WORKSHOPPROCEEDINGS

AssessingtheBenefitsofAvoidedClimate
Change:CostBenefitAnalysisandBeyond

AddressingClimateChangethrough
aRiskManagementLens

GaryW.Yohe
WesleyanUniversity

May2010


ThisworkshopwasmadepossiblethroughagenerousgrantfromtheEnergyFoundation.
EnergyFoundation
301BatterySt.
SanFrancisco,CA94111

WorkshopSpeakers
DavidAnthoff,EileenClaussen,KristieEbi,ChrisHope,RichardHowarth,Anthony
Janetos,DinaKruger,JamesLester,MichaelMacCracken,MichaelMastrandrea,
SteveNewbold,BrianONeill,JonORiordan,ChristopherPyke,MarthaRoberts,
SteveRose,JoelSmith,PaulWatkiss,GaryYohe

ProjectDirectors
SteveSeidel
JanetPeace

ProjectManager
JayGulledge

ProductionEditor
L.JeremyRichardson

ContentEditors
JayGulledge,L.JeremyRichardson,LiwaywayAdkins,SteveSeidel

SuggestedCitation
Yohe,G.W.2010.AddressingClimateChangethroughaRiskManagementLens.In
AssessingtheBenefitsofAvoidedClimateChange:CostBenefitAnalysisandBeyond.
Gulledge,J.,L.J.Richardson,L.Adkins,andS.Seidel(eds.),ProceedingsofWorkshop
onAssessingtheBenefitsofAvoidedClimateChange,March1617,2009.Pew
CenteronGlobalClimateChange:Arlington,VA.p.201231.Availableat:
http://www.pewclimate.org/events/2009/benefitsworkshop.

Thecompleteworkshopproceedings,includingvideoof17expertpresentations,thissummaryreport,
andindividualoffprintsofexpertpapersareavailablefreeofchargefromthePewCenteronGlobal
ClimateChangeathttp://www.pewclimate.org/events/2009/benefitsworkshop.

May2010

PewCenteronGlobalClimateChange
2101WilsonBlvd.,Suite550
Arlington,VA22201

202 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop

AddressingClimateChangethrough
aRiskManagementLens

GaryW.Yohe
WesleyanUniversity

Abstract
IntheSummaryforPolicymakersoftheSynthesisReportforitsFourthAssessment,the
IntergovernmentalPanelonClimateChange(IPCC)achievedunanimousagreementfrom
signatorycountriesoftheFrameworkConventionthat,Respondingtoclimatechange
involvesaniterativeriskmanagementprocessthatincludesbothadaptationandmitigation,
andtakesintoaccountclimatechangedamages,cobenefits,sustainability,equityand
attitudestorisk(IPCC2007c,pg22;emphasisadded).Byacceptingthiskeysentence,
governmentsrecognizedforthefirsttime,thattheirnegotiationsandassociatedpolicy
deliberationsmust,individuallyandcollectively,beinformedbyviewsoftheclimate
problemdrawnthroughthelensofreducingrisk.
Asnewasthisperspectivemightbeforclimatepolicynegotiators,riskmanagementis
alreadywidelyusedbypolicymakersinotherdecisionmakingprocesses,suchasdesigning
socialsafetynetprograms,monetarypolicy,andforeignpolicy.Eventhoughgovernments
andsomesegmentsofthepolicycommunityarecomfortablewiththeriskmanagement
paradigm,however,theclimatechangeresearchandassessmentcommunityhad
heretoforebeenslowtocatchon.Thispaperpresentsafirstattempttodeconstructthe
applicationofariskbasedparadigmtoclimatechangebyconsideringthecriticalphrases
thatarehighlightedabove,offeringinsightsintowhatwedoanddonotknowineachcase.
Perhapsmostimportantly,thetypicalcostbenefitanalysisusedtomakedecisionsin
establishingregulationsmaynotbefullyappropriatefortheclimateproblembecause,toa
largedegree,manydamagescannotbeexpressedmonetarilyandbecauseuncertaintyisso
pervasive.Toavoidbeinghamstrungbythesefundamentalcomplications,traditional
policyanalysesneedtobesupplementedbyriskbasedexplorationsthatcanmore
appropriatelyhandlelowprobabilityeventsandmoreeasilyhandlelargeconsequences
calibratedinnonmonetarymetrics.Inshort,adoptingariskbasedperspectivewillbring
newclaritytoourunderstandingofthediversityandcomplexityoftheclimateproblem.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 203


Introduction
SincetheSynthesisReportoftheFourthAssessmentReport(IPCC,2007c)wasapproved
bytheIntergovernmentalPanelonClimateChangeinNovember2007,muchoftheworlds
attentionhasfocusedontheeconomiccostsofmitigation,speciesextinctions,extreme
weatherevents,andotherimpactsthatwerehighlightedinthepreviouslyapprovedreport
ofWorkingGroupsIIandIII(IPCC,2007a&2007b).Althoughtheseareimportant
examplesofthesocietyandnaturesvulnerabilitytoclimatechange,thekeypolicy
developmentoftheFourthAssessmenthasreceivedlittleattention:inafewparagraphs
thatappeartowardtheendoftheSummaryforPolicymakersoftheSynthesisReport,
governmentsacceptedriskastheunifyingthemeofthisandfutureassessments.Because
theworldsgovernmentsunanimouslyapprovedtheSummaryforPolicymakerswordby
word,theyhaveallagreedthatrisknotjustimpactsortheirderivativevulnerabilities
mattersmosttothemastheyconsiderhowtorespondtotheclimateproblem.Asan
expressionofthisparadigmshift,governmentsembracedafundamentalinsightofthe
FourthAssessment:
Respondingtoclimatechangeinvolvesaniterativeriskmanagementprocess
thatincludesbothadaptationandmitigation,andtakesintoaccountclimate
changedamages,cobenefits,sustainability,equityandattitudestorisk.
(IPCC2007c,pg22;emphasisadded)
Governmentsarebeginningtounderstandthattheriskassociatedwithanypossibleevent
dependsbothonitslikelihoodanditspotentialconsequences.Thisisthedefinitionofrisk
thatfinanceministershavebeenusingfordecades,soitisnotsurprisingthatmany
governmentsdelegationsunderstandtheconcept.Perhapstheonlysurprisewasthatit
tooksolongforgovernmentstorecognizethattheyshouldlookatclimatechangethrough
thesamelensthroughwhichtheyviewsocialsafetynetprograms,monetarypolicy,and
foreignpolicy.
Althoughgovernmentsandsomesegmentsofthepolicycommunitymaybecomfortable
withtheriskmanagementparadigm,theclimatechangeresearchandassessment
communityhasheretoforefailedtocatchon.Ifresearchersaretocontributefurthertothe
assessmentsthatarethefoundationsofglobalpolicydeliberations,theymustmakerapid
progressinprovidingpolicyrelevant,rigorous,scientificinsightintohowtomakesenseof
theIPCCsconclusionquotedabove.
Thispaperpresentsafirstattempttodeconstructtheparadigmbytakingeachitalicized
phraseinturnandofferinginsightsintowhatwedoanddonotknowineachcase.We
knowquiteabitaboutsomeofthesecriticalphrases,andfutureresearchagendaswill
focusonapplying,extending,andcommunicatingthatknowledge.Forothers,however,
researchintonewapproachesisrequired.Section1beginswithabriefdiscussionofhow
iterativeprocessesmightbeappliedtotheclimatearena.Section2focusesattentiononthe
needforsupplementingtraditionalcostbenefitapproacheswithriskmanagement

204 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


techniques.TheinsightsofferedhereareamplifiedinSections3and4byunderscoringthe
rolesofbothadaptationandmitigationinariskreductionportfolioandbyhighlightingthe
boundariesofourunderstandingofthebenefitsandcobenefitsofsuchaportfolio.Section
5offerssomepreliminarythoughtsabouttherolesofsustainability,equity,andattitudes
towardriskinevaluatingandsynthesizingthevalueandcostsoffutureiterativepolicy
decisions.Finally,theconcludingsectionreviewswhatwedoanddonotknowabouthow
toapplyanalytictechniquestothesyntheticconclusion.

1. Iterative
Althoughclimatechangeisalongtermproblemthatwillrequiresustainedpolicyaction
foracenturyorlonger,itisunlikelythatwewillbeabletosetclimatepolicytodayforthe
entire21stcentury.Manyuncertaintiesaresoprofoundthattheywillnotberesolvedsoon
and,insomecases,mayonlyberesolvedinhindsight.Aclassicexampleofthisconundrum
isclimatesensitivitytheincreaseinglobalmeantemperaturethatiscausedbyadoubling
ofcarbondioxideconcentrationsfrompreindustriallevels.Currentunderstanding,as
reportedinIPCC(2007a,pg65),putsthelikelyrangeofthiscriticalparameterat24.5oC,
buthighervaluesarepossibleanditiswidelyacceptedthattimelyreductionsinthis
uncertaintyareunlikely.1AsreportedinRoeandBaker(2007),forexample,the
probabilityoflargetemperatureincreasesisrelativelyinsensitivetodecreasesin
uncertaintiesassociatedwiththeunderlyingclimateprocesses.AllenandFrame(2007)
respondedbyarguingthatitwaspointlessforpolicymakerstocountonnarrowingthis
fundamentaluncertainty.Asaresult,apolicyresponsethatdelaysimmediateactionin
favorofwaitingfortheresultsofacrashresearchprogramtonarrowtherangeisnot
viable.Moreover,weshouldnotanticipatethatwewillbeabletosetlongtermpoliciesin
concreteatanytimeintheforeseeablefuture.Itfollowsthatwemustbegintoconstructa
processbywhichinterimtargetsandobjectiveswillbeinformedbylongtermgoalsinways
thatnecessaryadjustmentscanbemadeinanefficientmanner(e.g.,Yoheetal.,2004).This
isasimpleandlogicalconclusion,butdifficulttomakeoperational.
Domesticandinternationalbankingandfinancialsystemsprovidesomeevidencethat
iterativepolicymakingcanbeaccomplishedonamacroscale(e.g.,StiglitzandWalsh,
2005;chapters32and33).Forexample,centralbanksfrequentlysettrajectoriesfor
growthinthemoneysupplywhentheyexpectnormaleconomicactivityovera
foreseeablefuture;theyworkwithinanannouncedtimeperiodthatdefinesprecisely
whentheyexpecttomakethenextroundofpolicydecisions.Sincetheydonothaveexact
controloverthemoneysupply,however,centralbanksalsoreservetherighttointervene

