POLICY RESEARCH WORKING PAPER 1315 Global Tradable Carbon Ag'obaItadablrmit regime designed to minimize Permits, Participation the costs of stabilizngworid, * 1 r ~~~~~~~~~~~~- -carbon enMsions from fossil Incentives, and Transfers .boeissions 19am -fuel combustion at 1987 l1evels-bytheyear2000- BjonzLarsen: - : :-- B and to be acceptbe to both Ar.war Shah -ac n -P- A'ZWT Shh OECD and non-OECO -countries.- The World Bank Policy Research Department: Public Economics Division June 1994 I POLICY RESEARCH WOKKING PAPER 1315 Summary findings Most OECD countries have committed themselves to To encourage participation, it is proposed that each stabilizing their carbon emissions at 1990 levels by the non-OECD country be allocated permits equivalent to its year 2000, and some to reducing emissions to 80-90 projccted baseline emissions - and that OECD countries percent of 1990 levels by the years 2005 or 2010. be allocated permits equivalent to the world emissions Most non-OECD countries are reluctant to reduce target minus the permit allocations to the non-OECD emissions to combat global climate change. They argue countries. Such a scheme recognizes that OECD that suck policies would forestall their development, that countries have a higher willingness to pay for increasing the stock of greenhouse gases in the air is primarily from reductions and that non-OECD countries have a smaller historical emissions from OECD countries and the historical "global emissions debt.' former Soviet Union, and that those countries should be Under that proposed regime, Larsen and Shah find chat made to bear the cost of such abatement policies. the cost of emissions reductions for OECD countries Larsen and Shah evaluate alternative permit allocations would be about 50 percenr lower than unilateral for a global tradeable permit regime designed to reductions would be, and that non-OECD countries minimize the costs of stabilizing world carbon en-issions would also realize substantial net gains from participating from fossil fuel combustion at 1987 levels by the year in such a global treaty. 2000. Moreover, that global treaty would be 68 percent less They find that an important cross-section of countries cosrly worldwide than would realizing the same target would have little incentive to participate in a treaty based through unilateral reductions by the OECD countries. on such widely discussed forms of permit allocations as permit allocations by GDP, by population, or by some combination of the two. This paper - a produca of the Public Economics Division, Policy Research Department - is part of a larger effort in the department to examine economic policy instruments for global environmental protection. The study was funded by the BAnk's Research Support Budget under the research project "World Energy Subsidies" (RPO 677-28). Copies of this paper are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Carlina Jones, room N10-063, extension 37754 (21 pages). June 1994. The Policy Resarh Workig Paper Series dssemiates the rfdn of unit i prss to enwcage the cxhangc of idas about development issua An objective of theseries isto gntte findgs out quiy, even if the presnstations ar kss than fulty polisheS The papers carry the names of the audhors and shoud be used and cited accordingly. 7Th finings. nterpretatons, and cordusions are the auhors' oww and shodd not be attributed to the World Bank, its Excutive Board ofD;recrors, or any of its inerber counties Produced by the Policy Research Dissemination Center Global Tradeable Carbon Permits, Participation Incentives and Transfers by Bjorn Larsen, World Bank and Anwar Shah, World Bank The authors are grateful to an anonymous reviewer and the editor for helpful comments on an earlier version of this paper. Global Tradeable Carbon Permits, Participation Incentives and Transfers 1. Introduction This paper examines the feasibility of alternative tradeablc permits regimes to stabilize global carbon dioxide emissions at their 1987 level by the year 2000.' An optimal policy response to global climate change would be a cooperative agreement in which each country reduces emissions until the marginal costs of reductions for each country are equal to the sum of marginal benefits of reductions across all countries (Hoel 1990, 1991a,b). However, reliable estimates of marginal benefits of greenhouse gas reductions are simply not available, and therefore, one can only strive for cost efficiency in meeting emissions targets. Some countries may consider the costs of greenhouse gas reductions to exceed potential benefits and therefore may not be willing to participate in a treaty unless suitably compensated. Furthermore, a treaty must also address the flee rider problem. Most OECD countries have committed to stabilize their carbon emissions at 1990 levels by year 2000, and some to reduce emissions to 80-90% of 1990 levels by year 2005 or 2010. However, most non-OECD countries are reluctant to reduce enmssions to combat global climate change arguing instead that such policies would forestall their development and tha the stock of greenhouse gases in the atmosphere is prmarly due to historical emissions from OECD countries and the former Soviet Union and it is they who should be made to bear the costs of such abatement policies. This paper evaluates alternative permit allocations for a global tradeabte permit regime that mimizes costs of stabilizing world carbon emissions from fossil fuel combustion at 1987 levels by the year 2000. Particular attention is paid to the issue of inducing broad participation in a treaty. Alternative 'According to Nordhaus (1991a). carbon dioxide emissions are responsible fbr about 80% of global wanning potntial associated with the accumulation of all greenhouse gases in the atmosphere. permit allocations are evaluated in terms of costs or benefits of participating in a global permit system as well as efficiency gains of inducing broader participation. The paper is organized as follows. Section 2 is concerned with current and projected carbon emissions. Section 3 details marginal cost functions and determination of equilibrium pennit prices. Section 4 evaluates the costs and benefits of alternative permit allocations and incentives to participate inan intemational treaty. A final sectionsummarizes main conclusions. 