How global climate policy could affect competitiveness

(1)

ContentslistsavailableatScienceDirect

Energy

Economics

j o ur na l h o me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / e n e e c o

How

global

climate

policy

could

affect

competitiveness

夽

Hauke

Ward

a,b,c,∗

_,

_Jan

_Christoph

_Steckel

b,c

_,

_Michael

_Jakob

b a_Institute_of_{Environmental}_Sciences_(CML),_Department_of_Industrial_Ecology,_Leiden_University,_the_Netherlands b_Mercator_Research_Institute_on_Global_Commons_and_Climate_Change,_Torgauer_Straße₁₂_–_15,_10829,_Berlin,_Germany c_Potsdam_Institute_for_Climate_Impact_Research,_Leibniz_Association,_Postfach₆₀₁₂_03,_14412,_Potsdam,_Germany

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received18January2019

Receivedinrevisedform4October2019 Accepted14October2019

Availableonline2November2019 JELclassiﬁcation: C67 F16 F18 L50 L60 Q56 Keywords: Globalcarbonprice Input-outputanalysis Energy-intensive Trade-exposedindustries Competitiveness Labormarket Globalsupplychains

a

b

s

t

r

a

c

t

Aglobaluniformcarbonpricewouldbeeconomicallyefficientandatthesametimeavoid ‘carbon-leakage’.Still,itwillaffectthecompetitivenessofspecificindustries,economicactivityandemployment acrosscountries.Thispaperassessesshort-termeconomicshocksfollowingtheintroductionofaglobal carbonpricethatwouldbeinlinewiththeParisAgreement.BasedontheWorldInput-OutputDatabase (WIOD),wetracethecarboncontentoffinaloutputthroughglobalsupplychains.Thisallowsusto estimatehowpricesofthefinaloutputwouldreacttotheintroductionofaglobalcarbonprice.We findthatimpactsonindustrialcompetitivenessarehighlyheterogeneousacrossregionsandeconomic sectors.ThecompetitivepositionofBrazil,Japan,theUSAandadvancedeconomiesoftheEUislikely toimprove,whereasindustriesandlabormarketsinnewlyindustrializingAsianeconomiesaswellas EasternEuropearelikelytoexperiencesubstantialadverseimpacts.

1. Introduction

AchievingtheclimatetargetsoftheParisAgreement(UNFCCC, 2015)requires substantial reductions in global greenhousegas (GHG) emissions.Under thecurrent global climategovernance

architecture, members of the Paris Agreement pledged

volun-taryunilateralclimatetargetsandmitigationpoliciesasspecified in theirNationally Determined Contributions(NDCs). However, the implementationof ambitious unilateral climatepolicies by first moversoften facesresistance due to concerns about ‘car-bon leakage’, i.e. the fear that energy-intensive, trade-exposed (EITE)industriescouldrelocatetocountrieswithlaxerclimate poli-ciesorlessefficientproductiontechnologies.Countriesadopting ambitiousclimatepoliciesmightthenruntheriskofincurring sub-stantialcoststoreducetheiremissionswhileatleastsomeofthe emissionreductionsachievedaresimplytransferredelsewhere.

夽 PublicationofthissupplementwassupportedbyETHZürich,theUniversityof MünsterandEconomicsforEnergy.

∗ Correspondingauthorat:InstituteofEnvironmentalSciences(CML), Depart-mentofIndustrialEcology,LeidenUniversity,P.O.Box95182300RA,Leiden,the Netherlands.

E-mailaddress:h.ward@cml.leidenuniv.nl(H.Ward).

Economistshavefora longtime advocateda uniformglobal priceonGHGemissions(Cramtonetal.,2017;Edenhoferetal., 2015;High-LevelCommissiononCarbonPrices.,2017;Weitzman, 2014).Thiswouldbe themostcost-efficientpolicyinstrument, guaranteeingthatabatementtakesplaceforactivitiesforwhich emissionscanbereducedintheleast-costmanner.Moreover,a globalcarbon pricewould preventcarbonleakage, asno coun-trywouldbenefitfromanartificialcomparativeadvantagearising fromthelackofclimatepolicy.Yet,itposesahighercostburden onmorecarbon-intensiveproducers(FullertonandMuehlegger, 2019).Hence,auniformglobalcarbonpricecouldhaveimportant consequencesforindustrialcompetitiveness,nationalgrossvalue addedandemployment.Whiletheacademicliteraturehas exten-sivelyexaminedtheeffectofunilateralclimatemeasuresoncarbon leakageandcompetitiveness1_,_there_is,_to_our_knowledge,_only_very

limitedunderstandingoftheimplicationsofglobalclimatepolicy.

