iscussion Papr No. 06-018 Efficincy Losss from Ovrlapping Economic Instrumnts in Europan Carbon Emissions Rgulation Christoph Böhringr, Hnrik Koschl and Ulf Moslnr
iscussion Papr No. 06-018 Efficincy Losss from Ovrlapping Economic Instrumnts in Europan Carbon Emissions Rgulation Christoph Böhringr, Hnrik Koschl and Ulf Moslnr ownload this ZEW iscussion Papr from our ftp srvr: ftp://ftp.zw.d/pub/zw-docs/dp/dp06018.pdf i iscussion Paprs dinn inr möglichst schnlln Vrbritung von nurn Forschungsarbitn ds ZEW. i Biträg lign in allinigr Vrantwortung dr Autorn und stlln nicht notwndigrwis di Minung ds ZEW dar. iscussion Paprs ar intndd to mak rsults of ZEW rsarch promptly availabl to othr conomists in ordr to ncourag discussion and suggstions for rvisions. Th authors ar solly rsponsibl for th contnts which do not ncssarily rprsnt th opinion of th ZEW.
Non-Tchnical Summary Enrgy markts and nrgy-intnsiv industris in all EU mmbr stats spcially in Grmany ar subjct to a divrs st of policis rlatd to climat chang. Th main climat policy instrumnt within th Europan Union is th EU Emissions Trading Systm (EU ETS) which is in plac sinc 2005. In policy practic, th launch of a nw rgulation in gnral dos not (fully) rplac xisting policy instrumnts that may hav bn stablishd with similar policy objctivs. Thus, a nw rgulation oftn oprats in paralll and intracts with prvious rgulation. Thr ar svral considrations as to why a mix of policy instrumnts might b prfrabl to a singl instrumnt. iffrntiatd instrumnts can b justifid if thr ar additional or multipl objctivs to b implmntd by th policy, such as political, social or tchnology-rlatd critria that may conflict with pur (static) fficincy considrations. Scond-bst rgims, which ar charactrisd by initial markt distortions or imprfctions,.g., provid a gnral argumnt for diffrntiatd rgulation. Such rgims includ situations with uncrtainty, xtrnal knowldg spillovrs, initial tax distortions, markt powr, transaction costs, tc. In climat policy dsign, sctor-spcific diffrncs in transaction costs hav,.g., bn usd as an argumnt for applying diffrnt climat policy instrumnts to diffrnt conomic sctors. Th following analysis abstracts to a larg xtnt from markt distortions and instad focuss on th static fficincy implications of mission taxs imposd on nrgy-intnsiv sctors that ar in addition subjct to th EU ETS. Th potntial fficincy losss from a simultanous application of both instrumnts in qualitativ and quantitativ trms ar analysd within a partial quilibrium framwork for th EU. It turns out that thos firms within th EU ETS which at th sam tim ar subjct to domstic nrgy or carbon taxs will abat infficintly much whil othr firms within th EU ETS will bnfit from lowr intrnational mission prmit prics. Th sam logic disprovs th argumnt that additional national mission taxs will rduc infficincis in abatmnt supposd to b rsulting from allowanc (ovr-) allocation. In ssnc, unilatral mission taxs within th EU ETS ar cologically inffctiv and subsidis nt prmit buyrs. Thus, all firms that ar subjct to missions trading and any CO 2 mission taxs at th sam tim should b xmpt from th lattr. Th forgon tax rvnu could b gnratd by auctioning a small fraction of th prmits instad. This would b chapr for th missions trading sctors as a whol and could b compatibl vn with th tight auctioning rstrictions of th EU dirctiv.
