The Political Economy of Weak Treaties PDF
Preview The Political Economy of Weak Treaties
The Political Economy of Weak Treaties Marco Battaglini CornellUniversityandEinaudiInstituteforEconomicsandFinance Ba˚rd Harstad UniversityofOsloandFrischCentreforEconomicResearch Inrecentdecades,democraticcountrieshavenegotiatedhundredsof internationaltreatiesandagreements.Thispaperanalyzestheequilib- riumdesignoftreatiesnegotiatedbypoliticalincumbentsseekingre- election.Weshowthatincumbentsarepronetonegotiatetreatiesthat are“weak,”inthattheymayormaynotbecompliedwith:thismakesit possibletodifferentiatethealternativecandidatesinawaythatfavors the incumbent. We also show that political economy considerations leadtooverambitioustreatiesthatrelytoomuchontechnologyinstead ofsanctionstomotivatecompliance.Ourtheorycanrationalizeseveral puzzlesassociatedwithtreaties. I. Introduction The presence of public goods and externalities has always been one of thefundamentalmarketfailuresmotivatinggovernmentalintervention. Analogously,cross-borderexternalitiessuggestthatregulatoryauthority Wehavebenefittedfromcommentsbytheeditor,threereferees,GuriBang,Stephen Coate,JonHovi,MassimoMorelli,EleonoraPatacchini,TorstenPersson,RagnarTorvik, andseminarparticipantsatCornell,theUniversityofGeneva,Harvard,theLondonSchool ofEconomics,theNorwegianUniversityofScienceandTechnology(NTNU),Rice,the StanfordGraduateSchoolofBusiness,theStockholmSchoolofEconomics,UppsalaUni- versity,the2018WorldCongressofEnvironmentalandResourceEconomistsmeeting,the 2017NottinghamInterdisciplinaryCentreforEconomicandPoliticalResearchconference, the2016EuropeanEconomicAssociationmeeting,andthe2016Warwickpoliticaleconomy conferenceinVenice.WealsothankAngelaCools,TorjeHegna,andEstebanMendezChacon ElectronicallypublishedJanuary16,2020 [JournalofPoliticalEconomy,2020,vol.128,no.2] ©2020byTheUniversityofChicago.Allrightsreserved.0022-3808/2020/12802-0003$10.00 544 political economyof weak treaties 545 shouldbecentralizedorthat,attheleast,countrieswouldbenefitfrom negotiatinginternational treaties.Externalitiesarethusexpectedtoin- fluencethedesign ofpoliticalinstitutions whentheinstitutionsareen- dogenous,andtheyarecertainlyendogenouswhenitcomestointerna- tionaltreatiesandbodies,asthesearebeingnegotiatedanddesignedat multiplehigh-levelpolicymeetingseveryyear.Inthepastfewdecades,de- velopedanddevelopingcountrieshavenegotiatedhundredsofinterna- tionalenvironmentalagreements(IEAs),forexample.1 Economictheoryhasfailedinexplainingtheobservedpatternoftrea- ties,inourview.Aslongasthereisnosupranationalgovernmentinplace, anindividualcountryhasanincentivetofreerideinsteadofparticipating inmultilateral agreements.Afterall,manyoftheagreementsintendto provideregionalorglobalpublicgoods.Environmentalagreements,for example,havetargetedawiderangeofgoals,fromforestpreservationand water management to the regulation of transboundary pollution. Since standard game theory predicts free riding and small cooperative coali- tions,theriseofIEAsisapuzzletomanyeconomists.Inasurveyonthe“eco- nomicsofclimatepolicy,”KolstadandToman(2005,1605)refertothe riseofIEAsasthe“paradoxofinternationalagreements.”2 TwofeaturesofIEAsthathavesofarattractedlittleattentionsuggest that the paradox should be qualified. The first is the fact that IEAs are surprisinglyweakagreements:theygenerallydonotincludeeffectiveen- forcementormonitoringmechanisms. Thelackof enforcementis only partiallyexplainedbythelackofthird-partyenforcementinglobalpoli- tics; after all, the countries could sign treaties where noncompliance is metbytradesanctions(asintradeandarmscontroltreaties).3Thesecond strikingfeatureofIEAsisthatmanyofthem,includingsomeofthemost prominent,aregenerallyseenasineffective.4Thesetwofactssuggestthat foroutstandingresearchassistanceandFrankAzevedoforhelpontheediting.Harstad’s partoftheresearchreceivedfundingfromtheEuropeanResearchCouncil(ERC)under theEuropeanUnion’sHorizon2020researchandinnovationprogram(grantagreement no.683031). 