WPC lUCB4 %v| 0 0< 0<F 0? 0? 0? 0<? 0{ 0<n 0< 0G 0D- 0Qq 0 0 0 D 0 i!%F&'l) D+,=, 0- 0C- AM-# .U<//U<k// 0Em0 AY0 1 8c B%2+2Dc3-45U<9dnZ9f:n:d8;a<<<U02?nYb? 0rB-Cw@D&EfFyFdI9GaK!TcK!dfKdwLQU<Qf1Rd$3RaWVfkVamV!ti!l!oV!pr!Px!dx!|!0!pԀ!D!pT!Ď!ԑ!PV" X! !`!0 !pP!p ),0 Z\ <School of the Environment 0 .   h:Default ParaDefault Paragraph FontXXXW%\  `*Times New RomanTTW                                W%\  `*Times New RomanTTW8w,toc 1toc 1   4                                5+ ` hp x (#58w,toc 2toc 2   4                                 5+ ` hp x (#58z,toc 3toc 3   4                                 5+ ` hp x (#58z,toc 4toc 4   4 <                                5+ ` hp x (#58z,toc 5toc 5   4<                                5+ ` hp x (#58w,toc 6toc 6   4                                5+ ` hp x (#58cX,toc 7toc 7                                  8w,toc 8toc 8   4                                5+ ` hp x (#58w,toc 9toc 9   4                                5+ ` hp x (#5@z0index 1index 1   4                                 5+ ` hp x (#5@w0index 2index 2   4                                 5+ ` hp x (#5Pt8toa headingtoa heading                                  5+ ` hp x (#5@0captioncaption  XXXW%\  `*Times New RomanTTW                                W%\  `*Times New RomanTTW^:_Equation Ca_Equation CaptionXXXW%\  `*Times New RomanTTW                                W%\  `*Times New RomanTTW($$  0  (2K($ 8 !  W XX  #   XX?+ 4 hDL!X?{g % (  1  )    K(%  XX   X3. 4 #XX{XXInpractice,onewouldalsoneedtoincludedifferentialdomestictaxationand C subsidypoliciesasinfluencestotheexpenditure,GNPandrentchanges.OuranalysisincorporatestheseeffectsaswellinthewelfaremeasuresdescribedfortheamendedHarrisonRutherfordWootonmodeldescribedbelow.Alltaxescollectedineachregionareredistributedbacktotheindividualhouseholdrepresentingallconsumersinthatregion.#XX{X#X# _ K(%  XX   X4. 4 #XX{XXThisconditionreliesonapartialequilibriumevaluationofasmallopeneconomy # andassumesaCES(constantelasticityofsubstitution)productionfunctionforthesingleindexofallmarketoutputintermsofasinglemarketedinputandpollutionservices.WiththeassumptionthattheFrischmoneyflexibilityofincomeisindependentoftheeffectsofpollutiononpreferences,LopezshowsthattherelationshipbetweenpollutionandincomewilldependontheinverseoftheelasticityofsubstitutioninproductionincomparisontotheFrischindex.#XX{X#X#  K(%  XX   X5. 4 #XX{XXOneexampleofthisroleforenvironmentalresourcesisfoundinevaluatingthe * netbenefitsofcommandandcontrolversusincentivebasedregulations.See + Oates,PortneyandMcGartland[1989].#XX{X#X# < K(%  XX   X6. 4 #XX{XXWiththenestedArmingtonassumptionimportersareassumedtominimizeimport  costssubjecttoanimportaggregationfunctionlinkinginthefirstnestdomesticconsumptionofthecommoditywithaCESaggregateoftheimportsofthesamecommodityfromallregions.#XX{X#X#  K(%  XX   X7. 4 #XX{XXSulfuroxideistheonlytransboundarypollutant.Itseffectsarelimitedtoone   morbidityrelationship(chestdiscomfort),notamortalityeffect.Thedispersioncoefficientfor30xfromUKtoGermanyisderivedastheproductoftheshareoftheUK's30xemissionsreceivedbyGermany(0.015)andtheS0xdispersioncoefficientforGermany(.01128ppm/ton).#XX{X#X#'dxd Level 1 Level 2 Level 3 Level 4 Level 5(e2$ 8 !  W ($     A<< cHP LaserJet 5M (PCL),'3 pet ^-(4 HP5MPCL.PRS,\,\H !L- 0&D `$Times NewRoman%\  `$Times NewRoman Level 1 Level 2 Level 3 Level 4 Level 5($ (    ) $%M << deUU BT(q,-,~OVERLINE{u})~=~e(~p(~q,~!,~OVERLINE{u}),~!,~OVERLINE{u})~-~g  ~(p(q,~!,~OVERLINE{u}),~!)~#-~(--1)~!*~m~-~(1-~w)(p-p*)SUP{T}q-~ d BT~=~e~(p~(q,~!,~ZSUB{d}SUP{!},ZSUB{f}SUP{!},~OVERLINE{u}),~!,  ~ZSUB{d}SUP{!},ZSUB{f}SUP{!},~OVERLINE{u})~-~e~(p~(q,~OVERLINE x !,~ZSUB{d}SUP{*},~ZSUB{f}SUP{*},~OVERLINE{u}),OVERLINE{!},~ZSUB @ {d}SUP{*},~ZSUB{f}SUP{*},~OVERLINE{u}) #   XX?+ 4 hDL!X?gg ##Xd#  K(%  XX   X2. 4 #XX{XXThisframeworkistheoneoftenusedindescribingenvironmentalcosting.Itis ) importanttoacknowledgethatthe"costs"addedbyprivateincrementalcostsrepresenttheaggregateincrementalwillingnesstopaytoavoidtheexternalitiesassociatedwitheachlevelofproduction.Thus,thisarisesfromtheaffectedpeople'spreferencesforpollutionabatement.ToderivethelinkbetweenS)SԀand  SSEwouldrequireunderstandinghoweachsectorrespondstocontrolasrequired  torealizethenetpollutionarisingfromeachlevelofoutput.SeeSmith[1992]foradiscussionofsomeoftheseassumptionsinthecontextofapplyingenvironmentalcostingtoagriculture.#XX{X#X#%\  `$Times NewRoman;;oo};3;oo};[S;(P;1;)A;(;1#;) ;];-+;!;-b;!;m ;m  d-~~~~~(1-w)~[(p-pSUP{*})SUP{T}q~-~(OVERLINE{p}-pSUP{*})SUP{T}q];f;y;oo>};;oo};(;1(;);[;(w;)4;(;) ;];w%;p;poo}T;qs;p;poo }TI ;qsTable_A 2 K(%  XX   X8. 4 CopelandandTaylor[1995b]haveformulatedatwocountrymodelwith  productionexternalitiesthatencouragespatialseparationofactivitiesandtradethatlendstoenvironmentaldegradation.ThelogicoftheanalysisiscomparabletoHelpmanandKrugman's[1985]analysisoftheeffectsofincreasingreturnsoftrade.* `CG Times%\  `*Times New RomanTTX\  P6G;P%\  `*Times New RomanTTXXj\  P6G;XP%\  `$Times NewRomanTTX\  P6G;P%\  `$Times NewRomanTTXXj\  P6G;XP%\  `$Times NewRomanTTG\  P6G;P(J $    K(%  XX  #XX{XX  X1. 4 Foragoodsummaryofthisdiscussionaswellascitationstoempiricalliterature n bearingonthistopicseeRepetto[1995].#XX{X#X# }E -~~~~~g~(~p(q,~!,~ZSUB{d}SUP{!},~ZSUB{f}SUP{!},~OVERLINE{u}),  ~ZSUB{d}SUP{!},~ZSUB{f},SUP{!},~!)~-~g~(~p(q,~OVERLINE{!},~Z x SUB{d}SUP{*},~ZSUB{f}SUP{*},~OVERLINE{u}),~ZSUB{d}SUP{*},~ZSUB{f}SUP{*},~OVERLINE{!}) o-~~~~~[~(--1)~!SUP{*}m~-~(~OVERLINE{-}-1~)!SUP{*}OVERLINE{m}~]BT$q|uepqJu6 u g p q uU;m;w ;p ;poom }T ;q(w, ,)?((,,),, )C ( (h ,J ,)T,6)o;(l;1;);(;1_ ;) ;(. ;)-h!T! !!;-G;! ;|J6  ] ;;;%;T;1 ; ; ;Table_As #   XX?+ 4 hDL!X?ddoo!oo!oo !oo !U doooooooofdBTdedpdqndZoo9dPdZoo9f{dug dZoo 9dI dZoo 9ft du de dpdqCdZoo9d{dZoo9fduVdZoo9ddZoo9fdud(d(d,d,'d, d,d)d,d, d, d, d) d(]d(d,d, d,Ad, d)Kd,d,d,Td,d)d!d!ad!td!{t at #   XX?+ 4 hDL!X?{g (2) &D `$Times NewRomandoo!oo!oo !ooi ! doozoooooo,dgdp~dq%dZoo9dPdZoo9f{dudZoo9d dZoo .f dg dpFdqdZoo]9d%dZoo9f]dugdZoo9ddZoo9fcd(?d(d,d,d, d,d)d,> d,@ d, d,w d)+ d(d(d,{d,d,d,d)d,-d,ed,Gd)Cd! d! d!d!{ ]Table_ATable_B)\  p`&Times New Roman,4  p.Times New RomanBoldS)*f9 p`2Times New RomanItalica) p`CG Times<6X9`+ CourierLegal[42 p`SymbolY p`Wingdings26X`#0Line DrawScalable-LGl,_ p CG TimesBoldT3y.D8D,,DI5+T47RCR,,XRqqg]S57WCW,,?XW_T6W!1(1,,h175RCR,,XRT88RCR,,XR ! {]S9?xxx,,XxS:1gCg,,;XgݸT ;9C,,hX <?dddXXXd3T=x/D8D,,DT>H)!),,,)T?V"1(1,,h1T@a$7,7,,7TA:!!,,!Br5ddd,,dSC~)V8V,,;VݸT Dw08,,h Er5SSSXXS3F7WCW,,XWs6RCR,,XR%.194 6,|  x a Jx6UnitedKingdom   19.85 3|C"u,^09CSS999S]+9+/SSSSSSSSSS//]]]Ixnnxg]xx9?xgxx]xn]gxxxxg9/9MS9ISISI9SS//S/SSSS9?/SSxSSIP!PZ9ZM999+999999S9SxIxIxIxIxIlnIgIgIgIgI9/9/9/9/xSxSxSxSxSxSxSxSxSxSxIxSxRxSxSxS]SxIxIxInInInInIxigIgIgIgIxSxSxSxSxSxSxSxS9/9/9/9/u]?xSg/gCg/g9g/xSxbxSxSxSxSxn9n9n9]?]?]?]?gFg/g/xSxSxSxSxSxSxxSgIgIgIxSg/xSn9]?g/xSxSxixSxSgI9//xS4+SS/uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuNA.SSxSSJJSJSSSSSS8SSSSSSSSSxJxJxJxJxJorJiJiJiJiJ8.8.8.8.{SxSxSxSxS{S{S{S{SxSxJ{SxSxSxS{S`SxIxJxIqIqIrJrJ{dgIiJiJgIxSxSxSxSxSxS{S{S8.8.8.8.z]AuSg/gZ?Z?Z?YAgYAi.xSxS{dxS{SiJ8..xS488SS88/NxxxSSS8JDDSSSSSS;SSSS;88VVS++SSddSSxSc]]8VS;xxSxWxxS唔S88xdxxxxxxxxxx8SxS]SxoS8SxJS`xlxxxxxxxxxxMxxxxxxodxGcxxxxxxxSxxxxxxxJxxxxJxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx8xxx8xxx8xxx8xxxxxxxxxxxxxxdi]d]oJiAlJ{SxJ8.uJo]]{JoSxJxd`SdSSiJxJodx]dddxdiA{S8..xJxJJddJdddd888SSSSSSSddddddddddddddddddddddAAAAAAASSSSSSSSSSSSSSSSSSSSSS............JJJJJJJJJJJJJJJJJJJdddddddddddddddddddddd`SJ8"u,^"(22TN"""28"2222222222888,\HBBH>8HH"&H>XHH8HB8>HH^HH>"".2",2,2,"222N2222"&22H22,006"6."""""""""2"2H,H,H,H,H,XAB,>,>,>,>,""""H2H2H2H2H2H2H2H2H2H2H,H2H1H2H2H282H,H,H,B,B,B,B,H?>,>,>,>,H2H2H2H2H2H2H2H2""""F8&H2>>(>>">H2H;H2H2H2H2XHB"B"B"8&8&8&8&>*>>H2H2H2H2H2H2^HH2>,>,>,H2>H2B"8&>H2H2H?H2H2>,"H2422/FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFN$<<$.2",2222`2 LL2 LL2L"",,2d""<22b$<<$88d",TdLL\dd2Ld<:628b<<BBBB186666668HHH8dd22d2bbbbG(:"8*02H"u,ae^$(8<><q*WWxxxWWkkxxx<xxxxkkxxxk<((xxxxxWHxkWWWWWWWWWW(x(-((((((((((---#J:55:2-:::2F::-:5-2::K::2%(#(#(#(((>((((((:((#&&++%((:#:#:#:#:#F45#2#2#2#2#:(:(:(:(:(:(:(:(:(:(:#:(:':(:(:(-(:#:#:#5#5#5#5#:22#2#2#2#:(:(:(:(:(:(:(:(8-:(22 222:(:/:(:(:(:(F:555----2"22:(:(:(:(:(:(K::(2#2#2#:(2:(5-2:(:(:2:(:(2#:(4((/888888888888888888888888888888888888888888888888N00%(#((((M(==(==(=##(P0((N00--P#CP==JPP(=P0.+(-N0055551-+++ +++-:::-PP((P(NNNN9 .-"&(:"/6=Dz/)@.%\  `*Times New RomanTT'  ; <School of the Environment 0 .   "h;]09V99V^+^+0VVVVVVVVVV00^^^L9R"R8OVq^^^z0^88^^^zzzzggE^/zU*FvggzzzzKggqU^^^zzzvvv^VUUUUUUUUgBBBBBBBBBBBB|lr^gFiUiKiU|gY98|^v^c|YoU|^^_^egKiKvcY|^vv|liK|g988|^vccvljgzllllllllllllllllllllllKKKKKKKgggggggggggggggggggggg888888888888^^^^^^^ccccccccccccvvvvvvvvvvvvvvvvvvvvvv"$[dپak{u{{{AMuu]daqqZZnn{{xu{{M{aZZ5M5M҅P?k" "u,ae^;C]ddCCCdCCCCddddddddddCCȲdxN`xoȐCCCddCdoYoYFdo8Co8odooYNCodddYdddddddddCddddddddodddddϐYYYYYN8N8N8N8oddddooooddpddddxodddXXYYXYYXddddddooL8N8N8N8p`o8P8N8ppodd┐XXYpLoNpLoNPDCopoopoȐdXYXo8oYoNCdddoYN88d4CCddCC/NdddCdUUddddddFddddFCCssd44ddxxddd~ooCsdFdsd岲dCCȐxȲCddodȐȅdCdYdsȐ`ȐȐȮxȐUvŐdȐddCCCCؐxoxoYNYYYN8YooYdYxxdxddYYxoxxxxNYN88YYYxxYxxxxCCCdddddddxxxxxxxxxxxxxxxxxxxxxxNNNNNNNYYYYYYYYYYYYYYYYYYYYYY888888888888YYYYYYYYYYYYYYYYYYYxxxxxxxxxxxxxxxxxxxxxxxdYC"u,ae^;C`ddCCCdCCCCddddddddddCCȰdxxxsCYoxxdoxxooCCCddCddYdY8dd88Y8ddddLL8dYYYLYdYddddddCdddddddddxdxdxdxdxdYxYxYxYxYC8C8C8C8dddddddddoYxddddoYdxdxdxdxdXXYYxxXxYxYxXddddddddD8C8C8C8dYYp8pHo8p8p8dxddddxLxLxLdLdLdLdLpHp8o8ddddddoYpLpLpLdo8dxLdLo8oYoYxddxYC88d4CCddCC/NdddCd]]ddddddFddddFCCddd88ddxxdddkddCddFddd尰dCCȐxȰCdxdodȐȅdCdYdsȐ]ȐȐȤxȐUvŐdȐYYCCCCxxxoxoYxLoYdYC8YooYdYxxdxddoYoYxxoxxxxxxLdC88YoYYxxYxxxxCCCdddddddxxxxxxxxxxxxxxxxxxxxxxLLLLLLLdddddddddddddddddddddd888888888888YYYYYYYYYYYYYYYYYYYxxxxxxxxxxxxxxxxxxxxxxsdYC@BLMNPQRSTUVWXYZ]^_`a)\  p`&Times New Roman,4  p.Times New RomanBoldS)*f9 p`2Times New RomanItalica) p`CG Times<6X9`+ CourierLegal[42 p`SymbolY p`Wingdings26X`#0Line DrawScalable-LGl,_ p CG TimesBold)4&_ p`$CG TimesItalicE 8 !  W 7d7-<ћ 4    <      D      L Copyright1995 4    <      D      L Revised8/22/95  <( 4 <DL!X<XX    EnvironmentalandTradePolicies:SomeMethodologicalLessons    V.KerrySmithandJ.AndrsEspinosa*  ?+ 4 <DL!4X?I. 4 Introduction h   4 Tradeandtheenvironmenthasbeenafrontpageissueformostofthenineties.1 ـBothpopularandprofessionalpublicationshaveofferedlessonsabouthowpoliciesshouldbemodifiedtoharmonizebothsetsofobjectives. 9 $  1      ׀Thisinteresthas,inturn, ` stimulatedaconsiderablebodyofnewconceptualresearch(seeKrutilla[1991],Chichilnisky[1994],Copeland[1994],CopelandandTaylor[1994],[1995a],andLopez[1994]asexamples).Wehavelearned,asaresultoftheresearchtodate,thatdomesticenvironmentalregulationswilleitherincreaseortheywilldecreasethe  internationalcompetitivenessoftheproductstradedbythecountryundertakingthem.Bycontrast,weknowthatremovingtradebarrierswilleitherimproveoritwillreduce P!  thelevelofwellbeingofthetypicalhouseholdinthecountryreducingtradebarriers.Finally,whiletheverdicthereisnotasclearcutasinthesefirsttwoareas,itseemslikelyresearchcurrentlyunderwaywillsoonestablishthateffortstolinktradeandenvironmentalpolicieswilleitherenhanceordegradethecurrentlevelsofwellbeing (#( experiencedindevelopedeconomiesfromthelevelsreachedusingthecurrentindependentapproachestotheseproblems. D,',  4 Whilethissummaryispresented,inpart,to"amuse"thereader,itisnonethelessareasonablyaccuratedescriptionofwherewestand.Moreover,itisnotanindictmentoftheresearchtodate.Rather,itreflectswhatislearnedfromconceptualresearchinmostareas thedevilisinthedetails.Thatis,thespecific & v circumstancesofeachcountry seconomyinrelationtotheworldmarketsrelevantfor  L itsproductsmustbeconsideredbeforeclearcutconclusionscanbeoffered.Thus,wecanagreethat...8 4   4 Theinternationalmarkettransmitsandenlargestheexternalitiesoftheglobalcommons.Nopolicythatignoresthisconnectioncanwork.(Chichilnisky[1994],p.865). 44 8 4   44 Nonetheless,thisagreementdoesnotpreventpolicymakersfromrejectingmostproposalsforgreaterlinkageinthesepolicies.Moreover,theircautionmaywellbewarrantedbasedontheavailableempiricalevidence.AsRepetto[1995]observed,thereisnosolidempiricalevidencethatincreasingthestringencyofenvironmentalregulationsnecessarilyhasdetrimentaleffectsontradeandinvestment.Ofcourse,thisdoesnotmeanthatcoordinationshouldberuledout.Rather,itheightenstheimportanceofunderstandingthefeaturesofeachcase,beforeattemptingtooffergeneralizationsaboutwhentherewillbetheadvantagestolinkingtradeandenvironmentalpolicies. 4 Researchonthesequestionsisalsolikelytohavepayoffsinotherareas.Indeed,wewillarguethatimportantmethodologicalissueshaveemergedfromtheintermediateresultsoftheresearchtodateontradeandtheenvironment.Inwhatfollows,wesummarizetheseinsights,andillustratetheirimportancewithanewCGE *>&* -<7XXdd7  -<7XXdd7  modelthatincorporateslocalandtransboundaryexternalities.UsingKrutilla[1991]andAnderson s[1992]adaptationoftheconventionalappliedwelfareframework,the  firstsectionillustrateshowenvironmentalandtradepoliciesbecomeintertwined.Thisgraphicalanalysisalsoprovidesaplatformforourdescriptionofhowassumptionsofrecenttheoreticalandcomputationalmodelsinfluencethesestudies'conclusionsaboutenvironmentalandtradepolicy.ThethirdsectiondescribestheEspinosaSmith[1995]extensiontotheHarrisonRutherfordWooton[1989]modelofmostoftheeconomiescurrentlycomprisingtheEuropeanUniontoincludeenvironmentalexternalitiesandhowthatframeworkcanbeusedtogaugetheimportanceofseveralofthe simplifyingdetailsofearlierresearch.Whilemostoftheempiricaldetailinthismodelconsidersdevelopedeconomies,thegenericlessonsderivedfromexaminingtheimplicationsofincorporatingenvironmentalresourceswithinpreferencesareequallyrelevanttothedevelopingcontext.Theserelationshipsarediscussedintheclosingsectionofthepaperaswellasthroughoutourdescriptionofthefindings.ThelastsectionalsousestheresultsfromouranalysistosuggestacorollarytoKrugman's[1991]argumentsabouttheimportanceofgeographyforinternationaleconomicinteractions.II. 4 ASimplifiedStoryofInterdependence t% $ Л 4 UsingtheKrutillaandAndersondemandandsupplyformat,itispossibletodevelopastraightforwardcategorizationofthecurrentliterature sdescriptionoftrade *:&* andtheenvironment.Considerfirst,thecaseofasmall,openeconomyexportingaproducttotheworldmarket.ThisisrepresentedinsimpletermsinFigure1.D)DԀand  S)SԀdescribethedomesticdemandandsupplyandOPwtheworldprice. Small T impliesthiscountrysactionsalonewillnotaffecttheworldmarket.Underautarky,themarketequilibriumwouldbeatthepointXwithdomesticprice,OPx,and  H quantityOQx.IfSSEdefinesthesupplyfunctionreflectingthefullsocialcostof    producingQ,(i.e.,includingbothincrementalprivateandanyincrementalsocialcostsarisingfromexternalitiesfromtheproductionofQ), 1 . $  2      ׀thenthewelfarelossassociated n  withfailingtoadoptefficientenvironmentpoliciesisgivenbyXEH.SHXistheaddedsocialcostsduetoproducingQx.Mostoftheselossescountagainstthe b economicsurpluswewouldhaveattributedtoproducingOQx(i.e.,thetriangleSXDis 4 economicsurplusandSXEtheamountsubtractedduetoaddedenvironmentalcosts). 4 Ifthiseconomyisopenedtotheworldmarket,producersnowfacetheworldprice,OPwandwouldexport(i.e.,OPw>OPx).TheynowproduceQw,sellingOQA Z onthedomesticmarketandQAQwontheworldmarket.Socialcostsdueto ,| environmentalexternalitiesincreasebyHXCG.Someofthisincreasedisplacestheincreasedrent(AXC)earnedbyproducersandthenetoutcomedependsontherelativesizeofABE(theremainingrent)andBGC(theincrementalenvironmentalcostthatdoesnotdisplacetherentchange).Anyexportrestrictionsintendedtoreduceincentivestoparticipateintheworldmarketthatfailtoreflectthepresenceofenvironmentalcostsmayalterthejudgmentonrentlossesversusenvironmentalcosts,dependingontheincrementalprivateandsocialcosts. *:&*  4 Thereareanumberofissuesthatcannotbeaddressedinthistypeofasimple,onemarketevaluation.Forexample,thisdiagramdoesnottakeaccountoftheeffectofeachpolicyonthepositionoftheprivatesupply,socialcost,andprivatedemandfunctions.Thetreatmentofthesegeneralequilibriumeffectsiswhatdistinguishesmostofthenewconceptualliteraturefromthissimpleapproach.Therearealsothe"details"aboutthefactorsunderlyingthesefunctionsthathaveimplicationsrelevanttocomplexdomesticenvironmentalpolicies.Asaresult,wefirstconsiderthefactorsthatcanbeimportanttotheresultswithsmallopenmodelsandthenturntotheeffectsofgeneralequilibriumfeedbacks. 4  4 A.   ExtensionstoKrutillaAndersonFramework   4  4 Copeland[1994]andLopez[1994]considerthepotentialformultipleeffectsinfluencingeitherwelfareorthepollution/incomerelationshipinasingleeconomy.Generalequilibriumadjustmentsarenotconsidered.Copelandproposesasmall,openeconomywheretheprimaryfocusofhisevaluationisageneralizationoftheanalysistoincludeseveralmarketsandprograms.Thus,heusesaframeworkthatprovidesaproperaccountingoftheprice,policyinducedrent,andamenitychangesrequiredtoproperlymeasureHicksiangainsandlossesforexogenouseffects.Therearenofeedbackeffectsofdomesticpolicythrougheitherinternationalmarketsortheglobalcommonsbecausethedomesticeconomyissmallrelativetotheworldmarket,andpollutionislocalized. 4 Whileitisnotdescribedintheseterms,Copeland'sevaluationofpoliciesuses *:&* aHicksiancompensatingvariationmeasure.HebaseshisevaluationonmarginalchangesintheAndersonNeary[1992]balanceoftradefunctionwithchangesineitherenvironmental(representedaseffluentcharges)ortrade(importtariffandexportsubsidies)policies.ThisfunctionreducestotheHicksiancompensatingvariationwithadjustmentsforchangesinincome(orrents)duetoexistingpolicies.  $  3        H  4 Intermsofourdiagramforthecaseofasmallopeneconomy,Copeland'smodelgeneralizestheanalysisinthreeways.Itallowsformultiplecommodities,(consistentlyaccountingfortheeffectsoftradeandenvironmentalpolicyforeach);expressesthewelfaremeasuresinHicksianterms;and,perhapsmostimportantlyfromtheperspectiveofourlaterdiscussionofgeneralequilibriumeffectsoftradeandenvironmentalpolicy,allowsforindirectconsumereffectsbetweenpollutionandmarketedcommodities.Thus,hisframeworkallowsamarginalchangeinpollutiontoshiftthedemandfunctionforthemarketedgood.Intermsofourdiagram,thisapproachrecognizesthattheintroductionofpoliciestakingaccountoftheexternalitiesinherentintheSSEfunctioncould,inprinciple,leadtoshiftsinD)D.Thiseffectis ,| ignoredinvirtuallyalltheotheranalysesoftradeandenvironment. 4 Lopezalsoconsidersasmallopeneconomyandprovidesamoredetaileddescriptionofdomesticpollutiongeneration,astheequivalentofusinganenvironmentalinputtoproducemarketedoutputalongwithlaborandcapital.However,theenvironmentalinputisassumedtobeweaklyseparablefromconventionalinputs.Hisfocusisonhowgrowthinincome,whetherduetotechnologicalchangeortradeliberalization,willinfluenceenvironmentalquality. *:&* Intheabsenceofstockeffectsontheenvironment,inhisschemeincomegrowth(andgreaterconventionalfactorusage)willinitiallyincreasepollutionandthencauseittodeclineasincomegrows.Thisresultallowsforarolefortheenvironmentinpreferences,butrequiresthatFrisch'sindexoftheimportanceofincomeeffectsbeseparablefrompollution.  $  4      ׀Thislaterpointisimportantbecauseincomechangesdonot  H influencetherateofsubstitutionbetweenmarketedgoodsandamenities.Hismodeldoesallowincometoincreasewillingnesstopayforpollutionreduction(orequivalentlyforamenities).IntermsofFigure1,theprimaryfocusofhismodelisinthedetailsunderlyingS)SԀandtheconnectiontoSSE. @  4 Theserestrictionsarealsorelevanttousingthemodeltounderstandtheroleofcoordinationbetweentradeandenvironmentalpoliciesinthedevelopingaswellasthedevelopedcontexts.Forexample,animportantbyproductofLopez'smodelisthedemonstrationthatincomeeffectscanleaddifferencesinthelevelsofenvironmentalqualityselectedatdifferentincomelevels.Developedanddevelopingcountriescanappearasdifferentpointsonaparabolicrelationshipbetweenpollutionandincome,withpollutioninitiallyincreasingwithincomeandthendeclining(comparabletotheϜGrossmanandKrueger[1993]empiricallinkbetweenpollutionandGDP).IntheLopezanalysis,thisresultfollowsfromacomparisonoftheimportanceofsubstitutionintwocontextsasthelevelofincomegrows.Thefirstoftheseinvolvestheenvironment'simportancetopeople'spreferencesandthesecondconsidersitsroleinproductionatdifferentlevelsofoutput. 4 Theserelationshipsbecomemorecomplexiftheenvironmentinfluences *:&* people'sdemandsformarketedgoodsand,simultaneously,affectssomemarketedinputsdifferentlyfromothers.EarlierresearchbyLopezwithNikletschek[1991],forexample,illustrateshowthepositivecontributionsofenvironmentalresources(e.g.,naturalbiomass)toatraditionalsectorcanhavethesetypesofeffectsonproduction.Theyfoundthatdirectproductiontaxestotakeaccountoftrade'sdetrimentaleffectsontheenvironmentwouldbenecessarytoenhancewelfarewithtradeliberalization. 4 WiththeKrutillaAndersonframework,theinclusionofsocialcostsrequiresconsideringtheaddedincrementalcostsduetoexternalitiessothattheyareproportionaltoincreasesinmarketedoutputs.Lopez'smodelfocusesonthesimplificationsunderlyingtheKrutillaAndersontypeofanalysisandhowtheyinfluencewhatislikelytobeobserved(acrosscountries)fromaprocessthatviewspollutionlevelsasoutcomesofthetradeoffbetweenmarketedgoodsandamenitiesashouseholdsexperiencegrowthinrealincome.B. 4 IncorporatingGeneralEquilibriumFeedbacks ,|  4 Theremainingpapersinthisareaconsiderdifferentapproachesfortreatingthegeneralequilibriuminteractionsbetweenmarketandnonmarketoutcomes.Chichilnisky'sanalysisusesaconventionalgeneralequilibriumframeworkwithtwocountries,(designatedtheNorthandSouth).TheNorthisintendedtorepresentadevelopedeconomywithawellestablishedsystemofpropertyrightstoallresources,whiletheSouthdesignatesadevelopingeconomywithadifferentsetofconditionsdescribingtheavailabilityoffactorinputs.Hermodelreliesonthedifferencesin *:&* accessconditionsbetweenprivateandopenaccessresources(asfactorinputs)toalterthesupplyfunctionsfortradeablegoods.Becauseendowmentsarenotfixedandrespondtoprices,thesupplyofmarketedoutputsproducedusingtheenvironmentalresource'sserviceswillbegreaterunderopenaccessconditions.Themodelallowsforthegeneralequilibriuminteractionbetweenproductandfactormarketconditionsinbotheconomies.Establishingprivatepropertyrightstotheenvironmentalresourceofferstheonlyfeasiblepolicyresponsetotheoverexploitationinducedbytheaccessconditionstotheenvironmentalresources.Tradesimplyincreasestheproblem,evenwhenpoliciesdesignedtoraiseotherfactorcostsareconsidered. 4 Intermsoffigure1thisframeworkisananalysisofhowaccessconditionstoenvironmentalresourcescaninfluenceS)S.Therearenopublicgoodsorexternalities, 4 asidefromtheinefficienciesduetoopenaccess.WhileChichilnisky'sexamplesintroduceamenities,themodeldoesnot.Indeed,privatepropertyrightssucceedbecausetherearenopreferencerelatedeffectsservedbyprotectingtheenvironmental Z resource.Thus,SSEdoesnotexistinthisframework. 0  4 CopelandandTaylor[1994]adoptedasimilartwocountry(North/South)settingbutintroduceexternalities,localtoeachcountry'sdecisions.Thus,thismodeladdsgeneralequilibriumeffectstothesocialcostissuesidentifiedinfigure1withSSE.Itallowsfor:interactionsbetweenthetwocountries,endogenous(toeach x% $ country)pollutiontaxes,andacontinuumofgoods,differingbothintheirpropensitytogeneratepollutionandimportancetoconsumers.ThemodelassumeCobbDouglastechnologyandCobbDouglaspreferencesinmarketedgoods.Domesticpollutionis *>&* treatedasstronglyseparablefrommarketedgoods. 4 AswithLopez,incomeeffectsarethekeysourceofdifferencesinenvironmentalpolicyand,inturn,leadtothemotivationfortrade.IncreasedpollutionintheSouthprovidesthewayitcanincreasethegainstoberealizedfromtrade.LowerincomesintheSouthassurethatthistradeispossiblebecausepollutioncontrolislesshighlyvalued.IfthedifferencesintheSouth'sabilitytoabsorbpollutioncorrespondtoincomedifferences(e.g.,thelessdevelopedcountryhasbothlowerincomelevelsandlowerexistingpollution)thesedifferencesininitialconditionswouldsimplyreinforcewhatthemodeldisplays.TheChichilniskyandtheCopelandTaylormodelseitherignoreorassumeseparabilitybetweenmarketedgoodsandtheenvironment.Asaresult,theyfailtotakeaccountofthesamesubstitutioneffects(i.e.,arisingthroughpreferences)thatwerenotedinourdiscussionofLopez.Pollutionitselfmayalterthecompositionofgoodsdemanded. 4 TheCopelandTaylorframeworkshiftsthefocusofattentionfromthegoods'marketsandtheeffectsoftradeonthem,asillustratedinfigure1fortheonecommoditycase,tothe"market"forpollution.Intheirfirstpaper,thegeneralequilibriumeffectsoftradeareseenthroughthedomestic"market"forpollution.Tradeinfluencesboththesupplyofpollutionandexertseffectsonincometoshifttheincrementalvalueofpollutionreduction.Inasubsequentpaper(CopelandandTaylor[1995a])theyallowfortransboundarypollutionandmultiplecountries.Separability(ofpreferencesinaggregatepollution)impliesthatthepollutioninducedeffectsofgreaterproductioninonecountrynowarenotlimitedtothatcountry.Theyactasan *:&* offsettotheincreasedincomeandlowercostsformarketedgoodsfromtrade.Allpollutioninanycountryisapurepublicgood(bad).Buttheseeffectsarisewithequalforcetoallcountries.Astheauthorsnote,theresultsparallelwhatwefindintheliteratureoncharitablecontributions,(seeBergstrom,Blume,andVarian[1986]),policychangesthatexpandaggregateoutputdonotreduceworldwidepollution. 4 BarbierandRaucher[1984]consideraframeworkthatrelatesmorespecificallytotheproblemsofdevelopingeconomiesbyevaluatingtheeffectsoftradeonincentivesfortropicaldeforestation.Usinganoptimalcontrolmodeltheyallowforstockexternalitiestoinfluencetradepolicyforbothsmallandlargeeconomies.AswithChichilniskyandCopelandandTaylor,themodelassumesseparablepreferences,sotherearenopreferencerelatedsubstitutioneffectsinfluencinghowtheexternalityineachmodelinfluencesoptimaltradepolicy.̛ 4 C. ComputableGeneralEquilibriumApproaches Z  4 Aframeworkwithpurepublicgoodshasbeentheprimaryspecificationusedforthosecomputablegeneralequilibrium(CGE)modelsthatintroducednonmarketenvironmentalresources.(seeBallardandMedema[1993],andPiggott,Whalley,andϜWigle[1992]).Environmentalresourcesareallowedtoinfluencethelevelofhouseholdwellbeingbutnotthepreferencerelatedsubstitutionsbetweengoodsduetochangesintheseresources.Equallyimportant,thesemodelsdonothaveanexplicitspatialdimension.Theirstructuretreatstheenvironmentalresourceasauniformpublicgood.ThisspecificationiscomparabletotheformusedinCopelandand *:&* Taylor'smultiplecountrymodelwithglobalpollution. 