Efficincy Losss from Ovrlapping Economic Instrumnts in Europan Carbon Emissions Rgulation Christoph Böhringr a,b, Hnrik Koschl a, Ulf Moslnr a a Cntr for Europan Economic Rsarch (ZEW), Mannhim, Grmany b Cntr for Environmntal Economics, Univrsity of Hidlbrg, Grmany Fbruary 2006 Abstract: Enrgy markts and nrgy-intnsiv industris in all EU mmbr stats spcially in Grmany ar subjct to a divrs st of policis rlatd to climat chang. W analys th potntial fficincy losss from simultanous application of mission taxs and missions trading in qualitativ and quantitativ trms within a partial quilibrium framwork for th EU. It turns out that thos firms within th EU Emissions Trading Schm (EU ETS) which at th sam tim ar subjct to domstic nrgy or carbon taxs will abat infficintly much whil othr firms within th EU ETS will bnfit from lowr intrnational mission prmit prics. Th sam logic disprovs th argumnt that additional national mission taxs will rduc infficincis in abatmnt supposd to b rsulting from allowanc (ovr-) allocation. In ssnc, unilatral mission taxs within th EU ETS ar cologically inffctiv and subsidis nt prmit buyrs. Thus, all firms that ar subjct to missions trading and any CO 2 mission taxs at th sam tim should b xmpt from th lattr. Th forgon tax rvnu could b gnratd by auctioning a small fraction of th prmits instad. This would b chapr for th missions trading sctors as a whol and could b compatibl vn with th tight auctioning rstrictions of th EU dirctiv. Kywords: missions trading, mission taxs, National Allocation Plans JEL Classifications: 61, H21, H22, Q58 Corrspondnc: Ulf Moslnr ZEW, P.O.Box 10 34 43, -68034 Mannhim, Grmany phon: +49 621 1235 209, fax: +49 621 1235 226, mail: moslnr@zw.d
1. Introduction Th main climat policy instrumnt within th Europan Union is th EU Emissions Trading Systm (EU ETS), in plac sinc 2005 (EU 2003, 2004). In policy practic, th launch of a nw rgulation in gnral dos not (fully) rplac xisting policy instrumnts that may hav bn stablishd with similar policy objctivs. Thus, a nw rgulation oftn oprats in paralll and intracts with prvious rgulation (s Johnston, 2003, or Sorrll and Sijm, 2005, for an ovrviw). Th objctiv of this papr is to shd som light on th conomic fficincy implications associatd with th co-xistnc of th EU ETS and mission or nrgy taxs lvid at th mmbr stat lvl. According to basic conomic principls, th us of a mix of policis in ordr to pursu a singl policy objctiv will b at bst rdundant and at worst countrproductiv (Johnston, 2003). If thr is an fficint instrumnt to implmnt an nvironmntal targt, it maks littl sns to introduc an additional on. Nvrthlss, it is in th natur of nvironmntal policy in a fdral systm such as th EU that instrumnts introducd on a Europan lvl ar complmntd by instrumnts of th mmbr stats. From a mor subtl thortical point of rasoning, thr ar svral considrations as to why a mix of policy instrumnts might nvrthlss b prfrabl to a singl instrumnt. iffrntiatd instrumnts can b justifid if thr ar additional or multipl objctivs to b implmntd by th policy, such as political, social or tchnology-rlatd critria that may conflict with pur (static) fficincy considrations. Scond-bst rgims, which ar charactrisd by initial markt distortions or imprfctions,.g., provid a gnral argumnt for diffrntiatd rgulation. Such rgims includ situations with uncrtainty, xtrnal knowldg spillovrs, initial tax distortions, markt powr, transaction costs, tc. In climat policy dsign, sctor-spcific diffrncs in transaction costs hav,.g., bn usd as an argumnt for applying diffrnt climat policy instrumnts to diffrnt conomic sctors. Th following analysis abstracts to a larg xtnt from markt distortions and instad focuss on th static fficincy implications of mission taxs imposd on nrgy-intnsiv sctors that ar in addition subjct to th EU ETS. In this contxt, two popular argumnts in favour of an additional mission tax on rsidual missions from ETS sctors will b critically discussd and finally disprovd in this papr: 1. A tax would hlp to bring down th missions. This would b cologically bnficial and hlp to rach th mission targt of th EU burdn sharing agrmnt. 2. Givn that th allocation of prmits to th trading sctors has bn vry gnrous, th marginal abatmnt costs (in th trading sctors) can b xpctd to b lowr than in 1
