Speech

As Easy as P.I.E.: Productivity, Innovation, and Education AnnualTechnologyTransferShowcasefortheUniversityofMissouriSystem St.Louis,Missouri April25,2002 How economies grow and prosper is essarilyreflectofficialpositionsoftheFederal one of the central questions of eco- ReserveSystem.Iappreciatecommentsprovided nomics. At least since the time of bymycolleaguesintheResearchDivisionatthe Adam Smith, economists have rec- FederalReserveBankofSt.Louis.RichardG. ognized that enhancing living standards is as AndersonandKevinL.Kliesenprovidedespe- easy as P.I.E.: combine productivity, innovation, ciallyvaluableassistance.However,Itakefull andeducation.Productivitygrowthisthecritical responsibilityforerrors. factor that determines future living standards. Such growth, in turn, depends on the birth of new ideas—innovation and invention—and our OUR PRODUCTIVITY EXPERIENCE abilitytoturnsuchideasintousabletechnology— Oureconomyisadynamic,ever-changing thatis,technologytransfer.Both,inturn,depend system.Asaresult,productivitygrowthebbsand on education. flows,oftenforreasonsonlyimperfectlyunder- Speedinginnovationthroughgovernment stood.Yet,alleconomistsappreciatethatthe assistanceisnotanewidea.In1974,forexample, interactionamongproductivity,innovationand theU.S.government’svariedresearchlaboratories educationiscrucialtomaximizingeconomic joinedtogethertoformaconsortiumtopromote growth. technologytransfer;in1986theireffortswerecodi- Thereisnowlittledoubtthatthepaceof fiedintofederallawbytheFederalTechnology productivitygrowthintheUnitedStatesrose TransferAct.Anumberofnewapplicabletech- duringthelastdecade.From1995to2001,non- nologieshaveresultedfromthispublic-private farmlaborproductivitygrewatanannualrateof liaison,includingnewteststorapidlyidentify nearly2.5percent,morethanapercentagepoint foodcontaminationandnewchemicals(spun fasterthanthedisappointingperformanceseen offfromtheNASAspaceexplorationprogram) from1973to1995.Increasesinbusinessefficiency toincreasethecoolingcapacityofyourautomo- boostedexpectationsofthefuturegrowthofcor- bileairconditioner.Numerousotherexamplesare porateearnings.Increasedefficiencyalsolowered availableontheconsortium’sInternetwebsite. unitlaborcostsandallowedsteadyincreasesin But,mytopictonightisnottospeakofindi- wageswithouttriggeringhigherinflation.Inturn, vidualnewtechnologies.Rather,Iwilldiscuss higherexpectedearningsfedintohigherequity howeconomists—andespeciallypolicymakers— prices.Boththehigherrealearnings—earnings thinkoftechnology.Insodoing,Iamgoingto afteradjustmentforinflation—andincreased focusmyremarksonproductivitygrowth,aresult wealthsupportedstronggainsintheaverage oftechnologytransfer. household’sstandardofliving.Ahigherstandard Beforeproceeding,Iwanttoemphasizethat oflivingispreciselywhatwedesirefromeffective theviewsIexpressherearemineanddonotnec- technologytransfer. 1 ECONOMICEDUCATION Whatcausedtheseeventsofthe1990s?In themeineconomichistory.Economichistorians theearly1990s,innovationsintheproductionof seemtoagreethatlargegainsinlivingstandards microprocessorsandrelatedsemi-conductors donotarisefromspecificinnovationsorinven- allowedsharpdecreasesinthepricesofcomput- tionsbut,rather,fromtheapplicationofsuch ingandtelecommunicationsequipment.Inturn, innovationsbythoseseekingtocapitalizeonthem. businessesaggressivelyreorganizedtheirmanage- ThenotedhistorianAngusMaddisonhasargued mentinformationsystems.Inaddition,other