4 Toourknowledge,theonlymodelthathasattemptedtodevelopamoreexplicittreatmentoflocalandtransboundarypollutionisPerroniandWigle[forthcoming].Theirframeworkallowsforbothlocalandglobaleffects.Itincorporatesseparateregionalabatementdecisions,butassumesthesametechnology.Aconstantelasticityoftransformation(CET)functionisusedtoaggregatenetlocalandnetglobalpollutionabatementactivities.Withineachregion,localenvironmentalqualityisdeterminedbythelevelofoutputofthesectorsgeneratingemissions,decisionsaboutabatementactivities,aregionspecificconstantelasticitydamagefunctionlinkingaggregatenet(i.e.,afterabatement)emissionstoquality,andtheinitialstockofqualityassumedpresentintheregion.Controloverthemixofglobalversuslocalenvironmentalqualityisrealizedthroughtheabatementfunction.Theemissionsgeneration,damagefunctions,andenvironmentqualityendowmentsdifferacrossregions.Globalenvironmentalqualityisdescribedinasimilarway,buttheparametersofdamagefunctionsareassumedtobethesameacrossregions.Itisdifficulttoevaluatehowthemodel'simpliedresponseinthefinalmixoflocalandglobaldamagestopolicyaccordswiththeassumptionsgenerallyusedindomesticpolicyevaluations. 4 PreferencesarespecifiedasanestedCESfunctionwiththepollutioncompositeandgoodsaggregateinthetopnest.Thisimpliessubstitutionbetweengoodswillnotbeinfluencedbythelevelofpollution(localorglobal).Nonetheless,thestructureisaninterestingsteptowardmorerealism. *:&*  4  4 D. Overview   4 Ourreviewoftherecentresearchontradeandtheenvironmentadoptedthemodelingperspectiveusedinenvironmentaleconomicstoevaluatewhat"details"shouldbejudgedasimportantinevaluatingtheworktodate.First,themodels,boththeoreticalandcomputational,haveconsistentlyassumedseparablepreferences.Thisapproachseparatestheeffectsoftheenvironmentonwellbeingfrompeople'sdecisionsmadeaboutdifferentmixesofprivategoods.Aswewilldiscussinmoredetailbelow,therearenopreferencerelatedfeedbackeffectsofexternalitiestransmittedoutsidemarketsontheoutcomesobservedinmarkets.Thisapproachisinconsistentwithallrevealedpreferencemethodsfornonmarketvaluation.Itistheabsenceofseparability,andindeed,specificassumptionslinkingprivategoodstopublicgoodsthathavebeenthemostimportantaspectofthepreferredmethodstoestimatetheeconomicvaluesofnonmarketedcommodities. 4 Second,inactualapplications,wegenerallyfindthatenvironmentalexternalitiesarisethroughsometypeofspatialinteraction.Thatis,activitiesinonelocationhaveeffectsonotheractivitiesinthatlocationaswellasonothersoutsidetheimmediatearea.Virtuallyallpolicyresponsessuggestthatthedetailsofboththeactivityandthetransmissionmechanismmatter.  $  5      ׀Recently,CopelandandTaylor t% $ [1995b]havedemonstratedhowspatialdifferencesintheeffectsofexternalitiesonproductioncangiverisetononconvexitiesakintothoseduetoincreasingreturns(seeHelpmanandKrugman[1985]).FromanappliedCGEperspective,onlyPerroniand *:&* Wigle[forthcoming]acknowledgedthistypeofconnection.However,aswenoted,theirspecificationoftheseconnectionsmayoversimplifymattersinimportantways. 4 Finally,attemptstolinktradeandenvironmentalpoliciesmustbeevaluatedwithinaframeworkthatacknowledgesthecomplexityoftraderestrictionspresentinmostcountries.Equallyimportant,theymustallowwelfarejudgmentstotakeaccountofthegeneralequilibriumeffectsthatchangebothrelativepricesandincomes(rents)thatcanarisefromexistingpolicies.AsChichilnisky(andothersbeforeher)acknowledge,thesetypesofincreasesincomplexityaredifficulttoincorporateinameaningfulwayintheoreticalmodels.WhiletheycanbeaccommodatedinCGEanalyses,theymustbeintroducedsothatthespecificationmatchessomeexistingbaseline. 4 Thereasonforthisapproachisdirect.Itisrelativelyeasytoacknowledgethatcomplexityinthenatureofregulationsandpolicyinducedpricewedgeswillmatter.Becauseeachspecificformwillchangeoutcomes,itmakeslittlesensetocatalogallpossibledistortions,manyofwhichdon'texist.Instead,themostproductiverouteformodeldesignwouldcallforselectingaspecific"realistic"descriptionofapolicyregimeandcalibratetheanalysistodealwithit.Clearly,thespecificcalibrationinfluencesboththeparametersandtheconclusionsdrawnfromthatmodel.Nonetheless,byselectinganexistingmodelandincorporatingenvironmentaleffectswithinit,thereisareadybenchmarkforcomparison.Forthisreason,weselectedarespecifiedversionofanestablishedmodelforthemajoreconomiesintheEuropeanUnion(EU).Themodelwasrecalibratedtoincludelocalandtransboundary *:&* pollutantsaswellasthehealtheffectstheycauseusingdamagefunctionsbasedonthosecurrentlyusedinregulatingthesepollutantsintheU.S.̜III. 4 DevelopingaNonMarketCGEModel    Л 4 InthissectionwedescribetheHarrison,Rutherford,Wootonmodel(HRW1)andtheadaptationstoittoincorporatethreeairpollutantssulphuroxides(SOx),asa @ transboundaryexternality;andparticulatematter(PM)andnitrogenoxides(NOx)as b sourcesoflocalexternalities.Themodellinksthepollutantstoproductionintherelevantsectorsofeachcountry;incorporatesasimplediffusionsystemforthetransboundarypollutant;andintroducesthemorbidityrelatedhealtheffectsfromairpollutionasnonseparableeffectsonpreferences.Mortalityimpactsaretreatedasseparableeffectsonindividualwellbeing.Theamendmentsweremadetoincorporatecurrentestimatesofhealthdamagesbothmorbidityandmortalityfromtheseairpollutants.Thedesignalsopermitsanevaluationoftheimplicationsof:(a)omittingtheenvironmentfromevaluationsofchangesintradepolicy;(b)ignoringthejointeffectsoflocalandtransboundarypollutants;and(c)restrictingpreferencestobeseparableinenvironmentalresources. 4 Afterdescribingthestructureofthemodelandouramendments,weoutlinethewelfaremeasureusedtoevaluateascenarioinvolvingbothtradeliberalizationwith *:&* environmentaldegradation.Ouranalysisconsidersthreedifferentwaysthatenvironmentaldamagesmighthavebeenintroducedintothemodel.Thefirstofthesedirectlycalibratesthepreferenceandproductionstructuretomeetbaseyearconditions,includinghealtheffectsastranslatingparametersonaStoneGearyspecificationofpreferences.Thesecondtreatstheanalysisofenvironmentaldamagesasanexogenous  H valuationtaskwithcomputationaftertheCGEmodelissolvedforamarketequilibriuminmarketedgoods.ThisfollowsthestrategyrecentlyproposedbyBoyd,Krutilla,andViscusi[1995].Thelastconsidershealthdamagesimpactsonthesourceforhouseholdincomethatendogenouslyreduceeffectivelaborendowmentavailabletohouseholds.  A. 4 TheRevisedHarrison,Rutherford,WootonModelHRW1    4  HRW1wascalibratedtorepresentconditionsintheEUin1985.ItincludeselevenregionswitheightoftheEUcountriesGermany,France,Italy,theNetherlands,Belgium,UnitedKingdom,Denmark,andIreland,aswellastheUnitedStates,Japanandanaggregateregionthatincludesallothercountries.Themodeldistinguishessixaggregatecommodities(agricultureandfood;miningandenergy;nondurablemanufacturing;durablemanufacturing;constructionandtransportation;andservices)andthreefactorinputs(land,labor,andcapital).Factorsarenottradedoutsidethehomeregion.ConsumptiongoodstradeusinganestedArmingtonstructure.  $  6       )l$(  4 Productionischaracterizedusingconstantreturnstoscale,nested,production *>&* functions.Thesefunctionsassumeaconstantelasticityofsubstitutionintheprimaryfactors(landentersonlytheagriculturalsector)andalineartechnologyinintermediateinputs.Eachregion's(country)demandisrepresentedbyasingleconsumerwithaCobbDouglasutilityfunctioninthesixconsumptiongoods.Thebudgetconstraintisspecifiedbytheaggregateconsumer'sexogenousendowmentofprimaryfactors,netrevenuesfromtaxesandlevelofforeigninvestment(assumedtoequalalumpsumtransferinthemodel).Themodelincludesadetailedrepresentationofdomestictaxes,tradebarriers,governmentandEUprograms.Forexample,allregions(excepttheU.S.)havevalueaddedtaxesontheprimaryfactorsrepresentedasafixedadvaloremtaxrate.ProducersintheEUcountriesalsoreceivesubsidiesonproductionaimedatfosteringexports.ThenetEUtransfersareequaltothevalueofreceiptsfrominterventionpurchasesandvariableagriculturalimportlevies(duetotheCommonAgriculturalPolicy)lessthecostofexportsubsidiesandthenetcontributionstotheEU. 4 OurmodelusestheHRW1structureinexactlytheformdevelopedandreportedinvariouspublications(i.e.,Harrisonet.al.[1989,1991]).Wehavemadethreemodificationstothemodel:(a)replacedtheCobbDouglaswithStoneGearyutilityfunctionsfortheaggregateconsumerineachregion;(theCobbDouglasspecificationtakestheformU=#XX{XX#{X#X##XX#|#ixi#XX{X#X{##XX{XXԀ# {X#Xi##XX?#|##XX{X#X|##XX{XXԀ#XX{X#X}#,whileaStoneGearydisplaceseach t% $ commoditybyatranslationorsubsistenceparameter,i,soU={XX##XX#i󛀀(xi󀄜i)# {X#Xi#XX?