Thisarticleisalsoavailableonlineat:
www.elsevier.com/locate/ecolecon
SPECIALISSUE:TheDynamicsandValueofEcosystemServices:Integrating
EconomicandEcologicalPerspectives
Economicandecologicalconceptsforvaluingecosystem
services
StephenC.Farbera,*,RobertCostanzab,1,3,MatthewA.Wilsonc,2,3bGraduateSchoolofPublicandInternationalAffaris,Uni6ersityofPittsburgh,Pittsburgh,PA15260,USA
CenterforEn6ironmentalScienceandDepartmentofBiology,InstituteforEcologicalEconomics,Uni6ersityofMaryland,
Box38,1WilliamsSt.,Solomons,MD20688-0038,USA
cInstituteForEcologicalEconomics,Uni6ersityofMaryland,0216SymonsHall,CollegePark,MD20742-5585,USA
aAbstract
Thepurposeofthisspecialissueistoelucidateconceptsofvalueandmethodsofvaluationthatwillassistinguidinghumandecisionsvis-a`-visecosystems.Theconceptofecosystemservicevaluecanbeausefulguidewhendistinguishingandmeasuringwheretrade-offsbetweensocietyandtherestofnaturearepossibleandwheretheycanbemadetoenhancehumanwelfareinasustainablemanner.Whilewin-winopportunitiesforhumanactivitieswithintheenvironmentmayexist,theyalsoappeartobeincreasinglyscarceina‘full’globalecological–economicsystem.Thismakesvaluationallthemoreessentialforguidingfuturehumanactivity.Thispaperprovidessomehistory,background,andcontextformanyoftheissuesaddressedbytheremainingpapersinthisspecialissue.Itspurposeistoplacebotheconomicandecologicalmeaningsofvalue,andtheirrespectivevaluationmethods,inacomparativecontext,highlightingstrengths,weaknessandaddressingquestionsthatarisefromtheirintegration.©2002ElsevierScienceB.V.Allrightsreserved.
Keywords:Economicvaluation;ecologicalvaluation;ecologicalservices;valuation
1.Definitions
Theterms‘valuesystem,’‘value’,and‘valua-tion’havearangeofmeaningsindifferentdisci-plines.Inthispaper,weprovideapractical
*Correspondingauthor.Tel.:+1-412-648-7602
E-mailaddresses:eofarb@vms.cis.pitt.edu(S.C.Farber),costza@cbl.umces.edu(R.Costanza),mwilson@cbl.umces.edu(M.A.Wilson).1Tel.:+1-410-326-7263.2Tel.:+1-301-405-8952.3Asof09/01/2002,thisauthorcanbereachedattheGundInstituteforEcologicalEconomics,UniversityofVermont,SchoolofNaturalResources,GeorgeD.AikenCenter,BurlingtonVT05405–0088,USA.
synthesisoftheseconceptsinordertoaddresstheissueofvaluationofecosystemservices.Wewanttobeclearabouthowweusethesetermsthrough-outouranalysis.‘Valuesystems’refertointrapsy-chicconstellationsofnormsandpreceptsthatguidehumanjudgmentandaction.Theyrefertothenormativeandmoralframeworkspeopleusetoassignimportanceandnecessitytotheirbeliefsandactions.Because‘valuesystems’framehowpeopleassignrightstothingsandactivities,theyalsoimplypracticalobjectivesandactions.Weusetheterm‘value’tomeanthecontributionofanactionorobjecttouser-specifiedgoals,objec-tivesorconditions(Costanza,2000).Aspecificvalueofthatactionorobjectistightlycoupled
0921-8009/02/$-seefrontmatter©2002ElsevierScienceB.V.Allrightsreserved.PII:S0921-8009(02)00088-5
376S.C.Farberetal./EcologicalEconomics41(2002)375–392withauser’svaluesystembecausethelatterdeter-minestherelativeimportanceofanactionorobjecttootherswithintheperceivedworld.Wedefine‘valuation’astheprocessofexpressingavalueforaparticularactionorobject.Inthecurrentcontext,ecosystemvaluationrepresentstheprocessofexpressingavalueforecosystemgoodsorservices(i.e.biodiversity,floodprotec-tion,recreationalopportunity),therebyprovidingtheopportunityforscientificobservationandmeasurement.
Thedistinctionbetweenintrinsicandinstru-mentalvalueisanimportantone(Goulderetal.,1997).Ontheonehand,someindividualsmightmaintainavaluesysteminwhichecosystemsorspecieshaveintrinsicrightstoahealthful,sustain-ingconditionthatisonaparwithhumanrightstosatisfaction.Thevalueofanyactionorobjectismeasuredbyitscontributiontomaintainingthehealthandintegrityofanecosystemorspecies,perse,irrespectiveofhumansatisfaction.SomeinterpretLeopoldandAldo(1949)landethicasconstitutinganintrinsicvaluesystem,wheresomethingis‘rightwhenittendstopreservetheintegrity,stabilityandbeautyofthebioticcom-munity.Itiswrongwhenittendsotherwise.’Ontheotherhand,instrumentalvaluesreflectthedifferencethatsomethingmakestosatisfactionofhumanpreferences.Instrumentalvalues,suchaseconomicvalues,arefundamentallyanthropocen-tricinnature.Policiestowardtheenvironmentwillalwaystendtobebasedonamixofintrinsicandinstrumentalvaluesystems.Inthispaper,wedealwithboth.
2.Economicconceptsofvalue
Thehistoryofeconomicthoughtisrepletewithstrugglestoestablishthemeaningofvalue;whatisitandhowisitmeasured.Aristotlefirstdistin-guishedbetweenvalueinuseandvalueinex-change.Theparadoxofuseversusexchangevalueremainedunresolveduntilthe16thcentury(SchumpeterandJoseph,1978).Thediamond–waterparadoxobservedthatwhilewaterhasinfi-niteorindefinitevalue,beingnecessaryforlife,itsexchangevalueislow;yetunessentialdiamonds
bearahighexchangevalue.Followingthisobser-vation,therewaswidespreadrecognitionofthedistinctionbetweenexchangevalueandusevalueofgoods.Galianidefinedvaluetomeanarelationofsubjectiveequivalencebetweenaquantityofonecommodityandaquantityofanother.HenotedthatthisvaluedependsonUtilityandScarcity(utilitaetrarita)(SchumpeterandJoseph,1978).Twohundredyearslater,AdamSmithdistinguishedbetweenexchangevalueandusevalueofgoodsbycitingthediamond-waterparadox,butusedittodismissusevalueasabasisforexchangevalue.Smithformulatedacostofproductiontheoryofvalue,wherebywages,profitandrentarethethreeoriginalsourcesofexchangevalue.Inhisfamousbeaver–deerexam-plehesuggestedalabortheoryofexchangevalue:ifittakestwicethelabortokillabeaverthantokilladeer,onebeaverwillsellforasmuchastwodeer.Healsosuggestedalabor-disutilitytheoryofexchangevalue,notingthatgoodsexchangebasedupontheunpleasantnessofthelaborrequiredtobringthegoodstomarket.However,itissignifi-canttonotethatSmithlimitedhislabortheoryto‘thatearlyandrudestateofsocietywhichpre-cedesboththeaccumulationofstockandtheappropriationofland’.Inotherwords,whenlaboristheonlyscarcefactor,goodswillex-changebasedupontheratiooflaboruse(Schum-peterandJoseph,1978).
Inadditiontoformulatinghishypothesisre-gardingtheoriginsofexchangevalue,Smithsoughttoestablishaunitofmeasureofvalue,orwhathetermedtherealmeasureorrealpriceofagood.Heproposedthat‘labouralone…nevervaryinginitsownvalue,isalonetheultimateandrealstandard’ofthevaluesofallcommodities.Hencelaborcouldbeanumeraire,andithadspecialpropertiesofinvariantvalue(SchumpeterandJoseph,1978).
Ricardoalsosoughtaninvariantunitofmea-sureforvalue.Hefeltthattherewasnocommod-ity,includinglabor,whoseexchangevaluecouldserveasaninvariantstandardtomeasurethevariationinexchangevaluesofothercommodi-ties.Anditwasnotpossibletoaddupcommodi-tiestomeasurenationalwealthorproductionwithonlyexchangeratios.AccordingtoRicardo,
S.C.Farberetal./EcologicalEconomics41(2002)375–392377
thismeasuremustbeinvarianttochangesinrelativefactorrewards,i.e.capitalversuslabor,andbeacommoditywhosecapitalandlaborusedidnotvaryovertime,i.e.notechnologicalchange.Heproposedthatbothwheatandgoldpossessedtheseproperties(BlaugandMark,1968).Whilenotcreatingvaluetheycouldmea-surevalue.
