speeches · October 13, 2003

Speech

William Poole · President
Economic Growth and the Real Rate of Interest BryantCollege Providence,RhodeIsland October14,2003 My wife, Gerie, and I are absolutely tratethismorningonrealinterestrates,forthat delighted to be back in Rhode iswherethereallyinterestingstoryliestoday. Island, here at Bryant College. As Realinterestratesareinterestratesadjustedfor much as I love St. Louis, Little expectedinflation. Rhody,whereIlivedforalmostaquartercentury, Mostoftheexplanationofwhynominal sure feels like home. interestratesaresomuchbelowtheirlevelof20 IcametoRhodeIsland,toBrownUniversity, yearsagoisthatinflationandexpectedinflation in1974.Theeconomywasinthemiddleofa aremuchlowertodaythanovermostoftheperiod nastyrecession.Interestratesbeganfallinginmid- since,say,1970.However,althoughinflationand 1974—temporarilyasitturnedout—butatthat expectedinflationhavedrifteddownslightly timeratesaslowastoday’sseemedimpossible. overthelastfewyears,thebigstoryisthatreal TheratesinthemarketasIspeakareatlowsnot rateshavedeclineddramaticallysince1999. seensincethe1950sintheUnitedStates.Theone- Whatdeterminesrealinterestrates?Howare yearconstant-maturityTreasuryrate,forinstance, theseratesrelatedtothepaceofeconomicgrowth? averaged1.22percentduringthethirdquarterof Andwhatdoestrendproductivitycontributeto thisyear.Aloweryieldwaslastseenin1954,49 therealrateofinterest?Theseareseveralofthe issuesI’lldiscussthismorning. yearsago. Beforeproceeding,Iwanttoemphasizethat Inmostmarketdiscussion,andinsome theviewsIexpressherearemineanddonot FederalReservecommentaryaswell,thecurrent necessarilyreflectofficialpositionsoftheFederal lowinterestratesareascribedtoeasymonetary ReserveSystem.Ithankmycolleaguesatthe policy.Theargumentusuallygoessomethinglike FederalReserveBankofSt.Louisfortheircom- this:TheFederalOpenMarketCommittee(FOMC) ments.JamesB.Bullard,vicepresidentinthe lowereditstargetshort-termnominalinterestrate, Researchdivision,providedespeciallyvaluable thefederalfundsrate,toalowlevel,andallother assistance.However,Iretainfullresponsibility market-determinedinterestratesfollowed.While forerrors. thisdescriptioniscertainlyaccurateononelevel, itisoverlysimplisticandmasksanumberof importantissues.Inparticular,whatisitabout REAL INTEREST RATES theeconomytodaythatpermitstheFedtosetthe intendedfederalfundsrateatsuchalowlevel? IrvingFisherofYaleUniversitywasperhaps Noonedoubtsthatsettingthefundsrateat1 themostfamousAmericaneconomistinthefirst percentin1999wouldhaveproducedanexplo- halfofthetwentiethcentury.Oneofhismany siveinflation. contributionstoeconomicsciencewashis TheinterestratesI’vementionedarenominal analysisdecomposingthenominalinterestrate interestrates.Anominalratecontainsacompo- intoarealcomponentandaninflationcomponent. nentmeanttocompensateinvestorsforexpected Thisdecompositionwassopowerfulanidea inflationovertheholdingperiod.Iwillconcen- thatitremainswithustodayasafundamental 1 ECONOMICGROWTH principleforeconomistsanalyzinginterestrate inflationofabout2percentandarealyieldabit determination. above2percent.Fouryearsago,theexpectedreal Thedecompositionisnothardtounderstand. yieldonTreasurysecuritieswasintheneighbor- Considera1-yearTreasurysecuritywithayieldof hoodof4percent.Apparently,investorsdonot 1.22percent,theaverageinthethirdquarter. seegoodopportunitiestoearnasubstantialreal Supposethatoneyearfromnowthe12-month yieldtoday.Whathashappened? inflationrateturnsouttobe1.0percent.Then, thatinvestmentwillrealizeapaltry0.22percent realyield,orinflation-adjustedyield.Shouldthe THE REAL