Distributive fairness: A mutual recognition approach

Distributive fairness: A mutual recognition approach

Arild Underdal

* , Taoyuan Wei

aUniversityofOslo,DepartmentofPoliticalScience,P.O.Box1097,Blindern,0317Oslo,Norway

bCenterforInternationalClimateandEnvironmentalResearch–Oslo(CICERO),P.O.Box1129,Blindern,0318Oslo, Norway

1. Introduction

Cannormsofdistributivefairnessserveaspillarsofanewand more effective global climate change regime? A positive answerrequiresthatatleasttwoconditionsbemet.First,a small setofcompatible fairnessprinciples andoperational interpretationsoftheseprinciplesmustbeacceptedasvalid and relevantby acritical minimumofparticipating states.

Second,theseprinciplesandinterpretationsmustinfactserve asimportantpremisesforthesestates’policiesandpositions.

Good reasons for pessimism regarding both of these conditions are easy to find. Climate change mitigation combinesseveralfeaturesthatmakeitanextremelydemand- ing governancechallenge (Levinet al., 2012;Verweijet al., 2006; Victor, 2011). For many countries, large cuts in greenhouse gas (GHG) emissions are called for, requiring radicalchangesinimportantpoliciesandpractices.Verylong timelags,manyextendingwellbeyondonehuman genera- tion,existbetweenmitigationmeasures(involvingmoreor lesspredictablecostsforspecificgroups)andeffects(inthe formofmoreuncertainbenefitsfortheworld).Suchtimelags article info

Articlehistory:

Availableonline10April2015

Keywords:

Mitigation Fairness Equity

Responsibilities Capabilities Mutualrecognition

abstract

Cannormsofdistributivefairnessserveaspillarsofanewandmoreeffectiveglobalclimate regime?Threegeneralprinciples–responsibilities,capabilities(capacity),andneeds(or rights) –are frequentlyinvoked andrarelydisputed. Yet,parties’interpretationsoften diverge,reflectingconflictsofinterests.Todeterminehowmuchisatstake,wecompare–by meansofaglobalintegratedassessmentmodel(GRACE)–15legitimateinterpretationsof

‘responsibilities’and‘capabilities’intermsoftheirimplicationsforthemitigationobliga- tionsandcostsofsevenpotentiallypivotalactors.Mostoftheseinterpretationsyieldsimilar resultsformostactors.Inascenariowhereglobalemissionsin2030arereducedby20%

comparedtoabusiness-as-usualbaseline,mitigationcostsvarybylessthan1%ofGDPfor theUnited States,the EuropeanUnion,Japan,India, andChina.ForBrazilandRussia, however,varianceismuchlarger.Moreover,forallactors,mitigationcostsrisesteeplyas ambitionlevelsincrease.Undersuchcircumstances,searchingforasingle‘fairness-opti- mizing’formulaislikelytofail.Asnegotiatorsexploresystemsofvoluntarypledges,amore promisingapproachwouldconceiveoffairnessasamultidimensionalconstructandfoster accommodationthroughmutualrecognitionofalimitedrangeoflegitimatenorminter- pretations.

#2015TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

*Correspondingauthorat:UniversityofOslo,DepartmentofPoliticalScience,P.O.Box1097,Blindern,0317Oslo,Norway.

Tel.:+4722855241.

E-mailaddress:[email protected](A.Underdal).

Availableonlineatwww.sciencedirect.com

ScienceDirect

journalhomepage:www.elsevier.com/locate/envsci

http://dx.doi.org/10.1016/j.envsci.2015.03.009

1462-9011/# 2015The Authors. Publishedby ElsevierLtd.This is anopen access articleunder theCC BY-NC-ND license (http://

creativecommons.org/licenses/by-nc-nd/4.0/).

distortcost-benefitcalculationsbyleaving importantstake- holdersdisenfranchisedandfuturebenefitsunderrepresent- ed. Stark asymmetries between rich (polluters) and poor (victims)generate severeconflictsofinterest and ‘dampen cooperative efforts’ (Parks and Roberts, 2008, p. 621). In addition,strongcompetitioninworldmarketsandinterna- tionalpoliticstendstoreinforceparties’concernswithrelative gains and losses. Under such conditions, orchestrating effectivecooperationwouldbeatallorderforanyintergov- ernmentalorganization.ForthenegotiationsundertheUnited NationsFrameworkConventiononClimateChange(UNFCCC) –aninstitutioncombininguniversalparticipationwithavery demanding decision rule (consensus) and a distribution of implementation power that tilts in favour of the major emitters–thechallengeseemsoverwhelming.

Oneimportantimplicationofthis sombreassessmentis that searching for a common and precise formula that policymakers and diplomats can use to ‘derive’ a fair distributionofobligationsandrightsisnotlikelytosucceed.

Infact,intensivesearchforasingleauthoritative‘fairness- optimizing’formula may wellincrease the risk ofdeadlock (Bretschger,2013;ParksandRoberts,2008;Victor,2011).Partof theexplanationcanbefoundinglobalconferencediplomacy itself. Plenary sessions – in particular, those spotlighting political leaders – provide fertile ground for ideological posturing and for defending the interests of important domestic constituencies. Moreover, by establishing semi- permanent groups, theUN system‘may actuallyconstruct new lines of confrontationover and above the substance- based disagreements existing between countries’ (Castro etal.,2014,p.109).Theriskofsuchcounter-productiveeffects willlikelyincreasefurtherifa‘top-down’formulaapproach weretobepursuedatatimewhennegotiationsareturning towards‘bottom-up’pledgesofvoluntarycontributions.

Yet, extant research strongly indicates that fairness matters,particularly when dealingwithstark asymmetries between rich and poor (Dannenberg et al., 2010; Gampfer, 2014;Langeetal.,2010).Theclimatechangechallengebrings totheforefrontprofoundquestionsconcerningmoralrespon- sibility, mitigation and adaptation capacity, and people’s rightstotheglobalcommonsandtoeconomicdevelopment.

Althoughofteninvokedtolegitimizeandreinforceinterest- basedargumentsand positions,normsoffairnesscanalso servetoconstrainthepursuitofself-interestandtoprovide roadmaps for accommodation (Dannenberg et al., 2010;

Gampfer,2014;Langeetal.,2010).Someanalystsarguethat for an international agreement to be effective it ‘must be widelyperceivedasequitable’(WinklerandRajamani,2014,p.