WhileRicardohadseveralfollowers,includingJ.S.MillandMarx,labortheoriesofvalueandthepursuitofaninvariantstandardofvaluewanedinthelate19thcentury.Thiswaspartiallyinresponsetothelogicoftheutilitarians,suchasMenger,Gossen,JevonsandWalras,whoarguedthatexchangevaluewasbasedonbothutilityandscarcity(BlaugandMark,1968).Sraffa,anotedRicardianscholar,soughttoresurrecttheclassicalpursuitofatheoryofvalueindependentofde-mandorvalueinuse.Inhisbook,ProductionofCommoditiesbyMeansofCommodities:PreludetoaCritiqueofEconomicTheory,Sraffa(1960)es-tablishedconditionsunderwhichexchangeratiosbetweencommoditiescanbedeterminedbasedontheiruseinproduction;i.e.asetofcommoditypricesthatwouldexhaustthetotalproduct.Theseexchangeratioswerenotbasedonanyoptimalityormarginalityconditions.Instead,Sraffadividedcommoditiesintobasic(goodswhichenteredintoallproductionprocesses)andnon-basic,andshowedthataninvariantstandardofvaluewouldbeacombinationofbasiccommoditiesreflectingaverageinputproportionsinproduction.Thiscontrived‘commodity’wouldthenbeusableasameasureofnationalwealthorincome.3The‘marginal’revolutioninvaluetheoryorigi-natedwiththeconfluenceofseveralrelatedstreamsofeconomicthoughtinthe20thcentury.Mengerproposedthereweredifferentcategoriesofwantsordesires,suchasfood,shelter,clothing,etc.,thatcouldbeorderedintermsoftheirsub-jectiveimportance.Withineachcategory,thereis
3WhileacceptingSraffa(1960)mathematicalproof,somereviewers(Harrod,1961;Reder,1961)notedthattheexchangevalueswouldnotbeindependentofdemandasSraffaclaimed.ItwasfurthernotedthatSraffa’sdidnotconstituteapricetheoryinthesenseofestablishingtheprocessofpricedetermi-nation.
anorderedsequenceofdesiresforsuccessivein-crementsofeachgood.Hepostulatedthattheintensityofdesireforoneadditionalunitdeclineswithsuccessiveunitsofthegood(BlaugandMark,1968).Replacingtheterm‘desireforoneadditionalunit’withtheterm‘MarginalUtility,’wethushavetheeconomicprincipleofdiminish-ingmarginalutility.
Theideathatpeoplehavedifferent,butor-dered,categoriesofwantsordesiresraisesthecriticalissueofwhethertrade-offsexistbetweencategories.Ifindividuals‘weight’categories,itimpliesatrade-off.Atoneextreme,categoriesmaybelexicographicallyordered,likewordsinadictionary.Onelevelofwantmustbesatisfiedbeforealowerlevelbecomesrelevantinthepro-cessofvaluation.Therearenotrade-offsbetweenlevelsofwants.Forexample,theneedforcaloricintakeislikelysuperiortothatofrecreationalpleasure—nonumberofrecreationalopportuni-tieswilllikelysubstituteforaninsufficientdiet.Inthelexicographiccase,individualswouldusetheirmonetaryresourceshierarchically,satisfyinghigherorderwantsandneedsfirst.Whenahigherorderwantorneedisatrisk,theindividualwouldtakeresourcesawayfromlowerlevelonesuntilhigherlevelneedsweresatisfied.Lexicographicpreferencesdonotmeanmonetaryvaluationisimpossible,asindividualswouldstillbeabletostatehowmuchoftheirresourcestheywouldbewillingtosacrificeforagoodorservice;butitmaybealltheirresourcesifahighlevelneedisatrisk.
Moreproblematicforvaluationareinstanceswherebasicneedscannotbesatisfiedbythere-sourcesatanindividual’sdisposal—i.e.timeormoney.SimilartoMenger,Ekinsetal.(1992)suggestedtheuniversalityofbasichumanneeds,includingsubsistence,affection,protection,under-standing,leisure,identity,andfreedom.Althoughonecanimagineneedslikeaffectionbeing‘pur-chasable’withmoney,or‘freedom’beingpur-chasablebymigration,manyoftheseneedsmaynotbesatisfiedbymoneyortimebecauseindivid-ualssimplymaynotconsiderthemtobepur-chasablebymoneyortime.Thus,notonlyisitpossiblethattrade-offsbetweenneedswillnotbepossible,butsomeneedsmaynotbereducibletomoneyortime.
378S.C.Farberetal./EcologicalEconomics41(2002)375–392LancasterandKelvin(1971)introducedtheconceptofconsumptiontechnology,wherebycon-sumersconsidercharacteristicsofgoods.Forex-ample,foodmaybeevaluatedoncaloric,proteinorvitamincontent.Differentfoodsaresubsti-tutabledependingonthecompositionoftheircharacteristics.Peopleallocatetheirbudgetacrosscharacteristics,purchasinggoodsthatareefficientsourcesofdesiredcharacteristics.Thetechnologi-calinabilitytosubstitutecharacteristicsmayre-strictthemarginsonwhichenvironmentalgoodsandservicescanbevalued.Forexample,whilehealthmaybevalued,andindividualswouldbewillingtopayforit,thepropermixofcalories,proteinandvitaminsmaymakemarginalin-creasesordecrementsinoneofthesecharacteris-ticseitherveryhighlyvaluedorofverylowvalue.Buildingonthisinsight,multi-attributeutilitytheoryformalizestheutility-generatingtechnol-ogybyproposingthattotalutilityisafunctionofthecharacteristicsofgoodsorservices.Asimpleexamplewouldbewhereutility,U,fromfoodconsumptionisalinearfunctionofthecaloric,C,protein,P,andvitamin,V,content:U=aC+bP+cV.
(1)
Here,theparametersa,b,andcreflecttheweightingofthreefactorsindeterminingutilityforfoodconsumption.Whenutilitiesaremeasur-ableinmonetarywillingnesstopay(WTP)orwillingnesstoaccept(WTA)compensation,theseparametersrepresentthemarginalmonetaryvalueofeachcharacteristic.Thislogicformsthebasisforhedonicpricingmodelsofvaluation,discussedbelow,wherebythevalueofmarketgoods,sayahouse,dependsuponthecharacteristicsofthehouseanditslocation,aswellassurroundingenvironmentalamenitiesordisamenities.
Gossenproposedthatinordertomaximizesatisfactionfromagood,suchaslaborormoney,anindividualmustallocatethatgoodacrossdif-ferentusestoequatetheirmarginalutilitiesineachuse(BlaugandMark,1968).Hencemarginalutilitywouldprovideabasisforexplainingex-changevalue.Ifwetreatthingssuchasiron,cement,fertilizer,naturalagentsandlaborasincompleteconsumablegoods,themarginalutil-ityofthegoodstheyproducecanbeusedto
explaintheirexchangevalue.Thislogicestab-lishedafulltheoryofvalue.Italsodemonstratedthatexchangevaluescouldbebasedonusevalue.Whilethediamond–waterparadoxhadbeensolvedmanytimes,theclassicaleconomists,suchasSmithandRicardo,couldnotresolveitusingtheirlabortheoriesofvalue.Itwasresolvedonlybyrecognizingtheimportanceofutilityandscarcityindeterminingexchangevalues,andtheroleofmarginsinvaluedetermination.
Whiletheclassicaltheoristssoughtastandardphysicalcommodityunitformeasuringexchangevalue,neoclassicaltheoristsdidnotneedsuchacommodity.Asvaluewasassumedtobedeter-minedbyutilityonthemargin,andconsumerswereassumedtoallocatemoneyoptimallyacrossuses,themarginalutilityofmoneywasthesameforanindividualinallitsuses.Moneythusbecamethestandardunitofmeasure.
Thesignificanceofthemarginalutilitytheoryofvaluetotheevolvingconceptofecosystemservicevaluationisthatitcanbeusedtomeasureusevalues,notjustexchangevalues,inmonetaryunits.Thegeneraloptimizationmodeloflabor/leisureandconsumption/savinggiventimeandwealthconstraintswouldyieldequivalenciesofgoodsformoney,goodsfortime,andtimeformoney.Timeormoneycanthusbeusedasastandardofmeasureofusevalue;howmuchtimeormoneywillapersonwillinglysacrificetoob-taincommodityX?Insum,asthepursuitofaneconomictheoryofvaluetraversedthebroadmetaphysicalterrainofeconomicthought,theanswerappearstohavebeenfoundintheconceptofvalueinuse.4Theutility-basedvaluesofgoodsandservicesarereflectedinpeople’sWTPtoattainthem,ortheirWTAcompensationtoforegothem.WTPandWTAbecomemeasuresofthesevalues.Theymaybebasedonsmallmarginalchangesintheavailabilityofthesegoodsandservices,oronlargerchangesincludingtheircompleteabsenceor
4Sincethemarginalutilityofagooddependsuponhowmuchthepersonpossesses,wewouldexpectadifferencedependinguponwhetherthepersonisaskedhowmuchtheywouldsacrificetoobtainXorhowmuchwouldtheyacceptincompensationtoforegoX(seeHicks,1939).
S.C.Farberetal./EcologicalEconomics41(2002)375–392379
presence.ThesevaluationsarereflectedinFig.1.LetthecurveDrepresenttheWTPforeachunitofthegoodorservice,T,foranindividualorgroup.Thisisa‘Marginal’WTP.The‘Total’WTPforT0unitsofTistheaggregatedareasA+B.AreaAmaybeverylargeforgoodsorservicesthathavesomeutilitythresholdwherethegoodbecomesincreasinglyvaluableasitbecomesscarcer.Thisistrueformanyecologicalgoodsandservices,suchaslifesupportgoodslikeoxy-genandwater;the‘Marginal’valueisfinitebutthe‘Total’valueisindeterminate.Thisisthedistinctionthatliesbehindthediamond-waterparadoxnotedabove.
Exchange-basedvaluesarereflectedintheprices,P,atwhichthegoodsorservicesareexchanged.WhensupplyisT0,andtheitemissoldcompetitively,apricePisdeterminedwhichclearsthemarket.Thesepricesalsoreflectthe‘Marginal’valuationsplacedonavailablequanti-tiesaroundT0.Sopricesreflect‘Marginal’valueswhentherearemarketsforthegoodsorservices.5The‘Total’exchangevalueofT0isP×T0.Thisisanobservablemarketvaluewhentherearemar-ketstoobserve.Butwhentherearenosuchmarkets,Pmustbedeterminedindirectly,andP×T0wouldrepresentapseudo-marketvalue.Thiswouldbethe‘‘Total’’exchangevalueofthe
Fig.1.Utilityandexchangebasedvaluesofgoodsandser-vices.