1 ExamplesofthistypeofnegotiationaretheprotocolssignedundertheConventionon Long-RangeTransboundaryPollution,whichattemptstoreducesulfurandotherhazard- ous emissions having transboundary effects, or those signed under the United Nations FrameworkConventiononClimateChange,whichcommitsstatepartiestoreducegreen- housegas(GHG)emissions. 2 Naturally,alargebodyofliteraturehasbeendevotedtohighlightingandexplaining this paradox. See, e.g., Carraro and Siniscalco (1993), Barrett (1994), Dixit and Olson (2000),andBattagliniandHarstad(2016).Wereviewthisliteraturemoreextensivelyat theendofthissection. 3 TheMontrealProtocolof1997regulatingchlorineemissionsdamagingtheozonelayer, forinstance,didindeedpermittradesanctionstobeimposedonviolators.IPCC(2014, 1016)discussestradesanctionsforclimateagreementsandalsosuggeststhat,asanalterna- tive,“asanctioncouldtaketheformofatemporarysuspensionofmonetaryandtechnolog- icaltransfersifrecipientcountriesarefoundinnon-compliance.” 4 FinusandTjøtta(2003)andRingquistandKostadinova(2005)findthattheHelsinki andOsloProtocolshavenotgeneratedemissionreductionbeyondthelevelsthatwould 546 journal ofpolitical economy theparadoxmaybethatsomanycountriesarenegotiatingandsigning weakagreements,ratherthanthenumberofagreementsitself.Negotiat- ingtreatiesisanexpensiveandlaboriousprocess;signingtreatiesthatare eithernotratified(aswasthecasewiththeUnitedStatesandtheKyoto agreement)orratifiedandthenrenegedon(aswasthecasewithCanada andtheKyotoagreement)isevenmoredamaging.Wemaycallthisthe “paradoxofweakagreements.” It is immensely important to understand these puzzles and to shed lightonwhytreatiesarenoteffectivelyaddressingtheworld’smostchal- lengingproblems.Inourview,arealisticanalysisshouldaccountforthe factthatnegotiationsareheadedbypublicofficialsandpoliticianswho maybesubjecttoelectoralconcerns.Domesticpoliticaleconomyconsid- erations are of fundamental importance to any politician, and they are thuslikelytoinfluencetheequilibriumdesignofinternationaltreaties. Ourmainresultisthat,inthepresenceofsufficientlystrongreelection concerns,politicalincumbentsbenefitfromnegotiatingweakagreements thatleavetheultimatedecisiononcompliancetothewinnersoffutureelec- tions.Thus,equilibriumtreatiesarecharacterizedbyenforcementmech- anismsthatarelesseffectivethanoptimalandareindeedrepudiatedwith positiveprobability.Interestingly,thisisageneralphenomenonthatdoes notdependonthepreferencesoftheincumbentgovernmentthatnego- tiatestheagreements:relatively“green”and“brown”governmentsalike are affected by it. On the one hand, these political economy consider- ationsexplaintheunderprovisionofinternationalcooperationbyratio- nalizing weak agreements when strong agreements would be optimal. Ontheotherhand,theelectoralconcernsmayinducegovernmentstone- gotiate agreements even when having no agreement would be optimal. Thetheorycanthusrationalizewhydemocraticcountrieshavenegotiated hundredsofIEAsinthepastfewdecadesandwhytheremightbeanover- supplyofweaktreaties.Thisresultcomplementstheviewthat“political failures”weakenthecaseofgovernmentalregulationinthepresenceof marketfailures,byshowinghowstrategicpoliticsalsolimittheefficiency ofinternationaltreaties. Inourbenchmarkmodel,thepoliticalincumbentinthehomecoun- trynegotiatesatreatywithaforeigncountry(oragroupofforeigncoun- tries).Thetreatyisconsideredbecausethehomecountry’sactiongener- atesnegativeexternalitiesonothers.Ourmechanismpermits—butdoes notrequire—theretobeasymmetricexternalityfromtheforeigncoun- try onto the home country. The treaty specifies what the home country ought to do to reduce the externalities as well as the consequence if it have been