1_See_Böhringer_et_al.,₂₀₁₂_;_Branger_and_Quirion,₂₀₁₄_;_Carbone_and_Rivers,₂₀₁₇_; DechezleprêtreandSato,2017;Jakobetal.,2014;Forinetal.,2018;KimandKim, 2012;Sayginetal.,2011;Voigtetal.,2014.

https://doi.org/10.1016/j.eneco.2019.104549

(2)

Thispaperaimstoﬁllthisgapbyprovidinginsightsintothe short-termcompetitivenessimpactsinducedbya globalcarbon price.OuranalysisisbasedontheWorldInput-OutputDatabase (WIOD) (Timmer et al., 2015), a multi-regional Input-Output (MRIO)database.Weusethisdatatoderivethe‘NormalizedNet CarbonContent’(NNCC)ofallglobalsupplychains,i.e.the emis-sionsgeneratedperUSDofoutputfromagivensectorinagiven countryovertheentiresupplychain.Thisinformationcanbeused todeterminethepriceincreasethat wouldresultfromaglobal carbonpriceineachsectorandcountry.Thusitcanhelpto iden-tifysectorsandcountriesfacingthemostsevererisksofadverse impactsoncompetitiveness,valueaddedandemployment.

Vulnerableindustriesarethosethatdisplaybothasignificant relative price increase and a highcontribution to the national valueaddedwithinthecorrespondingeconomy.Thesesectorscan beexpectedtolosemarketshare,profitabilityandemployment. Our resultsindicate that a uniformglobal carbon price would, atleastintheshortterm,improvethecompetitivenessof coun-trieswithlow-carbonenergysystems(e.g.BrazilorFrance) and efficientproductiontechnologies(particularlyindustrialized coun-tries). By contrast, developing or transitioning economies with carbon-intensiveenergysystems, suchasChina,India and Rus-sia,canbeexpectedtoexperiencesubstantialpriceincreasesin someenergy-intensiveindustries.Overall,agloballyuniform car-bonpricingpolicywouldprobablymakeWesternindustriesmore competitive,althoughnegativeinitialimpactscanbeexpectedfor EasternEuropeanand fast-growingAsian countries2 _._The_most

severe impacts would probably affect low-skilled workers and henceprobablythepoorestsegmentsofsocieties.

Thispaperproceedsasfollows.Section2introducesthedataand methodology.Section3presentstheresults.Section4discusses policyimplicationsandconcludes.

2. Dataandmethods

Ouranalysesarebuiltonaninnovativewaytoestimatethe Nor-malizedNetCarbonContent(NNCC)ofglobalsupplychains.We buildouranalysesonmulti-regionalinputoutputmodelingbased ontheWorldInputOutputDatabase(WIOD).Inthissectionwe willﬁrstdescribethedataweuse.Afterwards,wewilldescribe howwecalculatetheNNCC.BasedontheNNCC,weestimatethe implicationsofimplementingaglobalcarbonpriceoncountries’ economicperformanceandlabormarket.Finally,wewillcritically discussthemethodologicalapproachtakeninthispaper.The fol-lowingsub-sectionsdescribetheunderlyingdata,themethodology toestimatethecarboncontentoftradeandhowitcanbeusedto assessthecompetitivenessimpactsofaglobalcarbonprice.We thendiscussthelimitationsofourapproach.

2.1. Data

We use the World Input-Output Database (WIOD) (Timmer

etal.,2015)andthecorresponding‘satellitedata’onenergy-related CO2emissionsforeachsectorineachregion3.Thedatasetreﬂects

41 regions and 35 sectors. WIODincludes the EU27 (which is

thecurrent EuropeanUnionwithout Croatia), aswellas major economies(includingAustralia,Canada,Japan,Mexico,SouthKorea andUnited States),and newlyindustrialized economies (Brazil, China,Indonesia,India,Russia,TaiwanandTurkey).Theseregions accountforapproximately85%ofworldGDP(Timmeretal.,2015). Theremainderisincludedinaresidualregion,referredtoasthe

2 _For_a_spatial_{distribution,}_see_Fig._A3_in_the_{Supplementary}_information 3 _We_do_not_consider_emissions_from_land_use_and_land_use_change.

‘RestoftheWorld’(ROW).4 _We_use_the_year_2009,_as_it

consti-tutesthemostrecentreleasethatincludesemissionsdata.Thishigh regionalandsectoralresolutionenablesadetailedunderstanding ofthecarboncontentofglobalsupplychains.

2.2. Estimatingthenormalizednetcarboncontent

Weapplyamodifiedsupply-sidefootprintanalysistoidentify thecarboncontentofglobalsupplychains.IntheMRIOdata, out-putfromacertainsectorinoneregionemployedasaninputfor production in a differentsector and/orcountryis described by theinter-industryflowmatrixZ ∈R(m∗n)×(m∗n)_,_with_m_being_the numberofsectorsandnbeingthenumberofregions.Final con-sumption,bysectorandcountry,ofallcountriesisdenotedbythe finaldemandmatrix Y ∈R(m∗n)×n_._The_elements_of_Z _are_given byZr,sr,s,whichdenotesallmonetaryflowsfromregionr,sectors toregionr’,sectors’Correspondingly,YconsistsofelementsYr

r,s, whichrepresenttheaggregatedmonetaryﬂowfromregionr,sector sintotheﬁnaldemandofregionr.