#);(b) F'"& introducednineairpollutioninducedmorbidityeffectsastranslatingeffectsoneachhousehold'ssubsistenceparameterforservices(thisistheaggregatethatHRW1 *:&* designatesforhealthexpenditures);and(c)introducedanexplicitsetofpollutiongenerationanddispersionmodelsforeachofthethreeairpollutantsidentifiedearlier.Asaresult,airpollutionwillaffectthemarginalratesofsubstitutionforconsumptiongoodsrelevanttotheseservices.Ourspecificationforairpollutioneffectsalsoallowsformortalityeffectsasaseparableshiftinpreferences.Themagnitudeofthiseffectisalsodirectlylinkedtoexistingestimatesofmortalityairpollutiondoseresponsemodels.Finally,monetarylossmeasuresuseestimatesofthevalueofstatisticallives.Eachofthesemodificationshasbeenintroducedsothatthemodel'sinitialcalibrationismaintained. 4 TherearefourimportantfeaturesofouradaptationtoHRW1toincorporateairpollution:(a)developingbenchmarkairpollutionmeasuresforeachsectorineachcountry;(b)specifyingtheairdiffusionsystemfortransfrontierpollutants(e.g.SOx);  (c)linkingtheambientconcentrationsofpollutantstomorbidityandmortalityeffects;and(d)monetizingthesehealthimpacts.Wewillsummarizethehighlightsassociatedwitheachstep. 4 EmissionsofPM,NOx,andSOxarebasedontheOECDEnvironmentalData N Compendium1989.Eachregionwasassignedthepopulationweightedestimateofthe !$  averageannualconcentrationforurbanandrurallocations.BecauseSOxisassumedto #" beatransfrontierpollutant,weused1991estimatesoftherelationshipbetweenemissionsanddepositionstoeachcountryfromeachsource(seeMarkandyaandRhodes[1992]).ThefirstthreecolumnsofTable1reportthebenchmarkconcentrationsusedforallthreepollutants.Thefourthcolumnreportstheshareof *B&* МSOxdepositionduetodomesticsources.Forlocalpollutants(PMandNOx),the  depositionisassumedtoequalemissionsandambientconcentrationcoefficientsareconstantmultiples,convertingdepositionunitstotheappropriateambientmeasures. 4 TheallocationofemissionstospecificproductionsectorswasbasedontheU.S.emissionsbyproductionsector.Theseemissionperunitofoutputrateswereappliedtoeachsector'soutputineachregionandtheresultingestimatesofaggregateemissionswererenormalizedbycountry,usingtheratioofactualtopredictedaggregateemissions. 4 ThethirdcomponentoftheprocessofincorporatingenvironmentalresourcesintotheHRW1involvescomputationofthehealtheffects.Theninemorbidityeffectsaregivenbypollutantasfollows:PM(bronchitis,chroniccough,croup,emphysema,upperrespiratorydisorders,andcoughepisodes),NOx(lowerrespiratorydisorders,eye  irritation),andSOx(chestdiscomfort).Thedoseresponsefunctionswerebasedonthe  МDesvousgeset.al.[1994,Vol1]studyofmorbidityeffects.Implementingthesemodelsrequiresthecorrectpollutionmeasure(andweusedconventionalconversionratiosintheseadaptations),alongwithabasepopulationtoapplytherevisedincidencerate,relevantforeachhealtheffect.Inmanycasesthesepopulationswerespecificdemographicgroups(ofteneitherolderpopulationsoryoungchildren).Ingeneral,itwasnotpossibletoidentifypopulationsatthislevelofdetail.Weusedtheoverallpopulationofeachregiontoestimateabaseincidencelevel.ForNOx,itwaspossible F'"& toadjustforthefractionofthepopulationunder15yearsofageindevelopingthehealthimpact. *:&*  4 Monetizationofthehealtheffectsrequiresestimatesfortheincrementalwillingnesstopayforeachhealthevent.TheseweretakenfromDesvousgeset.al.[1994,Vol.3].Becauseallofthevaluationestimateswerein1993U.S.dollars,theywererescaledto1980(thereferenceyearusedbyHRWtodeflatepricesintheirbenchmarkdataset)andconvertedtoeachcountry'scurrency.Mortalityeffectswereassumedduetoparticulatematterandused$3.5milliondollarestimateasthevalueofastatisticallife(in1989dollars). 4 Changesinpollutionareassumedtochangethelevelofthesubsistenceparameterfortheserviceaggregate(PollakandWales[1981]).ThistranslationwascalibratedtomeetthebenchmarksolutionforHRW1givenourestimatesoftheambientconcentrationsofeachpollutantineachcountryandthedoseresponsefunctions.ThesubsistenceparametersfortheStoneGearyspecificationineachcountryweredevelopedusingGamaletsos[1971]andLluchandPowell[1975].Basedontheseestimates,adjustedtoreflectrealGDPinthebenchmarkyearforHRWI,totaldemandforeachproductwasreducedbythevalueofitssubsistenceparameterforbenchmarkconditions(i.e.,includingestimatesofairpollution).TheremainingparametersoftheStoneGearyfunctionswerederivedusingHRW1'scalibrationmethod. 4 Byallowingthesubsistenceparameterforservicestobealinearfunctionofestimatedmorbiditylossesinthebenchmarkyear,theinterceptinthetranslatingequationwasusedtoadjusttheimpactofpollutantssothebenchmarkestimatecorrespondedforeachsubsistenceparameter.Mortalityeffectsweretreatedasa *:&* separableadjustmenttoutility.B. 4 MeasuringtheWelfareConsequencesofTradeandEnvironmentalPolicies T  4 Tomeasurethewelfareeffectsofchangesintradeorenvironmentalpolicies,itisimportanttoacknowledgethatexistingtradepolicies(e.g.,tariff,quota,andsubsidy)generateincomeforsomeeconomicagents.Policychangescanleadtoincomechanges.Thus,efficiencygainsarisingfromeffortstoremovedistortionsinproductorfactorpricesorreduceenvironmentalexternalitiescan,inageneralequilibriumframework,changetheserents.Toacknowledgetheminappliedwelfareanalysisimpliesweacceptasecondbestsituationinevaluatingpolicychanges. 4 FollowingAndersonandNeary[1992b]weadoptaBalanceofTradefunction(BT)toevaluatethewelfareeffectsofpolicychange.ThisisconsistentwithCopeland,exceptheassumedabalanceoftradeequilibriumandweallowfortradesurplusanddeficits().Equation(1)definestheBTfunctionintermsoftheHicksian Z expenditurefunction,e(.),theGNPoraggregaterevenuefunction,g(.),andtariffandquotarents.?+ 4 hDL!X?Ӝ&'>XHDz  p ! @Xdddddddd@x$_"!F !!      !!where 4 q   = h quotas(vector) t% $  4 -   = h proportionalmarkupsduetotariffs F'"&  4 0 4 )ud   0h ! !  = h levelofwellbeingforsinglehousehold(assumedto )h$( representallhouseholdsintheeconomy)*:&*h!h!  4 !   = h domesticprices(i.e.,!i=!i*+tiso!=!*-,   4     h withti=tariffonithcommodity,!*adiagonal   4     h matrixand-i=1+(ti/!i*))   L  T  4 !*   0 h  = h diagonalmatrixofworldpricesondiagonalandzeros & v elsewhere. h!h!  4 m   0 h  = h vectorofimportdemands(m=e!󀄀g!)  h!h!  4 p*   0 h  = h vectorofforeignpricesonquotarestricted   commodities(p=domesticpricevector) h!h!  4 p(q,!)   =virtualpricesofquotarestricted @ Ѐ 4     h goods b  4 (1w)   0 h  = h fractionofquotarentsthatreturntodomestic 4 residents(wisthesharethatgoestoforeigners) h!h!  4    = h thetradesurplusordeficit.   4 T   = h asasuperscriptidentifiesthetransposeofavector Z  4 IntroducingenvironmentalexternalitiesregulatedatsomelevelZdfordomestic N andZftransboundarysourcesaddsthesetwotermstoe(.),g(.),and,therefore,tothe !   virtualpricesfunctions,p(.). 4 Considernowasimultaneouschangeintariffrestrictionsfrom-to)-Ԁ t% $ (with-<)-)andfromZ0dtoZ*d,Zf0toZf*.Usingthebalanceoftradefunctionsto F'"& evaluatethewelfareimplicationsofthispolicychangewehaveequation(2).̜&*DCXHDz  p !@Xdddddddd@xE-_"*Fq0d !!*:&* *:&* (conventionalHicksiancompensatingvariation)̜&;HEXHDz  p ! @Xdddddddd@x_"F0d !!      !!(changeinthevalueoffactorendowmentshouseholdincome)̜&=0,WGCz  0 ! @Xdddddddd@x_"x_"F$ ; !!      !!(changeintariffincome)̜&23WGCz  0 ! @Xdddddddd@x_"x_"F ; !!      !!(changeinquotarents) 4 ReorganizingtermsweseethattheAndersonNearycriteriafortherestrictivenessoftradepolicyiscompletelyconsistentwithHicksianwelfaretheoryinthepresenceofmultiplerestrictions.Thefirstterminequation(2)istheHicksianwillingnesstopayforthecompositechange,andtheremainingtermsadjustforincomeandrentchangesduetothepolicychange.Thezeroprofitconditionalongwithanassumptionthatfactorsareimmobile(acharacteristicoftheHRWmodel)assuresthatchangesintheGNPfunctioncorrespondtothechangeinvalueofdomesticendowments. 4 Thisformulationalsoallowsconsiderationoftheeffectsofseparability.Withseparablepreferences,environmentalresourceswouldbeyetanotheradjustmenttotheincomechanges.TheywouldnotinfluencethevirtualpriceofquotasortheHicksianϜWTPforthepolicychangeindependentofthisincomeadjustment.Inthisgeneral *:&* formulationtheycan,inprincipleinfluencethepuresubstitutioneffectcapturedinthefirsttermaswellastheotherthree"income"relatedchanges. 4 Afinalissuethatthisexpressionillustratesconcernstheroleofthemeasuresofenvironmentalresources.Weassumedinequation(2)thatZdandZfenterboth & v householdexpenditureandGNPfunctionsinthesameway.Thisspecificationimplicitlyembedstheroleoftheenvironment,asamechanismconvertingemissionstoambientconcentrations,inoneofthesetwobehavioralfunctions.RestrictionsondomesticemissionsarerelevanttoZdaswellastoanycontributiondomestic n  emissionsmaketothelevelofthepollutantsinfluencedbytransboundarysources,andthustotheZfintheGNPfunction.Theambientenvironmentandthetransboundary b transferfunctioninfluencewhatisrelevantforthelevelofZfenteringtheexpenditure 4 function. 4 InpracticethelevelofZdintheGNPfunctionisnotthesameasthataffecting  theexpenditurefunction.Thereissomeconversionrelationship.Becausethereisonlyonehouseholdineachcountry,theapproachsimplyscalesemissionsinthecurrentversionofthemodel.Itisnonethelessoneofthewaysspatialdifferenceswithineachregioncouldbeintroducedinamodelwithoutrequiringpreferencedifferences.Thediffusionfunctionsserveastheequivalentofpreferencedifferences. 