Speech

Central Bank Transparency: Why and How? “HowTransparentShouldaCentralBankBe?” ThePhiladelphiaFedPolicyForum FederalReserveBankofPhiladelphia Philadelphia,Pennsylvania November30,2001 Since arriving at the St. Louis Fed, I’ve officialsinapoliticalsetting.Everydemocratic been asked a number of times what I societystrugglesinthepurgatorybetweenthe foundtobemybiggestchallenge.Onthe idealofgovernmentresponsivetocitizenconcerns monetary policy side of my responsibil- andthepracticeofamessyprocessresponsive ities, I reply without hesitation that my biggest towell-financedadvocates.Weareinstinctively challengehasbeentoexplainpolicytoanumber suspiciousofsecretdealsbecausewefearthat of different audiences. Given that experience, I theywillcomeattaxpayerexpenseorwillbenebelieve that the transparency issue is not posed fitundeservingspecialinterests.Theexperience properly by the title of this session, “How ofourdemocracy,andthatofotherdemocracies, Transparent Should a Central Bank Be?”; the demonstratesbeyondanydoubtthatthefearof issue for me is how in fact to be transparent and secretdealsisjustified.Webelievethatpolicy what being transparent really means. decisionsinthepublicinterestratherthaninthe I’lldividemyremarksintotwosections.First, privateinterestwillleadtoahealthierandfairer I’lldiscusswhyIthinktransparencyisimportant. society. Second,I’llconsidertheproblemofdeciding Ihastentoadd,however,thatasfarasIknow whatactionsandstatementsmightimprove specialinterestmotivationshavenothadan transparencyandhowwemightmeasurethe importantbearingonU.S.monetarypolicy. successofstepstobecomemoretransparent.I Becausemonetarypolicyisageneralizedpolicy mustemphasize,however,thatIdonothavesetinstrument,ithasinherentlylittlescopeforprotledviewsonthissubjectandbelievethatthere vidingspecialbenefitstonarroweconomicinterismuchroomforconstructivedevelopmentin ests.Idiscussedsomeoftheseissuesatgreater thisarea.Thetopicisalmostunexploredatan lengthseveralyearsagoinalectureentitled,“The empiricallevel. FederalReserveasaDemocraticInstitution.” Beforeproceeding,Iwanttoemphasizethat Transparencyinageneralsensesimplymeans theviewsIexpressherearemineanddonot providingthefullestexplanationpossibleof necessarilyreflectofficialpositionsoftheFederal policyactionsandtheconsiderationsunderlying ReserveSystem.Ithankmycolleaguesatthe them,inastimelyamanneraspossible.Anadvan- FederalReserveBankofSt.Louis,especially tageoftransparencyinthissense,familiarto BobRasche,fortheircomments,butIretainfull everyteacherandresearcher,isthatthebestway responsibilityforerrors. tobesureyouunderstandanissueistoexplain ittoothersinaclass,ajournalarticle,alecture, orinmeetingminutes.Transparencyisagreat WHY TRANSPARENCY? spurtodevelopingcoherentviews,andsurelyit Mostdiscussionsabouttransparencyare isbeneficialtopolicymakerstobecoherentin conditionedbynotionsofaccountabilityofpublic theirownthinking. 1

MONETARYPOLICYANDINFLATION Iremembermyconcerninyearspast,before marketconfidence.Thoseexpectationswillbe assumingmycurrentposition,thatrefusalof moreaccurateandconfidencemorecomplete publicofficialstoexplaintheiractionsoften themoreaccuratelythemarketunderstandswhat reflectedtheirownsensethattheirviewswere theFedisdoing. notverywellformed,ratherthananeedtohide Iknowofnofindinginthemacroeconomics unsavorymotivations.Ipersonallybelievethat literaturethatprovidesatheoreticalcase,or mycurrentsenseofobligationtoexplainmy empiricalsupport,fortheviewthatconfusionor viewspubliclyisasmuch,ormore,abenefitto uncertaintyintheprivatesectoraboutthedirecmeastomylistenersandreaders.Iregularlyuse tionofmonetarypolicyservestobetterachieve myspeechesasanopportunitytodigintoaspe- policyobjectives.Indeed,intherationalexpeccificpolicyissueandtoforcemyselftodevelop tationsliterature,inaccurateexpectationsabout anopiniononamatterIreallyoughttounderstand policyunambiguouslycreateinefficienciesinthe thoroughly. economy.Itakethispointseriously,andbelieve Mostdiscussionsoftransparencydonotseem