未来能源研究所-估算供应方化石燃料干预措施的减排量-50页VIP专享VIP免费

下载本文档

阅读 977
格式 pdf
大小 3.17 MB
约50页
2023-05-07
收藏
点赞(0)
海报
举报

Estimating the Emissions

Reductions from Supply-side

Fossil Fuel Interventions

Brian C. Prest, Harrison Fell, Deborah Gordon, and TJ Conway

Working Paper 23-11

April 2023

Resources for the Future i

About the Authors

Brian C. Prest is an economist and fellow at RFF specializing in the economics of

climate change, energy economics, and oil and gas supply. Prest uses economic

theory and econometrics to improve energy and environmental policies by assessing

their impacts on society. His recent work includes improving the scientific basis of the

social cost of carbon and economic modeling of various policies around oil and gas

supply. His research has been published in peer-reviewed journals such as Nature,

the Brookings Papers on Economic Activity, the Journal of the Association of

Environmental and Resource Economists, and the Journal of Environmental

Economics and Management. His work has also been featured in popular press outlets

including the Washington Post, the Wall Street Journal, the New York Times, Reuters,

the Associated Press, and Barron’s.

Harrison Fell is an associate professor in the Department of Agricultural and

Resource Economics at North Carolina State University. His research focuses on

environmental, energy, and natural resources economics, with an emphasis on policy-

relevant work. Dr. Fell’s recent work is related to energy systems with a focus on

renewable energy, emissions trading systems, and electricity regulation.

Deborah Gordon is a senior principal in RMI’s Climate Intelligence Program, where she

coleads the Oil and Gas Solutions Initiative. Gordon serves as a senior fellow at the

Watson Institute of International and Public Affairs at Brown University and is an

affiliate at the Brown Climate Solutions Lab. Her research spearheaded the

development of the Oil Climate Index plus Gas (OCI+), a first-of-its-kind analytic tool

that compares the life-cycle climate impact of global oil and gas resources. The OCI+

is the topic of Gordon’s new book, No Standard Oil (Oxford University Press, 2022).

TJ Conway is a principal in RMI’s Climate Intelligence Program. He plays a leading role

in the Oil and Gas Solutions Initiative, working with a range of stakeholders to help

accelerate oil and gas firms’ decarbonization efforts. Conway is also Professor of the

Practice at Georgetown University, where he teaches an energy course through the

Landegger Program in International Business Diplomacy at the Edmund A. Walsh

School of Foreign Service.

Acknowledgments

Prest and Fell gratefully acknowledge the support of Onyx Transition. We appreciate

thoughtful comments from Mike Braun, Seth Zimring, and Darius Nassiry.

Estimating the Emissions Reductions from Supply-side Fossil Fuel Interventions ii

About RFF

Resources for the Future (RFF) is an independent, nonprofit research institution in

Washington, DC. Its mission is to improve environmental, energy, and natural resource

decisions through impartial economic research and policy engagement. RFF is

committed to being the most widely trusted source of research insights and policy

solutions leading to a healthy environment and a thriving economy.

Working papers are research materials circulated by their authors for purposes of

information and discussion. They have not necessarily undergone formal peer review.

The views expressed here are those of the individual authors and may differ from

those of other RFF experts, its officers, or its directors.

Sharing Our Work

Our work is available for sharing and adaptation under an Attribution-NonCommercial-

NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license. You can copy and

redistribute our material in any medium or format; you must give appropriate credit,

provide a link to the license, and indicate if changes were made, and you may not

apply additional restrictions. You may do so in any reasonable manner, but not in any

way that suggests the licensor endorses you or your use. You may not use the

material for commercial purposes. If you remix, transform, or build upon the material,

you may not distribute the modified material. For more information, visit

https://creativecommons.org/licenses/by-nc-nd/4.0/.