that1820markedamajorturningpointineco- entrepreneursquicklyintroducedmachinetools, nomichistory.About1820,businessesbeganpro- material-handlingequipment,andsimilarcapital ductiveuseoftheinnovationsofthe18thcentury, goodsthatcontainedembeddedmicroprocessors. includingthesteamengine,therailroadlocomo- Someeconomistshaveequatedtheimportanceof tiveandchemicalprocesseslikebleaching.Ever thistechnologytransfer—thatis,thesuccessful sincethattime,theeconomy’sratioofcapitalto combinationofinnovationandtechnologysoas laborhastendedtoincreasefairlysteadily—and toimproveproductivity—withtheintroduction productivityalongwithit.Asaconsequence,in oftheelectricdynamoduringthelastpartofthe EnglandandtheUnitedStatesoutputpercapita 19thcenturyandthewidespreadadoptionofsci- begantodoubleapproximatelyevery40to50 entificagricultureduringtheearly20thcentury. years.Althoughmodest-soundingbytoday’s Letmeemphasizethattheproductivity standards,thisadvancewastrulyremarkable:in improvementwe’rediscussingrequiredinnova- total,duringthepreviousmillennium,worldwide tionsinhardware,insoftwareandinbusiness realGDPpercapitahadincreasedonlyabout50 processes.Anytwoofthesewithoutthethird percent. wouldhaveyieldeddisappointingresults. Productivitygrowthduringthe20thcentury Wal-MartCorporationisperhapsthemost followedthatofthe19thcentury—adoptionof widelydiscussedexampleofhowtheadoptionof importantinnovationsraisedproductivitygrowth. lower-costcommunicationsandinformationpro- Asaconsequence,U.S.outputpercapitabegan cessingequipmentcanmakepossiblefundamental todoubleaboutevery30years.Notsurprisingly, changesinbusinessmanagement.Wal-Mart’s theannualgrowthrateofnonfarmbusinesspro- informationsystemsdeliverdata—hour-by-hour, ductivityslowedduringtheDepression,averag- product-by-productandstore-by-store—bothto ingapproximately1½percentbetween1929and managementandtoWal-Mart’sdistributioncen- 1938.IntheUnitedStates,the“GoldenAge” ters.Suchinformationsystemsareexpensive,to seemstohavebeen1950to1973,whenproduc- besure—butthebenefitsseemtohaveexceeded tivityincreasedatalmosta3percentannualrate. thecosts.Lastyear,Wal-Martbecamethenation’s Around1973,however,U.S.productivity largestfirmmeasuredbyannualsales.Yet,Wal- growthbegantostall.From1973to1995,nonfarm Mart’ssuccessdidnotdependjustoninnovations laborproductivitygrewataDepression-like1.4 withintheconfinesofthatfirm.BecauseWal-Mart percentannualrate.Aswithmanysignificant encourageditssupplierstolinktoitsinformation economicevents,theredoesn’tseemtobeasimple system,thosesuppliershaveimprovedtheirinfor- explanationforthedecline.Norwastheproduc- mationandinventorymanagementsystems.In tivityslowdownrecognizedimmediately.By1976, turn,thesuppliersofthosefirmshavefoundit however,itwasevidentthatsomethingstructurally profitable,andoftennecessary,toimprovetheir significanthadhappenedtotheU.S.economy.In informationandinventorymanagementsystems, 1977,afullfourtofiveyearsaftertheslowdown andsoon.Throughthistieringprocess,Wal-Mart started,theCouncilofEconomicAdvisers,then itselfhasbecomeapowerfulengineoftechnology headedbyAlanGreenspan,trimmeditsestimate transferfortheentireU.S.economy. ofpotentialGDPgrowthfromabout4percentto Thecloserelationshipbetweentechnology 3.5percent.Butlatereventsweretoshowthat transferandeconomicprosperityisaprominent eventhatratewasmuchtoooptimistic. 