorandthegovernmentinasystemx4itnperiooicelectionsthattanendtielifeoftiegovernmentinthiscasetnegoverrmentmaybaQeashorterhorizonthantheprivateaector-rnexpectationalPhillipscurvedescribesthereatonsnipbetweenoutputandinflationeachperioo.
(5y=4-om—mwheremistheexpectedrate0finflation.
Considertirataone—periodgame.
Thepolicy—makersetstheinflationrate.
Underoiacretiontieexpectedinflationrateiatasenasgiven,implyingLate-iLoLik_ilyt4-bmJIfexpectationsarecorrect,tneinflationratewillbepoaitive,atthelevel(71m,aLoik_IlyswheresubscriptdrepresentsdiecretionM.
Notethattheinflationrateiahigherthelargerisb,andthusthegreatertheoutputgainfromunanticipatedinflation,thelargeristhedistortion(ix—liy*,andthesmallerisa(thelesscostlyisinfiationi.
Theimpliedvalueoftnelossfunctionunderdiscretionis181L=(k_l-y*2ll4ab2l261heequi1ibrunieevidentlyworse4crthegovernmenttandifithastnesameutility1unctcntheprivatesector)thanazeroinflationequilicriumThezeroinfiatonequilibrium,theprecomnitnentsolution,givesavalueoftneloseFunctionequaltoWhyintnisgamedoesthepolicy—makernotchooseaninflationrateofzero,therelyattainingLratterthanL7Undertherulesoftnegane,inwtcntheprivatesectorccmmiteitselfFirsttoagivenm,m=UisnctaNashequilicrium,Cncetheprivatesectorhascommitteditselftonthepolity—maserwillchoosethepositiverateofinflationimpliedcyTheinflationratelTpin17)isaNashequilibriumthatifexpecteccytneprivatesectorwillbeimplementedbythegcvernment""14thepclicy—maxsrcouicsomehowcommitherselftochccsingm=0,shecouldccteintnedistortevsecond—beetoutcomeU.
Fordiscussingreputetionelequilibriumwewantalsotocalculatetnainflationrateandvalueoftheutilityfunctioninthefoolingsolutioniflwhichindividualsexpectthepolicymakertocreatezeroinflaticncuttieinsteadactsopportunistically.
Withir=0,thecptmaldiscretionaryrateofinflationisfromfbI:4"tisalectheonlyNashequilibriue.
ItistemptingwhentalkingoFtneprivetesector"mcvimgfirst"totothimkcfitsettingitsexpectationstrategically.
Ifnweeaprivatesectorstrategicviriaele,itcouldoxsetatthevaluethatwouldfrom16)mducem0Acccrdngtotneaiqehnxthatwouldresultiny=kywandproduceafirstbeetsolution.
ButtnxtwcuidnotzeanequilioriumBecausemandrwouldbecifferent,inotnerwcrde,itieinconsistentczarguethatthepublicveeceratanegativenunzerincrcertoacnievem0't:',.
a'D(b—.
pSitntcorrsspordrg'4105oftnt1055tunctiOn5tti.
S44a0'Tnussucl.
a10.
,5ltdlIZ'tin'0pSflLjNotstnato,ais,mostly,aasaaaaofthsutilitygainfros.
nsspsctsoinflationagiv.
stntincrsassinoutputandathsutilityossfrosnignsrinflation.
vstins,'isistha4undassntasatofinsqualitisstonassonstratasti.
bsnsfitsofpr,cosaitsantsii;bap(.
5Tntaiscrstionarysol,t.
:npoducsstilt.
argsstloss,'ssultingasit005*rapositivsratsofintationwit'snooutputgair.
Tn.
rsfors,onsausps:'s,t'stpolicy—master.
so.
anitro;saazeroi,fati:nrats,to.
tt.
irRutbecauset:ss'.
.
nct,r.
513w5'metspisrnsantsuccasdeirt:3.
.
—;thapr.
vatssectortca's4flsritacts:o'sistsntly.
s.
i,tntgovsrsasrtistssptsoto,iolataipsctationsiftntpriiatssectormouldaslulledintosxpecting:sroinflation.
Instrivingtoobtainoutputgainsbyfoolirgto.
public,tiltgovernaantsuccssdsontyinraisingtntinflationratsaridproducingthsworstoftnttortsoutcomesin'13;.
Thsrsfors,ydlandandPrascottarguad,policy—saharashouldotconstrainSbyruls.
ThatwouldanablaUsatoattainHispracossittadsolution,adsittsdlynotUsbastpossibla,butbattarthantntdiscrationaryaltsrnativs.