2. Currea and Projected Carbon Emissions The sample countries were selected on the basis of their share of world emissions, per capita emissions, and/or their emissions per dollar of GDP in order to have a diverse representation. Nine lower income, five middle income, four emerging market economies, and all twenty-four OECD) countries are included (see Table 1)2. The sample countries accounted for almost 90% of world carbon emissions from fossi! ftiel Unibustion in 1987. The shares of world emissions from the OECD countries and the non-OECD sample countries were about 46% and 43 % respectively. The remaining non-OECD countries are aggregated and presented as the "rest of world". The share of world emissions is as high as 10% for China, 18% for the former Soviet republics (12% for Russia), and 22% for the United States. Per capita emissions vary from 30 kg in Bangladesh to 593-0 kg in Luxembourg. Emissions per dollar of GDP vary from 0.06 kg in Switzerland to 2.06 in North Korea. Projections of carbon emissions to the year 2000 for the same sample counties are presented in Table 2. Total world carbon emissions from fossil fuel combustion are projected to be 36% higher in year 2000 relative to the 1987 emission level. This implies an average annual growth rate of 2.4%. The share of world carbon emissions for the OECD countries are relatively unchanged -from 1987. However, China's share of world carbon emissions are projected to increase from 10% to 14V%. Emissions shares 2Some of these countries no longer exist, i.e. the Soviet Union, Yugoslavia and Czechosic*P-kia. 2 Table 1. Carbon Emissions and GDP Daa For Selected Countries - 1987 GDP per Carbon Perent of Carbon Cubon capita emissions world emissions enissions (US S) (00C tons) enilssions to GDP per capita (kg/USS) (kg/cap) BANGLADESH 166 3155 0.06% 3.18 30 NIGERIA 229 8962 0.16% 0.37 84 CHINA * 286 569900 10.38% 1.87 533 INDIA 322 145400 2.65% 0.57 182 PAJISTAN 325 13154 0.24% 0.39 128 INDONESIA 443 26258 0.48% 0.35 153 ZIMBABWE 598 4160 0.08% 0.77 463 EGYPT 709 19037 0.35% 0.54 380 NORTH KOREA 889 39203 0.71% 2.06 1834 MEXCO 1715 77200 1.41% 0.55 943 BRAZIL 2145 50300 0.92% 0.17 356 SOUTH AFRICA 2493 75900 1.38% 0.92 2292 VENEZUELA 2629 23307 0.42% 0.49 1276 SOUTH KOREA 3121 44900 0.82% 0.34 1067 POIAIND 1700 125700 2.29% 1.96 3338 YUGOSLAVIA 2700 32339 0.60% C.52 1403 USSR 2900 1012900 18.45% 1.23 3578 CZECHOSLOVAKIA 24W 64000 1.17% 1.71 4110 SWIZERLAND 26115 10330 0.19% 0.06 1580 ICELAND 21873 4S1 0.01% 0.09 M9S5 NORWAY 19963 8580 0.16% 0.10 2D48 DENMARK 19830 16610 0.30% 0.16 3238 SWEDEN 19257 15150 0.28% 0.09 1812 GERMANY 16754 26670D 4.86% 0.20 3427 FINLAND 18070 14420 0.26% 0.16 2925 LUXEMBOURG 16331 2200 0.04% 0.36 5930 FRANCE 15913 91000 1.66% 0.10 1636 AUSTRIA IS441 13000 0.24% 0.11 1717 THE NETHERLANDS 14521 35600 0.65% 0.17 2428 BELGIUM 14457 26000 0.47% 0.18 2637 ITALY 13176 97000 1.77% 0.13 1691 UNITED KINGDOM 12024 154000 2.80% 0.23 2707 IRELAND S353 7540 0.14% 0.25 2120 SPAIN 7452 43600 0.79% 0.15 1123 GREECE 4619 14370 0.26% 0.31 1437 PORTUGAL 3612 7700 0.14% 0.21 758 TURKEY 1293 34150 0.62% 0.50 649 JAPAN 19437 237100 4.32% 0.10 1942 UNITE STATES 18434 1246100 22.69% 028 5112 CANADA 16056 109100 1.99% 0.26 4221 AUSTRALIA 11364 63900 1.16% 0.35 3932 NEW ZEALAND 10749 S700 0.10% 0.16 1709 OECD 2520331 45.90% 0.20 3015 Rest of World 634394 1.1.55% 0.55 639 World 54910W 100.00% 0.34 1112 Sources: World Resource Instite; Wodd Bank. 3 Table 2. Projecdons of Carbon Emnission for Selected Countres - Year 2000 Carbon Percent of Carbon Carbon emissions world emissions emissions (000 tws) emissions per capita to GDP (kg/cap) (kgIUS$) BANGLADESH 8078 0.11% 58 0.29 NIGERIA 13161 0.17% 82 0.44 CHINA 1074631 14.28% 839 1.46 INDIA 291640 3.88% 294 0.60 PAKISTAN 24804 0.33% 161 0.39 INDONESIA 41066 0.55% 195 0.29 ZIMBABWE 5048 0.07% 397 0.66 EGYPT 40605 0.54% 642 0.61 NORTH KOREA 44617 0.59% 1614 2.06 MEXICO 113371 1.51% 109o 0.55 BRAZIL 89141 1.18% 506 0.18 SOUTH AFRICA 104629 1.39% 2351 0.98 VENEZUELA 30150 0.40% 1244 0.52 SOUTH KOREA 90059 1.20% 1905 0.28 POLAND 125700 1.67% 3210 1.73 YUGOSLAVIA 32839 0.44% 1298 0.46 USSR 1012900 13.46% 3441 1.08 CZECHOSLI;VAKIA 64000 0.85% 3953 1.50 SW_rZERLAND 13363 0.18% 1866 0.05 ICELAND 622 0.01% 2401 0.08 NORWAY 11099 0.15% 2515 0.09 DENMARK 214B7 0.29% 4081 0.14 SWEDEN 19598 0.26% 2169 0.08 GERMLANY 345005 4.59% 4376 0318 FINLAND 18654 0.25% 3639 0.14 LUXEMBOURG 2846 0.04% 7474 0.32 FRANCE 117718 1.56% 2009 0.09 AUSTRIA 16817 0.22% 2109 0.10 THE NETHERLANDS 46052 0.61% 2832 0.15 BELGIUM 33634 0.45% 3324 0.16 ITALY 125480 1.67% 2159 0.11 UNIT KINGDOM 199215 2.65% 3412 0.20 IRELAND 9754 0.13% 2638 0.22 SPAIN 56401 0-75% 1434 0.13 GREECE 18589 0.25% 1835 0.27 PORTUGAL 9961 0.13% 968 0.18 TURKEY 44177 0.59% 657 0.44 JAPAN 306714 4.08% 2416 0.09 UNITE STATES 1611963 21.43% 5886 0.24 CANADA 141132 1I8-% 4922 0.23 AUSTRALIA 82661 1.10% 4301 0.30 NEW ZEALAND 7374 0.10% 1993 0.14 ToM! OECD 3260317 43.34% 3611 0.17 ROW 1056313 14.04% 823 0.55 Wodd 7523069 100.00% 1254 0.31 ScuTe: Model Caluations. 