/50

下载本文档

文本预览下载提示常见问题

EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventionsBrianC.Prest,HarrisonFell,DeborahGordon,andTJConwayWorkingPaper23-11April2023ResourcesfortheFutureiAbouttheAuthorsBrianC.PrestisaneconomistandfellowatRFFspecializingintheeconomicsofclimatechange,energyeconomics,andoilandgassupply.Prestuseseconomictheoryandeconometricstoimproveenergyandenvironmentalpoliciesbyassessingtheirimpactsonsociety.Hisrecentworkincludesimprovingthescientificbasisofthesocialcostofcarbonandeconomicmodelingofvariouspoliciesaroundoilandgassupply.Hisresearchhasbeenpublishedinpeer-reviewedjournalssuchasNature,theBrookingsPapersonEconomicActivity,theJournaloftheAssociationofEnvironmentalandResourceEconomists,andtheJournalofEnvironmentalEconomicsandManagement.HisworkhasalsobeenfeaturedinpopularpressoutletsincludingtheWashingtonPost,theWallStreetJournal,theNewYorkTimes,Reuters,theAssociatedPress,andBarron’s.HarrisonFellisanassociateprofessorintheDepartmentofAgriculturalandResourceEconomicsatNorthCarolinaStateUniversity.Hisresearchfocusesonenvironmental,energy,andnaturalresourceseconomics,withanemphasisonpolicy-relevantwork.Dr.Fell’srecentworkisrelatedtoenergysystemswithafocusonrenewableenergy,emissionstradingsystems,andelectricityregulation.DeborahGordonisaseniorprincipalinRMI’sClimateIntelligenceProgram,whereshecoleadstheOilandGasSolutionsInitiative.GordonservesasaseniorfellowattheWatsonInstituteofInternationalandPublicAffairsatBrownUniversityandisanaffiliateattheBrownClimateSolutionsLab.HerresearchspearheadedthedevelopmentoftheOilClimateIndexplusGas(OCI+),afirst-of-its-kindanalytictoolthatcomparesthelife-cycleclimateimpactofglobaloilandgasresources.TheOCI+isthetopicofGordon’snewbook,NoStandardOil(OxfordUniversityPress,2022).TJConwayisaprincipalinRMI’sClimateIntelligenceProgram.HeplaysaleadingroleintheOilandGasSolutionsInitiative,workingwitharangeofstakeholderstohelpaccelerateoilandgasfirms’decarbonizationefforts.ConwayisalsoProfessorofthePracticeatGeorgetownUniversity,whereheteachesanenergycoursethroughtheLandeggerPrograminInternationalBusinessDiplomacyattheEdmundA.WalshSchoolofForeignService.AcknowledgmentsPrestandFellgratefullyacknowledgethesupportofOnyxTransition.WeappreciatethoughtfulcommentsfromMikeBraun,SethZimring,andDariusNassiry.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventionsiiAboutRFFResourcesfortheFuture(RFF)isanindependent,nonprofitresearchinstitutioninWashington,DC.Itsmissionistoimproveenvironmental,energy,andnaturalresourcedecisionsthroughimpartialeconomicresearchandpolicyengagement.RFFiscommittedtobeingthemostwidelytrustedsourceofresearchinsightsandpolicysolutionsleadingtoahealthyenvironmentandathrivingeconomy.Workingpapersareresearchmaterialscirculatedbytheirauthorsforpurposesofinformationanddiscussion.Theyhavenotnecessarilyundergoneformalpeerreview.TheviewsexpressedherearethoseoftheindividualauthorsandmaydifferfromthoseofotherRFFexperts,itsofficers,oritsdirectors.SharingOurWorkOurworkisavailableforsharingandadaptationunderanAttribution-NonCommercial-NoDerivatives4.0International(CCBY-NC-ND4.0)license.Youcancopyandredistributeourmaterialinanymediumorformat;youmustgiveappropriatecredit,providealinktothelicense,andindicateifchangesweremade,andyoumaynotapplyadditionalrestrictions.Youmaydosoinanyreasonablemanner,butnotinanywaythatsuggeststhelicensorendorsesyouoryouruse.Youmaynotusethematerialforcommercialpurposes.Ifyouremix,transform,orbuilduponthematerial,youmaynotdistributethemodifiedmaterial.Formoreinformation,visithttps://creativecommons.org/licenses/by-nc-nd/4.0/.ResourcesfortheFutureiiiAbstractSupply-sideinterventionsthatretirehighlyemittingfossilfuelassetshavereceivedincreasedattentionfrompolicymakersandprivateactorsalike.Yetconcernsaboutmarketleakage—whereinreducedsupplyfromonesourceispartiallyoffsetbyincreasedproductionfromothersources—haveraisedquestionsabouthowmuchemissionsreductionstheycanachieve.Inthispaper,weestimatetheeffectsofthesesupply-sideinterventionsonglobalemissions,accountingforbothmarketleakageaswellastherelativegreenhousegas(GHG)intensityofdifferentsourcesofsupply.WeaccountforuncertaintyinmarketleakageratesandtheemissionsintensitiesofthecurtailedandsubstitutesourcesofsupplythroughaMonteCarloanalysis,drawingonsupplyanddemandelasticitiesfromtheeconomicsliteratureandemissionsintensitydatafromthestate-of-the-artOilClimateIndexplusGas(OCI+)dataseton586oilandgasfieldsaroundtheworld.Wefindthatthelife-cycleemissionsreductionsfromsupply-sideinterventionsareontheorderofofthegrosslife-cycleemissionsofeachbarrelcurtailed,dependingontherelativeemissionsintensityofthecurtailedandsubstitutesourcesofsupply.Further,targetingthemostemissions-intensivesourcesofoilsupplycouldachieveyetfurtheremissionsreductions.HowonecomparesmethaneandCO2emissionsalsohasimportantconsequencesforwhichsourcestotarget.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventionsivContents1.Introduction12.DerivationoftheEmissionsImpactsofSupply-sideInterventions52.1.SubstitutiontoOtherFormsofEnergy102.2.WhenCouldLeakageBeZero?TheCaseofHotellingDynamics113.EstimatingMarketLeakage123.1.GeneralPrinciplesforEstimatingLeakageParameters123.2.EstimatesofDemandElasticities163.3.EstimatesofSupplyElasticities184.EstimatesofEmissionsIntensities195.QuantitativeApplication236.Conclusion317.References32Appendix36A.1.SensitivityApproachtoEmissionsIntensityCalculation36A.2.DetailedResultsundera20-YearGWP39A.3.TheEffectofDynamics41ResourcesfortheFuture11.IntroductionWithnumerousgovernmentsandcorporationsmakingincreasinglyambitiouscarbonandmethaneemissionspledges,meetingnow-ubiquitous“netzero”goalswillrequirereducingnotonlytheconsumptionoffossilfuels,butalsotheirproductionandtheiremissionsintensities.Windingdownfossilfuelproductionhaslongbeenapolicyleverconsideredintheacademicliterature(Harstad2012;Erickson,Lazarus,andPiggot2018;LazarusandvanAsselt2018;Asheimetal.2019;vanderPloegandRezai2020;P.NewellandSimms2020;Prest2021;PrestandStock2021).Moreover,thelife-cycleemissionsintensityofoilandgasproduction—includingScope1,2,and3emissions—hasbeenstudiedandfoundtobewide-ranging(Masnadietal.2018;Jingetal.2020;Gordon2023;Gordon,Tan,andFeldman2016).Inrecentyearstheseso-called“supply-side”climatepolicieshavegainedpopularityamongpolicymakers.Forexample,intheUnitedStates,PresidentJosephBidencampaignedonendingoilandgasleasingonfederallands,andhisadministrationhasslowedfederalleasingactivity.Morerecently,privateactorshavebeguntoproposemonetizingtheretirementofemissions-intensiveassets,startingwithcarbon.Forexample,newlyproposed“carbonretirementportfolios”(HandlerandBazilian2021)wouldpurchaseemissions-intensiveassetslikeoilandgaswellsorcoal-firedpowerplantssimplytoretirethem.TheresultingreductionsinGHGemissionscouldthenbemonetizedinmanyforms,includingbyattractingcapitalfromclimate-consciousinvestors,generatingandsellingcarboncredits,orreceivingdirectpaymentsfromgovernmentspertonofemissionsreduced.Forthosecarboncreditsorpoliciestobecredibleandeffective,theirnetimpactsonemissionsmustbeestimatedasrigorouslyaspossible.Evenifonecancompellinglydemonstratetheemissionsavoideddirectlyfromtheassetinquestion(say,anoilfield),thenetimpactsonglobalemissionswilldifferduetoemissions“marketleakage.”Theterm“marketleakage”asusedinthispaper,andintheeconomicsliteraturemorebroadly,describesthephenomenonwhereinreducedsupplyfromonesourceispartiallyoffsetbyincreasedproductionfromothersources.Thisreplacementproductionoffsetssome,butgenerallynotall,oftheemissionsbenefitsfromthesupply-sideintervention.Somemaketheextremeclaimthatmarketleakagefromsupply-sideinterventionsis100percent—forexample,implyingthattakingabarrelofoilproductionoffthemarkethasnonetimpactonoilconsumptionbecauseitisoffsetone-for-onebyincreasedproductionelsewhere.However,thisargumentiscontradictedbybasiceconomictheory.Ineconomictheory,thepriceandconsumptionofaproductsuchasoilaredrivenbytheintersectionofsupplyanddemandcurves.Ifasupply-sideinterventionremovesproductionfromaninframarginalfield(i.e.,afieldwithamarginalcostproductionbelowthecurrentmarketprice),thisshiftsthesupplycurvetotheleft,asshowninFigure1.Theshiftinthesupplycurveresultsinanew,higherequilibriumprice.AscanbeseeninFigure1,theretirementofaninframarginalfieldconcurrentlyEstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions2leadstoareductionintheequilibriumquantity,thoughthisreductionistypicallylessthanthequantityofoilretired.Figure1representsastylizedexample,butrealityismorecomplex,withsupplycurvesthatarenotsimplystraightlinesandthatchangeshapeovertimeasnewoilfieldsarediscoveredanddeveloped.However,Figure1nonethelessrepresentsthekeyfactorsdrivingmarketleakage:theslopesofsupplyanddemandcurvesaroundtheirpointofintersection.Figure1.StylizedIllustrationoftheLeakageMechanismFortheretirementoftheinframarginalfieldtohavenoeffectontheequilibriumquantity,demandforoilwouldhavetobeperfectlyinsensitivetoprice,meaningthedemandcurvedoesnotslopedownwards,butratherisaverticalline.1Thiscontentionisimplausible,particularlyinthecontextofincreasedavailabilityofoilsubstitutesincertaindemandsegments—suchaselectricvehicles(EVs),justoneexampleofhowconsumerscanrespondtopricechangesbyusinglesspetroleumproducts.The1Leakagecouldalsobe100percentifoilsupplywereinfinitelyelastic—thatis,aperfectlyflatsupplycurve—whichisevenmoreimplausiblebecauseitwouldimplythatthepriceofoilisaconstantvaluethatisunaffectedbyfluctuationsindemand.ResourcesfortheFuture3mechanismbywhichcurtailedoilsupplyreducesconsumptionisstraightforward:areductioninsupplyshiftsthesupplycurveupandtotheleft;thisincreasesthepriceofoil,2andsincedemandcurvesingeneralslopedownwards,oilconsumersrespondbyusingless(movingupandtotheleftalongthedemandcurve).Figure1showsthismechanisminasimplesupplyanddemanddiagram,demonstratingthatsome,butnotall,ofcurtailedsupplyisoffsetbyleakage;onlyintheunrealisticcasewherethedemandcurveisperfectlyvertical(“perfectlyinelastic”)willleakagebe100percent.Theresponsesofconsumers’quantitydemandedtopricesignalsarenotlimitedtotravelingless(say,throughworkingremotelyanadditionaldayoftheweek);therearealsootheravenuesofdemandresponse,suchasimprovingfuelefficiencyorswitchingtoalternativessuchasEVsorothermodesoftransit.Beyondthetransportationsector,oilisusedinpetrochemicalandotherindustrialapplicationsaswellasheating,allofwhichcouldfeaturetheirownchannelsofdemandresponse.Overall,unlessconsumersarecompletelyinsensitivetoprice,someoftheeffectofreducedsupplyismanifestbyreducedconsumption—thatis,leakageisincomplete,implyingthatoveralloilconsumptionandemissionsdecline.Whilethetheoreticalbasisforsomelevelofleakageisclear,itisdifficulttoempiricallyestimatebecauseitisgenerallynotdirectlyobservable.IfoneweretoreduceabarrelofsupplyfromawellinNewMexico,and,inresponse,aproducerrampsupitsproductionby,say,halfabarrelsomewhereacrossthecountryoreventheglobe,itisnotfeasibletoobservethisspecificresponse.Hence,estimatesofleakageratestypicallyrelyoneconomicmodelingoftheglobaloilmarket.Anadditionalchallengetomeasuringleakageisthatthe“leaked”replacementsupplymayhaveahigherorlowerGHGintensity,whichhasimplicationsforthenetimpactsofasupply-sideintervention.Forexample,ifarelativelyemissions-intensivesourceofoilproductioniscurtailedandthatsupplyispartiallysubstitutedbycleanersources,thiswouldreduceemissionsevenifmarketleakageis100percent.Hence,onemustalsoaccountfortherelativeemissionsintensityofdifferentsourcesofoilsupply.Duetothevariationincrudeoilqualityandtheemissionsintensityofsupplychains,oilsproducedfromdifferentsourceshavedifferentlife-cycleclimateimpacts(Masnadietal.2018;Gordon2021;RMI2022).Sincethereisachoiceofwheretoengageinsupply-sideinterventions,theemissionsintensityoftheoilcurtailedisinpartachoicevariable.Itisperhapsobviousthatsupply-sideinterventionsshouldthentargetoilswiththehighestemissionsintensitiesfirst,suchasCanadianoilsandsorheavycrudesinCaliforniaorelsewhere.2Inpractice,refineriesoperateasintermediariesbetweenproducersofcrudeoilanditsend-users—primarilytransportationfuelslikegasolineanddiesel—andrefinerscanadjusttomakedifferentproductswhenthecompositionofpetroleumproductdemandshifts.Nonetheless,thepriceofgasolineiscloselylinkedtothatofcrudeoil,sopriceincreasesinthemarketcrudeoilarelargelypassedontoconsumers.