5Unfortunately,thisisnotthecaseformanyunmarketedecologicalgoodsandservices—techniquesthateconomistshavedevelopedforassessingthe‘Marginal’valuesofgoodsareoutlinedbelowinSection5.
goodiftherewereamarketwithanavailablesupplyofT0.
Measuresofeconomicvaluearedesignedtoreflectthedifferencethatsomethingmakestosatisfactionofhumanpreferences.Ifsomethingisattainableonlyatacost,thenthedifferenceitmakestosatisfypreferencesisthedifferencebe-tweenitsutilityandthecostofattainingit.For-malconceptsofCompensatingandEquivalentVariationsareusedtoreflectthisdifference(VarianandHal,1992).Forexample,supposeinFig.1thatT0isavailableatacostofP.Underthesetermsofavailability,thewelfaredifferencemadebyT0isareaA.The‘Marginal’valuethatalterationsinavailabilitymaketowelfarewouldbereflectedbychangesinA.Usingtimberfromtreesasanexample,supposetimberisharvestedatacostofPperunitoftimber.Thevalueoftrees,perse,wouldberepresentedbyareaA,whichislessthanA+B.
Thusconceived,thebasicnotionofvaluethatguideseconomicthoughtisinherentlyan-thropocentric,orinstrumental.Whilevaluecangenerallymeanthecontributiontoagoal,objec-tive,desiredcondition,etc.,thementalmodelusedbyeconomistsisthatvalueisbasedonwantsatisfaction,pleasureorutilitygoals.Thingshavevalueinsofarastheypropelindividualstowardmeetingpleasureandneedobjectives.Valuesofobjectsintheenvironmentcanbeconsideredonthemargin,aswellasonthewhole;i.e.thevalueofoneadditionaltreeversusthevalueofalltrees.Whilevaluerelatestotheutilityofathing,theactualmeasurementofvaluerequiressomeobjec-tivemeasureofthedegreetowhichthethingimprovespleasure,well-being,andhappiness.Inafiniteworld,theresourcespeoplehaveavailabletomeettheirpersonalobjectivesarelimited.Economistshavethusdevelopedanex-tensivetheoryofhowpeoplebehaveinthepres-enceofconstraintsonfeasibleactivities(VarianandHal,1992).Theworkinghypothesisisthatpeoplemakedecisionsinordertooptimizesatis-faction,pleasureorutility.Thisoptimizational-waystakesplaceinthepresenceofconstraints,suchasincome,wealth,time,resourcesupply,etc.Optimizationthusyieldsadeterministicsetofpossibledecisionsinmostreal-worldsituations—whenconstraintschange,sodothedecisions.
380S.C.Farberetal./EcologicalEconomics41(2002)375–392Theessenceofthisperspectiveisthattheeco-nomicworldworkslargelydeterministically,mov-ingfromoneequilibriumtoanotherinrelativelystablefashion,andrespondstochangesincon-straintsinapredictablefashion.Thedetermina-tionofequilibriumisaresultantofconflictingforces,suchassupplyanddemand,orunlimitedwantsandlimitedmeans.Whiletherearein-stancesofinstability,disequilibriaandindeter-minism,thesearetreatedasexceptionsratherthantherule.
Sinceindividualscanbeobservedmakingchoicesbetweenobjectsinthemarketplacewhileoperatingwithinthelimitsofincomeandtime,economistshavedevelopedmeasuresofvalueasimputationsfromtheseobservedchoices.Whilemonetarymeasuresofvaluearenottheonlypossibleyardstick,theyareconvenientsincemanychoicesinvolvetheuseofmoney.Hence,ifyouareobservedtopay$10forabottleofwine,theimputationisthatyouvaluewinetobeatleast$10,andarewillingtomakeatrade-offof$10worthofotherthingstoobtainthatbottle.Themoneyitselfhasnointrinsicvalue,butrepresentsotherthingsyoucouldhavepurchased.Timeisoftenconsideredanotheryardstickofvalue;ifyouspend2hgolfing,theimputationisthatyouvaluethegolfexperiencetobeworthmorethan2hspentinotheractivities.Valueisthusaresultantoftheexpressedtastesandpreferencesofpersons,andthelimitedmeanswithwhichobjectscanbepursued.Asaresult,thescarcertheobjectofdesireis,thegreateritsvaluewillbeonthemargin.
Importantly,the‘technologies’ofpleasureandproductionallowforsomesubstitutionbetweenthings.Avarietyofgoodscaninducepleasureandarethustreatedconceptuallyasutilitysubsti-tutes.Abearmaysubstituteforanelkincon-sumption,hunting,andinawildlifeviewingexperienceeventhoughbearsandelkarenotsubstitutesintermsofecosystemfunction.Ontheproductionside,inputsarealsoconsideredtobesubstitutableforoneanother.Machinesandtech-nologycansubstituteforpeopleandnaturalin-puts.Clearly,economistsrecognizethattherelationsbetweengoodsandservicesareoftenmorecomplicatedthanthis.Formalnourishedpeople,sugarisnotechnologicalsubstituteforprotein,eventhoughtheybothprovidecalories.Asdiscussedearlier,preferencesmaybelexico-graphic—somethingsaremoreimportantthanothers,andcannotbesubstitutedforlowerlevelwantsorneeds.Ontheproductionside,nonum-beroflumbermenisasubstitutefortimberwhenthereisnotimber.Productionmayrequirecertaininputs,butatthesametimetheremaybesubsti-tutabilitybetweenothers.AsKrutillaandJohn(1967)suggests,theremaybeclosesubstitutesforconventionalnaturalresources,suchastimberandcoal,butnotfornaturalecologicalsystems.Theneoclassicalperspectivealsoassumesthattastesandpreferencesarefixedandgiven,andthatfundamentaleconomic‘problem’consistsofoptimallysatisfyingthosepreferences.Tastesandpreferencesusuallydonotchangerapidlyand,intheshortrun(i.e.1–2years),thisbasiceconomicassumptionisprobablynottoobad.Inthelongerrun,however,itdoesnotmakesensetoassumetastesandpreferencesarefixed.People’sprefer-encesdochangeoverlongertimeframesastheexistenceofarobustadvertisingindustryattests.Thisobservationisimportantbecausesustainabil-ityisaninherentlylong-runconceptandecosys-temservicesareexpectedtocontinueintothefarfuture.Thisfactisverydisturbingformanyeconomistsbecauseittakesawaytheeasydefini-tionofwhatisoptimal.Iftastesandpreferencesarefixedandgiven,thenwecanadoptastanceof‘consumersovereignty’andjustgivepeoplewhattheywant.Wedonothavetoknoworcarewhytheywantit;wejusthavetosatisfytheirprefer-encesefficiently.
However,ifpreferenceschangeovertimeandundertheinfluenceofeducation,advertising,changingculturalassumptions,andvariationsinabundanceandscarcity,etc.,weneedadifferentcriterionforwhatis‘optimal’.Moreover,wehavetofigureouthowpreferenceschange,howtheyrelatetothisnewcriterion,andhowtheycan,orshould,bechangedtosatisfythenewcriterion(Nortonetal.,1998).Onealternativeforthenewcriterionissustainabilityitself,ormorecom-pletelyasetofcriteria:sustainablescale(sizeoftheeconomicsubsystem),fairdistribution,andefficientallocation(Daly,1992).Thissetofcrite-
S.C.Farberetal./EcologicalEconomics41(2002)375–392381
riaimpliesatwo-tiereddecisionprocess(Page1977;DalyandCobb,1989;Nortonetal.,1998)offirstcomingtoasocialconsensusonasustain-ablescaleandfairdistributionand,second,usingthemarketplaceandothersocialinstitutionslikeeducationandadvertisingtoimplementthesede-cisions.Thismightbecalled‘communitysover-eignty’asopposedto‘consumersovereignty.’Itmakesmosteconomistsveryuncomfortabletostrayfromconsumersovereigntybecauseitraisesthequestion:iftastesandpreferencescanchange,thenwhoisgoingtodecidehowtochangethem?Thereisarealdangerthatatotalitariangovern-mentmightbeemployedtomanipulateprefer-encestoconformtothedesiresofaselecteliteratherthantheindividualsinsociety.
Here,twopointsneedtobekeptinmind:(1)preferencesarealreadybeingmanipulatedeveryday;and(2)wecanjustaseasilyapplyopendemocraticprinciplestotheproblemashiddenortotalitarianprinciplesindecidinghowtomanipu-latepreferences.Viewedinthislight,theafore-mentionedquestionistransformed:dowewantpreferencestobemanipulatedunconsciously,ei-therbyadictatorialgovernmentorbybigbusi-nessactingthroughadvertising?Ordowewanttoformulatepreferencesconsciouslybasedonsocialdialogueandconsensuswithahighergoalinmind?Eitherway,webelievethatthisissuecannolongerbeavoided,andisonethatwillbestbehandledusingopendemocraticprinciplesandin-novativethinking.Whichleadsusbacktotheroleofindividualpreferencesindeterminingvalue.Ifindividualpreferenceschangeinresponsetoedu-cation,advertising,andpeerpressurethenvaluecannotsolelyoriginatewithindividualprefer-ences.Valuesultimatelyoriginatefromwithintheconstellationofsharedgoalstowhichasocietyaspires—valuesystems—aswellastheavailabil-ityof‘productiontechnologies’thattransformthingsintosatisfactionofhumanneeds.