th rst of th conomy. A tax would bring th marginal abatmnt costs in th ETS sctors closr to th fficint lvl. Th tax could thrfor b usd as a scond-bst instrumnt to incras th fficincy of national or EU wid abatmnt. Th papr is structurd as follows: Chaptr 2 first introducs a simpl modl suitabl to dal with th EU ETS. This will b followd by an analytical discussion of th intraction of th missions trading systm and mission taxs outsid of and within th EU ETS. Chaptr 3 will thn prsnt a quantitativ numrical illustration for th formr EU 15. Th last chaptr will conclud th papr and summariz th main rsults. 2. A modl for missions rgulation in th EU mmbr stats In th following th ffcts of possibl ovrlaps of xisting CO 2 missions rgulations with th EU ETS ar illustratd on th basis of a simplifid partial modl for th mission prmit markt. Each mmbr stat is rprsntd by th abatmnt cost functions of two sctor groups. Th first on (IR sctors) includs all sctors and firms that ar subjct to th EU trading dirctiv (including th powr sctor, oil rfining, svral nrgy-intnsiv industris). Th scond on (NIR sctors) covrs all sctors outsid th scop of th dirctiv (including privat housholds, transport, trad). As it is prscribd by th EU trading dirctiv, th IR sctors ar allocatd thir mission budgt and can trad th prmits thraftr. In contrast, th NIR sctors do not participat in trad. In ordr to achiv thir national Kyoto targts, th mmbr stats ar rquird to tak complmntary action in ths sctors. In th following it is assumd that th mission rductions in th NIR sctors ar implmntd cost fficintly by mission taxs. 1 This sction starts with a brif commnt on th aggrgation of individual abatmnt cost functions into on aggrgat function (Sction 2.1). Sction 2.2 summariss th ky issus rlatd to dsigning th National Allocation Plans (NAPs) and points out how th initial prmit allocation rlats to th optimal us of taxs in th NIR sctors. Ths introductory rmarks ar followd by an analysis of how prmits and mission taxs intract within th Europan trading schm. This is at first don for idntical scops of tax and prmit rgulation in all EU mmbr stats (Sction 2.3) and thn for th cas that taxs ar applid in 1 Thortically, in ordr to xamin th rsulting aggrgat burdn on both sctor groups within a country, mor dtail on th us of tax rvnus is ncssary. Rvnu rcycling issus, howvr, will b ignord in th following. For rasons of simplicity, th analysis concntrats on th fficincy ffcts (in trms of th allocation of abatmnt masurs) on th lvl of mmbr stats or th whol EU. 2
th IR sctors in addition to th trading schm (Sction 2.4), also assuming first an fficint bas cas and thn a suboptimal initial prmit allocation. 2.1. Potntial fficincy loss from suboptimal aggrgation of marginal abatmnt costs A cntral rol in minimising th abatmnt costs across diffrnt sourcs is playd by th (marginal) abatmnt cost functions: If marginal abatmnt costs of two sourcs ar qual, thn thir common abatmnt is implmntd at minimum costs. Assum that th two sctors of th conomy, IR and NIR, ar rprsntd by th marginal abatmnt cost functions ' ' C ( ) and C ( ), rspctivly. Th aggrgatd marginal abatmnt cost function N N ' C ( ) can b drivd from th horizontal sum of th sparat marginal abatmnt cost functions: Th functions C ) and C ) ar aggrgatd to C ( ) in Figur 1. ( N ( N min C C min ( ) C ( ) max C ( N ) C ) N ( ΔC ( ) * * N Figur 1: Aggrgation of two marginal abatmnt cost functions. = * + * N Although this is common knowldg, on should kp in mind that C ( ) is alrady min th rsult of minimising th sum of abatmnt costs for a givn mission lvl, i.. at = + thr is C ) = C ( ). In principl, th aggrgatd abatmnt cost N ( N N function dpnds on both mission lvls according to C, ) = C ( ) + C ( ). Evn undr th constraint N singl sctors will caus th aggrgatd marginal abatmnt costs to incras. In th worst cas, all abatmnt is achivd xclusivly by th sctor with th highr marginal abatmnt 3 ( N N N = + th dviation from th optimal mission lvls of th
costs lading to C ( ). Th shadd ara in Figur 1 indicats th rang of th marginal max abatmnt costs if non-optimal allocation btwn th sctors is prmittd. This aggrgation problm should b rcalld whil discussing th EU ETS, whr svral parts of th conomy ar rgulatd in diffrnt ways. It is in fact at last plausibl that th mission rgulation rsulting from a political procss will hardly match th optimal distribution of abatmnt activitis among th two sctors. 