4 in most of the other low and middle income countries are also projected to increase significantly. Former Soviet republics' share is projected to decline from 18% to 13% as a result of economic restructuring. Emissions shares are also projected to decline in the eastern european countries. Carbon emissions per capita are projected to increase from 1112 kg to 1254 for the world as a whole, while emissions per dollar of GDP are projected to decline from 0.34 to 0.31 as a result of increased efficiency in energy use. 3. Marginal Cost Functions and Equilibrium Permit Price The purpose of a tradeable emissions permit regime is to limit aggregate emissions to a targeted level. We consider the year 2000 to which carbon emissions are projected as a baseline scenario of business as usual. The equilibrium permit price in year 2000 is determined by the aggregate permit level (world emissions in 1987) and the marginal cost function of emission reductions in each country3. At the present time, country specific marginal cost finctions .re not available for most non-OECD countries. Even for OECD countries estimated cost functions are mostly based on different models and assumptions. We have therefore chosen to apply the following marginal cost fimction (assuming zero intercept) derived by Nordhaus (199la,b) based on a survey of cost studies in OECD countries, R = 1- e.4ulomc (1) where R is percentage emissions reductions and MC is marginal costs of reductions (US$Iton)4. It is plausible that costs of reductions are lower in countries that have done very little to improve energy efficiency- In particular, there may be net economic gains from initial carbon reductions in countries with significant fossil fuel subsidies (Churchill and Saunders 1991, Larsen 1993, Larsen and Shah 1994, Larsen and Shah 1992a,b, Shah and L2rsen 1992a) such as the former Soviet Union, China and 3See Kosobud et at (1991), Quinn et al (1992) and Soulh et al (1991) for a discussion and estimates of welfare gains of initial allocations of multiperiod undated permits to allow intcrtemporal efficiency. 'The studies upon which this margial cost function is based ignore possible benefits of local pollution reductions associated with reductions in carbon intensive energies such as fossil fuels. For further discussion and empirical evidence see Gbmsrod et al (1992), Shah and Larsen (1992a,b) and Summers (1991). 5 Venezuela. We assume that all countries by year 2000 would have reached the point at which marginal costs are non-negative, and therefore, marginal costs of reductions are zero for the first unit of carbon emission. Each country is assumed to reduce emissions until the marginal costs of reductions are equal to the emission permit price'. The equilibrium permit price, given the cost function in (1), that clears the permit market and thus limits total world emissions to the 1987 level is estimated to equal US $58 per ton of carbon and is independent of any permit allocation. 4. Costs of Alternative Permit Allocations An important aspect of alternative permit allocation schemes is the international support they may expect to receive. It is reasonable to assume that a treaty will be signed primarily by nations that expect positive net benefits from participation. In view of the scientific uncertainty, at present, it is extremely difficult to incorporate benefits from greenhouse gas emission reductions in terms of damages averted by delaying the threshold of global warming. Thus the approach taken in this paper is to assume that some countries have a positive willingness to pay for emissions reductions while other countries have small or zero willingness to pay. Permit allocations must then be such that the countries with negligible or zero willingness to pay are induced to participate in a treaty. Furthermore, countries with a positive willingness to pay must also gain from such an allocation compared to the case wit emissions reductions exclusively in the latter countries6. In particular, it is assumed that OECD countries have positive willingness to pay for emission reductions, while non-OECDs have negligible or zero willingness to pay. Each country will reduce emissions unfil marginal costs of reductions are equal to the permit price of US $58 per ton of carbon. If the number of emissions permits is less (greater) than baseline emissions minus reductions, the country will purchase (sell) permits for US $58 per ton of carbon. The net costs 5For cost efficient national policy responses, see Pearce (1991) and Poterba (1991). 'Bohm (1992) considers tde impact on negotiated allocations in the case of no permit trade versus the case of allowing permit trade. Odier effects of unilateral reductions such as increased emissions in counties not participating in a treaty due to shif5s in the marginal benefit curve (Hoe[ 1991b) or changes in comparative advantage are also ignored. 6 or benefits in US dollars of alternative permit allocations, with linear approximation of the marginal cost function over the relevant range, are given by, CJ = P [EJ' - (Ej - AR)] - 0.5PAER (2) where P' is the permit price, EJP is the allocated permit volume, EJ is the baseline 2000 emissions, and Ej' is emissions reductions in country j. The first term is the costs (-) or revenues (+) from pennit purchases or sales, and the second term is the cost of emissions reductions with EJ given by (1) multiplied by baseline 2000 emissions. Thus given the permit price, baseline emissions and cost function, the net costs or benefits are determined by the permit allocation scheme. The permit allocations considered are: (i) by population; (ii) by GDP; (iii) by a combination of population and GDP; (iv) to each non-OECD country equivalent to its projected baseline emissions and to OECD countries as a group equivalent to the world emissions target minus non-OECD permits; and (v) unilateral reductions by the OECD countries. Permit allocation by population: The permit allocation by population is calculated for the year 2000 by, vJ = Er ( POPj I POR. ) (3) where Ejl is emissions permits allocated to country j (units of thousand tons), El is world permit level (EP = Ej EJ'), POPj is the projected population of country j, and POP,, is the projected world population. El is in this case equivalent to 1987 world emissions. Permit allocation per capita is the same in all countries and is given by ErIPOPW = 0.915 tons of carbon. Permit allocations relative to projected baseline emissions in the year 2000 are presented in Table 3. The ratio is greater than one, i.e., allocations are larger than projected emissions, for all the low- income sample countries expect for North Korea. The ratio is less than one for all the middle-income, emerging market economies and OECD countries, except for Brazil and Turkey. OECD countries as a group would only receive permits equivalent to 25% of their projected emissionLs. 