103).

In this paper, we ‘translate’ the UNFCCC principles of responsibilities and capabilities into 15 allocationschemes anduse aglobal integratedassessment model(GRACE,see Appendix)toexploretheimplicationsoftheseschemesforthe mitigationobligationsandcostsofsevenpotentiallypivotal actors:UnitedStates,EuropeanUnion,Japan,Russia,Brazil, China,andIndia.Webegin(Section2)withbrieflyreviewing extantresearchtoidentifybroadlyacceptedfairnessprinci- plesandlegitimateinterpretationsoftheseprinciplesforthe globaldistributionofmitigationobligations.InSection3,we apply these interpretations to our seven actors undertwo

alternativeglobalemissionreductiontargets.Wefirstexplore the implications of the 15 interpretations for the relative distributionofmitigationobligations(Section3.1)andmove on to estimate each actor’s costs of meeting its own obligations underthetwoglobalemissionreductiontargets (Section3.2).Inthefinalsection,weexploretheimplications of these results for fairness-promoting strategies in the UNFCCCnegotiations.Giventhestarkasymmetriesbetween rich and poor and the consensus rule of the UNFCCC conferences,we arguethatthe mostconstructivecontribu- tionstoafairandeffectiveagreementwilllikelycomefrom actors who conceive of fairness as a multidimensional construct,recognizealimitedrangeofnorminterpretations aslegitimate,andfosterpositivereciprocitythroughcoopera- tive(moreprecisely,‘integrative’)behaviour.

2. Fairness principles and operational interpretations

2.1. Normsandinterests

In theresearch literature,threegeneral observationsstand out.First,eventhoughabewilderingarrayoffairnesscriteria and arguments may seem to exist (see e.g. Klinsky and Dowlatabadi,2009,pp.97–98),theliteratureshowsconsider- ableconvergenceonasmallsetofbasicprinciples.Second, parties’relativeprioritiesand(operational)interpretationsof these principlestendtoreflectnational circumstancesand materialinterests(Carlssonetal.,2013;Langeetal.,2010).Not surprisingly,G77estimatesresponsibilityretrospectively–in someinstancesgoingbacktotheIndustrialRevolution–while theUnitedStatesattachesmoreimportancetorecenttrends and likely future trajectories. Where some interpretations yield significantly higher mitigation costs than others, material interests will likely trump fairness norms. Third, the two sets of premises seem to interact synergistically, meaning (a) that parties tend to favour fairnessprinciples and interpretations that are compatible with their own material interests, and (b) that any given principle and interpretation willlikely be more important in reinforcing the positions of parties that stand to gain from their application than inmodifying the positionsofparties that expecttolose.Combining(a)and(b),wecanseethatinhighly asymmetrical relationships,broadconsensusatthelevel of generalprinciplesneednotfacilitateagreementonaspecific deal(Underdaletal.,2012,p.487).Accordingly,attentiontothe operationalinterpretationofnormsisrequiredtounderstand whatpartiescangainorlose.

In this paper, theterm ‘fair’ refersto distributionsthat combinetwo keyelements: equaltreatmentofequal cases (here:equality),anddifferentialtreatmentofcasesthatdiffer significantly inimportant respects (here: equity). The latter requirementismostoftentranslatedintoasomewhatflexible notion ofproportionality. Sometimes, however, the range of variance is so wide that even a flexible interpretation of proportionality would leavethe poorestor weakest parties withburdenstheycannotreasonablybeexpectedtoshoulder.

In suchcases,amorecategoricalrule ofexemptionisoften introduced, relieving certain parties (temporarily) of any

substantive obligations for which they are not adequately compensated. One important lesson emerging from this literatureisthattoqualifyasfair,aclimateagreementmust combinenotionsofequality,proportionality,andexemption,as indicatedinTable1(seeRingiusetal.,2002).

In climate change negotiation documents and public statements, at least three interpretations of equity are frequentlyinvokedandrarelydisputed(seee.g.Mattooand Subramanian,2012).Theseinterpretationsrefertoaparty’s responsibilityforcausingdamage,itscapacitytocontributeto problem solving, and its need for (or right to) the goods or benefitsconcerned(Table2).Responsibilityisthebackboneof the polluter-pays principle, capacityis the keydifferential variableinschemesofprogressivetaxation,andneedisthe mostimportantcriterioninsocial-welfareprogrammes.The UNFCCCreferenceto‘commonbutdifferentiatedresponsibil- itiesandrespectivecapabilities’(CBDR&RC),acceptedby195 states, is a fairly succinct expression of this three-pillar platform.Intheclimatechangeliterature,noclearrankinghas beenestablishedamongthesecriteria,butawidelyaccepted interpretationseemstobethatinsofarasneeds(rights)refer tobasicgoodsorfundamentalhumanrights,theneeds(rights) criteriontrumpstheothers(Mu¨llerandMahadeva,2013,p.8).

A survey targeting respondents directly involved in the climate change negotiations showed strongest support for theprinciplesofresponsibility(supportbalance=+69%)and need(supportbalance=+57%)(Langeetal.,2007).¹

Thequestionhereishowalternativeinterpretationsofthe CBDR&RCprinciplesofresponsibilitiesandcapabilitieswould affectimportantcountries.Weaddressthisquestionfirstby exploringhowcountries’relativecontributionsvarywiththe exact interpretation of these principles, and second by estimating the impact of countries’ relative mitigation contributions on their national welfare, under two global ambition levels. To render these tasks manageable, the analysis is limited to seven key actors: Brazil, China, the EuropeanUnion,India,Japan,Russia,andtheUnitedStates.

Together,theseactorsaccountfornearlytwothirdsofworld GHGemissions(includingLUCF).Moreover,mostareseenas leadersoflargergroups ofcountries(Karlssonetal., 2011), indicatingthatanymitigationagreementsignedandratified byallsevenwilllikelybeacceptedbyalargemajorityofother statesaswell.