2.2. Th challng of dsigning allocation plans in th EU ETS It is on of th strngths of prmit trading rgims that in thory th fficincy is indpndnt of th initial prmit allocation. Th missions trading schm implmnts a givn mission targt at last costs but only among th sctors that ar trading th missions. Sinc th EU ETS dos not covr th whol conomy, but only th IR sctors, th ovrall fficincy of mission rduction is affctd by th initial allocation of mission rductions btwn IR and NIR sctors such as dfind in th NAP. In this contxt, th rgulator facs a grav information problm: To dsign an fficint NAP allowing for qualisation of marginal abatmnt costs across ths sctors and countris th rgulator nds to know th (marginal) abatmnt costs of th NIR sctors in all countris. Furthrmor, h has to anticipat th markt pric for pollution prmits in ordr to assign th fficint amounts to th national NIR sctors. Thrfor on sourc for infficincis in abatmnt may alrady b th allocation of th missions budgt to th sctors not subjct to missions trading according to th NAP Macro plan for sctors and branchs. In trms of th cost functions in Figur 1 this mans, th distribution of missions (or abatmnt) btwn th sctors IR and NIR can caus th total marginal abatmnt costs to dviat from th aggrgatd optimum C ( ). Obviously, at th macro lvl thr is a trad-off btwn issus of min comptitivnss, fficincy, and compnsation in dsigning NAPs (s Böhringr and Lang, 2005a, or Böhringr t al., 2005, for a comprhnsiv analysis). Howvr, dspit th cntral rol that th NIR sctors (in particular privat housholds and transport) play for th cost fficincy of th Europan CO 2 rduction policy, in most mmbr stats th discussion in th run-up to th NAPs was dominatd by th burdn on th IR sctors. According to this, many analyss conclud that th actual NAP Macro plans contain a too gnrous allocation to th IR sctors, lading to comparativly high rduction rquirmnts on th part of th NIR sctors whil laving th IR sctors with a comparativly lax rduction rquirmnt (.g. Btz t al., 2004). 4
2.3. Taxs and EU ETS with idntical scop in all EU mmbr stats W bgin with th assumption that th NIR sctors ar rgulatd sparatly (.g. by countryspcific mission taxs), whil th IR sctors in all EU mmbr stats ar comprhnsivly subjct to missions trading and additional uniform mission taxs at th sam tim. This cas of two ovrlapping mission rgulating rgims with idntical scop is wll known in thory and turns out to b rlativly straightforward: Th mittrs s both, th pric of th prmits and th tax, and abat until thir marginal abatmnt cost quals th mission tax plus th prmit pric. Thrfor this rgim dos not diffr substantially from th cas whr only prmits ar issud (or only a tax is lvid such that th givn lvl of missions is implmntd) unlss th tax is high nough such that th amount of prmits is not binding. Th prmit pric would thn b zro and th missions would b blow th lvl targtd by th numbr of prmits. In this cas th tax could also b viwd as a backstop to nsur a minimum of abatmnt fforts. Thrfor, unlss both instrumnts ar binding, tax and missions trading rgims with idntical scops do not ncssarily coincid with fundamntal fficincy problms. But it complicats th rgulation and is thrby likly to incras th transaction costs. Most of all, it has no cological ffct sinc th tax is compnsatd by a lowr prmit pric unlss th pric is zro and th prmit schm bcoms dispnsabl. 2.4. Additional taxs within th IR sctors in only svral EU mmbr stats Suppos again that th NIR sctors ar rgulatd sparatly. Howvr, in contrast to th prvious sction and mor closly mirroring rality, it is now assumd that th tax systm is not introducd EU-wid but only in on or svral countris. Actually, svral EU countris hav alrady introducd CO 2 or nrgy taxs during th last yars. In Grmany, for xampl, both th NIR and th IR sctors ar subjct to an nrgy tax with rducd tax rats for th IR sctors (s BMU 2003). First of all, it is should b sn clarly that an additional tax has no cological ffct itslf. This holds dspit th first of th two argumnts mntiond in th introduction. Rcall, 1. A tax would hlp to bring down th missions. This would b cologically bnficial and hlp to rach th mission targt of th EU burdn sharing agrmnt. Hr, it is hlpful to commnt on two aspcts. First, onc th NAPs ar stablishd, th mission budgt is allocatd to th IR sctors. Any rduction fforts in th IR sctors thraftr lad to prmit transfrs but thy ar not rlvant to th targt formulatd in th EU Burdn Sharing Agrmnt. Scond, on of th major strngths of a prmit trading systm is its cological ffctivnss as it is a quantity-basd instrumnt. Any pric basd mchanism 5
