research spotlight measuring the digital economy · the u.s. statistical system does not currently...

13
May 2019 Volume 99, Number 5 Research Spotlight Measuring the Digital Economy By Kevin Barefoot, Dave Curtis, William A. Jolliff, Jessica R. Nicholson, and Robert Omohundro In April 2019, the Bureau of Economic Analysis (BEA) released updated measures of the digital economy’s contributions to the overall U.S. economy. The measures include estimates of digital economy gross output, value added, employment, and compensation for the period from 1997 to 2017. The new estimates rely on the same methodology for measuring the digital economy as those used for the initial estimates BEA published in March 2018. The April 2019 update incorporated underlying data revised during the 2018 comprehensive update of the Industry Economic Accounts and expanded the time series. The March 2018 and April 2019 estimates are BEA’s initial efforts to lay the foundation for a digital economy satellite account and are the first set of estimates within the framework of the national accounts. Conceptually, a digital economy satellite account should include all goods and services related to the digital economy. However, BEA’s measures published to date are based on goods and services that are primarily digital. BEA is researching data and methods to overcome the challenges of measuring “partially digital” goods and services to expand the coverage of the digital economy measures later in 2019. The updated estimates show the digital economy accounted for 6.9 percent ($1,351.3 billion) of current-dollar gross domestic product (GDP) in 2017, up from 5.9 percent in 1997. Digital economy real value added grew at an average annual rate of 9.9 percent per year from 1998 to 2017, compared to 2.3 percent growth in the overall economy. BEA developed the digital economy estimates with support from the Commerce Department’s National Telecommunications and Information Administration. – 1 –

Upload: others

Post on 26-Jul-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

May2019 Volume99,Number5

ResearchSpotlightMeasuringtheDigitalEconomyByKevinBarefoot,DaveCurtis,WilliamA.Jolliff,JessicaR.Nicholson,andRobertOmohundro

InApril2019, theBureauofEconomicAnalysis (BEA)releasedupdatedmeasuresof thedigitaleconomy’s contributions to the overall U.S. economy. Themeasures include estimates of digitaleconomygrossoutput,valueadded,employment,andcompensationfortheperiodfrom1997to2017. The new estimates rely on the samemethodology formeasuring the digital economy asthose used for the initial estimates BEA published in March 2018. The April 2019 updateincorporated underlying data revised during the 2018 comprehensive update of the IndustryEconomicAccountsandexpandedthetimeseries.

TheMarch 2018 and April 2019 estimates are BEA’s initial efforts to lay the foundation for adigital economysatellite accountandare the first setof estimateswithin the frameworkof thenationalaccounts.Conceptually,adigitaleconomysatelliteaccountshouldincludeallgoodsandservicesrelatedtothedigitaleconomy.However,BEA’smeasurespublishedtodatearebasedongoodsandservicesthatareprimarilydigital.BEAisresearchingdataandmethodstoovercomethechallengesofmeasuring“partiallydigital”goodsandservices toexpand thecoverageof thedigitaleconomymeasureslaterin2019.

Theupdatedestimatesshowthedigitaleconomyaccounted for6.9percent($1,351.3billion)ofcurrent-dollar gross domestic product (GDP) in 2017, up from 5.9 percent in 1997. Digitaleconomyrealvalueaddedgrewatanaverageannual rateof9.9percentperyear from1998to2017,comparedto2.3percentgrowthintheoveralleconomy.

BEA developed the digital economy estimates with support from the Commerce Department’sNationalTelecommunicationsandInformationAdministration.

– 1 –

Page 2: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

Methodology

BEAprepared these statisticswithin the supply-use framework, followingmethodologyused intheproductionofBEAsatelliteaccounts,includingthoseontravelandtourism,artsandculturalproduction,andoutdoorrecreation.

Theestimationprocessincludedthreemainsteps:

1.Developingaconceptualdefinitionofthedigitaleconomy;2.Identifyinggoodsandserviceswithinthesupply-useframeworkrelevantformeasuringthedigitaleconomydefinedinthefirststep;and

3.Usingthesupply-useframeworktoidentifytheindustriesresponsibleforproducingthesegoodsandservices,andestimatetheoutput,valueadded,employment,compensationandothervariablesassociatedwiththisactivity.