1
IPCC(2007a)describestheequilibriumclimatesensitivity[as]ameasureoftheclimatesystemresponseto
sustainedradiativeforcing.Itisnotaprojectionbutisdefinedastheglobalaveragesurfacewarmingfollowinga
doublingofcarbondioxideconcentrations.Itislikelytobeintherange2Cto4.5Cwithabestestimateof
about3C,andisveryunlikelytobelessthan1.5C.Valuessubstantiallyhigherthan4.5Ccannotbeexcluded,
butagreementofmodelswithobservationsisnotasgoodforthosevalues.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 205


earlierthananticipatedifthemoneysupplyclimbsaboveorfallsbelowclearlydescribed
thresholdsthatdefineacceptablelevelsofvariability.Centralbankscanalsomonitor
exchangeratesinexactlythesameway.Inbothcases,actorsacrosstheeconomyknow
exactlyhowthecentralbankwillconductitsanalysesinanticipationofmakingscheduled
policyadjustments;thatis,theycananticipatemuchofwhatwillhappenduringthose
adjustmentsandbegintomakeappropriatechangesintheirownbehaviorsinadvanceof
thepolicychange.Theseactorsalsounderstandwhatthebankswilldoifunanticipated
adjustmentsareinitiatedbycrossingthetriggerthreshold;thatis,theycanalsodetect
earlywarningsignssothattheycanbegintorespondinadvanceofthesemore
unexpectedevents.Inotherwords,transparencyintheprocesscanlessenthecostsof
plannedorunplannedpolicyadjustments.
Theexperiencesofthesemonetarystructuressuggestatleastthreeevaluativecriteriathat
canbeappliedtotheclimatearena:(1)keeplongtermtargetoptionsavailableforaslong
aspossiblebyadoptinghedgingstrategiestoinformneartermactionsandidentifying
downstreamadjustmentthresholds,(2)minimizetheadjustmentcostsofregularly
implementedadjustments,and(3)minimizeadministrativecomplexityinbothbymaking
themastransparentandaspredictableaspossible.
Eventsinthefinancialmarketsthatmarkedthesecondhalfof2008clearlyindicatethat
difficultiescanstillarise,especiallywhenpolicyleversandwellunderstoodmonitoring
mechanismsbreakdown.Centralbanksmayhavebeenmonitoringthemoneysupply,
inflationrates,andexchangeratefluctuationintheearlypartof2008,butitwouldseem
thattheywerenotkeepingtrackofcomplexityinfinancialinstrumentsthatspread
enormousriskacrossarangeofunsuspectingandotherwisedebtburdenedcitizensand
institutions.
Potentiallyunforeseendifficultiesareperhapsevenmoreubiquitousanddangerousinthe
climatearena.Theenormousuncertaintythatstillcloudsourunderstandingoftheclimate
systemmeansthatclimatepolicymustbeimplementedwhilesimultaneouslymonitoring
impactsandvulnerabilityofhumanandnaturalsystems.Evenwhenweunderstand
specificclimateprocessesverywell,persistentandpotentiallyprofounduncertainties
aboutimpactscanproducefuzzythresholdsofdangerousinterferenceacrosskey
vulnerabilitiesthatcannotallbecalibratedindollars.Somearebestleftintermsof
millionsofpeopleatriskfromcoastalstormsorchangesinthereturntimesofthe
current100yearflood.Inotherwords,risksarepotentiallylargeandthepossibilityofa
bailoutmightbequitelow.Insightsdrawnfromourexperiencewithmonetarypolicy
showus,however,thatthiscomplexityshouldnotbeasourceofparalysis.Instead,they
showthefundamentalneedforiterativepoliciesthataredesignedbothtohedgeagainst
potentialcalamities(i.e.lowertheirlikelihoodswithfullunderstandingthatthereareno
guarantees)andtoadjustefficientlyinresponsetonewinformationabouttheclimateand
economicsystemsaswellasperformanceagainstneartermgoals.

206 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


2. Riskmanagementprocess
Havingrecognizedthatpervasiveuncertaintyinourunderstandingofboththeclimate
systemandfuturepoliticalsocialeconomicdevelopmentpathwaysrequiresiterative
climateresponses,wemustnowturntoframingtheunderlyinganalysisinwaysthatcan
informsuchanapproach.Herewecontrastconventionalbenefitcosttechniquesdesigned
todeterminetheoptimalpolicywithbroaderriskmanagementtechniquesdesignedto
hedgeagainstuncertainbutpotentiallyhighconsequenceoutcomesandallowformid
courseadjustmentsasneeded.
BeginningperhapswithNordhaus(1991),thedynamiclongtermversionofthestandard
benefitcostparadigmhasbeenthemainstayofeconomicanalysesofclimatepolicy
(particularlyonthemitigationside).Inapplyingthisapproach,researcherstrackeconomic
damagesthatwouldbeassociatedwithclimatechangeandcoststhatwouldbeassociated
withclimatepolicyovertimescalesthatextenddecadesorcenturiesintothefuture.They
calibratethesedamagesandcostsalongscenariosofeconomicdevelopmentandresource
availabilitythatrepresentarangeofpossible(butunpredictable)futures.Thedamages
andcostsaredisaggregatedacrosscountriesandregionstovaryingdegreesbydifferent
researchers,andbotharediscountedbacktopresentvalues.Inthisfinalstep,estimatesof
thepresentvalueofbenefitsandcostsarehighlysensitivetouncertainnaturalparameters
(e.g.,climatesensitivity)anduncertainpolicyparameters(e.g.,theassumeddiscountrate,
whichisextremelysensitivetoattitudestowardrisk,attitudestowardinequity,andinter
temporalimpatience,asdiscussedbelow).
Pointingoutthatsomebenefits(andevensomecosts)cannotbemonetized,many
researchersandcommentatorshavebecomeincreasinglycriticalofthisapproach.2In
response,practitionershaveopenedthedoortotrackingbenefitsandcostsintermsof
alternative,noneconomicmetrics,althoughsuchmetricshaveyettobeappliedto
regulatorypolicy.3Theyhavealsorecognizedproblemswithspecifyingappropriate
discountrates,copingwithuncertainty,andaccommodatingtheprofounddistributional
consequencesofclimatechange.4Earlyin2009,PresidentObamasignedExecutiveOrder
13497instructingtheDirectoroftheOfficeofManagementandBudgettoreviewthese
issueswithadvicefromregulatoryagencies.Whileprogresshasbeenslowatthefederal
level,thisactionopensthedoorfurthertononquantifiedcostsandbenefitsexpressedin

2
CritiquesofrelyingtooheavilyonlimitedbenefitcostanalysesincludeTol(2003),Yohe(2004,2006),and
Ackermanetal.(thisvolume).
3
Forexample,CircularA4[WhiteHouse(2003)]wasissuedbytheOfficeofManagementandBudgettoupdate
longstandinginstructionsthatdefinedthestandardsforgoodregulatoryanalysisexercisesthatworkfrom
statementsofneedandexplorationsofalternativeapproachestoproduceevaluationsofthebenefitsandcosts
quantitativeandqualitativeoftheproposedactionandthemainalternativesTheCircularleadswithan
explicitpresumptionagainsteconomicregulation.Mostofthetext,though,isdedicatedtoilluminatingbest
practicesforcircumstancesinwhichinterventionisdeemedwarranted.Itbeginsbyhighlightingbenefitcost
andcosteffectivenessanalysesasthesystematicframeworkswithinwhichtoidentifyandtoevaluatethe
likelyoutcomesofalternativeregulatorychoices.TheCongressionalBudgetOffice(2005)amplifiedthesepoints.
4
SeethepaperbyRoseinthisvolume.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 207


nonmonetaryterms,includingmaintainingriskthresholds,beingconsideredfor
regulatorydesign.Indeed,CityofNewYorkadoptedthisapproachwhenitstartedtothink
abouthowtoprotectitsenormouspublicandprivateinfrastructurefromgrowingclimate
risk;seeNPCC(2009)fordetails.
AssuggestedbytheIPCC(2007c),riskmanagementtechniquescanexplicitly
accommodatemany(butnotall)ofthesethornyissues.5Itsmoststraightforward
applicationstoclimatechangebegin,aselsewhere,withthestatisticaldefinitionofrisk
theprobabilityofaneventmultipliedbyitsconsequence.Inbenefitcostapproaches,all
consequencesarecalibrateddirectlyasanticipatedeconomicoutcomesthatareexpressed
inunitsofcurrency.Intheseapplications,anydollarlostorgainedinonepossibleoutcome
isworththesameasanyotherdollarlostorgainedinanyotheroutcome.Itfollowsthat
decisionmakersneedonlyworryaboutexpecteddollargainsorlossesregardlessofhow
goodorbadanyparticularoutcomemightbe.6Riskmanagementapproachesexpandthe
rangeofanalyticapplicabilitybyallowingconsequencestobecalibratedintermsofmore
generalwelfaremetrics.Thesemetricsmaydependonthesameoutcomesasbefore,but
theymakeitclearthatonedollarinonepossibleoutcomeisnotnecessarilyworththe
sameasonedollarinanotherpossibleoutcome.Metricsthatreflectaversiontorisk,for
example,holdthatanextradollargainedinagoodoutcomeisworthless,intermsof
welfare,thananextradollarlostinabadoutcome.Itfollowsthattheextremesofpossible
outcomesmatterinthesecases,anditisinthesecontextsthatpeoplebuyinsurance
and/oradopthedgingstrategiesagainstespeciallybadoutcomesevenifsuchstrategies
failinabenefitcostanalysis.Theydosobecausehedgingincreasesexpectedwelfare
(computedoverthewelfareimplicationsofthefullrangeofpossibleoutcomes)even
thoughitreducestheexpecteddollarvalueoftheassociatedoutcomes.7
Whatdoweknowabouthowtoapplythisinsightintheclimatearena?Accordingtothe
IPCC(2007a),weknowunequivocallythattheplanetiswarming.Wearenowvirtually

5
ThefoundationsfortheresultsthatfollowcanbefoundinRaiffaandSchlaiffer(2000).
6
Tobeprecise,lettherangeofpossibleoutcomesbecalibratedby{X1,,Xn}wherethesubscriptsindicate
financialvaluesinnpossiblefuturestatesoftheworld.Abenefitcostapproachlooksonlytocalculatingthe
expectedoutcomeacrossthesefutures.Thatis,if{1,,n}representthesubjectivelikelihoodsofeach
possibleoutcome,thenexpectedbenefitcostcalculationswouldfocusattentionexclusivelyonE{x} iXi.
7
Tocontinuewiththenotationoffootnote5,riskanalysisletstheconsequencesbecalibratedintermsofwelfare
andnotjustoutcomes.Itfollowsthattherelevantmeasureoftherangeofconsequencesis{U(X1),,U(Xn)}
whereU()isthewelfaremetric.Inthiscase,decisionmakersworryaboutexpectedwelfareandnotexpected
outcome;i.e.,theywouldfocusattentiononE{U(X)} iU(Xi).Iftheyareaversetorisk(soU()increaseswith
Xatadecreasingrate),thentheywouldbuyinsuranceeventhoughthepremiumstheypaylowersevery
possibleoutcome.Why?Becauseinsuranceguaranteesthattheywillbecompensatedtosomedegreeshoulda
reallybadoutcome(areallylowvalueforXi)materializes.Inotherwords,theywillinglysacrificeexpected
economicvaluecalculatedacrossallpossibleoutcomestoreducethepainthattheywouldfeelinthe
(potentiallyunlikely)eventthatasinglebadoutcomemightoccur;andtheyaresowillingbecausedoingso
increasestheirexpectedwelfare.Hedgingisavariantonthesamethemeinwhichdecisionmakerssacrifice
expectedeconomicvaluetoinvestinsomeactionthatworkstoreducethelikelihoodthatabadextremeevent
mightoccur.BothresultscanbederiveddirectlyfromtheobservationthatriskaversionmeansthatE{U(X)}<
U(E{X}).