7 From (2) and (3) the net costs or benefits relative to GDP are given by, Cj/GDPj = P l/gdpj [EPIPOP. - (1 - 0.5R)E/POP1] where gdpj is per capita GDP for country j. Given the permit allocation per capita and (1), we bhve, Cj/GDPj = r l/gdpj [0.915 - 0.865E1/P0PJ1 (4) From (4), we observe that the break-even point is 1.058 tons of carbon baseline emissions per capita. The net benefits (net costs) are decreasing (increasing) in per capita baseline emissions and decreasing (decreasing) in per capita GDP. The net benefits and costs of this allocation in the year 2000 are presented in Table 4. Brazil, Portugal, Turkey and all the low-income sample countries, except for North Korea, would benefit from this permit allocation scheme. In particular, the net benefits for Bangladesh and Nigeria would be more than 25% of GDP in year 2000, and more than 11 % for Pakistan. From (4) and Table 2 we see that this is because of the low per capita baseline emissions and the low per capita GDP. The net costs would be as large as more than 5% of GDP :or Poland and Czechoslovakia as a result of high per capita baseline emissions and relatively low per capita GDP. The net costs to the 013OCD as a group would be more than 0.6 % of GDP. The net costs are high for OEC0D countries such as Luxembourg, Australia, the United States and Canada because their baseline per capita emissions are significantly higher than in most other OECD countries. Given the net costs involved for most of the middle-income countries and particularly the emerging market economies, these countries are unlikely to participate in a global treaty based on a permit allocation by population. The sample countries in these groups are projected to contribute more than 20% of world emissions in the year 2000, and therefore participationby these groups is Vital for an 8 Table 3. Rauc of Permt AIlomiGon o Carbon Emllon (Yar 2000) Allocation Allocauon Non-ORCD by bvy fally covent population ODP BANGLADESH 15.75 0.77 1.0] NIOBRIA 11.17 0.51 1.00 CHINA 1.09 0.15 1.00 INDIA 3.12 0.37 1.00 PAKISTAN 5.70 0.7 1.00 INDONESIA 4.70 0.78 1.00 ZIMBABWE 2.30 0.34 1.00 EGYPT 1.43 0.37 1.00 NORTH KOREA 0.57 0.11 1.00 MEXICO 0.83 0.41 1.00 BRAZIL 1.81 1.27 1.00 SOUTH AFRICA 0.39 0.23 1.00 VENEZUELA 0.74 0.43 1.00 SOUTH KOREA 0.48 0.79 1.00 POLAND 0.29 0.13 1.00 YUGOSLAVIA 0.71 0.49 1.00 USSR 0.27 0.21 1.00 CZECHOSLOVAKIA 0.23 0.15 1.00 SWITZERLAND 0.49 4.22 0.38 ICELAND 0.38 2.85 0.38 NORWAY 0.36 2.49 0.38 DENMARK 0.22 1.56 0.38 SWEDEN 0,42 2.71 0.38 GERMANY 0.21 i.25 0.38 FINLAND 0.25 1.58 0.38 LUXEMBOURG 0.12 0.70 0.38 FRANCE 0.46 2.48 0.38 AUSTRIA OA3 2.29 0.38 THE NETHERLANDS 0.32 1.52 0.38 BELGIUM 0.28 1.40 0.38 ITALY 0.42 1.99 0.38 UNTED KINGDOM 0.27 1.13 0.38 IRELA 0.35 1.00 0.38 SPAIN 0.64 1.69 0.38 GREECE 0.50 0.82 0.38 PORTUGAL 0.95 1.22 0.38 TURKEY 1.39 0.51 0.38 JAPAN 0.38 2.55 0.38 UNITED STATES 0.16 0.92 0.38 CANADA 0.19 0.97 0.38 AUSTRALIA 0.21 0.74 0.38 NEW ZEALAND 0.46 1.60 0.38 Total OECD 0.25 1.28 0.38 ROW 1.11 0.41 1.00 World 0.73 0.73 0.73 Note:: a: Non-OECD couniesare allocated emissionsperritsequivalentto project baseline emissions. OECDcountries as a group are allocated permits equivalent to the targeted world emissions less permits allocated to non-OECD countries. Each OECD country receives pcrmits in propordon to their projected baseline emissions. Source: Authors' calculations. 9 Table 4. Net Costs (-) or Benefits t+l as a Percentage of GDP of nitenwadve Permit Alocadons (Year 2000). Allocadon Allocation Non-OECD OECD by population by GDP fully covered unilatemi reductions BANGLADESH 25.38% -0.16% 0.23% NIGERIA 26.63% -0.93% 0.35% CHINA 1.92% -6.06% 1.15% INDIA 7.92% -1.73% 0.47% PAKISTAN 11.11% -0.68% 0.31% INDONESIA 6A1% -0.14% 0.23% ZIMBABWE 5.58% -2.04% 0.52% EGYPT 1.98% -1.75% 0.48% NORTH KOREA -3.58% -9.10% 1.62% MEXCO J.10% -1.46% 0.43% BRAZIL 0.97% 0.42% 0.14% SOUTH AFRICA -2.72% -3.63% 0.77% VENEZUELA -0.39% -1.30% 0A41X SOUTH KORFA -0.64% 4.13% 0.22% POLAND -5.83% -7.39% 1.36% YUGOSLAVIA -0.42% -0.99% 0.36.% USSR -3.79% 4.16% 0.85% CZECHOSLOVAIaA -5.56% -6.28% 1.18% SWIT2JUMAND, -0.12% 1.04% -0.15% -0.30% ICELAND -0.22% 0.91% -0.22% -0.44% NORWAY -0.26% 0.85% -0.26% -0.51% DENMARK -0-54% 0.58% -0.41% -0.81% SWEDEN -0.21% 0.89% 4.24% -0.47% GERMNY -0.69% 0.40% .0.51% -1.02% FiNLAND -0.51% 0-59% -OAI% -0.80% LUXEMBOURG -1.38% -0.30% -0.91% -1.81% FRANCE -0.22% 0.85% -0.26% -0.51% AUSTRIA -0.25% 0.82% -028% -0.55% THE NEITERLANDS -0.47% 0.57% -0.42% -0.83% BELGIUM -0.55% 0.50% -OA6% -0.91% rrALY -0.29% 0.74% -0.32% -0.64% UNITED KINGDOM -0.69% 0.31% *0.56% -1.12% IRELAND -0.68% 0.18% -0.64% -1.26% SPAIN -0.18% 0.64% -0.38% -0.75% GREECE -0.58% -0.07% -0.78% -1.55% PORTUGAL 0.09% 0.38% -0.53% -1.04% TURKEY 1.36% -0.92% -1.26% -2.50% JAPAN -0.25% 0.87% 4.25%S 4.50% UNITED STATES -1.01% 0.08% -0.70% -1.38% CANADA 4.92% 0.14% -0.66% -1.31% AUSTRALIA -1.16% -0.23% -0.87% -1.72% NEW ZEALAND -0.33% 0.60% -0.40% -0.79% Total OECD -0.62% 0.43% -0.50% -0.9% ROW 0.80% -1.47% 0.43% Wodd -024% -0.24% 4.24% Source: Author's calculatios. 