Duringthesecondstepofthisprocess,BEAreviewedthedetailedgoodsandservicescategoriesfromthesupply-useframeworktoidentifydigitaleconomygoodsandservices. Somegoodsandservicescategoriesincludeamixofbothdigitalandnondigitalgoodsandservices.Forexample,thegoodscategoryelectronictoysandgames,includinghomevideogames(excludingcartridges,disks, and tapes) includesbothdigitalvideogamesandnondigitalelectronic toys.Conceptually,measures of the digital economy should include digital video games; however, due to data andresourceconstraints,theestimatespresentedhereincludethegoodsandservicescategoriesthatBEAconsidersprimarilydigital.

DefiningthedigitaleconomyThe digital economy is characterized by the rapidly changing nature of technology. Ideally, thedefinitionofthedigitaleconomywouldallowforthechangingnatureofwhatitencompassesovertime. Recognizing the large overlap between the information and communications technology(ICT)sectorandthedigitaleconomy,BEAuseditsICTsectorasthestartingpointformeasuringthe digital economy. To aid with identifying additional digital economy goods and services toinclude in theestimates,BEAreliedonanalystexpertiseandexistingdigitaleconomy literatureandstatistics,includingworkfromtheOrganisationforEconomicCo-operationandDevelopment(OECD). BEAincludesinthedigitaleconomy:(1)thedigital-enablinginfrastructureneededforacomputer network to exist and operate, (2) the digital transactions that take place using thatsystem(“e-commerce”),and(3)thecontentthatdigitaleconomyuserscreateandaccess(“digitalmedia”).

IdentifyingdigitaleconomygoodsandservicesUsing this definition and the existing detailed data from the supply-use tables (SUTs), BEAidentifiedgoodsandservicesforinclusioninthedigitaleconomyestimates.BEAclassifiesdataintheSUTsusingaframeworkthatincludesabout5,000categoriesofgoodsandservicesbasedontheNorthAmericanIndustrialClassificationSystem(NAICS).

1

2

3

4

– 2 –

Page 3: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

BEA selected over 200 goods and services categories for inclusion in the digital economyestimates.TheAppendixof theMarch2018workingpaper “DefiningandMeasuring theDigitalEconomy”listsallthegoodsandservicesincludedintheestimates.ThegoodsandservicesBEAincluded in the estimates are only a subset of all the goods and services identified in theconceptual definition of the digital economy. As noted in the introduction, some NAICS-basedgoodsandservicescategoriesincludedigitalgoodsandservicesaswellasnondigitalgoodsandservices, or they are only “partially digital.” While BEA’s conceptual definition of the digitaleconomy includes all digital goods and services, BEA included only primarily digital goods andservicescategoriesintheestimates.

Theseestimatesserveasthebaseforfutureestimatesthatwillalsoincludepartiallydigitalgoodsandservices.Includingthelatterrequiresadditionaldataandresearchtosplitouttheportionofthe commodity value that is in scope for the digital economy. The last section of this researchnotessomeoftheeffortsrelatedtomeasuringpartiallydigitalgoodsandservices.

Digital-enablinginfrastructureDigital-enabling infrastructure is comprised of the basic physical materials and organizationalarrangementsthatsupporttheexistenceanduseofcomputernetworksandthedigitaleconomy.Table A describes the sub-components included in digital-enabling infrastructure and indicatesBEA’scoverageofthesub-componentinthecurrentdigitaleconomyestimates.

BEAMethodologyforEstimatingSupply-UseTables

TheSUTsareanintegralandessentialelementoftheU.S.economicaccounts.First,theyarethebuildingblocksforothereconomicaccounts.ProminentamongtheseareBEA’sNationalIncomeandProductAccounts(NIPAs),whichfeaturetheestimatesofexpenditure-basedGDP.Second,theSUTsshowhowindustriesinteract;specifically,theyshowhowindustriesprovideinputto,anduseoutputfrom,eachothertoproduceGDP.Theyareacomplete,balancedsetofeconomicstatistics,andtheypresentafullaccountingofindustryandfinal-usetransactions.