COMPONENT AND inflationratebe2.0percent,thentherealyield ECONOMIC GROWTH willbe–0.78percent. Nowconsideraninvestorwhoiscontemplat- I’vebeendiscussingtherealyieldonTreasury ingbuyingaTreasurysecuritywithamaturityof securities.Treasurieshavezerodefaultrisk,or oneyear.Theinvestorwantstobecompensated atleastasclosetozeroaswe’llfind.However, foranydeclineinthepurchasingpowerofmoney becauseofarbitrage,other,generallyriskier,instru- overthetermoftheinvestment,butcan’tknow mentswillbepricedtoyieldasuitableamount therateofinflationinadvance.Thustheinvestor abovethisrisk-freerate.Inparticular,investments formsanexpectationastotherateofinflation.If inproductivecapitalshouldhavearealyield theexpectedrateofinflationishighrelativeto abovetherisk-freeyield,butthetwoyieldsshould thenominalyieldonthebond,theinvestorwill fluctuatetogetherovertimebecauseinvestors probablylookforsomeotherinvestment.This canchoosewhatinvestmentstomake.Whatdeter- phenomenonappearedinspadesinthelate1970s; minestherealyieldonproductivecapital?There fearinginflation,manyinvestorsshunnedbonds issomewell-establishedtheorythathelpsto andbidupthepricesofland,fineart,gold,and answerthisquestion.However,thetheoryI’m othergoodsexpectedtoretainvalue. abouttodiscussabstractsfromimportantcom- Thereislittlequestionthatexpectationsplay plicationsraisedbyuncertaintyandconnections acriticalroleindetermininginterestrates.Since amongcountriesthroughinternationalcapital inflationiscontrolledoverthemedium-tolong- flows.Growththeoryhelpsusunderstandfunda- termbytheFederalReserve,expectationsof mentaleconomicforcesbutshouldnotbeinter- futuremonetarypolicyareakeycomponentof pretedasprovidingprecisequantitativeguidance. observednominalinterestrates.Today,andquite Manypeoplethinkaboutwealthinfinancial consistentlyinrecentyears,investorshavebeen terms,butforsomepurposesthatviewcanbe confidentthattheFederalReservewillmaintain misleading.Everyfinancialassetissomeone’s lowandstableinflation.Forthatreason,inflation financialliability.Fortheworldeconomyasa expectationshavebeenquitelow.Basedonmarket whole,then,financialwealthnetsouttozero. readingsfromtheTreasury’sinflation-indexed Beyondfinancialwealth,Americanhouseholds securities,longer-terminflationexpectations ownthenation’scapitalstock,eitherdirectlyor todayareapproximately2percentbasedona throughownershipoffinancialclaimsonbusi- 10-yearmaturity. nesses.(Ofcourse,someoftheU.S.capitalstock Althoughinvestorswanttorealizeashigha isownedbyinvestorsabroad,butI’llneglectthis realyieldaspossible,giventheriskassumed, factasitisnotimportantfortheissueathand.) theymaynotinsomecircumstancesbeableto Thiscollectionofrealassets—buildings,machines, expectevenapositiverealyield.Today,theone- land,transportationequipment,highways,intan- yearnominalinterestrateisabout1¼percent giblecapital,andsoon—istheproductivecapital whileexpectedinflation,accordingtoseveral that,combinedwithlabor,producesnational measures,overthenextyearisabout1½percent. output.Thestatisticalmeasureofthatoutputis Overa10-yearhorizon,investorstodayexpect grossdomesticproduct(GDP).Calculationsvary, 2 EconomicGrowthandtheRealRateofInterest butestimatesoftheratioofthevalueofthecapital centinrecentyears.Estimatesofthelong-runrate stocktooneyear’sGDPareintheneighborhood oflaborforcegrowtharearound1percent.There- of3.So,withU.S.GDPincurrentdollarsrunning fore,avalueof3½percentisagood,back-of-the- atanannualrateofjustunder$11trillion,the envelopecalculationconcerningthelong-run U.S.capitalstockisworthsomethingapproaching