10 efficient global treaty. Furthermore, several large OECD countries may also be reluctant to participate given that they experience net costs of about 1% of GDP. Permit allocation by GDP The permit allocation by GDP is calculated for the year 2000 by using the following expression, EjP=Ep ( GDPjI / C-DP, ) (5) where GDPj is country j's GDP in millions of US dollars, GDPW is aggregate world GDP in the same units in year 2000, and EP is equivalent to 1987 world emissions. Permit allocation per dollar of GDP is given by E/(GDPW = 0.23 kg of carbon. Permit allocation relative to projected baseline emissions in the year 2000 are presented in Table 3. The ratio is less than one for all the low-income sample countries in contrast to the case of permit allocation by population. It is also less tha one for the middle-income (except for Brazil) and the emerging market economies as in the case of allocation by population. Thus, the high income countries of the OECD end up with more pennits than the baseline emissions, except for Luxembourg, Australia, the Unted States and Canada, and the relatively lower income OECD countries such as Greece and Turkey. From (1), (2), (5) and permit allocation per dollar of GDP, we have, CiGDPJ = P' [0.23 - 0.865E/GDPj] (6) From (6) we see that the break-even point is 0.26 kg of carbon baseline emissions per dollar of GDP, and the net benefits (net costs) are decreasing (increasing) m baseline emissions per dollar of GDP. The net benefits of this allocation in the year 2000 are presented in Table 4. All the OECD countries, except for Turkey, Luxembourg, Australia and Greece, would benefit under this permit allocation. Even though the United States and Canada would receive less pernits than their baseline emissions, they would sfill benefit from this allocation beouse their baseline emissions per dollar of GDP are less than 0.26 kg (see Table 2). All the low-income, middle-income countries (except Brazil) and the emerging market 11 economies would have net costs from this allocation. In particular, the net costs would be large for North Korea, China, Poland, Czechoslovakia and USSR because of their high baseline emissions per dollar of GDP. Given the estimated high net costs, most non-OECD countries are unlikely to participate in a global treaty based on a permiit allocation by GDP. Non-OECD countries as a group are projected to contnrbute more than 56% of world carbon emissions in the year 2000 and therefore represent a group whose participation is vital for an efficient global treaty. Thus a permit allocation by GDP appears even less likely to induce broad participation than an allocation by population. Furthermore, negotiations regarding an allocation by GDP will also have to resolve the problematic issue of determining the level of a common currency GDP. A Dermit allocation by a combination of orpulat!on and GDP: A permit allocation that is based on a combination of population and GDP is unlikely to induce any broader participation than an allocation by population because the middle-income countries and the emerging market economies will end up with net costs under both allocations and therefore under any allocation that is also a combination of the two (Table 4). A nermit allocation that fully covers non-OECD projected emissions: An altemative allocation, given that allocations by population or GDP are unlikely to induce participation from all important groups, is identified: non-OECD countries are allocated emissions permits equivalent to projected baseline emissions in year 2000, Epk = Ek (7) for each country k of the non-OECDs. The OECD countries as a group are allocated permits equal to the targeted world emiissions minus permit allocations to the non-OECDs, EPOE= E- EPO (8) 12 The permits to the OECDs as a group can be distributed to each OECD country by several allocation schemes. We have arbitrarily chosen that each OECD country is allocated permits in proportion to their projected baseline emissions, Al _ Es EOECD / EOECD(9) where A1 is baseline emissions for country 1 and EOECD is total OECD baseline emissions7. Permit allocations relative to projected baseline carbon emissions for the year 2000 are presented in Table 3. The ratio is equal to one for all the non-OECD countries, and 0.38 for all the OECD countries. From (1), (2) and (7), we have that the net benefits relative to GDP for the non-OECD countries are, CkIGDPk = 0.135PIGDPk (10) and from (1), (2) and (9), that net costs relative to GDP for the OECD countries are, Cl/GDP, = -0.485P'E/GDPI (11) These net benefits and costs for the year 2000 are presented in Table 4. They are increasing in baseline emissions per dollar of GDP. Thus net benefits relative to GDP are highest (more than one % of GDP) for North Korea, Poland, Czechoslovakia and China, and the net costs are highest (more than 0.85% of GDP) for Turkey, Luxembourg, Australia and Greece. The average costs to the OECD countries are estimated at 0.50% of GDP. All the non-OECD countries would benefit from this allocation,8 and are therefore far more. likely to participate in a global treaty under this permit allocation compared to the allocations by population, GDP or some combination of the two. 7An interesting alternative allocation is one in which permits are allocated so as to equlize cost relative to GDP for each OECD county. Given relatively large differences in per capia income, and perhaps marginal cost functions, this can give rise to a very different permit allocation than an allocation that is proportional to projected emissions (see Bohm and Lrsen 1994). $See Bohm and Larsen (1994) for an allocation under which non-ECD countries (eastern european countries and states of the fonner Soviet Union in their case) receive permits so as to cover the costs of emissions reductions, i.e., without net benefits. 