4 Aswenotedearlier,thePerroniWigleuseofconventionalCESfunctionstocapturethesedifferencesisoneofthewaysofrecognizingtheirimportance.Unfortunately,weknowverylittleaboutthecorrespondencebetweentheuseofthistypeofaveragingfunctionandthetypesofdiffusionmodelsusedbyenvironmental *:&* scientiststolinkambientconcentrationstoemissionsofdifferenttypesofpollutants.Bothinfluenceexcessdemandsandvirtualprices.Thus,thesimplifyingassumptionshighlightedearliercaninprincipleinfluenceallaspectsofawelfareevaluationofpolicychange.Separabilityassumptionsincorrectlyconfinetheirimpactstothetermsadjustingforchangesinincome.̛ 4 C. DotheAssumptionsMatter?    4 Toevaluatewhethertheassumptionsoftherecentresearchontradeandtheenvironmentinfluenceconclusionsaboutwelfareeffectsofliberalization,wehaveconsideredonecompositescenarioinfourdifferentways.Ourpolicychangeintroducesareciprocalreductioninnontariffbarriersby50%(from.20to.10)foralltradebetweentheUKandeachofitsEUtradingregionsinthegoodsfromthedurablemaufacturingsector.Itiscombinedwitha25%increaseintheemissionratesforallthreepollutantsforthissectorintheU.K.Wemightconsiderthesecondcomponentofthescenarioasanapproximatemethodfordescribingarelaxationinenvironmentalstandardstorepresenttheentryofmarginalplantsinresponsetothetradeliberalization. 4 Thefirstapproachforevaluatingthischangeignorestheenvironmentaleffectscompletely.LabeledNE,itfocusesonthemarketeffectsofthereductionintraderestrictiveness.ThesecondapproachadoptstheBoyd,Krutilla,andViscusi[1995]frameworkbykeepingtrackofthechangesinemissionsduetothecompositepolicyafterthemodelissolvedforanewvectorofgeneralequilibriumpriceswiththe *:&* reciprocalreductioninnontariffbarriersinthissectorforEUregions.Theincrementtothebalanceoftradefunctionwithenvironmenteffectsiscomputedbasedonthescenarioincomparisontothebaseline.Thechangesinemissions,correspondinghealtheffects(morbidityandmortality)arecomputedandvaluedatbaselineincrementalWTPs.Thesearetreatedastheequivalentofanexogenousincomechangeusingthefixed"prices"forthehealtheffects.Thisapproach(labeledBKV)imposesastrongerrestrictionthanseparabilityofpreferences(inenvironmentalandmarketedcommodities)becausereductionsinincomeduetothedamagesarenotallowedtoinfluencethecompositionoffinaldemandsformarketedgoodsandimplicitlythegeneralequilibriumlevelsofemissions.Thethirdapproach(labeledENDOW)treatsmorbidityeffectsasreducingtheavailablelaborendowment.Itmightseemplausibletointerpretthisapproachasa"pure"incomereduction.Thiswouldnotbecorrect.Thereducedendowmentsalterfactorandproductprices.Asaresult,substitutioneffectsdoaccompanythechangeinpollution.Theyariseinthiscasefromtheincrementalcostincreasesduetoincreasedlaborscarcityratherthanfromaspecifiedroleoftheenvironmentinpreferences. 4 Finally,thelastevaluationframework(labeledESforEspinosaSmith)recognizestheeffectsofmorbidityonthresholddemandsforservicesandasaresultreflectsthesubstitutioneffectsinducedforothermarketedgoodsaswellasanyincomeeffectsthroughrevaluationofendowmentsorchangesinrents(asgiveninequation(2)). 4 Tables2and3presenttheresultsforthetrade/environmentscenarioevaluated *:&* witheachofthefourdifferentstrategies.Table2comparestheirimpactsonproducerandconsumerpricesandtrade.Qualitatively,theresultsaresimilarwiththedirectionofchangeinallthesectoralpricesandtradesimilarregardlessofhowthepolicyisimplemented(i.e.,withandwithoutenvironmentalconsequencesaswellaswithandwithoutvariousformsofseparabilityimposed).Table3offersthemostdirectillustrationoftheeffectsoftheselaterassumptions.Togaugetheimplicationsofseparability,considerfirsttheimplicationsoftakingaccountoftheairpollutioninducedeffectsonmorbidity.ThisisgiveninthefirstrowofTable3labeledasIA.1.TakingtheESapproachasthe"correct"basisforevaluatingthecompositetradeandenvironmentscenario,wefindthatignoringthisaspectoftheenvironmentwouldcausea12percentoverstatementofthegainrelativetoGDP.ItshouldbenotedthatboththeBTandGDP(i.e.,thenumeratoranddenominatorintheratioreportedin  thetable)arechanginginthesecomparisons. 4 Generalequilibriumadjustmentsaffectthecompositionoffinaldemand,thelevelofemissions,andthevalueoffactorendowments.TheBKVapproachwouldoverstatethelosses.Thiserrorissixpercent,asseeninthebracketedtermintherow N labeledIA.1.Thisoutcomeiswhatshouldbeexpectedbecausetheirapproachignorestheprospectsforsubstitution. 4 Theresultsfromtreatingmorbidityeffectsasanendogenousadjustmenttothelaborendowmentmayatfirstseemsurprising,buttheyshouldnotbe.ThemeasureofwelfarechangeasapercentofGNPiscomparabletotheEScaseinpartbecauseofthespecificformoftheutilityfunctionandtheroleweassignedtopollution.Witha *>&* StoneGearyfunction,adjustmentstothesubsistenceparametersreducediscretionaryincome,duetotheincreaseddemandsforhealthservices.Thisfollowsbecausethesubsistenceparameterforthissectorincreaseswiththeincreaseinthemorbidityeffectsduetoairpollution.AtthebenchmarksolutionintheU.K,thesemorbidityinducedservicedemandswereaboutfivepercentoftheexpendituresonservices(andonepercentoftotalexpenditures). 4 Theendowmentapproachallowsthelaborendowmenttoadjustdownwithincreasedmorbidityeffectsduetoairpollution.Becausethisadjustmentisendogenous,factorpricesandfinalgoods'pricesalsoadjust,changingtheproductmix,andultimately,thepollutionlevelsandassociatedmorbidityeffects.Welfareeffectstakeaccountofthechangeinrelativepricesbetweenthetwoscenariosandtheassociatedadjustmentindemands. 4 Whenweconsiderthemortalityeffectsenteredasaseparableeffectonpreferences,thedistinctionsbetweenthenoenvironment(NE)andotherscenariosbecomesmorepronounced.MortalityeffectswerefivepercentofU.K.totalexpendituresinthebenchmarksolution.Herewefindthatitispossibletomisinterpretthenetbenefitsfromthechangeintradepolicy.Consideringtrade !   liberalizationalone,thereductioninnontariffbarriersenhanceswellbeing.If,aswepostulated,thereisanaccompanyingincreaseintheemissionratesforairpollutioninthissector,thenthereisanetwelfarelossconsideringboththeseparableandnonseparablehealthrelatedeffects.Onereasonforconsideringtheseelementsinstepsistorecognizethatsmallomissionsfromtheenvironmentalconsequencesoftrade *>&* policymaynotbeimportanttothemeasuresofthewelfareconsequencesofthechange.Ofcourse,thisisnotguaranteed.Thenatureofthechange,sectoraffected,andcharacterofenvironmentalimpactsallmatter.Thisisillustratedwithourexamplewhereeventhisrelativelysmallincreaseintotalemissions(from2.0%to6.7%)wassufficienttochangetheverdictonthepolicy. 4 Ofcourse,itisimportanttoacknowledgethatwepostulatedthe25%increase    inemissionratesfromthissector.ComparisonoftheentriesinthefirstcolumnundercategoryBwiththeotherthreecolumnsillustratehowtradeliberalizationaffectedemissionsintheabsenceofthisexogenouslyspecifiedincreaseinemissionratesforthissector.Particulatematterwouldhaveincreasedbyabouttwotenthsofapercent(incomparisontothe6.7%aggregateincreaseimpliedbythescenario). 4 WhiletheprimaryfocusoftheanalysisisontheU.K.,thepriceeffectsandtheenvironmentaleffectshaveimpactsbeyondthisregion.Germanyexperienceswelfaregainsforthetradeliberalizationwhichwouldalsobeoverstatediftheenvironmental ^ effectsarenottakenintoaccount.Thisoverstatementarisesprimarilyfromtheeffectsofincreasedoutputinthedurablemanufacturingsectoronemissionsand,toamorelimiteddegree,fromthetransboundaryeffectsofincreasedemissionsfromtheU.K.forGermany.  $  7      ׀Thedirectionoftheseeffectsbothformorbidityandmortality,andfor #" methodsofmeasurement,arecomparabletowhatwefoundwiththeU.K.Here,wedonothavethesimultaneousincreaseinpollutantsassumedtoaccompanytradeliberalizationintheU.K.'sdurablemanufacturingsector.Nonethelessthereisalargedifferenceintheevaluationofthepolicydependingonwhether(andhowmuchof)the *B&* environmentaleffectsarerecognized.Thedifferencesamongevaluationmethods,ϜBKV,endowment,andourapproacharemuchsmallerbecausetheoverallimportanceofthenonseparableenvironmentaleffectsissmallerasafractionofGDP. 4 Overall,thiscomparisonreinforcestheissuesidentifiedinourreviewoftheconceptualandCGEliteraturetodate.Assumingthatenvironmentalresourcesmakeseparablecontributionstopreferencesnotonlycontradictsthelogicusedinrevealedpreferenceapproachestononmarketvaluation,ithasanimportanteffectontherelativepriceimpactsandwelfaremeasuresderivedforcompositetradeandenvironmentpolicies.Equallyimportant,thereisamoresubtle,butimportantroleplayedbynonseparability.Thesespecificationsrecognizethejointnessinthegeneralequilibriumsolutionbetweenmarketandnonmarketinteractionsbetweeneconomic 4 agents.Whenenvironmentalresourcesinfluencethemarginalratesofsubstitution(inconsumption)formarketgoods,wecanexpectthatexternalitieswillbothaffectandbe  affectedbyfinalgoodchoices.Thefirstoftheseinfluencesisthroughtheinteraction b outsidemarketsandthesecondistheresultofmarketbasedsubstitutiononfinalproductionpatternsandthelevelofemissionsthatresultfromthem. 