ThecoreoftheSUTsconsistsoftwobasicnational-accountingtables—a“supply“tableanda“use“table.Thesupplytableshowsthecommoditiesthatareavailablefordomesticconsumption.Theusetableshowstheinputstoindustryproduction(intermediateinputs)andthecommoditiesthatfinalusersconsume.TheusetableisthemostfrequentlyrequestedtablebecauseofitsapplicationstotheestimatesofGDP.

BEAusesNAICStoclassifyindustries.TheUnitedStates,Canada,andMexicojointlydevelopedthisclassificationsystemwiththeaimofimprovingthecomparabilityoftheireconomicstatistics.NAICSclassifiesindustriesbasedontheirproductionprocesses.TheNAICScodescomprisesixdigits,whichreadingfromlefttoright,indicatethegeneralsectordowntoadetailedindustry.

TheU.S.statisticalsystemdoesnotcurrentlyhaveaseparateclassificationsystemforcommodities,whicharegroupsofsimilarproductsdefinedbythecharacteristicsoftheproduct(commodity)itselfratherthanbytheproductionprocess.Atpresent,BEAusesacommodityclassificationsystemtoassigneachcommoditythecodeoftheindustryinwhichthecommodityistheprimaryproduct.Thefoundationforthiscommodityclassificationsystemisthesix-digitNAICScode.

BEApreparesbenchmarkSUTsroughlyevery5yearsbasedonthehighestqualitysourcedata,notablytheU.S.CensusBureauEconomicCensus.Largelybecauseoftheirrichsourcedata,thebenchmarkSUTsarethemostimportantstatisticalsourceofinformationforcomprehensiveupdatesoftheNIPAsandarewidelyusedbyotherstatisticalagencies.BEAreleasedthemostrecentbenchmarkSUTsin2018.Theseaccountscover2012andusethe2012NAICSforclassification.

– 3 –

Page 4: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

TableA.Digital-EnablingInfrastructureintheDigitalEconomyEstimates

Infrastructuresub-component Sub-componentdescription

Currentlyincludedin

digitaleconomyestimates?

ComputerhardwareThemanufacturedphysicalelementsthatconstituteacomputersystem,including,butnotlimitedto,monitors,harddrives,semiconductors,wirelesscommunicationsproducts,andaudioandvisualequipmentproducts.

Includedalmostcomprehensively

SoftwareTheprogramsandotheroperatinginformationusedbydevicessuchaspersonalcomputersandcommercialservers,includingbothcommercialsoftwareandsoftwaredevelopedin-housebyfirmsfortheirownuse.

Includedalmostcomprehensively

Telecommunicationsequipmentandservices

Theequipmentandservicesrequiredforthedigitaltransmissionofinformationoveradistancebycable,telegraph,telephone,broadcasting,orsatellite.

Includedalmostcomprehensively

Structures

Theconstructionofbuildingswheredigitaleconomyproducerscreatedigitaleconomygoodsorsupplydigitaleconomyservices.Thestructurescategoryalsoincludesbuildingsthatprovidesupportservicestodigitalproducts.Thisincludestheconstructionofdatacenters,semiconductorfabricationplants,theinstallationsoffiberopticcables,switches,repeaters,andsoforth.

Notyetincluded;partofongoingwork

InternetofthingsInternet-enableddeviceslikeappliances,machinery,andcarswithembeddedhardwareallowingthemtocommunicatewitheachotherandconnecttotheinternet.

Notyetincluded;partofongoingwork

Supportservices Servicesnecessaryforthefunctionofdigitalinfrastructuresuchasdigitalconsultingservicesandcomputerrepairservices.

Includedalmostcomprehensively

Forbothstructuresandinternetofthings(IoT)infrastructure,BEAdoesnotcurrentlyhavedataavailabletoseparatedigitaleconomyactivityfromallotheractivity.ThecaseofIoTinfrastructurepresentsadditionalchallenges.Forexample, theconnectivityofan internet-enabledrefrigeratormayallowtheownertotrackandpurchasefooditemswhentheyarerunningloworrecordusageof the appliance.However, theprimary function of the refrigerator is to keep food cold, outputwhichBEAwouldnotclassifyasbeingpartofthedigitaleconomy.