within this rgim will not altr th cological ffctivnss. Additional taxs in som countris within th trading schm will lad to prmit xport to othr rgions of th trading schm whr no tax is applid. Th ovrall cological ffct will b zro, unlss th tax is sufficintly high and applid in a sufficint numbr of countris such that th prmit pric is drivn down to zro and th trading schm is not binding anyway. Th tax acts as a lowr bound to th abatmnt xpnditurs. Furthrmor, starting from an EU wid fficint allocation of mission prmits with qualisd marginal abatmnt costs, it is obvious that introducing an additional mission tax in th IR sctors in a singl country would typically produc infficincis (in Figur 1, th rsulting total marginal abatmnt cost curv would dviat from th minimum cost curv). Th tax drivs marginal abatmnt costs apart and introducs infficincis on both th mmbr stat and EU lvl. Th opn qustion is whthr rsults may chang whn two spcific cass ar considrd: ovr-allocation of prmits to th IR sctors (Sction 2.4.1) and markt powr of th country that introducs a tax to th IR sctors in addition to EU ETS (Sction 2.4.2). 2.4.1. Assuming ovr-allocation to th IR sctors Now assum an ovr-allocation of mission prmits to th sctors subjct to th missions trading schm in a country. Ovr-allocation to th IR sctors is dfind by an infficint allocation of mission rductions btwn th NIR and IR sctor, implying highr marginal abatmnt costs for th NIR sctors than for th IR sctors (th lattr ar xprssd by th x-post prmit pric). Rmmbr th scond argumnt formulatd in th introduction: 2. Th marginal abatmnt costs (in th IR sctors) ar by dfinition too low in cas of an ovr-allocation of prmits. A tax would bring th marginal abatmnt costs in th IR sctors closr to th xtrnal costs and closr to th fficint lvl. Th tax could thrfor b considrd as a scond-bst instrumnt to incras th fficincy of th national or EU wid abatmnt. Considr an opn conomy rprsntd by th marginal abatmnt cost function of thir IR and NIR sctors. According to th assumption of an ovr-allocation th national mission budgt is allocatd to both sctors such that th marginal abatmnt costs in th NIR sctors, τ N, xcd th prmit pric on th Europan markt, p. For now, assum that th prmit pric is not significantly influncd by th domstic action (small country assumption). This situation is illustratd in Figur 2. 6
C C ( N ) C ) C ( ) N ( min τ N p τ IR N tax tax Figur 2: Unilatral mission tax in th IR sctor. Th dark shadd triangl marks th fficincy costs. In addition to th prmit rgim th IR sctors ar confrontd with a taxτ IR. This lifts up th incntiv to rduc missions within th taxd rgion such that th marginal abatmnt costs of th IR sctors will qual p + τ. For a small opn conomy ths additional IR abatmnt fforts will lav th prmit pric unchangd. Th shadd aras in Figur 2 illustrat th tax-inducd fficincy ffcts. On th on hand, rducing missions from to tax will lad to a surplus du to th sal of prmits (light shadd rctangular ara). For rasons of simplicity, w assum that th us of ths rvnus at th firm lvl dos not hav any fficincy implications. On th othr hand th additional abatmnt costs ar givn by th ara undr th abatmnt curv (dark triangl plus light shadd rctangl). Pr saldo, th tax lads to additional fficincy costs, rprsntd by th dark shadd triangl. Th EU ETS implmnts any givn EU-wid targt for th IR sctors at minimum costs indpndnt of whthr th country-spcific NAP lads to ovr-allocation or not. An additional tax within th trading schm cannot chang th distribution btwn th IR and NIR sctors x post. It brings th marginal costs closr to what would hav bn th optimal lvl across all mittrs in th EU givn that th NAPs had bn dsignd in a cost-fficint mannr, but only in th rgion whr th tax is applid. u to th strict division of th IR 7
and NIR parts of th conomy, taxs do not act as an instrumnt to implmnt a scond-bst solution. Thrfor, a unilatrally introducd tax drivs apart th marginal abatmnt costs within th IR sctors of th diffrnt rgions and lads to fficincy losss. To conclud: An additional tax on firms of th IR sctors in a singl mmbr stat (which dos not hav any markt powr on th EU prmit markt) is costly for th mmbr stat, incrass th EU ovrall implmntation costs of th missions targt and has no cological ffct. 2.4.2. Assuming a larg conomy with markt powr W now rlax th assumption of a small mmbr stat. This mans that th tax-inducd incras of prmit supply on th EU prmit markt will lowr th EU-wid prmit pric. As in th cas of a small opn conomy, th additional tax will driv apart th marginal abatmnt costs within th IR sctors of th diffrnt countris and thrfor gnrat a cost burdn for th EU as a whol. A larg opn conomy may undr spcific assumptions bnfit from th introduction of a small tax. 2 If th mmbr stat imposing th tax (i) is a nt prmit importr (bfor and