ContentslistsavailableatScienceDirect
Advances in Applied Energy
journalhomepage:www.elsevier.com/locate/adapen
Heterogeneity of consumption-based carbon emissions and driving forces in Indian states
Qi Huang
a, Heran Zheng
b,∗, Jiashuo Li
a,∗, Jing Meng
c, Yunhui Liu
d, Zhenyu Wang
e, Ning Zhang
a, Yuan Li
a,∗, Dabo Guan
c,faInstitute of Blue and Green Development, Shandong University, Weihai, 264209, China
bIndustrial Ecology Programme, Department of Energy and Process Technology, Norwegian University of Science and Technology, Trondheim, 7010, Norway
cThe Bartlett School of Sustainable Construction, University College London, WC1H 0QB, London, UK
dResearch Centre for Contemporary Management, Tsinghua University, Beijing 100084, China
eSchool of Urban and Regional Science, Institute of Finance and Economics Research, Shanghai University of Finance and Economics, Shanghai 200433, China
fDepartment of Earth System Science, Tsinghua University, Beijing 100084, China
a r t i c le i n f o
Keywords:
Carbon flows India
Multi-regional input-output analysis Consumption-based emissions Low carbon transition
a b s t r a ct
Asthesecondmostpopulouscountryintheworld,Indiaisonthewaytorapidindustrializationandurbanization, possiblybecomingthenextcarbongiant.Withitsvastterritoryandhighregionalheterogeneityintermsof developmentstagesandpopulation,state-levelconsumption-basedemissionspatternsanddrivingforcesare criticalbutunfortunately,remainfarfromcompleted.Inthispaper,wefirstappliedamulti-regionalinput- outputmodeltoascertainheterogeneityinconsumption-basedemissionsandtrackcarbonflowsintheinter- statesupplychain,usingournewlyconstructedIndianmulti-stateinput-outputtablefor2015,basedonFlegg locationquotientmethod.Wefoundthathouseholdconsumptiondominatedconsumption-basedemissionsat statelevels,accountingfor60–78%oftotalconsumption-basedemissions,whileinvestment-ledemissionswere relativelyhigherindevelopedregionstheindevelopingregions.Morethan30%ofconsumption-basedemissions indevelopedstateswereimportedfromlessdevelopedstateswithhighercarbonintensity,indicatingalarge spillovereffect.InIndia’slowcarbontransition,policymakersshouldnotonlyfocusonalocalmitigationpolicy indevelopedstates,butoncarbonleakagefromthedevelopingstates,giventhesignificantheterogeneityin industrialdistributionandpopulation.Inter-statecooperationisrecommended,withdevelopedstatessubsidizing themitigationinthedevelopingstates,whichalsoentailsalowermarginalcosttolowcarbontransition.
1. Introduction
India,responsiblefor7%of globalCO2 emissionsin2018, isthe thirdlargestenergyconsumerandCO2emitterintheworld[1].Since 2010,India’semissionsincreasedfrom1750.56Mtto2621.29Mtin 2018,andwereresponsiblefor21.5%oftheglobalemissionincrement overthesameperiod[2].ThissharpincreaseofCO2emissionsinIn- diaismainlydrivenbyeconomicgrowth,andinparticularbytherapid industrializationofsomedevelopedstates[3,4].Suchperformancesug- geststhatIndiawillbethenextcarbongiant.Ontheonehand,ashome to1.3billionpeople,only32%ofIndia’spopulationlivesinurbanareas, whichmeansthatitislikelytoseetheworld’slargestshifttowardsur- banization(withanestimateofa50%urbanizationlevelby2050)[5]. Ontheotherhand,IndiaistheseventhlargestcountryintermsofGDP, havinganaverageannualgrowthrateof7.0%GDP[6]butemits1.9t CO2percapitaayear,eighttimeslowerthanthatoftheUnitedStates
∗Correspondingauthors.
E-mailaddresses:[email protected](H.Zheng),[email protected](J.Li),[email protected](Y.Li).
andfourtimeslowerthanthatofChina(asacomparabledeveloping country)[7].Thepotentialforindustrialization,andthecatch-upeffect inCO2percapita,willseeresource-intensiveandlabor-intensivepro- ductionsprevailinIndia,underthetransitionfromservice-orientedto industry-orientedandrelaxedenvironmentalstandards[8].Therefore, India’semissionswillincreaserapidlyandcontinuouslyoverthenext fewyears,andseethecountrybecomeoneofthemaincontributorsto globalemissions[9].
Tofacilitatealowcarbontransition,IndialauncheditsIntendedNa- tionallyDeterminedContributions(INDC)withapledgeofreducingthe emissionintensityofitsGDPby33–35%fromits2005levelby2030 [10].India hasalsoimplementedaNational ActionPlanon Climate Change(NAPCC),whichidentifieseight“NationalMissions” tofulfill itsINDCcommitments[11].Inaddition,StateActionPlansonClimate Change(SAPCC)forallstatesinIndiasetaseriesoftargetstolowercar- bonintensitythroughregionaleffortstoachievecoherencebetweenthe
https://doi.org/10.1016/j.adapen.2021.100039
Received8March2021;Receivedinrevisedform18May2021;Accepted20May2021 Availableonline25May2021
2666-7924/© 2021TheAuthor(s).PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/)
strategiesandactionsatnationalandsubnationallevel[12].Duetoits vastterritory,diversesocioeconomicbackgroundsandresourceendow- ments,thereishugeheterogeneitybetweenIndianstates.Forexample, lessdevelopedcentralandeasternstates,suchasformer“BIMAROU” states,1includingBihar,MadhyaPradesh,Rajasthan,OdishaandUttar Pradesh,arecharacterizedbyaweakereconomyandlowersocioeco- nomicstatus,butwithsufficientnaturalmineralresourcesandcheap labor[13].Inordertocomprehensivelyformulateregionalmitigation policies,apilotresearchprojectforfeasibilityandefficacyofinterstate emissiontradingshouldbeimplemented,offeringinsightsintothere- gionaldisparitiesdiscussedabove[14,15].
Previousstudieshavealreadyinvestigatedsubnationalproduction- based emissionsand emphasized its policy implications forregional mitigation[16–18].However,progressagainstemissionstargetscould alsobeevaluatedthroughtheuse ofconsumption-basedemissionin- ventoriesthatreallocateemissionsfromstateswhereproductsarepro- duced,tostateswhereproductsareultimatelyconsumed[19,20].Such consumption-basedaccountingofCO2emissionsmaybetterreflectthe responsibilityforpayingthecostsofemissionmitigation[21,22].Re- gardingresearchonconsumption-basedemissionsforIndia,moststud- iesfocusonthenationallevel,resultinginhighlyaggregateddataand absenceofregionalvariations[23,24].ForalargecountrylikeIndia, addressingspatialaggregationissuesandconsideringregionalhetero- geneities, willhelp clarify rootdriversand allocate emissionreduc- tionresponsibility toregions.Previousstudieshavealreadyproposed amulti-regionalinput-outputmodelforregionalemissionstudiesand haveexplainedthepatternof emissionflowsinothercountrieswith vastareasofterritory[25,26].Tothebestofourknowledge,India’s consumption-basedemissionsatastatelevelhavenotbeenquantified, leavingopenthequestionofhowtomitigateemissionsfromaregional andconsumption-basedperspective.