Inordertoassessthecarboncontentoffinalconsumption,we needtoderivetheLeontiefinverseL,whichspecifiestheinputs usedtogenerateaunitoffinaloutputovertheentiresupplychain (MillerandBlair,2009).Forthisreason,wedividetheelementsof Zbythecorrespondingtotalsectoraloutput Or,s=

r

s Zrs rs +

r Yr

r,s.Thetechnologymatrixthatreﬂectsallnecessarydirect inputstoproduceoneunitofoutputineachsectorisdenotedby A.TheLeontiefinverseisthencalculatedasL= (I−A)−1,whereI denotestheunitymatrix.Torelatemonetaryinputsfromdifferent sectorstoCO2emissions,weusedataonemissionFrs bysectors inregionr,whichwedividebysectoraloutputsOtoarriveatCO2 emissionsperUSD.

The total CO2 emissions ˆfrs associated withone unit of sec-tor s in region r for ﬁnal consumption can then be expressed as ˆfs r =

r

s fs rL r,s

r,s.Wecalltheemissionsassociatedwiththe entireglobalsupplychainofaunitofﬁnaloutputthe‘Normalized NetCarbonContent’(NNCC).NNCCallowsthecomparisonof asso-ciatedcarbonperunitofoutput,andrelatedpriceincreaseswithin sectorsandacrosscountries.Thisapproachyieldsasubstantially higherresolution thanaconventional carbonfootprint analysis, wheretwo vectorsaremultiplied(i.e.,(f·L)·Y),hence omitting importantsectoral,aswellasregional,details.

2.3. Assessingimpactsoncompetitivenessandemployment

Ouranalysisassumesthatagloballyuniformcarbonpriceis implementedand thattheassociated priceincreasesfor manu-facturedgoods are fullypassed through toconsumers (Kenkel, 2005), regardlessofwhetherthecarbon priceis appliedatthe pointofextraction,production orconsumption(Karstensen and Peters,2018).Thisisinlinewiththeobservationthatintheshort term,priceincreasesresultingfromfueltaxesaremainlyborneby consumers(ChouinardandPerloff,2004;MarionandMuehlegger, 2011).Weassumeperfectsubstitutabilityofeconomicoutputfrom agivensectoracrosscountries,i.e.thatallﬁnaloutputofagiven sectorisproducedfortheglobalmarket.Theshort-termimpacton industrialcompetitivenessisthendeterminedbythepriceincrease andtheassociated demand-sidereactionwithineach economic

(3)

sector5_._Our_analysis_considers_that_changes_in_the_demand_of_ﬁnal

goodswillhaverepercussionsforallsectorsformingpartoftheir globalsupplychain.Ifaglobalcarbonpricepisimplementedand passedontotheconsumer,ﬁnalgoodsfromsectorsinregionr willseeapriceincreaseofpr

s=p· ˆf s

rperUSD.Forouranalysis, weassumeaglobalcarbonpriceofUSD50pertCO2,whichwould beroughlyinlinewiththe2◦Ctarget(High-LevelCommissionon CarbonPrices.,2017).

Weassumethatallﬁnaloutputisexportedtotheglobalmarket, fromwhichitmaybeconsumedeitherintheregionwhereitwas produced,asdomesticconsumption,orinadifferentregion.We employapriceelasticityofexportvaluesofı=−2.84fromSolleder (2013)6_to_assess_changes_in_ﬁnal_output_resulting_from_these_price

changesbysectorandcountry.Weusethisparametertoproject changesinﬁnaloutputasafunctionofpricechangesrelativeto theglobalaverage.Thatis,weassumethatsectorsinregionsthat aresubjecttoapriceincreasethatishigher(lower)thantheglobal average,experiencealoss(gain)inmarketshare.Usingtheabove elasticitytomodifyeachentryoftheﬁnalconsumptionmatrixY givesusthematrixY∗.Theelementsofthismatrixarecalculatedas y∗rr,s=yr

r,s· (1+prs− ¯ps)ı,with ¯psbeingtheglobalmeanprice changeforsectors.Y∗canthenbeusedtocalculateallsectoral outputswithaglobalcarbonprice,O∗=LY∗.

ThematrixZcanalsobeusedtodeterminepaymentsfor pro-ductionfactors,suchaswagesandreturnstostakeholders,asthe differencebetweenoutputsfromonesectorandthissector’sinputs fromothereconomicsectors.Theinputsofsectorsofcountryr arecalculatedasIr,s=

r

s

Z_rr,s_,s.Hencethecorrespondingvalue

added(VAD)canbecalculatedasVADr,s=Or,s−Ir,s.TheVADper unitofoutputisthengivenby

v

adr,s=VADr,s/Or,s.

MultiplyingthenewlyderivedelementsofO∗withthe corre-spondingelements

v

adallowsustoprojectchangesinregionaland sectoralvalueadded.