28PreilpiCarDiscussion;HowpersuasiveisthierShouldweexpectdiscretionarypolicy—nakeraalwaystocnoosetheshort—runoptimalsoiutknn,ormighttreytakeintoaccounttheconeequenceaforfutureexpectationsofanycurrentdecisionstopursueshort—rungains'Beforepresentingnodeisw;threputation,webrieflyoiacusstnegeneralprobienofdynamicinocr]ciatencySocietiesdealroutinelyandcontinuouslywithsituation;invnicccynanicinconestencycouldoccurSoooinoividuais,nneaitninganeralandtnenationaldebtinparticulararestandinginvitat;onstosurprisetaxation,whichisrarelyeaplcitlyimposesImplicitsocialsacur;tyocl;gationsarehonoredandprotected.
Propertyrightaareprotectedc_flawanounosracocstobeessentialtoeconomicefficiencyinamarketenvironment.
Centralcaniswithdiscretionarypowerssuccessfullyrunlow—inflationpolicies;nseveralcountries,includingGermany,Japan,andSwitosrland,Thelaw,constitutionalorlessfundamental,isobviouslyon;snlutiuntothedynam;cinconsistencyproblem,Butnotallpotentialdynamicnoons;stancysituationsaredealtwithbythelaw.
Thisraisesthequastmcnsofwhicnissuesareandshouldbehandlesthroughthelawandwnicncydiscretionarypolicy—making.
Theabilitytodescribefuturecontingann;asfullynuscbeamimportamtelementinthischoice,4nEister11979)amdSchelling11984)arestimulatingreferences,dealinginpartaitnindivisualinconsistenciesandproblemsofself—control,"Thedynamicinconsistencyliteratureinmacroeconomicsnecbeanalmostexclusivelyconcernedwiththeallegedinflationarybiasofnacroeconomicpolicy.
SwitoerlandendSernanyaside,itisclearfromnistorytnatinflationarybiasisonlyasometinething.
AttneenosoftheNapoleonicandCivilWars,andWorldWarI,BritainandtheUnitedStatesdeflatedtogetbacktofiaedgoldparities,Theseepisodestoodeseraeactentoninthedynamicinconsistencyliterature.
Achallengefortnetheoryofdynamicinconsistencyistoeapla;nwhycountrieswereabletoinstitutenechan;snstosuppresstheirinflationarytiesuntiltheendofWorldaarI,anoccexplainwhytnebiasisworseinconecountriesthanothers,29potestiaUtf—aitfaiwaoftn.
itingaeottheconstitationtliewlaw—ruls—ois:'st.
onaqeistoviewpolic.
esasirvot.
ingatradsofbatweetis;ere.
tsofei'ibi1ityandti'scostsofoynaeicircorsisten:j.
Dependingor.
ti'soeiCythelegalsystsesf55anchoiceofthecoetsthatshouldOsattachedtoatteaptetoclangsit.
3iecrttionarpolicies,sunaseonetarjpolic4,canotciaigedattowcostsraesfinedbylawsic,aseucnoffiscalpo.
icyareclangeabeatgreatercost;rulesiadojconstititionallaws.
cvastierightsoprintsproperttorinterstatscoeeercsareirprisciplea&sochargeablebutatjetgreatercost.
Non—RulesS3pat.
ovei3r.
ema,o4reduciigtntinf'ationarysiasotntoasicmacph,seielopttOyogotft955,istoappointconservati'epohicysarers.
Supposesocietystossfsnctionis.
3ata.
oorbanner,osthepotic,enerSmeightoninf.
atzoninnitpersonaltossfunction.
Sivir.
;eucanindividaait.
iseiacretionresultsinaossforeocietjof141a:t'oe.
3LTnesorecorsereativetntpolicyeaner,thecloserthssocietycoeestoachievingtntp—ecoeeittedequ.
hibriue.
Thenotionofappointingconservativecentralbankersiscertainlysuggestiie.
Furtrerreasonstoappointtheerelatetothebenefitsofreputation,tobeesaeiiedbela.
notnernon—rulessoutioiistoputinpLaceincentivesforsuccesaorgovernmentstobenaveconsistently.
Thisistneapproachthatnatoeenfollowedinamodelin.
onnggove'neentdebtbutnotcapitalbyucaeandStoney(l9fliandii'aeodelincludinginadditioneonetarjpolicyoyPersscn,mRogoffslIS5;suggeststhIstradeoff,utichisenaeinedfartherinSectionVibelow.
CuciereanandMeltoer(l9Sbiincludssieilarconsiderationsintheiranalysisofagovernmentschoicebetweendiscretionandrules.
Perseon,andSvensson(IPES)ThegeneralprincipleexplainedbyPerasonandSversson)i984),istoplacethesucceseordecision—makersinasituattonwherethepenalty4croeviatingirontheprecommittedconsistentplanoaiatcesattnemarginthebenefitofdoingso.
Inprinciplesucharrangnentacanccnaceinanysituationswherethefullaetofstatesofnaturecantaapaufadenante.
TheLucas—StokeyandParseon—Evenesonsolutionraiseedelicatenasuas.
