POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise AUGUST 2012 • Number 86 JUN 010 • Numbe 18 71595 Trade and Climate Change: An Analytical Review of Key Issues Harun Onder The last decade has witnessed an increasing global awareness of human impact on the planet’s climate and its likely consequences. However, strategic and structural complexities hinder further compliance and participation in efforts to establish a global agreement for climate change mitigation. This induces economists and environmentalists to further investigate the two-way relationship between trade and climate change, that is, climate-related consequences of liberal- ized trade and possible benefits of using trade policy for climate change mitigation. The Impact of Trade on Climate the impact of a marginal change in trade on the emission level in a given economy can be decomposed into three major chan- There has been a rapid expansion of international trade and nels: scale effect, composition effect, and technique effect.1 deepened trade liberalization among countries within the last The net change in aggregate emissions in this particular econ- few decades. The volume of world trade is nearly 32 times omy will be determined by the interaction of these effects. greater than the 1950 level. The ratio of merchandise trade to The scale effect refers to an increase in the emissions due gross domestic product (GDP) has increased from less than to increasing level of economic activity, holding all other fac- 20 percent to more than 50 percent in less than half a century tors constant. In a simple case without any other structural (figure 1). This was partly facilitated by reductions in average change in the economy, trade opening will increase the level world tariffs at major export destinations; that is, tariffs were of production, transportation, and consumption of goods. reduced from 18 percent in Europe and 15 percent in North This will, in turn, drive up the level of pollution through America in the late 1950s to about 4 percent in North Atlan- greater emissions during these processes. Therefore, the sign tic countries by the end of the 20th century (Baldwin 2006). of the scale effect is generally positive when isolated, because In parallel, there has been a drastic increase in the carbon di- higher trade level is usually associated with higher levels of oxide (CO2) concentrations in the atmosphere, from about economic activity. Analogously, a reduction in trade between 310 ppm (parts per million) in the 1950s to about 390 ppm countries due to a hike in trade barriers would cause a con- by the end of the century (World Bank 2010). The simultane- traction in economic activities and decrease the level of emis- ous expansion of trade, greater trade liberalization and higher sions through the scale effect. pollution intensities, therefore, raise questions about the cli- The composition effect, on the other hand, refers to a change mate impacts of trade and trade liberalization. in the emission level because of a change in the relative shares Do Emissions Rise When Countries Trade More? of different goods in the aggregate production of a country. Emissions of green house gases (GHGs) do not necessarily Other things being equal, a marginal change in trade barriers rise when countries trade more. Trade theory suggests that will alter the relative prices of the goods produced in an econ- 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 1. Trends in Output, Openness, and Emissions: Correlation or Causality? necessary fiscal and regulatory mea- 45 60 sures will be implemented to induce the producers to reduce the GHG in- million kt (CO2), US$ trillion (2000) 40 tensity per unit of output in the pro- 50 duction process. 35 This analytical approach has merchandise trade 40 proven most useful in a single country 30 (RHS) % of GDP case to investigate the marginal impact 25 30 of a reduction in trade costs on aggre- gate emissions. The sign of overall im- 20 CO2 emissions 20 pact depends on relative magnitudes 15 of these three effects, which, in turn, 10 depend on country characteristics 10 such as emission policies and compar- ative advantages. A cross-country 5 0 comparison, therefore, does not neces- sarily reflect the causality between GDP openness and emissions unless it ac- Source: World Development Indicators Database. counts for country-specific factors. Nevertheless, it may provide some in- omy. This will, in turn, change the quantities of production in tuition without claiming such causality. Figure 2 shows an equilibrium. When goods have different degrees of carbon in- application where openness is positively correlated with ag- tensity, a change in the composition of production will affect gregate emissions, higher per capita GDP, higher share of in- the overall emission level. The sign of the composition effect is dustry in the economy, and higher emission intensity per unit determined by the comparative advantage of the economy. A of energy use. More careful case studies reveal mixed results reduction in trade barriers will increase the domestic price of for the sign and magnitude of trade’s impact on aggregate