13 Unilateral reductions by the OECD countries: As an alternative to allocations that potentially could induce participation from the non-OECD countries, the OECD countries could unilaterally reduce emissions as to stabilize world emissions at 1987 levels. In this case the same quantity of permits would be allocated to the OECDs as under the allocation in which the projected emissions of the non-OECDs are fully covered, and we assume that permits to each of the OECD countries continue to be allocated in proportion to their baseline projected emissions. However, the OECDs can now only trade permits among themselves, and not with the non-OECDs. In this case the equilibrium permit price is estimated at US $181 per ton of carbon as compared to US $58 if all countries participate. The costs relative to GDP for the year 2000 are presented in Table 4. The costs are on average 0.99% of GDP, or almost twice as high as under the allocation in which the projected emissions of the non-OECDs are fully covered. The cost efficiency gains to the world as a whole is about 68%, which is significantly higher than the cost reductions to the OECD countries under the permit allocation in which the non-OECDs receive permits equivalent to their projected emissions. This is because a proportion of efficiency gains accrue to the non-OECD countries. A _comarison of costs and benefits of altemaive permit allocations: We have evaluated five alternative permit allocations in terms of their costs and benefits to participating countries and their likely merits to induce broad participation from OECD and non-OECD countries. An allocation by GDP is rather unlikely to induce participation from the non-OECD countries given the high net costs involved, although this allocation would result in net benefits to most of the OECD countries. An allocation by population would be less costly for the OECD countries than unilateral reductions. However, most of the niddle income and emerging market economies would unlikely participate if permits are allocated relative to population. We observed that the allocation in which the projected baseline emissions were fully covered for the non-OECDs would involve less avenge net costs for the OECDs than if permits were allocated by population, or OECDs were to undertake 14 unilateral reductions. The non-OECD countries would realize net benefits from this allocation. This leads us to conclude that the allocation in which non-OECDs are fully covered in terms of their projected baseline emissions could be attractive to both OECD and non-OECD countries. However, a potential problem with the latter allocait,n scheme is that net benefits relative to GDP for no-OECDs would be larger for non-OECDs with higher emissions per dollar of GDP, such as China, North Korea and Poland and therefore, countries that have achieved higher energy or carbon emission efficiency at their own cost prior to an international treaty would not be rewarded. One may argue that this consideration is not important for non-OECD countries because those that have achieved higher eergy or carbon emission efficiency have done so on national interest grounds. OECDs with higher energy efficiency would nevertheless be rewarded. For instance, the costs relative to GDP would for the United States be more than twice as high as for France if the marginal cost function in (1) is reasonably correct for both countries. However, if marginal costs are lower in OECDs with relatively low carbon or energy efficiency, the cost difference between low and high carbon efficiency OECD countries would be less than in Table 4. The allocation scheme in which non-OECDs are fully covered is noteworthy as developing coutries are not forced to reduce emissions that may slow their development efforts and OECD countries do not assume an unrealistic cost burden of emissions reductions. An additional merit of this allocation scheme is that it recognizes that OECDs had the advantage of unresticted emissions during their own development path, while non-OECDs are far from having used their "quota". Smith (1991) developed a "natural debt index" to illustrate this point by examining data on accumulated historical emissions. Furthermore, an international treaty on greenhouse gas emissions reductions need be such that countries have an incentive to participate. A simple, but useful, perspective may be to have an allocation scheme at relatively low cost to OECDs but that allows large enough transfers to induce participation frm non- OECD countries. The allocation scheme presented here, in which projected emissions of non-OECD countries are fully covered, may satisfy such a condition. 15 The free rider problem: An optimal policy response to global warming would be a cooperative agreement in which each country reduces emissions until the marginal costs of reductions for each country are equal to the sum of marginal benefits of reductions across all countries. In this case the marginal costs of reductions for any given country are generally larger than the marginal benefits to that country alone. Thus the country is generally better off by not participating in the treaty, i.e. by being a free rider, if a similar treaty is established among the remaining countries. The free rider problem is therefore an issue as regards the permit allocations evaluated in this paper. We have estimated the costs and benefits of participating in a global treaty based on marginal costs of em;issions reductions and permit allocations, ignoring benefits (or possibly costs for some countries) of greenhouse gas reductions in terms of climate changes. In this case, countries for which partcipation involves net costs after emissions reductions and permit trade, would in general be better off by not participating. However, countries for which participation involves net benefits would be worse off by not participating. In fact these countries would have zero benefits if they do not participate. The permit allocation, of the ones we have evaluated, that we found most likely to induce the broadest participation, and at significantly lower costs for the OECDs tban unilateral reductions, was the one in which each non-OECD country is allocated permits equivalent to its projected baseline emissions and the OECDs are allocated permits equivalent to the targeted world emissions minus permits to the non- OECDs. Under this permit allocation all the non-OECD countries would be worse off if they do not participate in the treaty. However, the free rider problem would remain among the OECD countries because their participation would involve net costs. In particular, this could be a problem among dLe less richer OECD countries. These countries may have to receive a larger volume of permiits in order to ensure their participation. 