2 AsEasyasP.I.E.:Productivity,Innovation,andEducation Whatcausedthatproductivityslowdown? Economistsusestatisticalmodelsofthe The1977EconomicReportofthePresidentargued economy’sproductionprocesstoseparatethese thatthepermanentincreaseinrealenergyprices threecomponents.Measuringthefirsttwocom- followingtheAraboilembargolikelywasasignifi- ponentsisdifficultbutrelativelystraightforward cantfactor.Otherfactorsincludedhigherinflation, becausefirmsreporttothegovernmenteachyear adramaticescalationinenvironmentalandwork- boththeiremploymentandtheircapitalpurchases placeregulations,andaninfluxofalargenumber anddepreciation.Measuringtrueinnovation— ofpersonsintothelaborforceforthefirsttime, changesinourknowledgeofhowtodothings— especiallywomenandteenagers.Subsequent isdifficult.Infact,thisextremelyimportant economicresearchhasreachedessentiallythe componentistypicallymeasuredasaresidual— sameconclusion.Interestingly,though,theoil everythingelse—afteraccountingforotherfactors. shockstorycontinuestobefavoredbymany Despitethisdifficulty,therearemanythings economists. thatwedoknowaboutwhatfostersinnovation— Overall,realpercapitaincomeintheUnited andwhatdoesn’t.Writingchieflyaboutinforma- Stateshasincreasedmorethaneight-foldduring tiontechnology(IT),StanfordUniversity thelast200years.Thepacehasbeenuneven,both economistTimothyBresnahanhasarguedthat intimeandgeography,butithasbeenremarkable ITinnovationsbythemselvesareoflittlevalue nonetheless.Today,wecanonlyhopethatthe totheaggregateeconomy.Hearguesthatanother roughly2½percentannualgrowthinU.S.labor twokeydevelopmentsmustalsooccur.First,the productivitysince1995continues.Evena2per- inventionmustbean“enablingtechnology.” centgrowthtrendwouldbesuperiortotheecon- Thatis,itmustbeonethatcanbeusedinnumer- omy’sdismalproductivityperformanceduring ousapplications—thesearetheonesthateventu- the1970sandmuchofthe1980s. allywillboostgrowth.Second,themostvaluable innovationsarethosewithnetworkeffects,a typeofeconomicexternality.Inotherwords,the BOOSTING PRODUCTIVITY valuecreatedbyanITinnovationisrelatedto thebreadthofitsuseacrosstheeconomy.But, THROUGH INNOVATION thesebenefitsmaytakealongtimetoappear. Accordingtoofficialproductivitydatacom- Acommonexampleamongeconomichisto- piledbytheBureauofLaborStatistics(BLS),the riansisearly20thcenturyelectrification.Electri- post-1973slowdown—thatis,relativetothepro- ficationenhancedproductivitybyincreasing ductivitysurgethatoccurredfrom1948to1973— flexibilityandallowingmanufacturerstouse wasattributableentirelytoaslowdowninthe laborandcapitalmoreefficiently.Forexample, rateoftechnicalprogress,whateconomistscall electrificationenableduseofcontinuous-process totalfactorproductivity,orTFP. techniquessuchasthefactoryassemblyline. TounderstandwhatTFPmeans,notethat Efficiencyalsoimprovedwiththewidespread economistsattributeoutputgrowthtothreecom- adoptionof“unitdrive,”thatis,theuseofrela- ponents:increasesinlaborinput,growthinthe tivelyinexpensive,dedicatedelectricmotorsto nation’scapitalstock,whichincreasescapital powerindividualmachinesandtools,rather input,andeverythingelse.Wethinkofthecatchall thanusingasystemofshaftsandbeltspowered term“everythingelse”asreflectingadvancesin byasinglecentralengine.Unitdrivebrought knowledgebecausethisisthepartofoutputin savingsthroughreducedenergyusage,lesswear excessofwhatcanbeaccountedforbymeasured andtear,andmoreflexibleandefficientfactory inputsoflaborandcapital.Noself-respecting design.Electrificationalsoenhancedproductivity