aftr implmnting th tax), (ii) has rlativly flat marginal abatmnt costs in th IR sctors and (iii) gnrats a sufficint amount of mission rduction by th tax so that th markt pric for prmits falls substantially, thn th lowr import pric for prmits and th rducd amount of prmits to b importd can compnsat for th incrasd abatmnt fforts such that th country can mak a nt profit from introducing th tax. 3 Figur 3 illustrats this point. Th x ant prmit allocation to th IR sctors is dnotd by alloc. Bfor introducing th tax th mission lvl is such that th marginal abatmnt costs of th IR sctor qual th prmit markt pric p, i... Th introduction of a taxτ IR will act as an additional rduction incntiv on top of th prmit pric, th missions fall to tax and th lowr prmit dmand causs th pric to fall to th aftr-tax lvl p tax. 2 This ffct is wll known in th litratur on intrnational trad, whr an import duty in a big country may lowr domstic import dmand and world markt prics, lading to a nt gain for th country ( bggar-thynighbour-policy ). 3 Not that it is undtrmind whthr th sctor itslf bnfits. Th way th tax rvnus ar rcycld can hav an impact on th ovrall costs of a rgulation as is dmonstratd in Gouldr (1995) or Bovnbrg (1999). 8
C IR C ( ) X τ IR p p tax Y C A B B alloc tax Figur 3:A larg nt prmit importr may bnfit from a unilatral mission tax. IR W now compar th two situations bfor and aftr th tax is imposd. Bfor, th import xpnditurs of th country ar illustratd by th rctangular ara ABC. Th incrasd rduction ffort and th lowr prmit pric aftr th tax is introducd impos (abatmnt) costs on th IR sctor qual to th ara B BCX. On th othr hand th import xpnditurs ar rducd, and th nw xpnditurs ar rprsntd by th ara AB Y. Thrfor, whthr th country bnfits from th tax or not is now dtrmind by th dark and th light shadd aras. If th light shadd rctangular ara xcds th dark shadd triangular ara as is th cas in Figur 3 thn th country bnfits from th tax in th IR sctors. It is a mattr of quantitativ analysis whthr th conditions ar mt by any EU mmbr stat. Howvr, th conditions appar rstrictiv sinc countris with flat (marginal) abatmnt cost functions tnd to turn out prmit xportrs rathr than importrs. Morovr, it is qustionabl whthr any of th EU mmbr stats has a sufficint markt shar to substantially driv down th prmit pric by unilatral action. Grmany, for xampl, as th biggst markt, rprsnts roughly a quartr of th whol Europan prmit markt. 3. Quantitativ illustration for th EU 15 In this sction, th ffcts of th hypothtical introduction of an additional CO 2 tax in slctd countris will b analysd using a simpl numrical partial quilibrium modl of th EU carbon markt. Aftr a brif dscription of th modl paramtrisation in Sction 3.1 simulation rsults of a st of scnarios will b prsntd and discussd in Sction 3.2. 9
3.1. Prmit markt modl Th modl usd includs th EU 15 carbon markt. All mmbr stats ar rprsntd by sparat marginal abatmnt cost curvs for th sctors covrd by th missions trading dirctiv (IR) and th sctors not covrd (NIR). Marginal costs of mission abatmnt may vary considrably across countris and sctors du to diffrncs in carbon intnsity, initial nrgy pric lvls, or th availabl carbon substitution possibilitis. Basically, th drivation of continuous marginal abatmnt cost curvs rquirs a sufficintly larg numbr of discrt obsrvations for marginal abatmnt costs and th associatd mission rductions. Ths data can b gnratd by tchnologyorintd partial quilibrium modls of th nrgy systm (such as th POLES modl by Criqui and Mima, 2001, or th PRIMES modl by Capros t al., 1998) or by computabl gnral quilibrium (CGE) modls (s.g., Eyckmans t al., 2001). In this papr, a rducd form of complx CGE intractions in trms of marginal abatmnt cost curvs is gnratd. Strictly spaking, marginal abatmnt cost curvs for th IR and NIR sctors across EU countris ar drivd from th PACE modl - a standard multi-rgion, multi-sctor CGE modl for th EU conomy (for a dtaild algbraic xposition s Böhringr, 2002). PACE is basd on rcnt consistnt accounts of EU mmbr stats production and consumption, bilatral trad and nrgy flows for 1997 (as providd by th GTAP5-E databas s imaranan and Mcougall, 2002). With rspct to th analysis of carbon abatmnt policis, th sctors in th modl hav bn carfully slctd to kp th most carbon-intnsiv sctors in th availabl data as sparat as possibl. Th nrgy goods idntifid in th modl includ primary carrirs (coal, natural gas, crud oil) and scondary nrgy carrirs (rfind oil products and lctricity). Furthrmor, th modl faturs thr additional nrgy-intnsiv