2.2. Interpretationsofresponsibility

Normative theory distinguishes between an actor’s role in causingdamageandthatactor’smoralresponsibility(‘guilt’) forthedamageithascaused.Acausalroleisanecessarybut notasufficientconditionformoralresponsibility.Toassign moralresponsibility,onemustinadditionprovethatanactor (a)had,oratleastcouldhaveobtained,effectivecontrolover the harmfulactivities forwhich it stands accused, and (b) knew,oratleastcouldreasonablybeexpectedtohaveknown, the (risk of) damage caused by these activities (see e.g.

Aristotle,350BCE;Mu¨lleretal.,2009).Thecontrolrequirement limits the transferability of guilt across generations.² The availableknowledgerequirementimpliesthatthehistorical backlog of moral responsibility cannot go back to the Industrial Revolution, beginning around 1760. Most of the research literature seems to agree that there developed around 1970 a sufficiently solid and well-known scientific basisforsuspecting humanactivities ofbeing asignificant driverofclimatechange(MattooandSubramanian,2012,p.

1088);tobeonthesafeside,somesaycompellingevidencehas been available ‘atleast [...]since 1990’ (Parikhand Parikh, 2009, p. 4), the year the first IPCC assessment report was published.

Whatmaylegitimatelybetracedbacktoearlyperiodsof technologicalinnovationandeconomicgrowthareaccumulat- edcompetitiveadvantagesenjoyedalsobycurrentgeneration(s).

Particularly relevanttotheclimatechangenegotiationsare persistent advantages and benefits accumulated through unrestricteduseofEarth’scapacitytoabsorbGHGemissions.

Thiscapacityisaglobalcollectivegood,overexploitedbythe richNorthtoitsownadvantage.Theworld’spoor,manybeing innocentvictimsofclimatechange,canmakeastrongcasefor equal opportunitiesor adequatecompensation(Baer, 2013).

Onewayofrecognizingtheequal-opportunitiesclaimwould betoincludecompetitiveadvantagesaccumulatedbytherich Norththroughunrestricteduseofglobalcommonsasintegral elementsofcapabilities(implyinghighermitigationobligations for the North) and needs/rights (implying lower mitigation obligationsfortheSouth).

To apply the responsibility principle to climate change mitigation,furtherspecificationisrequiredinat leastthree respects. First, whichGHGs and human activitiesshouldbe includedinresponsibilityassessments(seedenElzenetal., 2013)?Sincethe humanimpacton theclimatesystemisa function of the weighted aggregate of all GHG emissions generatedbyhumanactivities,thedefaultoptionwouldbean equally comprehensive responsibility estimate. Practical problemsofemissionsaccountingandimpactmeasurement may,however,leadpartiestosettleforalesscomprehensive programme. Hence, this analysis considers two notions of comprehensiveness:CO2emissionsfromfossilfuelsonlyand totalemissionsofGHGs.

Second, who qualifies for exemption? In the research literature, the most common approach has been to grant exemptions to countries whose average income level falls belowacertain(official)povertyline(e.g.Baer,2013;Mu¨ller and Mahadeva, 2013). Since even poorcountries have rich Table1–Fairnessprinciplesandtheirvaliditydomains.

Fairnessprinciples Validitydomains Equality Relevantdifferencestoosmallto

benormativelysignificant Equity–proportionality Relevantdifferencesnormatively

significantbutnotverylarge Equity–exemption Relevantdifferencesverylarge Note:BuildsonRingiusetal.,2002.

1 The supportbalance is measuredas (very high+highsup- port) (low+verylowsupport).Inthesurvey,theresponsibility principlewaslabelled‘polluterpays’andtheneedprinciple‘poor losers’.

2Formoreorlesspermanentorganizations,suchasstates,this limitationmaybemodifiedbutnotdismissed.

peopleandrichcountrieshavepoorpeople,somehaveargued thatthedomesticdistributionofincomeorwealthshouldalso betakenintoaccount(e.g.Rao,2013).Theanalysisreportedhere applies the following exemption rules: Countries with per capitaCO2emissionsabovetheworld average(here:United States,EuropeanUnion,Japan,andRussia)haveproportional responsibilityforalltheirownemissions.Countriesemitting between50%and100%oftheworldaverage(here:Chinaand RestoftheWorld)areproportionallyresponsibleforemissions withinthatintervalonly.Countriesemitting<50%oftheworld average(here:BrazilandIndia)aregrantedfullexemption.

Third, for which time horizon should responsibility be estimated(seeFrimanandHjerpe,2014)?Thescopeconditions ofcontrolandknowledgelimitthebacklogofmoralguiltbut providenoguidanceregardinglikelyfutureemissiontrajec- tories.Whatcanbesaid,however,isthatdynamicupdating wouldberequiredtocapturesignificantchangesincountries’

shares of global emissions. For CO2 emissions, four partly overlappingtimehorizons areconsidered(1971–2009,1990–

2009,1971–2017,and1990–2017);forallGHGs,onlythe1990–

2010periodisincluded.

Some recent studies have made a case for replacing conventionalemissionsaccountingwithestimatesthatalso capturethecarbonembodiedininternationaltrade(seee.g.

Peters et al., 2012). Other studies have argued that since responsibilities and capabilities vary substantially within countries,theappropriatelevelofanalysiswouldbeindivid- ualsorhouseholdsratherthancountries(seee.g.Karthaetal., 2012).Threeofourallocationschemesaredesignedtoexplore plausible implications of these arguments for the global distributionofmitigationobligations.

2.3. Interpretationsofcapabilities

Strictly speaking, capabilities can be assessed only with referencetoaspecifictaskorfunction.Thus, somefactors importantinboostingmitigationcapacity–renewableenergy resourceendowments,forexample–arelessimportantfor adaptation.Moreover,acountry’scontributiontomitigating GHGemissionsmaytakedifferentforms–frompreservingor establishing sinksto transforming carbon-intensive energy systems–andthecapabilitiesrequiredwillsomewhatdepend on the kind of contribution made. Confronted with such complexity,researchers(andpolicymakers)havelookedfora simple capabilitiesconcept thatcan covera widerange of tasks and functions. ‘Capacity to pay’seems to meet this requirement at least as well as any other equally simple conceptualization, and GDP per capita has emerged as a

broadly acceptedindicatorforwhichstandardized dataare readilyavailable.Refinementshavebeensuggested,however.