risk-freerealrateofreturnintheU.S.rightnow. $33trillionrightnow. Inastandardmodel,thiswouldalsobethelong- Economistsanalyzetherelationshipbetween runrateofoutputgrowth. inputsandoutputusingtheconceptofaproduc- Letmetrytoprovidesomeinsight,atleast tionfunction.Theproductionfunctionshows qualitatively,intothisresult.First,fromahouse- howinputsofcapitalandlaborarecombinedto hold’spointofview,interestiscompensationfor produceoutput.Oneofthefeaturesofthemodern delayingconsumption.Ata4percentrateofinter- industrializedworldisproductivityimprove- est,forexample,delayingonedollar’sworthof ment—thatis,usingthesameinputstoproduce consumptionforoneyear,andsavinginstead,per- higheroutputofgoodsandservices.Technology mitsconsumptionof$1.04attheendoftheyear. isimprovingovertime.Thisfactisnomystery Nowconsidertheroleofpopulationgrowth. toanyonewhohaslivedinagrowingeconomy Agrowingpopulationneedsagrowingcapital formorethanafewyears.Weseetechnological stock—factorybuildings,machines,houses, improvementsallaroundus.Pricesofgoodslike schools,aircraft,highways,andsoforth.Ifthe TVsandcomputerscandeclinebecauseproducers capitalstockdoesnotgrow,thentheamountof areconstantlyfindingwaystocreatemoreoutput capitalpercapitadeclinesaspopulationgrows. frominputsoflaborandcapital. Toprovidethegrowingcapitalstock,peopleand Totaloutputgrowsfromlargerquantitiesof businessesmustsave—mustforegocurrentcon- inputsandfromhigheroutputperunitofinput. sumptionoutofcurrentproduction.Thehigher Becausethestandardoflivingismeasuredby istherateofpopulationgrowth,thehighermust consumptionpercapita,itisusefultomeasure betherateofsavingandinvestmenttoequipthe totaloutputperhouroflaborinput.Outputper growingpopulationwithcapital.Ahigherreal houroflaborinput,orlaborproductivity,rises interestratepersuadespeopletosavemoreto asmorecapitalisemployedrelativetolaborand providecapitalforthegrowingpopulation. astechnologicalchangeraisesoutputforany Theroleofproductivitygrowthcanbeana- givenamountoflaborandcapital. lyzedinsimilarfashion.Assume,aseconomists Iftherewerenoshockstotheeconomy,a usuallydo,thataworkerispaidhismarginalprod- constantpaceofproductivityimprovementcom- uct,atleastinequilibrium.Supposetheworker’s binedwithaconstantrateoflaborforcegrowth outputisworth$20perhourtodayand,because woulddetermineabalancedgrowthpathforthe ofproductivitygrowth,$21perhourinoneyear. economy.Alongthebalancedgrowthpath,all Wouldyouprefertoworkanhourwhenyourpay marketsintheeconomyareinequilibriumand is$20or$21,whichis5percenthigher?The expectationsofthefutureareconsistentwith answerisobvious.Butnotallworkcanbedelayed actualoutcomes.Andalsoalongthisbalanced tonextyear.Ahigherrealinterestratecanper- growthpath,householdsholdingthecapitalstock, suadepeopletoprovidelabornow.Workingan eitherdirectlyorthroughownershipofbusinesses, hournow,savingandinvestingthehourlyearn- wouldreceivearealreturn.Whatwouldthelevel ingsof$20at5percentinterestyields$21dollars ofthatrealreturnbe? inoneyear,thesameamountthatwouldbereal- Instandardmodelsofeconomicgrowth,this izedfromdelayingworkforoneyear. realreturnturnsouttobetherateofproductivity Inshort,thehigherispopulationgrowthand improvementplustherateoflaborforcegrowth. thehigherisproductivitygrowth,thehigherwill FortheUnitedStates,business-sectorproductivity betheequilibriumrealrateofinterest.Theresult hasimprovedatanannualrateofabout2½per- thattherealrateofinterestispreciselyequalto 3 ECONOMICGROWTH thesumofpopulationandproductivitygrowth thesefundamentalrealeconomicforces.We ratesdependsonthedetailsofthemodel,butthe ignorethesefundamentalsatourperil. generalqualitativeresultmakesalotofintuitive