a global stocktake of the paris pledges implications for energy systems and economy

TheJRC-POLESmodelisaglobalpartialequilibriumsimulation modeloftheenergysector,covering38regionsworld-wideplus theEU.Themodelcovers15fuelsupplybranches,30technologies inpowerproduction,6i

A global stocktake of the Paris pledges: Implications for energy systems

a

European Commission, Joint Research Centre (JRC), Seville, Spain

b

E3MLab, National Technical University of Athens, Greece

A R T I C L E I N F O

Article history:

Received 18 April 2016

Received in revised form 25 July 2016

Accepted 26 August 2016

Available online 13 September 2016

JEL classification:

C60

Q40

Q50

Keywords:

International climate negotiations

Mitigation policy

Paris agreement

INDC

Modelling

Global stock taking

A B S T R A C T

(http://creativecommons.org/licenses/by/4.0/)

1.Introduction

The twenty-first edition of the annual United Nations-led

conferenceonclimatechange(ConferenceoftheParties,COP21)

washeldinParisinDecember 2015.TheParisAgreementisan

importantstepforwardininternationalclimatechange

negotia-tions Itsmain merits include a legally binding2C target,the

introductionofa five-yearlyreviewprocessfrom2018onwards

withafirstglobalstocktakescheduledfor2023andanagreement

oninternationalclimatefinancing.Comparedtopreviouseditions

suchasCOP3inKyotoandCOP15inCopenhagen,thebottom-up

approach to climate change mitigation (introduced in Durban,

COP17in2011)wasafundamentalshiftinthenatureofthepolicy

process.Intherun-uptoCOP21,mostcountriessubmittedclimate

actionpledgeslabelled‘IntendedNationallyDetermined

Contri-butions'(INDCs).Thegreenhousegasemissionsofthecountries

emissionsin2010(UNFCCC,2016).Hence,incontrasttotheKyoto

protocol, the Paris pledges have a broad coverage in terms of

sufficient conditiontoavoid globalwarmingof more than 2C above pre-industrial levels by theend of the century, a target includedintheCopenhagenAccord(COP15)in2009 andinthe CancunAgreement(COP16)in2010.Pre-COPanalysesindicatethat theINDCsimplyanincreaseinglobaltemperaturesintherangeof 2.6–3.1C by2100 (Fawcett et al.,2015; Gütschowetal., 2015; Rogeljetal.,2016).Anotheroutstandingchallengeisthevoluntary natureofindividualcountries’emissionreductions.Onceratified, the Paris Agreement will be legally binding, but the INDCs of

reduction,itdoesnotincludeanyexplicitreferencetotheaviation andshippingsector

The outcomes of previous rounds of international climate changenegotiationshave beenassessedby variousstudies.For

Protocoldoesnotimplyacost-effectiveclimatechangemitigation policy and highlight the cost-reducing potential of emission

combinationofpermittradeandthepresenceof‘hotair'(dueto emissiontargetswellabovetheprojectedbusinessasusual)may

* Corresponding author.

E-mail addresses: toon.vandyck@ec.europa.eu , vandycktoon@hotmail.com

(T Vandyck).

http://dx.doi.org/10.1016/j.gloenvcha.2016.08.006

0959-3780/ã 2016 The Authors Published by Elsevier Ltd This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ).

Global Environmental Change 41 (2016) 46–63

ContentslistsavailableatScienceDirect

j o u r n a lh o m e p ag e :w w w e l s e vi e r c o m / l o c a t e / g l o en v c h a

strongly reduce the environmental effectiveness of the Kyoto

Protocol.Theanalyses ofthepledgesoftheCopenhagenAccord

basedonintegratedassessmentmodels(denElzenetal.,2011a,b;

vanVlietetal.,2012;Riahietal.,2015)andcomputablegeneral

equilibrium(CGE) models(Dellink etal.,2011; McKibbin etal.,

2011;Petersonetal.,2011;Saveynetal.,2011;Tianyuetal.,2016)

typicallyfindapolicycostbetween0and3%ofGDPcomparedtoa

baselinein2020fordifferentcostmetrics(abatementcost,GDP,

welfare) Pre-COP21assessments of theINDCs canbe foundin

Fawcettetal.(2015)andIEA(2015)

Thispaperassessestheenergy-relatedandeconomic

implica-tionsoftheclimatemitigationpoliciesembeddedintheINDCs.The

maincontributiontotheliteratureistwofold.First,wepresenta

timely, policy-relevant, global stocktake of the Paris mitigation

pledges that translates the outcome of thelatest international

climate negotiations into quantifiable changes in a range of

variablesincludingenergydemand,thecompositionofenergyand

electricityproduction,economicactivity,tradeandemployment

Thesecond contributionliesin the methodologicalframework,

presentedinthefollowingsection.Thecombinationofa

bottom-up,detailedenergysystemmodelandatop-downglobaleconomic

modelexploitsthecomplementaritiesbetweenbothandenables

anextensivestudyofclimatechangemitigationpolicies

The remainder of the paper is organised as follows After

presentingthemethodology, wedescribethescenariosstudied:

theReferencescenario,theINDCscenariocoveringthemitigation

componentoftheParispledgesandascenariothatislikelytoput

theworldontracktomeetthe2Ctarget.Resultsarepresentedin

Section4.Wehighlighttheimpactonenergyproduction,demand

presenthowthegapbetweentheINDCsandthe2Cpathwaycan

bebridged.Thefinalsectionconcludes

Theassessmentofclimatechangemitigationpoliciespresented

inthispaperbuildsonthecombinedmodellingeffortofadetailed,

europa.eu/jrc/en/poles)and aneconomy-wideComputable

Gen-eralEquilibrium(CGE)model(JRC-GEM-E3,https://ec.europa.eu/

jrc/en/gem-e3/) The models are harmonized along a common

Referencescenarioandaresoft-linkedtoexploit

complementar-itiesofa detailedrepresentation ofenergyproduction, demand

mechanisms including international trade, intermediate input

linksbetweenindustries,andrecyclingoftaxationrevenueonthe

otherhand.Assuch,thispaperaddressespartofthecritiqueon

standard modellingpractices put forwardbyRosen (2016) and

Rosen and Guenther (2016), particularly onthehigh degreeof

aggregationinmostintegratedassessmentmodels.Incontrastto

exercisesusingnumerousmodelsinordertoprovidearangeof

resultsforacommonsetofoutputvariables(Kriegleretal.,2013,

2015;Riahietal.,2015),thispaperemphasizesthatdifferentmodel

typescancontributecomplementarypartstoacomplexpuzzle

ThescenariosanalysedherebuildontheanalysesbyLabatetal

(2015), Kitous and Keramidas (2015) and Kitous et al (2016),

adopted by Russ et al (2009) and Saveyn et al (2011) The

bottom-uprepresentationofthepowersectorcontributestobutis

distinctfromtheliteraturereconcilingtop-downandbottom-up

informationwhilebuildingahighdegreeofenergysystemdetail

intoaCGEmodel(e.g.McFarlandetal.,2004;Hourcadeetal.,2006;

Sue Wing, 2008; Böhringer and Rutherford, 2008; Abrell and

Rausch,2016;LiandZhang,2016).Thefollowingparagraphsbriefly

describetheJRC-POLESmodel,theJRC-GEM-E3modelandtheway

inwhichthetwomodelsarecombined.Formoredetailedmodel descriptionswerefertoAppendicesAandB,theabove-mentioned modelwebsitesandthemathematicaldescriptionofJRC-GEM-E3

inCaprosetal.(2013) TheJRC-POLESmodelisaglobalpartialequilibriumsimulation modeloftheenergysector,covering38regionsworld-wideplus theEU.Themodelcovers15fuelsupplybranches,30technologies

inpowerproduction,6intransformation,15finaldemandsectors andcorrespondinggreenhousegasemissions.GDPisanexogenous inputintothemodel,whileendogenousresourceprices, endoge-nous global technological progress in electricity generation

mitigation policies discussed in the next section and listed in

AppendixCareimplementedbyintroducingcarbonpricesupto thelevelwhereemissionreductiontargetsaremet.Carbonprices affect the average energy prices, inducing energy efficiency responsesonthedemandside,andtherelativepricesofdifferent fuelsandtechnologies,leadingtoadjustmentsonboththedemand side (e.g fuel switch)and the supply side (e.g investments in renewables)

model.Themodeldescribestheeconomicbehaviourof

(exogenous) government policies,different typesof energy use

Inter-industryconnectionsareexplicitlyrepresentedvia interme-diateconsumption.Climatepoliciesareintroducedinthemodel

endoge-nouslyderivestheshadowpricestomeettheseconstraints,raising thecostofemission-intensiveinputsforfirmsandconsumptionof emission-intensive goods for households Emission reductions occurviathreemechanisms:areductioninoutputand consump-tion,substitutiontowardslow-carboninputsandgoodsand end-of-pipeabatementtechnologies

combinationofthetwomodelsinawaythatallowsforabroad assessmentwhilepreservingthedetailsandparticularstrengthsof each First,a Reference sharedbythe twomodelsis developed basedoncommonassumptionsforthe(exogenous)evolutionof two important factorswith regardsto climatechange: region-specificeconomic(GDP)andpopulationgrowth.Theevolutionof the sector composition of economic activity follows the same

developing countries based on historical data In addition, the

scenarioresultsofthedisaggregatedenergymodelfeedintothe economy-wideCGEmodeltomakeuseofthein-depthtreatment

of the energy system in JRC-POLES In particular, the totals of greenhouse gasemissionsderived fromthebottom-upanalysis determineregionalemissionconstraintsfortheeconomic assess-ment withJRC-GEM-E3.In addition,thesharesof thedifferent technologiesinelectricitygenerationinJRC-POLESareusedasan inputintheJRC-GEM-E3analyses.Thissoft-linkisenabledbythe splitofelectricitygenerationinto10technologiesinthe JRC-GEM-E3model.Asaresult,thetechnologymixinelectricitysupplyin theJRC-GEM-E3modelisconsistentwithanenhanced represen-tation of the specific features that characterize real-world

technology, capacity investment decisions, intermittency, re-gion-specificpotentialsofrenewableenergysources(per technol-ogy)andendogenoustechnologicalprogress.Changesinelectricity

technologies (e.g.solar panels) arenot considered explicitly in

includechangesincoal,oil,gasandelectricityvolumesandprices

(whichareendogenousinbothmodels).Futureworkcanfurther

exploretheseoptionsfortheintegrationofmodels

3.Scenarios

This section describes the three scenarios analysed in this

paper:theReference,theINDCscenariorepresentingthe

mitiga-tioncomponent of theParis pledgesand the2C scenario All

scenarioshaveidenticalassumptions onpopulationgrowth For

theEU,populationforecastsaretakenfromEuropeanCommission

(2013).Forallotherregions,populationprojectionsofUN(2015)