Inadditiontoincomeandeducation,timeplacesconstraintsonvaluecreation.Constraintsoftimeandintertemporalsubstitutabilitiescreatetemporalimplicationsforvalue.Economistspre-sumethatapresenttimepreferenceexistsduetolimitedtimehorizonsandconcernsforuncer-taintyinthefuture(Fisher,1930).Thismeans
individualswilldiscountvaluesofthingsinthefutureincomparisontothesamethingsinthepresent.IfIhaveanequalendowmentofapplesnowandayearfromnow,Iwouldplaceagreatervalueonhavinganapplenowthanonhavinganapple1yearfromnow.Theabilitytoconvertthingstomoneyinthepresenceofpositivefinan-cialinterestrateswill,therefore,resultinthe‘optimizingindividual’discountingthingsinthefuture.
Incontrasttoeconomists’traditionalassump-tionsofpositivetimepreferences,orpositivedis-countrates,psychologistssuggesttimepreferenceismorecomplicated.Forexample,LoewensteinandPrelec(1991)findthatinsomecircumstancespeoplebehaveasiftheyhavenegativetimeprefer-ence,preferringmoreinthefuturetomorenow.Theauthorssuggestthisisduetodread,theanticipationofsavoringbetterconditionsinthefuture,andtheaversiontoloss.However,thisnegativetimepreferencemaynotbeoperativewhenthetimeperiodisambiguous.Theimplica-tionsofsuchexperimentalresultsfordiscountinginenvironmentalpolicysettingsarenotclear,buttheydoraiseseriousquestionsaboutthestandardpracticeofdiscountingfutureenvironmentalbenefits(Clark,1973).
3.Ecologicalconceptsofvalue
‘Value’isatermthatmostecologistsandothernaturalscientistswouldprefernottouseatall,exceptperhapsinitscommonusageasareferencetothemagnitudeofanumber—e.g.‘thevalueofparameterbis9.32’.Usingthedefinitionofvalueprovidedearlier,ecosystemsandnon-humanspe-ciesarepresumednottobepursuinganycon-sciousgoals,and,therefore,theydonothavea‘valuesystem’.Likewise,onecannottalkabout‘value’asthedegreetowhichanitemcontributestoachievingagoalinthiscontextsincethereisnoconsciousgoalbeingpursued.Nevertheless,someconceptsofvalueareimportantinthenaturalsciences,andareinfactquitecommonlyused,andwetrybrieflytoelucidatethemhere.
Ifonelimitstheconceptofvaluetothedegreetowhichanitemcontributestoanobjectiveor
382S.C.Farberetal./EcologicalEconomics41(2002)375–392conditioninasystem,thenwecanseehownatu-ralscientistsusetheconceptofvalueallthetimetotalkaboutcausalrelationshipsbetweendiffer-entpartsofasystem.Forexample,onecouldtalkaboutthevalueofparticulartreespeciesincon-trollingsoilerosioninahighslopearea,orthevalueoffiresinrecyclingnutrientsinaforest.Thereareotherwaysinwhichtheconceptof‘value’isusedinthenaturalsciences.Forexam-ple,acoreorganizingprincipleofbiologyisevo-lutionbynaturalselection.Evolutioninnaturalsystemshasthreecomponents:(1)generationofgeneticvariationbyrandommutationsorsexualrecombination;(2)naturalselectionbyrelativereproductivesuccess;(3)transmissionviainfor-mationstoredinthegenes.Whilethisprocessdoesnotrequireconscious,goal-directedbehavioronthepartofanyofitsparticipants,onecanstillthinkoftheoverallprocessasbeing‘goal-di-rected’.The‘goal’of‘survival’isembeddedintheobjectivefunctionofnaturalselection.Whiletheprocessoccurswithoutconsciousnessofthisgoal,speciesasawholecanbeobservedtobehave‘asif’theywerepursuingthegoalofsurvival.Thus,oneoftenhearsevolutionarybiologiststalkaboutthe‘survivalvalue’ofparticulartraitsinorgan-isms.Naturalselectionmodels,whichmaximizethefitnessofspecies,arenotonlytestable,theybearclosesimilaritiestoeconomicutilitymaxi-mizationmodels(Low,2000).
Beyondthis,theideaof‘co-evolution’amongawholegroupofinteractingspecies(EhrlichandRaven,1964)raisesthepossibilitythatonespeciesis‘valuable’tothesurvivalofanotherspecies.Extendingthislogictotheco-evolutionofhu-mansandotherspecies,wecantalkofthe‘value’ofnaturalecosystemsandtheircomponentsintermsoftheircontributiontohumansurvival.Ecologistsandphysicalscientistshavealsopro-posedan‘energytheoryofvalue’,eithertocom-plementorreplacethestandardneoclassicaltheoryofvalue(Odum,1971,1983;Slesser,1973;Gilliland,1975;Costanza,1980;Clevelandetal.,1984;Halletal.,1992).Itisbasedonthermody-namicprincipleswheresolarenergyisconsideredtobetheonlyprimaryinputtotheglobalecosys-tem.ThistheoryofvaluerepresentsareturntotheclassicalideasofRicardoandSraffa(seeabove),butwithsomeimportantdistinctions.Theclassicaleconomistsrecognizedthatiftheycouldidentifya‘primary’inputtotheproductionpro-cessthentheycouldexplainexchangevaluesbasedonproductionrelationships.Theproblemwasthatneitherlabornoranyothersinglecom-moditywasreally‘primary’.
Theclassicaleconomistswerewritingbeforethephysicsofthermodynamicshadbeenfullydevel-oped.Energy—or,morecorrectly,‘free’or‘avail-able’energy—hasspecialcharacteristicswhichsatisfythecriteriafora‘primary’input:(1)En-ergyisubiquitous.(2)Itisapropertyofallofthecommoditiesproducedineconomicandecologicalsystems.(3)Whileothercommoditiescanprovidealternativesourcesfortheenergyrequiredtodrivesystems,theessentialpropertyofenergycannotbesubstitutedfor.Availableenergyisthustheonly‘basic’commodityandisultimatelytheonly‘scarce’factorofproduction,therebysatisfy-ingthecriteriaforaproduction-basedtheorythatcanexplainexchangevalues.
Energy-basedconceptsofvaluemustfollowthebasicprinciplesofenergyconversion.Thefirstlawofthermodynamicstellsusthatenergyandmatterareconserved.But,thislawessentiallyreferstoheatenergyandmechanicalwork(rawenergyorthebombcalorimeterenergy).Theabil-itytodoworkisrelatedtothedegreeoforganiza-tionororderofathingrelativetoitsenvironment,notitsrawenergycontent.Heatmustbeorganizedasatemperaturegradientbe-tweenahightemperaturesourceandalowtem-peraturesinkinorderforusefulworktobedone.Inasimilarfashion,complexmanufacturedgoodslikecarshaveanabilitytodoworkthatisnotrelatedtotheirrawenergycontent.Thesecondlawofthermodynamicstellsusthatusefulenergy(organization)alwaysdissipates(entropyordisor-deralwaysincreases)withinaclosedsystem.Inordertomaintainorganizedstructures(likeaneconomy)onemustconstantlyaddorganized,lowentropyenergyfromoutsidethesystem.
Estimatingtotal‘energy’consumptionforaneconomyisnotastraightforwardmatterbecausenotallfuelsareofthesamequality—i.e.theyvaryintheiravailableenergy,degreeoforganiza-tion,orabilitytodowork.Electricity,forexam-
S.C.Farberetal./EcologicalEconomics41(2002)375–392383
ple,ismoreversatileandcleanerinendusethanpetroleum,anditalsocostsmoreenergytopro-duce.Inaoil-firedpowerplantittakesfrom3–5kcalofoiltoproduceeachkcalofelectricity.Thus,addinguptherawheatequivalentsofthevariousformsoffuelconsumedbyaneconomywithoutaccountingforfuelqualitycanradicallydistortthepicture,especiallyifthemixoffueltypesischangingovertime.
Anenergytheoryofvaluepositsthat,atleastattheglobalscale,freeoravailableenergyfromthesun(pluspastsolarenergystoredasfossilfuelsandresidualheatfromtheearth’score)aretheonly‘primary’inputstothesystem.Labor,manufacturedcapital,andnaturalcapitalare‘in-termediateinputs’.Thus,onecouldbaseatheoryofvalueontheuseinproductionofavailableenergythatavoidstheproblemstheclassicaleconomistsencounteredwhentryingtoexplainexchangevaluesineconomicsystems.Therehavebeenafewattemptstoempiricallytestthistheoryusingbothtimeseriesdataandcross-sectionaldata.Studiesthathavetriedtoadjustforfuelqualityhaveshownaverycloserelationshipbe-tween‘availableenergy’consumptionandeco-nomicoutput.Clevelandetal.(1984)andmorerecentlyKaufmann(1992)haveshownthatal-mostallofthechangesinE/GNP(orE/GDP)ratiosintheUSandOECDcountriescanbeexplainedbychangesinfuelqualityandtheper-centofpersonalconsumptionexpenditures(PCE)spentdirectlyonfuel.ThelattereffectisduetothefactthatPCEisacomponentofGNPandspendingmoreonfueldirectlywillraiseGNPwithoutchangingrealeconomicoutput.Fig.2isanexampleoftheexplanatorypowerofthisrelationshipfortheUSeconomyfrom1932to1987.Muchoftheapparentgaininenergyeffi-ciency(decreasingE/GNPratio)isduetoshiftstohigherqualityfuels(likenaturalgasandprimaryelectricity)fromlowerqualityones(likecoal).RenewableenergysourcesaregenerallylowerqualityandshiftstothemmaycausesignificantincreasesintheE/GNPratio.