sense. TherealratecalculationIhavejustcompleted TREND BREAKS IN isforabalancedgrowthpathinaworldofno PRODUCTIVITY GROWTH uncertainty.Itisonlyabenchmarksince,of course,actualeconomiesmustadapttoshocks Giventhegrowththeoryresult,thatthereal allthetime.Unusualweatheraffectingagriculture, rateofinterestwillequaltherateofproductivity theriseandfallofindustries,changinggovern- growthplustherateofpopulationgrowth,itis mentpoliciesand,yes,warsallhaveimportant clearthatanalysisofchangesinproductivity impactsontheeconomy.Thustheeconomyisin growthisimportantforunderstandingchanges aconstantstateofadjustment,andprobablyonly inrealinterestrates.Butrecentbehaviorofinter- rarelyactuallyonabalancedgrowthpath.The estratesandproductivitygrowthonlydeepens paceoftechnologicalchangeitself,andtherate themystery.Productivitygrowthseemstohave ofgrowthofthelaborforce,arenotcompletely beenrisingatthesametimerealinterestrates smoothprocesses.Theusefulnessofthebalanced havebeenfalling.Overtheeightquartersending growthpathconceptisthatwhenimportantdis- withthesecondquarterofthisyear,productivity turbancesoccur,wehavesomeideaaboutthe growthfortheU.S.businesssectorhasaveraged statetowhichweexpecttheeconomytoreturn. 4.8percentperyear.Theaverageoverthelast12 Ithinkitisclearthatmanyobserversofthe quartersis3.6percent.Therateofproductivity world’sindustrializedcountrieshavesometype growthaveragedabout2½percentperyearin oftheorylikethisinmindwhenthinkingabout thesecondhalfofthe1990s.Onthiscalculation, interestratesandtheirrelationtofundamental thelong-termrealrateofinterestshouldhave factorsintheeconomy.Inparticular,itisnotsur- increasedbyabout1percentagepointafter1999, prisingthattheUnitedStates,withpositivepop- insteadoffallingbyabout2percentagepoints. ulationgrowthandrobustproductivitygrowth, BeforeItrytoaddressthisapparentanomaly, hasahigherrealrateofinterestthanacountry let’sgobackinhistoryafewdecades.Theperiod suchasJapan,whichfacestheprospectofalong- fromthelate1940stothelate1960swasagolden termdeclineinpopulationandhaslowerpro- eraforU.S.productivitygrowth.Whileactual ductivitygrowththanintheUnitedStates.Nor measurementsarevolatile,theaveragepaceof itissurprisingthatcapitalisflowingfromJapan, productivityimprovementwasclearlyhighduring wherereturnsandcapitalrequirementsarerela- thisperiod—perhapsintheneighborhoodof3 tivelylow,totheUnitedStateswherereturnsand percentperyear.Butsometimeinthelate1960s requirementsarerelativelyhigh. orearly1970s,productivitygrowthratesslowed Whataretheimplicationsforpolicymakers? substantially,awidelystudiedbutpoorlyunder- Clearly,wemustkeepasharpeyeonproductiv- stoodeventthatbecameknownastheproductivity ityimprovementswhenthinkingabouttheappro- priatelevelofnominalinterestrates,becausethe slowdown.Productivitygrowthratesfellbyhalf. realcomponentofthenominalrateisdrivenby Thisdeclineinproductivitygrowthhadadra- productivityinanimportantway.Someofthe maticimpactontheU.S.economy,asitmeant increasesinproductivityregisteredinthequar- thatrealincomefortheaveragehouseholdwould terlystatisticsrecentlyhavebeenquiteextraor- improvemuchlessrapidlythanearlier.Witha dinaryindeed.Theprospectsforcontinued productivitygrowthrateof3percent,realhouse- productivityimprovementalsoappeartobegood. holdincomedoublesevery25years;withpro- Eventually,butnooneknowswhen,itseems ductivitygrowthat1.5percent,realhousehold likelythatmarketinterestrateswillrisetoreflect incomedoublesevery48years. 