Finally,WIODalsoaccountsforsectorallaboruseLU(inhoursfor threedifferentlevelsofqualification,low-,medium-,and highly-skilled).WecalculatethesectorallaborintensityLIr,sq insectors, regionrofqualificationqperunitofoutputasLIr,sq =LUr,sq /Or,s. Onthisbasis wecanprojecthowchangesinfinal outputrelate tochangesinlabordemandforeachsectorsandregionLUq∗r,s= LIr,sq ·O∗r,s.Thesumofallsectoralchangesineachregiongivesa roughindicationofimpactsonregionallabormarkets.

2.4. Discussionofthemethodology

Althoughthe analysisallows the identiﬁcationof the initial impactsofaglobalcarbonpriceatarelativelyhighlevelofsectoral detail,itfacesanumberoflimitations.

First, sectoral resolution is restricted by the availability of underlyingdata.Althoughtherearedatasetswithhigherregional resolution(Aguiaretal.,2016;Lenzenet al.,2013), wedecided touseWIOD,asit accountsfor thelargestand mostadvanced countrieswithhighqualitynationalaccounting.Thisisanasset (Steen-Olsenetal.,2015).

Second, using monetary input-output (IO) data implicitly assumeshomogeneityofproductsacrossregions.Thisassumption

5_Given_this_assumption,_we_use_a_price_elasticity_of_export_values_to_simulate_the

short-termimpactofpricechangesinﬁnaldemand.Ouranalysisassumesthatall commoditiesproducedinagivensectoraresubjecttothesameelasticity(however, thecompositionofcommoditiesproducedineachsectordoesneednotbeidentical acrosscountries).

6_The_elasticity_refers_to_changes_in_trade_value_caused_by_an_export_tax._The

ı-valueisclosetothenumbersobservedbyHeadandMayer(2014),who,ina meta-analysis,ﬁndamedianpriceelasticityfortradegravityequationsof-3.19.

Fig.1.Short-termpriceincreasesacrosscountriesfromaglobalUSD50carbon pricebyeconomicsector.Boxesrepresent25thto75thpercentiles,blackcenter linesrefertomedians,whiskersineachdirectioncorrespondtothe10thpercentile (lowerbound)andthe90thpercentile(upperbound).AnoutlierforCoke,Reﬁned PetroleumandNuclearfuelshasbeenexcludedfromtheplot12_._The_boxplots

indi-catetheheterogeneityincarboncontentofglobalsupplychainsofﬁnaldemand goodsendingintherespectiveregionalsector.

canbequestioned(forfurtherdetailsseee.g.Alexeeva-Talebietal. (2012);Steen-Olsenetal.(2015)andWardetal.(2017)). Consider-ingdifferencesinproductsacrosssectorsandregionswouldadd additionaldetail toourapproach.In thesamevein,usingmore heterogenic elasticitiesacross sectors could help reﬁne results. However,giventhatproductsinacertainsectoraresimilaracross countries,itseemsunlikelythatthisextensionwouldoverturnthe maininsightsofouranalysis.

Third,thispaperaimstoidentifytheshort-termimplicationsof priceshocks.Thistypeofanalysisishighlyrelevantinassessingthe politicaleconomicimplicationsofdifferentpolicies(Fullertonand Muehlegger,2019),but doesnotappropriatelycapturedynamic adjustment effects in the long run. Long-term effects, suchas changesinglobalsupplychains,substitutionofcarbon-intensive inputs,orinducingtechnologicalchange,couldbeassessedwith computablegeneralequilibrium(CGE) models(seee.g.Carbone andRivers(2017)andMattooetal.(2009)).However,CGE mod-elsareunabletoappropriatelycaptureshort-termeffectsandthe (potentiallycostly)adaptationprocessesfromonestateof equilib-riumtoanother(FullertonandMuehlegger,2019).Ourapproach, bycontrast,considersglobalsupplychainsandinternational con-nectionsata higherlevel ofdetail thana CGE7_, _thus_providing

importantcomplementaryinformation.

3. Results

Thissectionﬁrstassessesshort-termpriceincreasesthatwould occur,byeconomicsectorandregion,inresponsetothe introduc-tionofaglobalcarbonprice.Wethenusethisinformationtoassess

12_This_outlier_is_Estonia_with_a_53%_projected_price_increase._It_is_likely_to_be_caused

bythecomparativelylargeimportanceofEstonianshaleoilproduction(IEA,2009), seeTableA1intheSIforfurtherdetail.

7_Typically_nested_production_functions,_which_are_used_for_CGE_modelling_have

(4)

Fig.2.ProjectedpriceincreasesduetoacarbonpriceofUSD50pertCO2vs.theshareofagivensectorintheregion’stotalvalueadded.Thedottedlinesshowtheaverage

priceincreaseacrossthecountriesconsideredandtheaverageregionalVADshare.OnlyregionswithaVADsharegreaterthantheglobalmeanareplotted.Valuesforprice increasesandregionalvalueaddedforallsectorsaregiveninTableA1andTableA7intheSI,respectively.

potentialimpactsoneconomicoutput,valueaddedand employ-ment.