Dynamicinconsistencydoeenotdieeppear,Ratherthesolutionaesunesthatttegovernmentwillnotviolatecertainexplicitobligations,suchaxrepayingthedebt,eventnougnthereisanincentivetodoso.
Withoutatheorycirsoutationsuchsolutionsnavetoberegardedasincomplete.
Weturnnowtomodelswith!
reputation.
V.
DLnamicin:onaistencandPputationalEQuriibrius.
ThetasicPhillips—curveexampleinviteetnequestionofahetnarhnadeciaon—nakercannot,bybehavingconsistently,reachabetterrsauihtheytheone—perioddiscretionaryoutcome.
Perhapsbyshowinglorbearanca,investinginreputation,acentralbankcaminducetheprivatesectortobelievetnatitwillnotproduceunexpectedxnflation,Thereindeedexietreputationalequilibriainwhichtnemonetaryauthorityiseapectedandinducedtobehaveconsistentlysolongaxtcoasso.
Thekeytoanalyzingsuchequilibriaisthespecificationofprivatesectorexpectations.
Supposethatthehorizonisinfiniteandthatpolicy—makershavethelossfunction(l)iin(4)Denotetheinflationrate(7)aesociatedwihhdiacretionarf2o&;t;titieoreperiodproblemoyn,meinflationrateasaociateiaitntieprecommittedmonetarypoliis'tp=westartsit'expectationsoasedontnvprivatesectoraewintnepolicy—maseraseitnerreiabieoropportuniattc.
iftneinflationrateraeeverosenantnir;othertnantheprecommitteorateofzero,tneexpectedinflationratefrortenonelioentrim,:tnegovernenbhasnithertoproo'ucettieprecommitteoinflationrate,mI,ti5expecteOtocontinuecoir;so,whytracepa—titularevpectationsnTheywiiturnouttooejustz+eO,orrattonal.
Butaxnmanycases,theyarenottheonlyratioraexpectatiors,oringametheoryterminology,perfect,eguliiria.
Theproblemonutp;erationaiexpectaton5solutionstointertemporaigamesiSwellknown,Besoswegiveanotherexampleofaconsietentsetofexpeotatons.
ovantneseexpectations,consideragovernmenttnathasaleafsmehavetconxistsnts/,nowconeiderngshatnertocontinueproducingtentn4latonorwhetherrathertofoolthepublic.
Ifitcrests,itgainsinthatperiooc15}Temptation=L0—u,9L÷1÷9;Itthenhastopayforitscheatingmybeingexpectedtoproducethediscretionarysolutionforever.
Ifthatisenatisexpected,tnatisthebestthingforthegovernmenttobo.
'helossfromdiscretionarypolicyinoneperioorelativetotneprecommittedequilibriumiscibiLoss=L—L,32Notethatbotr!
thetemptationandthelossareincreasingxne,Tnegainfromactingopportunisticallysthenequaltothetemptationminustnepresentdiscountedvalueofthelosethatstartsaperiodlwterx3amfromopportunism=Temptatmon—LosefdLtB—(l+$)Itilleb,]Thegovernmentwillactopportunisticallyifitha;averynighciscountrate,anditwillthenceexpectedtoandwillcehavathatwwyIceverysucceedingperiod,itwillkeepthemnflationrateatzerodiscountrateiSloworiffmehigh.
Theroleofindeterminingwnetcartnegovrnnentkeep;inflationatoeroappearsparsooxicelinthatwhenhigh,thesnort—rungainfromunanticipatedinfiatmoni5nigh.
ButsinceootnthegainandtneIcesareincreasinginf,theneteffectisprmnriindeterminate,annoependsonthecurvatureoftheicesfunction,Notethatinthiscertaintysetting,areputstionalequilibriummxposamoleonlymfthehorizonisinfinite,Otherwisethegovernmentwouloonsureinthelastperiodtoproducethediscretionaryoutcomewhatevertneprivatesectorsexpectation,andworkingbackwardswouldccenpeccectuonthesameinthefirstperiod.
BerroandGordonilTB3aproduceamotnare'wnplaofareputationalequilibriuminthistypeofmodel.
Tnamrexpectationsassumptionisthatifthegovernmentfailstoproducetheanpecteomnfiatmnnratethisperiod,theprivetesectorexpectsthediscretionaryinfietmunracenextperiod;iftheyproducetheexpectedinflationratethisparioo,hne7areexpectedtodosoagainnextperiod.
3,Thenestquestioniswhetherazeroirletionritecanbeejetisnecasanequtibriau.
TielosefroaopportunismLastsonlyonep.
rioo,oef3retnegosenmeitregeinsc'eeioilityinsisfaceswinViesueDeCie4Dnitinintheinitiaperiod.
Thusthegainfromopportunisminthiscaseis,ueng15:eqsiteis3am$Teiptetmor—ass.
t$san,.