208 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


certainthattheclimateischangingatacceleratingrates.Wealsoknowwithveryhigh
confidencethatanthropogenicemissionsaretheprincipalcause.Weevenhaveevidence
thatanthropogenicclimatechangewasthestrongestcontributortotheconditionsthat
createdthe2003heatwaveacrosscentralEuropethatcausedtensofthousandsof
prematuredeaths(Stottetal.,2004;IPCC,2007b).Wealsoknowfromthedire
consequencesofthe2003Europeanheatwave,the2004AsianTsunami,andHurricane
Katrinain2005,8thatbothdevelopinganddevelopedcountriesaresusceptibletothetypes
ofeventsthatareexpectedtooccurmoreoftenandwithgreaterintensityinthefuture
becauseofclimatechange.Thisknowledgealoneissufficienttoestablishtheseriousrisks
ofclimatechangeandtheneedtorespondinthenearterminwaysthatwillreducefuture
emissionsandtherebyamelioratethepaceandextentoffuturechange.Indeed,lookingat
uncertaintythroughariskmanagementlensmakesthecaseforneartermactionthrough
hedgingagainstallsortsofclimaterisksrisksthatcanbedenominatedintermsof
economicdamagesaswellasotherindicators,includingmorewidespreadhunger,water
stress,orgreaterhazardsfromcoastalstorms.Itthenfollowsfromsimpleeconomicsthat
actionshouldbeginimmediatelyinordertominimizetheexpectedcostofmeetingany
longtermobjective.
Asdiscussedpreviously,monetarypolicyprovidesarealworldillustrationofhowhedging
strategieshavebeenemployedatamacroscale.Ata2003symposiumonMonetaryPolicy
andUncertainty:AdaptingtoaChangingEconomy,9ChairmanoftheFederalReserve
BoardAlanGreenspan,illustratedthepoint:
Forexample,policyAmightbejudgedasbestadvancingthepolicymakers
objectives,conditionalonaparticularmodeloftheeconomy,butmightalsobe
seenashavingrelativelysevereadverseconsequencesifthetruestructureof
theeconomyturnsouttobeotherthattheoneassumed.Ontheotherhand,
policyBmightbesomewhatlesseffectiveundertheassumedbaselinemodel
butmightberelativelybenignintheeventthatthestructureoftheeconomy
turnsouttodifferfromthebaseline.Theseconsiderationshaveinclinedthe
FederalReservepolicymakerstowardpoliciesthatlimittheriskofdeflation
eventhoughthebaselineforecastsfrommostconventionalmodelswouldnot
projectsuchanevent.(Greenspan,2003,pg.4;emphasisadded)
Indeed,noneofthemodelsthatinformedFederalReservepolicywouldevenputa
probabilityonthechanceofdeflation.TheBoardsimplyknewthatitwasnotzeroandthat
theydidnotwanttheeconomytoenduretheconsequences.TheChairmanexpandedon

8
Notethattheseeventsneednotbelinkedtoclimatechangetoexposetheunderlyingvulnerabilitiestosimilar
eventsthatwouldbelinkedtoclimatechangeinthefuture.Forexample,althoughtheAsiantsunamiwas
causedbyanearthquake,itsimulatedastormsurgethatmightbeassociatedwithastrongtropicalcyclone.
9
MonetaryPolicyandUncertainty:AdaptingtoaChangingEconomy:AsymposiumsponsoredbytheFederal
ReserveBankofKansasCity,JacksonHole,Wyoming,August2830,2003.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 209


thisillustrationinhispresentationtotheAmericanEconomicAssociation(AEA)attheir
2004annualmeetinginSanDiego:
theconductofmonetarypolicyintheUnitedStateshascometoinvolve,atits
core,crucialelementsofriskmanagement.Thisconceptualframework
emphasizesunderstandingasmuchaspossiblethemanysourcesofriskand
uncertaintythatpolicymakersface,quantifyingthoseriskswhenpossible,and
assessingthecostsassociatedwitheachoftherisks.Thisframeworkalso
entails,inlightofthoserisks,astrategyforpolicydirectedatmaximizingthe
probabilitiesofachievingovertimeourgoals(Greenspan,2004,pg.37;
emphasisadded)
Clearly,theseviewsareconsistentwithanapproachthatwouldexpendsomeresources
overtheneartermtoavoidasignificantrisk(despitealowprobability)inthefuture.
Indeed,theChairmanusedsomefamiliarlanguagewhenhesummarizedhisposition:
Asthisepisodeillustrates(thedeflationhedgerecordedabove),policy
practitionersunderariskmanagementparadigmmay,attimes,beledto
undertakeactionsintendedtoprovideinsuranceagainstespeciallyadverse
outcomes.(Greenspan,2004,pg.37;emphasisadded)
Canourcurrentunderstandingoftheclimatesystemsupportasimilarapproachto
managingtherisksofclimatechange?Attheveryleast,weneedsomeinformationabout
consequencesofclimatechangethatwewouldliketoavoidandsomeinsightintothe
sensitivityoftheirlikelihoodtomitigation.Thesearefundamentalquestionsthatmustbe
addressedbeforeproceeding.
Manyauthorshaveprovidedinsightstosomeoralloftherequisitecomponents,i.e.,
estimatesofprobabilitiesofspecificoutcomesandquantificationsortheassociated
vulnerabilities.Somehave,forexample,comparedthecostsofmitigationwiththe
correspondingchangesinclimaterisks:
MastrandreaandSchneider(2004)usedthesimplifiedintegratedassessmentmodel
DICEtoassessthecostsofavoidingdangerousclimatechangeasdefinedby
assumptionsdrawnfromtheIPCCThirdAssessmentReport(IPCC,2001a,2001b).10
Websteretal.(2003)usedanintegratedmodelofintermediatecomplexityto
quantifythelikelihoodsofglobalwarmingfuturesin2100,beginningwithprojections
ofpopulation,economyandenergyuse.
Jones(2004a,2004b),Wigley(2004)andJonesandYohe(2008),presentframeworks
thatprobabilisticallyrelatestabilizedCO2concentrationswithequilibrium
temperature,althoughthesestudiesstoppedshortofrelatingtheirresultstoeither
thecostsofmitigationorthebenefitsofavoidingdamages.

10
Forbackgroundonintegratedassessmentmodels,seethepaperbyMastrandreainthisvolume.

210 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


Othershavebeguntoconsidertheroleoffuturelearningininformingrisk
assessment.BrianONeilleditedanentirevolumeofpapersdesignedtoexplorethe
roleoflearninginsettinglongtermmitigationstrategies(ONeill,2008).
Schlesingeretal.(2006)adoptedamorefocusedapproachbytrackingthelikelihood
ofacollapseoftheAtlanticthermohalinecirculation(THC)overthenextoneortwo
centuriesunderavarietyofmitigationassumptionsusingthreealternative
representationsofuncertaintyinclimatesensitivity.ZickfeldandBruckner(2008)
followedwithaninvestigationoftheimplicationsofalternativeemissionscorridors
onthesameTHCriskprofileusinganalternativeoceanmodel.
Takentogether,thesestudiesshowprogressintrackingthepotentialefficacyofmitigation
inreducingthelikelihoodofverynegativeoutcomes.
Theconceptofcertaintyequivalenceisusedinriskanalysistoconvertconsequences
calibratedinwelfaretermsintofinancialterms.Essentially,certaintyequivalenceaccounts
forriskaversionbyprovidinganestimateofhowmuchpeoplewouldbewillingtopayto
avoidtheriskysituation.11Ithasalsobeenemployedtoinformclimatechangemitigation
decisionsincaseswheretherelativelikelihoodsofvariouspossiblefuturescanbe
analyzed.Sternetal.(2006and2007)expresseddamagesintermsoflossesincertainty
equivalentpercapitaconsumptiondiscountedover200years.Usingtheuncertainty
analysiscapabilitiesofthesimplifiedintegratedassessment12modelPAGE2002,the
authorsaccommodatedenormousvariabilityinpercapitaconsumptionacross1000model
runsbycomputingmeanexpecteddiscountedutilitywithoutandwithclimatechangefor
threedifferentdamagecalibrations.Certaintyequivalentswithandwithoutclimatechange
werethencomputedforeachcalibration;i.e.,theauthorscalculatedthelevelofpercapita
consumptionwhich,ifitweretogrowwithcertaintyat1.3percentperyear(anassumed
naturalgrowthrate),wouldachievealevelofdiscountedutilityexactlyequaltothe
expecteddiscountedutilitiesjustdefined.Theeconomicvaluesofglobaldamage
attributabletoclimatechangedamagesunderalternativecalibrationswerethentakento
bethedifferencesbetweencertaintyequivalentswithandwithoutclimatechangeforthe
threecalibrations.Intheirsimplestform,thesecomputeddifferencesaresimplyestimates
ofthefractionofcurrentpercapitaconsumptionthattherepresentativecitizenwouldbe
willingtopaytoeliminatealloftheclimateriskcapturedbytheunderlyinganalysis.
Sternetal.(2006)estimatedwhattheirrepresentativecitizenwouldbewillingtopayto
avoidalldamagesassociatedwiththreedamagecalibrations,buttheyprovidedno

11
Thecertaintyequivalentofariskysituationisimplicitlydefinedastheoutcomethatwould,ifitcouldbe
guaranteed,achievealevelofwelfareorutilitythatisequaltotheexpectedwelfareorutilitycalculatedacross
allpossibilities.Returningtothenotationoffootnotes5and6,thecertaintyequivalentoutcomeXceisdefined
implicitlyasthesolutiontotheequationU(Xce) E{U(X)}.SinceXce<E{X}forriskaversedecisionmakers,the
differencebetweenacertaintyequivalentandanexpectedoutcome(i.e.,E{X}Xce)thereforerepresentsan
estimateofwhatpeoplewouldbewillingtopaytoavoidtheriskysituationaltogether.Inaddition,differences
incertaintyequivalentsfortwodistributionsofoutcomescanbeusedtotrackwhatpeoplewouldbewillingto
paytoreduceuncertainty.Forarecentapplicationofthisapproach,seeNewboldandDaigneaultinthisvolume.
12
Forbackgroundonintegratedassessmentmodels,seethepaperbyMastrandreainthisvolume.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 211


informationaboutwhattheirrepresentativecitizenwouldbewillingtopayforvarious
levelsofemissionreductionsandtheirassociatedreductionsindamages.TolandYohe
(2009)workedtofillthisgapusingamuchsimplermodel;Table1showstheirresults.
NoticeinthefirstrowthattheunregulatedpathiscalibratedtomatchtheSternbaseline
a5.3percentreductionincertaintyequivalentpercapitaconsumptionfromclimate
change.Correspondinglevelsofresidualdamage,expressedcomparablyintermsof
certaintyequivalencealongcostminimizingmitigationpathways,arethenreportedfor
concentrationthresholdsrangingallthewaydownto400ppm.Sincealloftheseresiduals
arepositive,noneoftheconsideredmitigationtargetsobviatestheneedforadaptation.

Table1:Estimatesofresidualeconomicdamagealongleastcostmitigation
pathwaysfromtheSternetal.(2006)baselineexpressedintermsofpercentage
changesincertaintyequivalentpercapitaconsumptionrelativetoscenarios
alongwhichclimatedoesnotchange.(Source:TolandYohe,2009)
AtmosphericConcentration CertaintyEquivalentPerCapitaConsumption
unregulated 5.3percent
750ppm 3.8percent
700ppm 3.4percent
650ppm 3.0percent
600ppm 2.6percent
550ppm 2.2percent
500ppm 1.7percent
450ppm 1.3percent
400ppm 0.8percent

Tosummarize,weknowthathumanitysgreenhousegasemissionsarechangingthe
climateinwaysthatarelikelytobedetrimentaltosocietyandthatsomeofthe
consequencescouldbecatastrophic.Wealsoknowthatthetimingandseverityofthese
changesareimpreciselyassociatedwithparticularsocioeconomicandemissionspathways.
Tobebrutallyhonest,pervasiveuncertaintyaboutthephysicalandeconomic
consequencesofclimatechangeunderminesthecredibilityofeconomicallyoptimal
policiesthatemergefromtraditionalbenefitcostcalculations.Sincethereisgoodevidence
tosuggestthatgettingtheoptimalpolicywrongcouldbeextremelyexpensive,itfollows
fromstraightforwardeconomicsthatacomplementaryapproachaimedat
managing/reducingriskisrequired.Itisimportanttorecognizethathedgingpoliciesthat
emergefromtheriskmanagementapproachwouldsacrificealittleinexpectedutility,but
thepayoffwouldbereductionsinthelikelihoodsofunacceptabledeclinesingeneral
welfaredeclinesthatwouldresultiftheoptimalpolicyshouldfail.