Here,ourstudyfocusesontheheterogeneityinconsumption-based accountingofCO2(fromcombustionoffossilfuels)atstatelevelinIndia andtracksCO2emissionflowswithinIndia,byconstructingastate-level multi-regionalinputoutputmodel(MRIO).DuetothelackofaMRIO table,theFlegglocationquotientmethodandgravitymodelareapplied tocompileanIndianMRIOtablein2015,whichincludes32statesand 50sectors.Thecontributionofthispaperistwofold:first,consumption- basedemissioninventoriesatstatelevelareestimatedforthefirsttime;
andsecond,wetrackCO2emissionflowamongstatesinIndiaandestab- lishthepatternofoutsourcingofCO2emissions.Ourfindingsdemon- stratethathouseholdconsumptioninaffluentstatesdominatetheto- talconsumption-basedemissionsinIndia.Thepatternofoutsourcing ofCO2emissionswithinIndiasuggeststhatanothercontributionthese affluentstatescouldmakeistointroducecleanertechnologiesintoless developedstatesviasomefinancialmechanism,suchasaninterregional cleandevelopmentmechanism.Ourfindingsandsuggestionsarealso madefortheimprovementofregionalenergyefficiencyandmanage- mentofenergyuse,duetotheconnectionbetweenenergyuseandCO2 emission.
Thepaperisorganizedasfollowing:Aftertheintroductionsection, wegiveabriefreviewoftheexistingliterature.Then,webrieflyin- troducetheaccountingframeworkofconsumption-basedemissionsin Indiaandtheunderlyingdata.Intheresultssection,weprovideade- scriptiveoverviewonconsumption-basedemissionsamong23regions (including3unionterritories,2aggregatedregions and18 states)in India,illustratingrankingandcompositionofconsumption-basedemis- sions.We thenrank emissionflows within Indiaandidentify which statesarekeynetemissionexportersandimporters.Aftertheresults section,wediscussthedomesticandinternationalpolicyimplicationof ourresults.Lastly,weconcludethisstudyandstatetheuncertainties andlimitations.
1 Note:RajasthanandOdishanolongerbelongtoBIMAROUafter2018.
1.1. Literaturereview
AstheIndianclimateissuehasbecomeanincreasingconcernfor 1.5°Cglobalmitigationtarget,aconsiderablebodyofresearchhasin- vestigatedthepotentialtrajectoriestoachievinglowcarbondevelop- mentandemissionmitigationpolicies[27].Intheexistingliterature, mostresearchhasfocusedonproduction-basedemissionmanagement.
Somescholarsfocusonspecificindustries,suchastheironandsteelin- dustry[28],theelectricitygenerationindustry[29],thecementindus- try[30,31].Williametal.haveinitialedaforwardprojectionofoutput fromironandcementindustriesandestimatedtheenergyandCO2emis- sionssavings,basedonabottom-upconservationsupplycurvemethod- ology[31].Someresearchersalsofocusonproduction-baseddrivenfac- torsof CO2 emissions.RapideconomicgrowthinIndiahasbeenthe dominatingdrivingforcecontributingtotheincreaseincarbonemis- sions,whileenergyintensityhasbeenthemainfactorreducingcarbon emissions[7,32].Meanwhile,researchondemand-sidemanagement hasgraduallyemerged.Somepapershavequantifiednationalindirect emissionsandcompleteemissions[24,33],andhaveidentifieddynamic orglobaldriversfromaconsumption-basedperspective[34-36].Forex- ample,Wangetal.havefoundthatsince2008,finaldemandhasbeen themajordrivingforcebehindtheincreaseincarbonemissionstrans- fersfromcapitalformationtohouseholdconsumption[34].Fewstudies haveinvestigatedregionalanalysistoillustratethespecificpolicyim- plications.Jemyungetal.havequantifiedregionalcarbonfootprintsof households,andhighlighttheneedtodifferentiateindividualresponsi- bilitiesforclimatemanagement[37].However,previousstudieshave notclarifiedtherootdrivers,takenintoaccountregionalheterogeneities intheindustrychain,whichistheproblemaddressedinthispaper.
2. Methodologyanddatasources 2.1. Accountingscope
Acleardefinitionofthescopeofemissionaccountingcouldavoid double-countinginemissionestimates.TheInternationalCouncilofLo- calEnvironmentalInitiativesinitialed3scopesofemissionboundaries [38]. Scope1includesdirectemissionsfrom fossil fuelscombustion withinterritoryboundaries.Scope2includesemissionsfromconsump- tionof importedelectricityfromupstreampowerplants. Scope3in- cludestheindirectemissionfromupstreamproductions,thatis,emis- sionsembodiedintrade[39].
In this study, carbon emissions were calculated using two ap- proaches:production-basedemissions(Scope1,SectoralApproach)and consumption-basedemissions(Scope3,RegionalInput-outputModel) [40]. Production-based carbon emissionsin this study only refer to energy-relatedemissionsfromthecombustionoffossilfuels,excluding emissionsfromindustryprocessesoragricultureproduction. Inaddi- tion,thisstudyonlyconsideredCO2emissionswithoutothernon-CO2 greenhousegasses(CH4,NO2),becauseoflackofregionalemissionfac- tors.Basedontheproduction-basedemissioninventories,weapplied theMRIOmodeltocalculateconsumption-basedemissions.
2.2. FlegglocationquotientmethodtocompiletheMRIO
Aninput-outputtable (IOT)basedonextensive surveyingby sta- tisticalinstitutionsisconsideredbetterthananon-survey-basedIOta- ble[41,42].Manypractitionersarealsocriticalofapplyingsomenon- surveymethodsforsimulatingIOtables[43,44].Somehybridmethods arefavoured,requiringfewerprimarydatabutretainingsignificantac- curacy.However,hybridapproachesarestillcostlyinsomedeveloping countriesbecauseoftheimperfectionofstatisticalsystems[45].There- fore,giventhelimiteddataresourcesinsub-nationalIOtablesinIndia, weusedtheLocationQuotient(LQ)method,anon-surveymethod,to compiletheMRIOtableforIndia’s32states.
InthestandardLQmethod,regionalinputcoefficientsarespecified asapiecewisefunctionofnationalinputcoefficientsandLQ.Mathemat- ically,itisexpressedas:
𝑎𝑟𝑟𝑖𝑗={ 𝑎𝑛𝑖𝑗, 𝐿𝑄𝑟𝑖𝑗>1
𝑎𝑛𝑖𝑗×𝐿𝑄𝑟𝑖𝑗,𝐿𝑄𝑟𝑖𝑗≤1 (1)
Where 𝑎𝑟𝑟𝑖𝑗 is intraregionalinputcoefficients; 𝑎𝑛𝑖𝑗 is national input coefficients;𝐿𝑄𝑟𝑖𝑗 is theproportionof regionalrequirementsofinput 𝑖purchasedfromwithinregion𝑟.𝑎𝑛𝑖𝑗canbeobtaineddirectlyfromthe nationalIOtable.𝐿𝑄𝑟𝑖𝑗canbederivedfromthestepsinsupportingin- formation(SI).