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions4Credibleestimatesofthetwokeycomponentsdeterminingtheemissionsimpactsofsupply-sideinterventions—marketleakageratesandrelativeemissionsintensities—arekeytoensuringthatestimatedemissionsreductionsarecredible.Thisisparticularlyimportantwhensuchcalculationshavedirectimplicationsformarketprices,carboncreditvolumes,andhencefinancialflowsbetweenmarketparticipants.Thispresentsachallenge:howonecancrediblyandtransparentlyestimatetheemissionsdisplacedbysupply-sideinterventionswhensuchcalculationsarenecessarilyabstractduetotheinfeasibilityofdirectlyobservingrealizedemissionsimpacts.Inthispaper,weusestandardtoolsofeconomicswhileinvokingminimalassumptionstoderiveamathematicalformularepresentingtheemissionsimpactsofasupply-sideintervention(say,removingonebarrelofoilsupplyfromthemarket).Wethenshowdifferentapplicationsofthatformulausingreasonablerangesofkeyinputs:namely,estimatesofsupplyanddemandelasticitiesandrelativeemissionsintensitiesofcurtailedandsubstitutesourcesofsupply.Sinceestimatesoftheseinputsareinherentlyuncertainandlikelytochangeovertime,nosinglenumbercancrediblyrepresenttheemissionsreductions;rather,weuseMonteCarloanalysistoestimatecentralvaluesandrangesfornetemissionimpactsofcurtailedoilsupply,reflectinguncertaintiesinthekeyinputs(supplyanddemandelasticitiesandemissionsintensityvalues)andpotentialsourcesofsubstitutesupply.Ingeneral,ourcentralestimatesfindnetemissionsreductionsregardlessofthesourceofcurtailedandsubstitutesupply,althoughthemagnitudesanduncertaintyrangesvaryconsiderablydependingonthosefactors.Wepresentestimatednetemissionsreductionsanduncertaintyrangesforagridoftypesandsourcesofcurtailedsupply(e.g.,heavyversuslightoil,sweetversussour)andpotentialsourcesofsubstitutesupply(e.g.,byregion).Forexample,ifproductionfromtherelativelyemissions-intensiveCanadianoilsandsiscurtailed,andthesourceofsubstituteproductionischosenrandomlyinproportiontoafield’sproduction,theexpectednetemissionsreductionsare357kilogramsofcarbondioxideperbarrelofoilequivalent3(usinga100-yearglobalwarmingpotential,[GWP]),orkgCO2e/boe,(95percentrange:178to523,accountingforuncertaintyintheleakagerateanduncertaintyintheemissionsintensityofsubstitutesupply).Ifweinsteadassumethatmarketleakageisdrivenbysimilarlyemissions-intensiveproductionfromotherCanadianoilsands,theestimatedemissionsreductionsareabout20percentsmaller:285kgCO2e/boe(95percentrange:86to471kgCO2e/boe).Forreference,theemissionsintensityofoilsandsisabout648kgCO2e/boe,suggestingnetemissionsreductionsontheorderofabouthalfofthegrosslife-cycleGHGemissionsofthecurtailedbarrel.Whiletheaboveresultsreflecta100-yearGWP,theemissionsreductionsundera20-yearGWP,whichgivesmoreemphasistomethaneemissions,wouldbelarger.3Allemissionsintensityvaluesinthispaperrepresentlife-cycleemissions,includingScope1,2,and3emissions.ResourcesfortheFuture5Ingeneral,ourcentralestimatesimplythatabarrelofcurtailedoilleadstonetemissionsreductionsregardlessofthetypeorlocationofthecurtailedbarrel.The95percentuncertaintyrangesnonethelessshowthatinsomenichecases,thereisasmallchancethatcurtailmentofrelativelycleansupplycouldincreaseemissionsifsubstituteproductioncomesfromrelativelyemissions-intensivesources.However,thesescenariosgenerallyinvolvelessplausiblesubstitutionpatterns,suchasanassumptionthatcurtailmentoflightoilproductionisreplacedsolelybyoilsandsproduction,ratherthanotherlightoilsources.OurMonteCarloanalysisfindsnetemissionsreductionswithahighdegreeofcertaintywhenthecurtailedsourcesofsupplyarehighlyemissions-intensiveoillikeorheavyoilsingeneral(99.7percent)orCanadianoilsandsinparticular(99.96percent).2.DerivationoftheEmissionsImpactsofSupply-sideInterventionsTheeconomicsofleakagecanbederivedusingthestandardeconomictoolsofsupplyanddemand.Wefocusonthemarketforcrudeoil,agloballypricedcommodity.Westartbyassumingthereare𝑁𝑁consumersandsuppliersofcrudeoil,indexedby𝑖𝑖,anddenotingthequantityofcrudeoilsuppliedfromsupplier𝑖𝑖as𝑞𝑞𝑖𝑖,𝑆𝑆(𝑝𝑝),whichisafunctionofthepriceofoildenoted𝑝𝑝.Similarly,wedenotethequantityofcrudeoildemandedas𝑞𝑞𝑖𝑖,𝐷𝐷(𝑝𝑝),whichissimilarlyafunctionoftheprice.Regardlessofmarketstructure,4forthemarkettoclearitmustbethecasethattotalsupplyequalstotaldemand:5�𝑞𝑞𝑖𝑖,𝑆𝑆(𝑝𝑝)𝑁𝑁𝑖𝑖=1=�𝑞𝑞𝑖𝑖,𝐷𝐷(𝑝𝑝)𝑁𝑁𝑖𝑖=14Alternativemodelsthathavebeenusedbyeconomiststostudytheoilmarketsincludeperfectcompetition,monopoly/oligopoly,andthedominantfirm/competitivefringemodel.Wemakenoassumptionhereaboutwhichofthosemodelsismostappropriateforunderstandingoilmarkets;weinsteadmakethefarmoregeneralassumptionthatsupplyanddemandcurvesaredifferentiablewithslopesoftheappropriatesigns.5Morespecifically,supplyanddemandmustequateoveralongenoughtimehorizon.Inpractice,supplyanddemandcanbeoutofbalanceinagivenyearduetoinventories,whichcanbebuiltupanddrawndownovertime.Theappendixconsidersageneralizationofthismodeltoinclude𝑇𝑇periods,findinganalogousresultstothatofthestaticmodel.(1)EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions6Theequilibriumpriceofoilisthevaluethatbringssupplyanddemandintoalignment.Next,weaugmentequation(1)witharetirementof𝑞𝑞�barrelsofin-the-money6supplyfromsomeregion.−𝑞𝑞�+�𝑞𝑞𝑖𝑖,𝑆𝑆(𝑝𝑝)𝑁𝑁𝑖𝑖=1=�𝑞𝑞𝑖𝑖,𝐷𝐷(𝑝𝑝)𝑁𝑁𝑖𝑖=1Theleft-handsideofequation(2)isglobalsupply,netoftheretirement,denoted𝑄𝑄𝑆𝑆=−𝑞𝑞�+∑𝑞𝑞𝑖𝑖,𝑆𝑆𝑖𝑖.Theright-handsideisglobaldemand,denoted𝑄𝑄𝐷𝐷=∑𝑞𝑞𝑖𝑖,𝐷𝐷𝑖𝑖.Equation(1)canbethoughtofasaspecialcaseofthisequation,butwithnocurtailedproduction(𝑞𝑞�=0).Theeffectonthepriceofoilfromthemarginalcurtailedbarrelcanbesolvedforanalyticallybydifferentiatingequation(2)withrespectto𝑞𝑞�:−1+�𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�𝑁𝑁𝑖𝑖=1=�𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�𝑁𝑁𝑖𝑖=1Solvingforthepriceeffect,𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�⁄,yields𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=1∑𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)−𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1Becausesupplycurvesslopeupward,𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)>0,anddemandcurvesslopedownward,𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)<0,theeffectofcurtailedproductionpushesupthepriceofoil:𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�⁄>0.Theimpactonproductionineachregionisgivenbytheproductofthispriceincreaseandtheslopeoftheregion’ssupplycurve.Withoutlossofgenerality,weassumethecurtailedsupplycomesfromregion1,implyingthefollowingexpressionsforthechangeineachregion’sproduction:𝑑𝑑𝑞𝑞𝑖𝑖,𝑆𝑆𝑑𝑑𝑞𝑞�=−1+𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑑𝑑𝑝𝑝𝑑𝑑𝑞𝑞�𝑓𝑓𝑓𝑓𝑓𝑓𝑖𝑖=1𝑑𝑑𝑞𝑞𝑖𝑖,𝑆𝑆𝑑𝑑𝑞𝑞�=𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�𝑓𝑓𝑓𝑓𝑓𝑓𝑖𝑖≠16Weassumethroughoutthatthecurtailedsupplyisinthemoney,meaningithasmarginalcostsbelowthemarketprice.Ifitwereoutofthemoney,theretirementwouldhavehappenedanyway,meaningthereducedsupplyisnotadditionalandtheemissionsimpactsaretriviallyzero.Whilethisisanuninterestingcasefromtheperspectiveofthismodel,theadditionalityassumptionisvitaltocrediblyclaimingemissionsimpactsinpractice.(2)ResourcesfortheFuture7Summingacrossregionsandsubstitutingtheformulaforimpactontheoilprice,𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=1∑𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)−𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1,givestheeffectsonglobalsupplyanddemand,𝑑𝑑𝑄𝑄𝑆𝑆𝑑𝑑𝑞𝑞�=𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=∑𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1∑𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)−𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1<0Thenumeratorisnegative(demandcurvesslopedownward),andthedenominatorispositive(supplycurvesslopeupward),meaningtheoveralleffectofthecurtailedproductiononglobaloilconsumptionisnegative.Moreover,theeffectdependsontheslopesofregionalsupplyanddemand.Thisequationcanbefurthersimplifiedtobeasimplefunctionofsupplyanddemandelasticities.First,denoteregion𝑖𝑖’ssupplyanddemandelasticitiesas𝜂𝜂𝑖𝑖and𝜀𝜀𝑖𝑖definedinthestandardway:𝜂𝜂𝑖𝑖=𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑝𝑝𝑞𝑞𝑖𝑖,𝑆𝑆⇒𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)=𝜂𝜂𝑖𝑖𝑞𝑞𝑖𝑖,𝑆𝑆𝑝𝑝𝜀𝜀𝑖𝑖=𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)𝑝𝑝𝑞𝑞𝑖𝑖,𝐷𝐷⇒𝑞𝑞𝑖𝑖,𝐷𝐷′(𝑝𝑝)=𝜀𝜀𝑖𝑖𝑞𝑞𝑖𝑖,𝐷𝐷𝑝𝑝.Pluggingthesetwointoequation(3)andfurthermultiplyinganddividingbytotalglobalsupplyanddemand(whicharethesame,𝑄𝑄𝑆𝑆=−𝑞𝑞�+∑𝑞𝑞𝑖𝑖,𝑆𝑆(𝑝𝑝)𝑁𝑁𝑖𝑖=1=∑𝑞𝑞𝑖𝑖,𝐷𝐷(𝑝𝑝)𝑁𝑁𝑖𝑖=1=𝑄𝑄𝐷𝐷=𝑄𝑄)yields𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=∑𝑤𝑤𝑖𝑖,𝐷𝐷𝜀𝜀𝑖𝑖𝑖𝑖∑𝑤𝑤𝑖𝑖,𝑆𝑆𝜂𝜂𝑖𝑖𝑖𝑖−∑𝑤𝑤𝑖𝑖,𝐷𝐷𝜀𝜀𝑖𝑖𝑖𝑖=𝜀𝜀𝜂𝜂−𝜀𝜀<0Where𝑤𝑤𝑖𝑖,𝐷𝐷=𝑞𝑞𝑖𝑖,𝐷𝐷/𝑄𝑄and𝑤𝑤𝑖𝑖,𝑆𝑆=𝑞𝑞𝑖𝑖,𝑆𝑆/𝑄𝑄areweightsrepresentingregion𝑖𝑖’sshareofglobaldemandandsupply,and𝜀𝜀and𝜂𝜂areglobalweightedaveragedemandandsupplyelasticities.Intuitively,dividingbythedenominatorofequation(4),1(𝜂𝜂−𝜀𝜀)⁄,representstheeffectofcurtailedsupplyonprices,whichislargestwhensupplyanddemandareinelastic(smallvaluesof𝜂𝜂and𝜀𝜀).Toderivetheeffectonconsumption,thisismultipliedbythecurvatureofthedemandcurve,𝜀𝜀.Notethatasaratiooftwoelasticityvalues,𝜀𝜀̅(𝜂𝜂−𝜀𝜀)⁄issomewherebetweenzeroand−1,indicatingthatonlyaportionoftheonebarrelofreducedsupplyisabsorbedbylowerdemand.Theremainderismadeupasleakagethroughthesupplyresponse.Thatis,thechangeinglobalconsumptionperbarrelcurtailedissimplyequaltothedirecteffect(minusonebarrel)offsetbytheleakage(plus𝐿𝐿barrels):𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=−1+𝐿𝐿(3)(4)EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions8Wecanusethistosolvefor𝐿𝐿algebraicallyas𝐿𝐿=𝜂𝜂𝜂𝜂−𝜀𝜀.Equation(5)impliesthatleakageisbetweenzeroandone,giventherequiredsignsoftheelasticities.Itiszerowhen𝜂𝜂=0(perfectlyinelasticsupply)or𝜀𝜀→−∞(perfectlyelasticdemand),anditisonewhen𝜂𝜂→∞(perfectlyelasticsupply)or𝜀𝜀=0(perfectlyinelasticdemand),althoughneithersituationisplausible.Whentheelasticitiesareofaboutthesamemagnitude,leakageisabout50percent.Whensupplyismoreelasticthandemand,themarketleakagerateisgreaterthan50percent.Theanalysishasthusfarbeenlimitedtoimpactsonsupplyanddemandofoil,withoutconsiderationofGHGemissions.Wedenotetheemissionsintensityofthecurtailedoilas𝑒𝑒𝑞𝑞�andthatofoilproducedinregion𝑖𝑖as𝑒𝑒𝑖𝑖(e.g.,inlife-cycletonsofCO2eperbarrelofoil).7Thentheimpactofanincrementalbarrelofcurtailedsupplyongloballife-cycleemissionsfromoil,definedas𝐸𝐸=∑𝐸𝐸𝑖𝑖𝑖𝑖=−𝑒𝑒𝑞𝑞�𝑞𝑞�+∑𝑒𝑒𝑖𝑖𝑞𝑞𝑖𝑖,𝑆𝑆(𝑝𝑝)𝑖𝑖,isgivenby:𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=−𝑒𝑒𝑞𝑞�+�𝑒𝑒𝑖𝑖𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�𝑖𝑖.Thisreflectsthereducedlife-cycleemissionsfromthecurtailedsource,𝑒𝑒𝑞𝑞�,offsetbytheemissions,𝑒𝑒𝑖𝑖,associatedwiththeincreasedquantitysuppliedfromeachofthe𝑖𝑖=1,…,𝑁𝑁regions,𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝),inducedbythechangeinthepriceofoil,𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�⁄.Further,equation(6)canbesimplifiedtobewrittenasafunctionofthreeparameters,𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=−𝑒𝑒𝑞𝑞�+𝑒𝑒𝐿𝐿,7The𝑒𝑒𝑖𝑖termrepresentsfulllife-cycleemissions,includingScope1,2,and3emissions.Theemissionsintensityofoilvariesprimarilytoupstreamandmidstreamsourcesofemissions,notdownstreamcombustionemissions.Thiscouldchangeifcarboncapturetechnologyisdeployedatscaleforoiluse.Weviewthisasunlikelyintheforeseeablefuture,ascarboncapturetechnologyhasprimarilyfocusedonabatingemissionsfromcoalandgasatpowerplants,ratherthanfromoilemissions,suchasthosefromvehicles.(5)(6)(7)ResourcesfortheFuture9where𝑒𝑒isthegloballyweightedaverageemissionsintensityofmarginalsupply.8Theweightsaretherelativecontributionsofeachproducingregiontomarginalsupply—thatis,therelativeslopesoftheirsupplycurves,asfollows:9𝑒𝑒=1∑𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑁𝑁𝑖𝑖=1�𝑒𝑒𝑖𝑖𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑖𝑖Equation(7)isthekeyequationforestimatingtheemissionsreductionsachievedperbarrelofoilcurtailed.Itskeyinputsare1.𝑒𝑒𝑞𝑞�:theemissionsintensityofthecurtailedoilsupply,2.𝑒𝑒:theweightedaverageemissionsintensityofmarginaloilsupply,and3.𝐿𝐿:themarketleakagerateofoil.Allthreeinputvariablesareinherentlyuncertain.Theleakagerate,𝐿𝐿,dependsonrelativeelasticitiesofglobalsupplyanddemandintheoilmarket.10Theemissionsintensityofcurtailedsupplyisprobablyeasiestvaluetoestimate,butisstillsubjecttouncertaintyduetoimperfectmeasurementofGHGemissions,inparticularupstreammethaneemissionsfromoilandgasinfrastructure.Theemissionsintensityofmarginaloilsupply,𝑒𝑒,isperhapsthemostdifficulttoestimatewithhighprecisionbecausedoingsoentailsnotonlyhavingemissionsintensityvaluesforallsourcesofmarginalsupply,butalsohavingestimatesofsource-specificsupplyelasticities.Anadditionaldifficultyisestimatingsupplyelasticitiesexante,beforeundevelopedfieldscomeonline.Inthesubsequentsections,wediscusspracticalapproachestochoosingparametervaluesforelasticitiesandemissionsintensityvaluesinequations(7)and(8).8Equation(7)canalsobewrittenequivalentlyas𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=−𝑒𝑒𝑞𝑞�(1−𝐿𝐿)+�𝑒𝑒−𝑒𝑒𝑞𝑞��𝐿𝐿.Thisformoftheequationseparatestheemissionsreductionowingtoreducedoilconsumptionnetofleakage,−𝑒𝑒𝑞𝑞�(1−𝐿𝐿),andtherelativeemissionsintensityofthereplacedbarrelsofconsumption�𝑒𝑒−𝑒𝑒𝑞𝑞��𝐿𝐿.9Wecanseethisisthecorrectvaluebecausepluggingthisexpressionintothepreviousequationyieldsthedefinitionoftheemissionsimpact:𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=−𝑒𝑒𝑞𝑞�+∑𝑒𝑒𝑖𝑖𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�𝑖𝑖,notingthat𝐿𝐿=∑𝑞𝑞𝑖𝑖,𝑆𝑆′(𝑝𝑝)𝑁𝑁𝑖𝑖=1𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�.10Thismodelisstatic,butinadynamicmodelparalleltothatofPrest(2022b),asimilarresultisobtainedwheretheelasticitiesarelong-runvalues(specifically,aintertemporallong-runaverage,seeAppendix).