Z6—s.
'6'i'uiJ4ithanignomecuantratetngovernmentwillproeacenignerthanexpectedinflationinperiodore.
Inperiodtwoitproducesthediscretionaryrates.
Itregainstnsta,ooui;whetwasexpected,buttnenpromptlyviolatesitinperiodtn'ee.
hastieinitialsetofexpectationswerenotrational.
eitnaLowsiaco.
.
ntrateandhighIthegovernmentwiltproducezer:inletion.
'IAb/accidentitwiretoviolatethatrule,tiepuolicwonideap.
cts.
tooefo.
sowedoyze'oinflation,andthegove'nmentwoldindeedacttnatway.
Thuswith.
ow6Viemisusedexpectationsarerational.
Fortnenig'iScase,BarroandGordonareabletoshowthatthereiaanequilibriaifIv1,natmeonlyifthegovernmentisnottooiepatient.
If6.
1,itwiltgotothediscretionarysolution.
Ifitisnottooiepatient,theequilibriainflationrateinthisreputetinielequitibrlueliesDetainthezeroinflationthatwouldbeattainableunderprecoemiteentandCmthatoccursmthegovernmentisentirelyshort-sightee.
'nenatureoftheruleseq4ilisriueisthatthegovernmentwillcarr1outthe'alebecausethereisnoaevantegetonotdoingso.
Ifitshouldby49ntheDarro—Gordonaodelzeroinflationisnotanequilihrii.
e,becauseofei4ferigaseseptionsontneutilityfunctioo.
34miscalculationdeviate,thinitwillnextperiodimplemente.
giventhatitise*pectedtodosoItregainscredibilityandthereafterishappytoimplementtneruleagain.
Theequilibriaisperfect——thoughitisfarfroeunique,alBarro—Sordonrecognizeanddiscuss.
NultioleEcuilibriatPerhapsthereareeanypossibleequilibriaintherealworld,anditispureaccidentthataparticularsituationexists.
Nonetheless,itwouldbepreferableiftheorycouldnarrowdowntheringsofpossibilities.
Thedescriptionofthepriveteactor'sresponsetothegoverneentedeviationasapunishmentraisesthehopethatthedesignofanopticalpunisheantstrategysillreducetheaultiplicityofequilibria.
Butunlesstheprivatesectoristhoughtofasasingleunion,Stisdifficulttoconceivenowitcanselectenopticalpunisheentasopposedtooptimallycalculatingexpectations.
Themoreproeisingrouteprobablyliesinnrichingthedescriptionoftheenvironmentinwhichthepolicyeakersandtheprivatesectoroperate.
&RandoeizinoSoverneentiTabellini(1983,1985),BackusandDriffill(i985,i985a)andBarro(1985)applytheKreps—Wilson(1982)reputationeodeltotheinflationproblem.
Tabellini,andBackusandDriffillconsiderasonopolyunionsettingwagesinagamewithasonetaryauthority.
Thealternativeassumption,eadebyBarro,isthatprivsteagentsarehoeogeneousandnotengagedinstrategicconsiderationsvis—a—visthepolicy—eaker.
TheunionversuscentralbankgamecaybeappropriateforEurope,butintheUnitedStatescontextthenotionthatprivateagentscannotcombineegeinstthemonetaryauthorityiscoreattractivethanthealternative.
35Thehorizonisfinn,.
Thepublicbitevestnerearetwopossioetypico4policystar.
tntstrongandthewin.
Thestrongneverin4lates.
Theweak'itstiecase.
atilityfactionastoepublic,istiny.
temptedtoproduceu—anticipatedinflation,butbypretendirgtobestrongcanOuisdupareputationforstrength.
Theseaspolicg—eaxerpotentiallyengagesinasuedrandomizingstrategy.
picsingaprobabilityofactingtough.
oralternately,producinginflationineacnperiod,andlettingthedicedecidethepolicycnoice.
Ifinanyperiodtntdicesasehisactweak,thepb&acunderstandsneisweak,andineachsuesequentperiodheobtainsartythediscretionaryoutcome.
Thegeneralformoftnesolutionisasfollows.
"Foralong—horizonproblemthepolicy—makerwillstartoutnotrandomizingatallandotproducinginflationitniswilLseephisrepi.
tationunchangedi.
Becauseprivateagentsareuncertainofthepolicy—makerstype,inflationisoeiowitsexpectedlevelallthistime,causingasmallrecession.
Eventaallytoeendbecoons,andthepolicy—eakerbeginstorandoeize.
Duringtnisperiodhisreputationisimprovingandtheprobabilityofplayingstrongisfaliirg.
Thentowardstheend,maybeonlyinthelastperiod,heinflatesforsa'e.
Oneresultthatemergeefromthieframeworktethatasthehorizongoestoinfinity,andprovidedthediscountrateisreasonable,thereputatienalquilibriuewithzeroinflationIcattained.