AmongthesetheOxfordCapabilitiesMeasurestandsoutasa strongcandidate,combiningnationalGDPandGDPpercapita figures with an index of ‘poverty intensity’ (Mu¨ller and Mahadeva,2013).Someresearchersfindthecapacity-to-pay frameworktoonarrowandpointtomoreinclusivemultidi- mensional constructs, in particular the UNDP’s Human DevelopmentIndex(Winkleretal.,2013,p.413).Theargument hasmerit;someimportantcapabilitycomponents–renew- ableenergyresourceendowmentsbeingoneobviousexample –arenotatallrepresentedbytheGDPpercapitaindicator.

Moreover, important tasks, such as de-carbonizing energy systems,callforcapacitytoinnovateandgovern,notmerely capacitytopay.

InresponsetoobjectionsraisedagainstrelyingonGDPper capitaastheonlyindicatorofrelevantcapabilities,twomore inclusiveindexesareconstructed.One,labelledtransformation capacity, is designed to measure countries’ economic and politicalabilitiesto‘de-carbonize’energysystems.Asdefined here,transformationcapacityisafunctionof(a)theconven- tionalGDPpercapitavariable(weight.6),(b)innovationcapacity (.2),(c)governancecapacity(.1),and(d)transparency(.1).The otherindex,labelledrenewableenergyresourceendowments,isa weighted aggregate ofcountries’ percapita endowments of solar(.4),wind(.2),bio(.2),andhydro(.2)energyresources.³

Tosumup,wehavetranslatedtheUNFCCCprinciplesof responsibilitiesandcapabilitiesinto15allocationschemes,of which11areinterpretationsofresponsibilitiesandfourare interpretations of capabilities. Table 3 provides a brief overviewoftheseschemes;moreinformationisprovidedin theonlinesupplementarymaterial.

2.4. Scenariosforassessmentofmitigationcosts

To estimate mitigation costs of alternative distributive schemes we need to specify mitigation targets (ambition levels).In thisanalysis, globalmitigationtargets aredeter- minedonthebasisofabusinessasusual(BAU)scenarioby usingaglobaleconomicmodel(GRACE)brieflydescribedinthe Appendix. We construct aBAU scenario,wherepopulation growth follows the 2010 version of the United Nations projection(UNPD,2011).TheBAUscenarioroughlyreproduces the regional GDPgrowth 2010–2030 asdepicted inthe New Table2–Commoninterpretationsofequity.

Focuson Objecttobeallocated

Costs(obligations) Benefits(rights)

Causesoftheproblem Responsibility(moralresponsibility/‘guilt’inhaving causedtheproblem)

Previouscontributions(tosolvingtheproblem)

Consequencesofthe solution(efforts)

Capabilities(capacitytocontributetoproblemsolving) Needfor(orrightto)thegoodsconcerned

Notes:BuildsonRingiusetal.(2002).The‘previouscontributions’argumentistheleastfrequentlyinvokedandthereforenotconsideredhere.

‘Need’isoftenoperationalizedastheinverseofGDPpercapita,themostcommonindicatorofcapabilities,and–assuch–includedhere.

3Someoftheestimatesofrenewableenergyresourceendow- mentsdonotmeetstrictstandardsofcomparability,socautionis requiredininterpretingtheresults.

Policies Scenario ofWorld EnergyOutlook 2010 (IEA, 2010).

GrowthofCO₂emissionsfromfossilfuelsfollowstotalGHG emissions in the reference scenario of the Climate Action Tracker(2012)byassumingaconstantshareofCO2emissions fromfossilfuelsintotalGHGemissions.Thatshareisalsoused toderiveCO2emissionsfromfossilfuelsinotherscenarios.

Besides the BAU scenario, we consider two mitigation scenarios of the Climate Action Tracker (2012): Scenario 1 representing the ‘More ambitious proposals & national policies not yet pledged internationally’ and Scenario 2 depicting an ‘Illustrative pathway likely holding warming below28C’.Forbothscenarios,wederiveglobalCO2emissions fromfossil fuels2010–2030(Fig.1a). Toachieve globalCO2

reductions in 2030 by 20% and 50% of the BAU levels in Scenarios1and2,respectively(Fig.1b),weassumeaglobal carbon market where regional CO2 prices are equalized through interregional carbon trade. This global market mechanism moves CO2 prices towards US$60 and US$340 pertonneinScenarios1and2,respectively(Fig.1c),reflecting thatmarginalcostsofCO2reductionincreasemarkedlyalong with tighter mitigation targets. Because the GRACE model assumesanidealworldinthesensethatbarriersareabsent, thecost-effectiveregional‘real’emissions(Fig.1d)canalways beachievednomatterwhichdistributiveschemeisselected (seee.g.Hoel, 1991). Indoingso, we isolatetheeffect ofa distributiveschemeonregionalmitigationcosts.

Inthisanalysis,weuseGDPchangescomparedtoBAUto indicate mitigation costs of the two scenarios and of alternativedistributiveschemes.

3. Results and discussion: how much is at stake?

3.1. Fairnessintermsofrelativecontributionstoglobal mitigation

Fig.2andTable 4summarize theresultsregarding relative contributions to global mitigation efforts. Four patterns

emergeasquiterobust.First,forallindicatorsofresponsibility and transformationcapacity,richcountriesmustcontribute muchmoretomitigatingclimatechangethanpoorcountries do.Second,thedistancebetween richandpoorcountriesis smaller for all capabilities-basedindicators(exceptGDPper capita)thanitisforemissions-basedindicatorsofresponsibili- ty.Thedifferencebetweenthesetwosetsofindicatorsreflects technologicaldevelopmentandchangesintheworldeconomy overthepastthreetofourdecades.Third,overall,responsibili- tiesmeasuredintermsofemissionsarenotmuchaffectedby thechoiceoftimehorizon(Spearmanrho=.94**–1.00**)orscope ofcomprehensiveness(.86*–.96**).Similarly,withinoursample ofactors,alltransformation-capacityindicatorsare strongly correlatedwitheachother(.71*–.96**)andtheaggregateindex itselfcorrelatespositivelywithresponsibilitiesmeasuredasper capitaCO2emissions(.82*–.86*).Theonlystrikingexceptionto this pattern isthe renewable energy resourceendowments index,whichaddsatrulydifferentdimensionsofarignoredin mostcapabilityassessments.Fourth,forallactorsexceptIndia, one or two ofour indicators deviate significantly from the prevailing pattern, and for Brazil and Russiadifferences in mitigationcostsarelarge.