4 Preferencerelatedimpactsprovideapotentiallyimportantfeedbackloopthathastodatebeenignoredbyexistingconceptualandcomputationalmodels.Intheseapproaches,theexclusivemessageseemstobeonewhereinternationalmarketsmagnifyexternaleffects.However,inamoregeneralapproach,theseexternaleffectscanchangethedemandsformarketedgoodsandservicesand,thereby,transmit $)t$( impactsoutsidemarketsthatnonethelesschangedomesticandinternationalresource *J&* allocationdecisions.IV. 4 Geography,Trade,andtheEnvironment  М̛ 4 Krugman's[1991]essaysonthefrontierissuesininternationaleconomicsarguethatsomeofthemostimportantresearchquestionsintradetheory,aswellasintheanalysisoftradepolicy,followfromgreaterrecognitionthatgeographymatterstoeconomicactivities.Inhismodels,thespatialorgeographicconcentrationofeconomicactivitiesarisesfromacollectionofforcesincreasingreturnstoacommonlocationinsomeproductionactivitiesthatmustbebalancedagainsttransportcoststothelargermarketsrequiredtotakeadvantageoftheseeconomies,aswellasbythedegreetowhichproductioncanbe"footloose." 4 Heusestwoprimaryapproachestoargueforgreaterattentionininternationaleconomicmodelstoeconomicgeography.Thefirstisabottomupframeworkthatlooksatthemicroforcesleadingtodifferentregionalpatternsofactivitythatmaytranscendseparateeconomiesastheyaretraditionallydefined.Heconcludeswithatopdownperspective,thatpostulatescountriesastheprimarysourceofrestrictionstothemovementoffactorsandgoods.Inhisanalysis,nationaleconomiesaresimplysourcesofinefficiencyinresourceallocations. 4 Inbothsituations,space(orlocation)mattersinprovidingexternaleconomies,butthefactorsgivingrisetoeithertheadvantagesorthecostsmustbepostulatedbytheanalyst.Increasingreturnsexisttheirsourceishumancapitalorotherforcesoftenoutsidethemodelandthusnotexplainedbyit.Analternativeexplanationfor *:&* thetendencieshefindsiswiththeenvironmentalresourcesthatregionallyandgloballytransmitexternaleffects. 5 $  8         4 Chichilnisky'sanalysisconcludedthatgreaterattentiontotradeandenvironmentinteractionswaswarrantedbecauseinternationalmarketsmagnify & v externalities.Ourdiscussionofhowenvironmentalexternalitiescaninfluencegeneralequilibriumevaluationsoftradeandenvironmentalpoliciesemphasizedthefeedback    loopor,moresimply,theimplicationsofinteractionsbetweenagentsthatariseoutside   marketsbecauseofoneormoreenvironmentalresources.Throughnonseparabilitiestheseinteractionscanplayadirectroleinthesubstitutionsthatinfluenceoutcomeswithinmarkets.Totheextentnonmarketedenvironmentalresourcesmakenonseparablecontributionstoproductionandpreferences,theyinfluencethesignalsformarketedgoodsfrombothsidesofmarkettransactions.Theseissuesarelikelytobeatleastasimportantfordevelopingeconomieswheretheenvironmentalresourcesoftenserveakeyroleinagriculture(e.g.LopezandNekletshek'snaturalbiomass)orasextractiveoutputsbutcouldalsocontribute(inapreservedstatus)towatershedprotection,microclimaticfunctions,andamenities(e.g.BarbierandRauscher). 4 Whatdeterminestheimportanceoftheseinteractionsoutsidemarkets?Oneexplanationisthatenvironmentalmediadefinetherelevantgeographicextentofthepositiveandnegativeinteractionsoutsidemarkets.Thiswasillustratedbythedifferencesbetweenlocalandtransboundarypollutants.Differencesintariff,nontariffbarriers,anddomestictaxesprovidethebasisforseparatingthe"regions"withintheϜHRW1model.Theyprovidetheprice"wedges"thatdefineeacheconomy.Without *F&* themsomeotherexogenousconstraintmustbeusedtoseparateordistinguishactivitieswithineachcountry.WithinCGEmodels,theserestrictionsprovidetheprimarylimitto"footloose"movementofproduction. 4 Oncetherestrictionsaredefined,however,itispossibletoconsidersomethingcompletelydifferentthediffusionofairpollutionthroughtransfercoefficientsthatdefine,inasimpleway,thebasisforlocalandtransboundaryfeedbacksintheinternationaleconomy.Butwecouldrethinkmattersandstartwiththeenvironmentaltransferfunctionand,then,considerhowthepricewedgesenter. 4 Thus,akeylessonthathasmoregeneralimplicationsbeyonddiscussionofcoordinationoftradeandenvironmentalpoliciesisthebasicpremiseofKrugman'sessays.Regionalanalysiscaninformdiscussionsoftheissuesrelevanttointernationaleconomicmodelingbecauseithelpstofocusontheforcesthatcreatepositiveandnegativeinteractions.Inhisapproachtherearestillomittedfactors,enteringas"accidents"ofhistoricalinteractionsthatgaverisetotheregions.Perhapsstartingwiththeroleoftheenvironmentalresourcesservingasonesourceofserviceswithpositiveandnegativeeffectswouldhelptakethe"mystery"outoftheoriginsofeconomicallymeaningfulregions.Transportcostsandfootloosebehaviorarenotthewholestory. 4 Therelevanceofthisperspectiveis,ofcourse,notlimitedtotheinteractionsoftradeandenvironmentalpolicy.Analysesofregionaleconomicissuesandnationalenvironmentalpoliciesareequallyplausiblecandidates.Hereinstitutionalrestrictionsprovidingfordifferencesinpricesarelessrelevant.Thepropertiesofdiffusion *:&* systems,whetherlinkingorseparatingactivities,nowhavetheopportunitytoinfluenceeconomicbehavior.Thus,weareleftwithanothersourceofendorsementforKrugman'srecommendationtolookatregionalpatternsofeconomicactivitiesfortheroleofconstraints(andincentives)whetherinstitutional,technological,or & v environmental,forthescaleanddispersionofeconomicactivities.  L    "  TR'33''3Letter (Landscape)3'T/XX{X#X@ : Table1:BenchmarkAirPollutionLevelsandtheMorbidityandMortalityLosses  @;MasSharesofServiceandTotalExpenditures   _______________________________________________________________________________________________________{[X` N hp xG"5%'),.8135@8:<H?AXhX{   `  N BenchmarkAirPollution   p      x    G" Benchmark '  ) Mortality k [X` h*p f!(#5%'),.8135@8:<H?AXN XpXG"5%'),X   `     h  *    p        f! MorbidityEffects '  ) Effects T *h14 dh,@dd ,@dd ,@dd ,@dd ,@dd ,@dd ,@dd ,@dd +  3)" =  Cx3Region 2)! ) 2PM(g/m3) 2)! ) 2SOx   (ppm) 2)! ) 2NOx   (g/m3) 2)! ) 2OwnDepositionRate 2)! ) 2ShareofTotalExp. 2)! ) 2ShareofServicesExp. 2)! ) 2ShareofTotalExp. C9!  x C ) axCGermany   H  59.55   H  .024   H  49.50   H  .42   H  .028   H  .112   H  .172 6, H  x a ax6France  Y  24.32  Y  .010  Y  22.34  Y  .42  Y  .013  Y  .049  Y  .045 6,Y  x a ax6Italy    83.85    .029    46.98    .65    .031    .146    .157 6,  x a ax6Netherlands  k  60.28  k  .010  k  46.05  k  .15  k  .025  k  .131  k  .110 6,k  x a ax6Belgium  |  91.40  |  .015  |  49.59  |  .31  |  .040  |  .216  |  .194 6,|  x a Jx6UnitedKingdom   19.85  - .014  - 31.98  - .83  - .013  - .046  - .051 6,- x J ax6Ireland  'w 16.32  'w .009  'w 25.16  'w .51  'w .006  'w .023  'w .014 6,'w x a ax6Denmark   60.61   .009   49.33   .15   .047   .255   .275 6, x a ax6U.S.A.  9 32.78  9 .006  9 53.46  9 1.00  9 .012  9 .031  9 .047 6,9 x a ax6Japan  J 35.02  J .008  J 39.15  J 1.00  J .023  J .069  J .058 =),J x a Zx=RestofWorld 2)! ) 248.40 2)!) 2.013 2)!) 246.19 2)!) 2.95 2)!) 2.047 2)!) 2.128 2)!) 2.3376XX{X/X1'% x Z ) 1  v# TRs3'Letter (Portrait)'33'Letter (Portrait)  '3Letter (Landscape)T"XXgX6X@I Table2    @$AComparisonofPriceandTradeEffectsofTradeandEnvironmentalPolicyChanges   ____________________________________________________________________________   `     h  *   50%NTBforUKDurables & v *<? d d@dd @dd @dd @dd @dd @dd @dd @dd 14, dd,8dde,8dde,8dde,ddu+  3)"  _ qx3UnitedKingdom 2)! ) 2NE 9)( ") 9BKVa 9)( ") 9ENDOWa 9)( ") 9ESa J@( " x q ) axJI.ConsumerPrices(%)   H  Ј   H  Ј   H  Ј   H  Ј 6, H  x a ax6ЀAgricultureandFood B8 Y  sh|??0.8515sh|??BЈ.8515 f\-Y  sh|??0.8515 sh|?? =yX5?0.8503=yX5?fЈ.8503 f\-Y  =yX5?0.8503 =yX5? k+ݓ?0.8618k+ݓ?fЈ.8618 f\-Y  k+ݓ?0.8618 k+ݓ? H}?0.8591H}?fЈ.8591 ZP8Y  H}?0.8591 x a H}? axZЀMiningandEnergy B8    q?0.9018 q?BЈ.9018 f\-   q?0.9018  q? sA?0.9003sA?fЈ.9003 f\-  sA?0.9003 sA? H.?0.9119H.?fЈ.9119 f\-  H.?0.9119 H.? ^)?0.9113^)?fЈ.9113 ZP8  ^)?0.9113 x a ^)? axZЀNonDurableManufacturing B8 k  H.?0.8124H.?BЈ.8124 e[-k  H.?0.8124 H.? ʡE?0.811ʡE?eЈ.8110 e[,k  ʡE?0.811 ʡE? &†W?0.8232&†W?eЈ.8232 f\-k  &†W?0.8232 &†W? H?0.8213H?fЈ.8213 ZP8k  H?0.8213 x a H? axZЀDurableManufacturing C9 |   0*-1.1979 0*CЄ1.1979 h^.|   0*-1.1979  0* X2ı.-1.1989X2ı.hЄ1.1989 f\.|  X2ı.-1.1989 X2ı.  ףp= -1.19 ףp= fЄ1.1900 f\,|   ףp= -1.19  ףp=  2U0*-1.19232U0*fЄ1.1923 [Q9|  2U0*-1.1923 x a 2U0* ax[ЀConstructionandTransportation C9 - H}8g-1.0877H}8gCЄ1.0877 g].