E-commerceBEAusestheterm“e-commerce”tobroadlydescribeallpurchasesandsalesofgoodsandservicesthatoccurovercomputernetworks.E-commercereflectsthenatureofatransactionforgoodsorservices.BEAconsiderse-commercetoincludedigitallyordered,digitallydelivered,orplatform-enabledtransactions.

E-commerce output is generallymeasured as thewholesale or retail trademargin on “digitallyordered”goodsandservicessoldovertheinternetorthroughsomeotherelectronicmarket.Themarginisequaltototalrevenueearnedfromonlinesaleslesstheproducercostofthegoodsandservices.BEA included in the estimates themargin activityof electronicmarket establishmentsand somenonmargin activity as described in tableB. For peer-to-peer (P2P) e-commerce, BEAcaptures theassociatedtransactions,but iscurrentlyresearchingdataandmethodologytosplitoutthevalueofthisoutputinscopeforthedigitaleconomy.

– 4 –

Page 5: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

TableB.E-CommerceintheDigitalEconomyEstimatesE-

commercesub-

component

Sub-componentdescription Currentlyincludedindigitaleconomyestimates?

Business-to-business(B2B)e-commerce

E-commercethatutilizestheinternetorotherelectronicmeanstoconducteitherintrafirmorinterfirmtransactionsofgoodsandservicesbybusinessestootherbusinesses.

IncludesthemarginonB2Bwholesaleactivityfromelectronicmarketestablishments.Activityfromothertypesofestablishmentsisnotincluded.AlsoincludesnonmarginoutputforB2Belectronicmarkets.

Business-to-consumer(B2C)e-commerce

E-commercethatutilizestheinternetorotherelectronicmeanstoconductthesaleofgoodsandservicesbybusinessestoconsumers,alsoknownasretaile-commerce.

IncludesthemarginonB2Cretailactivityfromelectronicmarketestablishments.Activityfromothertypesofestablishmentsisnotincluded.Alsoincludesnonmarginretailtradefromelectronicauctions.

Peer-to-peere-commerce

The“sharingeconomy,”orplatform-enablede-commerce,betweenconsumersthatinvolvestheexchangeofgoodsandservicesforafeeandfacilitatedthroughadigitalapplication.Includes,butisnotlimitedto,accommodationrentals,transportation/ride-hailing,deliveryandcourierservices,andconsumergoodsrental.

Notseparatelyidentifiedfromtraditionalindustries;partofongoingworktoseparatelyidentifywithintheaccount.

DigitalmediaIncreasingly,consumersforgothephysicalpurchaseorrentalofproductslikebooks,newspapers,music,andvideos,optinginsteadtoaccesstheseproductsonlineinadigitalformat.BEAdefinesdigitalmediaasthecontentthatpeoplecreate,access,store,orviewondigitaldevices.

TableCdescribeshowdigitalmediaiscapturedintheBEAdigitaleconomyestimates.Onlydirect-saledigitalmediaiscurrentlyincluded.BEA’snationalaccountsdatadoescapturetheadvertisingrevenueassociatedwith freedigitalmediasites;however,BEAdoesnotcurrentlyhavethedataneededtoidentifywhatportionofadvertisingrevenueisassociatedwiththesesites.

TableC.DigitalMediaintheDigitalEconomyEstimatesDigital

mediasub-component

Sub-componentdescription Currentlyincludedindigitaleconomyestimates?

Direct-saledigitalmedia

Businessesmayselldigitalproductstoconsumersforafee,eitheronanitem-by-itembasisorthroughasubscriptionservice.

Includesdatastreamingservices(bothsubscriptionandone-timepurchasesofcontent),internetpublishing,andinternetbroadcasting.

“Free”digitalmedia

Somecompaniesofferdigitalmediatoconsumersatnomonetarycost,suchasmuchofthecontentonYouTubeorFacebook.Typically,businessesofferingtheseservicesearnrevenuebysellingadvertisingspaceonthemarginsofthedigitalproduct.Thiscouldalsoincludepeer-to-peerdigitalmediaproducedbyconsumersforpeerconsumption.

Notseparatelyidentified;partofongoingwork.

Bigdata

Somecompaniesgeneratelargedatasetsaspartoftheirregularoperations.Thiscouldalsoincludedataonconsumerpreferencesorbehaviorcollectedthroughtheiruseofdigitalmedia.Companiesmayearnrevenuebysellingthisinformationorleveragingitinotherways.