Reference,theINDCscenarioand the2Cscenario,respectively,

greenhousegas emissionsand emissionintensitiesof GDP.The

trajectoryoftotalgreenhousegasemissionsineachofthethree

scenariosisdepictedinFig.1.Adetaileddescriptionofthepolicies

includedintheReference,theINDCscenarioandthe2Cscenario

canbefoundinAppendixCandintheonlineAppendix

TheReferenceservesasabenchmarkforcomparisonandbuilds

on various data sources and assumptions First, the Reference

includestheclimatepoliciesthatarecurrentlyimplementedor

announced,particularlyfor2020,withoutaddingnewadditional

policies(takingintoaccounttheinformationprovidedindenElzen

etal.,2015).Inmodellingterms,theexistingorannouncedcarbon

policiesarerepresentedbyacorrespondingcarbonprice.Carbon

valuesintheReferencearelow(EU)orzero(restoftheworld)in

2015.Furthermore,carbonvaluesrangebetween0and39US $

(2015)pertonneofCO2eintheyear2030.Second,growthofGross

DomesticProduct(GDP)intheReferenceisexogenousandbased

onforecastsbytheOECDEconomicOutlook(2013)andtheWorld

Bank (2014) Sector-specific growth pathsin the Reference are

considertheimpactsofchangingclimaticconditionsoneconomic

growth, as described in Fankhauser and Tol (2005) Third,the

growing scarcity of conventional oil resources and consequent

increasingmarketpowerofOPECdrivetheoilpriceupwardsover

time(endogenousinJRC-POLES).Theoilpricesinthemodelreflect

thelowlevelsobservedrecentlyandareprojectedtoreacharound

100US$2005in2030.Fourth,asaresultoftheabove-mentioned

assumptionsandpolicies,theglobalaverageenergyintensityof

GDPfollowsadownwardtrend,atarateobservedintheperiod

1995–2008,butslightlyfasterthantheaveragerateobservedover thepast25years( 1.4%peryear1990–2015, 1.7%peryear2015– 2030).Inadditiontotheimplementedpolicies,thisdecouplingis drivenbythepotential forenergy efficiency (especiallyin fast-growinglow-incomecountries)andtheincreasingtechnological maturityoflow-carbontechnologies.Fifth,themaindatasources

2014)andFAO-Stat(FAO,2014).A moredetailed descriptionof

AppendixA.Thelevelofglobalgreenhousegasemissionsinthe

considered,asillustratedinFig.1.For non-CO2 GHGs,marginal abatementcostcurvesarebasedonEMF21(Weyantetal.,2006),

USEPA(2013)and GLOBIOMforland use,land-usechangeand forestry(LULUCF)andagriculture(IIASA,2015a)

The INDCscenario represents theclimatechangemitigation pledgesmadebyindividualcountriesintherun-uptotheCOP21in

ambitions in the conditional INDCs, i.e including mitigation

provision of climatefinancing The financial transfers resulting fromtheGreenClimateFund(thefinancingmechanismunderthe UNFCCC)arenotpartoftheanalysishere,aslittleisknownabout the allocation of thefund at this point In the case where the mitigationpledgeswerealreadyreachedintheReferencescenario (asaresultofmarketforcesand technologicaldeployment),no additionaleffortwas required.The availableinformation in the INDCsistranslatedintoemissiontargets,whichareimplemented

inthemodelbyregion-specificeconomy-widecarbonprices.More detailontheincludedpoliciesisgiveninAppendixCandinthe online Appendix Implicitly and due to lack of more detailed information this assumes that policies are efficient within a region'sborders.Widelydifferingcarbonprices,rangingfrom0to

119 US $ (2015) in 2030, indicate that there is potential for

aggregate of GHG emissions stabilises around the level in the year2025(Fig.1).GHGintensityoftheeconomydecreasesatan

emissionsin2030aremorethan13%lowerthanintheReferencein

2030.Themainfocusoftheresultspresentedinthispaperlieson theyear2030,asmostoftheINDCsdonotextendbeyondthistime frame,Someoftheresults,however,consideratimehorizonupto theyear2050.Fortheseresults,weassumeacontinuedclimate changemitigationeffortsinallregionsafter2030.Inparticular,we assumethatpoliciesareintroducedsuchthattheyearlyrateof reductionofGHGintensity(GHGexcludingsinksperGDP;Sinks are defined as negative CO2 emissions from land-use related

managementcouldrepresent3GtCO2in2010andabout2GtCO2

in2050intheReference.However,duetosignificantuncertainty

on the historical estimates of sinks, they are generally not considered in the result section.) implied by the INDCs in the

2020–2030periodiscontinuedintheperiod2030–2050(global averagereductionrateof3.2%peryear)

The2Cscenarioconsidersapathwayofglobalgreenhousegas emissions that is likely to be consistent with limiting global temperatureincreaseto2Cbytheendofthecenturycomparedto levelsintheperiod1850–1900.Withatotalcarbonbudgetof1160

GtCO2over2011–2050andareductionofKyotogasesof72%in

scenario430–480ppmwithovershoot>0.4W/m2inIPCC(2014, AR5WGIIITable6.3)witha22–37%probabilityofexceedingthe

2Cwarmingtarget.AsillustratedinFig.1,GHGemissionsupto

Fig 1 Global greenhouse gas emissions in the Reference, the INDC and the 2  C

scenario Shaded areas represent the (median, 80th and 20th percentile per

temperature range of) scenarios included in the IPCC AR5 WGIII Scenario Database

( IIASA, 2015b ) Temperature ranges are based on IPCC (2014) with at least 60%

probability for the scenarios below 2  C, and 55% probability for staying between