Anotherwayoflookingattherelationshipbetweenavailableenergyandeconomicoutputusescross-sectionalratherthantime-seriesdata.Thisavoidssomeoftheproblemsassociatedwith
Fig.2.Theenergy/GNPratiofortheUSeconomyfrom1932to1987.Thepredictedratio(PRED)isbasedonaregressionmodelwithpercentofprimaryenergyfrompetroleum(%PET)fromelectricity(%ELEC)andpercentofPersonalConsump-tionExpendituresspentonfuel(%PCE)asindependentvari-ables(R2=0.96).FromClevelandetal.(1984),Kaufmann(1992).
changesinfuelmixanddistortionsinGNP.Forexample,Costanza(1980),CostanzaandHerendeen(1984)usedan87-sectorinput–outputmodeloftheUSeconomyfor1963,1967,and1973,modifiedtoincludehouseholdsandgovern-mentasendogenoussectors(toincludelaborandgovernmentenergycosts)toinvestigatetherela-tionshipbetweendirectandindirectenergycon-sumption(embodiedenergy)anddollarvalueofoutput.Theyfoundthatdollarvalueofsectoroutputwashighlycorrelated(R2=0.85–0.98)withembodiedenergy,thoughnotwithdirectenergyconsumptionorwithembodiedenergycal-culatedexcludinglaborandgovernmentenergycosts.Thus,ifonemakessomenecessaryadjust-mentstoestimatesofenergyconsumptioninor-dertobetterassess‘availableenergy’,itappearsthattheempiricallinkbetweenavailableenergyandeconomicvalueisratherstrong.
Someneoclassicaleconomistshavecriticizedtheenergytheoryofvalueasanattempttodefinevalueindependentofconsumerpreferences(seeHeuttner,1976).Thiscriticismisaxiomaticasthestatedpurposewastoestablishabiophysicalthe-oryofvaluenotcompletelydeterminedbysocialpreferences.Theenergytheoryofvalueover-comessomeoftheproblemswithproduction-
384S.C.Farberetal./EcologicalEconomics41(2002)375–392basedtheoriesofvalueencounteredbytheclassi-caleconomistsdiscussedearlieranddoesarea-sonablejobofexplainingexchangevaluesempiricallyinthefewcaseswhereithasbeentested.Despitethecontroversyandongoingde-bateaboutthevalidityofanenergytheoryofvalue(BrownandHerendeen,1996),itseemstobetheonlyreasonablysuccessfulattempttooper-ationalizeageneralbiophysicaltheoryofvalue(seePatterson,thisvolume).
4.Ecologicalthresholds,uncertainty,andeconomicvalue
Ecosystemscanbehighlynon-linearwithincer-tainregions,andchangescanbedramaticorirreversible(seeLimburgetal.,thisvolume).Theavailabilityofecosystemservicesmaybedramati-callyalteredatthesenon-linearpointsforonlyminorchangesinecosystemconditions.Avalu-ableserviceprovidedtohumansbynaturallyfunctioningecosystemsistheiravoidanceofad-versethresholdconditions,orwhatCiriacy-Wantrup(1963)referredtoas‘CriticalZones’forresourceconservation.Forexample,treesinaforestedecosystemprovideahydrologicserviceofmoderatingwaterflowsintostreamsduringpeakstormevents.AsFig.3belowshows,letussup-posethereisarelationshipbetweenthedensityoftreesinalandscapeandphysicalseverityofdownstreamflooding.Attreedensitiesexceedingthe‘CriticalThreshold,’marginalchangesinden-sitycanbeevaluatedusingmeasuressuchasexpectedincreasesinflooddamages.Underthismarginalregime,thereisasubstitutefornature’s
Fig.3.Thefloodprotectionvalueoftrees.
services,floodprotectionorpropertyreplace-ment.Belowthecriticalthreshold,however,floodseverityincreasessubstantiallyastreedensitydi-minishes.Economicvalueschangesubstantiallyforslightalterationsinecosystemconditionsbe-causehumanlivesandcommunitiesmaybeatsubstantialrisk.Undertheseconditions,tradi-tionalmonetarymeasuresofvaluemaynotbeabletoadequatelycapturetheimpactofseverefloods.Traditionalvaluationmethodsmaynotbeacceptableasmeasuresofthevaluesoftreesinproximityofthe‘CriticalThreshold’.
Duetotheprobabalisticnatureofstormevents,humansocietymaywishtomaintaintreedensitieswellinexcessofthecriticalthreshold,sayatT*.TherewouldbeawelfarelossiftreedensitiesfellbelowT*,andthislosswouldbeattributabletoboththemarginalincreaseinfloodseverityandtothefactthatnowthesystemisclosertoacatastrophe.Therewouldbeaninsurancepre-miumthatsocietywouldpaytoavoidsuchadramaticchangeinecologicalstates.Additionaltreeswouldhavevaluebothfortheirroleinreducingexpectedflooddamages,andasin-suranceforavoidinganaturalcatastrophe.
Theexampleaboveillustratesthatecosystemservicevaluehasbothefficiencyandsustainabilitycomponents.Inthelinear,marginalregion,wheretheactualstatesoftheeconomicandecologicalsystemsarenotdramaticallyaltered,thevaluesofchangingtreedensitiesarerationallybasedonefficiencygoals;inthiscaseavoidinghavingtorepairflooddamages.Inthenon-linear,non-mar-ginalregion,however,thevalueoftreesisasustainabilityvalue,astheyprotecttheeconomicandecologicalsystemsfromcollapse.Sustainabil-ityvaluesmaybemoreimportantthanefficiencyvaluesaroundandbelowthresholdlimits.Inshort,sustainabilityvaluesmaybelexicographi-callysuperiortoefficiencyvalues.
TheexampleinFig.3hasthepropertyofreversibility.EvenwhentreedensitiesfallbelowtheCriticalThresholdlevelandplacesocietyathighrisk,plantingmoretreesreversestheexpo-suretorisk.Thismaynotbethecasewithsomeecosystemconditions.Forexample,reductionsintreedensitiesbelowtheCriticalThresholdmayalterlandscapeconditionsforalongperiodof
S.C.Farberetal./EcologicalEconomics41(2002)375–392385
Fig.4.Floodprotectionvaluesoftreeswithecosystemirreversibilities.
timeevenaftertreedensitieshavebeenincreasedtopre-thresholdconditions.
InFig.4,theecological–economicsystemmovesalongpathcdratherthanbaoncethethresholdofirreversibilityhasbeenviolated.Thisirreversibilitywouldlikelyincreasethevaluesoci-etywouldpaytoavoidthethresholdcomparedwithconditionsofrelativelyeasyreversibility.Theinsurancevaluewouldincludenotonlyapre-miumtoavoidacatastrophe,butanoptionvaluetoavoidtheirreversibilityofflooding(ArrowandFisher,1974).
TherecentwildfireinLosAlamos,NewMex-ico,inthesummerof2000,providesadramatic,tragicexampleofthecatastrophesandirre-versibilitiesassociatedwithbeingnearcriticalthresholds.ThefirewasstartedasacontrolledburnofseveralhundredacresbytheUSNationalParkService.Yearsofimproperforestmanage-ment,suchasnaturalfiresuppressionandgrazingofunderstoryvegetationcreatedacircumstanceinwhichaminorchange,thesmallcontrolledburn,haddisastrousconsequences,destroying300homesandtemporarilydisplacing30000people.Tomakemattersworse,thedestructionofgroundcoverovernearly50000acreswilllikelypermanentlyaltersoilconditionsaserosionwill
beverysevere.Theformerforestsystemmayneverbereplicated.ThissituationissimilartoconditionsillustratedinFig.4.
Anotherexamplemaybethevalueoftreesinalandscape.InFig.5,alterationsintreedensitiesabovethe‘CriticalThreshold’levelonlymar-ginallychangethevisualappealofthelandscape.However,belowthiscriticalthresholdtheland-scapeisnolongeraforest;thestateofnatureisalteredsubstantially.Changesintreedensitiesabovethecriticalthresholdcanbevaluedonthemarginusingtraditionaleconomicvaluationtech-niques.However,supposetheforestisacriticalvisualelementtoacommunity,orthelossofforesthasdireimpactsonthestateofthelocal
Fig.5.Thevalueoftreesforvisualappeal.
386S.C.Farberetal./EcologicalEconomics41(2002)375–392economyorsocialfabric.Changesintreedensi-tiesbelowthecriticalthresholdmaynotbemean-ingfullyvaluedusingtraditionaltechniques.Furthermore,thesametypeofinsurancevalueasinthecaseoffloodingwillgiveapremiumtoremainingabovethecriticalthreshold.Giventhereareprobabalisticeventssuchasstormsandinfestations,thecommunitymaywishtokeepdensitiesaboveT*.Increasesindensityintheregionaroundthecriticalthresholdhavevaluebothforimprovingthevisualappealoftheland-scapeaswellasprovidinginsurancethattreedensitieswillnotfallbelowthisthreshold.
Inalloftheaboveexamples,criticalthresholdsinecosystemstructureorfunctiondonotneces-sarilyimplyeconomicthresholdsforvalues.Forexample,iffloodingimpactsoncommunitieswereneversevereorpeoplewerenothighlydependentontheexistenceofaforest,marginaleconomicvaluationmethodswouldbeappropriateacrossthewholerangeoftreedensities.Thisisinspiteofthefactthattheremaybethresholdsofdensi-tiesatwhichecosystemstructuresandfunctionsaresubstantiallyaltered.Thenaturalworldmaybeinanon-linear,non-marginalcondition,buttheeconomicworldremainsasmoothonewheresubstitutesreadilymitigatesignificantecosystemchange.Ofcourse,theoppositemaybetruealso—i.e.gradualchangesinnaturalconditionsmayleadtonon-linearchangesineconomiccon-ditions.Forexample,waterqualitymaygraduallyfallbelowcertainstandardsandalakeis‘sud-denly’closedtoswimmers.