(8)EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions102.1.SubstitutiontoOtherFormsofEnergyTheemissionsimpactofcurtailedoildemandgivenbyequation(7)representsreductionsinemissionsassociatedwithoilconsumption.Forthistorepresentacompleteaccountingofglobalemissions,thereductioninoilconsumptionmustrepresentfullfossilfueldemanddestruction(suchasimprovedfueleconomyorreducedvehiclemilestraveled)inadditiontosubstitutiontoalternativefuelsources(suchasEVs)thatmayentailtheirownemissions.Becauseoilisprimarilyusedasatransportationfuel,fewalternativeshaveexistedhistorically,meaningthatinthepast,price-inducedchangesinoilconsumptionlargelyreflecttruedemanddestruction,indicatingemissionsestimatescalculatedusingequation(7)basedonhistoricalparametersareindeedappropriate.However,iftheelasticitiesusedtocalibrateequation(7)begintoreflectfuturechangesintheavailabilityofoilsubstituteswithnon-zeroemissions,suchasEVs,thenequation(7)couldfailtoaccountfortheemissionsintensityofthosesubstitutes.Asathoughtexperiment,supposewidespreadadoptionofEVsmakesoildemandperfectlyelastic(𝜀𝜀→−∞),implyingzeroleakage(𝐿𝐿=0).Inthiscaseequation(7)reducesto−𝑒𝑒𝑞𝑞�,meaningone-for-onereductioninoilconsumptionfromthecurtailedsupply.Butthisfailstoaccountfortheemissionsfromtheelectricitypoweringthosevehicles,sotheemissionsreductionsfromequation(7)arecorrespondinglyoverstated.Moregenerally,abarrelofcurtailedsupplyleadsto(1−𝐿𝐿)barrelsofreducedoilconsumption.Denote𝑠𝑠astheshareofthatreductionthatistruefossilfueldemanddestruction,but(1−𝑠𝑠)issubstitutedtootherfuelsourceswithemissionsintensitydenoted𝑒𝑒𝑠𝑠,whichisinunitsoftonsofCO2eperbarrelofoilequivalent.Then,acompleteaccountingoftotalemissionswouldrequireadding(1−𝐿𝐿)(1−𝑠𝑠)𝑒𝑒𝑠𝑠toequation(7).Thatis,foreachbarrelofreducedconsumption,representedby(1−𝐿𝐿),anamountequalto(1−𝑠𝑠)ofthatbarrelissubstitutedtoanon-zeroemissionssource,withcorrespondingemissionrate𝑒𝑒𝑠𝑠.Thisadditionaltermislikelytobesmall.Inthehistoricalexperience,therehavebeenfewsubstitutesforoilconsumptionintransportation,meaning𝑠𝑠haslikelybeenclosetoone.Inourempiricalapplication,webasedemandelasticitiesonthehistoricalliterature,meaningourleakageestimatesshouldlargelyreflecttruefossilfueldemanddestruction.Inotherwords,if𝑠𝑠isclosetoone,thentheadditional(1−𝐿𝐿)(1−𝑠𝑠)𝑒𝑒𝑠𝑠termiseffectivelyzeroandneednotbeconsideredinthecalculation,whichcorrespondstotheassumptionimplicitinequation(7).However,thismaychangeinthefuture—forexample,withincreasedadoptionofEVsorothersubstitutesforoil.Suchadevelopmentwould,ononehand,reduceleakagebyintroducinganotherchannelbywhichdemandcanrespond(largeremissionsreductionsduetolessleakageinequation(7)),but,ontheotherhand,introduceemissionsfromsubstitutedfuel(smalleremissionsreductions).ResourcesfortheFuture11Theimportanceofsuchsubstitutionwoulddependontheemissionsintensityofthesubstitutefuel,𝑒𝑒𝑠𝑠.Giventhatoilsubstituteswillalmostuniformlyhaveloweremissionsintensitiesthanoil(e.g.,EVs,heatpumps),theneteffectofaccountingfornewsubstitutesovertimewillalmostsurelyyieldgreateremissionsreductionsthanimpliedbyequation(7)calibratedtohistoricaldata.Asanexample,passengerelectricvehicleshaveaper-mileemissionsintensityofroughlyone-fourthofthatofgasolinevehiclesatthecurrentemissionsintensityoftheUSelectricgrid.11Thissuggestsavalueof𝑒𝑒𝑠𝑠ofapproximatelyone-fourthofthevalueforoil,althoughthisvaluemayvaryacrossspaceandtime,particularlyiftheemissionsintensityofelectricityisdeclinesinthefuture,orifweseeimprovementsinEVenergyefficiency.Thissuggeststhisadditionaltermislikelytobesmallinthenear-term,atleastuntilbreakthroughsinpolicy,economics,ortechnologyrelevanttooilsubstitutesemergeinthefuture.Onnet,however,suchdevelopmentswouldleadtogreateremissionsreductionsthanimpliedby(7),asthereductioninleakagewouldbelargerthantheemissionsproducedbyEVs.Onecouldalternativelymitigateconcernsaboutemissionssubstitutionbysimultaneouslycurtailingboththesupplyanddemandforoilintandeminequalmeasure,asdiscussedbyPrest(2022a).Whilesuchanapproachmaybepromisingasapolicyeffort—forexample,byincreasingthestringencyoffueleconomyregulations—therearenotobviouschannelsforprivateactorstodrivereductionsinoildemandforthepurposeofgeneratingoffsets.Further,anaccurateaccountingofthoseeffectswouldrequirematchingthecharacteristicsofthecurtailedsupplyanddemand(e.g.,product-specificelasticitiesandemissionsintensities).Ifthosecharacteristicsdiffer,onewouldthenneedtoengageinacomplicatedaccountingadjustingforthosedifferences,whichisnotastraightforwardtask.2.2.WhenCouldLeakageBeZero?TheCaseofHotellingDynamicsAneconomicmodelfrequentlyusedtostudyresourceextractionistheHotellingmodel(Hotelling1931).IntheHotellingmodel,afixedamountofanexhaustibleresourceisextractedovertime,andthesupplierseekstooptimizethetimingofthatextraction.Inmostversionsofthemodel,alloftheresourceisextractedeventually,implyingthatchangesinpricescanonlyalterthetimeprofileofproduction.Themodelinthispaper,bycontrast,doesnotfeaturesuchclassicalHotelling-styledynamics,whichweargueisjustifiedbytheirweakempiricalsupport(Krautkraemer1998;SladeandThille2009;CairnsandSmith2019;Cairns,Davis,andSmith2021).Forexample,CairnsandDavis(2019)arguethatHotelling-stylemodels“yieldunsoundresultsthatshouldnotbeusedforpolicyevaluation.”Nonetheless,giventheirlonghistory,itisworthconsideringwhattheconsequencesofaHotellingmodelwouldbeforleakage.11E.g.,110gramsofCO2/milefora2023TeslaModel3,versus410grams/mileforanaveragenewgasolinevehicle.https://www.fueleconomy.gov/feg/Find.do?year=2023&vehicleId=46016&zipCode=20036&action=bt3EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions12Inourmodel,leakageoccursbecausecurtailedsupplyleadstohigheroilprices,whichinducesmoreproductionfromothersources.HowdoesthisstorychangeifthoseothersourcesarebelievedtoexhibitHotellingdynamics?SinceallofthedepletableresourceisextractedeventuallyinaHotellingmodel,changesinpricescanonlyalterthetimeprofileofproduction.Thismeansthatpricechangesinducedbycurtailedsupplymayalterthetimingbutnotthecumulativeamountofproductionfromothersources.Thiswouldimplyzeroleakageinthelongrun.12Thisextremeoutcomeseemsunrealistic,assuchamodelassumesthatnofossilfuelsareleftinthegroundinthelongrun,despitepolicyandtechnologicaldevelopmentspushingagainstsuchanoutcome.Ifsufficientlycheapalternativeenergysourcesareeventuallydevelopedthatpriceoutreserves,somefossilfuelswillbeleftintheground,withtheiramountdeterminedinlargepartbythepricemechanismsincludedthispaper’smodel—thatis,whichdeterminingreservesarepricedoutofthemarket.Insummary,Hotelling-styledynamicshaveweakempiricalsupport,haveunrealisticimplicationsforleakage,andareunlikelytochangethefundamentalmechanismsinthispaper’smodel.Seesection3.2inPrest(2022b)foralongerdiscussion.3.EstimatingMarketLeakageAsshownabove,marketleakage—theshareofcurtailedproductionthatisreplacedbyothersourcesofsupply—isdeterminedbytherelativeelasticitiesofsupplyanddemand.Thoseelasticitiesarenotknownforcertainbecauseofepistemicuncertaintyaboutoilmarketsastheyexisttodayandbecausethoseelasticitiesarelikelytochangeovertime.However,reasonablerangesandcentralestimatesofthoseelasticitiescanbeassembledfromquantitativemodelingfromtheacademicliteratureandothersources.Inthissection,wefirstpresentsomegeneralprinciplesfordeterminingappropriatevaluesforsupplyanddemandelasticities.Then,wepresentareviewofelasticityestimatesfromtheacademicliteraturethatmeetthosecriteria.3.1.GeneralPrinciplesforEstimatingLeakageParametersFourkeyquestionsarisewhenconsideringappropriateelasticityvaluestouseinequation(5):geographicscale,temporalscale,estimationmethodologyandcredibility,andfrequencyofupdate.Wediscussthemeachinturn.12Further,eveniftheHotellingmodelweretoholdandhigherpricesmerelyaccelerateextraction,thisstillleadstoleakageintheneartermandtheconsequentundesirablenear-termaccelerationinglobaltemperaturerise.ResourcesfortheFuture13Geographicscale.Themarketforcrudeoilisglobalinnature,meaningtherelevantelasticitiesforequation(5)arethereforegloballyrepresentativevalues.13Thus,totheextentpermittedbytheavailableevidence,analystsshouldendeavortouseonlygloballyrepresentativevalues.Whilemanystudiesestimateelasticitiesforspecificregions,thesemaynotnecessarilyrepresenttheglobalaverageiftheregioninquestionhasmoreorlesselasticsupplyordemand.ThisisparticularlyimportantgivenhowtheUSshaleboomincreasedprice-responsivenessthere(MasonandRoberts2018;R.G.Newell,Prest,andVissing2019;R.G.NewellandPrest2019;GilbertandRoberts2020),suggestingsupplyelasticitiesestimatedfortheUnitedStateslikelyexceedtheglobalaverageandarethusinappropriatetousedirectlyinequation(5).Temporalscale.Supplyanddemandelasticitiesaretypicallydifferentiatedbetweenshort-runandlong-runvalues.Short-runelasticitiestypicallyrefertoresponsesonthetimehorizonofafewmonthsorless.Estimatesofshort-runelasticitiestendtobesmallandoftenindistinguishablefromzerogiventhelackoftimeformarketparticipantstoadjustbehavior.Long-runelasticitiesnaturallytendtobelargerandoftenmuchlarger.Thestaticmodelaboveintentionallydoesnotmakeanexplicitassumptionabouttherelevanttimeframe,butitcanbethoughtofasrepresentingalengthoftimeoverwhichoildemandissufficientlyfungible,inthesensethatabarrelcanbeshiftedfromonepointintimetoanotherthroughstorage.Assubstantialvolumesofphysicaloilstorageexist,thiscanbethoughtofasafairlylongtimeperiod.Theimplicationisthattheelasticitiesinequation(5)shouldsimilarlyrepresentlong-runelasticities.Inaddition,thekeyleakageresultinequation(5)generalizestoamodelthataccountsforintertemporaldynamics,wheretheappropriateelasticityvaluesarelong-runweightedaverages(seeAppendix).Forthesereasons,long-runelasticitiesaretheappropriatevaluestouseinleakagecalculations.Estimationmethodologyandcredibility.Awidevarietyofmethodshavebeenappliedtoestimateoilsupplyanddemandelasticities.Theseincludetime-serieseconometricapproaches(KilianandMurphy2014),microeconometricmethods(Levin,Lewis,andWolak2017;Coglianeseetal.2017),andstructuralmodelling(BodensteinandGuerrieri2011;BalkeandBrown2018).Ingeneral,the“credibilityrevolution”ineconomicsoverthepastseveraldecadescouldsuggestplacinggreaterweightonmorerecentstudiesthatincorporaterecentadvancesineconometricmethods.However,doingsoinaquantitativesenseremainssubjective.Credibilityoftheresultingestimatesisalsoimportant,assomeindividualstudiesmayproduceimplausibleestimates,suchasestimatesoftheincorrectsign(e.g.,adownwardslopingsupplycurve)duetoidiosyncrasiesinastudy’sassumptions,methods,ordata.Sincesuchoutlierscanskewtheresults,somedegreeofexpertjudgmentisnecessarytoremoveimplausibleestimatesfromconsideration,butcautioniswarrantedwhendoingso,withageneralpreferencefora“lighttouch”review.13Thatis,appropriatelyproduction-weightedaveragesupplyelasticities,orconsumption-weighteddemandelasticities.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions14Ideally,commissionsofexpertswouldbeconvenedtobuildconsensusonreasonablesourcesofestimates,includingrecommendationsaroundbestpractices.Thereisprecedentforthisinothersettings,suchastheEnvironmentalProtectionAgency’s(EPA)commissionontheassessmentofcontingentvaluationmethods(Cummings,Brookshire,andSchulze1986)andbytheNationalAcademiesofSciences,Engineering,andMedicine(e.g.,NASEM(2017),whichwediscussinmoredetailbelow).Carboncreditregistriesmayopttoconvenetheirownsuchpanelstoestablishguidelinesforevaluatingestimates.Shortofthis,apracticalapproachcouldentailarecurringreviewoftheliteratureonelasticitiesofoilsupplyanddemand,aswedidinthisstudy,toreflectnewresearch.Frequencyofupdate.Asabove,elasticityestimatesshouldreflectlong-runvaluesbasedonthebestscientificunderstandingoftheshapeoffutureoilmarkets.However,thatbestunderstandingmaychangeovertimeformanyreasons.Newresearchmayproducebetterestimatesofsupplyanddemandelasticities.Developmentsinpolicyortechnologymayalsoaltertheexpectedtrajectoriesofthoseelasticities.Asonesimpleexample,policydevelopmentsmayacceleratethetransitiontoEVs,implyinglargerdemandelasticitiesthanpreviouslyanticipated.Similarly,theemissionsintensityofvariouskindsofoilcouldalsochangeovertimegiven,forexample,theGlobalMethanePledgeannouncedin2021.14Thismayyieldmoreorlessemissionsreductionsthanweoriginallyestimated,leadingtounder-orover-crediting.Therefore,amechanismisneededtoupdateelasticityestimatesandemissionsdataovertimeandadjustcreditingaccordingly.Thisraisesthequestionofhowfrequentlyoneshouldupdatetheparametervaluestoreflectachangingevidencebaseandpolicylandscape.Insomecases,theremaybeclearbreakthroughsinpolicy,economics,ortechnologythatsuggestrevisitingkeyestimates.Oneexampleinadifferentsetting—electricity—isthe2022passageoftheInflationReductionActintheUnitedStates,whichpromptedvariousmodelinggroupstoupdatetheirmodelstoreflectnewpolicies.15ShouldEVadoptionbecomeconsiderablymoreenticingthrough,forexample,lowercoststopurchaseandoperateorimprovedperformance,EVswillbecomeamoreviablesubstituteforgasoline-poweredvehicles.This,inturn,wouldmakedemandforoilmoreelastic,therebyreducingleakageandimplyingpreviousestimatesofemissionsreductionswereconservativelyunderstated.Ontheotherhand,arevolutioninindustrystructure,suchastheshalerevolutionorgrowthofnon-OPECsupplyfollowingthespikesinoilpricesinthe1970s,wouldmakesupplymoreresponsivethanoriginallyanticipated,implyingmoreleakagethanestimated.Eventslikethesewouldneedtobequitelargetowarrantarevisionofleakageestimates,andsowearguethattheywouldbereadilyapparentiftheycometopass.14https://ec.europa.eu/commission/presscorner/detail/en/statement_21_576615See,e.g.,https://www.rff.org/events