4 EconomicGrowthandtheRealRateofInterest Thedeclineinproductivitygrowthshould increasesinproductivitygrowthmighthave havehadadirectimpactontherisk-freerealrate effectsoninflationandmonetarypolicyopposite ofinterest,causingittofallinthe1970sbyan tothosecreatedbythesurpriseshortfallinpro- amountaboutequaltothedeclineinproductivity ductivitygrowthinthe1970s. growth.Butnominalinterestratesroseduring Duringtheearly1990s,U.S.inflationrates the1970s,becauseinflationincreased.The werehigherthantheyaretoday,andtheeconomy increasesininflationweretoolargeandvolatile wasinthemidstofarecoveryfromthe1990-91 toallowustoseethedeclineinrealinterestrates recession.Atsomepointintherecoveryprocess, associatedwiththeproductivityslowdown. usuallydated1995,thetrendproductivitygrowth Isitacoincidencethatinflationroseaspro- rateincreased.Again,becausethedatacontaina ductivitygrowthslowedstartingintheearly lotofnoiseandbecausemeasurementissuesare 1970s?Therearecertainlyreasonstobelievethat quitedifficult,itwasfarfromclearuntilseveral partoftheinflationrun-upwasinfactduetothe yearsaftertheeventthatthetrendpaceofpro- slowdowninproductivitygrowth.Declining ductivityimprovementhadincreasedtoabout productivitygrowthnotonlyputsdirectupward 2½percent.Realhouseholdincomewoulddou- pressureongoodspricesbutalso,byslowingtrend bleevery30yearsatthispaceofproductivity outputgrowth,makesitalltooeasyforpolicy- improvement. makerstomisreadthestateoftheeconomy.Mone- Duringthelasttwoyears,nonfarmbusiness tarypolicywastooexpansionaryinthe1970s, sectorproductivitygrowthhasaveragedabout inpartbecausepolicymakersoverestimatedthe 4¾percentatanannualrate.Atthisrate,real economy’scapacitytogrow.Becauseproductivity householdincomewoulddouble,astonishingly, isnotoriouslydifficulttomeasure,andavailable every16years. statisticalmeasuresarenoisy,therealizationthat Twoyearsistooshortofatimetoreachfirm productivitygrowthwassloweronasustained conclusionsaboutpersistentchangesinthetrend basis,andnotjustasashort-runaberration,did paceofproductivitygrowth.Andcertainly,some notoccurforsomeyearsaftertheslowdown oftheproductivitygainsseemtobeconnected began.Inthemeantime,expansionarymonetary withanunwillingnessoffirmstohireadditional policybuiltaninflationproblemintotheeconomy. laboroverthelasttwoyearswhiletheywaitfor Failuretorecognizetheproductivitygrowthslow- economicconditionstoimprove.Totheextent downiscertainlynotthewholestoryofmonetary thisistrue,productivitygainsmaybelowerwhen policymistakesinthe1970s,butprobablypart employmentstartstogrowatasustainedhealthy ofthestory. pace. Inanyevent,giventheeconomicinstability Still,theearlierlessonswithtrendproductiv- createdbyinflationinthe1970sanddisinflation itychangesandtheimpactthosechangeshave intheearly1980s,wecertainlycannotcharacter- onrealinterestratesandonthelevelofoutput izetheU.S.economyduringthisperiodasgrow- consistentwithbalancedgrowthsuggestcaution. ingalongabalancedgrowthpath.Realratesof Thereisadistinctpossibilitythattheunderlying interestwerelowerinthe1970sandhigherin paceofproductivitygrowthhasincreasedagain. the1980s,butthesechangesseemtohavehad Ifso,andiftheFOMCdoesnotmakeappropriate moretodowithinflation-forecastingerrorsand policyadjustments,inflationcoulddriftaway cyclicalprocessessetupbyinflationthanwith fromcurrentlowlevels.Theshort-runimpactof changesinlong-runproductivitygrowth. higherproductivitygrowthmightbetolowerthe The1990sbroughtapleasantandunexpected inflationrate,asrisingoutputplacesdownward surprise:Thetrendrateofproductivitygrowth pressureonprices.However,overthelongerrun, rosesignificantly.And,asI’veemphasizedalready, higherproductivitygrowthwillprobablyrequire productivitygrowthhasbeenevenstronger