This analysisleaves uswith threeimportant messages.

First,mostinterpretationsexaminedhereyieldsimilarresults for most but not all actors. Some policy implications of fairness principles seem, in other words, to be broadly accepted. Second, no single indicator can represent all legitimateinterpretationsoftheprinciplesofresponsibilities and capabilities.Third,nosingledichotomy– suchasthat between‘rich’and‘poor’,or‘developed’and‘developing’–can adequatelycapturethefullrangeofvarianceonresponsibility andcapabilityindicatorsfoundamongthecountriespartici- patingintheUNFCCCnegotiations.

3.2. Fairnessintermsofmitigationcosts

Here,threeobservationsstandout.First,marginalmitigation costsrisesteeplyforallparties(regions)asoverallambitions increase (Fig. 3). In Scenario 1, mitigation costs for the Table3–Twofairnessprinciplesandthe15interpretationsexaminedinouranalysis.

Fairnessprinciple Interpretations Operationalspecifications

Responsibilities TotalCO2emissionsfromfossilfuels:fourinterpretations,ref.

toalternativetimehorizons

Timehorizons:1971–2009;1990–2009;1971–

2017;1990–2017 PercapitaemissionsofCO2fromfossilfuels:four

interpretations,ref.toalternativetimehorizons

Asabove

TotalCO2emissions,includingcarbonembodiedinint.trade:

twointerpretations,ref.toalternativetimehorizons

Timehorizons:1990–2010;1990–2017

Totalemissionsofallgreenhousegases(inCO2equivalents):

oneinterpretation

Timehorizon,1990–2010

Capacities Energysystems’transformationcapacity:oneinterpretation, weightedaggregate(index)

Indexweights:GDP/capita(.6),innovation (.2),governance(.1),transparency(.1) Domesticdistributionofindividualwealth(adultsonly):one

interpretation

Indexdistinguishingfourlevelsofwealth, weighted0(<1000USD)to5(>100,000USD) Renewableenergyresourceendowmentspercapita:one

interpretation,weightedaggregate(index)

Indexweights:Solar(.4),wind(.2),bio(.2), hydro(.2)

Renewableenergyresourceendowmentspercapita+GDP/

capita:oneinterpretation

Indexweights:Renewables/capita(.5),GDP/

capita(.5) Note:Foradditionalinformation,seesupplementarymaterial,TABLESSM1–5.

EuropeanUnion,theUnitedStates,Japan,andChinaare,for allprinciplesandoperationalinterpretations,withintherange of1%ofGDP.InScenario2,correspondingfiguresforthese partiesareabout10timeshigher.Thisdramaticincreaseisnot visibleintheresponsibility-andcapability-basedestimatesof countries’relativecontributionsshowninFig.2.Theincrease is, however, profoundly important for the climate change negotiations,makinganewglobalagreementattheScenario2 levelmuchhardertoreach–andevenmoresotoimplement– thananagreementattheScenario1levelwouldbe.

Second,inbothscenarios,mitigationcostsvaryconsider- ablyamongparties,consistentwithpatternsfoundinother studies(e.g.MattooandSubramanian,2012).Amongdevelop- ing countries,India benefits under all principles and inter- pretations,andChinabenefitsfromsomeinterpretationsbut onlyinScenario2.Bycontrast,Brazilrisksconsiderablelosses

inbothscenarios.Substantialdifferencesarefoundalsowithin thegroupofdevelopedcountries.Inbothscenarios,theUnited States,theEuropeanUnion,andJapansufferonlymodestlosses comparedtoRussia,whichconsistentlystands outwiththe highestrelativemitigationcostsofallsevenparties.Russia’s vulnerability is due toits heavy dependence on fossil fuel extraction.Worthnoticing,ouranalysisindicatesthatmitiga- tioncostscanvaryasmuchwithinthegroupsof‘developing’

and‘developed’countriesasbetweenthesegroups.

Third,inthelessambitiousScenario1,fourparties–China, India,theEuropeanUnion,andtheUnitedStates–areonly moderately affected by the choices examined here among fairness principles and interpretations. Forthe other three parties,mitigationcostsvarymuchmore.Inoursample,Brazil provides the mostdramatic illustration, withdeviationsin GDPrangingfrom 0.6to 4.0%ofBAUinScenario1andfrom Fig.1–Globalmitigationscenarios:(a)GlobalCO2emissionsfromfossilcombustion,(b)globalmitigationtargetsofCO2

emissionscomparedtoBAU,(c)globalCO2prices,and(d)‘real’CO2reductionsin2030comparedtoBAU.

1.8 to 30.5% in Scenario 2. These huge differences are because Brazil qualifies for exemption by responsibility measuredintermsofCO2emissionsbutscores(much)higher on most capabilities indicators. Large differences can be expectedalsoforotherdevelopingcountriesthatqualifyfor fullexemptionby(some)responsibilitycriteriabutthathave at least some ofthe capabilities required to contribute to mitigation.Overall,exemptionrulesarecriticaldeterminants ofobligationsandcosts.

4. Conclusion: outlining a ‘mutual recognition’ approach

Ouranalysisshowsthatthefairnessprinciplesofresponsi- bilitiesandcapabilitiesyieldsimilarresultsformostactors, atleastinScenario1,butnotforall.Itfurthermoreshows

that although most operational interpretations of each principle correlate positively, no single indicator can representalllegitimateinterpretationsofeitherprinciple.

Thesefindings haveimportantimplicationsfortherole(s) thatnormsofdistributivefairnesscanplayintheUNFCCC negotiations.Giventhestarkasymmetriesexistingbetween richand poor,theamount ofcontrolthatmajoremitters haveoveremissioncuts,andtheUNFCCCsystem’slimited capacity to integrate andaggregatedivergent preferences (Keohane and Victor, 2011), search for an integrated

‘fairness-optimizing’formulamaywellincreasetheriskof deadlock.WhattheUNFCCCnegotiationsseemtoneedas attention turns towards bottom-up announcements of voluntary contributions (Kallbekken et al., 2014; Rietig, 2014)isamoreecumenicalapproachthatcanhelpparties buildagreementondiversityandfosterpositivereciprocity throughcooperativebehaviour.