Further,intraregionalintermediatetransactions,𝑧𝑟𝑟𝑖𝑗,canbeobtained from𝑎𝑟𝑟𝑖𝑗𝑥𝑟𝑗.Inaddition,interregionalintermediatetransactions,𝑧𝑟𝑠𝑖𝑗 (𝑠≠ 𝑟),shouldalsobeestimated.Thetotalamountofinterregionalinterme- diatetransactions( ∑
𝑠,𝑟;𝑠≠𝑟𝑧𝑟𝑠𝑖𝑗)shouldbeconsistentwiththeresidualofthe nationaltotalamountofintermediatetransactionsthathavenotbeen allocatedtointraregionaltransaction.Theresidualcanbeexpressedas:
𝑟𝑒𝑠𝑖𝑑𝑢𝑎𝑙𝑖𝑗=𝑧𝑖𝑗−∑
𝑟 𝑧𝑟𝑟𝑖𝑗 (2)
Obviously,theresidualisnon-negativebecausetheintraregionalin- putcoefficientsareboundedbynationalinputcoefficients.Theresidual thenneedstobedistributedamongtheinterregionaltransactionswith 𝑏𝑟𝑠𝑖𝑗.Here,𝑏𝑟𝑠𝑖𝑗 referstoaproxy(oraninitialestimate)ofinterregional transactionsoraninitialestimate.𝑑𝑟𝑠indicatesthedistancebetween region𝑟andregion𝑠.
𝑏𝑟𝑠𝑖𝑗={
𝑥𝑟𝑖𝑥𝑠𝑗
𝑑𝑟𝑠 ,𝑓𝑜𝑟𝑠≠𝑟
0, 𝑓𝑜𝑟𝑠=𝑟 (3)
Then,consistentestimationsareconstructedwiththehelpofakey parameterof𝑔𝑖𝑗𝑟𝑠,scalingthe𝑏𝑟𝑠𝑖𝑗,sothattheysumtounity:
𝑔𝑟𝑠𝑖𝑗 = 𝑏𝑟𝑠𝑖𝑗
∑𝑟,𝑠𝑏𝑟𝑠𝑖𝑗 (4)
Therefore,theinterregionalintermediatetransactionestimatescan besimplyproportionaltothesizeofthesendingsectorandthereceiving sector.
𝑧𝑟𝑠𝑖𝑗 =𝑔𝑟𝑠𝑖𝑗 ×𝑟𝑒𝑠𝑖𝑑𝑢𝑎𝑙𝑖𝑗 (5)
Initialestimates of finaldemand,imports andexport of each re- gionareshowninsupportinginformation(SI).Theseestimatesbased onsectoralvalueadded,output,samplesurveydatasets,andetc.,do notmeetthe“doublesumconstraints”,inwhichtherowandcolumn totalsmatchtheknownvaluesinnationalIOT[46].Theminimumen- tropyapproachisusedtoadjusttheinitialestimatestoensureagreement withthesummedconstraints.Theprocedureofminimizingthesquared distancestendstopreservethestructureoftheinitialestimatesasmuch aspossiblewithaminimumnumberofnecessarychangestorestorethe rowandcolumnsumstotheknownvalues.
min𝑆= ∑
𝑖,𝑗,𝑠,𝑟 (𝑧𝑠𝑟𝑖𝑗−̄𝑧𝑠𝑟𝑖𝑗)2
𝑤𝑠𝑟𝑖𝑗𝑧𝑠𝑟𝑖𝑗 +∑
𝑖,𝑟 (𝑥𝑟𝑖−̄𝑥𝑟𝑖)2
𝑥𝑟𝑖 +∑
𝑖,𝑟 (𝑣𝑟𝑖−̄𝑣𝑟𝑖)2
𝑣𝑟𝑖 + ∑
𝑖,𝑟,ℎ (𝑦𝑟𝑖ℎ−̄𝑦𝑟𝑖ℎ)2
𝑦𝑟𝑖ℎ
+∑
𝑖,𝑟 (𝑚𝑟𝑖−̄𝑚𝑟𝑖)2
𝑚𝑟𝑖 +∑
𝑖,𝑟 (𝑒𝑟𝑖−̄𝑒𝑟𝑖)2
𝑒𝑟𝑖
(6)
∑S.t.
𝑖,𝑠𝑧𝑠𝑟𝑖𝑗+𝑚𝑟𝑗+𝑣𝑟𝑗=𝑥𝑟𝑖 (7)
∑
𝑗,𝑟𝑧𝑠𝑟𝑖𝑗+∑
ℎ 𝑦𝑟𝑖ℎ+𝑒𝑟𝑖=𝑥𝑟𝑖 (8)
2.3. Consumption-basedaccountingofCO2emissions
Themultiregionalinputoutputmodelistheacceptedtoolemployed toaccountforconsumption-basedemissions,whichillustratestheinter- regionandinter-industryrelationshipsalongsupplychainsandinterre- gionaltrades[47].Specifically,consumption-basedCO2emissionscan
comprehensivelypresentemissionsdirectlyandindirectlyembodiedin thefinalconsumption[48].Inthisstudy,consumption-basedemissions incertainregionsincludesemissionsfromlocalproductionandemis- sionsembodiedindomesticimports,excludingemissionsembodiedin internationalimports.
IntheMRIOmodel,thetotaloutputofeachindustryineachregion, x,stimulatedbyavectoroffinaldemand,y,isgivenby
𝐱=(𝐈−𝐀)−1𝐲 (9)
Where (𝐈−𝐀)−1 is the Leontief inverse (matrix of total require- ments), with𝐿𝑟𝑠𝑖𝑗 indicatingthetotaloutputof industry𝑖in region𝑟 thatisstimulatedbyaunitofdemandof𝑗inregion𝑠;where𝐀isthe directrequirementmatrix,with𝑎𝑟𝑠𝑖𝑗 indicatingtheamountofdirectin- putsfromindustry𝑖inregion𝑟requiredtogenerateoneunitoutputof 𝑗inregion𝑠;where𝐲isthefinaldemandmatrixwith𝑦𝑟𝑠𝑖 indicatingthe finaldemandforgoodsofindustry𝑖inregion𝑠fromregion𝑟;where𝐱is thetotaloutputmatrixwith𝑥𝑟𝑖indicatingthetotaloutputofindustry𝑖 inregionr.Usingfamiliarmatrixnotationanddroppingthesubscripts, wehavethefollowing:
𝐀=
⎡⎢
⎢⎢
⎢⎣
𝐀11 𝐀12 ⋯ 𝐀1𝑛 𝐀21 𝐀22 𝐀2𝑛
⋮ ⋱ ⋮
𝐀𝑛1 𝐀𝑛2 ⋯ 𝐀𝑛𝑛
⎤⎥
⎥⎥
⎥⎦
;𝐲=
⎡⎢
⎢⎢
⎢⎣
𝐲11 𝐲12 ⋯ 𝐲1𝑛 𝐲21 𝐲22 𝐲2𝑛
⋮ ⋱ ⋮
𝐲𝑛1 𝐲𝑛2 ⋯ 𝐲𝑛𝑛
⎤⎥
⎥⎥
⎥⎦
;𝐱=
⎡⎢
⎢⎢
⎢⎣ 𝐱1 𝐱2 𝐱⋮𝒏
⎤⎥
⎥⎥
⎥⎦ .