disciplinewouldevernameanimportantconcept byimprovingfactorylightingandsafety. “everythingelse,”sointheproductivityliterature, Butthisprocessdidn’toccurovernight.Firms thistermisreferredtoas“totalfactorproductivity.” arereluctanttoscrapoldtechnologies,typically 3 ECONOMICEDUCATION embodiedinexpensiveplantandequipment, rately,attheBoardofGovernorsinWashington, merelyontheunprovenpromiseofnewerones. D.C.,ChairmanGreenspansawtantalizingevi- Hence,thesebenefitsaredelayedbysubstantial denceofapickupinproductivitygrowththat adjustmentcostsrelatedtoreorganizingtheway seemedsimplyinconsistentwithwhatofficial ofdoingbusiness—whatBresnahancalls“co- dataindicated.Inhisview,thelinkagesbetween invention”costs.Initially,earlyadoptersprove reporteddataonprofits,pricesandcostsdidnot thatthenewtechniqueswork.Later,aswider addupthewayeconomictheorysuggested.The adoptionoftheseinnovationscreatessomeecon- picturechangedwithsubsequentrevisionsto omyofscaleintheproductionofthenewequip- thedata—inparticular,theincorporationofsoft- ment,thecostofthenewtechnologydecreases, wareasafixedinvestmentintheGDPaccounts providingafurtherincentiveforfirmstofinally in1999—andeconometricworkbyseveralecon- taketheleaptothenewertechnologies. omists.ThisresearchshowedthatChairman Recognizingthesetrendsseemseasywith Greenspan’sintuitionwasessentiallycorrect. thebenefitofhindsight.Itisnothardtofindthe Mypointisnottorejointhatdebatebut successes;thefalsestartsandoutrightfailures rathertoemphasizethatthebenefitsofenabling mayneverappearinthehistoricalrecord.Inprac- technologiesoftenevolveslowly,andtheeco- tice,however,parsingcurrenteconomicdata nomicshiftsthattheycausemaybedifficultto doesnotreadilyyieldcluestoemergingtrends recognizeinthedata.Thereisnoeasywayto thatmaybetheresultofpastinnovations.The distinguishnewtrendsfromtemporaryaberrations mostobviousexampleinrecentyearswasthe inexistingtrends.Weshouldnot,forexample, policydebatefromabout1995to1998.Onthe dismissthepromiseofe-commerceorbusiness- onesideweretheso-called“NewEconomy” to-businessapplicationssimplybecausethey apostles,thosewhobelievedthatinnovations haveyettotakeoff.Iamnotmakingaforecast associatedwiththemicrochiphadpermanently onewayortheother,butemphasizingthathis- increasedthegrowthrateoflaborproductivity. torysuggestsamplereasontobecautiousinboth Accordingtothisgroup,theeconomy’spotential directions. growthhadrisentoapproximately3to3.5per- cent—implyingalong-runproductivitygrowth rateof2to2.5percent(theremaining1percent BOOSTING PRODUCTIVITY growthattributabletolaborforcegrowth).Some THROUGH EDUCATION NewEconomyadvocatesapparentlyhadmuch highergrowthratesinmind,althoughtheytypi- Sofar,Ihavefocusedontechnology.But,how callydidnotcommittospecificestimates.Onthe doinnovationandtechnologytransferoccur? othersidewerethe“traditionalists,”thosewho And,cangovernmentsdoanythingtoencourage believedthatrealGDPgrowthwasbeginningto morerapidtechnologicalprogressandeconomic risemainlybecauseofcyclicaldynamics(gradual growth?Duringtheindustrialrevolutionsofthe re-employmentofslackresources),orothertem- 18thand19thcenturies,forexample,private poraryfactors,andthatoncethosebenefitshad individualsandfirmsproducedmostinventions beenexhausted,theeconomywouldbebacktoa anddid“technologytransfer”largelywithout longerruntrendgrowthrateofabout2.5percent, governmentsubsidiesordirection.Although ashadbeenexperiencedfrom1973. economistsarefarfromhavingacompleteunder- Atthetime,officialdataseemedsquarely