Criticalthresholdswhereecologicalconditionsanddynamicsareuncertainrequirevaluationun-deruncertainty.Uncertaintymayrangefromknowingtheprobabilitiesofconditionsandtheirvalues,toonlybeingabletoidentifythecondi-tionsbutnottheirprobabilities.Thereareseveralmethodsfordealingwithsuchvaluationdilem-mas.Forexample,supposeanecosystemunderStateAwouldprovide$200inservices,butinStateBwouldprovide$0inservices.Iftheprob-abilityofeachstateoccurringis0.5,theexpectedvalueoftheecosystemservicesis$100.Anexam-plewouldbethestormprotectionvalueofanacreofcoastalbarrierislandsunderahurricane(StateA)ornohurricane(StateB).
Table1
Netincomefromcoastalstormdamageswithandwithoutbarrierislands
BarrierislandconditionsBarrierislandpresent
NobarrierislandStormoccursI−C+$200I−CNostorm
I
I
ExpectedvalueI+0.5($200−C)I+0.5(−C)WorstcaseI−C+$200
I−C
Undertheseconditions,thevaluationofecosys-temservicesisnotquitesosimple.Individualsmaybeaversetorisksandalossmaybeweightedmoreheavilythanagainofcomparablemagni-tude.Giventhis,whatwouldbetheWTPtopreservethebarrierislands;whataretheyworth?Theanswerdependsonwhetherthedecisionmakerisriskaverse.Therearetwouncontrollablestates:StormversusNoStorm,eachoccurringwitha0.5probability.Therearetwoecologicalconditions:BarrierIslandandNoBarrierIsland.WhenbaseincomesareIandthebasedamagesfromastormwithnobarrierprotectionareC,thematrixshownbelowinTable1representsnetincomeconditionsunderthestormandbarrierislandoptions.
Usingexpectedvalues,thevalueofthebarrierislandsis$100.However,theWTPtomaintaintheislandsisgivenby:
0.5U(I−C+$200−WTP)+0.5U(I−WTP)=0.5U(I−C)+0.5U(I)
(2)
oranamountsuchthattheexpectedutility,netofWTP,ofmaintainingthebarrierislandsjustequalstheexpectedutilitywithouttheislands.Itcanbeshownthatunderconditionsofriskaver-sion,wheretheutilityfunctionisconcave,WTPwouldbegreaterthantheexpectedvalueoftheloss,$100inthiscase,butlessthanthefulldamageof$200.6TheexcessofWTPovertheexpectedvalueofthelossisthe‘premium’that
6WhenWTP=$100,rearrangingtermsshowsthatU(I−C+100)−U(I−C)\\U(I)−U(I−100)foranyconcaveutil-ityfunction.
S.C.Farberetal./EcologicalEconomics41(2002)375–392387
riskaverterswouldpayratherthanriskafullloss.
Thisexamplecanbegeneralizedinseveralways.First,ifthebarrierislandshavesomeaddi-tionalvalue,suchasrecreationaloraestheticen-joyment,thevalueoftheislandsmeasuredbyWTPwouldbeadditi6etothestormprotectionvalues.Second,whenalteringecologicalcondi-tionsincreasestheprobabilitythatalosswilloccur,riskavertingindividualsshouldbewillingtopaysomethingtoavoidtheincreaseinproba-bilityofloss.ThisWTPwouldreflectwhattheecosystemisworthinsofarasinsuringagainstcrossingthresholdsandencounteringadverseirre-versibleconditions.
Whenuncertaintyconsistsofnotknowingtheprobabilitiesofvariousecologicalstates,e.g.ofahurricane,theabovematrixcanbeusedtoillus-tratevaluationunderthispurerisksituation.Ifsocietyisriskaverse,ausefuldecisionruleistoassumetheworstwilloccur,andseektominimizetheworst-casescenario.Forexample,maintainingthebarrierislandsresultsinaworst-casescenariounderahurricaneofI−C+$200;whiletheworstcaseifislandsarenotmaintainedisI−C.Theimpliedvalueofthebarrierislandsforplan-ningpurposesis$200underthisriskaversedeci-sionrule,associetywouldbewillingtopayupto$200tomaintaintheseislands.
Inthisexample,theabilitytoestimatestormdamagecostsavingsprovidesguidancetovaluingtheresourceaswellasdevelopingadecisionrule.Asimpledecisionrulewouldbetomaintainandconserveanecosystemservicewhenthecostofdoingsoisnottoogreat.Thisisthe‘safemini-mumstandard’proposedbyCiriacy-Wantrup(1963),andelaboratedonbyothers(Bishop,1978).Underthisstandard,conservationpracticesavoidCiriacy–Wantrup’s‘CriticalZone’ofdra-matic,irreversiblechangeinecosystems.Forex-ample,soilconservationwouldavoidgulleysormaintainamaximumacceptableerosionrate;forestconservationwouldestablishamaximumdeforestationrate;rangelandsconservationwouldmaintainaminimumlevelofplantmaterialaftergrazing;orspeciesconservationwouldestablishaminimumbreedingstockorhabitatcondition.Thesestandardsareecologicallybased,noteco-
nomic;althoughviolatingthemmaybeprudentiftheeconomiccostsaretoohigh.
5.Conflictsbetweeneconomicandecologicalvalues
Wealsorecognizethateconomicandecologicalmeasuresofvaluemayattimesbeatoddswithoneanother.Ashumansareonlyoneofmanyspeciesinanecosystem,thevaluestheyplaceonecosystemfunctions,structuresandprocessesmaydiffersignificantlyfromthevaluesofthoseecosystemcharacteristicstospeciesorthemainte-nance(health)oftheecosystemitself.Theintrin-sicvaluesofnaturalsystemfeaturesandprocesseswithinthenaturalsystemitselfmaypossessdiffer-entabundanceandfunctionalvaluepropertiesthantheircorrespondingeconomicvalues.Dimin-ishingreturnsandutilitywouldsuggestsomeeconomicsaturationinthedemandsforparticularecosystemservicesandconditions.Forexample,themarginaleconomicvalueforadditionalsun-lightmaybezeroorpossiblynegative—skincan-cerfromexcessivesunlight,excessiveheat,etc.Thedifferencesbetweenecologicalandeco-nomicvaluesrelatetotherelativeabundanceofecosystemserviceswithinnaturallyfunctioningecosystemsandeconomies.Clearly,aservicecanbemoreabundantorscarceinonethananother.Whileitislikelythatspecificecosystemstructuresandprocesseshavesomefunctionalroleinanecosystem,and,therefore,have‘value,’theymaynothavedirectorindirectvalueinmarketeconomies.Theremaybeinstanceswhereanecosystemissoisolatedfromhumaneconomicactivitythatwhathappensinitisirrelevanttohumanactivity,evenwhenallpossiblespatialandtemporalconnectionsareconsidered—i.e.onlytheintrinsicvalueremains.Ofcourse,ashumanscontinuetoincreasinglyinhabittheplanet,theseinstancesbecomeincreasinglyrare.Asourunder-standingofconnectionsbetweenandwithinecosystemsexpands,wefindmoreandmorein-stancesofsignificantimplicationsforhumanbe-ings.Thesechangingconditionsinknowledgemakeitincreasinglyincumbentuponustoavoidthequickdismissalofisolatedorpresumably
388S.C.Farberetal./EcologicalEconomics41(2002)375–392economicallyirrelevantecosystemsortheirprop-ertiesasirrelevanttohumanwelfare.
6.Economicvaluationmethods
Theexchangevalueofecosystemservicesisthetradingratiosforthoseservices.Whenservicesaredirectlytradableinnormalmarkets,thepriceistheexchangevalue.Theexchange-based,wel-farevalueofanaturalgoodorserviceisitsmarketpricenetofthecostofbringingthatservicetomarket.Forexample,theexchange-basedvalueoftimbertosocietyisits‘stumpagerate,’whichisthemarketpriceoftimbernetofharvestandtimeallocatedmanagementcosts.Ex-change-basedvaluationisrelativelysimple,astradesexistfromwhichtomeasurevalues.
Marketpricesreflectthevaluationofgoodsandservices,butonlyonthemargin.Forexam-ple,thepriceofaboardfootoftimberreflectswhatanotherboardfootisworthtobuyers.Itdoesnotreflectthewholevalueofalltimberusedbythebuyers.Youmaypayonly$2perboardfootandthatisallyouwouldbewillingtopayforthelastof,say,the5000boardfeetyoubuy.Butyoumaybewillingtopayconsiderablymorethan$10000fortheopportunitytobuyall5000boardfeet.Ofcourse,thetimberreflectsonlyaportionofthefullsocialvalueofatree,whichalsoprovidesanarrayofservicessuchassoilamendmentandstabilization,waterstorageandfloodcontrol,specieshabitat,aesthetics,climatecontrol,etc.Inlimitedcases,marketsforenviron-mentalserviceshavebeenformedthattendtoreflectthevaluationsofthoseservices(ChichilniskyandHeal,1998).
Whileexchangevaluerequiresmarketsorob-servabletrades,thesocialvalueofservicesismuchmorebroadanddifficulttomeasure.Thesesocialvaluesarewhathavecapturedtheattentionofenvironmentalandresourceeconomists.Theyhavedevelopedanumberoftechniquesforvalu-ingecosystemservices(Freeman,1993;KoppandSmith,1993).Theunderlyingconceptsforsocialvaluesthateconomistshavedevelopedarewhatasocietywouldbewillingandabletopayforaservice,WTP,orwhatitwouldbewillingto
accepttoforegothatservice,WTA.Thetwovaluationconceptsmaydiffersubstantiallyinpractice(Hannemann,1991).