Tocalculatetheembodiedemissionsinthegoodsandservices,we extendtheMRIOmodelwithenvironmentalextensionbydefiningdirect emissionintensity,whichindicatesCO2emissionsperunitofeconomic outputforallindustriesandregions.Mathematically,itisexpressedas:
𝑘𝑟𝑖=𝑒𝑟𝑖∕𝑥𝑟𝑖,where𝑒𝑟𝑖indicatestheCO2emissionemittedfromindustry𝑖 inregion𝑟.Therefore,totalconsumption-basedemissionsembodiedin goodsandservicesusedforfinaldemandcanbeestimatedby:
𝐶𝑂𝑟2=𝐤(𝐈−𝐀)−1𝐲𝑟 (10)
where𝐤isavectorofemissionintensityforallindustriesinallregions;
𝐲𝑟isthefinaldemand includingfourcategoriesforregion𝑟fromall regions.
Inturn,thetotalembodiedemissionsinexportsfromregion𝑟to region𝑠canbeestimatedby:
𝐄𝐱𝐩𝑟𝑠=𝐤̂𝑟(𝐈−𝐀)−1𝐲𝑠 (11) where𝐤𝑟representsthedirectemissionintensityforregion𝑟butzero forallotherregions;similarly,𝐲𝑠indicatesthefinaldemandincluding fourcategoriesforregion𝑠fromallregions.𝐤̂𝑟isthediagonalmatrixof 𝐤𝑟.
2.4. Datasources
Theproduction-basedcarbonemissionsfromfossilfuels(e.g.coal, petroleumandgas)forregionsinIndiacamefromtheGHGPlatform ofIndia(GHGPI),whichisthemostcomprehensiveGHG(greenhouse gas)emissiondatasetsatastatelevelinIndia(seeTable1andTable2).
Theseemissiondataincluded35regionsand22sectors(SITableS2).
TheMulti-RegionalInput-OutputTable(MRIOT)usedinthispaperwas compiledbasedon2015IndianSupply-UseTable(SUT)atanational level.TheMRIOTcontains32regionsand50economicindustriesfor eachregion.Inaddition,importsfromothercountriesappearedasasin- glerowvectorontheinputsideinMRIOT,andtheintermediatetransac- tionmatrixandfinaluseonlyincludedgoodsandservicesproduceddo- mestically.Hence,wemainlydiscusstheconsumption-basedemissions supportedbydomesticproduction,excludingimportedemissionsfrom othercountries.Tofacilitatediscussion,severalstateswithlowerpopu- lationdensityandGDP(seeSITableS1),wereclusteredintogeographic regions(i.e.HimachalPradesh,JammuandKashmir.Uttarakhandare aggregatedintoNorthregion,andAssam,SikkimandTripuraareaggre- gatedintoNortheastregion).BasicdataforcompilationoftheMRIOT camefromCEICdatasets,AnnualSurveyofIndustry(ASI)datasetsand
Fig. 1. a shows the ranking of state-level consumption-based emissions. b shows the rankingofstate-levelconsumption-basedemis- sionsintensityperGDP.cshowstherankingof state-levelconsumption-basedemissionsinten- sitypercapita.Forallbarchart,lengthofbar indicatesitssizeaccordingtothex-coordinate, andcolorofbarscorrespondstostateGDPper capitafromtheleastdevelopedstatesinyellow tothemostdevelopedstatesinblue.
theMinistryofStatisticsandProgrammeImplementation(MoSPI).De- tailedinformationabouthowwecompiledtheMRIOTcanbefoundin theSupportingInformation(SI).
3. Results
3.1. Theheterogeneousdistributionofconsumption-basedemissionsfor Indiain2015
In2015,Indiaemitted1766MtCO2fromtheburningoffossilfu- els,accountingfor6%ofglobalemissions.83%ofemissionswerecon- sumedby domesticfinaldemand andtheremaining17%, drivenby exporttoothercountries.Fig.1ranksconsumption-basedemissionsin 2015forall23regionsinIndia.Somestateswhereservicesandlight industriesaresubstantiallydeveloped,displayhighconsumption-based emissions.Forexample,thehighestconsumption-basedCO2emissions of195.1MtwasinMaharashtra,(accountingfor13.5%ofthetotal), whichismainlyduetothestatehavingthelargestGDPinIndia(ac- countingfor14%ofnationalGDP).Morethan70%ofitsGDPisat- tributedtoservicesandlightindustries.Similarly,96.7Mtemissionin Karnatakaand128.7Mtemissionin TamilNadualsoareassociated withhighlevelsoftotalGDP.Bycontrast,somestatessuchasMadhya Pradesh,ChhattisgarhandOdisha,havelowerconsumption-basedemis- sions,whichisassociatedwithbeinglesspopulousandhavinglowerper capitaGDP.Theeconomyinthesestatesisheavilydependentoncoal useandenergyintensiveactivitiessuchasenergyindustry,heavyindus- tryormaterialsmanufacturing,withtheresultthatconsumption-based emissionsaresignificantlylowerthanproduction-basedones.However, whenstandardizedbysocioeconomicfactors(GDPandpopulation),the rankingchangesdrastically(seeFig.1).Uttar Pradesh,asthesecond largestconsumption-basedemissionsconsumer, hasamiddlelevelof emissionintensitybyGDPandalowerlevelofemissionintensityper capita.This is becauseUttar Pradeshisthemost populousstateand haslowerGDPpercapita.Inotherwords,theconsumption-basedcar- bonintensity(emissionsperunitofGDP)isgreatestinlessdeveloped statesincentralandeastareas.Bycontrast,thehighGDPandmorede- velopedstatesinthewest,southandunionterritories,tendtobethe leastcarbon intensive.Forexample,Odisha,whose value-addediron andsteelaccountedfor18%ofitsGDP,hadthehighestconsumption- basedemissionintensity,27.04gCO2perRs,whichismorethandouble theconsumption-basedemissionintensityofMaharashtra.Conversely, percapitaconsumption-basedcarbonemissionsindevelopedstatesand unionterritoriessuchasGoa,Delhi,HaryanaandPuducherry,aremore thanthreetimesthatofOdisha,MadhyaPradesh,andJharkhand.Afflu- entstateswithaGDPpercapitaofaround200,000Rsormoresuchas
GoaandDelhi,haveconsumption-basedemissionspercapitaofaround 3tonnes.Incontrast,Jharkhand,aneasternstatewithaGDPpercapita ofjustaround60,000Rs,hasconsumption-basedemissionspercapita ofaround1tonne.