pol- the exported good. Therefore, more of the national resources lution (box 1). will be devoted to the production of the exported good. If the The methodology introduced here has been proven use- exported good is relatively GHG intensive, then the overall ful in investigating seemingly simple but inherently complex emission will increase, hence the composition effect will be problems. An important one in this case is the impact on positive. If the country has a comparative advantage in the global pollution when countries with different income and “clean� industries, however, trade liberalization will cause a production characteristics trade more with each other. The further expansion of the clean industries and reduce the level following sections will analyze this and elaborate on how to of aggregate pollution. Therefore, the scale and composition employ the decomposition tool to answer important ques- effects work against each other in the case of “clean� good ex- tions on the climate change mitigation policy agenda. porters, whereas they enhance each other to increase the level Do Countries Pollute More or Pollute Less When They of pollution in the case of “dirty� good exporters. Become Richer? Finally, the technique effect refers to the impact of an in- Countries may emit more GHGs as they become richer, or crease in trade on emission level due to a change in produc- they may emit less. It is also possible that they emit more tion methods. Other things being equal, this effect represents GHGs as they get richer in the earlier stages, and emit less the variation in the derived demand for the polluting input later. The latter case, if correct, could provide a basis for the per unit of the final good or service. There are two major chan- arguments that support the strategic delay in developing nels through which trade reduces emissions per unit of out- country mitigation policies. put: first, trade facilitates technology transfer, hence, cleaner The Environmental Kuznets Curve (EKC) is a hypothe- technologies become available for adoption by domestic pro- sis that emphasizes the non-monotonicity of the relationship ducers. Developing countries might benefit significantly between income and pollution. Accordingly, the aggregate from increased access to modern technology embodied in im- level of pollution exhibits an inverted U shape when graphed ports. Second, trade can reduce the emission per unit of out- against GDP: the level of pollution rises when a poor country put through consumer preferences and political processes. takes off, and eventually goes down after a threshold level of Since environmental quality is considered to be a normal income. Figure 3 shows a cross-country correlation of GDP good, the demand for it will increase when trade raises in- per capita and CO2 emissions per capita. The emissions ex- come. If politicians are responsive to public demand, then hibit a partial inverted U shape; however, this does not neces- 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 2. A Cross-Country Analogy of the Decomposition Methodology Open economies emit more CO2 per person… …one reason is because they produce and consume more… 40 80 coef�cient of coef�cient of 35 correlation = 0.15 (metric tons per capita) correlation = 0.29 70 (thousands, 2005 $) GDP per capita, PPP 30 60 CO2 emissions 25 50 20 40 15 30 10 20 5 0 10 0 50 100 150 200 0 merchandise trade (% of GDP) 0 50 100 150 200 merchandise trade (% of GDP) …another reason is because they produce more energy- …and finally, because they employ more polluting technologies intensive goods relative to other goods… and/or inputs. coef�cient of 6 70 coef�cient of correlation = 0.33 of oil equivalent energy use) 60 correlation = 0.15 5 CO2 intensity (kg per kg industry, value added 50 4 (% of GDP) 40 30 3 20 2 10 1 0 0 50 100 150 200 0 merchandise trade (% of GDP) 0 50 100 150 200 merchandise trade (% of GDP) Source: WDI Database. Notes: The top two panels include 161 sample countries, the bottom left quadrant includes 141 countries, and the bottom right quadrant includes 102 countries. The main criteria for country selection were data availability in each category. Moreover, oil-rich countries (countries with oil rents greater than 30 percent of GDP) were also excluded. Two outliers (Singapore and Hong Kong SAR, China) with too high trade openness were also excluded without a significant impact on correlations presented here. Box 1. Measuring the Impact of Trade The empirical evidence on the net effect of trade on aggregate emissions is mixed. In an attempt to evaluate the environmental impli- cations of the North American Free Trade Agreement (NAFTA), Grossman and Krueger (1992) find that the sulfur dioxide (SO2) in the air increases with per capita GDP at low levels of national income, and decreases at high levels across nations: the turning point comes around US$5,000 (1985 values). In the case of Mexico, this implies that overall emissions may decrease after the trade agreement, because trade liberalization was expected to increase the national income beyond the specified threshold, and because the lower trade costs would induce