thatitprovidesacompellingcaseforpolicymaktogomuchbeyondafaiththat“governmentin erstoprovideasmuchinformationaboutpolicy thesunshine”willyieldthedesiredresults.How- astheypossiblycan.ButIdowanttoreemphasize ever,naiveconfidencein“sunshine”issurely thatanaiveapproachtodisclosurecandamage inadequate.Oneissueisthatpolicymeetingsin thedeliberativeprocesswithoutprovidinginforthesunshinearesimplydifferentmeetingsthan mationofgenuinevaluetothemarkets. onesheldconfidentially.Adifferentmeetingin thesunshineisnotnecessaryabettermeeting fromtheperspectiveofservingthepublicinterest HOW TO ACHIEVE well. TRANSPARENCY? Sunshineadvocatesexpectsunshinemeetings todifferfromclosedmeetingsinthatpolicy- FromwhatI’vejustsaid,theanswerto“Why makersareassumednottocutdealscontraryto transparency?”isthatwehopetoachievebetter thepublicinterestwhentheymeetinthesun- policyresults.Buthowtoachievegreatertransshine.Idonotknowofastudyofthesubject,but parency,orwhetheranyparticularstepwillbe mycasualimpressionisthatthereisnoevidence constructive,isfarfromsimple. thatthereislessspecialinterestlegislationnow ConsiderthepossibilityoftelevisingFOMC thanthereusedtobebeforeenactmentofvarious meetings,soallcanobservetheproceedings. piecesofsunshinelegislation,suchastheFreedom Oneissueisthemismatchbetweenthetechnical ofInformationAct. levelofthemeetingandtheknowledgeofthe Fromaneconomist’sperspective,theissue audience.IfIreallywanttoconveyinformation of“Whytransparency?”seemsprettystraight- onaparticularsubjecteffectively,I’llgiveadifferforward.Putsimply,policyactionsaffectthe entlecturetoafreshmaneconomicsclassthan economythroughexpectationsinthefinancial toagraduateseminar.Monetarypolicyneedsto marketsandintheeconomymoregenerally. beconductedatthehighestpossibletechnical WhattheFOMCdoesmostdirectlyistosetatar- level;ageneralaudienceismorelikelytobecongetforthefederalfundsrate.Expectationsabout fusedthanenlightenedbywatchinganFOMC thelevelofthattargetinthefuturedetermine meetinglive.Mostviewerswouldgetlittleoutof interestratesallalongtheyieldcurve.Expecta- watchingadiscussionoftechnicaleconometric tionsabouthowtheFedwillrespondtovarious issues—whichdoarisefromtimetotimein possibleeventsdetermine,amongotherthings, FOMCmeetings—andmightwellmisinterpret theexpectedinflationrate.Assumingthatthe suchdiscussion.Perhapsweshouldn’tworry Fedhassoundpolicyobjectives,itssuccess toomuchaboutthisissue,astheaudienceforan dependscriticallyonmarketexpectationsand FOMCmeetingwouldprobablybeprettysmall 2

CentralBankTransparency:WhyandHow? afterafewsuchepisodes.Idonotthinkwewould FOMCdeliberationsareextremelyopenandthat competeverysuccessfullywithdaytimetelevision! issuesarethoroughlyexplored.Idonotthink Ofcourse,atelevisedFOMCmeetingwould thatdisclosingthetranscriptwithafive-yearlag notbethesamemeetingweholdtoday.Forone inhibitsmydiscussion,andbelievethattobethe thing,thedeliberationscouldnotincludedis- caseformostothermembersaswell.Ialsobelieve cussionofinformationobtainedunderpledgeof thatthetranscriptprovidesavaluablerecordfor confidentiality.Informationfromindividualfirms scholarsandIstronglysupportthecurrentsystem doesplayausefulroleinpolicymaking,andthe ofreleasinglightlyeditedtranscripts. Fedcouldnotobtainsuchinformationwithout So,myanswertowhytransparencyisthat maintainingitsconfidentiality.Moreover,there weexpectbetterpublicpolicyoutcomesfroma isampleevidencethatpeopleintelevisedmeet- transparentprocess,butthemoreIconsiderhow ingsbehavedifferentlythanthoseinclosedmeet- transparency,themoredifficulttheissueseems. ings.Someparticipantsmighthaveatendency ThecurrentsystemofreleasingtheFOMCtrantograndstandfortheaudienceandtoavoiddis- scriptwithafive-yearlagworkswell.Idonot cussingdifficultorcontroversialissues.Ina believethatmonetarypolicyactionsmotivated completelyopenmeeting,itishardforsome byspecialinterestswereaprobleminthefirst participantstotryoutideasandthenadmitthat place,andsothetranscriptshavenothadabeartheargumentsofferedbyothersarebetter.Itis ingonthatissue.Butthetranscriptshaveserved