3.1. Impactsonindustries

Fig.1 assessestheshort-termeffectofa globalcarbonprice ofUSD50pertCO2 onthepricesoffinaloutputacrosssectors. Foranygiveneconomicsector,thereisasubstantialvariationin termsof NNCCfor finaloutput,which canbeattributedto dif-ferencesin productiontechnologiesand energy systemswithin supplychains.Thisvariationresultsinsubstantialdifferencesin pricechangesoffinaloutputacrosscountriesforanygiven eco-nomicsector.Insomecountries,priceincreasesformetalswouldbe morethan20%,andmorethan15%forchemicals,plastic,and(fossil andnuclear)fuels.Thehighestlevelsofheterogeneityacross coun-triesareidentifiedforCoke,RefinedPetroleumandNuclearFuelas wellasOtherNon-MetallicMinerals(refertoTableA1inthe Sup-plementaryinformationforfurtherdetails).Thelowest(average) priceincreasesareprojectedforTextilesandTextileProducts, Man-ufacturingnec8_,_Electrical_and_Optical_Equipment,_and_Transport

Equipmentand Machinerynec.Fromaregionalperspective,the highestpriceincreasesareprojectedinnewlyindustrializingAsian economies,suchasChina,IndonesiaandIndia.OtherEastAsian economies(suchasKoreaandTaiwan)aswellasEasternEuropean states(suchasRussiaor Bulgaria)alsodisplaycarbon-intensive supplyandproductionchains.Acarbonpricewouldresultin com-parativelylargeincreasesinthepricesoftheﬁnaloutputofthese regions.Incontrast,economieswiththeleastcarbon-intensive sup-plyandproductionchains(severalEUmembersaswellasBrazil) wouldseeonlymodestpriceincreases.

8 _nec=_not_elsewhere_classiﬁed

Forthreeofthefourmostcarbon-intensivesectors(Fuel,Other Non-Metallic-Minerals,ChemicalsandBasicMetals)andthe major-ityoftheothersectors,valueaddedispositivelycorrelatedwith carbon-intensity(seeTableA10intheSI).Thisobservationsuggests that countries withmore carbon-intensive production patterns wouldbeover-proportionally affectedbya globalcarbon price. Suchapolicywouldplacethehighestburdenonthemostimportant economicsectors.

Figs.2and3relatepriceincreasesthatwouldresultfromaglobal priceof50USDpertCO2inacertainsectortothatsector’s eco-nomicimportance.InFig.2economicimportanceismeasuredby thepercentagecontributionofaparticularsectortothenational value-added(VAD).Fig.3considersVADrelativetoglobalVADfor eachparticularsector(i.e.aproxyfortheproducingcountry’sshare oftheglobalmarket)9_._The_dashed_lines_represent_global_averages.

Henceregionalsectorsofhighdomestic(Fig.2)orglobal(Fig.3) relevancewithintheglobalsamplearelocatedintherighthalf. Regionalsectorsinthetop-rightquadrantarealsoprojectedtobe subjecttostrongpriceincreases.

Forinstance,forChina’sTextilesector,weprojectarelatively largepriceincreaseofabout5%.Thissectoraccountsforabout3% ofChina’stotalVAD,andabout30%ofglobalvalueaddedinthe Textilesector.Incomparison,forTurkey(forwhichTextilesaccount forabout5%oftotalnationalVAD)thepriceincreasewouldbeonly about2%,whichisbelowtheglobalaveragepriceincreaseforthis sector.Hence,wewouldexpectaglobalcarbonpricetogeneratean increaseinTurkey’sglobalmarketshareinTextiles,butadecline

9_An_additional_{investigation}_of_{competitiveness}_impacts_using_“Revealed

(5)

Fig.3.ProjectedpriceincreasesduetoacarbonpriceofUSD50pertCO2vs.theshareofaregionintheglobalvalueaddedforagivensector.Thedottedlinesshowthe

averagepriceincreaseacrosscountriesandtheaverageVADshare.OnlyregionswithaVADsharelargerthantheworldmeanareplotted.Valuesforpriceincreasesand sectoralsharesinglobalvalueaddedaregiveninTableA1andTableA6intheSI,respectively.

inChina’sshare.Likewise,forMetals,Chinawouldexperiencea priceincreaseofmorethan10%,whereaspricesinmostEuropean countriesaswellasJapanandBrazilwouldincreasebylessthan theglobalaverage.

Overall, the most severely affected economic activities are locatedindevelopingandEasternEuropeancountries.Although, forChina,manysectorsrepresentarelativelysmallVADsharein thenationaleconomy,thecorrespondingprojectedpriceincrease couldleadtosigniﬁcantshiftsintheglobaleconomyinabsolute terms.Duetoitssize,Chinaisthelargestglobalproducerin numer-ouseconomicsectors.TheAsianeconomiesofoursample(withthe exceptionofJapan)wouldbeparticularlynegativelyimpactedbya globalcarbonprice10_._By_contrast,_US_industries_with_a_large_share

inglobalVADtendtobecleanerthanaveragesupplychains.This indicatesthattheintroductionofaglobalcarbonpricewouldboost theseindustries’comparativeadvantageandresultinincreasing marketshares.