production to shift toward labor-intensive and cleaner industries in which Mexico has a comparative advan- tage. However, Grossman and Krueger do not provide a decomposed estimate of the scale and technique effects. Antweiler, Copeland, and Taylor (2001) find that a 1 percent increase in GDP per capita due to trade liberalization decreases SO2 by about 1 percent by estimating the effects separately. This result is driven mainly by the technique effect always dominating the scale effect. A 1 percent increase in the economic activity raises SO2 concentration by 0.25 to 0.5 percent for an average country, whereas the emission concentration is driven down by 1.25 to 1.5 percent by the technique effect. The composition effect is reported to be country specific and relatively small, hence not affecting the result. Mani and Cunha (2011), on the other hand, estimate that the scale effect, which drives pollution intensities up, dominates the composition and technique effects in the case of the Dominican Republic–Central America Trade Agreement with the United States, albeit by a small margin. Sources: Antweiler, Copeland, and Taylor (2001); Grossman and Krueger (1992); Mani and Cunha (2011). 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 3. Is There a Kuznets Curve for CO2 Emissions? large one in the later stages, if in- come elasticity of marginal damage 20 due to pollution is increasing in in- come. This condition is essentially true when the willingness to pay for 15 (metric tons per capita) climate quality increases as income rises. And if this willingness is suffi- CO2 emissions ciently strong, the technique effect 10 will dominate the scale effect in rich economies. In the absence of a com- position effect due to neutral 5 growth, the aggregate pollution will increase in the beginning of the growth process, but will decrease 0 later. 0 10 20 30 40 50 The mechanisms above can gen- GDP per capita, PPP erate the EKC without considering (thousands, 2005 $) the impact of trade. However, autar- kic explanations of the EKC lack Source: World Development Indicators Database. Note: The sample covers 167 countries, excluding the countries with missing data, main oil and natural gas exporters, and some important strategic interac- two outliers (Luxembourg and Trinidad and Tobago). Best fitting line assumes a quadratic functional form, where the R2 is tions between countries. When 0.67 (the R in the case of a linear fit is equal to 0.60). 2 trade links economies in different stages of their development, what is sarily prove the existence of an EKC. An important caveat true for one country may not be true for others anymore; in here is that other country-specific characteristics that may others words, there might be a fallacy of composition. The potentially explain both GDP per capita and emissions per next section discusses this possibility. capita are unaccounted for. Does Trade Enable Rich Countries to Become Cleaner at A possible explanation for why there might be an EKC the Expense of Poorer Countries? could be the changing composition of production in the It is theoretically possible that both countries reduce their course of growth.2 Suppose the emission policy is fixed, the emissions, hence global pollution is reduced, when trade is polluting sector is capital intensive, and countries grow pri- liberalized. Therefore, trade may provide an alternative abate- marily via capital accumulation in the earlier phases of devel- ment mechanism. However, the impact of trade liberalization opment and via human capital improvement later. As capital on the aggregate pollution level in an economy depends on becomes more abundant relative to the effective labor in the the comparative advantage of the country. A marginal reduc- earlier stages, the economy will produce more of the polluting tion in trade barriers increases the international and domestic good relative to the clean good. Therefore, both the scale ef- prices of the good in which the country’s comparative advan- fect and composition effect reinforce each other in increasing tage lies. Therefore, a greater share of national resources is de- the aggregate pollution during the capital-augmented growth. voted to production of the exported good. Subsequently, as effective labor becomes more abundant In an open economy, the comparative advantage is shaped through human capital accumulation, more of the clean good by two forces: the pollution haven effect and the factor endow- is produced in the economy. At this stage, the pollution will ment effect, where the former refers to policy aspects of com- increase because of the scale effect and decrease because of parative advantage and the latter is related to the relative re- the composition effect. If the transition from capital-aug- source abundance (De Melo and Mathys 2010). If the mented growth to human capital–augmented growth is sub- polluting sector is capital intensive, then, all else being equal, stantial, then the composition effect will dominate and the a capital-abundant country with relatively loose pollution aggregate pollution will decrease. regulations will have a comparative advantage in the dirty in- Other factors that could generate the EKC are consumer dustry. Suppose the emission intensities are fixed by regula- preferences along with responsive emission policies. Suppose tion, then trade liberalization increases the aggregate emis- income grows because of factor neutral technological change, sions in a dirty-good-exporting country since both the scale and governments set the environmental policy to maximize effect and the composition effect work in favor of it. However, domestic welfare.3 Then, a marginal increase in income will it might reduce the aggregate pollution in the clean-good-ex- generate a small technique effect in the earlier stages and a porting country if the composition effect is strong enough. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Pollution haven and factor endowment effects do not Can Trade Policy “Level the Playing Field�? always reinforce each other. Developed countries are both Taxing GHG emissions is one of the most preferred options in capital abundant and more likely to have stricter climate implementing domestic mitigation policies.5 Typically such a policies. In this case, if the relative capital abundance in the tax is based on the content of fossil fuels embodied in produc- developed country is smaller in comparison to the relative tion. Therefore, the greater the fossil fuel dependency, the differences in income effects on pollution regulation, then higher the tax burden on a given sector in the domestic econo- the reductions in pollution (pollution haven effect) are my. In the absence of global compliance, however, unilateral greater in magnitude than the increments in it (factor en- taxation raises concerns on leakage and competitiveness. dowment effect). Therefore the developed country exports Leakage refers to an increase in GHG emissions in non- the clean good. In contrast, when the developed country is participating countries caused by an increase in the restric- sufficiently capital abundant, the factor endowment effect tions of emissions in participating countries. This can happen will dominate the pollution haven effect, and trade will through several channels: firms in regulated economies might cause an increase in the aggregate pollution in the developed shrink, while the ones in unregulated economies expand; country. Moreover, if the consumer preferences are such firms might relocate from regulated to unregulated econo- that the pollution policy is very responsive to income chang- mies; taxes in regulated economies decrease the international es, then trade can reduce the aggregate pollution both in and domestic prices of energy-intensive goods in unregulated developed and in developing countries.4 economies, boosting the domestic demand. Even though the The bottom line of the analysis is that different countries estimates of leakage magnitudes based on different comput- are affected differently by increased trade and greater trade erized general equilibrium algorithms vary, and there are sig- liberalization based on their relative factor endowments, pol- nificant results showing that the magnitude of leakage might icy choices, and the rigidity of these policies. be negligible, the concerns about leakage persist (Krishna 2010). Competitiveness, on the other hand, refers to a reduc- The Role of Trade Policy in Mitigating tion in market access for the firms in regulated economies Climate Change compared to the firms from unregulated countries. This is Current multilateral efforts to reduce GHG emissions are mainly driven by the cost disadvantages for the firms in par- considered to be ineffective because of insufficient participa- ticipating countries because of the higher emission taxes. tion and lack of enforcement (Barrett 2008). Governments Border tax measures (BTMs) are proposed as a mecha- differ in their willingness to impose limitations on activities nism to address the leakage and competitiveness concerns by with harmful climate impacts due to major complexities: leveling the playing field. BTMs might be implemented through first, projected harmful impacts of climate change and mitiga- a carbon-based tax on imports and subsidy on exports.6 Trade tion and adaptation costs differ across countries. Second, theory suggests that a uniform tax on both domestic and im- global climate is a public good; therefore, there is a free-riding ported goods is trade neutral if exporters are reimbursed opportunity for each country when others pay for it. Third, (Grossman 1980; Lockwood and Whalley 2008). However, there may be disadvantages for the first movers; that is, the when domestic goods and imported goods have different car- impact of efforts in one country may be undone fully or part- bon content because of different production processes, then ly by others if they do not participate. Fourth, any restriction the neutrality is no longer valid, and policy makers face a di- on current economic activity imposes definite instant costs, lemma. A BTM that is calculated using the carbon content of whereas the benefits may be realized in the future and are domestic production methods will not be as effective as a subject to stochastic