3.2. Theheterogeneousdriversofconsumption-basedemissionsinIndia
Consumption-basedemissionscanbesupportedbydomesticproduc- tionsandimportedgoodsandservices(showninFig.2a),whichexplains wheretheemissionscomefrom.Consumption-basedemissionscanalso be triggeredby differentdrivers,such asfourcategoriesof finalde- mand(showninFig.2b)anddifferentsectoralfinaldemand (shown in Fig.2c),whichdemonstratetheheterogeneityofdriversinIndian states.In23statesorregionsofIndia,withtheexceptionofOdishaand Chhattisgarh,morethan50%ofconsumption-basedemissionswereat- tributedtoimportsfromexternalregions,whichindicatesthatin2015, consumptionreliedheavilyonforeignproductsasopposedtolocalprod- ucts.Theaboveresultsareconsistentwithpreviousstudieswhichstate thatapproximately40–80%ofemissionsinJapanwereattributedtoim- ports[49].Forexample,Delhi(94%),thetopstateintermsofsharesof importedemissions,correspondedwithitsoverwhelmingconsumption- based emissionsthat werethree-fold its production-basedemissions.
ChhattisgarhandOdisha,bycontrast,accountedfortheminimumper- centages ofimportedemissions,reachingalowlevelof 40–50%,be- causetheyaretypicalproduction-basedemissionproducers.Fromthe perspectiveofcategoriesoffinaldemand,householdconsumptionin- duced thelargest shareof consumption-based emission, followedby fixed capital formation.The results are consistentwith previous re- searches onconsumption-basedemissionsinIndiaatthenation-level [33].Morethan60%ofconsumption-basedemissionsineachstatewere triggeredbyhouseholds,especiallyinChhattisgarh(78%)andinOdisha (75%).Emissionsinducedbycapitalformationaccountedforalarger proportionoftotalconsumption-basedemissioninmoreaffluentstates, suchasthetopthreeconsumersofTamilNadu,with33%,Maharash- traandUttarPradesh,with27%,thanlessdevelopedstatesofOdisha andChhattisgarh,with18%,andMadhyaPradesh,with22%.Anun- derlyingexplanationisthatindustrialsectorsinlessdevelopedstates withhighemissionintensity,consumedmorebutinvestedlessinnew high-techgrowthandurbanization.Incontrast,development models inmoredevelopedstates,includingrapideconomicgrowthassociated withrapidurbanization,pushinvestmentintransportationinfrastruc- tureandconstruction,whichaccountedfor70%oftotalcapitalforma- tioninIndia.Fromtheperspectiveoftheconsumption-basedemissions ofdifferentsectors,services,transport,energyindustryandconstruction arethemaincontributors.Fig.2cshowsregionaldisparitiesbetween
Fig. 2.Composition of consumption-based emissionbystates,alldividedintoemissions fromlocalproductionandemission fromdo- mesticimports(a),11sectors(c)andfourfinal demandcategories(b)(Unit:MtCO2).
Table1
Listofabbreviation.
Abbreviation Item
GDP Gross domestic product
INDC Intended Nationally Determined Contributions NAPCC National Action Plan on Climate Change SAPCC State Action Plan on Climate Change
BIMAROU Bihar, Madhya Pradesh, Rajasthan, Odisha and Uttar Pradesh MRIO Multi-regional input output
MRIOT Multi-regional input output table
IO Input output
IOT Input output table CCS Carbon capture and storage CDM Clean development mechanism LQ Location Quotient
FLQ Flegg Location Quotient GHGPI Greenhouse gas Platform of India GHG Greenhouse gas
SUT Supply -Use Table ASI Annual Industry Survey
MoSPI Ministry of Statistics and Programme Implementation CEIC China entrepreneur Investment Club
sectoralconsumption-basedemissions.Energyindustryaccountedfora largeshareofconsumption-basedemissionsinheavy-industryintensive states,whileservicesandequipmentindustryarethemaindriversof consumption-basedemissionsindevelopedstates.Forinstance,energy industryaccountedfor53%and17%oftotalemissionsinChhattisgarh andMaharashtra,respectively,whileservicesandequipmentindustry accountedfor15%and37%oftotalemissionsinthesetwostates.
3.3. Spillovereffectsfrominter-stateemissionflowsacrossIndia
Table3A,showsthelargestgrossemissionflowswithinIndia,re- flectingthepatternsofCO2outsourcing.MaharashtraandTamilNadu
arethemostpopulousandaffluentregionsinIndia,accountingfornine ofthetop10emissioninflows(seveninflowstoMaharashtra,twoin- flowstoTamil).Former“BIMAROU” stateswithlowersocioeconomic status,accountedforsixofthetop10emissionoutflows.Forexample, 40%ofconsumption-basedemissioninMaharashtraareimportedfrom Gujarat(13%),MadhyaPradesh(7%),Chhattisgarh(6%),Odisha(7%) andUttarPradesh(7%).
Table3BandDshowemissionflowsinducedbyfixedcapitalforma- tionandhouseholdconsumption,respectively.Similartogrossemission inflows,Maharashtraisalsothemajordestinationofthelargestemis- sioninflowstriggeredbyhouseholdconsumption.Peoplelivinginlo- cationswiththelargestinflowshavemuchhigherpercapitahousehold consumptionthanpeoplelivinginotherstatesthatexportemissionin- tensivegoodstomoreaffluentregions.Thisindicatesthatthehigher levelsofhouseholdconsumptionindevelopedstatesaresupportedby carbonintensiveproductioninlessdeveloped,heavyindustry-oriented regions.Inkeepingwithrapidgrowth,moredevelopedregions,suchas MaharashtraandTamilNadu,investalargershareofGDP(27–33%)to promoteurbanizationthanlessdevelopedstates(only18–22%)ineast andcentralareasofIndia,resultinginahigherlevelofpercapitafixed capitalformation.However,thelargestemissioninflowsembodiedin capitalformationnormallyoriginatefromlessdevelopedstates,such asOdishaandChhattisgarh,tothosedevelopedstates.Thisismainly causedbythedifferenceinindustrialstructureandendowmentsofnat- uralresources.Forexample,thecentralandeasternregionshavethe highestreservesofmetalandnon-metalmineralsandalowerposition inthesupplychains.
Inaddition,ourresultsshowthatstateswithhigheremissioninten- sityofproductiontendtobenetemissionexporters.Table3C,shows thatthelargestnetinflowsoriginatefromhighintensitystatestolow intensitystates(comparisonsofsectoralintensityareshowninSITa- bleS4).PerGDPemissionintensityconsistsofenergyintensity(energy
Table2
Datasourcesofkeysocioeconomicinformationatastatelevel.