3.2. Short-termeffectsonemploymentandGDP

Toassesspotentialshort-termeffectsofaglobalcarbonpriceon sectoralemploymentandnationalGDP,wecalculatethechanges inﬁnaloutputthatwouldoccurasaresultofpricechanges assum-ingthattheindustriesconcerneddonotadjusttheirproduction

10_An_additional_market_{concentration}_analysis_is_done_in_the_SI.

methodsandworkersareunabletoﬁndemploymentelsewhere

(seeSection2.3).Theresultsofthisanalysisshouldberegarded astheupperboundarymeasureoftheeconomicoutputandlabor forceunderpressure,ratherthananaccurateprojectionofGDP andjoblossesresultingfromaglobalcarbonprice.Understanding theextentoftheeffectsonGDPandthelaborforceunder pres-suremightbeofconsiderablepoliticalimportance;eventhough actuallossesinproductionandemploymentcanbeexpectedto reduce,theprospectoflosingafractionofoutputandemployment inagivenindustryislikely tobesufﬁcienttogeneratepolitical resistanceamongworkersandshareholders.

Fig.4analyzesthetotalexpectedshort-termeffectsintodirect and indirecteffects. Directeffectsarethose thatwouldonly be caused by changes in ﬁnal output. Indirect effects result from the additional changes in global supply chains that affect all downstreamsectorsproducingintermediateinputs.Valuesgreater (lower)thanone,denotepotentialgains(losses)inregionalGDP. For instance, a value of 1.01 denotes a potentialgain of 1% of regionalGDP.Overall,directeffectsdominateanddeterminethe directionofthetotaleffectformostregions.Totaleffectsrangefrom GDPunderpressureofabout3%forChinaandIndia,topotential gainsofabout1%forIreland.

(6)

Fig.4.DirectandindirecteffectsofaglobalcarbonpriceofUSD50pertCO2onGDP.Theﬁguredistinguishesbetweendirecteffects(referringtosectoralﬁnaldemand

changesbyconsumers)andindirecteffects(referringtodemandchangesbydownstreamsectors).Thesizeofdotsreferstothetotaleffectinpercentagepoints,whichis obtainedbymultiplyingthetwofactors(directandindirect)andconvertingtheresulttoapercentagechange.TableA1intheSIprovidesvalues.

Fig.5.Projectedpressureonlabormarketsresultingfromaglobalcarbonpriceof50USDpertCO2.Theresultsconsiderthreedifferentlevelsoflaborqualiﬁcation,i.e.

high,mediumandlow.Thebarsrefertothelefty-axis.Impactswhichareindicatedbydotsrefertotherighty-axis.Theresultspresentaceterisparibusimpactandmight thereforeserveasanupperboundoftheimpactonthejobmarket.

and2%ofthetotalworkforcerespectively11_._Moreover,_jobs_are

impactedindifferentwaysdependingonthequaliﬁcationlevel.In thelargemajority(>75%)ofallcountries,low-skilledjobsare pro-jectedtoseethelargestrelativechanges.Thisisprobablybecause theproportionoflowqualiﬁedlaborisgreaterthanaverageinthe sectorsinvestigated,seeTableA8intheSI.InChina,morethan 2%ofthelow-skilledworkforcewouldbeunderpressure.Indiais anexception,asresultsindicatethatthemedium-skillworkforce wouldbemostseverelyaffected.Nevertheless,morethan3%of thelow-skilledworkforcewouldalsobeunderpressureinthis country.

4. Discussionandconclusions

Itisfrequentlyarguedthatunilateralclimatepolicymightlead toadditionalpressureonEITEindustriesandrelatedjobs,

partic-11 _This_is_{approximately}_the_same_proportion_of_the_workforce_that_has_been

affectedintheUSasaresultofChina’sentryintotheWorldTradeOrganization (Acemogluetal.,2016).

ularlyinindustrializedcountries.Thispaperdemonstratesthata globaluniformcarbonpricewould,however,simplyshiftthe prob-lemtodevelopingcountries.Thesenegativelyimpactedcountries andindustriesmightthereforebereluctanttoacceptbinding cli-matemeasuresorrefusetoconstructivelyengageininternational climatenegotiations.