achieved without an agreement. Aakvik and Tjøtta (2011) find no evidence fortheeffectivenessoftheHelsinkiandOsloagreementsinreducingsulfuremissions. Vollenweider(2013)findsnoevidenceofnetenvironmentalbenefitsfortheGothenburg Protocolof1999. political economyof weak treaties 547 doesnot.Afterthenegotiation,anelectiondecideswhethertheincum- bentpartycontinuesto beinchargeoris replaced.Atthis stage,voters discernwhichpartyisbetter,giventhetreatythatisnegotiatedinthefirst period: the relatively “green” party, which has more environmentally friendlypreferencesthanthemedianvoterdoes,ortherelatively“brown” party, which has less environmentally friendly preferences than the me- dian voter does. At thelast stageof thegame, theelectedpartydecides whetheror notto comply withthe treaty, facing theoptionsnegotiated atthefirststageofthegame.We haveastrongtreatyif,nomatter which partyisinpowerinthefollowingperiods,theagreementiscompliedwith. Wehaveaweaktreatyifitincludessanctionsthatarenotsufficientlystrong to guarantee compliance (and so may be violated if the brown party is elected). Weusethissimplemodeltostudyhowelectoralincentivesshapethe typeofagreementthatissigned(weakvs.strong),thesizeandscopeof the agreement, and the incentives to invest in green technologies. Re- garding the type of agreement, we first show that signing an IEA may ormaynotbeoptimalfromasocialpointofview(dependingonthepref- erencesandthecostoftheenvironmentalpolicy);however,iftheIEAis signed,itshouldalwaysbestrong.Nevertheless,whenreelectionincen- tives are sufficiently important, the equilibrium IEA is always weak and thusnotalwayscompliedwith,regardlessofwhetherthefirst-periodin- cumbentwasgreenorbrown.Tounderstandtheintuitionbehindthisre- sult,notethatwithnoagreementorwithastrongagreement,theincum- bentandthechallengerareidentical(inthisrespect)fromthemedian voter’spointofview,becausetheywouldbehaveinthesamewayafterthe election:inthefirstcasebecausetherewouldbenoagreementtoimple- ment,inthesecondbecausebothofthemwouldimplementtheagree- ment.Whenthetreatyisweak,however,theagreementisenforcedonly ifthegreenpartyiselected.Thekeyinsightofouranalysisisthattheme- dianvoter’spreferencesdependonthedetailsoftheagreements:theme- dianvoterpreferscomplianceifthesanctionisrelativelyseverebut not otherwise.Usingthisinsight,weshowthatbothpartiescandesignaweak agreementthatgivesthemanadvantageintheelection.Thegreenparty designs a weak treaty in which the median voter wants implementation ex post and implementation is guaranteed only if the incumbent is re- elected;thebrownpartydesignsatreatyinwhichthemedianvoterdoes notwantimplementationandimplementationcanbeavoidedonlyifthe incumbentisreelected. Regarding the depth of treaties, we show that electoral incentives in- duce a novel “overshooting effect,” according to which the incumbent tendstomakeenvironmentalcommitmentsthat,besidesbeingweakasdis- cussedabove,arelargerthanwhatwouldbechosenwithoutelectoralin- centives.Thisphenomenon,again,isremarkablebecauseitcharacterizes 548 journal ofpolitical economy bothgreenandbrownincumbents.Asweexplainmoreextensivelyinsec- tionIII.A,thisphenomenonoccursbecausetheincumbentattemptsto compensatewithsizeforthefactthatthetreatymightnotbefullycom- pliedwith.Thelargerthepoliticalofficerentis,theweakerthetreatyis inthismodel. We also endogenize the compliance cost by allowing for investments intechnologies.Sinceinvestmentsreducethecostofcomplianceinthe second period, the treaty is “strong” (in that it will always be complied with) ifthe first-period incumbent invests a lot, but thetreaty is “weak” (inthatonlythegreenpartywillcomplyinthesecondperiod)ifthein- vestment level is lower. This extension demonstrates that weak treaties emerge in equilibrium even if the sanction level is exogenous, low, or