higherinterestrates.Ifmonetarypolicyadjust- overthelastthreeyears.Unexpected,persistent mentsdonotkeepupwitharisingequilibrium 5 ECONOMICGROWTH realrateofinterest,thentheinflationratemay anunsustainableinvestmentboom,whichwas ultimatelyrise.However,mybestguessisthat particularlymarkedinthetelecomanddot-com thecurrentstanceofpolicymakesadequate industries.Overthelastseveralyears,realrates allowance,offsettingtheriskthatthecurrent havebeenlowastheinvestmentslumphasgradu- lowinflationwillturnintoaperiodofdeflation, allyworkedoffexcesscapitalstockandasaseries andFedvigilanceoverthelongerrunwill,I ofshocks—9/11terroristattacks,majorbankrupt- believe,keepinflationundercontrol. cies,andcorporategovernancescandals—created apervasiveairofcaution.Theseareoscillations aroundtheeconomy’slong-rungrowthpath. SHORT-TERM REAL INTEREST Short-runadjustmentsinmonetarypolicyhave beenofenormoushelpindampingtheoscilla- RATES tions,butitisbeyondtheFOMC’spoweratthe I’veconcentratedontherealrateofinterest currentstateofknowledgetoeliminatetheoscil- onlonger-termbondsandonphysicalcapital. lationsaltogether. Nevertheless,itishardtoignorethefactthatata Asimportantasshort-runpolicyadjustments one-yearmaturity,therealrateofinterestisabout are,iftheFOMCfocusesonlyontransitorydis- zerotoday.Thatrate,Ibelieve,shouldbeviewed turbancesandfailstounderstandthelonger-run asatransitorysituationduringtheprocessof forces,itrunstheriskoffallingbehindasthe economicadjustmentreflectingfundamental economyconvergestowarditsbalanced-growth longer-termforces.Amongthesefundamental path.Thatispartofthestorybehindpolicymis- forces,therateofproductivitygrowthiscertainly takesinthe1970s.Idonotknowwhatthetrend atthetopofmylist. rateofproductivitygrowthwillturnouttobe, Policymakersdonothavetheluxuryofcal- butifisintheneighborhoodof3percent,thena culatinganinterestratetheoreticallyconsistent reasonableguessontheequilibriumrealrateof withabalancedgrowthpath,asIdidearlier,and interestisabout4percent—3percentfrompro- thensimplyprovidingthatinterestratetothe ductivitygrowthplus1percentfrompopulation market.Itisusefultounderstandtheeconomics growth.Inthemeantime,theFed’staskistopro- behindthelong-runbalancedgrowthpath,but moteasrapidareturntothelong-rungrowth thoseconsiderationsprovidelittlepracticalguid- pathasisconsistentwithmaintainingarateof anceforshort-runmonetarypolicy.Becauseof inflationthatislowandsteady. short-andmedium-termshocks,theeconomy I’veofferedyoumytakeonwherethestance canwanderawayfromitssteady-stategrowth ofmonetarypolicyistodayandonhowtheappro- pathandinterestratescandepartfromtheirlong- priatepolicystanceislikelytoevolveinthe runequilibriumvalues.TheFedwouldhinder futureaseventsunfold.I’veemphasizedthat ratherthansupportareturntotheequilibrium fundamentalrealforcesaffectingtherealrateof pathwereittodetermineitsinterestratestance interestarelikelytobecriticaltothelong-run onlyonthebasisofthelong-runequilibrium. pathofinterestrates,butthatthetimingofinter- Fouryearsago,realinterestrateswerehigh estratechangesishighlyuncertain.NowI’dbe becauseoftheincreaseinproductivitygrowthin pleasedtotakeyourquestions. themid1990s,andalsobecauseexcessiveopti- mism—excessiveinretrospect,anyway—ledto 6
Cite this document
APA
William Poole (2003, October 13). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_20031014_poole
BibTeX
@misc{wtfs_speech_20031014_poole,
  author = {William Poole},
  title = {Speech},
  year = {2003},
  month = {Oct},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_20031014_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}