Items (by states) Data sources
Value added of agriculture and services-2015 year CEIC datasets
Output of agriculture-2015 year Ministry of Statistics and Programme Implementation Output/Value added/capital formation/stock change of industries-2015 year Annual Survey of Industry
Population/Area-2015 year Office of Registrar General of India, Ministry of Home Affairs Monthly per capita consumer expenditure-2015 year National Sample Survey Organization
Revenue Expenditure-2015 year Reserve Bank of India
Table3
Thelargestinterstateemissionflowsembodiedintrade,householdandcapitalformation.
A . Emission flows embodied in trade
(Net emission embodied in trade) B . Emission flows embodied in capital formation
(Capital formation per capita)
Origin Destination Origin Destination
24.9 Gu ( − 2.2) Ma (47.9) 6.5 Od (15.5) TN (53.5)
18.9 Ma (47.9) Gu ( − 2.2) 6.4 Gu (37.4) Ma (40.0)
14.4 AP ( − 14.9) Ma (47.9) 5.9 Od (15.5) Ma (40.0)
14.3 AP (( − 14.9) TN (34.9) 5.0 AP (26.1) Ma (40.0)
13.3 UP ( − 16.1) Ma (47.9) 4.7 Ch (24.0) Ma (40.0)
13.0 MP ( − 31.4) Ma (47.9) 4.3 Od (15.5) AP (26.1)
12.9 Od ( − 75.5) Ma (47.9) 4.2 Od (15.5) UP (14.4)
12.4 Ch ( − 71.4) Ma (47.9) 4.2 AP (26.1) Ma (40.0)
12.4 Od ( − 75.5) TN (34.9) 4.2 Ka (41.5) Ma (40.0)
11.4 Ka (18.3) Ma (47.9) 4.2 Ka (41.5) TN (53.5)
C . Net emission flows embodied in trade
(Emission intensity per GDP) D . Emission flows embodied in household
(Household consumption per capita)
Origin Destination Origin Destination
11.0 Od (51.7) Ma (10.4) 16.8 Gu (91.2) Ma (77.4)
10.8 Od (51.7) TN (10.9) 14.3 Ma (77.4) Gu (91.2)
10.4 Ch (68.4) Ma (10.4) 9.3 AP (65.1) Ma (77.4)
7.9 Ch (68.4) TN (10.9) 9.0 UP (28.9) Ma (77.4)
7.4 MP (25.0) Ma (10.4) 8.8 MP (38.6) Ma (77.4)
7.3 Ch (68.4) UP (17.0) 8.5 AP (65.1) TN (84.4)
6.2 Od (51.7) UP (17.0) 7.6 Ma (77.4) Ka (92.5)
6.2 Od (51.7) WB (16.5) 7.3 TN (84.4) Ma (77.4)
6.2 AP (15.6) TN (10.9) 6.9 Ch (38.4) Ma (77.4)
6.1 Ch (68.4) AP (15.6) 6.8 UP (28.9) Gu (91.2)
Note: Abbreviation: Gu-Gujarat, Ma-Maharashtra, AP-Andhra Pradesh, TN-Tamil Nadu, UP-Uttar Pradesh, Ch- Chhattisgarh,MP-MadhyaPradesh,Od-Odisha,Ka-Karnataka.Unit:Emissionflows(MtCO2),Capitalformationper capita(Rs/person),Householdconsumptionpercapita(Rs/person),EmissionintensityperGDP(gCO2/Rs).
consumptionperGDP)andcarbonintensity(CO2emissionperenergy consumption).Inthecaseofenergyintensity,developedstatesorunion territories,suchasMaharashtraandDelhi,entailalowerlevelofenergy intensitythanthoseoflessdevelopedstates,duetoadvancedtechnology forproductionandgreatersharesoflow-energybuthigh-value-added productsinmoredevelopedregions.Inanothercaseofcarboninten- sity,stateslocatedinnorthornortheastIndiatendtoconsumeagreater shareoflow-carbonenergysources(i.e.solarpower,windpowerand nuclearpower,seeSITableS3)thanstateslocatedineastandcentral partsofIndia,butrelyheavilyonheavyindustry.
3.4. TheheterogeneousdriversofemissionflowsacrossIndia
Fig.3shows,atthesectorallevel,theimportedemissionsandex- portedemissions of each state or region. For some service-oriented states,thelargestshareofembodiedemissionsin importscanbe at- tributed totheservicesector.Inaddition,there is arelativelylarge amountofembodiedemissionsintheimportsoflightindustryinthese states(e.g.Maharashtra,TamilNadu,DelhiandKarnataka),followed byconstruction,transportandfood-relatedindustries.Forsomeheavy industry-orientedstates(e.g.Chhattisgarh,OdishaandJharkhand),the largestshareofembodiedemissioninexportscanbeattributedtocon- struction,followedbylightindustry.
Intermsofnetemissionflowsbetweeninflows(emissionsembod- iedinimports)andoutflows(emissionsembodiedinexports)ofeach
region,thenetemissionexportersarehighlycentralized,withthetop fourexportersaccountingfor80%oftotalnetoutflowinIndia,while thenetemissionimportersaremoredecentralized.Thelargestnetemis- sionimportersaredominatedbytheaffluentstatesandhighlyurban- izedunionterritoriessuchasDelhi,MaharashtraandTamilNadu.Some lessdevelopedregions,suchasthoseinthenorth(includingHimachal Pradesh,JammuandKashmir,Uttarakhand)andnortheast(including Assam,SikkimandTripura)alsoimportlargeembodiedemissions,pro- ducedelsewhereinIndia,withrespectivenetimportedemissionsof26.4 and11.0Mt.Theunderlyingexplanationisthatthereisacleaneren- ergymixinthoseregions,wherealargeshareoftheelectricitysupplyis bornebysolarandwindpower.Inaddition,lowerlevelsofindustrializa- tioninthoseregionspromptmoreimportsofhigh-techindustrialgoods fromexternalregions[50].Normalizingnetemissionflowsperunitof capitaandperGDP,furtherhighlightstheimbalanceofemissionflows fromlessdevelopedeasternregionstothedevelopedsouthandwestre- gionsorhighly-urbanizedunionterritories.Forexample,perunitGDP carbon intensityof netemissionexportersis greatestin Chhattisgarh (34.1gCO2perRs),Odisha(25.6gperRs),andJharkhand(10.3gper Rs),duetotheprevalenceofemissionintensiveproducts(i.e.ironand steel,non-metalminerals,electricitymainlysupportedbyfossilfuel)ex- portedfromthesestates.Meanwhile,perunitcapitacarbonintensityof netemissionimportersisgreatestinDelhi(1.94TonneCO2percapita), Chandigarh(1.50Tonnepercapita)andGoa(1.50Tonnepercapita) duetothehigherpercapitaconsumptioninaffluentregions.