Notseparatelyidentified;partofongoingwork.

– 5 –

Page 6: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

IdentifyingdigitaleconomyindustriesandpreparingresultsBEAestimatednominalvalueadded,output,compensation,andemploymentbyindustryforthedigital economy. After identifying the goods and services included in the digital economy, BEAidentified the industries that produce these goods and services using the supply table. Digitaleconomygrossoutputbyindustryrepresentsthetotalvalueofin-scopegrossoutputproducedbyeachindustryacrossalldigitaleconomygoodsandservices.Valueaddedforthedigitaleconomyis derived from the relationship between the industry output for the digital economy and totalindustryoutput.Thismeanstheratioof intermediateconsumptionassociatedwiththeindustryoutput for the digital economy is assumed to be the same as the ratio of total industryintermediateconsumptiontototalindustryoutput.Compensationandemploymentforthedigitaleconomy are derived through the same procedure as value added. Specifically, the ratio of anindustry’s digital economy output to total output is applied to total employment andcompensationfortheindustry.

BEA prepared price and quantity indexes for digital economy gross output and value added inthree steps. First, gross output indexes are derived by deflating each digital good and serviceproducedbyanindustrythatisincludedaspartofitsgrossoutputfromthesupplytable.Second,BEAderivedindexesforintermediateinputsbydeflatingallcommoditiesfromtheusetablethatare consumed by the industry as intermediate inputs in the production of digital goods andservices. Domestic and international sources of intermediate inputs are deflated separately byusingtheimportproportionality,orcomparability,assumption.Third,BEAcalculatedindexesforvalueaddedbyindustryusingthedouble-deflationmethodinwhichrealvalueaddediscomputedas thedifferencebetweenrealgrossoutputandreal intermediate inputswithinaFisher index-numberframework.

Results

GrossdomesticproductorvalueaddedBEAestimatesshowthedigitaleconomyhasbeenabrightspot in theU.S.economy.GDP is thevalueofthegoodsandservicesproducedbythenation’seconomylessthevalueofthegoodsandservices usedup in production. GDPby industry, or value added, is ameasure of an industry’scontribution tooverallGDP.Thedigital economyaccounted for6.9percent ($1,351.3billion)ofcurrent-dollarGDP($19,485.4billion)in2017,upfrom5.9percentin1997.WhencomparedwithtraditionalU.S.industriesorsectors,thedigitaleconomyrankedjustbelowprofessional,scientific,and technical services,whichaccounted for7.4percent ($1,450.0billion)ofcurrent-dollarGDP,and just abovewholesale trade, which accounted for 6.0 percent ($1,174.1 billion) of current-dollarGDP(chart1).

– 6 –

Page 7: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

In 2017, digital economy real (inflation-adjusted) value added totaled $1,483.5 billion. BEAestimates show thedigital economywasanengineofGDPgrowth throughout theperiod thesestatistics cover. From 1998 to 2017, real value added for the digital economy outpaced overallgrowthintheeconomyeachyearandmitigatedthedownturninGDPduringtherecessionin2008and 2009 (chart 2). Excluding the recessionary years of 2008 and 2009, digital economy realvalue-added, year-over-yeargrowth for thisperiodwasonaverage four timeshigher than totalGDPgrowth.Digitaleconomyrealvalueaddedgrewatanaverageannualrateof9.9percentperyearfrom1998to2017,comparedto2.3percentgrowthintheoveralleconomy.

The relative strength of the digital economy led it to consistently contributemore to economicgrowththanitsshareoftheeconomy.Forexample,in2017,thedigitaleconomyaccountedforjust6.9 percent of the economy; however, real value-added growth of 8.3 percent in the digitaleconomyaccounted for25percent,or0.55percentagepoint,of the total2.2percentgrowth inrealGDP.

– 7 –

Page 8: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

Overtheentire1998to2017period, thedigitaleconomyexperiencedrealvalue-addedaverageannualgrowthof9.9percent,comparedto2.3percentforthetotaleconomy(chart3).Withinthedigital economy, the hardware component grew the fastest, with 18.4 percent average annualgrowth over the period. However, the growth rate in hardware varied notably between thebeginning and end of the periodmeasured, ranging from 33.6 percent average annual growthbetween1998and2007,downto5.0percentaverageannualgrowthfrom2008to2017.Whilehardwaregrewthefastestovertheperiodfrom1998to2017,growthslowedtoonly5.0percentat an average annual rate in the most recent decade, the slowest of any digital economycomponent.