- H}8g-1.0877 H}8g V_?1.0859V_?g1.0859 e[-- V_?1.0859 V_? V-?1.101V-?e1.1010 e[,- V-?1.101 V-? j+?1.0996j+?e1.0996 ZP8- j+?1.0996 x a j+? axZЀServices B8 > =U?1.1669=U?B1.1669 f\-> =U?1.1669 =U? ??1.1651??f1.1651 f\-> ??1.1651 ?? Cl?1.1815Cl?f1.1815 e[-> Cl?1.1815 Cl? w/?1.179w/?e1.1790 YO7> w/?1.179 x a w/? axYII.ProducerPrices(%)   Ј   Ј   Ј   Ј 6, x a ax6ЀAgricultureandFood B8 P Gr?1.1074Gr?B1.1074 f\-P Gr?1.1074 Gr? W2ı?1.1059W2ı?f1.1059 e[-P W2ı?1.1059 W2ı? V-?1.121V-?e1.1210 e[,P V-?1.121 V-? zG?1.1175zG?e1.1175 ZP8P zG?1.1175 x a zG? axZЀMiningandEnergy B8 a TR'?1.1681TR'?B1.1681 f\-a TR'?1.1681 TR'? a2U0*?1.1663a2U0*?f1.1663 f\-a a2U0*?1.1663 a2U0*? <,Ԛ?1.1813<,Ԛ?f1.1813 f\-a <,Ԛ?1.1813 <,Ԛ? S?1.1805S?f1.1805 ZP8a S?1.1805 x a S? axZЀNonDurableManufacturing B8  &1?1.0695&1?B1.0695 f\- &1?1.0695 &1? _L?1.0677_L?f1.0677 e[- _L?1.0677 _L? MbX?1.084MbX?e1.0840 e[, MbX?1.084 MbX? :M?1.0813:M?e1.0813 ZP8 :M?1.0813 x a :M? axZЀDurableManufacturing B8 #s  q?0.9984q?BЈ.9984 f\-#s  q?0.9984 q? ۊe?0.9967ۊe?fЈ.9967 f\-#s  ۊe?0.9967 ۊe? 46 Ј  > Ј  > Ј  > Ј 6,> x a ax6Ѐ%Emissions   Ј   Ј   Ј   Ј 6, x a ax6ЀPM A7 P S㥛?0.161S㥛?AЈ.161 dZ,P S㥛?0.161 S㥛? B`"@6.714B`"@d6.714 dZ,P B`"@6.714 B`"@ sh|@6.687sh|@d6.687 dZ,P sh|@6.687 sh|@ ʡ@6.693ʡ@d6.693 YO7P ʡ@6.693 x a ʡ@ axYЀNOx B8 a "~j-0.111"~jBЄ.111 e[-a "~j-0.111 "~j V-@2.087V-@e2.087 dZ,a V-@2.087 V-@ rh|@2.061rh|@d2.061 dZ,a rh|@2.061 rh|@ Pn@2.064Pn@d2.064 YO7a Pn@2.064 x a Pn@ axYЀSOx B8  J +-0.098J +BЄ.098 e[- J +-0.098 J + "~j@2.427"~j@e2.427 dZ, "~j@2.427 "~j@ K7@2.402K7@d2.402 dZ, K7@2.402 K7@ sh|?@2.406sh|?@d2.406 YO7 sh|?@2.406 x a sh|?@ axYЀ%HealthEffects  #s  Ј  #s  Ј  #s  Ј  #s  Ј 6,#s  x a ax6ЀMorbidity  " Є A7 " X9v?1.654X9v?A1.654 dZ," X9v?1.654 X9v? RQ?1.645RQ?d1.645 dZ," RQ?1.645 RQ? '1Z?1.647'1Z?d1.647 YO7" '1Z?1.647 x a '1Z? axYЀMortality  5$ Є A7 5$ B`"@6.714B`"@A6.714 dZ,5$ B`"@6.714 B`"@ sh|@6.687sh|@d6.687 dZ,5$ sh|@6.687 sh|@ ʡ@6.693ʡ@d6.693 YO75$ ʡ@6.693 x a ʡ@ axYII.Germany  F & Ј  F & Ј  F & Ј  F & Ј 6,F & x a ax6ЀA.WelfareChange  !( Ј  !( Ј  !( Ј  !( Ј 6,!( x a ax6Ѐ1.BT/GDP(%) B8 #X* S!uq?0.0134S!uq?BЈ.0134 f\-#X* S!uq?0.0134 S!uq? ( 0?0.0123( 0?fЈ.0123 e[-#X* ( 0?0.0123 ( 0? ~jt?0.012~jt?eЈ.0120 e[,#X* ~jt?0.012 ~jt? @Lj?0.0121@Lj?eЈ.0121 ZP8#X* @Lj?0.0121 x a @Lj? axZЀ(excludingmortality)  i$, [0.11]  i$, [0.02]  i$, [0.00]  i$, Є 6,i$, x a ax6Ѐ2.BT/GDP(%) B8 %!. S!uq?0.0134S!uq?BЈ.0134 e[-%!. S!uq?0.0134 S!uq? y&1|?0.007y&1|?eЈ.0070 e[,%!. y&1|?0.007 y&1|? S!uq{?0.0067S!uq{?eЈ.0067 e[-%!. S!uq{?0.0067 S!uq{? y&1|?0.007y&1|?eЈ.0070G=;%!. y&1|?0.007 x a  y&1|? G*JK dd dd8dde8dde8dde8ddeGI, dd,8dde,8dde,8dde,8dde+  3)" &"0 qx3Ѐ(includingmortality) 2)!+'{"1) 2[0.91] 2)!+'{"1) 2[0.00] 2)!+'{"1) 2[0.04] 2)!+'{"1) 2Є1'%+'{"1 x q ) 1aWhenthe50%reciprocalreductioninnontariffbarriersfordurablesisimplementedfor $(t#3 thesecomputationstheemissionratesforallairpollutantsgeneratedbytheUKdurablesectorareincreasedby25%.  )F%5 @ GReferences0   Anderson,Kym,1992, TheStandardWelfareEconomicsofPoliciesAffectingTradeandtheEnvironmentinTheGreeningofWorldTradeIssueseditedbyKymAnderson k andRichardBlackhurst(AnnArbor;MI,UniversityofMichiganPress). (#(# 0   Anderson,J.E.andJ.P.Neary,1992a, TradeReformUnderQuotas,PartialRentRetentionandTariffs,Econometrica60(January):5776. a(#(# 0   Anderson,J.E.andJ.P.Neary,1992b, ANewApproachtoEvaluatingTradePolicy,WorkingPaperWPS1022,InternationalEconomicsDepartment,TheWorldBank,November. (#(# 0   Ballard,C.L.andS.G.Medema,1993. TheMarginalEfficiencyEffectsofTaxesandSubsidiesinthePresenceofExternalities:AComputationalGeneralEquilibriumApproach,JournalofPublicEconomics52(September):199216[ (#(#   XBarbier,EdwardB.andMichaelRauscher,1994,"Trade,TropicalReforestationandPolicyInterventions,"EnvironmentalandResourceEconomics,4(January):7590. h 0   Bergman,Lars,1993. CarbonTaxationinaSmallOpenEconomy:SectoralDifferentiation < vs.InternationalCoordination,unpublishedpaper,StockholmSchoolofEconomics. (#(# 0   0(#(#  Bergstrom,Theodore,LawrenceBlumandHalVarian,1986,"OnthePrivateProvisionofPublicGoods,"JournalofPublicEconomics,29(February):2550.(#(# 0   Boyd,Roy,KerryKrutillaandW.KipViscusi,1995,"EnergyTaxationasaPolicyInstrumenttoReduceCO2Emissions:ANetBenefitAnalysis,"Journalof M EnvironmentalEconomicsandManagement,29(July):124.8(#(# 0   Chichilnisky,Graciela,1994, NorthSouthTradeandtheGlobalEnvironment,American  \ EconomicReview,84.(September):851875. G(#(# Copeland,BrianR.,1994,"InternationalTradeandtheEnvironment:PolicyReforminaPollutedSmallOpenEconomy,"JournalofEnvironmentalEconomicsand #" Management,26(January):4465. $# 0   Copeland,BrianR.andM.ScottTaylor,1994, NorthSouthTradeandtheEnvironment,QuarterlyJournalofEconomics,109(August):755787.\'"&(#(# 0   Copeland,BrianR.andM.ScottTaylor,1995a, TradeandTransboundaryPollution,AmericanEconomicReview(inpress).*i%)(#(# Copeland,BrianR.andM.ScottTaylor,1995b,"Trade,Nonconvexitiesandthe +='+ Environment,"DepartmentofEconomics,UniversityofBritishColumbia,UBCdiscussionpaperno.9519,July.0   Desvousges,W.H.,H.S.Banzhaf,J.S.Smith,F.R.JohnsonandK.N.Wilson,1994.AssessingEnvironmentalExternalityCostsforElectricityGenerationinWisconsin, T Vol.Iand3(ResearchTrianglePark,N.C.:ResearchTriangleInstitute).= (#(# 0   Espinosa,J.AndrsandV.KerrySmith,1995,"MeasuringtheEnvironmentalConsequencesofTradePolicy:ANonMarketCGEAnalysis,"AmericanJournalof  J AriculturalEconomics77(August):772777. 5 (#(# 0   Freeman,A.MyrickIII,1993.TheMeasurementofEnvironmentalandResourceValues:   TheoryandMethods.(Washington,D.C.:ResourcesfortheFuture). (#(# 0   0(#(#  Gamalestsos,T.,1971.InternationalComparisonsofConsumerExpenditurePatterns:An x  EconometricAnalysis,(unpublishedPh.D.thesis,UniversityofWisconsin,Madison).c (#(# 0   Grossman,G.M.andA.B.Krueger,1993, EnvironmentalImpactsofaNorthAmericanFreeTradeAgreement,inP.Garber,editorTheMexicoU.S.FreeTrade  p Agreement,(Cambridge,Ma.:MITPress). [(#(# 0   Harrison,G.W.,T.F.RutherfordandJ.Wooton,1989."TheEconomicImpactoftheEuropeanCommunity,"AmericanEconomicReviewPapersandProceedings79  (May):288294. (#(# 0   0(#(#  0(#(#  Harrison,G.W.,T.F.RutherfordandI.Wooton,1991."AnEmpiricalDatabaseforaGeneralEquilibriumModeloftheEuropeanCommunity."EmpiricalEconomics16 n (January):95120. (#(# 0   0(#(#  0(#(#  0(#(#  0(#(#  0(#(#  0(#(#  0(#(#  0(#(#  0(#(#  0(#(#  Hazilla,M.andR.J.Kopp,1990."SocialCostofEnvironmentalQualityRegulations:AGeneralEquilibriumAnalysis."JournalofPoliticalEconomy98(August):853873. d(#(# Helpman,E.,andPO.Krugman,1985,MarketStructureandForeignTrade(Cambridge, !8  Ma.:MITPress).0   Krutilla,Kerry,1991, EnvironmentalRegulationinanOpenEconomy,Journalof $# EnvironmentalEconomicsandManagement20(March):127142.% $(#(# Krugman,Paul,1991,GeographyandTrade(Cambridge,Ma.:MITPress). d'"& 0   Lluch,C.andA.A.Powell,1975."InternationalComparisonsofExpenditurePatterns."EuropeanEconomicReview15(4):786797.!*q%)(#(# 0   Lopez,RamAn,1994, TheEnvironmentasaFactorofProduction:TheEffectsof +E'+ EconomicGrowthandTradeLiberalization,JournalofEnvironmentalEconomics  andManagement,27(September)163184.(#(# Lopez,Ram;nandMarioNikletschek,1991,"DualEconomicGrowthinPoorTropicalAreas,"JournalofDevelopmentEconomics,36:189211. X 0   Markandya,A.andB.Rhodes,1992.ExternalCostsofFuelCycles:AnImpactPathway , | ApproachEconomicValuation,unpublishedreportforEC/USFuelCycleStudy,  g Metroeconomica,Ltd.,June. (#(# 0   Oates,WallaceE.,PaulR.PortneyandAlbertM.McGarland,1989,"TheNetBenefitsofIncentiveBasedRegulation:ACaseStudyofEnvironmentalStandardSetting,"AmericanEconomicReview,59,(December):12331242. (#(# 0   Perroni,C.andR.M.Wigle,forthcoming."InternationalTradeandEnvironmentalQuality:HowImportantaretheLinkages?"CanadianJournalofEconomics.c (#(# 0   Piggott,J.,J.WhalleyandR.Wigle,1992."InternationalLinkagesandCarbonReduction 7 Intiatives,"inK.AndersonandR.Blackhurst,editors,TheGreeningofWorldTrade  p Issues(AnnArbor:UniversityofMichiganPress). [(#(# 0   Pollak,R.A.andT.J.Wales,1981."DemographicVariableinDemandAnalysis." / Econometrica49(November):15331551.(#(# 0   Repetto,Robert,1995,Jobs,CompetitivenessandEnvironmentalRegulation:WhatArethe  RealIssues?(Washington,D.C.:WorldResourcesInstitute).(#(# 0   Smith,V.Kerry,1992,"EnvironmentalCostingforAgriculture:WillitBeStandardFareintheFarmBillof2000?"AmericanJournalofAgriculturalEconomics,74, F (December):10761088. (#(#   !<    Footnotes ~*ArtsandSciencesProfessorofEnvironmentalEconomics,DukeUniversityandResourcesfortheFutureUniversityFellowandEconometrician,TRSRiskManagement,AmericanExpressCompanyrespectively.ThanksaredueGlennHarrisonforprovidingthedataandcodetoimplementHRWIaswellasforhelpingustounderstandhowtouseit,toananomymousreferee,EdwardBarbierandCharlesPerringsformosthelpfulcommentsonanearlierdraft,andtoPaulaRubioforpreparingandeditingseveralversionsofthispaper.PartialsupportforSmith'scontributiontothisresearchwasprovidedbytheNationalOceanicandAtmosphericAdministrationthroughGrantNo.NA46GP0466.EspinosasresearchwassupportedbyResourcesfortheFuture.Theconclusionspresentedarethoseoftheauthorsandnottheiraffiliatedorganizations.̜