identicaltozero.Thisextensionalsoallowsustodistinguishbetweenex- ternalenforcement(i.e.,asanction)andinternalenforcement(i.e.,tech- nology) and to predict how the two instruments will be combined. We showthat,evenifastrongtreatywithsanctionsisfirstbest,thelikelypo- liticaleconomyequilibriumisaweaktreatythatis(partially)enforcedby technology. Finally,weshowthatourresultsgeneralizetoafullydynamicsetting. Indeed,weshowthatwhenweakIEAsareexpectedinthefuture,anin- cumbent maybeeven morelikely select aweak treaty today. Ourworkconnectsandcontributestotwostrandsofliterature:thelit- eratureonenvironmentalagreementsandtheliteratureonthepolitical economyofcommitments.Theimportanceofpoliticaleconomyandre- election concerns has certainly been established when they regard do- mestic political decisions. For example, Persson and Svensson (1989), AghionandBolton(1990),andAlesinaandTabellini(1990)havehigh- lightedhowpublicdebtcanbeusedinthissensetolimitexpendituresof futuregovernments;BesleyandCoate(1998)studyhowfiscalpolicyin- vestmentsinpublicinfrastructurecanbeusedtoaffecttheoutcomeoffu- tureelections;BiaisandPerotti(2002)showhowprivatizationcanbeused to manipulate the preferences of the median voter; and Robinson and Torvik (2005) argue that inefficient local infrastructures may intend to influenceelections.Thus,thereisasolidtraditionforassumingthatre- electionconcernsinfluencepoliticaldecisions.Thepapersabovefocus ondomesticpoliticaldecisions,however.Weconnectthedomesticreelec- tionconcerntodecisionsintheinternationalarenainordertoexplain theobservedpuzzlesontreaties. Arecentpaperthathasstudiedtheeffectofelectionsonthesuccessof international treaties is by Buisseret and Bernhardt (2018). They con- sider a two-period model in which period 1 has a weight in the agents’ utilitiesequalto12dandperiod2aweightequaltod.Focusingonthe timingofelectionsasmeasuredbyd(theydefineanelectiontobe“close” to the signature of an international agreement if d is large), they argue political economyof weak treaties 549 thatwhetheranagreementissigneddependsonhowhostiletheincum- bent is and how close an election: if the incumbent is hostile, then the agreementissignedonlyiftheelectionisnotveryclose;iftheincumbent isfriendly,thenoneissignedonlyiftheelectionissufficientlyclose.Our analysisdiffersinthatwefocusonhowtheinternationalagreementisde- signed(weakvs.strong),andwehighlightageneraltendencytowardweak agreements,independentofthepreferencesoftheincumbent.5 Thetraditionalliteratureoninternationalagreementshasstudiedthe incentivesforcountriestoparticipateinthepresenceoffreeriding(Hoel 1992; Carraroand Siniscalco 1993;Barrett 1994, 2003; Dixit and Olson 2000).6Thetypicalpredictioninthisliteratureisthatthecoalitionsizeis very small, because of the benefits of free riding. This prediction has theeffectofmotivatingtheviewthattherisingnumberofIEAsisapara- doxicalphenomenon.Twoassumptionshavecharacterizedmostofthese analyses:first,thatcountriesactasindividualagentswithnointernalpol- itics and second, that once established, IEAs fully enforce their provi- sions.Bothassumptionsarerelaxedinourpaper. InBattagliniandHarstad(2016),weshowedthatincompletecontracts canbebeneficial,astheyhelpinducemorecountriestoparticipate.We followedtheliteratureinassumingthatoncesigned,thetreatyisalways ratified and respected. The incompleteness of the IEAs referred to the factthatagreementsdeliberatelydonotregulateimportantaspectsofthe environmentalissue,suchasinvestmentsingreentechnologies.Inthepres- entpaper,westudyarelatedbutdifferentissue,namely,whycountriesne- gotiatetreatiesspecifyingactionsthattheyknowwillnotnecessarilybere- spected, even when they could sign“strongcontracts” thatwouldalways berespected.Werefertotheseas“weaktreaties,”aconceptthatisdistinct fromtheconceptofincompletecontracts.7Furthermore,thekeymecha- nismofthepresentpaperisthatthetreatyisdesignedbyself-interested