intheIPCCScenarioDatabasewithaprobabilityofstayingbelow

2Cofatleast60%,andfallbelowthe20thpercentileinthe2040–

2050period.Apeak inworldaggregateGHGemissionsappears

aroundtheyear2020(Fig.1).Thespecificationofthe2Cscenario

considersconvergenceofcarbonprices,henceimplicitlyassumes

enhancedeconomicefficiencyformitigationeffortsandenhanced

technologydiffusionduetointernationalcollaborationovertime

Formiddle-andhigh-incomeregions,carbonpricesconvergeto

around53US$(2015)in2030,whichcorrespondswiththehighest

levelofcarbonvaluesintheINDCscenario(excludingRepublicof

emissionsare reduced in thecountriesand sectors where it is

cheapesttodoso.However,the2Cscenariostudiedinthispaper

allows for a two-track climate policy, acknowledging political

responsibilities” as included in the United Nations Framework

ConventiononClimateChange,negotiatedattheRioEarthSummit

in1992.Inparticular,carbonvaluesoflow-incomecountries(with

incomepercapitain2030lowerthan10000US$PPP,including

India,IndonesiaandanumberofcountriesinSub-SaharaAfrican,

CentralAmerica, South-EastAsia andthePacific) convergetoa

levelofaround26US$(2015)in2030,whichisapproximatelyhalf

ofthecarbonvalueinhigh-incomeregionsandbringsglobalGHG

emissionsonthepathwaydescribedaboveandillustratedinFig.1

Amoreelaborateassessmentofpotentialburdensharing

agree-mentsandtheunderlyingethicalprinciplesisoutsidethescopeof

thispaper(seeBabonneauetal.,2016,Marcuccietal.,2016and

Roseetal.,2016,foradiscussionontheequitydimensioninthe

contextoftheParisAgreement).Importantly,allregionscontribute

tothereductionsinGHGemissionsandtheintensitiesofclimate

actions – and, correspondingly, the carbon prices – gradually

increaseovertime.Forallcountries,wetaketheeffortintheINDC scenarioasalowerboundforthe2Cscenario.Therefore,the2C scenarioassumesacooperativesettingwithglobalparticipationin which free-riding is not considered Total GHG emissions are around27%lowerthanintheReferencein2030.Accordingly,GHG intensityoftheeconomydecreasesatmorethandoubletherateof thepast25years( 3.9%peryearovertheperiod2015–2030)

Table1summarizesthemainassumptionsbehindtheanalysis The lasttwo columnspresentthe inputsfor theINDCand 2C scenarios.ThepercentagechangesofGHGemissionsfrom2005to

2030intheINDCscenarioarebasedontheINDCssubmittedby individualcountries.Thelastcolumnindicateswhetheraregion wasincluded(basedonGDPpercapita)inthegroupofcountries forwhichcarbonpricesareassumedtoconvergetohighorlow levelsof53and26US$(2015)in2030.TheRestofCentraland SouthAmericaisaregionthataggregatescountriesofbothgroups (withChileinthehigh-incomegroup),hencetheoverallcarbon pricewillliebetweenthehighandthelowvalues

4.Results Thissectionpresentstheresultsofthenumericalsimulations

discussestheimpactof theclimatechangemitigationscenarios

onthecompositionofenergydemand.Next,wezoominonthe greenhousegasemissionpathsbygastypeandbyemittingsector

Wepayparticularattentiontotheelectricity productionsector Thesecondpartpresentstheeconomy-wideresults,highlighting thedifferentiationofimpactsacrossregionsandsectors

Animportantcaveatforallresultspresentedhereisthatthe scenariosdonotconsiderthe(avoided)damagesfrom(mitigating) climatechange(Rosen,2016).Forstudiesontheimpactofclimate

Table 1

The main characteristics of the Reference, the INDC scenario and the 2C scenario.

GHG a

Yearly GDP growth rate

GHG/GDP b

a

Greenhouse gas emissions are expressed in Gt CO2 e and exclude emissions from LULUCF and bunkers.

b

GHG/GDP is expressed in t CO2 e /US$(2005) PPP.

c ‘High', ‘Intermediate' and ‘Low' carbon values converge to 53, 45 and 26 US $ (2015) in 2030 respectively Carbon values of Republic of Korea and New Zealand are higher



change,werefertoOECD(2015)foraglobalassessmentand to

Ciscaretal.(2014)andHouseretal.(2014)forstudiesonthelevel

oftheEuropeanUnionandtheUnitedStates,respectively

4.1.Energydemand

Fuelcombustionisoneofthemainsourcesofgreenhousegas

emissions.Hence,policiesthatenvisagerestrictingemissionswill

haveanimpactontheaggregatelevelandcompositionofenergy

consumption Carbon pricing raises the price of energy, which

leadstoadecreaseoftotalenergydemandby3.8%(9.2%)and8.6%

(33.6%)intheINDCand2Cscenariosrespectivelyin2030(2050)

comparedtotheReference.Thisresultindicatestheimportanceof

energyefficiencyasacontributortoemissionreductions.Table2

decomposesthechangeinaggregateenergydemandbyfueltype

andillustratesthesubstitutionbetweenprimaryenergysources

ThelatterisdrivenbycarbonpricingbasedonaCO2equivalent

basis, which affects relative prices of different fuel types and

incentivizessubstitutiontowardslow-carbonenergysources

TheINDCshaveanegligibleimpactonglobaloilandnaturalgas

consumption.Thedemandforsolidfuels –coalandlignite–is

Hence,replacing solid fuels bynon-fossil fuels is animportant

element for climate change mitigation policies In contrast to

increasingvolumesofglobalcoalconsumptionintheReference

(comparedto2010,a41%increasein2030,73%in2050),thelevels

remainroughly constant in the2020–2030period in the INDC

scenario.Theseresultsareconsistentwiththefindingspresented

byIEA(2015)

Table 2 furthermoreindicatesthat the2C scenario implies

substantialreductionsinworlddemandforoilandgasfrom2025

onwards.GoingfromtheINDCstoapathwaythatislikelytolimit

globalwarmingto2C impliesanincreasedrate ofdecreaseof

solid fuel consumption, despite allowing for the possibility of

CarbonCaptureandStorage(CCS).ThecontributionofCCSwillbe

discussedin moredetail in Section 4.4 The impacts shown in

Table2areinlinewiththeresultspresentedbyBaueretal.(2015),

temperature increase of 2C with several models and with a

focusonfossilfuelmarkets

4.2.Emissionreductionsbygreenhousegas

Carbondioxide(CO2)istheprimaryanthropogenicgreenhouse

gas,coveringaroundthreequartersofglobalGHGemissions(in

CO2equivalentterms,IPCC,2014).However,theresultsillustrated

inFig 2 showthat both theINDC and the2C scenarioimply

implementcarbonpricesthatareuniform(onaCO2-equivalent

basis)acrossthedifferenttypesofgases.Hence,cost-minimising producerswilldeterminetherelative contributionsof different gases totheoverall emission reduction in anefficient manner, using least-cost options before more expensive alternatives In particular,theunderlyingsector-andregion-specifictechnology options(forCO2)andmarginalabatementcostcurves(fornon-CO2

emissionsandCO2emissionsinagriculture)leadtodifferenttime

considered

The INDC scenarioleads tostrongreductions inhydro

reveals thefactthat theemissionsof thesegasesarerelatively

(European Commission, 2012) The reduction of nitrous oxide (N2O)emissionsisoneofthemorecostlyoptions:acost-effective

approximately8%lowerthanthelevelsintheReferencein2030 The emission reduction profiles in the 2C scenario show strongerreductionsfor allgases.Interestingly,theemissionsof HFCsarereducedatafasterratethanintheINDCscenarioupto