Fig.2–Regionalcontributionstoglobalmitigationaspercentageoftheworldtotal.Note:Thenumbersabovethebarsshow themaximumsharesandthenumbersbelowshowtheminimumshares.Thediscretepointsaremeansharesof correspondingsubgroups.DetailsofthedistributionofcontributionstoglobalmitigationareprovidedinTableSM-1inthe OnlineSupplementaryMaterial.

Table4–Operationalspecificationsofnorminterpretationsgivingthelowestandthehighestmitigationobligations(and costs)foreachcountry.

Actor Lowestobligations/costs Highestobligations/costs

UnitedStates R:allGHGemissions,1990–2010 R:CO2emissionspercapita,1971–2009

EuropeanUnion C:renewableenergyendow.percapita R:CO2emissionspercapita,1971–2009

Japan C:renewableenergyendow.percapita C:individualwealth

Russia R:CO21990–2017,totalconsumption C:renewableenergyendow.percapita

Brazil R:CO2emissionspercapita(exemption) C:renewableenergyendow.percapita

China R:CO2emissionspercapita,1971–2009 R:CO2totalemissions,1990–2017

India R:CO2emissionspercapita(exemption) R:CO21990–2017,totalconsumption

Note:Rindicates responsibilities,Ccapabilities. Italicsmeanthatthis scorediffers substantiallyfromthescores observedforall other indicatorsforthisparticularactor(byafactor<5forthe‘lowest’column,orafactor>2forthe‘highest’column).Totalconsumptionincludes alsocarbonembodiedininternationaltrade.

Amutualrecognitionapproachisdesignedtodojustthat.It doessobynarrowingtherangeoflegitimateprinciplesand interpretationsinaccordancewiththeguidanceprovidedby normativetheory.In addition, the approachbuilds ontwo principalinsightsofsocialscienceresearch.One,pioneeredby Simon(1947), saysthat indealingwith complexproblems, decision-makers usually adopt a technique of (sequential) satisficingratherthan (synoptic)optimizing.Applied tothe UNFCCC negotiations, this insight suggests that fairness principles and interpretations may best be understood as filters blocking‘unfair’ options. Theother insightsays that concernswithfairnessandlegitimacyapplynotmerelytothe outcome but also to procedure and behaviour (e.g. Rawls, 1971).Infact,arecentlypublishedstudyofinternationaltrade negotiations(AlbinandDruckman,2014,p.1)findsthat‘[T]he correlationsbetweenproceduraljusticeandeffectivenessare verystrong,andsignificantlystrongerthanbetweendistribu- tivejusticeandeffectiveness’.Thisisanimportantreminder thatmutualtrustandrespectmustsometimesbegrownand that the process may take years of patient and careful cultivation.Thecoreofthemutualrecognitionapproachis asetofbehaviouralrulesthatmayhelppartiesdoso.

Thefirstandmostfundamentalrulecallsuponallparties toacceptasmallsetofbasicfairnessprinciples,andalimited range of interpretations of each of these principles, as legitimatepremises foraninternationalagreement. ‘Legiti- mate’heremeansbroadlyconsistent(a)withtheFramework

Convention’sCBDR&RCplatform,and(b)withtheguidance provided by relevant normative theory, as summarized in Section 2. Taken together, this set of principles and interpretationsmayserveasacommonandsomewhatelastic framework for working towards a distribution of commit- ments that recognizes ‘... the simultaneous presence of multiplevalid,andsometimesconflictingways,offraminga problem’(BrugnachandIngram,2012,p.61).

Second,indetermining whichprinciplesand interpreta- tions to accept as legitimate, parties apply a standard of reciprocity.Inthiscontext,reciprocityimplies(a)acceptance thatanyfairnessprinciple–andanyinterpretationofsucha principle–thatapartyadvocatesmaylegitimatelybeinvoked byanyotherparty,and(b)recognitionofanyotherprincipleor interpretationthatapartywouldlikelyhavesupportedhadit found itself incircumstances similar tothose ofthe party invoking that principle or interpretation. According to requirement (a), anyone claiming, for example, a right to developmentmustgrantthesamerighttoallothers,including previousandfuturegenerations.Accordingto(b),partiesmust admit that their ownconceptions of fairness are to some extent influenced by self-interest and allow others to be similarlyself-interested.Thesekindsofreciprocityrulesmay berefinedby,forexample,grantingpartiestherighttoclaim exemption from an indicator that yields a particularly unfavourableoutcome,faroutsidetherangedefinedbyother indicators (see Table 3). Conversely, parties may agree to Fig.3–DeviationsofGDPfromBAUin2030:(a)Scenario1and(b)Scenario2.Note:Theverticallinesshowtherangesofthe deviationsforthedistributiveschemesofeachscenarioandthe‘‘X’’sshowthesimpleaverageofdeviationsinthe15 cases.Themostpositivedeviationsareshownabovethelinesandthemostnegativebelow.Detailsoftheregional mitigationcostsareprovidedinTableSM-6intheOnlineSupplementaryMaterial.

constrainrelianceonsingleindicatorsthatyieldexceptionally positiveoutcomes.Suchruleswilllikelybecontroversialbut– ifsupportedbyconferencepresidents,committeechairs,and externalreviewpanels– theycan atleast callattention to importantquestionsofreciprocityandhelppartiesacknowl- edgetheirownbiases.

Third,inassessingalternativeconceptionsofdistributive fairness,partiesrecognizethattheyalloperateunderfeasibility constraints, political aswellas technologicaland economic.

Only measures that pass all these constraints can be effectivelyimplemented.Moreover,asstalematecontinues, partiesfaceanevermorepressingdilemmabetweenfairness andeffectiveness:insistingonimmediateand‘perfect’equity willalmostcertainlyleadtoGHGemissionswellabovethe thresholdIPCCconsidersprudenttoavoid‘dangerousanthro- pogenic interference with the climate system’ (UNFCCC, Article2).Mutualrecognitionoffeasibility constraintsmay helppartiesseemeritsinconstructiveexplorationofagree- mentsinvolvingchangeinrelativecontributionsovertime.

Fourth,sincesomeparties’scoresoncapabilitiesaswellas responsibilitiescriteriachangeovertime,adynamicagreement includingprovisionsforregularupdatingwillhavesignificant advantagesoverstaticarrangements.Whateveritsmeritsat the time of its invention, the frozen dichotomy between Annex I countries and the rest of the world is becoming increasingly inadequate as an expression of present and futurevarianceinresponsibilities,capabilities,andneeds.