Thereasoningissimilartothatebsveithepenaltyforrevealingyourweaknesdisaverylongperiodofinferiorperformance.
"Atwo—period.
saepleisworkedoutinfulldetailinFischertl9Bo.
.
Neithertheelegancenorthesuggesttveneeeofthetraps—Wilsonconstructoarbedented,Buttheanalysis,byt000singentirelyortheweakpolicy—makerwrohasmadeitthroughwithoutirflattngdrawsattentionawajIronthetmplausibtlityoftneunderlyingviewofthepolicy-makersactxcns.
Itisdilficuictobelieveamodelofreputationtnwhichacentralturkcreatestrviattonbecausethedtcefellonewayratherthanarother5z;WhileFedpoltcythatconditionsonthestockmarketortheexchangerutstsinoutcomestocnasttc,theFedisrotinthosecasespurelyranoonicing.
yj,Flexicility1mbvie5amotelsoftheprecedingtwosectionsappeartoprovettecieardomiranteofrulesoverdiscretion;reputationalmodelsstowthatp011:/makersunderdiscretionmayoutwillotnecessariiy,probcetneoptimaloutcometnatrulesensure.
howevertreissueisnotclosed.
Aoaa,cargumertforactivistpolicyvsthatthepolity—taxers;annandiecertainoisturoancesmoreflexiblyandmorecreaplycnancanmyrvadprivateagents.
crnstancetmeretsnogoodreasonwhyaahftintnedemand4ormoneysnouldbetransmittedtoprvcescausingalleconomicagentstoadvstpricesandsageswhenthemoney—treatinga4tborltlestanrespondinstead,Oneoftnemostimportantarguments4ordiscretionarypolicyisthatitleavesthepolicy—makertheflexibilitytorespondrapvolytocontingenciesnotforeseenornotdescribableintnepotentialrule.
Supposethatadisturbanceisooserveobythepolicysaxereacnperiod,afterprivatesectorexpectationshavebeendetermined,andttattnenatureofthedisturoancecannotbedescribedinthemonetaryrule.
wegeneralicethesupplyfunctionflto:i19}yyeable—e)+aHereseadisturbancewithexpectationosro,tnatsnotaeriallycorrelatedandtnatisnotknowntoprivateagentswhentheynasatneirwagedecisions.
ifedonotshowtinesubscripts.
Denotethevaroanceofcbyb.
ThesociallossfunctionisnowtheexpectationofLI1inepuation35flhisexampleiscloselyrelatedtotheanalysisbyRogoffvi98535ThisonetaryauthorityisInapositiontorespondtorealizationsofa,butr,representingwags—sitting,isdeterminedbsforeasmown.
menisnopre—coesiteentandnoconsiderationofreputation,sothatthediscretionarysolutionischosenitchperiod.
Theinflitionriteis(20)s(a.
o)bC(k—i)yior—a]implying5isthesassasundercertainty,namelyr(b/aflk—flysand(20)e(b/a)(k—i)yi—(atb')'bsInthissolutiontheaonetaryauthorityrespondstosupplyshocks,allowingthistoiffectbothoutputandinflationsanadversesupplyshockbothraisestheinflationratsandreducesoutputbelowthenaturalvate.
tZTheexplitedvalueofthelossfunctionundertheseconditionsiscalculatedas(21)LIt)(it,)th_l)ayea(l$)te)SIba/aSupposealternativelythattheeonetaryauthorityhadnodiscretionandthatthemoneysupplywasheldrigidlyconstant.
Accusealsothatthequantitytheoryholds,with(22)esya.
p$ywhenpisthe(logarithmofthe)pricelevel,andtheaoneystockIssetatthslevelwhichisexpectedtoproducep0.
Suppos.
thatlastperiodthe(logaritheofthe)pricelevelwaszero.
Theexpectedpricelevelandinflationratethisperiodarealsozero.
°Iaeaccusingthatthetargetlevelofoutputdoesnotchangewiththesupplyshock.
Thatassuaptiondoesnotaffectthebasicpointbeingmadehere.
TheexpectedvalueofthelossfunctionunderaconstantmoneyrulecantenoemnosntooeEslZy.
kital#bTnefirsttermi5largerunderdiscretionrefvertirgtneiasi:oytsmicinconsistencyresult.
Thesecondtermislargerundertneconstantmoneyrule,reflectingtnecenefitstosocietyofflexiblemonetarypolicy.
Thereisttusabasictradeoffbetweenthegainsromdynamicconsistencyandtnelossufflexibilityznimposingamonetaryrule.
tneextentthatthecentrabanknsaalongerhorizonthanoneperiod——anothismayisonereasontnattnelawattemptstoisolatecentraloanxmanagement'—ompoliticalpressures——itmaybeaoietoestablishareputat:onthatservestnesamepurposeasamonetaryrule.