politicians to influence elections. Domestic politics were absent in the 5 Another paper that has explicitly analyzed the relationship between international agreementsandelectionsisbyPerssonandTabellini(1992).Theyareinterestedintheef- fectsofEuropeanfiscalpolicyintegrationandnotethattheincreaseincapitalmobilityas- sociatedwithEuropeanintegrationinduces,ceterisparibus,areductionofcapitaltaxation. Anticipatingthis,voterstendtoelectpolicymakerswhomaychoosealeveloftaxationthat ishigherorlowerthanthelevelpreferredbythemedianvoter.Theauthorsmodelinstitu- tionalchangesasanexogenousreductioninthecostofcapitalmobility,sotheydonotex- plainthedesignoftheinternationalagreement. 6 SeeMaggiandMorelli(2006)forastudyofself-enforcinginternationalorganizations inamoregeneralcontext.Harstad,Lancia,andRusso(2018)analyzehowtechnologyin- vestmentsmaketreatiesself-enforcinginarepeated-gamecontext,whiletechnologysolves atimeinconsistencyprobleminHarstad(2018). 7 Inthepresentpaper,weakIEAsarenotatallincomplete,sincetheyincludealltherel- evantaspectsoftheagreement.Inthebaselinemodelofsec.II,greeninvestmentsarenot allowed,sothecontractsignediscompletebydefinition.Intheextensioninsec.III.B,we allowforgreeninvestments,andalsotheseinvestmentsarepartoftheagreement 550 journal ofpolitical economy 2016paper,wherethedrivingforcewashold-upproblemswhencountries negotiated. Recent research has also endogenized the government’s preferences byallowingforstrategicdelegationorlobbygroups’influence.8Thereis naturallyalsoalargebodyofliteraturestudyingtherelationshipbetween internationalandnationalpoliticsmoregenerally.Ineconomics,interna- tional cooperation has sometimes been viewed as collusion between in- cumbents,ruiningbeneficialtaxcompetition(Rogoff1985;Kehoe1989), whileelectionsallowvoterstodelegatestrategicallybeforepoliciesareset ornegotiated(PerssonandTabellini1995surveytheearlyliteratureon suchdouble-edgedincentives).Inpoliticalscience,so-calledtwo-levelgames havebeenanalyzedinwhichnationsnegotiatebeforethetreatymustbe ratifieddomestically(Putnam1988;Evans,Jacobson,andPutnam1993). Putnamstressedthatdomesticconflictsbetweendifferentpartiesarenec- essaryforinternationalagreementsandtheirratificationstosucceed,since oneparty,oftentheminority,canthencolludewiththeforeigncountryto haveapolicyimplementedthatneitherofthetwowouldhavebeensuc- cessfulatimplementingalone.9Weshowthatevenwhenalldomesticpar- tiesfindthepolicycostly,theagreementmaystillbesigned—anddesigned inaninefficientwayinordertoinfluencefutureelections. Inthispaper,weattempttoshiftthefocusoftheliteraturefromsimply explainingparticipationinaself-enforcingagreementandinsteadana- lyzetheverynatureoftheagreement.Thisanalysisnotonlyrationalizes thestylizedfactsmentionedabovebutalsoopensanumberofnewques- tionsthathavenotbeenstudiedtodate. Thepaperisorganizedasfollows.Thenextsectionpresentsthebench- markversionofourmodel,inwhichtreatyandabatementdecisionsare zero-onevariables.Wederiveourmainresultsinthispedagogicalsetting, inwhichtheunderlyingintuitionismosttransparent.SectionIIIextends thisbasicmodelinthreedirections:insectionIII.A,weallowthecoun- tries to choose the depth and scope of the negotiation; section III.B 8 On“strategicdelegation,”researchershavestudiedhowvoters(oragenericprincipal) choosethecharacteristicsofthenegotiatorwhenbargainingoverenvironmentalprotection inordertogainabargainingadvantage:see,forinstance,PerssonandTabellini(1992), Segendorff(1998),Eckert(2003),Buchholz,Haupt,andPeters(2005),andHarstad(2008, 2010).Onlobbying,seeGrossmanandHelpman(1994),Haffoudhi(2005),Altamirano- Cabrera,Weikard,andHaffoudhi(2007),orDietz,Marchiori,andTavoni (2012). Maggi andRodríguez-Clare(2007)examinehowtradeagreementscanbeusedascommitmentde- vicestolimitdemandsfromlobbyists.Ourcontributiontothisliteratureistoanalyzehow electoralconcernsinfluenceandexplainthedesignofinternationaltreaties. 