variations. BTM based on actual (foreign) carbon content in reducing the Trade policy is, therefore, suggested as a mechanism to emissions. However, the latter method has several problems, facilitate compliance and participation in global mitigation and it is not clear whether it would be compatible with pre- and also supplement domestic measures to internalize the vailing multilateral rules under World Trade Organization cost of climate distortions. In particular, trade policy is con- (WTO) legislation or not.7 Moreover, the implementation of sidered to have four roles under a global climate regime (De this method may not be feasible because many imported Melo and Mathys 2010): address leakage and competitiveness goods are composite goods with many different inputs pro- issues; generate sanctions against nonparticipation and non- duced in different locations under different mitigation poli- compliance; help reach global efficiency in mitigation policies cies. Mattoo et al. (2009) show that the latter policy would by facilitating the separation of abatement location and the have significantly worse impacts on developing country man- bearer of the cost of abatement; and, finally, maintain a free ufacturing and exports than the former. This result is magni- trade regime that is crucial for technology transfers. The fol- fied for countries with higher export shares in GDP and high- lowing sections investigate these roles. er fossil fuel dependency. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise In light of these observations, trade measures at the bor- the projected costs of adopting mitigation policies and not der may be a costly and highly inefficient tool for leveling the adopting the policies but facing trade sanctions. playing field in the case of pollution policies. However, anoth- How Can Trade Policy Help Separate the Location of er option on the desk, which avoids some of the above men- Abatement and Bearer of the Cost? tioned implementation issues, is using border measures only Trade policy can ensure tradability of emission allowances in the case of noncompliance in mitigation policies. The next and clean development types of mechanisms, which are cru- section will consider this alternative. cial for the global efficiency of climate change mitigation. The Can Trade Sanctions Facilitate Compliance in Multilateral World Development Report (World Bank 2010) on climate Efforts for Climate Change Mitigation? change suggests “acting now, acting together, and acting differ- Using trade sanctions against nonmembers and noncompliers ently.� An important rationale behind acting together but dif- in multilateral climate change mitigation agreements is essen- ferently is the need to separate the location of abatement and tially equivalent to linking the two agreements. The main the bearer of the cost of abatement. Obviously, the marginal purpose is to transfer enforcement power from the trade cost of abatement would be different among countries de- agreement to the climate change mitigation agreement. How- pending on the current emission intensities and availability ever, this method has not been tested for WTO compatibility. of alternatives. Moreover, given that capital investments are Moreover, trade sanctions might not have the desired impact irreversible to a certain extent, saving future emissions might hoped for on nonmembers.8 cost less than reducing the current emissions as the “no-re- The Kyoto Protocol is considered to have inadequate par- gret� development idea suggests. This approach emphasizes ticipation and compliance because of missing enforcement the importance of mitigation efforts in developing countries. mechanisms. Even though the commitments are legally bind- However, global efficiency does not necessarily imply equity. ing, any sanctions against a noncompliant party are required Unilaterally optimal mitigation efforts in developing coun- to be approved by the same party, as in the General Agree- tries would be different than globally efficient levels. There- ment on Tariffs and Trade (GATT). In comparison, the Mon- fore, they would need to be compensated for the difference. treal Protocol enabled trade sanctions among the parties and Trade policy can help provide the necessary financing for against the nonparties to the protocol, which have received developing country mitigation through international trad- credit in ensuring the success of the protocol (Barrett 2008). ability of carbon allowances and technology transfers. The However, it is not obvious how the trade sanction threat actu- Kyoto Protocol introduced the Clean Development Mecha- ally functioned in altering the incentives in the latter case. nism (CDM) for catalyzing mitigation in developing coun- There are significant differences between a comprehensive tries: the signatories are allowed to meet their commitment climate change mitigation agreement and a relatively modest targets by investing in emission reduction projects in other attempt at addressing ozone depletion in terms of scale, costs, countries. From the developing country perspective, this and composition of the impact on economies. Moreover, the