3.BothAdaptationandMitigation
Thediscussionhasthusfarframedmitigationasamechanismbywhichclimateriskscan
bereduced.Thisinitialfocusisappropriatebecauseadaptivecapacitycanbeoverwhelmed
evenwithinthemiddlerangeofprojectedwarmingindevelopedanddevelopingcountries

212 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


alike(IPCC,2007b,Chapter20).However,adaptationtounavoidablechangeisalso
requiredsincewewouldbecommittedtoanother0.6oCofwarmingevenifgreenhousegas
emissionshadfallenpermanentlytozerointheyear2000(IPCC,2007b).Itistherefore
essentialtoconsidertherolesofbothadaptationandmitigationinsettinglongrange
climatestabilizationgoalsandtranslatingthosegoalsintoshorttermobjectives(interms
of,forexample,emissionspeakingpointsandthetimingofadaptationinvestments).To
makethesyntheticstatementoftheIPCC(2007c)operational,wemustalsoshowhow
adaptationcanbeengagedinaniterativeprocessdesignedtomanageriskandhowthe
needforadaptationcanbeinfluencedbyinvestmentinmitigation.
Sincetheseissueshavenotyetreceivedmuchattention,wewillofferasimpleapplied
examplethatfocusesattentiononthevulnerabilityofNewYorkCitytoseverecoastal
stormsasaproofofconcept.Inthisexample,the100yearcoastalfloodinganomalyfor
NewYorkCity(asjudgedbyFEMAin2005)ischosentorepresenthowsuchvulnerability
mightbeexperienced.ItbuildsdirectlyonrecentworkbyKirshenetal.(2008)inwhich
returntimesofthecurrent100yearfloodingeventarecorrelatedwithprospectivelevels
ofsealevelrise.Itisimportanttonotethattheeffectsofchangingintensitiesor
frequenciesofcoastalstormswerenotconsidered.Onlytheeffectofrisingsealevelson
stormsurgesassociatedwithstormsthatnowoccurmorefrequentlywereconsidered(for
example,thecurrent25or50yearanomaliesthatwill,withrisingsealevel,portray
inundationpatternsnowassociatedwiththe100yearanomaly).
AlternativetrajectoriesoffuturesealevelrisearoundNewYorkCitywerederivedfrom4
alternativeemissionsscenariosreported,alongwithsubjectiveprobabilitiesoftheir
relativelikelihoods,inYoheetal.(1996)across9alternativeclimatesensitivities.13Figure
1showstheresultsofsuperimposingtheresultingprobabilityweightedsealevelrise
scenariosontheKirshenresultsforfloodreturnintervals.Giventhisinformation,a
decisionmakingplannerwhoreportedthata40yearreturntimewasthelowerboundof
hisorhercomfortzonecouldseean80percentchancethatthisthresholdwouldbe
breachedwithina2025planninghorizonwithvirtualcertaintybeyond2035.This
realizationcouldeasilytriggeranynumberofadaptiveresponsesthatcouldrangefrom
significantinvestmentinprotectiontoplannedretreatfromthesea(highlyunlikelyin
downtownManhattan,butmorelikelyforsomemoreresidentialandexposed
communities).If,however,ourplannerwerecomfortableafterhavingtakensome
preliminaryprotectiveaction,withalowerreturntimelike20yearsforthecurrent100
yearanomaly,thenthelikelihoodoffallingbelowthislowerthresholdwouldbeamore
tolerable20percentin2025and30percentin2030.Thesubjectivelikelihoodofcrossing
thecriticalreturntimethresholdwould,though,jumptomorethan60percentby2035.It
followsthattheoriginalurgencyofthemoreriskaverseplannerwouldbediminished,but

13
TheclimatesensitivitydistributionappliedhereisdrawnfromYoheetal.(2004);itisadiscreterepresentation
ofthedistributionreportedinAndronovaandSchlesinger(2001).

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 213


notbymuch;putanotherway,adaptationmightworkforashortwhile,butitwouldnotbe
sufficientoverthelongterm.
Figure2,drawnfromYohe(2009)displayssomeevidenceaboutthesensitivityofthe
returntimeofthe100yearfloodinganomalytoalternativemitigationpathwaysforthe
year2050.Theunabatedplotaddssomedetailtothe2050distributionrecordedinFigure
1.TheIPCC(450)mitigationscenarioaddsabout10yearstothemedianreturntime
roughlyequivalent,accordingtoFigure1,toa10or15yeardelayincrossingthe5050
riskthreshold.Notice,however,thatthemorecosteffectiveWREmitigationpathway(see
Wigleyetal.,1996),whichallowsemissionstopeaklaterattheexpenseofsharper
reductionsthereafter,resultsinasmallertimedelaythantheearlieractingIPCCscenario.
Twoinsightsfromtheseresultsarethat(1)slowingemissionsbuysmoretimefor
planning,financing,andimplementingadaptation,and(2)thetimingofemissions
reductions(i.e.earliervs.laterpeaking)forgivenstabilizationconcentration(e.g.,450ppm
CO2e)affectshowmuchtimethemitigationeffortbuys.Hence,thetimingofmitigation
effortscaninfluencetheurgencywithwhichadaptationmightbepursued.Differentlevels
ofmitigationeffortcouldevenalterwhichadaptationoptionswouldbefeasible
Itshouldbeclearfromthispreliminaryworkthatriskprofilescanportrayawiderangeof
vulnerabilitiesovertimeevenifthosevulnerabilitiescannotbeexpressedintermsofasingle
(monetary)metric.Theycan,therefore,beenormouslyvaluableinconsideringandprioritizing
investmentsinadaptationacrossmultiplesectorsand/ormultiplelocations.Theycanalsobe
usedtodisplaythesensitivityofrisks,withandwithoutadaptation,tovariousmitigation
pathways,althoughintegratingthecontentofmanyindividualriskprofilesandscalingthem
uptothemacroscaleatwhichmitigationdecisionsaremaderemainsproblematic.Acollection
ofvulnerabilitystudiesdrawnfromawidesampleofkeyvulnerabilitiescannonetheless
providethosedecisionswithinformationthatishiddeninsimplecalculationsofaggregate
economicbenefits.Suchcollectionscouldtherebyinformpoliticaldeliberationsaboutwhat
mightconstitutedangerousanthropogenicinterferencewiththeclimatesystemmorefully.

214 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


Figure1:TheRelativeLikelihoodsthattheReturnTimeofthe2005calibrated100YearAnomalyWill
BeSmallerthantheSpecifiedPlanningHorizoninSelectedYears.Anypointonanylineindicates,with
itsverticallocationandfortheidentifiedyearinthefuture,thelikelihoodthatthereturntimeofthe
100yearstormwillbesmallerthanthevalueidentifiedbyitshorizontallocation.Forexample,thethird
triangleuptheredlineshowsthatby2035itismorethan60percentlikelythatthereturntimewillbe
lessthan20years.

100
Likelihood that Return Time is Smaller than the Indicated

90

80
Threshold (Subjective Percentage)

70
2025
60 2030
2035
50
2040

40 2045
2050
30

20

10

0
0 5 10 15 20 25 30 35 40 45
Return Tim e Threshold

Figure2:TheRelativeLikelihoodsthattheReturnTimeofthecurrent100YearAnomalyWillBe
SmallerthanthePlanningHorizonin2050.Thelinesindicatethelikelihoodthatthereturntimeofthe
100yearstormalongunabatedandtwomitigationtrajectorieswillbesmallerthantheindicated
thresholdin2050;bothstabilizeconcentrationsofgreenhousegasesat450ppmvinCO2equivalents
alongtwodifferentemissionstrajectories.Theslowerpaceofearlyreductionsalongthemorecost
effectiveWREtrajectoryreducestheefficacyofmitigationtoslowthereductioninreturntimes.

90
Indicated Threshold (Subjective Percentage
Likelihood that Return Time is Smaller than t

80

70

60

50 Unabated
WRE (450)
40 IPCC (450)

30

20

10

0
0 5 10 15 20 25 30
Return Time Threshold (2050)

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 215


4.Damagesandcobenefits
Weturnnowtoconsidermetricsbywhichthedamagesofclimatechange,thecostsof
mitigation,andthepotentialforcobenefitsacrossthetwohavebeenexpressed.Thefirst
subsectionindicatesthatexistingestimatesofeconomicdamagesfromclimatechange
havefailedtoaddressmanyofthedimensionsofpossiblenonmarginalclimateimpacts
(thatis,impactsthatinvolvelargeand/orsuddenchanges).Theseconddescribesthekey
vulnerabilitiesidentifiedfromthecurrentliteratureinFourthassessmentReportofthe
IPCC(2007b);thepointhereisthatmanyofthesevulnerabilitiescannotbemonetized
(thatis,theycanonlybecalibratedinnonmonetarymetrics).Afinalsubsectiondiscusses
thelatestcontributionbyMartinWeitzmantothedebateatheoreticalresultthat
questionsourabilitytoaccommodateprofounduncertaintiesthatstretchtheunderlying
probabilitydistributionsintoregionswhereextremeandambiguousconsequencesmight
occur.

4.1 Missingimpacts
ThematrixdisplayedinFigure3,derivedfromasimilarfigureinDowningandWatkiss
(2003),summarizesthestateoftheartinanalyzingtheeconomicimpactofclimatechange
andthereforetheeconomicbenefitofclimatepolicy;italsoappearsinYoheandTirpak
(2008)fromwhichmuchofthissubsectionisdrawn.Thecolumnsaredividedverticallyby
thedegreetowhichthecomplicationofuncertaintyinclimatechangescienceiscaptured
bybenefitsanalysis.Theybeginwithcoverageofprojectionsofrelativelysmoothclimate
changetrends(e.g.,averagetemperature,sealevelrise),moveontoconsiderationsofthe
boundedrisksofextremeweatherevents(e.g.,largescaleprecipitationeventsand
droughts)andclimatevariabilityalongthosetrends,andendwithrepresentationsof
possibleabruptchangeand/orabruptimpacts.Therowsaredividedhorizontallybythe
degreetowhichthecorrespondingimpactscanbecalibratedinmonetaryterms.They
beginontheleftwithcoverageofmarketimpacts,moveontoconsiderationsofnonmarket
impacts,andendwithevaluationsofsociallycontingentimpacts(e.g.multiplestresses
leadingtofamineandmigration)acrossmultiplemetricsthatcannotalwaysbequantified
ineconomicterms.14
Takenasawhole,thediagramsuggeststhatmuchoftheexistingresearchhasfocusedon
marketimpactsalongrelativelysmoothscenariosofclimatechange;i.e.,mostofour
knowledgeabouttheeconomiccostsofclimatechangehasemergedfromarea1alone.In
thiscontext,researchershavenotedtheimportanceofsitespecificity,thepathdependence
ofclimateimpactsandtheadaptivecapacityofvarioussystems.Whilecoverageisgreatest

14
Theentriesinthematrixaremeanttobeillustrative;andtheyarenotmeanttosuggesttheexclusivelocationof
particularsectorslikeagricultureandforestry.Thereare,forexample,impactsinthosesectorsderivedfrom
projectionsoflongtermtrends.Theyareshownintheboundedriskcategorytodemonstrateadditionaland
perhapsdominatesensitivitytoclimatedrivenvariabilityandextremeweatherevents.

216 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


inarea1,thisdiversityofcontextmeansthatcoverageofevenmarketsectorimpactsisfar
fromcomprehensive.

Figure3:CoverageofExistingEconomicAnalysisoftheImpactsofClimateChangeRelatedRisks.Most
existingstudieshavebeenlimitedtomarketbasedsectors,thoughafewhavemovedbeyondregionIto
includenonmarketimpactsalongprojectedtrends(regionIV),boundedrisksinmarketandnonmarket
sectors(regionsIIandV)andabruptchangetoselectedmarketsectors(regionIII).