VII.
Confljg_Conments,Incrulesversusdiscretiondebateinmonetarypolicyisatleastl50yearsold.
therehasinthattimebeennomonetarysystemwhicnoperatedwithouttheexerciseofsuostantialdiscretionaryauthority——tobesuremoresoatsometimes,suchasafterthecollapseofthegoldstandard,amdinsomecountries,thanothers.
Thepre—1977argueentsofprincipleforruleslackedanyconvincingdemonstrationthatrulesmigntsysteeaticallybebetterthandiscretion.
thatdemonstrationcamewiththedynamicinconsistencyliterature,however,giventnepossiblebenefitsofthefiesibilityofmonetarypolicyunderdiscretion,andtheroleofreputation,thedynamicinconsistencyliteraturedoesnotestablishthesuperiorityofrules.
4)Inthinkingacoutmonetarypolicysnorules,ii15usefultodiscusswhothemonetaryoclicymakerswillbe3.
Forconcreteness,consicertneexampleoftneUnitedStates.
Atoneextremethedecisiononmonetarypolicycouldbemaceonthebasisofcurrentknowledgeandenshrinedintheconstitution.
Thatisunlikelytohappen,giventhedifficultiesofaner!
din;theconstitution.
Nor,giventheuncertaintiesovermonetarypolicy,shouldseriouseconomistsargueforsuchanamendment,itsurelyillbehoovesaprofessionthatcompletelyfailedtoanticipatethevariaoiiityofrealexchangeratesunderafloatingexchangeratesystemtobelievethattiscapable0fspecifyingamonetarypolicythatshoulobechangedonlytnroughtnetortuousprocessofconstitutionalamendment.
ThechoiceintheUnitecStatesisthusnotbetweenamonetarypolicydsternineobyruleforalltime,anddiscretion,butbetweenamonetarypolicyspecifiedbytheCongrsvsxnuonechosenbytheFed,CurrentU.
S.
fiscalpolicydoesnotsuggesttnattnsCongresswoulodoabetterjobofchoosingmonetarypolicythantneFed,thoughtnatisnottosaythattheFedcannotdobetter.
szIncommentingonanearlierdraftofthischapter,MiltonFriedmanstated"ThemajorcommentistheomissionofwhatIhaveincreasinglyconetoregardashamletonthisissueCruiesversusdiscretion),namelythepublicchoiceperspective.
Toillustrate,.
.
.
youtalkaboutalossfunctionforthepolicymakerchatincludeseoielbinflationanocxvdeviationofrealouputfromatargetlevel.
Ifwebringthisdowntoearth,thesearelikelytobeonlyveryinoirectlyrelatedtotherealobjectivesoftheactualpolicymakers.
Fromrevealedpreference,Isuspectthatbyfarandawaytnetwomostimportantvariablesintneirlossfunctionareavoidingaccountabilityontheonehandandacnavnqpublicprestigeontheother,AlossfunctionthatcontainstncsetwoelementsasitsmainargumentwillIbelievecomefarclosertoracionalioingthebehavioroftheFederalReserveoverthepast73yearsthanonesuchasyouhaveused.
"41TheChicagoschool5.
phasisonrates'srsusdiscretionwas.
istsading,isNilitsesphisisontnedesirao:tztjofrulesisscn.
neroisacrtiru.
.
sofmonetarypolicies,seesglvirgsoreaiscritiontothecentrszDartthinothers.
ItisdiffIculttoattachsuc'ivirtuetoarate,merelybeciaseitiseruts,ifitproducespooreconomicperformance.
Accordingly,sore#alsabasthinti'srutesversa.
seiec'etiondsflteistiesuostintiediscassionofalternativeeonetarypoliciesthataccompanied.
t.
Thatdiscussionhassputteredsincebeingoeriiledoytheeconoeetricevaluationcritique,hatistooimportantto01suppresseeeachlonger.
42REFERENCES.
Andersen,LeonallC.
andKeithM.
Carison11970).
"AMonetarcatModelforEconomicStabilization",(April),7—25.
Backue,DavidandJohnDriffill(1985)"RationalExpectationsandPciicyCredibilityFollowingaChangeiflRegime",ReviewofEconomicStudies,52,2(April),21l—222(IRESa).
"InflationandReputation",Review,75,5(June),530—538,Sagehot,Walter)1906).
Loflar.
London:Kegan,Paul,Trench,TrubnerSerro,RobertJ.
(1978).
"UnanticipatedMoney,Output,andthePricaLevelntheUnitedStates",LnalojPoliticjLEconon,(August),549—SBL.
(1985),"ReputationinaModelofMonetaryPolicywithincompleteinformation",manuscript,UniversityofRochester.
andDavidGordon(1983).