9 ArelatedlineofworkhasbeenpursuedbyFearon(1998a),whohasstudiedarmscon- trolagreementsastwo-stepprocessesinwhichfirstadealisnegotiatedinawarofattrition andthenitisimplementedinarepeated“enforcementgame.”Ratherthanstudyingthe strengthoftheresultingdeals,Fearonfocusesontheeffectofthetimehorizononthe lengthofthenegotiations.SeealsoFearon(1998b)forageneralreviewoftheliterature oninternationalrelations. political economyof weak treaties 551 allowsforinvestmentsingreentechnologyandrelatestheirchoicetothe strength of the treaty and the choice of sanctions; and finally, in sec- tionIII.C, when the time horizon is infinite, we show that weak agree- ments may be more likely to emerge today if they are also expected in thefuture.SectionIII.Dbrieflydiscussesotherextensions.SectionIV.A discusseshowtheanalysisshedslightontheroleofdomesticpoliticsin shaping international negotiations inthreerecent examples, including theKyotoProtocolandtheParisAgreement.SectionIV.Bpresentsafirst attempttotestsomeofthepredictionsofthetheorybyusingalargepanel ofenvironmentaltreatiessignedinthepast40years.Afteraconcluding section,theappendixpresentsimportantproofs,whileanonlineappen- dixpresentsotherproofsanddetailsthedata. II. The BasicModel and Result A. The Model Webeginouranalysisbypresentingasimpleworkhorsemodelwithtwo periods and two sets of countries: the home country H and the foreign country F (F can be the set of other countries). Country H’s action, or “emission,”generatesanexternalitye ≥ 0ontheforeigncountry.Coun- tryH,however,canabatepollutionandeliminatetheexternalitybyincur- ringacost.AlthoughabatementmayhavesomevaluealsotoH,theresults aresimplertoexpressifweassumethatthenetcostofabatingispositive for political parties. Section III generalizes the model and allows abate- menttobenonbinary,amongotherthings. Thetwocountriescannegotiateatreaty.Thetreatyspecifieswhether Hshouldabateandtheconsequenceifitdoesnot.Iftheconsequenceis just a reputational loss, then its severity would depend on whether the treatyislegallybinding,andsoon,buteventradesanctionsontreatyvi- olatorsarepermittedbytheMontrealProtocol.10Thecostofthesanction toHiss ≥ 0,andF’scostofimposingthesanctionisgs.Ifg > 0,thenF dislikesimposingthesanction(as,e.g.,whensisimposedbyrestricting tradewithF).Ifg < 0,thenFbenefitsfromimposingthesanction,per- hapsbecauseittakestheformofamonetarytransfer.Weallowgtobepos- itiveornegative,butweassumethatg ≥ 21,sothatthereisadeadweight lossð11gÞs ≥ 0whenthesanctionisimposed.11 Bothwhennegotiatingthetreatyandwhendecidingwhethertocom- ply,thehomecountry’sdecisionsaremadebyoneoftwopoliticalparties. Partiesandvotershaveheterogeneouspreferencesregardingenvironmen- talconservationandregulation.Thus,thenetcost(i.e.,theregulatorycost 10 SeeArticle4oftheProtocoland,foramoreextensivediscussion,Barrett(2003).See alsothediscussiononsanctionsbyIPCC(2014),mentionedinsec.I. 11 Naturally,ifthesanctionisapuremonetarytransfer,thenweshouldexpectg 521. 552 journal ofpolitical economy minus the environmental benefit) of additional abatement is c > 0 for G thepoliticalpartythatisrelativelygreen,whileitisc > c forthepolitical B G partythatisrelativelybrown.Thecostforthemedianvoter,M,isinbe- tween: c ∈ðc ,c Þ. Of course, there may also be parties and elections M G B abroad,buttheywillnotbeimportantforouranalysis. Thetimingofthegameisasfollows.First,inperiod1,F’sandH’sin- cumbentgovernmentsnegotiates.Second,anelectiondetermineswhether theincumbentremainsinpowerorisreplaced.Finally,thewinnerofthe electiondecideswhethertocomplyorfacethesanctions.Wenowexplain eachstepinmoredetail. 