standingoftheseissues,economicanalysispro- alignedwiththetraditionalistsandmainstream videssomeguidance. forecasters.Despitepersistentlystrongoutput First,governmentshould“donoharm.” growth,andhencepersistentlyone-sidedforecast Excessiveregulationandrigiditycanstiflethe errors,mostforecastersprojectedareturntothe transformationofinnovationsintoapplicable oldtrendgrowth.InsidetheFed,ormoreaccu- technology.Manyanalystshavenotedthatfew 4 AsEasyasP.I.E.:Productivity,Innovation,andEducation othercountriesenjoyedariseofproductivity withlotsofsubsidiesforinputs—likeresearch growthduringthe1990sasrapidasdidtheUnited andeducation—thatareusedintheinnovation States.Inpart,theexplanationforsuchadiffer- process.” encemaylieintherelativelyless-regulated,more ManyeconomichistorianscredittheU.S. flexible,andmorecompetitivenatureofU.S. highereducationsystemforourtechnological marketsandbusiness.TheUnitedStatesdoesa prowess.TheMorrillActof1862createdland goodjobofencouragingentrepreneurs. grantuniversities,therebystimulatingteaching Encouragingentrepreneursseemssimpleuntil andresearchinbothagricultureandengineering. weconsiderthatnewtechnologycreateslosers WithinadecadeaftertheAct’spassage,thenum- alongwithwinners.Thetransferofnewtechnolo- berofengineeringschoolswentfrom6to70, gies—suchasgrowinguseofthesteamengine, andlaterto126schoolsby1917.In1870,U.S. electricity,theinternalcombustionengine,and engineeringschoolsgraduated100students;in themicrochip—changestherelativefortunesof 1917,theygraduated4,300.Asearlyas1890,the numerousfirmsand,inturn,therelativedemand ratioofuniversitystudentsper1,000primary forvarioustypesoflabor.Asaresult,wagesof schoolstudentsintheUnitedStateswastwoto someworkerswilltendtoincreaserapidly— threetimesthatofanyothercountry.Aslateas whileearningsandjobsinotherindustrieswill 1914theUnitedStateswaswellbehindEurope contract.Governmentleadersmustresisttheurge inscientificagriculture.Agenerationlater,we to“save”thelatterindustrieslest,bysodoing, weretheworldleader.Today,ourhighereduca- theyforeclosegainsfortheoveralleconomy. tionsystemiscalledupontoprovidethenew Whilenoonelikestoobservelayoffsand talenttomaintainourtechnologyleadership. businessclosings,thesemaysignalthefuture Overthelast25yearsorso,thecollegewagepre- directionoftheeconomy.Governmentmustbe mium—thewagesofcollegegraduatesrelativeto cautiousnottointerferewiththesesignals.Itis thoseofhighschoolgraduates—hasjumped25 particularlydamagingwhengovernmentsprotect percent.Alittlemorethanadecadeago,about existingjobsbystiflinginnovationandblocking 39percentofthepopulation25yearsandolder entryofnewproducts,servicesandproducers. hadsomeformofcollegeeducation;in2000,the Second,governmentmustprovideasecure proportionhadrisento50percent.Technological systemofprivatepropertyrights,includingpro- progress—turninginnovationsintoapplicable tectionforintellectualcapital.DouglassNorth, technology—simultaneouslydependsonawell- thenotedWashingtonUniversityeconomichisto- educatedlaborforceandincreasesthedemand rianandNobellaureate,hasarguedthatanation’s forhighereducation. institutions,includingitsgovernment,areamong themostfundamentaldeterminantsofeconomic growth.Economicperformancetendstobebetter, IMPLICATIONS FOR MONETARY heargues,whengovernmentinterventioninpri- POLICY vatemarketsisminimalexceptfortheenforce- mentofprivatepropertyrights.Secureproperty