2030,butconvergetowards2050.Thisresultindicatesthatthe INDCsexploitnearlythefullpotentialofHFCemissionreductions Furthermore,Fig.2illustratesawidegapbetweenthereductions

ofCO2inbothscenarios:theINDCsleadtoalevelofCO2emissions thatisapproximately30%lowerthanthelevelintheReferencein

2050,whilethe2CpathwaystudiedheresuggestsalevelofCO2

emissionsaroundonethird ofthelevelin theINDC scenarioin 2050

4.3.Emissionreductionsbysector Theprevious sectiondecomposed theaggregateGHG reduc-tionsintogas-specificabatementprofilesovertime.Asecondway

todisentangletheemissionreductionsisonasector-specificbasis

Fig.3presentsemissionsreductionsin2030disaggregatedintosix categories: electricity generation, the energy sectors, industry, landuse,land-usechangeandforestry(LULUCF),agricultureand

anaggregatecategoryforbuildings,transportandwaste.Anumber

ofinsightscanbedeductedfromtheJRC-POLESmodelsimulations First, the power sector emergesas the main contributor to emissionreductionsinbothINDCand2Cscenarios.A transfor-mationoftheelectricityproductionsectorcoversmorethanathird

oftheemissionreductionsbetweentheReferenceandtheINDCin

2030.Inaddition,thepowergenerationsectorbridgesaround31%

of the gapbetweenthe INDCs and the2C scenario.The next

electricitysectorisachieved

Table 2

Changes in primary energy demand (total and by fuel type) in the INDC and the 2  C

scenarios, expressed as% change from the Reference Non-fossil fuels include

renewables and electricity generated by nuclear power plants.

Fig 2 Emission reduction by type of greenhouse gas in the INDC and the 2  C scenario CO 2 emissions included LULUCF but exclude sinks Greenhouse gases shown are carbon dioxide (CO 2 ), methane (CH 4 ), nitrous oxide (N 2 O), hydrofluor-ocarbons (HFCs) and other fluorinate gases (F-gases).

shiftawayfromemission-intensivefossilfuels (inlinewiththe

previoussection)isthemaindriverofemissionreductionsinthe

energy sector In the numerical simulations presented here, a

carbonpriceonaCO2-equivalentbasisprovidestheincentivesfor

thischange.GreenhousegasesotherthanCO2representmorethan

halfoftheemissionreductionsinthiscategory,whichistoalarge

extentduetoreductioninmethaneemissionfromtheproduction

offossilfuels

Third,decreasinggreenhousegasemissionsintheindustrial

sectorisanon-negligiblepossibility,representing17%oftheGHG

process CO2 emissions in the steel, non-metallic minerals and

chemicalsectors,andothergreenhousegasemissions(N2O,HFCs,

PFCsandSF6)inindustrialsectorssuchasthealuminiumsector.In

linewiththeprevioussection,theabatementpotentialofnon-CO2

greenhousegasesistoalargeextentusedintheINDCscenario,

while further emission reductions to reach the 2C pathway

mainlyrelyondecreasingCO2emissions.Asaconsequence,the

contributionofindustrialsectorstobridgethegapbetweenthe

INDCandthe2Cscenariofallsto13%

Fourth,reductionsinCO2 emissionsfromLULUCF(excluding

sinks) coveraround 11% in theINDC scenario Whensinks are

included, CO2 emissions fall by 1.4 Gt in the INDC scenario

comparedtotheReference,aresultthatiscomparablewiththe

numberof1.6GtobtainedbyGrassiandDentener(2015).Moving towardsa2Cpathwayimpliesamoresubstantialcontributionof

CO2reductioninLULUCF.Someregionswithasignificantshareof emissions fromLULUCF haverelativelyunambitious INDCs.For these regions, reducing CO2 emissions from LULUCF are cost-effective options In addition, due to a relatively flat marginal abatementcostcurve,avoideddeforestationbecomesan impor-tantsourceofemissionreductionsinreachingthe2Ctarget

improvements in energy efficiency beyond what is realized in theReference)andafuelshiftinthebuildingandtransportsector and areduction ofmethaneemissionsinwasteandagriculture sectors(seeIPCC,2014,Chapters10and11,respectively,foramore in-depthdiscussionofthetechnologicaloptions)togethercover aroundonefifthofthetotaldecreaseofGHG

4.4.Electricitygeneration

contributionofthepowersectortotheglobalemissionreductions Thissectionzoomsinonthetechnologycompositionofelectricity productioninthedifferentscenariosin2030and2050,presented

inFig.4

Afirstresultisthathighercarbonpriceslowerthetotallevelof electricityconsumption.Bothin2030andin2050,theINDCand

2C scenarios slightly reduce global electricity consumption compared to the Reference This result illustrates that energy

45 50 55 60 65

Gt CO2e

CO2 Non-CO2

31%

Energ y se ctor Industry Land use (c hang e), for estry Buildings, tr ansport, waste Agricul ture

18%

17%

11%

10%

9%

17%

13%

17%

13%

10%

Fig 3 Sector contributions to greenhouse gas emission reductions in 2030 The percentage above the bars indicates the share in reductions between scenarios CO 2 emissions exclude sinks The darker, lower end of the bar represents CO 2 reductions, while the upper part in a lighter colour shows the reductions in non-CO 2 greenhouse gases Non-CO 2 emission reductions in electricity generation and CO 2 emissions in agriculture are hardly visible, while emission reductions from land use, land use change and forestry (LULUCF) only cover CO 2 emissions Energy sector emissions include greenhouse gases emitted during extraction, production, transformation (e.g refining) and transport of energy fuels and associated fugitive emissions.