Software and telecommunications exhibited slower average annual growth in the more recentyears,buttoamuchlesserextent,withaverageannualgrowthfallingabout4percentagepointsbetween the first and second decades in the time series for both components. For software,growthfellfrom9.9percentaverageannualgrowthfrom1998to2007to5.8percentfrom2008

– 8 –

Page 9: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

to2017,whiletelecommunicationsfell from9.4percentto5.2percentoverthesameperiod.E-commerce and digital media exhibitedmore stable average annual growth (12.2 percent from1998to2007and11.1percentfrom2008to2017).Supportserviceswastheonlycomponentthatgrewfasterduringthesecondhalfoftheperiod,withaverageannualgrowthof7.0percentfrom1998to2007and8.8percentfrom2008to2017.

Overall, average annual growth of 18.9 percent in real value added of digital goods outpaceddigitalservices,whichmeasured7.7percent.Inthelaterdecade,from2008to2017,theaverageannualgrowthratesfordigitalgoodsanddigitalservicesweremorealignedat8.0percentand6.9percent,respectively.

From1997to2017,digitaleconomycurrent-dollarvalueaddedaccountedforanaverageof6.1percentoftotalU.S.current-dollarGDPeachyear(chart4).ThedigitaleconomyshareoftotalGDPhas generally been rising, and in 2017, it accounted for 6.9 percent of the total economy, thehighestshareoverthistimespan.Inthisyear,digitaleconomycurrent-dollarvalueaddedtotaled$1,351.3billion.

Incurrentdollars,digitaleconomyvalueaddeddecreasedin2001;thiswastheonlyyearintheseries inwhich year-over-year, current-dollar growthwas negative. From1998 to 2001, digitaleconomypricesrapidlydecreased.Thevalue-addedpriceindexfellby7.3percentfrom2000to2001,with the index forhardware fallingby33.9percent as thequality of computerhardwareincreasedrapidly.

In 1997, digital-enabling infrastructure, or the hardware, software, telecommunications, andsupportservicesthatcomprisethedigitaleconomy,accountedfor95.0percent($480.1billion)ofdigital economy current-dollar total value added ($505.3 billion) (chart 5). In 2017, digital-enabling infrastructure accounted for $1,143.4 billion (84.6 percent) of the total estimated$1,351.3 billion in digital economy current-dollar value added. E-commerce and digital mediagrew fromaccounting for just5.0percent ($25.2billion)of thedigitaleconomyto15.4percent($207.8billion)overthe20-yeartimespan.

– 9 –

Page 10: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

While e-commerce and digital media became a larger share of digital economy current-dollarvalueadded, therewerealso shifts in the compositionof the components thatmakeupdigital-enablinginfrastructure.In1997,telecommunicationsaccountedfor40.7percent($205.5billion)oftotaldigitaleconomycurrent-dollarvalueadded.By2017,telecommunicationscurrent-dollarvalueaddedhadgrownto$339.6billion,butitsshareoftheoveralldigitaleconomyhadfallento25.1 percent. Hardware declined in its total value and in its contribution to digital economycurrent-dollarvalueadded,accountingfor19.8percent($100.3billion)ofthetotal in1997andjust7.1percent($96.3billion)in2017.Supportservicesandsoftware,theothertwocomponentsofdigital-enablinginfrastructure,contributedmoretodigitaleconomycurrent-dollarvalueaddedin2017comparedto1997,growingfrom23.1percentto34.7percentand11.4percentto17.7percent,respectively.

GrossoutputGross output is ameasure of sales or revenue from production formost industries. Real grossoutput for thedigital economygrewatanannual rateof6.3percent from1998 to2017, fasterthanthetotaleconomy,whichgrewatanaverageannualrateof2.1percent.Thecompoundeffectofthefasteroutputgrowthinthedigitaleconomyrelativetotheoveralleconomyisclearlyseenwhenoutput is indexedtoabaseyear(chart6).Sincetheendof themostrecentrecession, theoutput growth from the digital economy has continued to outpace the overall economy. Since2010, digital economy real gross output growth averaged 2.5 percent per year and outpacedaverageannualrealgrossoutputgrowthof1.0percentintheoverallU.S.economy,wideningthedistancebetweenthegrossoutputindexesdisplayedinchart6.