Finally,wemakenoclaimthattheapproachoutlinedabove willalwaysbemoreeffectivethanaformulaapproachwillbe.

Clearly,onceagreed,asingleformulacanprovideguidance that is more precise. Moreover, as the European Union experienceindicates,aformulaapproachhasafairchance ofbeingadoptedandsuccessfullyimplementedwhere(a)the group of parties is fairly small and homogeneous, (b) the organizationservingthisgrouphashighinstitutionalcapaci- ty,(c) theimplementation powerbalance tiltsinfavour of enthusiasts,and(d)theproblemitselfscoreslowonpolitical- malignancyscales.Ourmainargumentsare(a)thatUNFCCC negotiations occur in a very different and much more demanding setting, and (b) that because they increasingly turntowardsindividualpledgesofemissionreductiontargets and measures, a ‘matching’ approach is needed to bring principlesofproceduralanddistributivefairnesstoserveas premisesforformulatingandassessingpledges.Themutual recognitionapproachisdesignedspecificallyforsuchhighly demandingsettings wheremutualtrustand understanding mustbecultivatedandconfirmedthroughfairproceduresand cooperativebehaviourvis-a`-visotherparties(seeLejanoand FernandezdeCastro,2014).

Acknowledgements

WegratefullyacknowledgefundingfromtheResearchCouncil ofNorwaythroughGrant209701/E20forCICEP.WethankFrikk Nesje for highly competent research assistance, Robbie Andrews and Glenn Peters for sharing data with us, and FrankAzevedoforexcellentediting.SolveigGlomsrød,Steffen Kallbekken, Ha˚kon Sælen, and two anonymous reviewers haveprovidedveryusefulcommentsandsuggestions.

Appendix A. The GRACE model

Thisstudyusesamulti-sector,multi-regional,recursively dynamicglobalcomputablegeneralequilibriummodelGRACE (Aaheim and Rive, 2005). GRACE stands for the Global Responses to Anthropogenic Change in the Environment.

The modelhasbeen applied tostudies on climateimpact, adaptation, mitigation, and related policy analysis (e.g.

Aaheim et al., 2012; Eskelandet al., 2012; Glomsrød et al., 2013;Rypdaletal.,2007).

Thisversion ofGRACE dividesthe worldinto 8regions:

United States,EuropeanUnion, Japan,Russia,China, India, Brazil, and the Rest of the World. The regional economy includes 15 production activities (Table 3, Glomsrød et al., 2013).ThemodeliscalibratedaroundtheGTAPv7database, with2004asabaseyear(BadriandWalmsley,2008).

In a region, the exogenous endowments of productive resources(i.e.labour,capital,andnaturalresources)arefully usedforproductioninayear.Labourcanflowfreelyfromone activitytoanother,whereascapitalandnaturalresourcesare activity-specific. Producers pursue profitmaximization and consumerspursueutilitymaximization.Bilateraltradeallows substitutionamongregionalcontributions.Regionalincome includestheremunerationforproductiveresourcesandtaxes.

Savingsasafixedshareofincomeareusedforinvestments such that the changes in rates of return on capital are equalizedforall regions.Thenewcapital formedfromthe investmentandcapitaldepreciationinaregionisallocatedto activitiessuch thattheirratesofreturn areequalized. The capital existing at the beginning of the previous year is activity-specific.

Economicgrowthismainlydrivenbysavingsandinvest- ments,butisalsodeterminedbypopulationgrowth,changein the availability of natural resources, and technological change.Theregional ratesoftechnological change arethe sameforallsimulationcases.

Appendix B. Supplementary data

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.

envsci.2015.03.009.

references

Aaheim,A.,Amundsen,H.,Dokken,T.,Wei,T.,2012.Impacts andadaptationtoclimatechangeinEuropeaneconomies.

GlobalEnviron.ChangeHum.PolicyDimens.22,959–968.

Aaheim,A.,Rive,N.,2005.AModelforGlobalResponsesto AnthropogenicChangesintheEnvironment(GRACE), Report.CICERO,Oslo,Norway.

Albin,C.,Druckman,D.,2014.Proceduresmatter:justiceand effectivenessininternationaltradenegotiations.Eur.J.Int.

Relat.20,1014–1042.

Aristotle,350BCE.NicomacheanEthics.ClarendonPress, Oxford(TranslatedbyW.D.Rossin1908).

Badri,N.G.,Walmsley,T.L.,2008.GlobalTrade,Assistance,and Production:TheGTAP7DataBase.CenterforGlobalTrade

Analysis,PurdueUniversity.https://www.gtap.agecon.

purdue.edu/databases/v7/v7_doco.asp.

Baer,P.,2013.Thegreenhousedevelopmentrightsframework forglobalburdensharing:reflectiononprinciplesand prospects.WileyInterdiscip.Rev.:Clim.Change4, 61–71.

Bretschger,L.,2013.Climatepolicyandequityprinciples:fair burdensharinginadynamicworld.Environ.Dev.Econ.18, 517–536.

Brugnach,M.,Ingram,H.,2012.Ambiguity:thechallengeof knowinganddecidingtogether.Environ.Sci.Policy15, 60–71.

Carlsson,F.,Kataria,M.,Krupnick,A.,Lampi,E.,Lo¨fgren,A˚ ., Qin,P.,Sterner,T.,2013.Afairshare:burden-sharing preferencesintheUnitedStatesandChina.Resour.Energy Econ.35,1–17.

Castro,P.,Ho¨rnlein,L.,Michaelowa,K.,2014.Constructedpeer groupsandpathdependenceininternationalorganizations:

thecaseoftheinternationalclimatechangenegotiations.

GlobalEnviron.Change25,109–120.

ClimateActionTracker,2012.GlobalGHGEmissionsin GtCO2eq.ClimateActionTracker..

Dannenberg,A.,Sturm,B.,Vogt,C.,2010.Doequitypreferences matterforclimatenegotiators?Anexperimental

investigation. Environ.Resour.Econ.47,91–109.

denElzen,M.J.,Olivier,J.J.,Ho¨hne,N.,Janssens-Maenhout,G., 2013.Countries’contributionstoclimatechange:effectof accountingforallgreenhousegases,recenttrends,basic needsandtechnologicalprogress.Clim.Change121,397–

412.