/rff-live/future-generation-exploring-the-new-baseline-for-electricity-in-the-presence-of-the-inflation-reduction-act/ResourcesfortheFuture15Absentaclearchangesuchasthis,itwouldbesensibletoestablisharoutineforrevisitingelasticityandemissionsintensityestimatesonsomepredeterminedrecurringbasis.Thereisatrade-offbetweenupdatingestimatesinresponsetoeverynewdevelopmentwhilealsoprovidingapredictableandthoroughprocessforthoseupdates.A2017reportbytheNationalAcademiesofSciences,Engineering,andMedicine(NASEM2017)addressedthisissueinthecontextofrecurringupdatestotheUSgovernment’sofficialestimatesofthesocialcostofgreenhousegases.Whileitisnotaperfectparalleltothequestionofhowoftentoupdatetheparametersunderlyingleakageestimates,aconsiderationofthegeneralprinciplesisnonethelessinstructive.TheNASEMreportrecommendedanupdatecycleofapproximatelyfiveyears,whichprovidesenoughtimeforthedevelopmentofevolvingresearch,aswellasathoroughsynthesisofit.TheproposedNASEMcycleinvolvesthreesteps,thefirstofwhichinvolvesthetechnicalprocessofcomprehensiveupdatestothemodelingprocessthatincorporatesrapidlygrowingareasofresearchinclimatescienceandimpacts.16Duetothehighlycomplex,multidisciplinarymodelingneedsunderlyingthesocialcostofcarbon,thatfirststepwasenvisionedtotaketwotothreeyearstocomplete.Thesecondstepentailsobtaininginputandcommentontheproposedestimatesfromscientificandtechnicalcommunitiesandstakeholders,whichwouldbeincorporatedintoafinalizedestimate.Thisstepisenvisionedtotakesixmonthstooneyear.Finally,thethirdstep“involvesathoroughindependentscientificassessmentofthe…estimationprocess,inordertotrackandassessnewscientificliteratureovertimeandmakerecommendationsforfutureimprovementsandresearch.”Turningbacktotheconsiderationsinupdatingtheparametersunderlyingleakageestimates,atimeframeofnomorethanfiveyearsseemssimilarlyappropriatealbeitduetotwooffsettingconsiderations.Ontheonehand,theNASEM’sprocessmaybeoverlyelaborateinthecontextofleakagebecausetherearefeweruncertaintiesandparametersinvolvedinleakageestimationasopposedtothesocialcostofcarbon,whichdependsonmorecomplexmodelsentailingthousandsofuncertainparametersandassumptionsspanningmanydisciplines.Thatconsiderationwouldsuggestasimplerandperhapsmorefrequentupdateprocessforleakageestimation.Ontheotherhand,partoftheNASEM’smotivationreflectedthefactthatclimateimpactsliteraturehasbeengrowingrapidly,suggestingmorefrequentupdatesarenecessarytokeeppace.Bycontrast,studiesestimatingsupplyanddemandelasticitiesforoilhavenotexperiencedthesamepaceofgrowth,andtheirestimateshavenotchangedrapidlyovertime,suggestinglessfrequentupdates.Onnet,thesetwoconsiderationssuggestthatanupdatecycleoccurringnolessfrequentlythaneveryfiveyearsmaybeappropriate,althoughunexpecteddevelopmentsthataltersubstantiallyoilmarketsand/oremissionsintensitiescouldwarrantmorefrequentreexaminations.16NASEMenvisionedthatthisupdateprocesswouldbeledbyaninteragencyworkinggrouphousedintheUSfederalgovernment,sincetheusecasefortheestimateswasforgovernmentanalyses.Thislogicalsoholdstotheextentleakageestimatesaresimilarlybeingusedingovernmentanalyses,suchasbytheDepartmentofInterior’sanalysisofoilandgasleasingdecisions,orinregulatoryoffsetprograms.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions163.2.EstimatesofDemandElasticitiesThereisalargeliteratureineconomicsestimatingelasticitiesofdemandforoilandgasoline,butnotallestimatesareusefulinestimatingleakage.First,manystudies,particularlyinthoseusingstructuralvectorautoregressionmethods,focusonestimatingveryshort-runelasticities,whicharenotappropriateforlong-runleakagecalculations.Further,manyempiricalestimatesestimatethepriceelasticityofdemandforgasoline,ratherthanforcrudeoil,whicharerelatedbutneverthelessdistinct.AsnotedbyHamilton(2009),becausecrudeoilamountstoabouthalfoftheretailcostofgasoline,thepriceelasticityofcrudeoildemandshouldbeabouthalfasbigastheelasticityforgasoline.Forthisreason,elasticitiesestimatedforgasolinemustbedividedbytwotoconverttoacrudeoilelasticityappropriateforuseinequation(5).Inlightoftheseconsiderations,weconductedareviewoftheeconomicliteratureestimatingoilorgasolinedemandelasticities.Wegavepreferencetostudiesthatwere1)peerreviewedorotherwisefromanauthoritativeacademicsource,2)werenotsimplycitationsofotherpapers,and3)presentedgloballyrepresentativeelasticities,althoughwealsoincludedrigorousstudiesestimatingelasticitiesforcountriesthatarelargeconsumersofoil,suchastheUnitedStates.Thisyieldedmorethan30studiesthatprovideddemandelasticityestimates,fromwhichweexcludedstudiesthatonlypresentveryshort-runelasticities(typicallywithatimehorizonofonemonth).Fromeachoftheremaining25studies,weextractedasinglecentralestimate17toavoidgivinggreaterweighttostudiesthatreportmultiplevalues.Whenarangeofestimateswaspresented,wetooktheaverage.Whereappropriate,weconvertedgasolineelasticitiestocrudeoilelasticitiesbydividingbytwo.Inoneinstance,astudyreportedelasticitiesforOECDandnon-OECDdemand,fromwhichwecalculatedaglobalconsumption-weightedaveragevalueassuminga53percentOECDconsumptionshare.18Fromtheseresulting25estimates,weremovedthreeimplausiblysmallandimplausiblylargeestimatesof-0.04,-0.05,-1.21,eachofwhichdiffersfromtheclosestremainingestimatebymorethan70percent.Theremaining22estimatesareshowninTable1inorderofpublicationyear.Theseelasticitiesrangefrom-0.12(Serletis,Timilsina,andVasetsky2010)to-0.72(Brown1998),withasimpleaveragevalueof-0.35andastandarddeviationof0.15.Morethanhalf(59percent)ofthestudieshavebeenpublishedsince2010.17Inprinciple,wewouldliketoincorporatenotonlycentralestimatesbutalsouncertaintyrangesthrough,forexample,Bayesianaveragingacrossestimates.Inpractice,thisnotpossiblebecausemanystudiesdonotreportstandarderrors.18https://www.eia.gov/finance/markets/crudeoil/demand-oecd.phpResourcesfortheFuture17Table1.EstimatesofthePriceElasticityofDemandforCrudeOilStudyCentralPeer-reviewed?DahlandSterner(1991)-0.43YesHausmanandNewey(1995)-0.40YesBrown(1998)-0.72No(FederalReserveBankofDallasreport)YatchewandNo(2001)-0.45YesGatelyandHuntington(2002)-0.42YesGrahamandGlaister(2002)-0.39YesCooper(2003)-0.32YesGoodwin,Dargay,andHanly(2004)-0.32YesBrons,Nijkamp,Pels,andRietveld(2008)-0.42YesSerletis,Timilsina,andVasetsky(2010)-0.12YesBodensteinandGuerrieri(2011)-0.42No(FederalReservediscussionpaper)Dahl(2012)-0.32YesLinandZeng(2013)-0.17YesBrown,Mason,Krupnick,andMares(2014)-0.45No(RFFreport)Dahl(2014)-0.38No(ColoradoSchoolofMinesworkingpaper)KilianandMurphy(2014)-0.26YesLevin,Lewis,andWolak(2017)-0.16YesCoglianese,Davis,Kilian,andStock(2017)-0.19YesKrupnick,Morgenstern,Balke,Brown,Herrera,andMohan(2017)-0.53No(RFFreport)BalkeandBrown(2018)-0.51YesHuntington(2019)-0.15YesKnittelandTanaka(2019)-0.19No(NBERworkingpaper)Simpleaverage-0.35Note:Indicateselasticityafterdividingbytwotoconvertfromgasolinetocrudeoilelasticities.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions183.3.EstimatesofSupplyElasticitiesWeconductedananalogousliteraturereviewtocollectestimatesforglobaloilsupplyelasticities.Thisliteratureisfarsmallerthanthatfordemandelasticities.Toavoidrelyingsolelyonaverysmallnumberofstudies,whichwouldgiveafalsesenseofconfidenceinthelikelyrangeofsupplyelasticities,weweremoreaccommodatingregardingacceptablestudiesforinclusioninourset(forexamplebyincludingestimatesbasedonsimulationmodels,suchasGreeneandLieby(2006)).ThisreviewresultedinnineestimatesshowninTable2,whichrangebetween0.25(Krichene2002)and0.55(BalkeandBrown2018),withasimpleaverageof0.42andastandarddeviationof0.10.Aswiththedemandelasticities,morethanhalf(55percent)ofthesestudieshavebeenpublishedsince2010.Table2.EstimatesofthePriceElasticityofSupplyofCrudeOilStudyCentralPeer-reviewed?Huntington(1994)0.40YesBrown(1998)0.43No(FederalReserveBankofDallasreport)Krichene(2002)0.25YesGreeneandLeiby(2006)0.46No(OakRidgeNationalLabmodeldocumentation)Coyle,DeBacker,andPrisinzano(2012)0.29YesBrown,Mason,Krupnick,andMares(2014)0.40No(RFFreport)Krupnick,Morgenstern,Balke,Brown,Herrera,andMohan(2017)0.51No(RFFreport)BalkeandBrown(2018)0.55YesPrest(2022b)0.47YesSimpleaverage0.42Usingthesimpleaveragevaluesoftheaboveestimatesfortheelasticitiesofsupply(0.42)anddemand(-0.35)givesafirst-orderapproximationoftheexpectedmarketleakagerategivenbyequation(5):𝐿𝐿≈0.420.42−(−0.35)=55%ResourcesfortheFuture19Whilethisapproximationdoesnotreflectthefullrangeofuncertaintyinsupplyanddemandelasticities,itremainsasimplebenchmark.Aformalassessmentofthecentralestimateoftheleakagerateanduncertaintyarounditrequiresanexplicittreatmentofuncertaintyintheseparametervalues.Inourquantitativeapplication,weundertakeaMonteCarloexerciseinwhichwesamplefromtheelasticitiesinTable1andTable2,aswellasuncertaintyinemissionsintensitiesoftheleakedproduction,whichwediscussnext.4.EstimatesofEmissionsIntensitiesWedrawupondatafromthe2023OilClimateIndexplusGas(OCI+)data,whichisadataproductdevelopedbyresearchersatRMI.TheOCI+isabottom-upsystemstoolthatalsousesinputtop-downmeasurements(ahybridapproach)toquantifyemissionsfromoilandgasproduction,processing,refining,shipping,andenduses.TheOCI+usesthreeunderlyingmodelstoassessGHGemissionsfromtheoilandgasvaluechainsegments.Theproductionmodel,OilProductionGreenhouseGasEmissionsEstimator(OPGEE)residesatStanfordUniversity.Therefiningmodel,PetroleumRefineryLife-CycleInventoryModel(PRELIM)residesattheUniversityofCalgary.Andtheenduseconsumptionmodel,OPEM(OilandGasProductsEmissionsModule)residesatRMI.TheOCI+anditsunderlyingmodelshavebeenpeer-reviewedandinternationallycitedandappliedinenergypolicydecisionmakingforoveradecade,aspartiallylistedintheOCI+webtool“Studies”tab.19Fordetailsoneachmodelandmodelinginputsandassumptions,refertotheOCI+Methodology.20TheOCI+providesannualfield-leveltime-seriesestimatesoflife-cycleGHGemissionsfor586oilandgasfieldsrepresentingtwo-thirdsofglobalsupplyfrom2015–2022.WhileOCI+hasdetailedestimatesofemissionsintensitiesbyproductandstageofproduction(upstream,midstream,downstream),weuseeachfield’stotallife-cycleemissionsintensity,whichismeasuredinkilogramsofcarbondioxideequivalentemissionsperbarreloilequivalent(kgCO2e/boe),presentingkeyresultsunderboth100-yearand20-yearglobalwarmingpotentials(GWPs).Werefertotheseestimatesasfield-specificemissionsintensityvalues,correspondingto𝑒𝑒𝑖𝑖valuesinequation(8).Whilewewillfocuson100-yearGWPsforourmainanalysis,wealsodemonstratethesensitivityofthemainresultstousinga20-yearGWP,whichgivesgreateremphasistomethane-intensivefields.Equation(8)alsorequiresinformationonoilproductiontoreflectthefactthatfieldsshouldbeweightedintheanalysisinproportiontotheirsize.Asoilsupplyisthefocusofthispaper,wefocussolelyonoilproductionasourmeasureoffieldsize,ratherthanonoilandgasproductiontogether.21However,theOCI+dataonlypresentscategoricaldataonfield-levelproduction,placingeachfieldintoaoneoffivebandsofcrudeoil19Seehttps://ociplus.rmi.org/about/studies.20Seehttps://ociplus.rmi.org/methodology.21Modelingthecurtailmentofgassupplyraisesotherissuesthatarebeyondthescopeofthispaper,suchastheinteractionbetweengasandcoaldemand.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions20production:“VeryLow”(0–5kb/d),“Low”(5–50kb/d),“Medium”(50–250kb/d),“High”(250–500kb/d),or“VeryHigh”(>500kb/d).Asweneedquantitativevalues,weapproximatefield-levelproductionasthemidpointofeachbin;forthe“VeryHigh”bandthathasnomidpoint,weuseavalueofonemillionbarrelsperday(twicethatband’slowpoint).Wedothiscalculationforeachfieldandyear,andthenaggregateproductionandemissionstothefieldlevelacrosstime.22Figure2depictstheper-barrelemissionsintensityestimatesundera100-yearGWP,orderedfromlowesttohighest,versusourapproximationoffield-levelproduction,withbarscoloredbyresourcetype.23Ingeneral,fieldswithheavieroiltendtohavehigherlife-cycleemissions,withtheCanadianoilsandsindicatedonthegraphasanexample,averaging648kgCO2e/boe,comparedtotheglobalaveragevalueof518kgCO2e/boe.ThePermianBasinisalsoindicatedonthegraph,withanaverageemissionsintensityof524kgCO2e/boethatisveryclosetotheglobalaverageunderthe100-yearGWP.TheweightedaverageemissionsintensitiesforselectedsetsoffieldsareshowninTable3;wediscussthesecategoriesinmoredetailinthenextsection.Whilethereisvariationinthelife-cycleemissionsintensity,thecurveFigure2isfairlyflatexceptaroundthehighestandlowestemissionsintensityfields,whichhighlightsthevalueoftargetingsupply-sideinterventionstowardsthemostemissivefields.Thisstabilityoweslargelytotwofactors.First,the100-yearGWPputslessemphasisonmethane-intensivefieldsthana20-yearGWPwould.Second,end-useemissions—whichincludeemissionsfromcombustionofthefinalproduct—accountforanaverageof75percentoftotallife-cycleemissionsundera100-yearGWP,24andend-useemissionsvarymuchlessacrossfieldsthandoproduction,shipping,andrefiningemissions.Inaddition,thereisevenfurthersimilarityinlife-cycleemissionsintensitieswithinaresourcetype,suchaslightoil,whichisimportantwhenconsideringthepossibilitythatcurtailedoilproductionmaybesubstitutedbyoilofasimilarquality.Figure3showsfield-levelemissionsintensitiesundera20-yearGWP,whichresultsinmorevariationinemissionsintensityduetothegreateremphasisonmethaneemissionsintensity,whichvariesmuchmoreacrossfieldsthanCO2emissionsintensitydoes.Thisbringsmanylightoilfields(graybarsinFigure3)thatareconsideredlow-emittingundera100-yearGWPtothehigherendoftheemissionsintensitycurve.Notably,witha20-yearGWPthePermianBasinisnowamongthemostemissions-intensiveoilplay,averaging678kgCO2e/boe,whichiswithin0.5%oftheemissionsintensityofCanadianoilsands,ascontrastedwithbeingaboutaverageundera100-yearGWP(seeTable3).Inthenextsection,wecombinethisOCI+datawiththeelasticityestimatesdiscussedabovetoconductaMonteCarloanalysisandassessemissionsreductionsgivenbyequation(7).22Weuseasimpleaverageoffield-levelproductionovertimeandproduction-weightedaverageoffield-levelemissionsintensity.23Forsimplicity,inthisfigureweaggregateOCI+’s10resourcetypesintofour(gasandlight/medium/heavyoil,wherecondensateisincludedinthelightoilcategory).24Theanalogousfigurefora20-yearGWPis68percent.ResourcesfortheFuture21Figure2.Field-levelLife-cycleGHGEmissionsIntensityEstimatesversusApproximateCumulativeProduction,100-YearGWPFigure3.Field-levelLife-cycleGHGEmissionsEstimatesversusApproximateCumulativeProduction,20-YearGWPEstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions22Table3.Life-cycleEmissionsIntensitiesofSelectedSetsofFieldsEmissionsintensity(kgCO2e/boe)Fieldtype100-yearGWP20-yearGWPAll(marketaverage)518580Oilsands648692PermianBasin524678Lightoil498568Mediumoil512565Heavyoil612666Sweetoil539604Souroil518576RegionOPEC507567NorthAmerica540614Central/SouthAmericaandCaribbean571632Europe476516NorthAfrica530621OtherAfrica532609MiddleEast487538RussiaandCentralAsia508576East/SoutheastAsia561612Notes:Allfiguresrepresentweightedaveragesfromfieldscorrespondingtothegivencategory,whereweuseourapproximationsofcrudeoilproductionasweights.ResourcesfortheFuture235.QuantitativeApplicationInthissection,wecombinetherangesofsupplyanddemandelasticitieswiththeOCI+dataonfield-levelemissionsintensitiestoconductaseriesofMonteCarlosimulationstoestimatethenetemissionsreductionsthatcouldbeachievedbycurtailingoilsupplyandhowthoseestimatesvarybythetypeofoilcurtailedandsubstituted.Thefirstsetofinputsaredistributionsofsupplyanddemandelasticities.Asdiscussedabove,theempiricalresearchonoilpriceelasticitiesofsupplyanddemandhasproducedarangeoflong-runestimates.Toempiricallyexplorethedegreeofsensitivityofemissionleakageunderthisrangeofelasticityestimatesandemissionintensitiesofthepotentiallyleakedoil,wefirstconductaMonteCarlofocusedontheleakagerate.Webeginbyformingadistributionofmarketleakagerates—thatis,equation(5)—bysamplingfromsupplyanddemandelasticities.Wecreatethisleakage-ratedistributionthroughtwodifferentsamplingmethods.Inthefirstmethod,wetake10,000draws,withreplacement,fromthelistofdemandandsupplyelasticitiesgiveninTable1andTable2,respectively.Witheachdraw𝑖𝑖ofan𝜀𝜀𝑖𝑖and𝜂𝜂𝑖𝑖value,weformaleakageratevalue,𝐿𝐿𝑖𝑖,basedonequation(5),resultingin10,000leakagerateestimates.Forthesecondmethod,wespecifyparameterizeddistributionsofsupplyanddemandelasticities.Specifically,forthesupplyanddemandelasticities,respectively,wecalculatethemeans(𝜂𝜂̅,𝜀𝜀̅)andstandarddeviations(𝜎𝜎𝜂𝜂,𝜎𝜎𝜀𝜀)basedontheindividualelasticityestimatesgiveninTables1and2.Wethenassumethatsupplyanddemandelasticitiesarebothdistributedastruncatednormaldistributionswiththerespectivesupplyanddemandelasticitydistributionsof𝑇𝑇𝑇𝑇(𝜂𝜂̅,𝜎𝜎𝜂𝜂,0,∞)and𝑇𝑇𝑇𝑇(𝜀𝜀̅,𝜎𝜎𝜀𝜀,−∞,0),witheachsetofelasticitiesconstrainedtohavetheappropriatesign.Inbothapproaches,weassumenocorrelationbetweensupplyanddemandelasticitiesbecausetheunderlyingfactorsdrivingthetwoarelargelyunrelated.Supplyelasticitiesaredrivenbythecoststructuresofoilproduction,whichareplausiblyunrelatedtothedriversofdemandelasticities—behavioralresponses(e.g.,changesinvehiclemilesdriven)andtheavailabilityofsubstitutes(e.g.,electricvehicles).Figure4showshistogramsofthetwoleakageratesunderthetwosamplingmethods.Thetwomethodsleadtosimilardistributionsofleakagerates,withalmostidenticalmeans(56.0percentversus55.7percent).25Giventhissimilarity,wepresentresultsbelowbasedonthe“samplingwithreplacement”oftheindividualelasticitiesapproachbecauseitmoreexplicitlyrepresentstheelasticityliterature,ratherthanapproximatesit.WhileFigure4showsmarketleakage,𝐿𝐿,asinequation(5),thatisonlyonecomponentoftheformulaforemissionsreductionsinequation(7).Theotherkeycomponentsaretheemissionsintensitiesofthecurtailedandsubstitutesources.25Theprimarydifferencebetweenthetwoapproachesisthatthetruncatednormalapproachproducesafatter-taileddistributionoftheleakagerate,owingtoextremedrawsofparameterizedelasticitydistributionsthatareoutsideoftherangeoftheelasticityestimatesfromtheliterature(e.g.,ademandelasticitythatiseffectivelyzero).EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions24Whilethecurtailedsourceofsupplyisgenerallyknown,thesourcesofsubstitutesupplyandtheiremissionsintensitiesaremoreuncertain.Becauseitisnotgenerallyfeasibletoempiricallyestimatethesourcesofsubstitutesupply(e.g.,itisgenerallynotpossibletoempiricallyestimatewithhistoricdatahowmuchagivenfieldXwillincreaseproductioniffieldYiscurtailed),wefocuswithaneutraldefaultassumptionthatsubstitutesupplycancomefromanyfieldinproportiontoitsproductionshare.Werefertothisasthe“marketaverage”casebecause,onaverage,thisapproachtreatssubstitutesupplyattheproduction-weightedaverageemissionsintensity.Wealsoassessthesensitivityofthisapproachtoalternativeassumptionsaboutpotentialsourcesofsubstitutesupply.Forexample,inonecaseweassumethatsubstitutesupplycomessolelyfromOPEC.Thesescenarioscouldrepresentaproxyforcoststructuresthatcouldleadsomesourcesofsupplytobemoreprice-responsivedueto,say,varyingcoststructures.Forexample,Caldara,Cavallo,andIacoviello(2019)findthatOPECisabouttwiceasprice-responsivethanaverage.26Similarly,onshoreNorthAmericansupplyisestimatedtobemoreprice-responsivethantheglobalaverage(R.G.NewellandPrest2019;R.G.Newell,Prest,andVissing2019;Prest2022b),andmanysuchfieldsalsohavebelow-averageemissionsintensitiesintheOCI+data.Acrossouralternativescenarios,wesimultaneouslyvarytwodimensions:theoilproductionthatiscurtailed,andthesourceofsubstitutesupply.Ineachcase,wedefineacategoryoffieldsunderconsiderationforcurtailmentandrandomlysample—10,000timeswithreplacement—afieldfromthatcategory(e.g.,CanadianoilsandsorPermianBasinfields),withsamplingweightsequaltotheirapproximatedproduction.Weseparatelydefineacategoryoffieldswheresubstituteproductionmayariseandanalogouslyrandomlysamplefromthosefields.Foreachsetof“curtailedsources”and“substitutesources,”thisyields10,000estimatesofcurtailedemissionsintensity𝑒𝑒𝑞𝑞�and“leaked”emissionsintensity𝑒𝑒𝑖𝑖,inwhichthesubstituteproductioncomesonlyfromfield𝑖𝑖.Thatis,ineachdraw,allfieldsexceptthesampledonearedroppedfromequation(8),resultingin𝑒𝑒=𝑒𝑒𝑖𝑖equation(7),andsimilarlysofor𝑒𝑒𝑞𝑞�.26ThatstudyisnotincludedinTable2becauseitonlyestimatesshort-runsupplyelasticities.ResourcesfortheFuture25Figure4.HistogramsofSimulatedMarketLeakageRatesunderTwoSamplingApproachesTheassumptionthatallsubstitutesupplycomesfromasinglerandomlysampledfieldisconservativefromthestandpointofuncertaintyinemissionsavoided.Thisincreasestheuncertaintyrangeofouremissionsestimatesrelativetoanalternativeassumptionthatallfieldsinthe“substitutionsources”categorycontributeasmallamounttothetotalleakedproduction.Thisalternativeassumptionwouldimplyusingtheweightedaverageemissionsintensitywithinacategoryfor𝑒𝑒̅in(7).Suchanapproachwouldyieldsimilarcentralestimatesbutanarroweruncertaintyrangethanthosepresentedbelow.Wepresentresultsusingthealternativeweighted-averageemissionsintensityapproachintheAppendix,showingthatthekeytakeawaysincentralestimatesandshareofsamplesyieldingnetemissionsreductionsarenonethelesssimilar,althoughthe95percentrangesarenarrower.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions26With10,000samplesofemissionsintensityvaluesforbothcurtailed(𝑒𝑒𝑞𝑞�)andsubstitutesupply(𝑒𝑒̅)inhand,wethencouplethemwith10,000independentdrawsoftheleakagerate𝐿𝐿showninthetoppanelofFigure4tocalculatethenetemissionschangeusingequation(7).Weconductthisexerciseforeachpairofcategoriesofcurtailedandsubstitutesupply.Wechoseoursetsof“curtailed”and“substitute”sourcesbasedonavarietyoffactors.Inthesimplestcase,werandomlysampleoneofthe586fieldsintheOCI+data,withalikelihoodinproportiontoitsapproximatedcrudeoilproduction,tobetheonethatiscurtailedorthatrampsupproductioninresponsetothecurtailedbarrel.Wecallthisthe“marketaverage”casebecauseinexpectationitcorrespondstocurtailedandsubstitutesupplyhavingamarket-averageemissionsintensity.Sincethecurtailedoilisachoicevariablefortheentityretiringoilassets,wefocusonkeyareasunderconsiderationsuchasNorthAmerica,includingthehighlyemissiveCanadianoilsandsandPermianBasin.Wealsoconsiderbroadcategoriesofoiltypes—light,medium,andheavy,orsourversussweet.Thisyieldsninecategoriesofsourcesofcurtailedsupply.Forpossiblesubstitutesources,weincludeallninecategoriesusedforthecurtailedsources,plusanadditionaleightregionaldimensionsbecausethemarginalsourceofsupplymaydependinsomemanneronphysicallocation(dueto,say,totheconnectednessofmarketsandenergyinfrastructure).Forinstance,theretirementofafieldinNorthAmericamaypromptmoreproductionfromotherNorthAmericanfieldsastheretirementmayopenmorepipelineorrailtransportationoptionsandperhapsmorelocalrefiningcapacity.WealsoincludeOPECasacategory,givenitshistoricalrelevanceasaswingsupplier.Inaddition,thetypeofoilanditsrefiningneedswouldalsoplayarole.Thiscouldoccurbecauserefiningcapacitytendstobegearedtowardscertaintypesofoil.Thus,theretirementofafieldthatproduces,forexample,heavyoilmayinducemoreheavyoilproduction,astherewouldbesomenewlycreatedrefiningcapacityforthattypeofoil.Altogether,werun1.53millionsimulations(reflecting10,000drawseachforninecategoriesofcurtailedsupplyand17categoriesofsubstitutesupply).WesummarizeasubsetofthesesimulationsinTable4,whichpresentstheexpectedvalue(average)ofthenetchangeinemissionsfromthisexerciseforeachoftheninecategoriesofcurtailedsourcesandthepercentofthe10,000drawsthatleadtonetemissionreductions.Forthistable,wefocusontwopotentialsourcesofsubstitutesupply:onecaseinwhichallsourcesarecandidatesforsubstitutesupply,andanotherinwhichallsubstituteproductionmustcomefromthesamecategoryofsourcesasthecurtailedsupply.TheprimarytakeawaysfromTable4arethat,first,acrossmanyregions,oiltypes,andtheGWPused,ourestimatesofaveragenetemissionreductionsfromthecurtailmentofabarrelofoilisgenerallyabout40–50percentoftheemissionsintensityoftheretiredbarreland,second,almostallofourMonteCarlodrawsleadtonetemissionreductions.ResourcesfortheFuture27Thefactthatvirtuallyallourdrawsleadtoemissionreductionshighlightsthefactthatwhilethereisconsiderablevariationinleakageparameters𝐿𝐿andcross-fieldemissionsintensity,thevariationinthecombinationofthoseparametersrarelyleadstoscenariosthatpredictemissionincreaseswiththecurtailmentofabarrel,andthisisrelativelyinsensitivetowhattypeofoilbarreliscurtailed.Thisresultisparticularlystrikinggivenourconservativeapproachofsamplingfield-levelemissionsintensities,whichlikelyoverstatestheuncertaintyrangesofnetemissionsimpacts.Finally,usinga20-yearGWPleadstomodestlylargerCO2ereductions,astheemissionsintensityofcurtailedsupplyishigher,butthiseffectissomewhatoffsetbymoreemissivesubstitutesupply.Exceptforthemethane-intensivePermianBasin,theneteffectofthesetwoforcesismodest.Forthisreason,fortheremainderofthepaperwefocusonresultsonlyusingthe100-yearGWP,althoughinterestedreaderscanseethefull20-yearGWPresultsintheAppendix.Table4.NetEmissionsChangeperCurtailedBarrelwhenSubstituteProduction’sEmissionsIntensityReflectsAllFieldsorSameasCurtailedSourceSubstituteproductionfrom:Allsources(marketaverage)SamesourcesAveragenetemissionschange(kgCO2e/boe)Shareofdrawswithemissionsreductions(%)Averagenetemissionschange(kgCO2e/boe)Shareofdrawswithemissionsreductions(%)Curtailedsource100-yearGWP20-yearGWP100-yearGWP20-yearGWP100-yearGWP20-yearGWP100-yearGWP20-yearGWPAll(marketaverage)-230-25798.7898.47-230-25798.7898.57NorthAmerica-249-28999.0899.12-238-27198.1498.73Oilsands-357-36799.9799.88-285-30599.9399.96PermianBasin-233-35399.3999.87-230-299100100Lightoil-208-24398.5798.13-220-25199.6298.60Mediumoil-220-23899.0598.48-224-24899.6299.25Heavyoil-322-34199.7999.69-270-29398.6698.97Sweetoil-248-27999.1498.88-237-26599.1198.41Souroil-227-25098.9398.45-228-25398.5798.20EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions28WepresentabroadersetofsubstitutionsourcesandtheresultingimplicationsforemissionreductionsfromthecurtailmentofabarrelinacertainareaoroiltypeinTables5,6,and7,allofwhichusea100-yearGWP.Table5showstheexpectedvalue(average)ofthenetchangeinemissionsfromthisexercise,withthecurtailedsourcesshownintherowsandthesubstitutesourcesincolumns.Importantly,ourMonteCarloexerciseallowsustoassessuncertaintyaroundthesecentralestimatesbyproviding95percentileranges,whichareshowninTable6(forthe97.5thpercentile)andTable7(forthe2.5thpercentile).Largernegativenumbersindicatelargeremissionsreductions,withthegreenshadingshowninproportiontothosereductions.Thecellinthefirstrowandcolumnshowsthe“marketaverage”case(notingthatthemarket-averageemissionsintensityis518kgCO2e/boe,seeTable3).Inthiscase,thecentralestimateforthenetreductioninemissionsis230kgCO2e/boe(95percentrange:37to453kgCO2e/boe),whichreflectsthereducednetemissionsofabout45percentofthataverageemissionsintensity.Thisisconsistentwitha55percentleakagerate(i.e.,onebarrelretirediscounteractedbya0.55-barrelincreaseinsupplyfromotherproducers).Table5.AverageNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),100-YearGWPResourcesfortheFuture29Turningtothesecondrow,curtailingNorthAmericanoilhasaslightlylargerimpactonemissions(249kgCO2e/boe)becauseNorthAmericanoilproductionisslightly(~4percent)moreemissions-intensivethanaverage.Thecurtailmentofoilsandsinthethirdrowshowsmuchlargeremissionsreductionsof357kgCO2e/boeinthefirstcolumn(95percentrange:178to523kgCO2e/boe),owingtoitshighemissionsintensity.However,ifthecurtailmentofabarrelfromanoilsandsfieldpromptsincreasesinproduction(leakage)fromotheroilsandsfields,theemissionsreductionsareasomewhatmoremodest285kgCO2e/boe(95percentrange:86to471).Forreference,theemissionsintensityofoilsandsisabout648kgCO2e/boe,suggestingnetemissionsreductionsofabouthalfofthegrossemissionsintensityofthecurtailedbarrel.Lookingacrossthecolumns,weseethatthenetemissionsreductionsarelargelyinsensitivetothesourceofsubstitutesupply,exceptinthefewcaseswherethatsubstitutesourcehasemissionsintensitiesthatareextremelyhigh(oilsands)orextremelylow(Europe).Thisresultowestotherelativeflatnessoftheemissionsintensitycurveandsuggeststhatthedefaultuseofa“market-average”emissionsintensityofsubstitutesupplyisareasonableapproximation,unlessthereisclearreasontothinkthatthecurtailmentofaspecificfieldislikelytodriveproductionbyaspecificsourceorregion,suchasduetoidiosyncraticmarketdynamicslikepipelineconstraints.Intermsofthesignoftheimpacts,ourcentralestimatesimplynetemissionsreductionsacrosstheboard.Whiletherangeofmagnitudesareindeedwide,thereislittleuncertaintyinthesignoftheimpactsonemissions,particularlyforhighlyemissions-intensivefields.Consideringthe,roughlyspeaking,“worst-casescenario”97.5thpercentileinTable6,theoilsandsrowuniformlyshowsnetreductionsinemissionsregardlessofthesubstitutesourceofsupply.Theworst-casescenariosdoshowthatinsomecasesthereisasmallchancecurtailmentofrelativelycleansupplycouldincreaseemissions.Thiscouldariseif1)curtailedsupplyisrelativelyclean,2)theleakagerateishigh(generallyabove75percent),and3)leakedproductioncomesfromrelativelyemissions-intensivesources(seethecellsinwhiteandredinTable6)However,thesescenariosgenerallyinvolvelessplausiblescenarios,suchasanassumptionthatcurtailmentoflightoilproductionisreplacedsolelybyoilsandsproduction,ratherthanothersourcesoflightoil.OurMonteCarloanalysisfindsnetemissionsreductionswithahighdegreeofcertaintywhenthecurtailedsourcesofsupplyarehighlyemissions-intensiveoillikeheavyoils(99.7percent)orCanadianoilsands(99.96percent).Acrossallofour1.53millionsimulations,wefindnetemissionsreductionsin98.75percentofcases.IntheAppendix,weshowasensitivityanalysisusingthealternativeapproachofusingweighted-averageemissionsintensityvaluesforeachcategoryofcurtailedandsubstitutesupplyinsteadoffield-levelsampling.Thatapproachyieldsessentiallyidenticalcentralestimatesandhighercertaintyofnetemissionsreductions;inthatcase,acrossall1.53millionsimulationsusingthe100-yearGWP,wefindnetreductionsin99.9percentofcases.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions30Table6.97.5thPercentileofNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),100-YearGWPTable7.2.5thPercentileofNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),100-YearGWPResourcesfortheFuture316.ConclusionSupply-sideinterventionsthatretirehighlyemittingfossilfuelassetshavereceivedincreasedattentionfrompolicymakersandprivateactorsalike.Yetconcernsaboutleakageofproductiontoothersourcesofsupplyhaveraisedquestionsabouthowmuchemissionsreductionstheycanachieve.Inthispaper,weestimatetheeffectsofthesesupply-sideinterventionsonglobalemissions,accountingforbothmarketleakageaswellastherelativeemissionsintensityofdifferentsourcesofsupply.WeaccountforuncertaintyinsuchleakageratesandtheemissionsintensitiesofthecurtailedandsubstitutesourcesofsupplythroughaMonteCarloanalysis,drawingonkeysupplyanddemandelasticitiesfromtheeconomicsliteratureandemissionsintensityestimatesfromthestate-of-the-artOCI+dataseton586oilandgasfieldsaroundtheworld.Wefindthattheemissionsreductionsfromsupply-sideinterventionsareontheorderof40–50percentofthegrossemissionsofeachbarrelcurtailed,dependingontherelativeemissionsintensityofthecurtailedandsubstitutesourcesofsupply.Whiletheprecisemagnitudeoftheemissionsreductionsachievedspanaconsiderablerange,ingeneral,ourresultimplythatsupply-sideinterventionsarehighlylikelytoreducegreenhousegasemissionsonnet,with98.75percentofthescenariosweconsideryieldingnetreductions.Further,targetingsupply-sideinterventionsathighlyemissions-intensiveheavyoilsislikelytohavegreaterimpactonemissions.Forexample,curtailingCanadianoilsands,whichhaveanaverageintensityof648kgCO2e/boe,isexpectedtoyieldemissionsreductionsof357kgCO2e/boe(95percentrange:178to523kgCO2e/boe)ifleakagecomesfromproductionwithmarket-averageemissionsintensity.However,ifleakedproductionfromthecurtailmentofoilsandsisprimarilyintheformofmoreproductioninotheroilsandsfields,theemissionsreductionsaresomewhatsmaller:285kgCO2e/boe(95%range:86to471kgCO2e/boe).Thesecentralvaluessuggestnetemissionsreductionsroughlyone-halfaslargeastheemissionsintensityofthecurtailedsourceofsupply,aresultwhichvariesdependingonthesourcesofcurtailedandsubstitutesupply.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions327.ReferencesAsheim,GeirB.,TaranFæhn,KarineNyborg,MadsGreaker,CathrineHagem,BårdHarstad,MichaelO.Hoel,DiderikLund,andKnutE.Rosendahl.2019.“TheCaseforaSupply-SideClimateTreaty.”Science365(6451):325–27.https://doi.org/10.1126/science.aax5011.Balke,NathanS.,andStephenP.A.Brown.2018.“OilSupplyShocksandtheUSEconomy:AnEstimatedDSGEModel.”EnergyPolicy116(May):357–72.https://doi.org/10.1016/j.enpol.2018.02.027.Bodenstein,Martin,andLucaGuerrieri.2011.“OilEfficiency,Demand,andPrices:ATaleofUpsandDowns,”October.https://www.federalreserve.gov/econres/ifdp/oil-efficiency-demand-and-prices-a-tale-of-ups-and-downs.htm.Brons,Martijn,PeterNijkamp,EricPels,andPietRietveld.2008.“AMeta-AnalysisofthePriceElasticityofGasolineDemand.ASURApproach.”EnergyEconomics30(5):2105–22.https://doi.org/10.1016/j.eneco.2007.08.004.Brown,StephenPA.1998.“GlobalWarmingPolicy:SomeEconomicImplications.”EconomicReview,FourthQuarter.FederalReserveBankofDallas.Brown,StephenPA,CharlesMason,AlanKrupnick,andJanMares.2014.“CrudeBehavior:HowLiftingtheExportBanReducesGasolinePricesintheUnitedStates.”IssueBrief.ResourcesfortheFuture.https://www.rff.org/publications/issue-briefs/crude-behavior-how-lifting-the-export-ban-reduces-gasoline-prices-in-the-united-states/.Cairns,RobertD.,GrahamA.Davis,andJamesL.Smith.2021.“LettertotheEditoronBelfiori(EEEPVol.10(1)):FossilFuelSubsidies,theGreenParadoxandtheFiscalParadox.”EconomicsofEnergy&EnvironmentalPolicy10(1).https://www.proquest.com/openview/21726c77481ec8ceca3d5e7aa4083533/1?pq-origsite=gscholar&cbl=2069346.Cairns,RobertD.,andJamesL.Smith.2019.“TheGreenParadox,AHotellingCuldeSac.”EconomicsofEnergy&EnvironmentalPolicy8(2).https://doi.org/10.5547/2160-5890.8.2.rcai.Caldara,Dario,MicheleCavallo,andMatteoIacoviello.2019.“OilPriceElasticitiesandOilPriceFluctuations.”JournalofMonetaryEconomics103(May):1–20.https://doi.org/10.1016/j.jmoneco.2018.08.004.Coglianese,John,LucasW.Davis,LutzKilian,andJamesH.Stock.2017.“Anticipation,TaxAvoidance,andthePriceElasticityofGasolineDemand.”JournalofAppliedEconometrics32(1):1–15.https://doi.org/10.1002/jae.2500.Cooper,JohnC.B.2003.“PriceElasticityofDemandforCrudeOil:Estimatesfor23Countries.”OPECReview27(1):1–8.https://doi.org/10.1111/1468-0076.00121.Coyle,David,JasonDeBacker,andRichardPrisinzano.2012.“EstimatingtheSupplyandDemandofGasolineUsingTaxData.”EnergyEconomics34(1):195–200.https://doi.org/10.1016/j.eneco.2011.07.011.Cummings,RonaldG.,DavidS.Brookshire,andWilliamD.Schulze.1986.ValuingEnvironmentalGoods:AnAssessmentoftheContingentValuationMethod.Rowman&Allanheld.ResourcesfortheFuture33Dahl,CarolA.2012.“MeasuringGlobalGasolineandDieselPriceandIncomeElasticities.”EnergyPolicy,ModelingTransport(Energy)DemandandPolicies,41(February):2–13.https://doi.org/10.1016/j.enpol.2010.11.055.———.2014.“WhatDoWeKnowaboutGasolineDemandElasticities?”ColoradoSchoolofMinesWorkingPapers,November.https://ideas.repec.org//p/mns/wpaper/wp201411.html.Dahl,Carol,andThomasSterner.1991.“AnalysingGasolineDemandElasticities:ASurvey.”EnergyEconomics13(3):203–10.https://doi.org/10.1016/0140-9883(91)90021-Q.Erickson,Peter,MichaelLazarus,andGeorgiaPiggot.2018.“LimitingFossilFuelProductionasthenextBigStepinClimatePolicy.”NatureClimateChange8(12):1037–43.https://doi.org/10.1038/s41558-018-0337-0.Gately,Dermot,andHiliardG.Huntington.2002.“TheAsymmetricEffectsofChangesinPriceandIncomeonEnergyandOilDemand.”TheEnergyJournal23(1).https://doi.org/10.5547/ISSN0195-6574-EJ-Vol23-No1-2.Gilbert,Ben,andGavinRoberts.2020.“Drill-BitParity:Supply-SideLinksinOilandGasMarkets.”JournaloftheAssociationofEnvironmentalandResourceEconomists7(4):619–58.https://doi.org/10.1086/708160.Goodwin,Phil,JoyceDargay,andMarkHanly.2004.“ElasticitiesofRoadTrafficandFuelConsumptionwithRespecttoPriceandIncome:AReview.”TransportReviews24(3):275–92.https://doi.org/10.1080/0144164042000181725.Gordon,Deborah.2021.NoStandardOil:ManagingAbundantPetroleuminaWarmingWorld.OxfordUniversityPress.———.2023.“WeNeedtoTalkAboutMethane.”InTheFletcherForumofWorldAffairs,47:5–10.TheFletcherSchoolofLawandDiplomacy.Gordon,Deborah,EugenePerkHanTan,andJeffreyFeldman.2016.“ChangingOils,ChangingManagement.”JournalofIndustrialEcology20(4):673–75.Graham,DanielJ.,andStephenGlaister.2002.“TheDemandforAutomobileFuel:ASurveyofElasticities.”JournalofTransportEconomicsandPolicy36(1):1–25.Greene,DavidL.,andPaulNewsomeLeiby.2006.“TheOilSecurityMetricsModel:AToolforEvaluatingtheProspectiveOilSecurityBenefitsofDOE’sEnergyEfficiencyandRenewableEnergyR&DPrograms.”ORNL/TM-2006/505.OakRidgeNationalLab.https://doi.org/10.2172/930748.Hamilton,JamesD.2009.“CausesandConsequencesoftheOilShockof2007–08.”BrookingsPapersonEconomicActivity2009:215–61.Handler,Brad,andMorganBazilian.2021.“ExploringCarbonRetirementPortfolios.”PayneCommentarySeries,ColoradoSchoolofMines,July.https://repository.mines.edu/handle/11124/14109.Harstad,Bård.2012.“BuyCoal!ACaseforSupply-SideEnvironmentalPolicy.”JournalofPoliticalEconomy120(1):77–115.https://doi.org/10.1086/665405.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions34Hausman,JerryA.,andWhitneyK.Newey.1995.“NonparametricEstimationofExactConsumersSurplusandDeadweightLoss.”Econometrica63(6):1445–76.https://doi.org/10.2307/2171777.Hotelling,Harold.1931.“TheEconomicsofExhaustibleResources.”JournalofPoliticalEconomy39(2):137–75.Huntington,HillardG.1994.“OilPriceForecastinginthe1980s:WhatWentWrong?”TheEnergyJournal15(2).https://doi.org/10.5547/ISSN0195-6574-EJ-Vol15-No2-1.Huntington,HillardG.,JamesJ.Barrios,andVipinArora.2019.“ReviewofKeyInternationalDemandElasticitiesforMajorIndustrializingEconomies.”EnergyPolicy133(October):110878.https://doi.org/10.1016/j.enpol.2019.110878.Jing,Liang,HassanM.El-Houjeiri,Jean-ChristopheMonfort,AdamR.Brandt,MohammadS.Masnadi,DeborahGordon,andJouleA.Bergerson.2020.“CarbonIntensityofGlobalCrudeOilRefiningandMitigationPotential.”NatureClimateChange10(6):526–32.https://doi.org/10.1038/s41558-020-0775-3.Kilian,Lutz,andDanielP.Murphy.2014.“TheRoleofInventoriesandSpeculativeTradingintheGlobalMarketforCrudeOil.”JournalofAppliedEconometrics29(3):454–78.https://doi.org/10.1002/jae.2322.Knittel,Christopher,andShinsukeTanaka.2019.“DrivingBehaviorandthePriceofGasoline:EvidencefromFueling-LevelMicroData.”w26488.NationalBureauofEconomicResearch.https://doi.org/10.3386/w26488.Krautkraemer,JeffreyA.1998.“NonrenewableResourceScarcity.”JournalofEconomicLiterature36(4):2065–2107.Krichene,Noureddine.2002.“WorldCrudeOilandNaturalGas:ADemandandSupplyModel.”EnergyEconomics24(6):557–76.https://doi.org/10.1016/S0140-9883(02)00061-0.Krupnick,Alan,RichardMorgenstern,NathanBalke,StephenPABrown,andAnaMaríaHerrera.2017.“OilSupplyShocks,USGrossDomesticProduct,andtheOilSecurityPremium.”Report.ResourcesfortheFuture.Lazarus,Michael,andHarrovanAsselt.2018.“FossilFuelSupplyandClimatePolicy:ExploringtheRoadLessTaken.”ClimaticChange150(1):1–13.https://doi.org/10.1007/s10584-018-2266-3.Levin,Laurence,MatthewS.Lewis,andFrankA.Wolak.2017.“HighFrequencyEvidenceontheDemandforGasoline.”AmericanEconomicJournal:EconomicPolicy9(3):314–47.https://doi.org/10.1257/pol.20140093.Lin,C.-Y.Cynthia,andJieyin(Jean)Zeng.2013.“TheElasticityofDemandforGasolineinChina.”EnergyPolicy59(August):189–97.https://doi.org/10.1016/j.enpol.2013.03.020.Masnadi,MohammadS.,HassanM.El-Houjeiri,DominikSchunack,YunpoLi,JacobG.Englander,AlhassanBadahdah,Jean-ChristopheMonfort,etal.2018.“GlobalCarbonIntensityofCrudeOilProduction.”Science361(6405):851–53.https://doi.org/10.1126/science.aar6859.Mason,CharlesF.,andGavinRoberts.2018.