Theeconomicvaluationmethodologyessen-tiallyconstructsWTPforaservice;orconstructstheadequatecompensationforaserviceloss,representingWTA.Supposetheserviceisfloodcontrolprovidedbyawetland.Supposedamagesfromfloodingwere$1million.Societywouldthenbewillingtopay$100000toreducetheprobabil-ityoffloodingby10%ifthesociety,asawhole,isriskneutral.Supposethewetlandsreduceflood-ingprobabilitiesby20%.Whenwetlandsservicesarefree,societyreceives$200000millioninser-vicesfornothing.Inprinciple,theownerofawetlandprovidingsuchaservicecouldcaptureuptothisamountofsocialvalueiftherewasacapturemechanism.Marketsforresourceservicesprovidecapturemechanisms.Theyworkrelativelywellfor‘private’goods,whereownerscandenyaccesstotheserviceifpaymentsarenotmadeandifmakingaccessavailabletoonepersonessen-tiallymakesitunavailabletoothers.Rawmateri-alsandfoodproductionaregoodexamplesofthese‘private’goodsorservices.
Manyecosystemservicesdonotqualifyformarkettradingbecausetheyarenot‘private’innature.Forexample,floodprotectionservicesofwetlandsortrees,oncemadeavailabletoonepersonmayindirectlybecomeavailabletoall.WetlandsandforestownerscouldnotcaptureallthepotentialsocialWTPforthisservice.
Whentherearenoexplicitmarketsforservices,wemustresorttomoreindirectmeansofassess-ingeconomicvalues.Avarietyofvaluationtech-niquescanbeusedtoestablishtheWTPorWTAfortheseservices.Therearesixmajorecosystemserviceeconomicvaluationtechniqueswhenmar-ketvaluationsdonotadequatelycapturesocialvalue:
AvoidedavoidcostsCostthat(AC):wouldserviceshavebeenallowincurredsocietytointheabsenceofthoseservices;floodcontrolavoidspropertydamagesorwastetreatmentbywetlandsavoidshealthcosts.
Replacementplacedwithman-madeCost(RC):systems;servicesnaturalcouldbewastere-treatmentcanbereplacedwithcostlytreatmentsystems.
S.C.Farberetal./EcologicalEconomics41(2002)375–392389
FactorIncome(FI):servicesprovidefortheenhancementofincomes;waterqualityim-provementsincreasecommercialfisheriescatchandincomesoffishermen.
TravelCost(TC):servicedemandmayrequiretravel,whosecostscanreflecttheimpliedvalueoftheservice;recreationareasattractdistantvisitorswhosevalueplacedonthatareamustbeatleastwhattheywerewillingtopaytotraveltoit.
HedonicPricing(HP):servicedemandmaybereflectedinthepricespeoplewillpayforasso-ciatedgoods;housingpricesatbeachesexceedpricesofinlandhomes.
ContingentValuation(CV):servicedemandmaybeelicitedbyposinghypotheticalscenar-iosthatinvolvesomevaluationofalternatives;peoplewouldbewillingtopayforincreasedfishcatchordeerbag.
Eachofthesemethodshasitsstrengthsandweaknesses.Also,eachservicehasanappropriatesetofvaluationtechniques.Someservicesmayrequirethatseveraltechniquesbeusedjointly.Forexample,therecreationalvalueofanecosys-temwillincludenotonlythevaluethatvisitingrecreationistsplaceonthesite(TC),butthein-creasedincomesassociatedwithsiteuse(FI).Alexanderetal.(1998)havesuggestedanextremeFIforvaluingglobalecosystemservices,measur-ingtherentsthatahypotheticalmonopolisticownerofnature’sservicescouldchargetheworld’seconomy.Forexample,anextrememea-sureofrentfromallnaturalsystemserviceswouldbethedifferencebetweenglobalGDPandaglobalsubsistenceincome.ThepaperbydeGroot,WilsonandBoumans(thisvolume)dis-cussestheappropriatetechniquesforvaluingdif-ferentecosystemservices.
Somevaluationtechniques,whileintuitivelyap-pealing,maymisrepresentWTPorWTAvalua-tionconceptsincertaincircumstances.ThisisespeciallyaproblemwhenusingReplacementCost(RC)methods.Theremaybecircumstanceswhenthesocialbenefitsthatmaybelostwhenecosystemservicesareunavailablearelessthanthecostofreplacementofthoseservices;orwhenthebenefitsgainedfromenhancedservicesarelessthanalternativemeansofprovidingthoseser-vices.Forexample,theAvoidedCostofillnessunderanecosystemenhancement,suchaswet-landstreatmentofwaste,maybelessthanthecostofcomparablewastetreatmentfacilities.Inthiscase,AvoidedCostisamoreappropriatemeasureofvaluethanReplacementCost.TheReplacementCostmeasureofvalueoftheworld’scoralreefsmayfarexceedthemeasureofbenefits.
7.ThechallengeofaggregatingeconomicvaluesThetraditionalprocedureofeconomicvalua-tionistoestablishindividual-basedvaluesusingoneofthemethodsdescribedinSection5above.Isolatedindividualvaluesarethenaggregatedtorepresentasocially-relevantunit—acommunity,astate,anation,ortheentireplanet.Thisisappropriatewhentheservicesprovidedarepurelyindividuallyenjoyed,asisthecasefor‘private’goodsandservicesthatarenotsharedandwheretherearenosubstantialpositiveornegative(ex-ternality)impactsofoneperson’suseonanother.Thisisalsothecasefor‘public’goodswhereenjoymentremainsindividual-basedwithoutex-ternalityimpacts.Anexamplewouldbetherecre-ationalenjoymentofanuncongestedforest.
Isolated,individual-basedvaluationandaggre-gationarenotappropriate,however,ininstanceswheregroupvaluesmayhingeongroupinterac-tions,wherepreferenceformationispartiallyasocialprocess,wheresharedknowledgeisimpor-tant,andwhereitemsvaluedhavesubstantialinterpersonalorsocialimplications.Valuingaforestfortimber,orevenrecreationisappropri-atelyanindividual-basedprocess.However,othervaluesoftheforestmaybemorecommunal,arenotwell-definedinpreferencefunctions,orhavesubstantialinterpersonalimpacts.Forexample,thevalueofforeststoacommunitywhosesocialsystem,folklore,etc.areintimatelydependentonthemismorethanthesumofindependentper-sonalvalues.
Oneapproachtoecosystemservicevaluationthathasgainedincreasingrecognitionintheliter-atureissmallgroupdeliberation(Jacobs,1997;BlameyandJames,1999;CooteandLenaghan,1997).Derivedfrompoliticaltheory,thisevolving
390S.C.Farberetal./EcologicalEconomics41(2002)375–392setoftechniquesarefoundedonprinciplesofdeliberativedemocracyandtheassumptionthatpublicdecisionmakingshouldresult,notfromtheaggregationofseparatelymeasuredindivid-ualpreferences,butfromaprocessofopenpublicdebate(Fishkin,1991;Dryzek,1987;Habermas,1984).Thus,theapplicationofaparticipatorydemocracyapproachtoenviron-mentalissuesestablishestwovaliditycriteriathatsetitapartfromtraditionalnon-marketvaluationapproaches:decentralizedformsofen-vironmentalpolicyformulationandthedirectinvolvementofnon-expertsinsmalldecision-makinggroups(seeWilsonandHowarth,thisvolume).
Thebasicideaisthatsmallgroupsofcitizenscanbebroughttogethertodeliberateaboutthesocialvalueofpublicgoodsandthatthe‘con-sensus’valuesderivedinthisopenforumcanthenbeusedtoguideenvironmentalpublicpol-icy(Jacobs,1997).Inthismanner,discursivemethodssuchascitizens’juries(CooteandLenaghan,1997),consensusconferences(JamesandBlamey,1999),anddeliberativeCVtech-niques(Sagoff,1998)haveincreasinglybeenproposedandusedinNorthAmerica,Europe,andAustraliatoinformenvironmentaldecisionmaking.Oneassumptioncommontoallthesetechniquesisthatdeliberativebodiesofcitizenscanrenderinformedjudgmentsaboutenviron-mentalgoodsnotsimplyintermsoftheirownpersonalutility,butalsoforsocietyasawhole.Thepurposeofdeliberationisto‘reachagree-mentonwhatshouldbedonebyoronthebehalfofsocietyasawhole’(Jacobs,1997).Insum,opendiscourseisassumedtoperforma‘correctivefunction’wheneachcitizenalonehasincompleteinformation,butactingtogetherwithotherscanpiecetogetheramorecompletepic-tureoftruesocialvalueforecosystemgoodsandservices.
Forexample,wemightconsidertherecentlyproposeddeliberative,or‘group’CVtechnique(Sagoff,1998;Jacobs,1997).WhilethereisalongtraditionofgroupresearchinCV,thegoalofsuchresearchhasgenerallybeentousefocusgroupstoincreasethecontent-validityofhypo-
theticalscenariosanddiagnosepotentialprob-lemsthatindividualrespondentsmayhavewiththepaymentvehicle(MitchellandCarson,1989).WithagroupCV,ontheotherhand,theexplicitgoalwouldbetoderiveagroup-consen-susvaluefortheecologicalgoodorserviceinquestion.Thevaluationexerciseis,therefore,conductedinamannersimilartoaconventionalCVsurvey—usinghypotheticalscenariosandre-alisticpaymentvehicles—withthekeydifferencebeingthatvalueelicitationisnotdonethroughprivatequestioningbutthroughgroupdiscus-sionandconsensusbuildiing.Thus,thedelibera-tiveCVapproachtreatssmallgroupdeliberationnotasadiagnostictool,butasanexplicitmechanismforvalueelicitation.