Eskeland,G.S.,Rive,N.A.,Mideksa,T.K.,2012.Europe’s climategoalsandtheelectricitysector.EnergyPolicy41, 200–211.

Friman,M.,Hjerpe,M.,2014.Agreement,significance,and understandingsofhistoricalresponsibilityinclimatechange negotiations.Clim.Policy1–19.

Gampfer,R.,2014.Doindividualscareaboutfairnessinburden sharingforclimatechangemitigation?Evidencefromalab experiment. Clim.Change124,65–77.

Glomsrød,S.,Wei,T.,Alfsen,K.,2013.Pledgesforclimate mitigation:theeffectsoftheCopenhagenaccordonCO2

emissionsandmitigationcosts.Mitig.Adapt.Strateg.Global Change18,619–636.

Hoel,M.,1991.Efficientinternationalagreementsforreducing emissionsofCO2.EnergyJ.12,93–107.

IEA,2010.WorldEnergyOutlook2010.InternationalEnergy Agency..

Kallbekken,S.,Sælen,H.,Underdal,A.,2014.Equityand SpectrumofMitigationCommitmentsinthe2015 Agreement.TemaNord2014..

Karlsson,C.,Parker,C.,Hjerpe,M.,Linne´r,B.-O.,2011.Looking forleaders:perceptionsofclimatechangeleadershipamong climatechangenegotiationparticipants.GlobalEnviron.

Polit.11,89–107.

Kartha,S.,Athanasiou,T.,Baer,P.,2012,September.Thenorth–

southdivide,equityanddevelopment–theneedfor trustbuildingforemergencymobilization.In:WhatNext VolumeIII:Climate,DevelopmentandEquity;alsopublished asDevelopmentDialogue.No.61.DagHammarskjo¨ld FoundationandWhatNextForum,Uppsala,Sweden.

Keohane,R.O.,Victor,D.G.,2011.Theregimecomplexfor climatechange.Perspect.Polit.9,7–23.

Klinsky,S.,Dowlatabadi,H.,2009.Conceptualizationsofjustice inclimatepolicy.Clim.Policy9,88–108.

Lange,A.,Lo¨schel,A.,Vogt,C.,Ziegler,A.,2010.Ontheself- interesteduseofequityininternationalclimate negotiations.Eur.Econ.Rev.54,359–375.

Lange,A.,Vogt,C.,Ziegler,A.,2007.Ontheimportanceofequity ininternationalclimatepolicy:anempiricalanalysis.Energy Econ.29,545–562.

Lejano,R.P.,FernandezdeCastro,F.,2014.Norm,network,and commons:theinvisiblehandofcommunity.Environ.Sci.

Policy36,73–85.

Levin,K.,Cashore,B.,Bernstein,S.,Auld,G.,2012.Overcoming thetragedyofsuperwickedproblems:constrainingour futureselvestoameliorateglobalclimatechange.PolicySci.

45,123–152.

Mattoo,A.,Subramanian,A.,2012.Equityinclimatechange:an analyticalreview.WorldDev.40,1083–1097.

Mu¨ller,B.,Ho¨hne,N.,Ellermann,C.,2009.Differentiating (historic)responsibilitiesforclimatechange.Clim.Policy9, 593–611.

Mu¨ller,B.,Mahadeva,L.,2013.TheOxfordApproach:

Operationalizing‘RespectiveCapabilities’.European CapacityBuildingInitiative,Oxford.www.eurocapacity.org.

Parikh,J.,Parikh,K.,2009.ClimateChange:AParkingPlace ModelforaJustGlobalCompact..

Parks,B.C.,Roberts,J.T.,2008.Inequalityandtheglobalclimate regime:breakingthenorth–southimpasse.Camb.Rev.Int.

Aff.21,621–648.

Peters,G.P.,Davis,S.J.,Andrew,R.M.,2012.Asynthesisofcarbon ininternationaltrade.Biogeosci.Discuss.9,3949–4023.

Rao,N.,2013.Internationalandintranationalequityinsharing climatechangemitigationburdens.Int.Environ.Agreem.

1–18.

Rawls,J.,1971.ATheoryofJustice.OxfordUniversityPress, Oxford.

Rietig,K.,2014.Reinforcementofmultilevelgovernance dynamics:creatingmomentumforincreasingambitionsin internationalclimatenegotiations.Int.Environ.Agreem.14, 371–389.

Ringius,L.,Torvanger,A.,Underdal,A.,2002.Burdensharing andfairnessprinciplesininternationalclimatepolicy.Int.

Environ.Agreem.2,1–22.

Rypdal,K.,Rive,N.,A˚ stro¨m,S.,Karvosenoja,N.,Aunan,K.,Bak, J.L.,Kupiainen,K.,Kukkonen,J.,2007.Nordicairqualityco- benefitsfromEuropeanpost-2012climatepolicies.Energy Policy35,6309–6322.

Simon,H.A.,1947.AdministrativeBehavior:AStudyof Decision-MakingProcessesinAdministrativeOrganizations.

BasicBooks,NewYork.

Underdal,A.,Hovi,J.,Kallbekken,S.,Skodvin,T.,2012.Can conditionalcommitmentsbreaktheclimatechange negotiationsdeadlock? Int.Polit.Sci.Rev.33,475–493.

UNPD,2011.WorldPopulationProspects:The2010Revision.

UnitedNationsPopulationDivision,NewYork.

Verweij,M.,Douglas,M.,Ellis,R.,Engel,C.,Hendriks,F., Lohmann,S.,Ney,S.,Rayner,S.,Thompson,M.,2006.

Clumsysolutionsforacomplexworld:thecaseofclimate change.PublicAdm.84,817–843.

Victor,D.G.,2011.GlobalWarmingGridlock:CreatingMore EffectiveStrategiesforProtectingthePlanet.Cambridge UniversityPress,Cambridge.

Winkler,H.,Letete,T.,Marquard,A.,2013.Equitableaccessto sustainabledevelopment:operationalizingkeycriteria.

Clim.Policy13,411–432.

Winkler,H.,Rajamani,L.,2014.CBDR&RCinaregimeapplicable toall.Clim.Policy14,102–121.