0 20000 40000 60000

Other Solar Wind Hydro Biomass Nuclear CCS Gas Oil Coal 2050

Fig 4 Electricity generation by technology in the Reference, the INDC scenario and the 2C scenario in 2030 and 2050 at global level Carbon Capture and Storage (CCS) covers coal-, gas- and biomass-fired electricity generation with CCS Other technologies include geothermal electricity, wave and tidal energy, and (stationary) hydrogen fuel cells Units are expressed in terawatt hour of electricity (TWhe).

totalenergydemand,mainlyinthebuildingandtransportsector

after2030,leadingtolowerelectricityconsumptionlevelsoverall

By2030,theINDCsleadtoatransformationofthepowersector

throughasubstitutionfromfossilfuelstolow-carbon

technolo-gies In the Reference, fossil fuels account for around 60% of

electricityproduction.Thisnumberreducesto53%and47%inthe

INDCand2Cscenario,respectively.Thedecreaseintheshareof

increasingshareoflow-carbontechnologies,mainlynuclearand

windenergy,butalsobiomass,hydroandsolar.Gas-firedpower

Referenceaswellasinbothscenarios

Inthelongerrun(2050),CarbonCaptureandStoragebecomes

animportanttechnologyforclimatechangemitigationpolicy.In

the2Cscenario,electricitygenerationfromcoalwithoutCCSis

closetozero.Inaddition,carbonpricesleadtomoreelectricity

being generated from nuclear, solar, wind, biomass and other

(geothermal,tidal,hydrogen)energycomparedtotheReference

The2Cscenarioimpliessubstantialinvestmentsinwindandsolar

capacity,whichunlocks(endogenous)technologicalprogressfor

thesetechnologies.Asaresult,windandespeciallysolarpower

becomes morecompetitive in the 2C, and consequently gains

marketshare

Fig 5 sheds more light on the technological progress in

electricity production technologies (in the 2C scenario; the

ReferenceandtheINDCcurvesfollowasimilartrendininvestment

costs) Incorporating technological change can have important

implicationsfortheoptimalemissiontrajectory.Aspointedoutby

vanderZwaanetal.(2002),includingtechnologicalimprovement

in climatechangemodelling maylead tofaster deployment of

renewables.TheJRC-POLESmodelincludestechnologicalprogress

re-sponsetothecumulativeinstalledcapacitiesonagloballevel.Fora

broaderdiscussionontheapproachesusedintheliterature,we

refertoLöschel(2002)andGillinghametal.(2008).Atwo-factor

learning-by-researchinthePOLESmodelisdescribedinCriquietal.(2015)

presentedinvanderZwaanetal.(2013)andvanSluisveldetal (2015).Thetechnologicalprogressinelectricitygenerationfrom solarstandsoutfromFig.5.Furthermore,theinvestmentcostsof oil and gas power plant installation decrease, but represent a smallerfractionoftotalcostsduetohighervariablecostsoffuel input

4.5.Macro-economiccosts Thissectionandthetwosectionsthatfollowconcentrateonthe economicimpactofclimatechangemitigationpolicies.Notethat thescenarioshereimplementadomesticemissiontradingscheme

butwithoutinternationaltradeofpermits.Section4.7considers carbon taxes and studies alternativerevenue recycling mecha-nisms

TheresultsoftheINDCscenariosuggestthattheParispledges haveonlyalimitedimpactonworldaggregateGDPof 0.42%.The

2Cscenarioimposesstrongerconstraintsonemissions,leadingto moresubstantialtransformationseconomy-wide.Thisisreflected

inareductionofglobaleconomicoutputlevelsof 0.72% Fourcommentstoframetheseresultsareinorder.First,yearly growthratesremainhigh:the2.98%yearlygrowthofglobaloutput levelintheReferencefor theperiod 2020–2030isonlyslightly reduced to2.93% and 2.90% in theINDC and 2C, respectively Hence, climate mitigation policies are compatible with robust

thatweonlyassessthecostsideofmitigationpolicyanddonot incorporatetheavoideddamagesofclimatechange.The JRC-GEM-E3modelisbasedonoptimisingbehaviouroffirmsand house-holdsundermyopicexpectations.Inabsenceofthemodellingof damagesofclimatechange,imposingGHGemissionrestrictionsin themodel implies thatagents havefeweroptions tomaximise profits or welfare Therefore, the resultsshould be seen as an assessmentoftheabatementcostandshouldnotbeconfusedwith theresultofacost-benefitanalysis.Third,theseresultsareinline withIPCC(2014),asshowninFig.6below.Foreachofthemodels involvedwithendogenousGDP,Fig.6(panela,left-handside)plots the model- and scenario-specific changein GDPaggregated at global levelagainst thecorrespondingreduction in greenhouse

emissions are expressed here relative tothe respective model references or baselines Results from different projects are included,includingEMF27 (Weyantetal.,2014), EMF22(Clarke and Weyant, 2009), AMPERE(Kriegleret al., 2015)and LIMITS (Kriegleretal.,2013;Tavonietal.,2014).TheFigureshowsaclear relationbetweenabatementeffortandcost,butwithsubstantial heterogeneityduetodifferingassumptionse.g.onavailabilityof technologies.Theright-handsideofFig.6(panelb)illustratesthat higheremissionlevelsintheReferencerequirestrongeremission reductionsrelativetothis Referenceinorder tomeet thesame targetfortemperatureincrease(indicatedbythecoloursinFig.6) Someofthereferencesorbaselinesdonotincludethepoliciesthat arecurrentlyinplace,whichexplainswhytheemissionlevelsin theReferenceoftheanalysispresentedinthispaperarerelatively low.Fourth,byimplementingregion-specificemissionreduction targetsbasedontheresultsoftheJRC-POLESmodeloptimization exercise in the 2C scenario,we get differentcarbon prices in variousregions.Anefficientscenariowithauniformglobalcarbon priceislikelytoleadtoalowercostestimateonaglobalaverage