1. TheNegotiations Wemaketwoimportantassumptionsaboutthenegotiationsinperiod1. First,weassumethatthetwopartiescanusesidetransferswhennegoti- atingthetreaty.Thisimpliesthattheequilibriumlevelofswillsimplybe thesthatmaximizesthesumoftwonegotiators’expectedpayoffs.Anad- vantageofthisassumptionisthatthemodelremainsrelevantwhetheror not there is also a symmetric problem where F emits, harming H. If we assumethattheeffectofF’semissionsontheharmgeneratedbyH’semis- sionsisnottoolarge,thenthetwoproblemscanbeseparatedandconsid- eredindependently.12Second,weassumethatHandFarefullycommit- ted toimposingthesanctionifH doesnotcomply.SectionIII.B proves thatinvestmentintechnologyisonewayoffacilitatingcommitment,and section III.C shows when complying with the sanction is incentive com- patibleinadynamicframework,whilesectionIII.Darguesthatourresults continuetoholdevenifscanberenegotiated. 2. Elections After the treaty has been negotiated, there is an election in the home country.Theoutcomeoftheelectionisdeterminedbythemedianvoter, M,whovotesforthecandidatedeliveringthehighestexpectedpayoff.Spe- cifically,Mreelectsthefirst-periodincumbenti ∈fB,Ggifui 2u2i > d, M M whereui (u2i)isM’sexpectedpayoffwhenelectingi(2i ≠ i),whiledis M M somerelativepopularityshockinfavorofthechallenger.Thepopularity shock, realized after the treaty is signed, can refer to the importance ofotherpolicydifferencesnotexplicitlymodeledhere.Weassumedto be uniformly distributed on ½2z=j,ð12zÞ=j(cid:2), where z ≥ 1=2, implying boththatthedensityoftheshockisjandthattheincumbentwinswith 12 This assumption may be realistic for climate change, where the marginal long-run harmofemissioncannotbeinfluencedmuchbyanindividualcountry’sshort-termemis- sionlevel.Theassumptionishardertodefendforsecurityandarmscontrolagreements, however. political economyof weak treaties 553 probabilityz ≥ 1=2ifui 5 u2i.Theincumbencyadvantageistherefore M M measuredbyz2ð1=2Þ ≥ 0.Westartbyassumingthatthevarianceinthe shockissufficientlylargethatreelectionprobabilitiesareinteriorin(0,1). Asisshownbelow,thispropertyisguaranteedifthedensityoftheshock issosmallthat (cid:1) (cid:3) 12z 12z j < min , : (1) c 2c c 2c B M M G 3. Compliance Atthefinalstageofthegame,thenewly(re)electedpolicymakerj ∈fB, Ggdecideswhethertocomplywiththetreaty.Bycomparingthetwocosts, thesecond-periodincumbentfindsitoptimaltocomplyifandonlyifthe sanctionsislargerthanthecosttoj,c.Ifs >(cid:1)s ; c > c ,bothofthepar- j B G tieswillcomplywiththetreaty,sowehavewhatwecallastrongtreaty.If insteads < s ; c < c ,thennoneofthepartieswillcomplywiththetreaty, G B sowehaveanineffectivetreaty.Ifs ∈½s,(cid:1)s(cid:2),thenthetreatywillbecomplied withifthesecond-periodincumbentisG,butnotifBisinpower.13Since thistreatymayormaynotbecompliedwith,wenameitaweaktreaty. 4. Objective Functions Thepayoffsareinlinewiththediscussionabove.IfHcomplies,thenF receivese > 0whileeveryi ∈fB,M,Ggpaysthecompliancecostc > 0.If i H does not comply, then F imposes the sanction at cost gs, where s > 0 measures the cost for every individual in H. In addition, the second- period incumbent j ∈fB,Gg enjoys the office rent R ≥ 0 as the benefit ofstayinginoffice.(Asimilarofficerentforthefirstperiodissunkand wouldnotinfluencetheanalysis.)Theproofsintheappendixallowtheof- ficerent,R,tobeconditionedontheidentityofthesecond-periodincum- bentj,andtheonlineappendixpermitstheofficerenttobeconditioned onwhetherjcomplies.Thesecontingenciesdonotinfluencethebasicre- sult,andtheyarethusabstractedfromhere. Payoffs Condition M Opponenti Incumbentj F Hcomplies 2c 2c 2c 1R e M i j Hdoesnotcomply 2s 2s 2s1R 2gs 13 Note that G has multiple best responses when s5s, while B has multiple best re- sponseswhens5(cid:1)s.However,itisshownbelow(seen.15)thatineverysubgame-perfect equilibrium(SPE),Gcomplieswhens 5s,whileBwillnevercomplywhens 5(cid:1)s.Thus, thesetofweaktreatiesisclosedinequilibrium.
Description:The list of books you might like