16 Variations in mareinal cost functions: A final consideration is possible deviations from the assumption of uniform marginal cost functions. We will consider one special case: suppose that the OECD countries still have the same marginal cost functions as (1), but that the non-OECD countries have lower marginal costs of emissions reductions than the OECD countries because, on average, OECDs have already exploited the lower cost options of improving their energy efficiency and reduce reliance on fossil fuels. With lower marginal costs in non-OECDs the equilibrium permit price will be lower, given the same world emissions target and baseline emissions. Emissions reductions would then partially shift from OECDs to non-OECDs where costs are lower. Thus the total cost to the OECDs of achieving the world emission target will be lower. This implies that the costs to the OECDs, as presented in Table 4, actually are overstated if the marginal costs in the OECDs are no higher and the marginal costs in the non-OECDs are lower. This is illustrated in Figure -1 with a linear approximation of the marginal cost function in (1). For the equilibrium permit price P, and marginal cost functions MC' and MCk for the OECDs and-non-OECDs respectively, the total costs to the OECDs are the area ABDE. In case the marginal costs in the non- OECDs are lower, say MC, the costs reductions for the OECDs are the area CDFG. Lower marginal costs in the nonOECD counties would also impact on their net benefits or costs. Under the allocation in which the non-OECDs receive permits equivalent to their projected emissions, and their marginal cost functions are given by MCk, their net benefits are the area BDF. If margina costs are lower, say MC, the net benefits are BCG. This area could be larger or smaller than BDF, depending on the decline in permit price relative to the increase in carbon reductions in the non-OECDs. In fact, net benefits will increase if the elasticity of the MCI curve (OECD) is elastic. The elasticity of the curve in Figure 1, i.e. equation (1) as estimated by Nordhaus (1991a), is inelastic in the relevant range, and thus net benefits to the non-OECDs would be lower than presented in Table 4 if marginal costs of emissions reductions for the non-OECDs are lower than assumed. 17 Costs/Benefits of Emissions Reductions and Permit Trade Figure 1 k MC MC | E Mc X P ~~~~~~~~...................................... ........... :F. . .............. tw............... A B OECD reduotlono Non-OECD reduotions S. Condusions This paper has evaluated altemative tradeable pernit allocations in a global permit regime for stabilization of world emissions at 1987 levels by the year 2000. An allocation by population is likely to be unacceptable to most middle income countries as well as to the emerging market economies of eastern Europe and the former Soviet Union because of the large net costs involved. An allocation by CDP will most likely be unacceptable to almost all non-OECD countries because they would end up bearing a large proportion of the costs of emissions reductions. Furthermore, an allocation based on a combination of population and GDP is unlikely to induce participation from most middle income and emerging market economies because of the net costs involved under allocations by both population and GDP. An alternative allocation scheme entails that non-OECD countries are allocated permits equal to their projected emissions, and that OECD countries are allocated permits equal to the world emissions target minus the permit allocations to the non-OECD countries. Such a scheme may be able to generate sufficient support for a viable international agreement on greenhouse gas emission reductions as it represents a reasonable allocation scheme given that the OECDs have a higher willingness to pay for eiission reductions, and the fict that non-OECDs are only responsible for a significantly lesser part of the stock of greenhouse gases in the atmosphere. Under this permit allocation, non-OECD countries would benefit from participation, and the net costs to the OECD countries would be only one-half of the costs of unilateral OECD reductions. The cost savings would be even larger if marginal costs of reductions in the non-OECD countries are lower than in the OECD countries. 19 REFERENCES Bohm, Peter (1992) "Distributional Implications of Allowing International Trade in C02 Emission Quotas." The World Economy, 15(1):107-114. Bohm, Peter and Larsen, Bjom (1994) "Fairness in a Tradeable-Permit Treaty for Carbon Emissions Reductions in Europe and the former Soviet Union." Forthcoming in Enviromnental and Resource Economics. Churchill, A.A. and Saunders, R.J. (1991) "Global Wanting and The Developing World." Finance and Development, June (28)(2):28-31. Glomsrod, S., Vennemo, H. and Johnson, T. (1992) "Stabilization of Emissions of C02: A Computable General Equilibrium Assessment." Scandinavian Journal of Economics, 94(1), 53-69. Hoel, Michael (1990) "C02 and the Greenhouse Effect: A Game Theoretic Exploration." University of Oslo. Hoel, Michael (1991a) "Efficient International Agreements for Reducing Emissions of C02." The Energy Journal, 12(2):93-107. Hoel, Michael (199 lb) "Global Environmental Problems: The Effects of Unilateral Actions Taken by One Country. Journal of Envirorunental Economics and Manasement, 20(l):55-70. Kosobud, R.F., South, D.W., Daly, T.A. and Quinn, K.G. (1991) "Tradeable C02 Emission Permits for Cost-Effective Control of Global Warming." Presented atthe 13th Annual North American Conference of International Association for Energy Economics. Chicago, Illinois, November 18- 20, 1991. Larsen, Bjom (1993) "World Fossil Fuel Subsidies and Global Carbon Emissions in a Model with Interfuel Substitution." World Bank Policy Research Working Paper Series, No. 1256. Washington, D.C. Larsen, Bjorn and Shah, Anwar (1994) "Energy Pricing and Taxation Options for Combatting the "Greenhouse Effect". In Akihiro Amano, editor, Climate Chan,e: Policy Instruments and their Implications. Intergovenmuental Panel on Climate Change, Tsukuba, Japan. Larsen, Bjorn and Shah, Anwar (1992a) "World Energy Subsidies and Global Carbon Emissions." World Bank Policy Research Working Paper Series No 1002. Washington, D.C. Larsen, B. and Shah, A. (1992b) "Combatting the Greenhouse Effect.' Finance and Development, December, 29(4):20-23. The World Bank. Washington, D.C. Nordhaus, William D. (199la) "The Cost of Slowing Climate Change: a Survey." The Enery Journal, 12(1):37-64. Nordhaus, William D. (1991b) "To Slow Or Not To Slow: The Economics of The Greenhouse Effect." The Economic Journal, 101(July):920-37. 