particularlydifficulttoanalyzeunpleasantpos- toenhanceknowledgeofthepolicydecision sibilitiesinpublic,suchasthataparticularpolicy processandthathasbeenconstructive. actionmighthavetheeffectofincreasingtherisk Probablymoreimportantthanthetranscripts ofrecession. inrecentyearswasthedecisioninFebruary1994 Anyonewhohasheldapolicymakingposition toreleasethepolicydecisionpromptlyattheconknowshowimportantconfidentialityissuesare. clusionofeachFOMCmeeting.Mycolleaguesin TheFederalReservefacedasuddenandsevere theSt.LouisFedresearchdivisionandIhave confrontationwiththisissueinOctober1993, conductedsomeresearchontheeffectsofthat wheninformationconcerningthelong-standing step.Thebottomlinefromthisresearchisthat practiceoftapingFOMCmeetingsbecamepublic. promptdisclosureofpolicyactionssignificantly MostFOMCmemberswereunawarethatthe improvedtheaccuracyofmarketforecastsof meetingswerebeingtaped;thepublishedtran- policyactions. scriptoftheFOMCconferencecallofOctober15, TheFOMCbegandisclosingitsso-called 1993,demonstratesvividlytheanguishanduncer- “bias”in1999andlaterrevampedthelanguage taintyabouthowtoproceed.Itisclearfromthat ofthatstatement.Moreover,disclosureofthe transcriptthatthememberswereunanimousin policyactionisnowaccompaniedbyashort believingthatquickreleaseofmeetingtranscripts statement.Ihavenotexaminedempiricallythe wouldseverelyinhibitfreeexchangeofideas effectivenessofthesestepsandhaveanopen andinformationanddamagethedeliberative mindastowhethertheyhaveinfactimproved process.Manymemberswereconcernedthat publicunderstandingofpolicyactionsorthe releaseoftranscriptsevenafteralagofthreeto accuracyofmarketforecastsoffuturepolicy fiveyearswouldalsobeharmful. actions.Ithinkitimportanttomaintainanempiri- I’mtold—buthavenofirsthandknowledge— calperspectiveontheseissues,forthereisno thatafterthetapingofFOMCmeetingsbecame otherwaytodecidewhetheraparticularstepin publiclyknown,themeetingschangedsignifi- thenameoftransparencyisinfacteffective. cantlyforatime.Membersreadfromprepared Differencesincentralbankdisclosurepracstatementsandgive-and-takediscussiondidnot ticesofferagreatopportunitytoconductresearch occurasreadilyasithadpreviously.Duringthe onhowbesttopursueanefforttoincreasetranstimeI’vebeeninSt.Louis,myimpressionisthat parency.Manypoliciesandpracticesdifferamong 3

MONETARYPOLICYANDINFLATION theworld’scentralbanks.Ifwearetomakegenuineprogressondeterminingwhatworksand whatdoesn’twithregardtodisclosure,weneed todigintotheresearchopportunitypresentedby thesedifferentdisclosurepractices. I’llfinishbymentioningtwospecificideas forimprovingFedtransparency.Oneisthatthe Fedcouldannounceanexplicitinflationobjective,expressedeitherasapointtargetoratarget range.Ihavelongbelievedthatsuchastepwould beuseful,andIadvocatedanexplicitinflation objectivenottoolongafterarrivinginSt.Louis. Anumberofothercentralbankshaveexplicit inflationobjectives;however,comparingU.S. andforeigninflationexperiencedoesnotprovide aconvincingcasethatanannouncedobjectiveis necessarytomaintainalowrateofinflation.Thus, althoughIdofavoranannouncedinflation objective,itisclearthatanempiricalcaseisnot adecisiveconsideration;transparencyissues canneverberesolvedentirelythroughempirical investigation. AsecondspecificsuggestionisthattheFOMC mightconsiderreducingitsend-of-meetingstatementtorelativelysimpleboilerplatelanguage. Thissuggestionmayappeartobeastepbackward, buttheissueisthatthecurrentstatementisopen toavarietyofmarketinterpretations,andthe uncertaintyaboutexactlywhatthestatement meansmaynotbehelpfultothecauseofclear communication.Boilerplatelanguagewitharelativelyfewoptionsmighthave,orcometohave,a settledmeaningthatwouldreducemarketuncertainty.Thisexampleisalsointerestingbecause itillustratesthegeneralpointI’memphasizing thathowtransparencyisnotasimpleissue. Insummary,thecaseforwhytransparency isclear.Transparencypromotesaccountability, improvesmarketefficiencyandprobably improvestheclarityofpolicymakingitself.How transparencyisjustplainhard.Itiseasytofind communicationsgaps,butnotatalleasytofill them. 4