Theresultspresentedinthispaperprovideaninitialbasisforthe developmentofsector-speciﬁcpoliciestolowerthevulnerability ofhighlyrelevantindustriestoaglobalcarbonprice,ortoincrease theircapacity todeal with the associated price increases.This mightbeofspecialimportanceforpoorercountriesconsidering thatmanufacturinghasbeenthe“mainavenueofrapideconomic convergence”, i.e. of high importance for developing countries (Rodrik,2015,p.28).Forcountrieswithcarbon-intensive electric-itysupply,e.g.China,theimpactscouldbesigniﬁcantlyreduced by de-carbonizing electricity production. Other alternativesare

to promote efﬁcient technologies that reduce the demand for

(7)

Furthermore, a rational and successful climatepolicy might need toensure that losers are compensated(Trebilcock, 2014; UNEP,2017).Recyclingcarbontaxrevenuescouldprovidesuch opportunities(Goulder,1995;Klenertetal.,2018).Learningfrom pastexperiencesof(successful)structuraltransformationswithin developedeconomiescanalsohelpgenerateappropriatemeasures toensurea‘justtransition’(Smith,2017)towardsadecarbonized globaleconomy.

Statementofcontribution

HWcontributed theinitialresearchidea.HW developedthe analysisframeworkandanalyzeddatawithinputsbyMJandJCS. HW,MJandJCSwrotethepaper.

Acknowledgement

The authors gratefully acknowledges funding by the

Ger-man Federal Ministry of Education and Research (BMBF) grant

“Klimapolitische Maßnahmen und Transformationspfade zur

BegrenzungderglobalenErwärmungauf1,5◦C(PEP1p5)”, fund-ingcode01LS1610B.The authorsalsowant tothankAlexander Rohlf, Brigitte Knopf, Peter Ward, the participants of the 8th AtlanticWorkshoponEnergyandEnvironmentalEconomics,the participantsof theDIEworkshop onSustainableProductionand Consumptionandanonymousrefereesforvaluablecommentsand suggestions.

AppendixA. Supplementarydata

Supplementarymaterial relatedto thisarticle canbe found, intheonlineversion,atdoi:https://doi.org/10.1016/j.eneco.2019. 104549.

References

Acemoglu,D.,Autor,D.,Dorn,D.,Hanson,G.H.,2016.Importcompetitionandthe greatUSemploymentsagofthe2000s.J.LaborEcon.34,141–198.

Aguiar,A.,Narayanan,B.,McDougall,R.,2016.AnoverviewoftheGTAP9database. J.Glob.Econ.Anal.1,181–208.

Alexeeva-Talebi,V.,Boehringer,C.,Löschel,A.,Voigt,S.,2012.Thevalue-addedof sectoraldisaggregation:implicationsoncompetitiveconsequencesofclimate changepolicies.EnergyEcon.34,127–142.

Böhringer,C.,Balistreri,E.J.,Rutherford,T.F.,2012.Theroleofbordercarbon adjust-mentinunilateralclimatepolicy:overviewofanEnergyModelingForumstudy (EMF29).EnergyEcon.34,97–110.

Branger,F.,Quirion,P.,2014.Wouldbordercarbonadjustmentspreventcarbon leak-ageandheavyindustrycompetitivenesslosses?Insightsfromameta-analysis ofrecenteconomicstudies.Ecol.Econ.99,29–39.

Carbone,J.C.,Rivers,N.,2017.Theimpactsofunilateralclimatepolicyon competi-tiveness:evidencefromcomputablegeneralequilibriummodels.Rev.Environ. Econ.Policy11,24–42.

Chouinard,H.,Perloff,J.M.,2004.Incidenceoffederalandstategasolinetaxes.Econ. Lett.83,55–60,http://dx.doi.org/10.1016/j.econlet.2003.10.004.

Cramton,P.,MacKay,D.J.C.,Ockenfels,A.,Stoft,S.(Eds.),2017.GlobalCarbonPricing. MITPress,Cambridge,MA,USA.

Dechezleprêtre,A.,Sato,M.,2017.Theimpactsofenvironmentalregulationson competitiveness.Rev.Environ.Econ.Policy11,183–206.

Edenhofer,O.,Jakob,M.,Creutzig,F.,Flachsland,C.,Fuss,S.,Kowarsch,M.,Lessmann, K.,Mattauch,L.,Siegmeier,J.,Steckel,J.C.,2015.Closingtheemissionpricegap. Glob.Environ.Change31,132–143,http://dx.doi.org/10.1016/j.gloenvcha.2015. 01.003.

Forin,S.,Radebach,A.,Steckel,J.C.,Ward,H.,2018.Theeffectofindustry delocal-izationonglobalenergyuse:aglobalsectoralperspective.EnergyEcon.70, 233–243.

Fullerton,D.,Muehlegger,E.,2019.Whobearstheeconomiccostsof environmen-talregulations?Rev.Environ.Econ.Policy13,62–82,http://dx.doi.org/10.1093/ reep/rey023.

Gerlagh,R.,Kuik,O.,2014.Spillorleak?Carbonleakagewithinternational technol-ogyspillovers:aCGEanalysis.EnergyEcon.45,381–388.

Goulder,L.H.,1995.Environmentaltaxationandthedoubledividend:areader’s guide.Int.TaxFinance2,157–193.

Head,K.,Mayer,T.,Chapter32014.Gravityequations:workhorse,Toolkit,and cookbook.Handb.Int.Econ.4,131–195, http://dx.doi.org/10.1016/B978-0-444-54314-1.00003-3.