non-nrgy sctors (iron and stl, papr, pulp and printing, non-frrous mtals) whos installations in addition to th scondary nrgy branchs (rfind oil products and lctricity) ar subjct to th EU missions trading systm. Th rmaining manufacturrs and srvics ar aggrgatd to a composit industry that producs a non-nrgy-intnsiv macro good, which togthr with final dmand capturs th activitis (NIR sgmnts) that ar not includd in th EU trading systm. To gnrat th rducd form modl, a squnc of carbon tax scnarios for ach rgion is prformd whr uniform carbon taxs (starting from 0 to 200 pr ton of carbon in stps of 1 ) ar imposd. Th outcom is a larg numbr of marginal abatmnt costs, i.., carbon taxs, and th associatd mission rductions in IR and NIR sctors. Thn a last-squar fit by a polynomial of third dgr is applid (s Böhringr t al., 2005). 10
Whn prsnting th following simulation rsults on has to kp in mind that th quantitativ modl rsults mainly dpnds on th spcification of th (marginal) abatmnt cost curvs togthr with th ffctiv rduction rquirmnts for IR and NIR sctors. Furthrmor, it should b considrd, that markt intractions and spillovr ffcts ar nglctd. Apart from trms-of-trad ffcts, othr potntially important gnral quilibrium intractions concrn rvnu-rcycling. This is, howvr, not th subjct of this analysis. For th prsnt illustrativ purposs whr markt shars and th shap of marginal abatmnt cost functions mattr most, a partial prmit markt modl sms to b justifid. 3.2. Scnario for implmnting additional CO 2 taxs in th IR sctor This sction illustrats th ffct of a unilatral tax on th missions of th IR sctors in Grmany, a country with a comparativly larg markt shar on th CO 2 markt. Th initial allocation of th mission prmits rprsnts a homognous 90 prcnt of th baslin missions to th IR sctors of all mmbr stats. By doing this th abatmnt rquirmnt in th NIR sctors is sufficintly strict such that th ncssary abatmnt would lad to marginal abatmnt costs substantially abov th prmit pric on th markt (this was dnotd as an ovr-allocation to th IR sctors, s Sction 2.4.1). Thn a unilatral mission tax in th IR sctors in Grmany is introducd which incrass stpwis from 0 to 5 pr ton of CO 2 missions. Figur 4 shows, as was xpctd from th analysis in th prvious sction, that th ovrall EU-wid complianc costs incras with th tax rat. Thrfor, dspit th prsumd ovrallocation, unilatral taxs which mov th marginal abatmnt fforts in Grmany closr to what thy would hav bn in th optimum do hav a ngativ ffct on th EU as a whol. 11
Figur 4: Excss cost burdn of a unilatral tax on th EU as a whol. Figur 5: Rsulting prmit pric and marginal abatmnt costs if a unilatral tax is introducd. 12
Individual countris could on th othr hand bnfit from th unilatral tax in Grmany. This is du to th falling prmit pric that is inducd by th incrasd supply from th additional Grman abatmnt. Figur 5 shows th prmit pric (falling curv) and th marginal abatmnt fforts in Grmany composd of th markt prmit pric plus th mission tax (rising curv). Th falling prmit pric can b xpctd to bnfit th nt importrs of prmits. This xplains that som countris may bnfit, whil othrs los from th unilatral tax. Figur 6 illustrats th complianc costs of diffrnt countris with a rising unilatral tax in Grmany. Not, that th incras of th national complianc costs rlativ to th cas without a unilatral mission tax can b largr for a country that xports prmits than for th country that introducs th tax itslf. Figur 6: Cost implications on th sparat mmbr stats. 4. Concluding rmarks This papr shds som light on th dbat of th appropriat instrumnt mix in climat policy. In particular it analyss whthr th Europan Union missions trading schm should b 13
complmntd by additional mission taxs. It turns out that combinations of a quantity basd instrumnt such as a prmit trading schm with a pric basd instrumnt such as mission taxs (or, in principl, subsidis) hav to b dsignd carfully. For th Europan prmit trading rgim on has to tak into account som of its spcific dsign issus. Ths ar (i) th limitd scop of th prmit trading schm and (ii) th multi-jurisdictional framwork govrning parts of th trading schm and most of th rgulation complmntary to th prmit trading. Emission taxs ar typically a mattr of th mmbr stats lgislation. Any additional rgulation within th sctors covrd by th missions trading dirctiv has to considr that it will act within an cologically ffctiv rgim, maning that th ovrall missions ar dtrmind by th amount of prmits on th markt. Onc th initial allocation of th prmits is fixd, additional