Finally,Icometomyfourththoughtofthe rights,includingclearownershipofintellectual evening:thegovernment,andmorespecifically, propertyviapatentsandcopyrights,encourage theFederalReserve,mustfollowsoundmacro- entrepreneurshipandtechnologytransfer. economicpoliciesconsistentwithalow,stable Third,governmentmustsponsorastrong rateofinflation. andwidelyavailablesystemofhighereducation. Thestrengthanddurationofthecurrenteco- EconomistPaulRomer,aleadinggrowthexpert nomicexpansionwillultimatelydependonthe atStanfordUniversity,hasarguedthat“…the performanceoftheinflationrate.Lowandstable realsuccessofAmericaneconomicpolicyhas inflationreducesuncertaintyregardingthefuture beentohavemoderatelystrongpropertyrights healthoftheeconomyand,inturn,encourages 5 ECONOMICEDUCATION entrepreneurshipandrisk-taking.Highandvari- cannotberuledout.Acceptingtheforecastof2 ableinflationincreasesrisk,whichinducescau- to2½percenttrendproductivitygrowth,then tionamongentrepreneursandventurecapitalists. theeconomy’slong-rungrowthtrack,assuming Theconsequenceislessinnovationandless thatthelaborforceincreasesapproximately1 applicationofknowninnovations. percenteachyear,isapproximately3to3.5per- Monetarypolicymakingrequiresanestimate cent,aboutapercentagepointhigherthanthe ofthepotentialgrowthrateoftheeconomy trackthatprevailedbetween1973and1995. becauseitgivesusasenseofhowfasttheeconomy Maintainingthehighertrackwillraisetheliving cangrowwithoutdevelopinginflationaryimbal- standardsoffuturegenerationsofAmericans,as ances.Growthmorerapidthanthelong-runpath wellasthoseincountrieswetradewith.Butthis cangenerateimbalancesthatthreatenlong-run outcomecanonlycometopasssolongasinflation sustainedprosperity.Yet,nopolicymakerwants remainslowandstable. tounnecessarilyslowaboomingeconomyifthe Yet,wemustbemodest.Ourunderstanding economy’sperformancereflectsanacceleration ofthedeterminantsofproductivitygrowthistoo ofproductivity.Productivityincreasesarethe imprecisetojustifyfirmconvictionsaboutany largestpartofoureconomy’slong-rungrowth. productivitygrowthforecastoverthenearterm, Eventherecent,mildeconomicslowdownseems muchlessthelongrun.Givenourincomplete tohavedonelittletoslowproductivity’sacceler- knowledge,therefore,itisimportantthatwenot ationthatstartedinthemid1990s.Recentdata lockourselvesintoamonetarypolicythatdepends arehighlyencouraging;fourth-quarterproduc- onanyparticularrateofproductivitygrowth. tivitygrowthwasmorethan5percent,andfirst Instead,policymakersmustbeonguardthatan quartergrowthcouldevenbeashighasaremark- increaseininflationdoesnotderailtheeconomy’s able8percent.Asaresult,unitlaborcostshave long-rungrowthcombinationofinnovation,pro- decreased,corporateprofitshaveincreased,and ductivityandeducation. businessinvestmentspendingisrebounding. I’llfinishwiththisobservation:itisalot Manyanalystsnowbelievethattheeconomy’s easier—awholeloteasier—tobeapolicymaker sustainableproductivitygrowthrateisapproxi- inanenvironmentofstrongproductivitygrowth mately2to2½percent.Amodestlyhigherrate thaninoneofstagnation. 6

William Poole (2002, April 24). Speech. Speeches, Federal Reserve. https://whenthefedspeaks.com/doc/speech_20020425_poole

@misc{wtfs_speech_20020425_poole,
  author = {William Poole},
  title = {Speech},
  year = {2002},
  month = {Apr},
  howpublished = {Speeches, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/speech_20020425_poole},
  note = {Retrieved via When the Fed Speaks corpus}
}