High-LevelCommissiononCarbonPrices,2017.ReportoftheHigh-Level Com-missiononCarbonPrices.WorldBank,WashingtonD.C,URL:https://www. carbonpricingleadership.org/report-of-the-highlevel-commission-on-carbon-prices(accessed2019-06-23).

IEA,[WWWDocument]URLhttps://www.iea.org/statistics/statisticssearch/report/ ?year=2009&country=ESTONIA&product=Coal(accessed2018-04-20)2009.Est. Coal2009.

Jakob,M.,Steckel,J.C.,Edenhofer,O.,2014.Consumption-versusproduction-based emissionpolicies.Annu.Rev.Resour.Econ.6,297–318,http://dx.doi.org/10. 1146/annurev-resource-100913-012342.

Karstensen,J.,Peters,G.P.,2018.Distributionsofcarbonpricingonextraction, com-bustionandconsumptionoffossilfuelsintheglobalsupply-chain.Environ.Res. Lett.13,http://dx.doi.org/10.1088/1748-9326/aa94a3.

Kenkel,D.S., 2005.Arealcohol taxhikes fullypassed throughtoprices? Evi-dencefromAlaska. Am.Econ.Rev.95,273–277,http://dx.doi.org/10.1257/ 000282805774670284.

Kim,K.,Kim,Y.,2012.InternationalcomparisonofindustrialCO2emissiontrends andtheenergyefﬁciencyparadoxutilizingproduction-baseddecomposition. EnergyEcon.34,1724–1741.

Klenert,D.,Mattauch,L.,Combet,E.C.,Edenhofer,O.,Hepburn,C.,Rafaty,R.,Stern, N.,2018.Makingcarbonpricingwork.Nat.Clim.Change8,669–677.

Koesler,S.,Pothen,F.,2013.TheBasicWIODCGEModel:AComputableGeneral EquilibriumModelBasedontheWorldInput-OutputDatabase.ZEWDok.NR, 13-04.

Koesler,S.,Schymura,M.,2015.Substitutionelasticitiesinaconstantelasticityof substitutionframework–empiricalestimatesusingNonlinearLeastSquares. Econ.Syst.Res.27,101–121.

Lenzen,M.,Moran,D.,Kanemoto,K.,Geschke,A.,2013.BuildingEora:amulti-region input-outputdatabaseathighcountryandsectorresolution.Econ.Syst.Res.25, 20–49,http://dx.doi.org/10.1080/09535314.2013.769938.

Marion,J.,Muehlegger,E.,2011.Fueltaxincidenceandsupplyconditions.J.Public Econ.95,1202–1212,http://dx.doi.org/10.1016/j.jpubeco.2011.04.003. Mattoo,A.,Subramanian,A.,vanderMensbrugghe,D.,Hu,J.,2009.Reconciling

climatechangeandtradepolicy.Cent.Glob.Dev.Work.Pap.,189.

Miller,R.E.,Blair,P.D.,2009.Input-OutputAnalysis,secondedition.Cambridge Uni-versityPress,Boston.

Rodrik,D.,2015.Prematuredeindustrialization.J.Econ.Growth21,1–33,http://dx. doi.org/10.1007/s10887-015-9122-3.

Saygin,D.,Worrell,E.,Patel,M.K.,Gielen,D.J.,2011.Benchmarkingtheenergyuse ofenergy-intensiveindustriesinindustrializedandindevelopingcountries. Energy36,6661–6673.

Smith,S.,2017.JustTransition.OECDRep.

Solleder,O.,2013.Tradeeffectsofexporttaxes.Grad.Inst.Int.Dev.Stud.Work.Pap., 08/2013.

Steen-Olsen,K.,Owen,A.,Hertwich,E.G.,Lenzen,M.,2015.Effectsofsector aggre-gationonCO2multipliersinmultiregionalinput–outputanalyses.Econ.Syst. Res.26,284–302.

Timmer,M.,Dietzenbacher,E.,Los,B.,Stehrer,R.,Vries,G.J.,2015.Anillustrated userguidetotheworldinput–outputdatabase:thecaseofglobalautomotive production.Rev.Int.Econ.23,575–605.

Trebilcock,M.J.,2014.DealingWithLosers.OxfordUniversityPress,Oxford.

UNEP, URL:www.unenvironment.org/resources/emissions-gap-report (accessed 2018-04-20)2017.TheEmissionsGapReport2017.

UNFCCC,2015.ParisAgreement.

Voigt,S.,DeCian,E.,Schymura,M.,Verdolini,E.,2014.Energyintensity develop-mentsin40majoreconomies:Structuralchangeortechnologyimprovement? EnergyEcon.41,47–62.

Ward,H.,Radebach,A.,Vierhaus,I.,Fügenschuh,A.,Steckel,J.C.,2017.Reducing globalCO2emissionswiththetechnologieswehave.Resour.EnergyEcon.49, 201–217.