Fig.3. Emissionembodiedinimportsandex- portsbystates,alldividedinto11sectors(Unit:
MtCO2).
4. Discussion
Aspotentialgiantemitter,India’sclimateactionstoaddressclimate changearecriticaltothe1.5°Cglobalmitigationtarget[51].Thelargest challengetoseekingacleanerpathwaytoindustrializationorurbaniza- tionin Indiais thatthereis nosuccessful previousmodeltofollow, amongeitherdevelopedcountriesorindustrializedemergingcountries likeChina.MitigationinavastcountrylikeIndiacannotbeachieved withoutconsideringtheregionalheterogeneityinindustriesandsocioe- conomicstages.Likemostcountrieswithvastterritory,regionalemis- siondisparitiesinIndiaalsoshowthatthemostaffluentregionsdomi- natethetotalconsumption-basedemissions.Forexample,consumption- basedCO2emissionsinMaharashtrawiththelargestGDPinIndia,ac- countedfor13.5%ofthenationaltotal.Theoveralleconomicgrowth rateofthestateisexpectedtobe atacompoundannualgrowthrate of7%until2051,andislikelytoremainthehighestcontributor,ac- countingforabout12.8%of nationalGDPby2051[12],whichmay correspondtoitspositionaslargestcontributortototalemissions.High economicgrowthintheaffluentstateswouldundoubtedlyincreasetheir consumption-basedemissions,especiallycarbonflowsembodiedinthe supplychainsfrom otherunder-developedstateswithhigher carbon intensity.Thespillovereffectscouldunderminemitigationefforts,ob- servedinotherdevelopingcountries,andrequirethepolicymakersto shiftthescaleofmitigationmorebroadlywiththeparticipationofmulti- states,especiallyfor thoseinextricablyconnectedin termsof supply chains.
Exploringthegreengrowthwayrequirescooperationtobalanceeco- nomicgrowthandinequalityreduction.Inacarbon-constrainedfuture, developedcountries’effortstobecomecarbonneutral(e.g.capturing carbonandstoring)[52],willsavethecarbonquotaforindustrializa- tionfordevelopingcountries,suchasIndia.Similarly,thecarbonquota shouldbealsoallocatedatthesub-nationallevelbetweenaffluentand developedstatessuchasDelhi,GoaandMaharashtra,andlessdevelop- ingstatessuchasOdishaandTelangana.Therefore,theseaffluentstates shouldalsosetambitioustargetsofmitigation(e.g.carbonneutrality)to provideroomforlessdevelopedstates.Ourstudyfoundthathousehold consumptionwasthelargest contributortoconsumption-based emis- sionsatstatelevel.Morethan60%oftotalemissionsin2015weretrig- geredbyembodiedemissionsin householdsineach state,whichare highlyassociatedwithtransport,foodprocessing andenergy. Alow- carbonlifestyleofwealthierhouseholdsiscritical.Consideringthecom- parativeadvantagesinwealthierareas,plantingtreesorcarboncapture
andstorage(CCS)inwealthierareasisnotareasonableapproachto achievingcarbonneutrality.Connectingpoorregionsandrichregions throughacarboncreditmechanismwillprovideopportunityforacar- bonsink[53].
By contrast, capital formation, as the largest contributor to consumption-basedemissionsinmostdevelopingcountries[54],only accountedfor20–30%oftotalregionalconsumption-basedemissions in India.Thisreversewas mainlycausedbyIndia’slarge population butlowerlevelofurbaninfrastructureexpansion.However,withrapid industrializationsincethefinancialcrisisin2008,Indiaasthesecond largestpopulatedcountry,alsohastheworld’slargestpotentialforur- banization,whichwillmakecapitalformationdominantinIndiaagain.
Wealsofoundthatmoreurbanizedstatesarecharacterizedbyahigher proportionofemissionsembodiedincapitalformation,forexample33%
inTamilNadu,27%inMaharashtrabutonly18%inOdishaandChhat- tisgarh.Intermsofpercentage,inlessdevelopedstates,theurbanpopu- lationissignificantlylowerthanthenationalaverage.Therefore,theses statesareinauniquepositiontochartoutanurbanizationpaththat learnsfromthemistakesorexperiencesofothermoredevelopedstates.
Giventheclimatechangedimension,thesestatesshouldgofurtherby definingaclimate-friendlyurbanizationpath.Forexample,developing anenergyefficientbuildingandtransportinfrastructureintheprimary stagesofurbanization,couldbealow-costwaytosubstantiallyreduce futureenergyusecausedbythepotentialexpansionofurbanization.
Interregionaltrade inIndia,triggeredbycomparativeadvantages andregionalheterogeneity,bringseconomicbenefitsbothfordeveloped states(lowercosts)anddevelopingstates(GDPgrowth).Meanwhile,the highcarbonflowsembodiedininter-statetradeindicatetheconsider- ablespillovereffects.Tobemorespecific,householdconsumptioninde- velopedstatessuchasMaharashtraandTamilNadu,weresupportedby outsourcingemissionstosurroundingregions,whilecapitalformation wasmainlysupportedbyemissionsoccurringinlessdevelopedstates thatareheavilyreliantonheavyindustry.Thepatternofoutsourcing ofCO2showsthetradebetweendevelopedstatesanddevelopingstates tendtobecarbon-intensiveandinvolvelowvalue-addedgoods,suchas metalandnon-metalindustry.Inaddition,wefoundthatthenetemis- sionflowstendedtooriginatefromhighintensitystateswithtolowin- tensitystates.ThemajormechanismforinterregionaltradetodriveCO2 emissionsrestsondifferencesofproduction-basedemissionintensitybe- tweenimportersandexporters[55].Carbon-intensivemanufacturingin lessdevelopedstates,suchasformer“BIMAROU” states,entailsdrasti- callymoreemissionsthanmakingthesameproductsinthemostafflu-
entstates,usingadvancedtechnology;thus,interstatetradeincreases India’sCO2emissions.Wealsofoundthatnetemissionexportersare highlycentralizedandmainlylocatedintheeast.Restructuringthein- dustrialsectoramongIndianstatesgoesagainstexistingcomparative advantage.Therefore,thereisapotentialopportunitytodecarbonize, byimprovingemissionintensityinthemainnetexportersinIndia,in whichmoreenergy-efficienttechnologiescanbeinstalled.Suchaction willupgradetheindustrychainandenhancethevalueaddedofprod- uctsindevelopedstates.Giventhatconsumptionindevelopedstates issupportedbycarbon intensiveproductioninlessdevelopedheavy industry-orientedregions,suchlowmarginal costapproachescanbe supportedbytheaffluentregions’effortstointroduceaccessible,high- technologiesintolessdevelopedstatesinIndia.Forexample,aninterre- gionalcleandevelopmentmechanism(CDM)willincentivisedeveloped netimporterswithcleaner productionprocessestobuycarbon emis- sionpermits[56],byinvestinginlessdevelopednetexporterstouse advancedtechnologiesandupgradedproductions.