serves both the financing of mitigation and the adoption of trade sanctions were never implemented thanks to a broad cleaner technology.11 Eliminating the tariff and nontariff bar- participation in and compliance with the Montreal Protocol. riers against imports and foreign direct investment (FDI) in- Therefore, the legitimacy of the measures has not been tested flow might significantly improve the transfer of technology under current WTO legislation.9 through the CDM and similar mechanisms. A further ambiguity regarding the efficiency of trade However, Mattoo et al. (2009) point to an important ca- sanctions in promoting participation and compliance con- veat regarding the climate-motivated financial transfers from cerns the net impact of particular implementation mecha- developed countries to developing countries. Transfers with nisms on nonmembers. A carbon-based tariff on imports from carbon tradability can induce a decline in developing country nonmembers would affect the exporting country significantly. manufacturing output and exports via Dutch disease–type However, this cost might be small in comparison to the loss of mechanisms. Therefore, it is important to establish mecha- GDP growth had the exporter ratified the mitigation agree- nisms that would foster the efficient use of financial inflows. ment and employed the entailed mitigation policies.10 Obvi- In particular, the diminishing productive capacity in develop- ously, this would be less of an issue if the exporter is export ing countries because of the narrowing scope for traditional dependent or its exports more carbon intensive, because the technologies should be replaced with climate-friendly pro- impact of sanctions would then be greater. On the other hand, ductive capacity. The following role for trade policy addresses the opportunity cost of compliance in climate change mitiga- this issue. tion policies will also be higher when the economy is more Will Trade Be Free After All? carbon (energy) intensive and loses a greater share of GDP Trade measures carry a significant risk of triggering waves of growth because of the climate-motivated regulations. As a re- protectionist policies. Therefore, implementation procedures sult, the unilaterally optimal policies are to be determined by need to be well defined under a multilateral framework. It is 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise very likely that possible trade measures will be challenged at Acknowledgments the WTO. However, if the climate-motivated trade measures The author is grateful to Ian Gillson, Paulo Bastos, Daniel are not ruled out by the WTO panel, then the developing Lederman, Eduardo Ley, Apurva Sanghi, Kazi Mahbub-al countries would be tempted to increase their current tariff Matin, Stefanie Sieber, and Giovanna Bua for valuable com- rates on imports from the respective partners to the bound rates without facing any legal restraint. Even when trade mea- ments and suggestions. sures emanate from a multilateral agreement that developing About the Author countries themselves ratify, this will not exempt them from the risk of initiating a trade war. There are many technical Harun Onder is a Young Professional in the World Bank’s Pov- complexities in the implementation of trade measures that erty Reduction and Economic Management (PREM) Network, may be observed imperfectly and misinterpreted. Among Economic Policy and Debt Department. these complexities are the difficulty in assessing product-spe- Notes cific emissions and unobservable technical regulations in the country of origin that would affect the imported goods’ quali- 1. For a detailed analysis, see Grossman and Krueger (1992) fication for a border measure (WTO-UNEP 2009). Any of and Copeland and Taylor (2003). these problems might induce a government to take retaliatory 2. This helps eliminate the technique effect; for a detailed action when it mistakenly believes that the partner is pursu- analysis of alternative cases, see Copeland and Taylor (2003). ing a protectionist policy.12 3. This might be a regulation that determines the optimal pol- Hence, multilateral investigations are necessary for joint- lution intensity per unit of output. A technology change is fac- ly accepted trade and climate policies. Careful and detailed tor neutral (or Hicks neutral) when the ratio of marginal pro- definitions of implementation tools and procedures are cru- ductivities of factors do not change for a given ratio of factors. cial in preventing the undesired protectionist consequences 4. See Copeland and Taylor (2003) for model details. of trade policies. 5. Alternatively, the same outcome can be reached by cap- Concluding Remarks and-trade policies, where a fixed amount of emission allow- ances are traded in the market (Mattoo et al. 2009). However, This note summarized a methodology that is particularly use- a tax system has advantages over cap-and-trade policy, such as ful in considering the impact of further trade liberalization verifiability (De Melo and Mathys 2010). on pollution and answering questions on how trade policy 6. In the case of cap-and-trade policies, importers would be can be used to address the challenges in global efforts to miti- required to submit emission allowances. gate climate change. 