rgulation within th trading sctors can hav distributional ffcts or it can altr th prmit pric, but it will not chang th ovrall missions. Th only xcption would b a prmit pric drivn down to zro in cas of comprhnsiv and sufficintly high mission taxs. Emission taxs can b usd in an EU wid fficint mannr to rgulat th missions in th sctors that ar not covrd by th trading schm. Whn applid within th trading sctors, howvr, unilatral taxs in mmbr stats incras th EU ovrall costs of implmnting a givn missions targt. Typically such taxs also incras th costs for th country that introducs th tax sinc th additional abatmnt xpnditurs xcd th rvnus gnratd from th mission prmits mad availabl by th rduction. Th cas of an additional unilatral tax in th missions trading sctors in Grmany was illustratd using a simpl partial markt modl for missions trading. In th modl a fictitious unilatral tax on th IR sctors in Grmany of 5 pr ton of CO 2 would incras th complianc cost in th whol EU by about half a prcnt. For individual countris th ffct rangs from incrasing th costs by about 15 prcnt to rducing complianc costs by about 10 prcnt rlativ to th cas without a tax. Obviously, an xportr dos not profit if h introducs a unilatral tax which brings down th prmit pric. On th othr hand unilatral mission taxs can b bnficial for th country introducing th tax whil still bing costly for th rst if th country has a larg shar in th prmit markt, has comparativly flat marginal abatmnt costs in th sctors subjct to missions trading, and is at th sam tim a nt prmit importr. Th rason is that th rducd domstic prmit dmand can lowr th markt pric for prmit and thrby lowr th country s import xpnditurs. 14
Thrfor taxs within th part of th conomy that is rgulatd by th prmit trading systm should b handld with grat car. If thy ar usd thy should b justifid by othr rasons than implmnting th commitmnts of th first Kyoto priod in a cost fficint mannr. Rfrncs Btz, R., W. Eichhammr, and J. Schlich (2004). signing National Allocation Plans for EU Emissions Trading - A First Analysis of th Outcom, Enrgy and Environmnt, 15 (3), 375-423. BMU (2003). i Ökologisch Sturrform: Einstig, Fortführung und Fortntwicklung zur ökologischn Finanzrform, Brlin. Böhringr, C. (2002). Industry-lvl Emission Trading btwn Powr Producrs in th EU, Applid Economics, 34 (4): 523-533. Böhringr, C. and A. Lang (2005). Mission Impossibl!? On th Harmonization of National Allocation Plans undr th EU Emissions Trading irctiv, Journal of Rgulatory Economics, 27 (1): 51-71. Bovnbrg, A.L. (1999). Grn Tax Rforms and th oubl ividnd: An Updatd Radr's Guid, Intrnational Tax and Public Financ, 6: 421-443. Capros, P., L. Mantzos,. Kolokotsas, N. Ioannou, T. Gorgakopoulos, A. Filippopoulitis, and Y. Antoniou (1998). Th PRIMES Enrgy Systm Modl Rfrnc Manual, National Tchnical Univrsity of Athns. ocumnt as pr rviwd by th Europan Commission, irctorat Gnral for Rsarch. Criqui, P. and S. Mima (2001). Th Europan Grnhous Gas Tradabl Emission Prmit Systm: Som Policy Issus Idntifid with th POLES-ASPEN Modl, ENER Bulltin, 23: 51-55. imaranan, B. and R.A. Mcougall (2002). Global Trad, Assistanc and Production: Th GTAP 5 ata Bas. Wst Lafaytt, Cntr for Global Trad Analysis. Purdu Univrsity. EU (2003a). irctiv Establishing a Schm for Grnhous Gas Emission Allowanc Trading within th Community and Amnding Council irctiv 96/61/EC. Europan Commission, Brussls. Availabl at: http://uropa.u.int/ur-lx/fr/com/pdf/2001/ fr_501pc0581.pdf EU (2004). irctiv 2004/101/EC, amnding irctiv 2003/87/EC stablishing a schm for grnhous gas mission allowanc trading within th Community, in rspct of th Kyoto Protocol s projct mchanisms. Europan Comission, Brussls. Availabl at: http://uropa.u.int/comm/nvoirnmnt/climat/mission/pdf/dir_2004_101_n.pdf 15
Eyckmans, J.,. van Rgmortr, and V. van Stnbrgh (2001). Is Kyoto Fatally Flawd? An Analysis with MacGEM, Working Papr Sris Faculty of Economics, No. 2001-18, Catholic Univrsity of Luvn, Luvn. Gouldr, L.H. (1995). Environmntal Taxation and th oubl ividnd: A Radr's Guid, Intrnational Tax and Public Financ, 2: 157-183. Johnston, N. (2003). Th Us of Tradabl Prmits in Combination with Othr Environmntal Policy Instrumnts, Rport ENV/EPOC/WPNEP(2002)28/Final, OEC, Paris. Klppr, G. and S. Ptrson (2002). On th Robustnss of Marginal Abatmnt Cost Curvs: Th influnc of World Enrgy Prics, Kil Working Papr No. 1138. Institut for World Economics, Kil. Sorrll, S. and J. Sijm (2003). Carbon Trading in th Policy Mix, Oxford Rviw of Economic Policy 19 (3), 420-437. 16