20 Pearce, David (1991) "The Role of Carbon Taxes in Adjusting to Global Warmiing." The Economic Journal, 101(July):938-48. Poterba, James M. (1991) "Tax Policy to Combat Global Warming: On Designing a Carbon Tax." In R. Dornbusch and J.M. Poterba (eds.): Global Warmine: Economic Policy Response. MIT Press, Cambridge, Massachusetts. Quinn, K., McGrath, D.T. and Kosobud, R. (1992) "A Cost-Effective C02 Emission Permit System for Global Climate Change Mitigation." Presented at The Fifty-Sixth Annual Meeting of the Midwest Economics Association. Chicago, Illinois, March 26-28, 1992. Shah, Anwar and Larsen, Bjom (1992a) "Carbon Taxes, the Greenhouse Effect and Developing Countries." World Bank Policy Research Working Paper Series No 957. The World Bank. Shah, Anwar and Larsen, Bjorn (1992b) "Global Warming, Carbon taxes and Developing Countries." Paper presented at the 1992 Annual Meetings of the American Economic Association, January 3, tNew Orleans, USA. Smith, Kirk R. (1991) "Allocating Responsibility for Global Warming: The Natural Debt Index." Ambio, 20(2):95-96. April. South, D., Kosobud, R.F. and Quinn, K.G. (1991) "Greenhouse Gas Emissions Control by Economic Incentives: Survey and Analysis. " Presented at the 13th Annual North American Conference of International Association for Energy Economics. Chicago, Illinois, November 18-20, 1991. Summers, Lawrence H. (1991) "The Case For Corrective Taxation." The National Tax Journal vol. XLIV, No. 3:289-292. World Bank (1992) World Development Report 1992. Washington, D.C. World Resources Institute (1990) World Resources 1990-91. Washington, D.C. 21 Policy Research Working Paper Series Contact Title Author Date lor paper WPS1292 Services as a Major Source of Growth William Easterly April 1994 C. Rollison In Russia and Other Former Soviet Martha de Melo 84768 States Gur Ofer WPS1293 Product Slandards, Imperfect Glenn Harrison April 1994 N. Artis Competition, and Completion of the Thomas Rutherford 38C10 Market in the European Union David Tarr WPS1 294 Regulations, lnstitutions, and Hadi Salehi Esfahani April 1994 B. Moore Economic Performance: The Political 35261 Economy of the Philippines' Telecommunications Sector WPS1295 Why Higher Fiscal Spending Persists Bruno Boccara April 1994 M. Pfeiffenberger When a Boom in Primary Commodities Ends 34963 WPS1296 Earnings-Related Mandatory Salvador Valdes-Prieto April 1994 H. Rizkalla Pensions: Concepts for Design 84766 WPS1297 How Relative Prices Affect Fuel Charles C. Guo May 1994 C. Jones Use Patterns in Manufacturing: James R. Tybout 37699 Plant-Level Evidence from Chile WPS1298 Capital Goods imports, the Real Luis Serven May 1994 E. Khine Exchange Rate, and the Current Account 37471 WPS1299 Fiscal Policy in Classical and Klaus Schmidt-Hebbel May 1994 E. Khine Keynesian Open Economies Luis Serven 37471 WPS1300 Dynamic Response to Extemal Klaus Schmidt-Hebbel May 1994 E. Khine Shocks in Classical and Keynesian Luis Serven 37471 Economies WPS1301 Estimating the Health Effects of Bart Ostro May 1994 C. Jones Air Pollutants: A Method with an 37699 Application to Jakarta WPS1302 Sustainability: Ethical Foundations Geir B. Asheim May 1994 C. Jones and Economic Properties 37699 WPS1303 Conflict and Cooperation in Managing Scott Barrett May 1994 C. Jones Intemational Water Resources 37699 WPS1304 Informnal Gold Mining and Mercury Dan Biller May 1994 D. Biller Pollution in Brazil 37568 WPS1305 Informnation and Price Determination Nemat Shafik June 1994 A. Yideru Under Mass Privatization 36067 Policy Research Working Piper Seriea Contact Title Author Date for paper WPS1306 Capital Flows and Long-Term Ibrahim A. Elbadawl June 1994 R. Martin Equillbrium Real Exchange Rates Ralmundo Soto 39065 In Chiie WPS1307 How Taxation Affects Foreign Direct Joosung Jun June 1994 S. King-Watson Investment (Country Soecific Evidence) 31047 WPS1308 Ownership and Corporate Control In Brian Pinto June 1994 M. Kam-Choong Poland: Why Stato Firms Defied the Sweder van Wljnbergen 39618 Odds WPS1309 Is Demand for Polluting Goods Gunnar S. Eskeland June 1994 C. Jones Manageable? An Econometric Study Tarhan N. Feyzioglu 37699 of Car Ownership and Use In Mexico WPS1310 China's Economic Reforms: Pointors Justin Yilu Lin Juno 1994 C. Spooner for Other Economies In Transition Fang Cal 30464 Zhou U WPS1311 The Supply Response to Exchange Mustapha Rouls June 1994 J. Schwartz Rate Reform In Sub-Saharan Africa Weshah Razzak 32250 (Empirical Evidence) Cados Mollinedo WPSI312 The New Wave of Private Capital Eduardo Fernandez-Arias June 1994 A. Tutt Inflows: Push or Pull? 31047 WPS1313 New Estimates of Total Factor Vikram Nehru June 1994 M. Coleridge- Productivity Growth for Developing Ashok Dhareshwar Taylor and Industrial Countries 33704 WPS1314 The Significance of the Europe Bartlomiej Kaminski June 1994 M. Patefia Agreements" for Central European 37947 Industrial Exports WPS1315 Global Tradable Carbon Permnits, Bjom Larsen June 1994 C. Jones Particpation Incentives, and AnwarShah 37754 Transfers WPS1316 Preserving the CFA Zone: Macroeco- Shantayanan Devarajan June 1994 C. Jones nomic Coordination After the Michael Walton 37699 Devaluation