"APositiveTheoryofMonetaryPolicyinaNaturalRateModel",Journalo+PoliticalEconpy,91,4)Mug,569—610.
)1983a).
"Rules,DiscretionandReputationinaModelofMonetaryPolicy",JournalofMonetaryEconomics,12,1)July,101—122,Ssan,CharlesP.
)1983).
"TargetingNominalIncome:AnAppraisal",EconomicJournal,soa-819.
Eianchard,OliviarJ.
andStanleyFischer)1988).
LecturesinMacroeconomics,forthcoming,MITPress.
Bloonfield,Arthur1.
11959).
Monetay_Po1icyy.
pdar_thQJattjpLgStandard:1880—1914.
Newyork:FederalReserveSank.
Bordo,MichaelD.
andAnnaJ,Schwartz(1984).
AClassicalGoldstandard,1821—1931,UniversityofChicagoPress.
Brainard,William11967).
"UncertaintyandtheEffectivenessofPolicy",Proceeglagi(May),411-425.
Calvo,Guillermo(1978).
"OntheTimeConsistencyofOptimalPolicyinaMonetaryEconomy",Econometrca,46,6(Nov.
),1411—1426.
Canooneri,Matthew8.
11985).
"MonetaryPolicyGamesandtheRoleofPrivateInformation",AmericanEconomicReview,75,5)Dec.
,iOSO—1070,43Coope',J.
PhillipannStanleyFiscner'l9i).
"StochasticSimulatioro"MonetaryRulesinTwoMacroeconometricMods.
s,JournaloftnemericanStatist:caiAssociation,(Dec.
750—781.
.
MonetaryandFiscalPolicyinthePAlyStochastccSt.
LouisModel",JournalofMoney1CrezitandBarain;,Fee1—iC.
Cukiermar,vea.
1985;.
"CentralBan;BenaviorandCredibility——SoneRecentDeveloomentenaruscript,TelAvivUniversity.
_noAllanH.
Meltcar)19861.
APositive'TheoryofDiscretorary°olcy,theCostofDemocraticGovernmentanntneBenefitsofaConetitton",4ortncominginEconomicinquiry.
Diamond,Pete'A,lRB5.
.
ignoranceandMonetaryPolicy,m:meo,M.
I.
T.
Dutton,Jorn1754.
InsBankofEnglandandtneRulesoftheGameundertneInternationalGobsStandard",inBordoanoScnwartc,ceditors,oq.
cit.
Eichengreen,Barry1985)editor).
ThiQQLLiAQardinLftrLA4Hist[y,NewYork:Metnuen.
Eleter,Jonl979,.
andtheSirens.
CamoridgeJniveretyFree;.
Fetter,FrankA.
iRbS(.
theDevE1qkrnentofBritienMonetarOrtnodoat.
Camoriage,Mass:MarvaroUniversityPress,Fiscner,Stanley,15).
DynamicInconsistency,CooperationandtheBenevolentDissemblingGovernment",JournalofEconomi_p_jpjpeandControl,2,1(Feb.
;,93—107.
;l986.
TimeConsistentMonetaryandFiscalPolicies:ASurvey",mimeo,M.
I.
T.
MonetaryPolicyandPerformanceintheU.
S.
,JapanandEurope,1973—198;,forthcomingnMonetaryandEconomicStudies.
and3.
PhillipCooper(1973).
"StabclizationPolicyand.
,age',Journalo4PoliticalEconoey,;Juiy/Aug.
)947—877.
Fisher,Irving;1920.
StabilizingtheDollar.
NewYorkMacmillan,lOOtMax.
.
Newhaven:CityPrintingCompany.
Friedman,Milton(1948).
"AMonetaryandFiscalFrameworxforEconomccStability",reprintedinMiltonFriedman,EssaysinPositiveEconomics,ChicagoUniversityPress,1953.
44—(1959),AProj!
mfpr±onetarLStabthtLNewYork;ForchamUniversityPress'19b2).
"ShouldThereBeArIndependentMonetaryAuthority",inLelandB.
Yeagsr(editor)InSearchofaMonej43jConstitutionCambridge!
Mass;karvardUniversityPress.
"TheLaginEffectofMonetaryPolicy",inMiltonFriedman,The_OptimumtityofMong.
Chicago;Aldine.
(1972)"TheCaseforaMonetaryRuleNewsweek,Feoruary7,p67.
andAnnaJ.
Schaartz(1963),AMonetarBtpfygjStQ±hAI!
ISt_jLjBt9bUPrincetonUniversityPress.
Goodhart,CharlesA.
E.
(1972).
LAQiDspIAAg_n1.
London;WeidenfeldandNicholson.
Hall,RobertE.
(1983).
"MonetaryStrategywithanElasticPriceStandard",inPrice_StabjjjandjlkiIcRolic',KansasCity;Fedsr;iReserveBank,37—l59.
Hart,Albert3.
(1952).