8.Conclusions
Theconceptsof‘value’,‘valuesystem’,and‘valuation’havemanymeaningsandinterpreta-tionsandalonghistoryinseveraldisciplines.Wehaveprovidedasurveyofsomeofthesemeaningsastheyrelatetotheissueofecosys-temservicevaluationtoserveasbackgroundandintroductiontotheremainingpapersinthisspecialissue.Thereisclearlynotone‘correct’setofconceptsortechniquestoaddressthisim-portantissue.Rather,thereisaneedforcon-ceptualpluralismandthinking‘outsidethebox.’Thatiswhattheremainingpapersinthisspecialissueattempttodo.Whiletheybreaksomenewgroundandaddresstheissuesininterestingnewways,itisclearthatmuchadditionalworkre-mainstobedone.Afteralongandinterestinghistory,theissueof‘value’isnowgoingthroughaperiodofdevelopmentthatshouldhelpustomakebetter,andmoresustainable,decisions,notonlyasindividuals,butalsoasgroups,communities,andasstewardsoftheen-tireplanet.
Acknowledgements
ThisworkwasconductedaspartoftheWork-ingGroupontheValueoftheWorld’sEcosystem
S.C.Farberetal./EcologicalEconomics41(2002)375–392391
ServicesandNaturalCapital;TowardaDynamic,IntegratedApproachsupportedbytheNationalCenterforEcologicalAnalysisandSynthesis,aCenterfundedbyNSF(GrantcDEB-0072909),theUniversityofCalifornia,andtheSantaBar-baracampus.Additionalsupportwasalsopro-videdforthePostdoctoralAssociate,MatthewA.Wilson,intheGroup.
References
Alexander,A.,List,J.A.,Margolis,M.,d’Arge,R.C.,1998.
Amethodforvaluingglobalecosystemservices.Ecologi-calEconomics27(2),161–170.
Arrow,K.,Fisher,A.C.,1974.Environmentalpreservation,
uncertaintyandirreversibility.QuarterlyJournalofEco-nomics88,312–319.
Bishop,RichardC.,1978.Endangeredspeciesanduncer-tainty:theeconomicsofasafeminimumstandard.Amer-icanJournalofAgriculturalEconomics60(1),10–18.Blamey,R.K.,Rosemary,F.J.,1999.‘Citizens’Juries—An
AlternativeoranInputtoEnvironmentalCost-BenefitAnalysis’.ConferenceoftheAustralianandNewZealandSocietyforEcologicalEconomics,Brisbane,Australia,7July,GriffithUniversity.
Blaug,M.,1968.EconomicTheoryinRetrospect.Irwin,
Homewood,IL.
Brown,M.T.,Herendeen,R.A.,1996.EmbodiedEnergy
AnalysisandEnergyAnalysis:AComparativeView.Eco-logicalEconomics19,219–235.
Chichilnisky,G.,Heal,G.,1998.Economicreturnsfromthe
biosphere.Nature391(6668),629–630.
Ciriacy-Wantrup,S.V.,1963.ResourceConservation:Eco-nomicsandPolicies.DivisionofAgriculturalSciences,UniversityofCalifornia,UniversityofCaliforniaPress.Clark,C.W.,1973.Theeconomicsofoverexploitation.Sci-ence181,630–634.
Cleveland,C.J.,Costanza,R.,Hall,C.A.S.,Kaufmann,R.,
1984.EnergyandtheUSeconomy:abiophysicalperspec-tive.Science225,890–897.
Coote,A.,Jo,L.,1997.CitizenJuries:TheoryintoPractice.
InstituteforPublicPolicyResearch,London.
Costanza,R.,1980.Embodiedenergyandeconomicvalua-tion.Science210,1219–1224.
Costanza,R.,2000.Socialgoalsandthevaluationofecosys-temservices.Ecosystems3,4–10.
Costanza,R.,Herendeen,R.A.,1984.Embodiedenergyand
economicvalueintheUnitedStateseconomy:1963,1967and1972.ResourcesandEnergy6,129–163.
Daly,H.E.,1992.Allocation,distribution,andscale:towards
aneconomicsthatisefficient,just,andsustainable.EcologialEconomics6,185–193.
Daly,H.E.,Cobb,J.B.,1989.FortheCommonGood:Redi-rectingtheEconomyTowardCommunity,theEnviron-mentandaSustainableFuture.BeaconPress,Boston.DryzekJohn,S.,1987.RationalEcology:Environmentand
PoliticalEconomy.BasilBlackwell,NewYork.
Ehrlich,P.,Raven,P.,1964.Butterfliesandplants:astudyin
coevolution.Evolution8,586–608.
Ekins,P.,Max-Neef,M.(Eds.),1992.Real-LifeEconomics.
Routledge,London.
Fisher,Irving,1930.TheTheoryofInterest.Macmillan,New
York.
FishkinJames,S.,1991.DemocracyandDeliberation.Yale
UniversityPress,NewHaven.
Freeman,M.,1993.TheMeasurementofEnvironmentaland
ResourceValues:TheoryandMethods.ResourcesfortheFuture,Washington,DC.
Gilliland,M.W.,1975.Energyanalysisandpublicpolicy.
Science189,1051–1056.
Goulder,L.H.,Donald,K.,1997.Valuingecosystemservices:
philosophicalbasesandempiricalmethods.In:Daily,G.C.(Ed.),Nature’sServices:SocietalDependenceonNaturalEcosystems.IslandPress,Washington,DC,pp.23–48.
Habermas,J.,1984.TheTheoryofCommunicativeAction.
BeaconPress,BostonMA.
Hall,C.A.S.,Cleveland,C.J.,Kaufmann,K.,1992.Energy
andResourceQuality:TheEcologyoftheEconomicPro-cess.UniversityofColoradoPress,Colorado.
Michael,H.W.,1991.WillingnesstoPayandWillingnessto
Accept:HowMuchCanTheyDiffer.AmericanEco-nomicReview81(3),635–647.
Harrod,R.,1961.Bookreview.TheEconomicJournal71,
783–787.
Heuttner,D.A.,1976.Netenergyanalysis:aneconomicas-sessment.Science192,101–104.
Hicks,J.R.,1939.ValueandCapital:AnInquiryIntoSome
FundamentalPrinciplesofEconomicTheory.OxfordUniversityPress,London.
James,R.F.,Blamey,R.K.,1999.Citizenparticipation—
somerecentAustraliandevelopments.PacificScienceConference,Sydney,Australia,4July.
Jacobs,M.,1997.Environmentalvaluation,deliberative
democracyandpublicdecision-making.In:Foster,J.(Ed.),ValuingNature:Economics,EthicsandEnviron-ment.Rutledge,London,UK,pp.211–231.
Kaufmann,R.K.,1992.Abiophysicalanalysisoftheenergy/
realGDPratio:implicationsforsubstitutionandtechni-calchange.EcologicalEconomics6,35–56.
Kopp,R.J.,Smith,V.K.,1993.ValuingNaturalAssets:The
EconomicsofNaturalResourceDamageAssessment.Re-sourcesfortheFuture,Washington,DC.
Krutilla,John,V.,1967.Conservationreconsidered.Ameri-canEconomicReview57(4),777–786.
Lancaster,Kelvin,1971.ConsumerDemand:ANewAp-proach.ColumbiaUniversityPress,NewYork.
392S.C.Farberetal./EcologicalEconomics41(2002)375–392Leopold,Aldo,1949.ASandCountyAlmanac.OxfordUniver-sityPress,NewYork.
Loewenstein,G.,Prelec,D.,1991.Negativetimepreference.
AmericanEconomicReview81(2),347–352.
Low,B.S.,2000.WhySexMatters:ADarwinianLookatHuman
Behavior.PrincetonUniversityPress,Princeton,NJ.
Mitchell,R.C.,Carson,R.T.,1989.UsingSurveysforValue
PublicGoods:TheContingentValuationMethod.ResourcesfortheFuture,Washington,DC.
Norton,B.,Costanza,R.,Bishop,R.,1998.Theevolutionof
preferences:why‘sovereign’preferencesmaynotleadtosustainablepoliciesandwhattodoaboutit.EcologicalEconomics24,193–211.
Odum,H.T.,1971.Environment,PowerandSociety.Wiley,New
York.
Odum,H.T.,1983.SystemsEcology:AnIntroduction.Wiley,
NewYork.
Page,T.,1977.ConservationandEconomicEfficiency.Johns
HopkinsUniversityPress,Baltimore.
Reder,M.,1961.Bookreview.AmericanEconomicReview51
(4),688–725.
Sagoff,M.,1998.Aggregationanddeliberationinvaluing
environmentalpublicgoods:alookbeyondcontingentvaluation.EcologicalEconomics24,213–230.
Schumpeter,Joseph,A.,1978.HistoryofEconomicAnalysis.
OxfordUniversityPress,NewYork.
Slesser,M.,1973.Energyanalysisinpolicymaking.NewScientist
58,328–330.
Sraffa,P.,1960.ProductionofCommoditiesbyMeansof
Commodities:PreludetoaCritiqueofEconomicTheory.CambridgeUniversityPress,Cambridge.
Varian,Hal,R.,1992.MicroeconomicAnalysis.W.W.Norton,
NewYork.
因篇幅问题不能全部显示，请点此查看更多更全内容