Uncertainty in Valuation
Market Non Market Socially
Contingent
Uncertainty
Projection I IV VII
in
(e.g., sea level Coastal Heat stress Regional costs
Predicting Rise) protection Loss of wetland Investment
Climate Loss of dryland
Change Energy
(heating/cooling)

Bounded II V VIII
Risks
Agriculture Ecosystem Comparative
(e.g. droughts,

change advantage &
floods, storms) Water
market
Biodiversity
Variability structures
Loss of life
(drought, flood,
storms) Secondary
social effects

System III VI IX
change &
surprises Above, plus Higher order Regional
collapse
(e.g. major Significant loss Social effects
events) of land and Regional
resources collapse
Non-marginal
effects
Source:YoheandTirpak(2008),derivedfromDowningandWatkiss(2003).

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 217


AlthoughalimitedliteratureexistsforeconomicimpactsinareasIIV,nostudyhas
attemptedacomprehensiveanalysis,eitherforlackofdataorfortheinabilitytomonetize
damagesforcertaincategoriesofimpacts.15Theexistingliteraturehasalmostnothingto
sayaboutimpactsandvulnerabilitycalibratedinthenonmarketimpactsofabruptchange
inandthemultiplemetricsofsociallycontingentimpacts(areasVIIX).Throughthese
sociallycontingentvulnerabilities,climateimpactsinoneplace(e.g.,thedeveloping
countries)canbefeltelsewhere(e.g.,intheUnitedStatesortherestofthedeveloped
world).Allcalculationsofthepotentialbenefitsofclimatepolicycompletelyignorethese
elementsofclimaterisk.Itdoesnotnecessarilyfollow,however,thatattemptstocalibrate
thesevulnerabilitiesintermsofeconomicdamagesshouldbethefocusofnewresearch.
Indeed,itishere,perhapsmostcritically,thatmultiplemetricsofclimaterelatedriskmust
beaccommodatedsothatourpolicydiscussionsaremorefullyinformedaboutwhatmight
happen.
Despitetheseshortcomingsinthecoverageofourimpactsanalysisandconcernsaboutour
understandingoftheclimatesystem,researchersand,policymakersarenowrequiredto
usetheresultsofanalysesthatemanatelargelyfromareaIofFigure3toconduct
assessmentsofoptimalclimatepoliciesandtocomputeestimatesofthesocialcostofCO2
andothergreenhousegases.Thesesocialcostsareestimatedbytrackingthedamage
causedovertimebyreleasinganadditionaltonofagreenhousegaslikeCO2intothe
atmosphereanddiscountingthoseestimatesbacktotheyearofitsemission.Thatistosay,
thesocialcostofcarbonrepresentsthemarginalcostofcarbonemissions;alternatively,it
representsthemarginalbenefitofaunitofcarbonemissionsreduction.
Estimatesofthesocialcostofcarbonthatarecurrentlyavailableinthepublishedliterature
varywidely.IPCC(2007b)wasinformedbyanearlysurveyconductedbyTol(2005),which
reportedthatfully12percentofthenavailablepublishedestimateswerebelow$0(i.e.the
impactsofclimatechangewereestimatedtoproduceanetpositiveeconomicbenefit).
Theirmedianwas$13pertonofcarbon,andtheirmeanwas$85perton.Tol(2007)offers
anupdatedsurveyofmorethan200estimates.Hisnewresultsshowamedianforpeer
reviewedestimateswitha3percentpurerateoftimepreferenceandwithoutequity
weights(i.e.norecognitionthatadollarofharmeffectsthepoormorethantherich)of$20
pertonofcarbonwithameanof$23pertonofcarbon.Moreover,hereportsa1percent
probabilitythatthesocialcostofcarboncouldbehigherthan$78pertongiventhesame
assumptions;andhenotesthattheestimatesincreaserapidlywithlowerdiscountrates
oneoftheprimaryreasonswhytherangeofestimatesofthesocialcostofcarbonisso
large.

15
ForfurtherdiscussionofdamageestimatesforareasIIVinFigure3,seetheappendixattheendofthispaper
andthepaperbyMastrandreainthisvolume.

218 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


Figure4:MultipleAnalyticFrameworksforClimatePolicyResearch.Characteristicsof
variousanalyticalapproachesarehighlighted.Generalapplicabilityincreasesfromtop
tobottomwiththeprospectofsupportinganalysesofdamagesandcobenefitsthat
would,weretheyavailable,begintopopulatethelowerrightsideofthematrix
depictedinFigure3andtherebyimproveourunderstandingofthefullrangeofissues.
(Source:JonesandYohe,2008).

Economic costs and benefits

- Including marginal analysis and benefit-cost analysis
Benefit-cost recognizing adaptation
- Sensitivity to filling in the matrix
framework - Uncertainty on parameters and policy
perspectives:
Conventional applications
Risk averaging and discounting
Global view for mitigation

Risk approach to standard events (climate variability)

and major events (abrupt changes)
- Risk analysis and risk portraits
- Pre-cautionary consideration of major
Risk framework events and tipping points, plus ecosystems,
and ecological functions
- Possibility of positive risks
- Include economic analysis for risk
- Include application to adaptation decisions
- Extend to social risk

Multi-attribute frameworks accounting for

multiple impacts and inequalities
- Disaggregated winners and losers
Social - Physical impacts (number of impacts)
framework - Human well-being and social indicators
- Inequality and distributional effects
- Health, settlement and welfare
- Development issues
- Social sciences

TheestimatesabovelargelyexcludeimpactspopulatingareasIIIXinFigure3,yetTols
surveyprovidesevidencethatassumptionsabouthowtoincludeevenpartialcoverageof
nonmarketdamagescandominateestimatesofmarketdamages.Assumptionsaboutwhat
mightemergefrommorethoroughinvestigationsofareasIVthroughVIofFigure3are
thereforecritical,evenifinferencefromalimitednumberofstudiesissuspect.Perhaps
evenmoretroublingistheobservationthatfewifanyoftheestimatesrecognizeabrupt
change(areasIII,VI,andIX);andnoneventureintoanythingcontainedintherighthand
column(areasVIIthroughIX).Ourcurrentinabilitytopopulatethelightlyshadedregions
ofFigure3withcredibleanalysesunderminesourabilitytocomputethesocialcostof
carbon,andthustheeconomicbenefitofclimatepolicy,withanyconfidence.
Figure4offerssomeinsightintohowsomeofthelightshadedareasinFigure3mightbe
accommodatedanalytically.Aftercharacterizingtraditionalbenefitcostandriskbased
approachesinitsfirsttworows,thelastrowdrawsattentiontoathirdtypeofanalysis:
multicriteriaapproachesdesignedtoilluminatevulnerabilitiesacrossthesocially
contingentimpactscalledoutbytherightcolumnofFigure3.Althoughpractical
approacheshaveyettobedeveloped,itislikelythatmuchofthisanalysiswouldidentify
thresholdsofsociallyunacceptableclimatechangeorclimatestress.Totheextentthatthis
istrue,theriskprofilesdescribedabovefortheriskmanagementperspectivecouldbe

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 219


applied.Multipleandpotentiallynonmonetarymetricshavealreadybeenaccommodated,
andmanyhavebeenexpressedintermsofthelikelihoodofcrossingcriticalthresholdsset
bynaturalsystems.Puttinghumansintothebusinessofdefiningcomparablethresholds
basedontheirvalues,institutions,andstateofknowledgeaddscomplicationtothe
analysis,butriskprofilescanaccommodatethesemetrics,aswell.

4.2 KeyVulnerabilitiesandMultipleMetrics
TheauthorsofChapter19ofIPCC(2007b)seizedonthecontentofFigure3(asportrayed
inChapter20)tounderscoretheneedformultipleimpactmetricsastheyexaminedand
identifiedkeyvulnerabilitiestoclimatechange.16Theybegantheirworkbyarguinghow
keyvulnerabilitiescouldbeidentifiedonthebasisofanumberofcriteriathatcouldbe
foundintheliterature:magnitude,timing,persistence/reversibility,thepotentialfor
adaptation,orlackthereof,distributionalaspects,likelihoodandimportanceofthe
impacts.Leavingthelastcriterion,importance,totheeyeofthedecisionmaking
beholder,theyofferedanillustrativelistbasedonnotonlytheirexpertassessmentsofthe
literature,butalsotheinsightsofferedbytheauthorsofthesectoralandregionalchapters
ofIPCC(2007b).
Thecontentoftheirworkhasbeenmosteffectivelycommunicatedthroughchangesinfive
aggregatereasonsforconcernfirstdevelopedforandpresentedintheIPCCsThird
AssessmentReport(IPCC,2001a,2001b).Thesemetrics,onlytwoofwhicharecalibrated
predominantly(butnolongerexclusively)intermsofeconomicmeasures,include:
Risktouniqueandthreatenedsystemsspeakstothepotentialforincreased
damagetoorirreversiblelossofuniqueandthreatenedsystemssuchascoral
reefs,tropicalglaciers,endangeredspecies,uniqueecosystems,biodiversity
hotspots,smallislandstates,andindigenouscommunities.Thereisnewand
strongerevidenceofobservedimpactsofclimatechangeonuniqueand
vulnerablesystems(suchaspolarandhighmountaincommunitiesand
ecosystems),withincreasinglevelsofadverseimpactsastemperaturesincrease
further.Anincreasingriskofspeciesextinctionandcoralreefdamageis
projectedwithhigherconfidencenowthanintheThirdAssessmentReport
(IPCC,2001b)aswarmingproceeds.
Riskofextremeweathereventstracksincreasesinextremeeventswith
substantialconsequencesforsocietiesandnaturalsystems.Examplesinclude
increaseinthefrequency,intensity,orconsequencesofheatwaves,floods,
droughts,wildfiresortropicalcyclones.Thereisnownewandstrongerevidence
ofthelikelihoodandlikelyimpactsofsuchchanges,suchastheIPCC(2007b)
conclusionthatitisnowmorelikelythannotthathumanactivityhas

16
Vulnerabilities,here,aredefinedasisnowmostusualintermsofexposuretoanticipatedimpactsand
associatedsensitivitiesthatcanbeamelioratedbyexercisingavailableadaptivecapacity.Sinceallthreeofthese
componentsofthevulnerability(exposure,sensitivityandadaptivecapacity)aresitespecificandpath
dependent,anabilitytoaccommodatethediversitynotedinsubsection2.1remainscritical.