"TheChicagoPlanofMonetaryReform",itFM.
LutzandL.
W.
Mints(editors),ReadingsinMonetaryTheoy.
L,ondonUsorgsAlienandUnwin.
Hawtrey,RalphS.
(19b2).
ACeiIjInkPate.
Newyork;AugustusM.
Kelleyreprintseries.
Holbrook,R.
S.
i1972(.
"OptimalEconomicPolicyandtheProblemofInstrumentlnstaoility",AmericanEconomicReview,(March),57—iS.
Kreps,David,andRobertWilson(1982).
"ReputationandImperfectCompetition",Journ!
L_cj_jc&.
fl2!
LcJQg.
L,27,2(Aug(,253—279.
Kydland,FinnE.
andEdwardC.
Prescott,(1977)"PulseRatherthanDiscretion;TheInconsistencyofOptimalPlans",JournalofPolitcaiEcggj.
y,85,3(June),473—492.
Lucas,RobertE.
(1976).
"EconometricPolicyEvaluation;ACritique",rCarnegie—RochesterConferenceSeriesonRublcPolicy,Vol.
1,ThePhiljJR!
CurveandLaborMarkets,l9—4b.
andNancyL.
Stokey(1983).
"OptimalFiscalandMonetaryPoloyonanEconomywithoutCapital",JournalofMmnetar,12,1(July),55—94.
McCallum,BennettT.
(l98l("PriceLevelDeterminacywithanInterestRateRuleandRationalExpectations",JournalofMonetaryEoonomios,8,F(Nov),319—329.
Mints,uioca,1945'.
AHistoryofBankingTheory,UniversityoLncag0Press.
MonetaLy_pjjcLjoraompetitiveocI!
tL.
'iewor:McGraw-4i'.
l.
Mishkin,Freoerc3.
1483i.
Macroeconomics.
UniversityofChicagoPress.
Morgan-Weob,Oharles(934;.
L9151LEALLVewyork:Macmilan,Nurkse,Pagnar'1544;.
ntL3inLQLQR1_91P(flac1LeagueohNationsArnoess,1973;.
Persson,Torstensnd,arsE.
0.
Svensson,;15S4).
"Time—ConsistertF.
scw(PolicyandOo,ernmentOasn—Fiow,JournalofMonetaryEconomics,14,7;Nov.
I35—4.
Persson,Mats,torsten°erssonannLarsE.
0.
Svensson1985i.
TimeConsistencyofFiscalandtionetary°oiicy",SeminarPaper*3C1,InstituteforInternationalEconomccStudies,Stocxnolm.
Poole,william.
1970.
OptimalChocceofMonetaryPolicyInstruments,naSimpleStncnasticMacroModel,QQarIro±Lf_919a91L51'34iMay),19T—ilo.
Prescott.
EdwardC.
1977).
ShouldControlTheoryBeUsedforEconomccStabiiization',inCarnegce—RochesterConferenceSeriesonPublicclccy,Vol.
7,QptimaRogoff,Kennetn'1985;.
"TheOptimalDegreeofCommitmenttoanIntermediateMonetaryTarget",j.
ffk1jRurn!
i_2tEc2a2.
!
ic5,100,4(Nov.
,llDS—l190.
Sayers,Richard3.
1958;.
Cent311__8ankftg9fflJ.
!
fl0P1.
Oxford:ClarendonPress.
Schelling,ThomasC.
(1984).
ChoiceCor,sgje,HarvardUniversityPress.
Selden,Richard7.
11932).
"StableMonetaryGrowth",inLelandB.
ieager,ed.
,InSearchofaMo5Ljcg3Jti9O.
Cambrioge,Mass:narvardUmiversityPress.
Simon;,HenryC.
)1948.
Ecogjg_PolicyforaFreeSociett.
universityofChicagoPress.
Tabellini,Guido(1983).
"AccommodativeMonetaryPolicyandCentralBaniReputation",manuscript,UniversityofCalifornia,LosAngeles.
46—)i985).
"CentralizedWageSettingandMonetaryPolicyinaReputatiomalEpuiltbriun",manuscript,UniversityoCuliiornia1LosAngeles.
Taylor,Jonn8,(1985)!
"WhetWouldNomcnalSNPTargettingDototheBusinessCycle"inCarnegie—RochesterConferenceSeriesonPublicPoiicyVol22ftn4!
LQLfli_QGfl4fliflLSfl,61-84.
"1ner,Jacob(1955).
StudiesintheTheoryofinternationalTradeLondon:George,Allen,anoUnwin.
Warourton,Clark(1966),Qressiojnfistion!
p4MonetarLPDiipj.
Baltimore:JohnsHopkinsPress.
LL8_EiLLLigntsLLftkaortsontionalPjnancs,TheConli+feCommitteeandtneMecmllanCommitteeReports.
tArnoPress,19Th)