Ontheotherhand,theresultspresentedheremayunderestimate thecostofclimatepoliciesinreality.Lobbygroups,overlappingor partial (e.g sector-specific instead of economy-wide) policies, institutional barriers,myopicpolicy-makersand theabsenceof

900

1000

1500

2000

2500

3000

4000

5000

6000

Solar central ized Solar dis tribut ed Wind offsho re Wind onsh ore Hydro Biomass Nuclear Biomass wit h C CS Gas wit h C CS Coal with C CS Gas Oil Coal

Investment cost

of ne w capac ity

US $ 2 015 / kW

Cumulat ive capacity insta lled ( GW)

Impl ied learning rates 2015 -2050 (%)

10

5 1 5 11

6 11 2

5 2 2

2010 2030 2050

12 9

Fig 5 Technological progress in electricity generation technologies in the 2  C

scenario from 2010 to 2050 The learning curves depicted here are based on a

learning-by-doing approach and show the relation in capacity investment cost and

installed capacity on the global level The representative technologies shown here

are conventional thermal turbines for coal, oil, gas and biomass; pressurized water

reactor generation III/III+ for nuclear; and large hydro installations Progress in

technologies with Carbon Capture and Storage (CCS) aggregates the learning in CCS

technology with the learning in the relevant coal-, gas- and biomass-fired electricity

generation technologies The learning rate is defined as the percentage cost

prices) could lead to suboptimal policies from an economic

efficiencypointofview

differentiation across regions and sectors The following two

sectionsthereforedisaggregatetheseresultstoprovidea better

effectsoftheINDCand2Cscenarios

4.6.Regionaleconomicimpact

OneofthemainnoveltiesoftheParisCOP21isthebottom-up

policyframework:countriesputforwardINDCsandconsequently

revealthe level ofambition of theirclimatechange mitigation

policies.Thebroadrangeofambitionlevelsislikelytotranslate

into economic impacts that differ substantially across regions

Differences in historical emission reduction efforts, energy

intensity,sectorcomposition,naturalresourceendowments,the

productionoffossilfuels,therelativeimportanceoftrade-exposed

additionalfactorsthatmaygiverisetoimpactvariationbetween regions.Alltheabove-mentionedaspectsarecapturedbythe JRC-GEM-E3analysis,ofwhichtheresultsaredisplayedinFig.7and

Table3

AfirstpointillustratedbytheINDCscenarioresultsisthata substantialnumberofregionsundertakesignificantclimateaction that leads to relatively small reductions in GDP (less than 1% reductionfromtheReferencein2030)comparedtotheReference However,theINDCscenarioshowsthatanumberofregionshave relativelyunambitioustargets,suchthattheiremissionlevelsare closetoorevenslightlyhigherthanintheReferencein2030.Some

oftheseregionsgainincompetitivenesscomparedtoregionswith more ambitious climate change mitigation policiesand conse-quentlyhavemarginallyhigherGDPlevelsthanintheReference.In themajorityoftheseregions,exportsincreaseorimportedgoods arereplacedwithdomesticallyproducedgoods(Table3).Hence, carbonleakageleadstoageographicalshiftofemission-intensive production

Fig 6 Impact on global aggregate GDP of the INDC and 2  C scenario in 2030 (JRC-GEM-E3 results) compared with results (of models with endogenous GDP) included in the IPCC AR5 WGIII Scenario Database ( IIASA, 2015b ) Each dot represents a model- and scenario-specific result, relative to the respective baselines Temperature ranges are based

on IPCC (2014) with at least 60% probability for the scenarios below 2  C, and 55% probability for staying between the ranges 2–3  C and above 3  C GHG reduction of the 2  C scenario and the INDC scenario cover emissions from energy, industry and agriculture, excluding LULUCF a) More stringent temperature targets require stronger emissions reductions leading to higher abatement costs b) Higher emission levels in the Reference or baseline imply stronger reductions relative to the Reference to meet a similar target for the rise in global temperature.

Fig 7 GDP impact by region in the INDC and 2  C scenario (% change from Reference in 2030) Colours reflect income groups as expressed by GDP per capita in 2010 (market prices, constant 2004 thousand US $) Some of the labels are omitted to improve the clarity of the figure; numerical results provided in Table 3 GHG emissions cover emissions from energy, industry and agriculture, excluding LULUCF a) Emission levels that deviate stronger from the Reference imply larger GDP impacts, although there is substantial