220 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


contributedtoobservedincreasesinheatwaves,intenseprecipitationevents,
andintensityoftropicalcyclones.
Distributionofimpactsandvulnerabilitiesconcernsdisparitiesofimpacts,i.e.
whetherthepooraremorevulnerablethanthewealthy.Someregions,
countries,andpopulationsfacegreaterharmfromclimatechangewhileother
regions,countries,orpopulationswouldbemuchlessharmedandsomemay
benefit.Newresearchfinds,forexample,thatthereisincreasedevidencethat
lowlatitudeandlessdevelopedareasin,forexample,dryareasandmegadeltas
generallyfacegreaterriskthanhigherlatitudeandmoredevelopedcountries.
Also,therewilllikelybedisparateimpactsevenfordifferentgroupswithin
developedcountries.
Aggregatedamagescovercomprehensivemeasuresofimpactsfromclimate
change.Impactsdistributedacrosstheglobecanbeaggregatedintoasingle
metricsuchasmonetarydamages,livesaffected,orliveslost.Newevidence
supportstheconclusionthatitislikelytherewillbehigherdamagesforagiven
levelofincreaseinaverageglobaltemperaturethanpreviouslythought,and
climatechangeoverthenextcenturywilllikelyadverselyimpacthundredsof
millionsofpeople.
Riskoflargescalediscontinuitiesrepresentsthelikelihoodthatcertain
phenomena(sometimescalledsingularitiesortippingpoints)wouldoccur,any
ofwhichmaybeaccompaniedbyverylargeimpacts,suchasthemeltingof
majoricesheets.Forexample,thereisnowbetterunderstandingthattheriskof
additionalcontributionstosealevelrisefrommeltingofboththeGreenlandand
Antarcticicesheetsmaybelargerthanprojectedbyicesheetmodelsassessedin
theIPCC(2007a,2007b)andthatseveralmetersofadditionalsealevelrise
couldoccuroncenturytimescales.
Figure5displaysthedifferencesintheburningembersthresholdsfortheseReasonsfor
Concernbetweenthe2001versionpublishedintheThirdAssessmentReport(IPCC,
2001a,2001b)andamorerecentinterpretationbymanyofthesameauthors(Smithetal.,
2009).Ingeneral,theauthorsjudgethatsignificantrisks,depictedbyredcoloringinthe
figure,occuratlowertemperaturesthanpreviouslyassumed.Theseauthorshaverevised
theirperceptionsofriskbasedonthepastdecadeofobservationsshowingthatboth
developinganddevelopedcountriesalikearemorevulnerabletoextremeweatherimpacts
thanpreviouslyrealizedandalsofromobservationsthattheclimatesystemmayreact
morestronglyandabruptlytowarming.
Thecriticalinsightstobederivedfromthisdiscussionisthatthenotionofrisk(asthe
productofprobabilityandconsequence)hasbeenfirmlyensconcedinthediscussionsof
impactsandbenefitsofclimatepolicyandthatinvestigationsofhowtorespondtoclimate
changehavebeguntorecognizethediversityofpotentialvulnerabilitiesbeyondthe
narroweconomicspectrumofaggregateandregionalimpactsthatcanbecalibratedin
currency.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 221


Figure5.Risksfromclimatechange,byReasonforConcern2001comparedwith2007.Climate
changeconsequencesareplottedagainstincreasesinglobalmeantemperature(C)after1990.Each
columncorrespondstoaspecificReasonforConcern(RFC).TheleftpaneldisplaystheRFCsfromthe
IPCCThirdAssessment(IPCC,2001a,2001b).TherightpanelpresentsupdatedRFCsasdescribedin
Smithetal.(2009)andrepresentsadditionalinformationaboutoutcomesordamagesassociatedwith
increasingglobalmeantemperature.Thecolorschemedepictsprogressivelyincreasinglevelsofrisk,
andshouldnotbeinterpretedasrepresenting"dangerousanthropogenicinterference,"whichisavalue
judgment.Thehistoricalperiod1900to1990warmedbyabout0.6Candledtosomeimpacts.This
figureaddressesonlyhowriskschangeasglobalmeantemperatureincreases,nothowrisksmight
changeatdifferentratesofwarming.Furthermore,itdoesnotaddresswhenimpactsmightberealized,
nordoesitaccountfortheeffectsofdifferentdevelopmentpathwaysonvulnerability.

OriginalReasonsforconcern(2001) UpdatedReasonsforconcern(2009)
5

GlobalWarmingRelativeto1990(C)
3

Future
2

Past
-0.6

4.3.AnotherReasonforConcernfromMartinWeitzman
Atfirstblush,Weitzman(2009)addsevenmorecomplexitytothevaluationproblemby
showingthatprofounduncertaintyaboutfundamentalparameterslikeclimatesensitivity
canoverwhelmanyeconomicestimateofclimatedamages.Inpractice,hisresultfollows
directlyfromourinabilitytoobservetheextremerangesofclimateimpactdistributions
withenoughfrequencytolearnanythingusefulabouttheirrelativelikelihoods.He
concludesthatuncertaintywilldominateanycalculationofexpectedclimatedamage
becauseevensystematiclearningovertimeaboutthecriticalvariablesisneverstrong
enoughtokeepexpectedmarginaldamagesfinite;andsohisresultargumentclearlycasts
doubtonresultsderivedfromeconomiccalibrationsofdamagesavoidedbymitigation.On
thepositiveside,hisresultindicatesthatthevalueofsometypesofinformationisfar

222 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


greater(andperhapsinfinitelygreater)thanthevalueofotherinformation.Itcan
thereforeoffersomeguidanceonwheretodevotescarceresearchresourcesinclimateand
policyscience.Moreover,offerssoundtheoreticfootingforageneralizedriskbased
approachdesignedexplicitlytoexamineandclarifythedefinitionoftolerableclimate
change.
Toexploretheimplicationsofhisresultalittlemorefully,itisappropriatetoputitmore
squarelyintothecontextofwhatweknowaboutclimatechange.Tol(2003),forexample,
workedwithinabenefitcostframeworkthatrecognizedmultipleregionswithandwithout
equityweighting.Hissimulationsacrossawiderangeofpossiblefuturesnotedthesmall
butnonzeroprobabilitythatutilityconsequencesofmarginalimpactscouldgrow
infinitelylargeinoneormoreregionswheresomenotimplausibleclimatefuturescould
driveeconomicactivitytosubsistencelevels.Aslongastheseregionsweregivennonzero
weightintheexpectedwelfarecalculation,theirplightwoulddominatethepolicycalculus
becauseexpectedmarginaldamageswouldapproachinfinity
Yohe(2003)suggestedthattheproblemhighlightedinTol(2003)couldbeovercomeby
implementingasecondpolicyinstrumentdesignedtomaintaineconomicactivityabove
subsistencelevelseverywhereaforeignaidprogramdesignedsimplytoprevent
economiccollapseanywhereinrealtime,evenifcollapsehappenstobetheresultofan
extremeclimateimpactsomeplaceintheworld.TolandYohe(2007)examinedthis
suggestionwithintheoriginalmodelingframeworkandfoundthat,withsufficientaid,the
issueofinfinitemarginaldamagecouldbeavoided.Whilethisworkdidnotenvision
globallydistributedextremesasreflectedinWeitzmanscharacterizationofuncertainty
surroundingclimatesensitivity,itnonethelesssuggestedthattimelysocialoreconomic
interventionsthateffectivelylopoffthethicktailsofregionalclimateimpactscould
undercutthepowerofhisresult.If,however,thecatastrophewerefeltglobally,then
virtuallyanyinsuranceorcompensationschemebasedontransfersfromwellofftoless
welloffregionswouldbreakdownbecausenondiversifiableriskwouldbeunbounded.It
ishere,therefore,thatageneralizedprecautionaryprinciplethelogicalimplicationof
Weitzmansinsightisanappropriateframefromwhichtoderiveapotentialresponse.
Yohe(2009)goessofarastosuggestthatitisherethattheanalogytohedgingagainst
deflationintheconductofmonetarypolicycarriesthemostweight.Indeed,Weitzmanhas
recentlyusedthisanalogytoargueforspendingupto3percentofGDPperyearfor
hedginginsurance(TheEconomist,2009).
YoheandTol(2009)nonethelesssuggestthatthepolicycommunityshouldinsteadaskthe
researchcommunitytodevelopgreaterunderstandingsofthefundamentalprocessesin
areasotherthanclimatesensitivityprocessesthatproducecatastrophicimpactsfrom
whateverclimatechangehappenstomaterialize,forexample.Eveniftheycannotrelyon
thescientificcommunitytoreducetherangeofpossiblescenariosinthetemperature
domain,theycouldaskitto(1)explorethetriggersofmoreregionalcatastrophe,(2)
identifytheparametersoffundamentalchangethatdefinethosetriggers,(3)contributeto

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 223


thedesignofmonitoringmechanismsthatcantrackthepaceofchangerelativetothese
triggers,and(4)conductsmallandlargescaleexperimentsinmodels,laboratoriesand
perhapstherealworldtolearnmoreabouttherelevantprocesses.
Riskprofilesofthesortdisplayedabovecanalsoprovidesomecriticalinsightsinintothe
practicalapplicabilityofWeitzmanswarningaboutthicktailsintheclimatesystem.One
caneasilyseeinFigure1,forexample,thatanydecisionmakerconcernedwithprotecting
NewYorkCityfromtherisksassociatedwithincreasesinthefrequencyofthecurrent100
yearfloodinganomalywouldbecomeacutelyconcernedwithinthenextdecadeortwo.
Onecan,aswell,seethatthisconcerndoesnotnecessarilydependonathorough
understandingofthedistributionofclimatesensitivity.Putanotherway,Weitzmans
insightcouldbelessrelevantincaseswhereclimatefuturesdrivenbythemiddleofthe
sensitivitydistributionproduceintolerableimpactsalittlebitfurtherintothefuturethan
theWeitzmanesqueextremesofthesamedistribution.Whyworryaboutthelow
probabilityextremeswhenevenhighprobabilityoutcomescouldbeintolerableafew
yearslater?
Tobeclear,thepointoffocusingonthelinksbetweenphysicalclimateprocessesand
potentiallycatastrophicimpactsataregionallevelisnottodismisstheneedforhedging
throughmitigationagainstcatastrophicgloballydistributedfuturesthatmightbehoused
intheextremesindistributionsofvariableslikeclimatesensitivity.Itis,instead,toinform
investmentsinadaptationthatcomplementglobalhedgingonthemitigationsideofthe
policyequation.

5.Sustainability,EquityandAttitudestowardRisk
Sustainability,equityandattitudestowardriskarecrosscuttingthemesthatpermeate
throughouteverythingnotedabove.Theabilityoftheresearchcommunitytoaccommodate
theirimplicationsintoanalyticaltechniquesisnotwelldeveloped,butitisnotdifficultto
demonstratethattheymatterandshouldthereforebeconsideredinriskmanagementand
policymaking.
Withregardtosustainability,forexample,therearesynergiesacrossthedeterminantsof
adaptiveandmitigativecapacitiesandtheprecursorsofsustainabledevelopment.Because
theymatchtoalargedegree,initiativesdesignedtopromoteprogresswithrespecttothe
MillenniumDevelopmentGoalscansupportclimatepolicy.Thenewsisnotallgood,
though,becausethepotentialforconflictingobjectivesisrealanddiversityconfounds
generalinsights.Withregardtothelater,whetherornotthelinksbetweenaneconomic
intervention(oranadaptation)anditsdesiredoutcomesarestrong,weak,oractuallyrun
countertoexpectedbenefitsisessentiallyanempiricalquestioninnearlyeveryinstance.
Andwhileitiswidelyknownthatunabatedclimatechangecanimpedeprogresstoward
achievingtheMillenniumDevelopmentGoals,forexample,thereissuchathingas

224 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


dangerousclimatepolicyadaptiveormitigativeprogramsthatretardeconomicgrowth
andtherebyundercuttheabilitytodevelopsustainably(e.g.,TolandYohe,2006).
Therelativeimportanceofequityandattitudestowardriskcan,perhaps,bestbedisplayed
formallybyexploringthedualrolesthattheyplayindeterminingtheproperdiscountrate
tobeappliedtomonetizeddamages.Inthisregard,itisessentialtorememberthatclimate
changeisalongtermproblemeveniftheappropriateapproachtodesigningpolicyisto
workiteratively.Greenhousegasemissionreductionsovertheneartermwouldmitigate
futuredamages,buttheywoulddolittletoalterthepresentclimateand/orthepresent
rateofchangeinclimateimpacts.Moreover,mitigationmustcontinuewellintothefuture
iflongtermobjectivesaretobeachievedandlongtermprogressistobesustained.Ina
costbenefitframework,therefore,thediscountedcostsofemissionabatementmustbe
justifiedbythediscountedbenefitsofavoidedimpactsinthefuture.Inariskmanagement
framework,thediscountedcostsofabatementmustbeminimizedsubjecttotheconstraint
ofachievingthedesiredreductionsinclimaterisk.Itfollowsfromeitherapproach,
therefore,thatanystatementaboutthedesirabilityofclimatepolicynecessarilycontainsa
valuejudgementabouttheimportanceoffuturegainsrelativetopresentandfuture
sacrifices.
Tounderstandwhy,itissufficienttorealizethatpeoplediscountfutureconsumptionfor
tworeasons.First,theyexpecttobecomericherinthefuture,andsotheyexpectan
additionaldollartobuylesshappinessthenthananadditionaldollarwouldbuytoday.In
economics,theamountofhappiness(orutility)anadditionaldollarcanbuyiscalledthe
marginalutilityofconsumption.Theinterestrateatwhichadollarwouldneedtogrowto
enticeitsownertoinvestitratherthanspendingittodayiscalledelasticityofmarginal
utilityofconsumption.Second,peopleareimpatient,preferringtoconsumenowrather
thanlater,regardlessoffuturecircumstances.Theinterestrateatwhichaninvesteddollar
wouldneedtogrowtoenticeitsownertoinvestinthefutureratherthanspending
impatientlyiscalledtherateofpuretimepreference.
Together,thesetwomotivesfordiscountingthefuturedrivethesocalledRamseydiscount
rate(denotedbyrbelow)thatwasdesignedtosustainoptimalsavingovertime(Ramsey,
1928).TheRamseyequationthereforeconsistsofthreecomponents:
r = + g
whereistherateofpuretimepreference,gisthegrowthrateofpercapitaconsumption,
andistheelasticityofmarginalutilityofconsumption.
Therateofpuretimepreferencecalibratesintertemporaltradingsothatindividualswho
anticipateconstantlevelsofconsumptionfromoneperiodtothenextwouldbewillingto
sacrificeonedollarofpresentconsumptionifheorshewouldbecompensatedwith$(1+
)ofextraconsumptioninthenextperiod.Highervaluesofreflecthigherdegreesof

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 225


impatiencebecausegreatercompensationwouldberequiredtocompensateforthelossof
$1incurrentconsumption.
Consumptionlevelsneednotbeconstantovertime,ofcourse,andthesecondterminthe
Ramseyequationworkstheimplicationofthisfactintothistradingcalculus.Whileg
measuresthegrowthrateofmaterialconsumption,greflectsthegrowthrateof
happinessmeasuredintermsofunderlyingpersonalutility.Ifconsumptionweretoclimb
byg100percentfromoneperiodtothenext,theneachfuturedollarwouldbeworthg
100percentless(assumingnoimpatienceso0).Itfollowsthatourindividualwould
considersacrificingonedollarincurrentconsumptiononlyifheorshecouldbe
compensatedbyanamountequalto$(1+g)inthefuture.
Incontemplatingwelfarebasedequivalenceofconsumptionovertime,itisnowclearthat
thistradingbasedaccommodationofgrowingconsumptionworksinexactlythesameway
asthepurerateoftimepreferenceindefiningtherateatwhichthefutureneedstobe
discounted.Putanotherway,ifoneconsideredempiricalestimatesforbothand,then
bothparametersshouldplayequallyimportantrolesindeterminingtheappropriate
discountrate.Perhapsbecauseimpatienceisintuitivelyclearwhiletheroleofthe
elasticityofmarginalutilityofconsumptionisnot,thedebateoverhowtodiscountthe
futurehasfocusedundueattentiononalmosttotheexclusionof.
Climatechangeisnotonlyalongtermproblem;itisalsoaveryuncertainproblemwiththe
potentialofreducingfutureconsumption(risk),andaproblemthatdifferentiallyaffects
peoplewithwidelydifferentincomes(inequity).Therateofpuretimepreferencespeaks
onlytothefirstcharacteristicoftheclimatepolicyproblemtheintergenerationaltime
scale.Theelasticityofmarginalutilityofconsumption,theparameter,speakstoallthree
characteristics(intergenerationalequity,theriskofuncertainbutnegativeoutcomes,and
differentialimpactsonpeoplewithdifferentincomes).First,itindicatespreciselythe
degreetowhichanadditionaldollarbringslessjoyasincomeincreasesforoneindividual.
Second,itcanbeinterpretedasameasureoftheutilityofanextradollarforarichperson
relativetotheutilityofanextradollarforapoorperson.Thisiswhyisoccasionally
referredtoastheparameterofinequityaversion.Third,itcanbeinterpretedasameasure
ofhowincreasesinconsumptionimprovewelfaremoreslowlythanreductionsin
consumptiondiminishwelfare.Thisiswhyisalsoreferredtoastheparameterofrisk
aversion;anditisinthisrolethatithelpsexplainwhyriskaversepeoplebuyinsurance.
Assuggestedintheopeningparagraphofthisfinalsection,thepurposeofthisbrief
discussionisnottoexplainexactlyhowsustainability,equity,andattitudestowardriskcan
beincorporatedintodeliberationsaboutclimatepolicy;thatisstillaworkinprogress.Itis,
instead,toconfirmthatthefirstprinciplesofriskmanagementapproachessupportthe
importancethatnegotiatorsplaceoneachconceptastheycontemplatehowtorespondto
nationsobligationsunderArticles2and4oftheUnitedNationsFrameworkConventionon
ClimateChangei.e.,toframeactionsthatwillhelpusavoiddangerousanthropogenic

226 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


interferencewiththeclimatesystemwhilehelpingthemostvulnerableamongustocope
withtheimpactsofresidualclimatechange.

6.ConcludingRemarks
IPCC(2007c)tellsusthatinsightsderivedfromariskbasedperspectiveshouldnowbe
insertedintopublicargumentsoverwhattodoaboutclimatechangeargumentsthat
haveheretoforetoooftenbeenstuckinafalsedichotomybetweenstrainedclaimsof
certainty(Theverdictisin,nowisthetimeforsignificantactionregardlessofcost,it
wontcostmuchanyway,etc),andimpassionedinvocationsthatgenericuncertainty
justifiesinaction(Climatechangeisuncertain,welackproof,mitigationistooexpensive,
R&Dalonewillsolvetheproblem,etc).Sensibledecisionsandprudentmanagementof
risksrequirethatweworkinthemurkyarenabetweenthesetwoextremesby
acknowledgingthatcopingwithuncertaintywillplayimportantrolesinboththe
identificationofpolicyobjectivesandthedesignofspecificpolicyinitiatives.Peopledonot
ignoreuncertaintywhenmakinginvestmentsandpurchasinginsurance,norshould
analystsandpolicymakersignoreuncertaintywhenassessingclimatechangepolicies.
Thevarioussectionsofthispapercanperhapsoffersomepreliminaryguidanceintohowto
findourwaythroughthemurk.Section2tellsusnottobetooambitiousto
acknowledgethatmidcoursecorrectionswillberequired;andsoitfollowsthatgreater
attentionhastobepaidtoexactlyhowtodesignaprocessbywhichthesecorrectionscan
beaccomplished.Section3tellsusthatthelensofriskmanagementcanbeproductivein
thisregard;somemacroscalepolicieshavealreadybeenframedintermsofhedging
againstparticularlytroublingpossibilities,buttherearenoguarantees.Section4tellsus
nottoexpectthatalloutcomescanorshouldbequantifiedinunitsofcurrency;benefitcost
analysesmaybethetraditionalstandardfordecisionanalysis,buttheymustbe
complementedbyriskbasedapproachesthatcan,whenuncertaintydominates,carrythe
dayaspoliciesaredesigned.Section5addstheambiguityofimbeddingclimatechoices
intodiscussionsofsustainabilitythatrecognizeattitudestowardinequalityandrisk.Every
participantinthepolicydiscussionsmustunderstandthathisorherattitudesaboutboth
inequality(acrosstimeandspace)anduncertaintyarevalueladenperspectivesthathave
farreachingconsequences.Difficultiesincreatingandinterpretingaggregateand
disaggregatedindicesofrisksurelypersist,butadoptingariskperspectivewillbringnew
claritytoourunderstandingofthediversityandcomplexityoftheclimateproblem.The
strengthofcollectionsofdirectorevenqualitativeprofilesofriskliesintheirabilityto
accommodatealternativemetricsofvulnerabilityand/orreasonsforconcerninwaysthat
allowcomparisonsacrosscontext.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 227


Acknowledgements
Thepaperbenefitedimmeasurablyfromtheconstructivecommentandcriticismofferedby
JayGulledgeandhiscolleaguesatthePewCenteronGlobalClimateChange,participantsat
theSymposiumonAssessingtheBenefitsofAvoidedClimateChangesponsoredbythePew
CenterinMarchof2009,andtwoanonymousreviewers.Theauthorthankseveryonefor
theirgraciousassistance.Errors,thoughdifficulttodetectgiventheubiquitousuncertainty
andenormousspatialandtemporalscalesoftheclimateproblem,willresidepermanently
withauthor.

Appendix
FurtherdiscussionofexistingdamageestimatesforclimatechangeimpactsinFigure3,
areasIIV.
Someinvestigators,notablyWestandDowlatabadi(1999),Yoheetal.(1999),andStrzepek
etal.(2001),havetriedtocapturethemarketbasedimplicationsofextremeeventswhose
intensitiesandfrequencieshaveorwillbealteredbyachangingclimate,buttheireffortsto
addcontenttoareaIIhavebeenmostsuccessfulwhenframedinthelimitingcontextof
impactthresholdsbeyondwhichclimatevariabilityproducesseveredamage.
NordhausandBoyer(2000)wereessentiallyaloneintheirinitialattempttoincorporate
abruptclimatechangeintoclimatedamageestimates;ofcourse,Sternetal.(2006,2007)as
wellasNordhaus(2008)contributedtothissmallliteratureinattemptstoexpandour
understandingofareaIII.Itisimportanttonote,however,thatnoneoftheirapproaches
areanchoredonrobustanalysesofeconomicdamagesthatmightbeproducedalongabrupt
climatechangescenarios.Theyare,insteadgenerallyinferredfromriskpremium
calculationsbasedonunderlyingutilitystructuresandratherarbitraryassumptions.
NordhausandBoyer(2000),forexample,assignedlowprobabilitiestolargeeconomic
costs(ontheorderof10percentofglobaleconomicactivity)forthemiddleofthiscentury.
Theseassumptionsallowedthemtoreportestimatesofthewillingnesstopaytoavoidsuch
riskestimatesthatareequivalenttothemaximumamountthatanindividualwouldbe
willingtopayforperfectinsurancethatwouldeliminate(atacost)allclimaterelated
uncertaintyaboutthefuture.Sincenosuchinsuranceisavailable,though,theseestimates
shouldbeviewedasindicesoftheeconomiccostofcatastrophicclimatechange.
Afewstudies,authoredforexamplebyNordhausandBoyer(2000),Tol(2002a,2002b),
Sternetal.(2006,2007),andNordhaus(2006)amongothers,havetriedtoincludesome
(butcertainlynotall)nonmarketimpactsdrivenbytrendsinclimatechange(areaIV).
Theirrepresentationsarenot,however,particularlycomprehensivesincedataarelimited
andestimationmethodsaresometimesextremelycontroversial.
Thesameauthorstriedtobringassessmentsofnonmarketimpactsofextremeweather
eventsintotheirintegratedassessmentsofclimatechange;thatis,theytriedtowork

228 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


constructivelyinareaV.Theireffortshave,however,alsobeenseverelylimitedbyapaucity
ofrobusteconomicestimatesofimpacts.LinkandTol(2004)madesomeprogressinthis
regard,butSternetal.(2006)wasthefirstattemptatcomprehensive(thoughmuch
criticized)inclusiontoattractmuchattention.Finally,theMillenniumEcosystem
Assessment(MEA,2005)alsocontributedtoareaIVandareaV,butthatworkstoppedwell
shortoftryingtoassigneconomicvaluestoecosystemservices.Moreover,whilevarious
workinggroupswithintheMEAprocessdevelopedscenarioswithinwhichthoseservices
producedutility,fewofthempaidmuchattentiontoclimatechangeasadriverofrisk.