“PriceElasticityofSupplyandProductivity:AnAnalysisofNaturalGasWellsinWyoming.”TheEnergyJournal39(01).https://doi.org/10.5547/01956574.39.SI1.cmas.ResourcesfortheFuture35NationalAcademiesofSciences,Engineering,andMedicine.2017.ValuingClimateChanges:UpdatingEstimationoftheSocialCostofCarbonDioxide.Washington,DC:NationalAcademiesPress.https://doi.org/10.17226/24651.NationalAcademiesofSciences,Engineering,andMedicine.2017.ValuingClimateDamages:UpdatingEstimationoftheSocialCostofCarbonDioxide.Washington,DC:NationalAcademiesPress.Newell,Peter,andAndrewSimms.2020.“TowardsaFossilFuelNon-ProliferationTreaty.”ClimatePolicy20(8):1043–54.https://doi.org/10.1080/14693062.2019.1636759.Newell,RichardG.,andBrianC.Prest.2019.“TheUnconventionalOilSupplyBoom:AggregatePriceResponsefromMicrodata.”TheEnergyJournal40(3).https://doi.org/10.5547/01956574.40.3.rnew.Newell,RichardG.,BrianC.Prest,andAshleyB.Vissing.2019.“TrophyHuntingversusManufacturingEnergy:ThePriceResponsivenessofShaleGas.”JournaloftheAssociationofEnvironmentalandResourceEconomists6(2):391–431.https://doi.org/10.1086/701531.Ploeg,Frederickvander,andArmonRezai.2020.“StrandedAssetsintheTransitiontoaCarbon-FreeEconomy.”AnnualReviewofResourceEconomics12(1):281–98.https://doi.org/10.1146/annurev-resource-110519-040938.Prest,BrianC.2021.“Supply-SideReformstoOilandGasProductiononFederalLands:ModelingtheImplicationsforClimateEmissions,Revenues,andProductionShifts.”ResourcesfortheFuture.https://www.rff.org/publications/working-papers/supply-side-reforms-oil-and-gas-production-federal-lands/.———.2022a.“Partners,NotRivals:ThePowerofParallelSupply-SideandDemand-SideClimatePolicy.”ResourcesfortheFuture.https://www.rff.org/publications/reports/partners-not-rivals-the-power-of-parallel-supply-side-and-demand-side-climate-policy/.———.2022b.“Supply-SideReformstoOilandGasProductiononFederalLands:ModelingtheImplicationsforCO2Emissions,FederalRevenues,andLeakage.”JournaloftheAssociationofEnvironmentalandResourceEconomists9(4):681–720.https://doi.org/10.1086/718963.Prest,BrianC.,andJamesH.Stock.2021.“ClimateRoyaltySurcharges.”WorkingPaper.WorkingPaperSeries.NationalBureauofEconomicResearch.https://doi.org/10.3386/w28564.RMI.2022.“OilClimateIndexplusGasv.1.0.”OilClimateIndexplusGas.2022.https://ociplus.rmi.org/.Serletis,Apostolos,GovindaR.Timilsina,andOlexandrVasetsky.2010.“InterfuelSubstitutionintheUnitedStates.”EnergyEconomics32(3):737–45.https://doi.org/10.1016/j.eneco.2010.01.013.Slade,MargaretE.,andHenryThille.2009.“WhitherHotelling:TestsoftheTheoryofExhaustibleResources.”AnnualReviewofResourceEconomics1(1):239–60.https://doi.org/10.1146/annurev.resource.050708.144223.Yatchew,Adonis,andJoungyeoAngelaNo.2001.“HouseholdGasolineDemandinCanada.”Econometrica69(6):1697–1709.EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions36AppendixA.1.SensitivityApproachtoEmissionsIntensityCalculationThisappendixpresentsasensitivityanalysisunderthealternativeassumptionthatallfieldswithinagivencategorycontributetocurtailedorsubstitutesupplyinproportiontotheirproduction,ratherthanthefield-specificsamplingofemissionsintensityusedinourmainresults.Thissensitivityanalysisfocusesonthe100-yearGWPtocorrespondwiththetablesinthemainresults.Thiseffectivelyreplacesthefield-specificdistributionofemissionsintensitieswiththeirproduction-weightedaverage.TheresultsareshowninTablesA1–A4,whichareanalogoustoTables4–7inthemaintext.ComparingtheaveragenetemissionschangesTableA1toTable4andTableA2toTable5,thisalternativeassumptionhasessentiallynoeffectontheaverageestimates,whichonlydifferveryslightlyduetorandomnessintheMonteCarlosampling.Butbyrestrictingtheuncertaintyrangeinemissionsintensity,itreducesthespreadinnetemissionsimpacts,whichisnownearlyentirelydrivenbyuncertaintyintheleakagerate.ThisyieldsnarroweruncertaintyrangesinTablesA3andA4ascomparedtoTables6and7andhighercertaintyofemissionsreductions.AmongthecurtailedsourcesconsideredinTableA1,wedonotfindevenasingledrawfeaturingnetemissionincreases.Acrossour1.53millionsimulationsusingthe100-yearGWP,wefindlessthan0.1percentofdrawsyieldingnetemissionsincreases.Thesedrawsrepresentcaseswithhighdrawsoftheleakagerate(uniformlyover77percent)andassumethatlowemissivecurtailedcategories(e.g.,lightoil)aresubstitutedbyhighlyemissiveonestoveryhighones(oilsandsorheavyoil),whichseemslikearelativelyimplausibleassumption.ResourcesfortheFuture37TableA1.NetEmissionsChangeperCurtailedBarrelwhenSubstituteProduction’sEmissionsIntensityReflectsAllFieldsorSameasCurtailedSource,NotSamplingfromField-levelEmissionsIntensitiesSubstituteproductionfrom:Allsources(marketaverage)SamesourcesAveragenetemissionschange(kgCO2e/boe)Shareofdrawswithemissionsreductions(%)Averagenetemissionschange(kgCO2e/boe)Shareofdrawswithemissionsreductions(%)Curtailedsource100-yearGWP20-yearGWP100-yearGWP20-yearGWP100-yearGWP20-yearGWP100-yearGWP20-yearGWPAll(marketaverage)-228-255100100-228-255100100NorthAmerica-250-289100100-238-270100100Oilsands-357-367100100-285-305100100PermianBasin-234-354100100-230-298100100Lightoil-208-243100100-219-250100100Mediumoil-222-240100100-225-249100100Heavyoil-322-342100100-269-293100100Sweetoil-249-279100100-237-266100100Souroil-228-251100100-228-253100100EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions38TableA2.AverageNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),100-YearGWP,NotSamplingIntensitiesTableA3.97.5thPercentileofNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),100-YearGWP,NotSamplingIntensitiesResourcesfortheFuture39TableA4.2.5thPercentileofNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),100-YearGWP,NotSamplingIntensitiesA.2.DetailedResultsundera20-YearGWPTableA5.AverageNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),20-YearGWPEstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions40TableA6.97.5thPercentileofNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),20-YearGWPTableA7.2.5thPercentileofNetEmissionsImpactsofOneBarrelofCurtailedSupply,byCurtailedSources(rows)andSubstituteSources(columns),20-YearGWPResourcesfortheFuture41A.3.TheEffectofDynamicsThemodelconsideredinthisstudyisstatic,butmoregenerallyoilisastorablecommodity,whichintroducesrelevantintertemporaldynamics.Thisisimportantwhenconsideringwhethertouseshort-runorlong-runvaluesforthekeyinputsintoequations(7)and(8),includingsupplyanddemandelasticitiesandtheemissionsintensitiesofmarginalsupply.Extendingthemodeltobedynamicrevealsaconceptuallysimilarresulttothosefoundinthemainequationsrevealsthattherelevantinputparametersshouldbebasedonlong-runvalues.Long-runelasticitiestendtobelargerthanshort-runones,whichhasambiguouseffectsontheleakagerate.However,itmaybereasonabletoexpectthelong-runmarginalsourceofsupplytobecleanerthantheshort-runvalue,giveninternationalgoalstocutmethaneemissionsovertime.Inthisappendix,wederiveresultsanalogoustoequations(7)and(8)butintheframeworkofadynamicmodel,wheresupplyanddemandarenowadditionallyindexedbytime,𝑡𝑡.Now,assumingoilcanbestoredforfutureusethroughinventories,themarket-clearingconditionisthatthesumofintertemporalsupplyanddemand,includingthebarrelwithheldtoday(nowindexedby𝑡𝑡=1),balance:−𝑞𝑞�1+��𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1𝑁𝑁𝑖𝑖=1=��𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1𝑁𝑁𝑖𝑖=1.Differentiatingthemarketclearingconditionwithrespectto𝑞𝑞�:−1+��𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�𝑁𝑁𝑖𝑖=1=��𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝)𝑇𝑇𝑡𝑡=1𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�𝑁𝑁𝑖𝑖=1Solvingforthepriceeffect,𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�⁄,yields𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�=1∑∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1−𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1Thisisexactlythesameformasthepriceeffectinthemaincase,exceptthatthevaluesthatmatterarethesumofmarginalsupplyanddemandacrossbothregionandtime.Theeffectontotalglobalsupplyanddemandinyear𝑡𝑡,𝑄𝑄𝑡𝑡,𝑆𝑆=∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝)𝑁𝑁𝑖𝑖=1and𝑄𝑄𝑡𝑡,𝑆𝑆=∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝)𝑁𝑁𝑖𝑖=1,are𝑑𝑑𝑄𝑄𝑡𝑡,𝑆𝑆𝑑𝑑𝑞𝑞�=∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑁𝑁𝑖𝑖=1∑∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1−𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1𝑑𝑑𝑄𝑄𝑡𝑡,𝐷𝐷𝑑𝑑𝑞𝑞�=∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1∑∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1−𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝)𝑁𝑁𝑖𝑖=1EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions42Notethatnowtheimpactsonglobalsupplyanddemandarethesameinthelongrun,buttheyneednotbethesameinanygivenyear.Justasinthestaticmodel,theimpactonsupplyandregionaldemandneednotbethesameforeveryindividualregion.However,thecumulativeeffectonglobaloilconsumptionisgivenbythesumofannualconsumptionovertime,𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=�𝑑𝑑𝑄𝑄𝑡𝑡,𝐷𝐷𝑑𝑑𝑞𝑞�𝑇𝑇𝑡𝑡=1=∑∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1𝑁𝑁𝑖𝑖=1∑∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑇𝑇𝑡𝑡=1−𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝𝑡𝑡)𝑁𝑁𝑖𝑖=1Asinthestaticmodel,thisequationcanbefurthersimplifiedtobeafunctionofsupplyanddemandelasticities.First,denoteregion𝑖𝑖’ssupplyanddemandtime-𝑡𝑡elasticitiesas𝜂𝜂𝑖𝑖,𝑡𝑡and𝜀𝜀𝑖𝑖,𝑡𝑡:𝜂𝜂𝑖𝑖,𝑡𝑡=𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑝𝑝𝑡𝑡𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆⇒𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)=𝜂𝜂𝑖𝑖,𝑡𝑡𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆𝑝𝑝𝑡𝑡𝜀𝜀𝑖𝑖,𝑡𝑡=𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝𝑡𝑡)𝑝𝑝𝑡𝑡𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷⇒𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷′(𝑝𝑝𝑡𝑡)=𝜀𝜀𝑖𝑖,𝑡𝑡𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷𝑝𝑝𝑡𝑡Pluggingthesetwointothepreviousequationandfurthermultiplyinganddividingbytotalglobalsupplyanddemand(whicharethesame,𝑄𝑄𝑆𝑆=−𝑞𝑞�+∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆(𝑝𝑝t)𝑖𝑖,𝑡𝑡=∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷(𝑝𝑝t)𝑖𝑖,𝑡𝑡=𝑄𝑄𝐷𝐷=𝑄𝑄)yields𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=∑𝑝𝑝𝑡𝑡−1∑𝑤𝑤𝑖𝑖,𝑡𝑡,𝐷𝐷𝜀𝜀𝑖𝑖,𝑡𝑡𝑁𝑁𝑖𝑖=1𝑇𝑇𝑡𝑡=1∑𝑝𝑝𝑡𝑡−1∑𝑤𝑤𝑖𝑖,𝑡𝑡,𝑆𝑆𝜂𝜂𝑖𝑖,𝑡𝑡𝑁𝑁𝑖𝑖=1−𝑤𝑤𝑖𝑖,𝑡𝑡,𝐷𝐷𝜀𝜀𝑖𝑖,𝑡𝑡𝑇𝑇𝑡𝑡=1where𝑤𝑤𝑖𝑖,𝑡𝑡,𝐷𝐷=𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷/𝑄𝑄𝐷𝐷and𝑤𝑤𝑖𝑖,𝑆𝑆=𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆/𝑄𝑄𝑆𝑆areweightsrepresentingtheshareofcumulativedemandandsupplycomingfromregion𝑖𝑖inyear𝑡𝑡.Ifthepricefollowsano-arbitrageconditionandrisesannuallyattherateofinterest,denoted𝑟𝑟(whichisalsoconsistentwithaHotellingpricepath),asin𝑝𝑝𝑡𝑡=𝑝𝑝1(1+𝑟𝑟)𝑡𝑡−1,thenthe𝑝𝑝𝑡𝑡−1termsareallreplacedby(1+𝑟𝑟)−(𝑡𝑡−1),asfollows:𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=∑(1+𝑟𝑟)−(𝑡𝑡−1)∑𝑤𝑤𝑖𝑖,𝑡𝑡,𝐷𝐷𝜀𝜀𝑖𝑖,𝑡𝑡𝑁𝑁𝑖𝑖=1𝑇𝑇𝑡𝑡=1∑(1+𝑟𝑟)−(𝑡𝑡−1)∑𝑤𝑤𝑖𝑖,𝑡𝑡,𝑆𝑆𝜂𝜂𝑖𝑖,𝑡𝑡𝑁𝑁𝑖𝑖=1−𝑤𝑤𝑖𝑖,𝑡𝑡,𝐷𝐷𝜀𝜀𝑖𝑖,𝑡𝑡𝑇𝑇𝑡𝑡=1ResourcesfortheFuture43Thisisconceptuallyanalogoustoequation(4)butinsteadofproduction-weightedaverageelasticities,theyaretemporallydiscountedproduction-weightedaverageelasticities.Asthediscountrateapproacheszero,thiscollapsestousingintertemporallyandregionallyaveragedelasticityvalues,asin𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=𝜀𝜀𝜂𝜂−𝜀𝜀wherenow𝜀𝜀=∑∑𝑤𝑤𝑖𝑖,𝑡𝑡,𝐷𝐷𝜀𝜀𝑖𝑖𝑁𝑁𝑖𝑖=1𝑇𝑇𝑡𝑡=1and𝜂𝜂=∑∑𝑤𝑤𝑖𝑖,𝑡𝑡,𝑆𝑆𝜂𝜂𝑖𝑖𝑁𝑁𝑖𝑖=1𝑇𝑇𝑡𝑡=1.Thisisthesameresultasinthestaticmodel,whichholdswhen𝑟𝑟=0or𝑇𝑇=1(inwhichthedynamicmodelcollapsestothestaticone).When𝑟𝑟>0,weshouldplacesomewhatgreaterweightonnear-termsupplyanddemandelasticities.Notealsothatwecanwritethisasawithin-yearregionalweightedaverage,whichisthenweightedbyyear𝑡𝑡’scontributiontototalsupplyanddemandovertime.𝑑𝑑𝑄𝑄𝐷𝐷𝑑𝑑𝑞𝑞�=∑(1+𝑟𝑟)−(𝑡𝑡−1)𝑄𝑄𝑡𝑡,𝐷𝐷𝑄𝑄𝐷𝐷∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷𝑄𝑄𝑡𝑡,𝐷𝐷𝜀𝜀𝑖𝑖𝑁𝑁𝑖𝑖=1𝑇𝑇𝑡𝑡=1∑(1+𝑟𝑟)−(𝑡𝑡−1)�𝑄𝑄𝑡𝑡,𝑆𝑆𝑄𝑄𝑆𝑆∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆𝑄𝑄𝑡𝑡,𝑆𝑆𝜂𝜂𝑖𝑖𝑁𝑁𝑖𝑖=1−𝑄𝑄𝑡𝑡,𝐷𝐷𝑄𝑄𝐷𝐷∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝐷𝐷𝑄𝑄𝑡𝑡,𝐷𝐷𝜀𝜀𝑖𝑖𝑁𝑁𝑖𝑖=1�𝑇𝑇𝑡𝑡=1Denotingtheemissionsintensityofregion𝑖𝑖’soilinperiod𝑡𝑡as𝑒𝑒𝑖𝑖,𝑡𝑡,cumulativeglobalemissionsaregivenby𝐸𝐸=−𝑒𝑒𝑞𝑞�+�𝑒𝑒𝑖𝑖,𝑡𝑡𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆(𝑝𝑝)𝑖𝑖,𝑡𝑡.Theimpactofcurtailedproductiononoverallemissionsarethen𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=−𝑒𝑒𝑞𝑞�+�𝑒𝑒𝑖𝑖,𝑡𝑡𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�.𝑖𝑖,𝑡𝑡Thiscanbere-writteninananalogousformtoequation(7)as𝑑𝑑𝑑𝑑𝑑𝑑𝑞𝑞�=−𝑒𝑒𝑞𝑞�+𝑒𝑒𝐿𝐿,EstimatingtheEmissionsReductionsfromSupply-sideFossilFuelInterventions44𝑒𝑒=1∑𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�𝑖𝑖,𝑡𝑡�𝑒𝑒𝑖𝑖,𝑡𝑡𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�𝑖𝑖,𝑡𝑡and𝐿𝐿=�𝑞𝑞𝑖𝑖,𝑡𝑡,𝑆𝑆′(𝑝𝑝𝑡𝑡)𝑖𝑖,𝑡𝑡𝑑𝑑𝑝𝑝𝑡𝑡𝑑𝑑𝑞𝑞�.Theseresultsthusmirrortheone-periodcaseinthemaintext,indicatingthattheresultsfromthesimpleone-periodmodelnonethelessextendtoa𝑇𝑇-periodcase.Wherenowtheweightedaverageemissionsintensityandleakageratevaluesreflecttheirintertemporalanalogues:ResourcesfortheFuture45

1、当您付费下载文档后，您只拥有了使用权限，并不意味着购买了版权，文档只能用于自身使用，不得用于其他商业用途（如 [转卖]进行直接盈利或[编辑后售卖]进行间接盈利）。
2、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。
3、如文档内容存在违规，或者侵犯商业秘密、侵犯著作权等，请点击“违规举报”。

碎片内容

碳中和的最新文档