inpanel a, left-handside) reveals a shiftdownand totheleft

implies stronger emission reductions, leading tomore sizeable

GDPimpactscomparedtotheReferencein2030.PanelbofFig.7

displaysthegreenhouse gasemission reductionsrelativetothe

levelsintheyear2010.ThisvisualizationshowsthattheINDCsof

comparedtohistoricallevels.Inaddition,theright-handsideof

Fig.7 illustrates clearly that the 2C target can be met while

allowinglow-incomeregionstoincreaseemissionsrelativetothe

levelsobservedin2010

Amoredetailedanalysisoftheresultsofthe2Cscenarioyields

anumberoffindings.First,fossilfuel-producingregions,suchas

SaudiArabiaandRussia,experiencearelativelystrongdropinGDP

comparedtotheReferencein2030.TheReferencedoesnotassume

atrend-breakingtransformation towardsadiversifiedeconomy,

suchthat economic activity in some countriesremains torely

heavilyon fossilfuel exports Asindicated inTable 2, the2C

pathwayleadstodemandreductionsforoil,gasandsolidfuels

Since these goods typically represent a substantial share of

economicactivityandexportsinsomeofthefossil-fuelproducing

regions,strong global climateaction appears tolowertheGDP

levelsinthesecountries.Second,theclimateambitionsinfluence

therelativecompetitive positions betweencountries Indiais a

countriesforwhichcarbonpricesconvergetorelativelylowlevels

(around26US $(2005)in2030) leadstocompetitivegains:an

increaseintheexportsofenergy-intensiveindustriesdriveGDPto

higherlevelsthanintheINDCscenarioin2030.Moregenerally,the

contributionof changesin trade balancetothechange inGDP

differsbyregionsandispositiveforsome,butnegativeforothers

Third,forsomeLatinAmericancountries,suchasArgentinaand

Brazil,the agricultureand consumer goods industry (including

foodproductionandprocessing)representasignificantshareof

economicactivityandarestronglyaffectedbyemissionreductions

policies.AsshowninSection4.3,agricultureisoneofthesectors with substantial (non-CO2) emission reduction potential The resultisthatthedropinGDPcomparedtotheReferencein2030is strongrelative tothereductionlevels forArgentina and Brazil Hence, sector-specific considerations are an important driver behindtheresults.Therefore,thenextsectiondisaggregatesthe globaleconomicimpactbysector

Investmentsonaveragearereducedless thantheotherGDP componentsas,despitethereductionofeconomicactivitydueto the reallocation of resources, the mitigation action is closely related tolow-carboninvestmentsin thepower, industrialand

decreasesmorethanGDPfornearlyallregionsasmostdomestic andinternationalpricesincreaseduetothecarbonpriceandthe reallocationofresourcesawayfromtheoptimalallocationofthe Referencescenario

contains substantialclimateaction,as indicatedinTable 1 The resultspresentedherethusonlylookattheimpactofadditional climatepolicies.Sinceambitiouslegislationisalreadyinplace,the ReferenceisclosetotheINDCscenariofortheEU.Inparticular,the Referenceincludesthe2020ClimateandEnergyPackage,which impliesa20%cutingreenhousegasemissionscomparedto1990,a

considersthe2030ClimateandEnergyFramework:40%reduction

ofGHGemissionscomparedto1990(43%comparedto2005inthe

comparedto2005innon-ETSsectors),27%renewablesinenergy consumption andan indicativetarget 27%forimprovements in energyefficiencycomparedtoprojectionsby2030

4.7.Sector-specificeffects Thissectiondisaggregatestheglobalresultsonasector-specific basis.Table4presentsoutputlevelsandchangesinemployment

Table 3

Macro-economic results of climate change mitigation; GHG changes exclude LULUCF.

assumeacommoncarbonpriceacrossallsectorswithinaregion

Thenotableexceptionis theEU, whereweimplementdifferent

targets between ETS and non-ETS sectors, as discussed in the

previoussection

Afirstobservationisthatrelativelystrongreductionsinoutput

and,correspondingly,employmentlevelsoccurinthefossilfuel

sectors:coal,(crude)oilandgas.Theseresultsareconsistentwith

Section4.1.Theunderlying explanationisthat strongerclimate

policiesleadtomoreefficientuseofenergyandtoashiftinthe

compositionoffuelconsumption.Energyefficiencyalsoleadstoa

lowerdemandforelectricity,which resultsinloweroutputand

employmentlevelsinthepowersector,inlinewithSection4.4

Table 4 shows the electricity supply sector as an aggregate of

generation, transmission and distribution, and illustrates that

sufficient tocompensate for the employment reduction due to

lower electricity demand and for the jobs lost in coal-based

electricity generation.The results hereconsider economy-wide

feedback mechanismsand inter-industry interactions via

inter-mediateinputs.Thereforetheyshouldbeseenascomplementary

withtheresultsinprevioussections

Second,energyintensivesectors, suchas ferrousmetals and

non-metallicmineralsareamongthesectorsthataremostaffected

bystrongerclimatepoliciesduetomoregreenhousegas-intensive

production inputstructures In addition, someof these sectors

greenhouse gases, as discussed in Section 4.3 Conversely, the

impactonoutputlevelsofrelativelylow-carbonservicesectorsis

smaller

Theresultsonemployment includeadditionalscenariosthat

explicitlyconsidertheimpactofrevenuerecyclingandalternative

scenarios withtaxrecycling (indicatedby‘Labour tax recycling:

yes'inTable4 therevenueraisedbycarbontaxesisusedtolower

existing distortionary labour taxes As a consequence, labour

becomesamoreattractiveinputintheproductionprocess,leading

tomorejobseconomy-wide:thejobdecreaseismitigatedfrom

0.34%to 0.26%intheINDCscenario,andfrom 0.74%to 0.66%

unem-ploymentratesaccordingtoawagecurvemechanism(indicated

empirical evidence (Blanchflowerand Oswald,1995), while the latterrepresentstheviewthatclimatepolicywillnotaffectthe fundamentaldeterminantsofunemploymentinthelongrun,such

BlanchardandKatz, 1997,forabroaderdiscussion).Theoutcomeof

transitionofjobsfromemission-intensivesectorstolow-carbon, serviceorientedsectors,inlinewiththefindingsofHafsteadand Williams(2016).ThejobtransitionisclearlyillustratedbyFig.8

(fixedunemploymentrate,withlabourtaxrecycling).Inaddition,

Fig.8showsthatthesectorsthatexperiencethestrongestnegative impactintermsofemploymentarenotnecessarilythesectorsthat providethelargestnumbersofjobs(indicatedbytheheightofthe barsinFig.8)

5.Conclusions Thispaperprovidesamodel-basedassessmentoftheINDCs,a centralelementintheglobalclimatechangenegotiationsheldin

Table 4

Sector-specific output and employment results in 2030.

Fig 8 Transition of jobs from energy-intensive sectors to more service-oriented sectors The employment impact per sector is shown for the 2  C scenario with carbon tax revenue recycling via lower labour taxes and fixed unemployment rates per region The length of the bars shows the percentage change relative to the Reference in 2030, while the height of the bars is scaled to reflect the employment levels in the Reference in 2030 As a result, the surface of the bars reflects the change