Duringthisperiodofeconomicrecovery,pricesfordigitaleconomygoodsandservicesdecreasedatanaverageannualrateof1.0percent(chart7).Pricesforallgoodsandservicesintheeconomyincreasedatanaverageannualrateof1.6percent.

– 10 –

Page 11: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

Althoughrealoutput inthedigitaleconomyacceleratedfasterthanrealoutput fortheeconomyoverall,thefallingpricesofdigitalgoodsandservicescausedcurrent-dollargrossoutputgrowthin thedigitaleconomyandtheoveralleconomyto trackmuchcloser(chart8).Digitaleconomynominal gross output reached $2.05 trillion in 2017, totaling 6.0 percent of total U.S. nominalgrossoutput.

EmploymentandcompensationIn2017,thedigitaleconomysupported5.1millionjobs,whichaccountedfor3.3percentoftotalU.S.employment(152.1millionjobs),comparabletothetransportationandwarehousingindustry(chart 9). Employees working in the digital economy earned $132,223 in average annualcompensation compared to $68,506 average annual compensationperworker for the totalU.S.economy.

– 11 –

Page 12: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

NextSteps

BEAisworkingtoexpandthecoverageofthedigitaleconomyestimatestoincludeadditionale-commerce activity and selected partially digital commodities. The feasibility of expanding thecoverage is conditional on data availability. BEA is also exploring data and methodologies toinclude activities of digital intermediaries, such as ride-hailing activity. At this time, cloudcomputingestimatesare fully included inBEAdigitaleconomyestimatesbutarenotaseparatelineitem,likehardwareorsoftware.SomeoftheNAICS-basedservicescategoriesincludeoutputrelated to cloud computing services aswell as unrelated output. BEA needs additional data toestimate the relevant portions to separately value clould computing in the digital economyestimates.

BEAactivelyparticipatesinOECDmeetingsondevelopingdigitalSUTswiththecontinuedgoalofimplementinginternationalguidanceonmeasuringthedigitaleconomy.BEAwillalignitsdigitaleconomy measures to international standards to the extent possible. BEA is assessing thefeasibilityofpublishingacompletesetofSUTsforthedigitaleconomyinlate2019.

– 12 –

Page 13: Research Spotlight Measuring the Digital Economy · The U.S. statistical system does not currently have a separate classification system for commodities, which are groups of similar

1. Formoreinformationonthemethodologyandsatelliteaccounts,see“MeasuringtheNation’sEconomy:AnIndustryPerspective.APrimeronBEA’sIndustryAccounts&rquo;.TheindustryaccountsareonecomponentoftheU.S.economicaccountsthatprovideinformationonthevalueandcompositionofoutputproducedintheUnitedStatesandonthetypesofincomegenerated.

2. BEAclassifiesgoodsandservicesusingasystembasedontheNorthAmericanIndustrialClassificationSystem.3. TheBEAICTsectorconsistsofcomputerandelectronicproductmanufacturing(excludingnavigational,measuring,electromedical,andcontrolinstrumentsmanufacturing);softwarepublishers;broadcastingandtelecommunications;dataprocessing,hosting,andrelatedservices;internetpublishingandbroadcastingandwebsearchportals;andcomputersystemsdesignandrelatedservices.BEA’sdefinitionisgenerallyconsistentwiththeinternationallyaccepteddefinitionoftheICTsectorusedanddevelopedbythestatisticalofficesoftheOrganisationforEconomicCo-operationandDevelopmentandUnitedNations.

4. ForinformationontheOECD’sdigitaleconomymeasurementwork,seehttp://www.oecd.org/sti/ieconomy/.BEAalsoparticipatesintheOECDworkinggrouponMeasuringGDPinaDigitalizedEconomy.

SurveyofCurrentBusinesswww.bea.gov/scb

[email protected](301) 278-9004

– 13 –