References
Allen,M.andD.Frame.2007.AbandontheQuest.Science328:582583.
Andronova,N.G.andM.E.Schlesinger.2001.ObjectiveEstimationoftheProbabilityDensityFunctionfor
ClimateSensitivity.JournalofGeophysicalResearch106:60522.
CongressionalBudgetOffice(CongressoftheUnitedStates).2005.UncertaintyinAnalyzingClimateChange:
PolicyImplications.Washington,D.C.36p.
Downing,T.andP.Watkiss,P.2003.TheMarginalSocialCostsofCarboninPolicyMaking:Applications,
UncertaintyandaPossibleRiskBasedApproach.PaperpresentedattheDEFRAInternational
SeminarontheSocialCostsofCarbon,July2003.
Greenspan,A.2003.Openingremarks,MonetaryPolicyandUncertainty:AdaptingtoaChangingEconomy.
InMonetaryPolicyandUncertainty:AdaptingtoaChangingEconomy,FederalReserveBankofKansas
City,pp.17.
Greenspan,A.2004.Riskanduncertaintyinmonetarypolicy.AmericanEconomicReview94:3340.
IntergovernmentalPanelonClimateChange(IPCC).2001a.ClimateChange2001:Impacts,Adaptation,and
Vulnerability.NewYork,CambridgeUniversityPress,2000pp.
IntergovernmentalPanelonClimateChange(IPCC).2001b.ClimateChange2001:SynthesisReport.A
ContributionofWorkingGroupsI,II,andIIItotheThirdAssessmentReportoftheIntergovernmentalPanel
onClimateChange.NewYork:CambridgeUniversityPress.398pp.
IntergovernmentalPanelonClimateChange(IPCC).2007a.ClimateChange2007:ThePhysicalScience
Basis,CambridgeUniversityPress,Cambridge.
IntergovernmentalPanelonClimateChange(IPCC).2007b.ClimateChange2007:Impacts,Adaptationand
Vulnerability,CambridgeUniversityPress,Cambridge.
IntergovernmentalPanelonClimateChange(IPCC).2007c.ClimateChange2007:SynthesisReport,
CambridgeUniversityPress,Cambridge.
Jones,R.N.2004a.ManagingClimateChangeRisks.InAgrawal,S.andCorfeeMorlot,J.(eds.),TheBenefitsof
ClimateChangePolicies:AnalyticalandFrameworkIssues,OECD,Paris,249298.
Jones,R.N.2004b.Incorporatingagencyintoclimatechangeriskassessments.ClimaticChange67:1336.
Jones,R.andG.Yohe.2008.ApplyingRiskAnalyticTechniquestotheIntegratedAssessmentofClimate
PolicyBenefits.IntegratedAssessmentJournal8:123149.
Kirshen,P.,C.Watson,E.Douglas,A.Gontz,J.Lee,andY.Tian.2008.CoastalFloodingintheNortheastern
UnitedStatesduetoClimateChange.MitigationandAdaptationStrategiesforGlobalChange.
Link,P.M.andR.S.J.Tol.2004.PossibleEconomicImpactsofaShutdownoftheThermohaline
Circulation:AnApplicationofFUND.PortugueseEconomicJournal3:99114.

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 229


Mastrandrea,M.D.andS.H.Schneider.2004.Probabilisticintegratedassessmentof"dangerous"climate
change.Science304:571575.
MillenniumEcosystemAssessment(MEA).2005.IslandPress,Washington,D.C.
http://www.millenniumassessment.org/en/Index.aspx.
NewYorkPanelonClimateChange(NPCC).2009.ClimateChangeandNewYorkCity:CreatingFlexible
AdaptationPathwaysAnnalsoftheNewYorkAcademyofSciences,NewYork.
Nordhaus,W.1991.ToSoworNottoSlow:theEconomicsoftheGreenhouseEffect.TheEconomicJournal
101:920937.
Nordhaus,W.2006.GeographyandMacroeconomics:NewDataandNewFindings,Proceedingsofthe
NationalAcademyofSciences(US)103:35103517.
Nordhaus,W.2008.AQuestionofBalance.YaleUniversityPress,NewHaven,234pages.
Nordhaus,W.andJ.Boyer.2000.WarmingtheWorldEconomicModelsofGlobalWarming.MITPress,
Cambridge,MA,232pages.
ONeill,B.2008.LearningandClimateChange:anIntroductionandOverview.ClimaticChange89:16.
Parson,E.2008.FiscalandRegulatoryApproachestoLimitingGreenhouseGasEmissions,NotesforKey
ElementsinBreakingtheGlobalClimateChangeDeadlock,OECDHeadquarters,www
personal.umich.edu/~parson/website/policy.
Raiffa,H.andR.Schlaiffer.2000.AppliedStatisticalDecisionTheory.WileyClassics,NewYork.
Ramsey,F.1928.AMathematicalTheoryofSavings.EconomicJournal38:543559.
Roe,G.andM.Baker.2007.WhyisClimateSensitivitysoUnpredictable?Science328:629632.
Schlesinger,M.,G.Yohe,J.Yin,N.Andronova,Malyshev,andB.Li.2006.AssessingtheRiskofaCollapseof
theAtlanticThermohalineCirculation.InAvoidingDangerousClimateChange(Schellnhuber,H.J.,
Cramer,W.Nakicenovic,N.Wigley,T.,andYohe,G.eds.),CambridgeUniversityPress.
Smith,J.B.,S.H.Schneider,M.Oppenheimer,G.W.Yohe,W.Hare,M.D.Mastrandrea,A.Patwardhan,I.
Burton,J.CorfeeMorlot,C.H.D.Magadza,H.M.Fussel,A.B.Pittock,A.Rahman,A.Suarez,andJ.P.van
Yperselen.2009.AssessingdangerousclimatechangethroughanupdateoftheIntergovernmentalPanel
onClimateChange(IPCC)reasonsforconcern.ProceedingsoftheNationalAcademyofScience
106(11):41334137,www.pnas.org_cgi_doi_10.1073_pnas.0812355106.

Stern,N.,S.Peters,V.Bakhshi,A.Bowen,C.Cameron,S.Catovsky,D.Crane,S.Cruickshank,S.Dietz,N.
Edmonson,S.L.Garbett,L.Hamid,G.Hoffman,D.Ingram,B.Jones,N.Patmore,H.Radcliffe,R.
Sathiyarajah,M.Stock,C.Taylor,T.Vernon,H.Wanjie,andD.Zenghelis(2007),Postscript,HMTreasury,
London.
Stern,N.H.,S.Peters,V.Bakhshi,A.Bowen,C.Cameron,S.Catovsky,D.Crane,S.Cruickshank,S.Dietz,
N.Edmonson,S.L.Garbett,L.Hamid,G.Hoffman,D.Ingram,B.Jones,N.Patmore,H.Radcliffe,R.Sathiyarajah,
M.Stock,C.Taylor,T.Vernon,H.Wanjie,andD.Zenghelis(2006),SternReview:TheEconomicsofClimate
ChangeCambridgeUniversityPress,Cambridge.
Stiglitz,J.andC.Walsh.2004.EconomicsW.W.Norton,NewYork.
Stott,P.A.,D.A.Stone,andM.R.Allen.2004.HumancontributiontotheEuropeanheatwaveof2003.Nature
432:610614.
Strzepek,K.,D.Yates,G.W.Yohe,R.S.J.Tol,andN.Mader.2001.Constructing"NotImplausible"Climateand
EconomicScenariosforEgypt.IntegratedAssessment2:139157.
TheEconomist.2009.IsItWorthIt?InGettingWarmerASpecialReportonClimateChangeandthe
CarbonEconomy,December5,2009,14pp.
Tol,R.S.J.2002a.EstimatesoftheDamageCostsofClimateChange:Part1:BenchmarkEstimates.
EnvironmentalandResourceEconomics21(1):4773.

230 Yohe:AddressingClimateChangethroughRiskManagement |PewBenefitsWorkshop


Tol,R.S.J.2002b.EstimatesoftheDamageCostsofClimateChange:Part2:DynamicEstimates,
EnvironmentalandResourceEconomics21(2):135160.
Tol,R.S.J.2003.IstheUncertaintyaboutClimateChangeTooLargeforExpectedCostBenefitAnalysis?
ClimaticChange56:265289.
Tol,R.S.J.andG.W.Yohe.2007.InfiniteUncertainty,ForgottenFeedbacks,andCostBenefitAnalysisof
ClimatePolicy.ClimaticChange83:429442.
Tol,R.S.J.andG.W.Yohe.2009.TheSternReview:ADeconstruction.EnergyPolicy37:10321040.
Walsh,C.E.2003.ImplicationsofaChangingEconomicStructurefortheStrategyofMonetaryPolicy.In
MonetaryPolicyandUncertainty:AdaptingtoaChangingEconomy,FederalReserveBankofKansasCity.
Webster,M.D.,C.E.Forest,J.Reilly,M.Babiker,D.Kicklighter,M.Mayer,R.Prinn,M.Sarofim,A.P.Sokolov,P.
Stone,andC.Wang.2003.UncertaintyAnalysisofClimateChangeandPolicyResponses.Climatic
Change,61:295320.
Weitzman,M.L.2009.OnModelingandInterpretingtheEconomicsofCatastrophicClimateChange.Review
ofEconomicsandStatistics91:119.
West,J.andH.Dowlatabadi.1999.OnAssessingEconomicImpactsofSeaLevelRiseonDeveloped
Coasts.InClimateChangeandRisk,London,Routledge.
WhiteHouse.2003.OMBCircularNo.A4,RegulatoryAnalysis,availableat
http://www.whitehouse.gov/omb/memoranda/m0321.html.
Wigley,T.,R.Richels,andJ.Edmonds.1996.EconomicandEnvironmentalChoicesintheStabilizationof
AtmosphericCO2Concentrations.Nature379:240243.
Wigley,T.M.L.2004.ModellingClimateChangeunderPolicyandNoPolicyPathways.InAgrawala,S.and
CorfeeMorlot,J.(eds.)TheBenefitsofClimateChangePolicies:AnalyticalandFrameworkIssues,OECD,
Paris,221248.
Yohe,G.W.2003.MoreTroubleforCostBenefitAnalysis.ClimaticChange56:235244.
Yohe,G.W.2004.SomeThoughtsonPerspective.GlobalEnvironmentalChange14:283286.
Yohe,G.W.2006.RepresentingDynamicUncertaintyinClimatePolicyDeliberations.Ambio35:3638.
Yohe,G.W.2009.TowardanIntegratedFrameworkDerivedfromaRiskManagementApproachtoClimate
ChangeClimaticChange95:325339.
Yohe,G.W.,M.Jacobsen,andT.Gapotochenko.1999.SpanningNotImplausibleFuturestoAssess
RelativeVulnerabilitytoClimateChangeandClimateVariability.GlobalEnvironmentalChange9:33
249.
Yohe,G.W.,N.Andronova,andM.Schlesinger.2004.ToHedgeorNotAgainstanUncertainClimateFuture.
Science306:415417.
Yohe,G.W.andD.Tirpak.2008.AResearchAgendatoImproveEconomicEstimatesoftheBenefitsof
ClimateChangePolicies.IntegratedAssessmentJournal8:117,2008.
Yohe,G.W.andR.S.J.Tol.2008.PrecautionandaDismalTheorem:ImplicationsforClimatePolicyand
ClimateResearch.Chapter7inRiskManagementinCommodityMarkets,HelyetteGeman(ed.),John
Wiley&SonsLtd.,Chichester,UK,pp9199.
Zickfeld,K.andT.Bruckner.2008.ReducingtheRiskofAtlanticThermohalineCirculationCollapse:
SensitivityAnalysisofEmissionsCorridors.ClimaticChange91:291316

PewBenefitsWorkshop|Yohe:AddressingClimateChangethroughRiskManagement 231