5. Conclusion
Inthestudy,weusedtheFlegglocationquotientmethodandgrav- itymodeltoconstructamulti-regionalinput-outputtableforIndiain 2015.Consumption-basedCO2emissionsatstatelevelforIndiain2015 werefirstmeasuredtoestablishregionaldisparitiesandsupplementre- gionalmitigationpolicies.Wefoundthatthedevelopedwesternregion dominatedIndia’scarbonfootprint.Householdconsumptiondominated consumption-basedemissionsinallstates,whileinvestment-ledemis- sionswererelativelyhigherindevelopedregionsthaninthedeveloping regions.Consumption-basedemissionsweresignificantlyredistributed amongIndianstatesviainterstatesupplychains.Wefoundthatstates notonly emitted CO2 emissionswithin territoryboundaries butalso imposedemissionsonotherregionsthroughinterregionaltrade.Espe- cially,carbon-intensiveproductionsineasternregionaresignificantly triggeredbydemandindevelopedwesternregion.
Mitigationpoliciesshouldfocusontheconnectionbetweendevel- opedanddevelopingstates.Alow-carbonlifestyleofwealthierhouse- holdsindevelopedstatesandplantingtreesorcarboncaptureandstor- ageindevelopedstates,viaacarboncreditmechanism(paymentfrom developedstates)willsavethecarbonquotafordevelopingstates.Seek- ingaclimate-friendlyurbanizationpathindevelopedstateswillnotonly reducelocalcarbonemissions,butalsofutureenergyuseandcarbon emissionscausedbythepotentialexpansionofurbanizationindevel- opingstates.Introducingaccessible,high-technologiesfromdeveloped statesviaaninterregionalcleandevelopmentmechanismwillimprove emissionintensityindevelopingstatesandalsobringabouthighvalue- addedindustries.Futureworkcanlinkourmulti-regionalinput-output tablewithglobalinput-outputtable(forexampleGlobalTradeAnalysis Project)totherelationshipbetweenIndianstatesandothercountries.A time-seriesresearch,suchasStructuraldecompositionanalysis,isalso proposedtoexploredynamicdriversindifferentregions.
5.1. Limitationanduncertainty
Thisstudyfacedanumberoflimitationsandassumptionsinthepro- cessofconductingtheresearch.First,weonlyconsideredconsumption- basedemissionsfrom domesticproductionin ourstudy.Thedataof importedgoodsatastatelevelinMRIOwerenotofficiallyavailable duringthestudyperiod.However,India,asa“manufacturer” inglobal supplychains,wascharacterizedbyhigherexportsthanimports.Con- sumptioninIndiawasmainlysupportedbydomesticgoods.Therefore, theuncertaintyaboutapossibledisconnectbetweenIndiaandtherestof theworld,waslimited.Second,ouranalysisfocusedonenergy-related CO2 emissionsexcludingothergreenhousegasses(forexample,CH4, NO2,emissionsfromagricultureproduction,andemissionsfromindus- trialprocesses),suchthatwemayundervaluetheimplicationofagri- cultureonmitigation.Third,theinterregionaltrade estimatedin this
studyisbasedontheFLQmodelandgravitymodelduetodataavailabil- ity,whichinevitablyleadstosomeuncertaintyregardinginterregional trade.
Dataandcodeavailability
Datasets for this research are available in: CEIC (CEIC is not accessible to public but accessible with licensing in the site) https://info.ceicdata.com/en-products-india-premium-database,MoSPI http://www.mospi.gov.in/publication/state-wise-and-item-wise-value- output-agriculture-forestry-and-fishing-2011–12–2017–18, GHGPI http://www.ghgplatform-india.org/economy-wide. India state MRIO datafor2015canbefoundinChinaEmissionAccountsandDatasets https://www.ceads.net/data/input_output_tables/ for free download.
TheMATLABandGAMScodesareavailableinTableS6.
Authorcontributions
H.Z.andJ.M.designedthestudy.Q.H.performedtheanalysisand preparedthemanuscript.Q.H.,H.Z.,J.M.andinterpretedthedata.D.G.
andJ.Lcoordinatedtheproject.Allauthorsparticipatedinwritingthe manuscript.
DeclarationofCompetingInterest
Theauthorsdeclarenocompetingfinancialornonfinancialinterests.
Acknowledgments
J.L.wassupportedbytheShandongUniversityInterdisciplinaryRe- searchandInnovationTeamofYoungScholarsandtheTaishanScholars Program.D.G.wassupportedbytheNationalNaturalScienceFounda- tionof China(41921005and91846301).N.Z. wassupportedbythe NationalKeyR&DProgramofChina(2018YFC0213600),theNational Natural Science Foundation of China 72033005,71822402. H.Z.
was supported by the Norwegian Research Council: 287690/F20.
J.M. was supported by the Natural Environment Research Council (NE/V002414/1).
Supplementarymaterials
Supplementarymaterialassociatedwiththisarticlecanbefound,in theonlineversion,atdoi:10.1016/j.adapen.2021.100039.
References
[1] Wang Q , Zhou YL . Evolution and drivers of production-based carbon emissions in China and India: differences and similarities. J Clean Prod 2020:277 .
[2] IEA Data services homepage. International Energy Agency; 2019 .
[3] Wang Q , Su M , Li RR . Toward to economic growth without emission growth:
the role of urbanization and industrialization in China and India. J Clean Prod 2018;205:499–511 .
[4] Sridhar KS . Urbanization and Carbon Emissions in India and China. Environ Urban Asia 2018;9:113–26 .
[5] United Nations Department of Economic and Social Affairs PD World urbanization prospects: the 2014 revision; 2014. Highlights .
[6] IMF World economic outlook database. International Monetary Fund; 2019 . [7] Ortega-Ruiz G , Mena-Nieto A , Garcia-Ramos JE . Is India on the right pathway to
reduce CO2 emissions? Decomposing an enlarged Kaya identity using the LMDI method for the period 1990-2016. Sci Total Environ 2020:737 .
[8] Pappas D , Chalvatzis KJ , Guan DB , Li X . Industrial Relocation and CO2 Emission Intensity: focus on the Potential Cross-Country Shift from China to India and SE Asia. Enrgy Proced 2017;142:2898–904 .
[9] Parikh KS , Parikh JK , Ghosh PP . Can India grow and live within a 1.5 degree CO2 emissions budget? Energ Policy 2018;120:24–37 .
[10] Ministry of Environment and Forests GoI. India’s Initial National Communication to the United Nations Framework Convention on Climate Change. 2012.
[11] Chandel SS , Shrivastva R , Sharma V , Ramasamy P . Overview of the initiatives in renewable energy sector under the national action plan on climate change in India.
Renew Sust Energ Rev 2016;54:866–73 .
[12] TERI Assessing climate change vulnerability and adaptation strategies for maharash- tra: maharashtra state adaptation action plan on climate change (MSAAPC). New Delhi: The Energy and Resources Institute; 2014. p.302