7. GATT has ruled against U.S. barriers based on the process- The main points highlighted in this note are: first, the es and production methods in the United States–Mexico dol- impacts of trade on GHG emissions are not uniform across phin–tuna case; however, a WTO panel ruled that measures countries, but are mainly driven by the differences in coun- can target processes and production methods in the United try characteristics such as underlying comparative advantag- States–India shrimp–turtle case. es. Second, the same country characteristics will also deter- 8. Chisik and Onder (2012) show that enabling cross retalia- mine whether GHG emissions would increase or decrease in tion by linking the agreements may actually reduce the coop- an economy as it grows. Therefore, delaying mitigation in a eration between governments if the policy actions are strate- given country may not necessarily lead to higher income and gic substitutes and the limited punishment understanding of lower emissions in that economy after the strategic delay. the GATT XXVIII rule prevails. Third, noncompliance by a country in climate change miti- gation does not automatically create a “pollution haven.� 9. Sampson (2000) suggests that trade sanctions cannot be Other factors, such as the factor endowment effect, may used against nonparties under the WTO legislation. dominate the pollution haven effect to reduce emissions in 10. Panagariya (2010) anticipates a loss of US$2.1 trillion in that country when trade barriers are lowered. Fourth, BTMs net present value terms in 2020 for India, if ratification causes can successfully level the playing field by addressing the leak- a 1 percentage point of loss in the current 8 percent growth age and competitiveness issues; however, there are nearly rate. prohibitive challenges in the implementation process and 11. Thirty-three percent of Clean Development Mechanism the measures themselves may fail to facilitate compliance. projects have claimed technology transfer (WTO-UNEP Finally, trade policy can effectively enable developing coun- 2009). tries to have access to clean technologies and finance for 12. Chisik (2010) shows that when degrees of information mitigation efforts. asymmetry differ across the issues, then using a policy tool in 7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise one area to punish a violation in the other reduces the welfare Grossman, G., and A. B. Krueger. 1992. “Environmental Impacts of a North of the parties by initiating a trade war, which would not occur American Free Trade Agreement.� CEPR Discussion Papers 644, London. if the issues were kept separated. Krishna, K. 2010. “Limiting Emissions and Trade: Some Basic Ideas.� NBER Working Paper Series w16147, Cambridge, MA. References Lockwood, B., and J. Whalley. 2008. “Carbon Motivated Border Tax Antweiler, W., B. Copeland, and S. Taylor. 2001. “Is Free Trade Good for the Adjustments: Old Wine in Green Bottles?� NBER Working Paper Series w14025, Cambridge, MA. Environment?� American Economic Review 91 (4): 877–908. Mani, M., and B. Cunha. 2011. “DR-CAFTA and the Environment.� In Baldwin, R. 2006. “Multilateralising Regionalism: Spaghetti Bowls and Getting the Most Out of Free Trade Agreements in Central America, ed. H. Building Blocks on the Path to Global Free Trade.� World Economy 29 Lopez and R. Shankar. Washington, DC: World Bank. (11): 1451–1518. Mattoo, A., A. Subramanian, D. van der Mensbrugghe, and J. He. 2009. Barrett, S. 2008. “Climate Treaties and Imperative of Enforcement.� Oxford “Can Global De-Carbonization Inhibit Developing Country Industri- Review of Economic Policy 24 (2): 239–58. alization?� World Bank Policy Research Working Paper Series 5121, Chisik, R. 2010. “Limited Incremental Linking and Unlinked Trade Agree- Washington, DC. ments.� Ryerson University Working Paper Series 023, Ontario. Panagariya, A., 2010. “Climate Change and India: Implications for Policy Chisik, R., and H. Onder. 2012. “Limiting Cross-Retaliation When Punish- Options.� University of Columbia, mimeo. ment Is Limited: How DSU Article 22.4 Complements GATT Article Sampson, G. 2000. Trade, Environment and the WTO: The Post-Seattle XXVIII.� Ryerson University Working Paper Series 025, Ontario. Agenda. Overseas Development Council and Johns Hopkins Press: Copeland, B., and S. Taylor. 2003. Trade and Environment: Theory and Evi- Washington, DC, and Baltimore. dence. Princeton University Press, Princeton. World Bank. 2010. World Development Report: Development and Climate De Melo, J., and N. A. Mathys. 2010. “Trade and Climate Change: The Chal- Change. Washington, DC. lenges Ahead.� CEPR Discussion Paper Series 3083, London. WTO (World Trade Organization), and UNEP (United Nations Environ- Grossman, G. 1980. “Border Tax Adjustments: Do They Distort Trade?� mental Programme). 2009. Trade and Climate Change Report. World Journal of International Economics 10(1): 117–28. Trade Organization, Geneva. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 8 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise