OCTOBER 2020 How could trade measures being considered to mitigate climate change affect LDC exports? Michael Friis Jensen © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Financial support for the project was provided by Umbrella Facility for Trade Trust Fund with contributions from the Governments of the United Kingdom (DFID), Sweden (Sida), Norway (Ministry of Foreign Affairs), Netherlands (Ministry of Foreign Affairs, and Switzerland (SECO). Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Working Paper How could trade measures being considered to mitigate climate change affect LDC exports? October 2020 1 Abstract Export led growth remains a critical route out of poverty for many least developed countries (LDCs). But in a world increasingly concerned about climate change and the contribution of international trade to global warming will measures being considered to mitigate emissions hurt the export opportunities of LDCs? This paper reviews the trade-related instruments being considered by both policymakers and business communities to mitigate climate change, identifying areas where LDCs may be affected. The paper explores the following key issues: (i) how a shift to low carbon transportation may affect LDC exports given their remote location from main markets; (ii) how LDCs’ exports will be influenced by their own domestic climate measures; (iii) whether mitigation instruments introduced by other governments that result in carbon border tax adjustments will significantly affect LDC exports; (iv) the importance of the nature of liberalization of trade in green goods; (v) how government sustainability standards in overseas market may affect trade, and finally (vi) the potential impact of private measures used by businesses for carbon management. The paper shows how climate change mitigation instruments will create challenges and provide new opportunities for LDC exports that will require attention to traditional trade policy issues but also to a deeper agenda relating to regulatory development and convergence. 2 Contents 1. Introduction 2. Will LDCs’ exports suffer in a carbon-constrained world? 3. LDCs’ own mitigation initiatives 4. Carbon border adjustment 5. Green liberalization 6. Sustainability standards 7. Business carbon management 8. Concluding discussion 3 How may trade measures being considered to mitigate climate change affect LDC exports? 1 1. Introduction About a quarter of global carbon emissions are embodied in international trade (Sakai and Barrett 2016) and regulating trade is therefore on the mind of climate policy practitioners. At the same time, trade is relatively more important for Least Developed Countries (LDCs), as reflected in higher than average export-to-GDP ratios. In a world increasingly concerned about climate change and becoming more willing to contemplate trade-related measures to fight it, should LDCs be worried? Will mitigation instruments hurt LDCs’ export opportunities? This paper addresses this issue by reviewing the trade-related instruments being considered by policymakers and business communities to mitigate climate change and identifying areas where LDCs may be affected. The report discusses the nature of the potential outcomes and how policy initiatives and development assistance may help to avoid or soften negative impacts and where possible to exploit new trade opportunities that may arise. Climate mitigation instruments are a large and diverse group of government policy and private business tools put in place by governments, NGOs, and firms, and even combinations of the three in multi-stakeholder initiatives. The paper does not intend to provide an exhaustive review but seeks to focus on the mitigation instruments most likely to affect LDC exports. First, we ask whether LDC exports will be hampered by LDCs’ remote location from their main markets when transport modes become low carbon. Second, we ask how LDCs’ exports will be influenced by their own domestic climate measures. Third, we ask how foreign mitigation instruments will affect LDC exports. We identify four types of mitigation measures: tariff cuts and other forms of liberalization of trade in green goods, carbon border tax adjustments, government sustainability standards, and private measures used by businesses for carbon management. One trade-related area not addressed in this paper is the emergence of markets for carbon offset projects. This market is particularly relevant because of the relatively low cost of emission reductions in LDCs. This is an important issue in its own right but does not fall within what is normally considered trade and development. Hence, except for a brief discussion of the Clean Development Mechanism (CDM) we leave the discussion of offsets to another time. Nevertheless, this paper provides a useful contribution in the context of offset markets by reviewing the development of carbon accounting methodologies to assess emission reductions, which is central to the operation of such markets. LDCs, as defined by the UN, are the focus of this paper. However, other developing countries share key characteristics of LDCs, such as vulnerability and dependence on a few exports, for example, 1 Report prepared for the World Bank project on Climate Change and Trade: Impacts and Opportunities for Least Developed Countries by Michael Friis Jensen, former staff member of the World Bank and currently working as an independent consultant with guidance and inputs from Paul Brenton and Vicky Chemutai, World Bank. 4 lower middle-income countries in Africa and small, island economies – and many of the observations and policy recommendations made here will be relevant for these countries too. Hence, many of the policy issues we will identify, for instance when discussing carbon accounting, are relevant for a larger group of lower middle-income developing countries. How LDCs’ exports might be affected by climate mitigation has not been widely studied. Keane (2012) provides an earlier assessment, written prior to the Paris Agreement, but neither the trade nor the climate change research communities have devoted much attention to the intersection between climate change, trade, and development.2 The novelty of many of the instruments involved is also a challenge. Researchers and practitioners have discussed carbon border adjustments for two decades yet very few have been implemented, thus there are equally few to evaluate. Governments apply sustainability standards in a multitude of areas, but their developmental impacts are not often analyzed. Indeed, many private business initiatives in carbon management are emerging, but most of these are very recent and developed in closed environments beyond the reach of analysts. With this focus and these limitations in mind, the paper is structured as follows. Section 2 asks whether LDCs will find it more difficult to export in a climate-constrained world. Section 3 looks at LDCs’ own mitigation initiatives. Section 4 examines carbon border adjustments, trade measures a country regulating carbon emissions domestically imposes to ensure the domestic regulations are not undercut by imports from other countries without equivalent regulation. Section 5 discusses green trade liberalization. Section 6 discusses sustainability standards. Sections 4, 5, and partly 6 concern government policy, while Section 7 focuses on business carbon management. This relates to a collection of private measures that many firms, primarily in rich countries, use to demonstrate that they seek carbon emission reductions in their operations and in the value chains through which they source their inputs and raw materials. Section 8 concludes and discusses policy recommendations. 2. Will LDCs’ exports suffer in a carbon-constrained world? Climate change mitigation includes a series of measures that are likely to increase the costs of transportation of traded goods, including LDC exports. There is a popular perception that the emissions caused by international transport are a hotspot in the supply chains and that they should be reduced to mitigate climate change. Exports from countries that are distant from global markets using high-emitting transport modes could be penalized, for example, fresh produce exports airlifted to market, or tourism in LDCs. A key technology question is whether LDCs will be able to exploit new carbon-efficient transport modes. A key policy question is whether LDCs, due to their low development levels and high need for growth, will be exempt from certain policy measures like carbon border measures. In the past, LDCs have enjoyed a special status in the world trade system and have received certain privileges and preferences relative to other countries. However, mitigation may also open new export markets for LDCs, exemplified by the potential for them to participate in producing and supplying zero-emission fuels.3 2 The intersection of climate change and trade policies, on the other hand, has been scrutinized by many legal scholars and trade economists, and the climate change and development debate is a very lively one, as is the trade and development debate. The overlap of all three agendas, however, has not been widely explored. 3 See https://blogs.worldbank.org/transport/zero-emission-shipping-whats-it-developing-countries 5 To discuss such questions, we briefly review the evidence on how transportation costs impact export volumes and structures, and subsequently look at how mitigation measures might affect costs. We then look at LDCs’ current export structures and how emissions are distributed in the global supply chains that LDCs are part of to gain an understanding of how vulnerable LDCs are to increased transportation costs caused by mitigation. We concentrate on air and sea transport but note that a number of LDCs and other developing countries are investing heavily in rail transport, which can be both more cost-effective and carbon reducing relative to road transport. Sea transport costs may be affected by a series of mitigation measures. Both the International Maritime Organization (IMO) and individual firms have been active in developing mitigation measures. In April 2018, the IMO announced a commitment of the shipping sector to reduce emissions by 50 percent of 2008 levels by 2050. Some firms set higher targets: the world’s biggest container shipper, Maersk, aims to be carbon neutral by 2050. Shippers also work collectively outside of the IMO. Maersk will work with industry players like Siemens, MAN, ABS, Mitsubishi, and Cargill to establish a research center of 100 scientists to support reduction in carbon emissions from shipping.4 Policymakers are considering instruments to incentivize the shipping industry to adopt low-carbon forms of sea transport. The Paris Agreement did not include specific language on shipping and air travel despite earlier drafts that had included relevant text (Hulac 2015). Some observers argue that this omission is because the two relevant UN agencies, the International Civil Aviation Organization (ICAO) and the IMO, are working with countries to mitigate the emissions of international transport. Of course, individual countries may still include targets for these two international activities in their national mitigation plans and might be able to act more effectively than the two UN agencies (Murphy 2018). The EU provides an example of countries contemplating shipping mitigation policies. The European Commission has proposed to include shipping in the EU Emission Trading System (ETS) if talks to reduce emissions with the IMO are not successful. This would put a price on emissions from shipping. Policymakers have also discussed establishing a fund to recycle the ETS revenue that shipping would generate to support low-emission shipping and to set binding targets for shipping companies to reduce carbon intensity by 40 percent from 2018 levels by 2030 (Stoefs 2020). Aviation, for both freight and passengers, is also under pressure to reduce emissions as exemplified, for example, by the consumer phenomenon ‘flight shame’ and no-fly campaigns.5 The ICAO has adopted a mitigation policy based on five elements: • Technology development, including the setting of emission standards and the introduction of biofuels. • Operational improvements through initiatives such as fuel efficiency monitoring and more direct flight paths. • Infrastructure improvements to make airports more fuel efficient. • Measuring all CO2 emissions on international routes. 4 https://www.weforum.org/agenda/2020/06/maersk-zero-carbon-shipping-sector-research-center-climate- change 5 https://www.theguardian.com/travel/2019/jan/26/why-i-only-take-one-holiday-flight-a-year-climate-change 6 • Capping CO2 emissions at their 2020 level and offsetting aviation’s CO2 emissions above the 2020 level through the Carbon Offsetting and Reduction Scheme for International Aviation (CORSIA). The ICAO has adopted a CO2 emissions standard which Larsson et al. (2019) describe as a performance standard with a maximum emissions level for all new aircrafts sold. They observe that the standard will only lead to modest reductions and is unlikely to contribute to emission reductions beyond what can be anticipated from the gradual improvements already occurring due to technological progress. CORSIA is the ICAO instrument that has attracted most attention. CORSIA is an offsetting scheme that is the centerpiece of the aviation response, at least in the short run, since – in contrast to shipping – new technologies that may mitigate the contribution of aviation to global warming are far from fully developed. ICAO members will be bound to offset part of their CO2 emissions through CORSIA. The European Commission expects that CORSIA will offset around 80 percent of airlines’ emissions above 2019 levels.6 A series of notable exemptions exist: • Airlines must offset only emissions above their 2019 level. • Airlines will only offset CO2 emissions and none of the other global warming impacts that aviation causes, such as NOx emission and emission of water vapor causing condensation trails. There is considerable uncertainty about the importance of non-CO2 impacts, yet many observers believe that the global warming impact of aviation is twice the impact caused by its CO2 emissions alone. • Airlines from 80 countries, representing 77 percent of international air traffic, have joined the deal’s voluntary first phase between 2021 and 2026. It becomes mandatory from 2027 for states with large aviation industries. The US has joined7 while China has not. • LDCs, Small Island Developing States (SIDS) and Landlocked Developing Countries (LLDCs) (Larsson et al. 2019). This exemption means that routes that depart from or land in any of these countries are exempt from the offsetting requirement. • CORSIA also exempts countries with very small aviation industries, very small operators, and very small planes, and flights for medical emergencies, disaster responses, and other humanitarian purposes. Furthermore, CORSIA only covers civil aviation. Presidential, police, customs, and military flights are excluded. As a result of these and other exceptions, Hedley et al. (2016) estimate that CORSIA will only cover about 40 percent of global aviation. National efforts might be more important than the ICAO policy. The EU formally included the aviation industry in its ETS from January 2012.8 Initially, all flights from and to EU airports were 6 Originally, the baseline for emission reductions was to be the average of emissions over 2019 and 2020. However, with COVID-19 drastically reducing air traffic and emissions in 2020, there would have been a substantial impact on the CORSIA baseline. In response the ICAO Council decided that the value of 2019 emissions shall be used for to avoid increasing the economic burden on the aviation industry of CORSIA implementation. https://www.icao.int/Newsroom/Pages/ICAO-Council-agrees-to-the-safeguard-adjustment- for-CORSIA-in-light-of-COVID19-pandemic.aspx 7 CORSIA, established by the ICAO, operates outside the Paris Agreement and, in principle, is not affected by the withdrawal of the US from this agreement. US airlines have expressed interest in participating in CORSIA. 8 The scheme covers the European Economic Area, thus the EU Member States plus Iceland, Norway, and Switzerland. 7 included, with the exception of LDCs. The inclusion of out-of-EU flights, however, gave rise to strong protests from the EU’s trading partners, like the US and China, and subsequently, the EU included only intra-EU flights. The European Commission has argued that it reduced the scope of its aviation policy to give ICAO time to agree a global measure. The exemption of flights out of Europe has been extended to 2024, after which the EU might decide to include all flights if it is not satisfied with the level of ambition and operation of CORSIA. Some studies discuss the costs of the emerging mitigation measures in international transport. Naturally, at this stage it is not easy to assess these as the policy and business instruments are still in discussion and new technologies are still under development and COVID-19 has (at least temporarily) changed travel habits. What is clear, though, is that the cost implications will be more manageable in shipping than in aviation. Shipping is already a relatively carbon-efficient way to move goods, while aviation is particularly carbon-intensive, and a range of options exist for shipping, such as lowering vessel speed and developing carbon-neutral fuels. In aviation, potential carbon- efficient technologies are harder to develop. Halim, Smith and Englert (2019) find that likely mitigation measures will increase the import price of goods only marginally. The study estimates that mitigation measure, such as those that increase the costs of bunker fuel by between US$10 to 50 per ton of CO2, might increase maritime transport costs by between 0.4 to 16 percent, yet this cost increase will increase the import price of goods by less than 1 percent. The finding might appear counterintuitive as shipping’s aggregate carbon footprint appears large. Observers estimate that the international shipping footprint is around 2-3 percent of total carbon emissions, about the same as Germany’s total carbon footprint. But shipping’s total is of course allocated to an enormous number of individual goods transported. The share of international transport in a goods carbon footprint is often small. Costa Rican coffee exemplifies the minor share of transport in commodity exports as evidenced in Table 1. Farm-level emissions are significant (20 percent) and caused mainly by fertilizers (19 percent) but consumers are responsible for nearly half of the total. International transport causes only 4 percent of total emissions along the value chain. The scope to reduce the carbon footprint of a commodity like coffee is much greater through changing consumer behavior than by reducing emissions from shipping. For example, Killian et al. (2013) emphasize that many coffee brewers in cafes and restaurants are constantly turned on, consuming considerable energy. Table 1 Carbon footprint of 1 kg of green coffee Stage Kg CO2e % In Costa Rica 1,50 31% Farm 1.02 20% Fertilizers 0.96 19% Mill 0.48 10% Exportation 0.27 6% International shipping* 0.19 4% In Europe 3.05 63% Roasting 0.19 4% Packaging 0.13 3% Distribution 0.15 3% Grinding and purchasing 0.29 6% 8 Consumption 2.15 45% Disposal 0.14 3% Total 4.82 100% * Costa Rica to Europe. Source: Killian et al. (2013). The impact of mitigation measures on aviation may be more severe than for shipping. The ICAO’s CORSIA will exempt LDCs but will affect a range of similar countries neighboring many LDCs. Kusova and Dufour (2019) find that 12 African countries intend to participate voluntarily, including the LDCs of Burkina Faso, Democratic Republic of Congo, Uganda, and Zambia.9 South Africa is expected to join when CORSIA becomes mandatory in 2027 and so might other larger African non-LDCs as their aviation sector exceeds the threshold. LDC freight and personal travel may be affected if they are routed through participating countries. The proposed mitigation measures do, however, appear modest. The EU’s ETS may include international aviation, implying higher costs than CORSIA. Today, the impact of the EU’s ETS even on flights that are included is minimal, due to the free allocation of many permits and the generally low price of permits that are sold. The environmental NGO, Transport & Environment, has calculated that the extra costs of an economy ticket on an Oslo-Rome flight would be less than €4 (US$ 4.5) per return ticket. The costs are low because about half of the allowances needed are allocated for free and the other half are sold at very low cost.10 When the EU first discussed including aviation in its ETS, initial research by the European Commission calculated modest increases for tickets to a typical developing country tourist destination. Good (2010) estimated that the scheme as initially proposed would have increased the costs of a return ticket to the Caribbean by €16 to €20. The impact on tourism, including in LDCs, is made particularly difficult to assess because the costs depend on the EU’s practice so far to grant allowances for free and on the volatile market price for allowances. Some observers forecast a negative impact in vulnerable developing countries. Bartels (2012) fears reductions in tourism demand. He finds that a destination like Barbados could see a 1-2 percent reduction in GDP because of its highly price-sensitive tourism industry and its high dependency on tourism. Fearing severe impact on vulnerable countries’ tourist industries, Keane (2012) criticizes the EU for only exempting LDCs from inclusion of aviation into the ETS and not other groups like Small and Vulnerable Economies and SIDS. The carbon footprint of air transportation of goods may be small relative to emissions at other points along the value chain. A number of studies of the horticulture sector have shown that imports by air from typically low-income countries that utilize sunshine, manual labor, and natural compost in production generate fewer emissions than produce from nearby countries that require heated greenhouses, tractors, and fertilizers. Again, emissions at the consumption stage often dominate. On the basis of these studies, a UK Minister of Trade noted during a debate on so-called food miles and aviation hotspots that “driving 6.5 miles to buy your shopping emits more carbon than flying a pack of Kenyan green beans to the United Kingdom”.11 9 Other countries are Cameroon, Equatorial Guinea, Gabon, Ghana, Kenya, Namibia, and Nigeria. 10 https://www.transportenvironment.org/state-aviation-ets 11 https://airlines.iata.org/analysis/cargos-carbon-emissions 9 3. LDCs’ own mitigation initiatives LDCs’ mitigation initiatives reflect their strong need for economic growth. Because growth is urgently needed to combat poverty, LDCs plan to let their total emissions continue to grow. But they will seek to shift to a less carbon-intensive growth path through mitigation activities. The LDCs’ contribution to the fight against climate change will thus be a reduction in future emissions as compared with a business-as-usual scenario and not as compared with current emissions. Planning total emission increases while shifting to a less carbon-intensive growth path is in line with LDCs’ historical responsibilities. These mitigation plans are most clearly articulated in Nationally Determined Contributions (NDCs) submitted following the Paris Agreement. This agreement requires that each country prepare NDCs outlining how the country plans to address climate change. The goal of the Agreement is to limit global warming to an increase of 2°C and to pursue efforts to limit the increase to 1.5°C. Each country determines its own ambition level in accordance with the principle of common but differentiated responsibilities. An LDC is free to choose how and to what degree it will participate in the fight against global warming. This flexibility was an important principle during the negotiations, without which it would have been difficult to get countries with different historical responsibilities for the greenhouse gases already emitted since the Industrial Revolution, and with different future development needs, to agree on a global climate change agreement. The Paris Agreement does not contain a mechanism to force countries to reach specific emissions targets by specific dates but leaves them discretion to set limits and dates for their own economies. Parties, including LDCs, are however legally bound to have their progress tracked by expert review and to determine ways to strengthen ambitions over time. By April 2016, 190 Parties to the Paris Agreement had communicated NDCs to the United Nations Framework Convention on Climate Change (UNFCCC), covering 97 percent of all Parties to the UNFCCC with a coverage of 95 percent of greenhouse gas emissions. Table 2 tabulates the commitments of the LDCs. Three aspects characterize most LDC NDCs. First, almost all their NDCs set goals as deviations from business-as-usual scenarios. Sao Tome & Principe, for instance, states that in 2005 “emissions were about 93 ktCO2eq, and BAU [business-as-usual] projection indicates that to fulfil the condition of the supposed growth parameters by 2030, the emissions will reach 240 ktCO2eq.” (Sao Tome & Principe 2015:6). Thus, any reduction goal is in relation to the forecast 2030 level of 240. The goal of Sao Tome & Principe and most other LDCs is to apply mitigation to shift from the business-as-usual scenario, in which it grows with its historical carbon intensity, to a less carbon-intensive growth path. Second, most LDC mitigation activities planned in the NDCs are conditional on foreign aid. LDCs split their reduction commitments in two parts, unconditional and conditional reductions. A country will implement unconditional reductions on its own, while it will only undertake conditional reductions if developed countries provide assistance in terms of finance, technical advice, or other support. As evidenced in Table 2, LDCs mainly make conditional reduction commitments. In general, LDCs and SIDS have a higher proportion of conditional commitments than other countries (Pauw et al. 2020). 10 Table 2. LDC mitigation commitments under the Paris Agreement Reduction targets # Country Total Unconditional Conditional Comments Afghanistan 13,6% 0,0% 13,6% Angola § 50,0% 35,0% 15,0% In industry, forestry and land use, agriculture, renewable energy, biofuels Bangladesh 20,0% 5,0% 15,0% In power, transport and industry Benin 16,2% 3,6% 12,6% Excludes forestry Bhutan To remain carbon neutral Bhutan is carbon neutral today and aims to remain so Burkina Faso 18,2% 6,6% 11,6% Burundi 25,0% 5,0% 20,0% Cambodia 26,7% 0,0% 26,7% Cambodia also plans to increase forest cover to 60% conditional on aid Central African Republic Targets expressed in kt-CO2 for 2030 & 2050 Chad 18,2% 71,0% In energy, agriculture & livestock, forestry & land use, waste Comoros 84,0% 84,0% In energy, agriculture, forestry and land use, waste DR Congo 18,2% 71,0% In energy, agriculture & livestock, land use & forestry, waste. Djibouti 40,0% 20,0% In energy, agriculture, waste, industrial processes Eritrea § 12,6% 38,5% Ethiopia 64,0% No conditional/unconditional split Gambia Sector (mainly cond.) targets expressed in tons CO2e In agriculture, energy, industry, transport, and waste Guinea Conditional sector targets expressed in tons of CO2e Lack of date prevents setting goals in some sectors Guinea-Bissau No targets due to lack of data; will mitigate if aided In reforestation, energy Haiti 31,0% 5,0% 26,0% In energy, agriculture, forestry, waste Kiribati 61,8% 12,8% 49,0% In energy, mangrove forest management Lao PDR Conditional project targets expressed in tons of CO2e In forestry, energy, transport Lesotho 35,0% 10,0% 25,0% In energy, industrial processes, agriculture, land use change and forestry, waste Liberia 15,0% 15,0% In renewable energy, cook stoves, biofuels, landfill Madagascar 14,0% 14,0% In energy, agriculture, forestry and other land use, waste Malawi No targets Lists projects in energy, industry, agriculture, forestry and land use, waste 11 Table 2. LDC mitigation commitments under the Paris Agreement (continued) Reduction targets # Country Total Unconditional Conditional Comments Mali Sector goals 29% agriculture, 31% energy, 21% forestry and other land use Mauritania 22,3% 2,7% 19,6% In energy, agriculture, forestry and land use, industry, waste Mozambique Reduction of 76,5 MT CO2 In energy, land use, forestry, waste Myanmar Lists project and policies conditional on aid In forestry, energy, transport, urbanization, waste, planning Nepal Project and policy targets mainly conditional In energy, forestry, transport, agriculture, waste, building. Niger 38,1% 3,5% 34,6% In agriculture, forestry and other land uses, energy Rwanda No goals, lists project and policies conditional on aid In energy, transport, Industry, waste and forestry Sao Tome & Principe 57 ktCO2eq In renewable energy. Sao Tome & Principe plans to remain a carbon sink country Senegal 15,0% 4,0% 11,0% In renewable energy, buildings, industry, appliances, clean cooking Sierra Leone No quantified goals, long term aim of low emissions Conditional on international support, Sierra Leone maintains low emissions Solomon Islands 45,0% 30,0% 15,0% Energy, transport, renewable energy, land use, land use change and forestry Somalia Lists project and policies conditional on aid Mainly in deforestation, land degradation, efficient charcoal kilns, stoves South Sudan § Lists project and policies conditional on aid In waste, wetlands, forestry, biomass, renewable energy, transport, agriculture Sudan Lists project and policies conditional on aid In energy, forestry, waste Tanzania 10-20% In energy, transport, waste management, forest sector Timor-Leste Lists project and policies conditional on aid In agriculture, energy, land use change & forestry, waste Togo 11,1% 31,1% In energy, agriculture, land use and forestry, buildings, health, coastal erosion Tuvalu Cut emissions from electricity (100%), energy (60%) In energy, agriculture, waste Uganda 22,0% 6,6% 15,4% In energy, forestry and land use change, wetlands Vanuatu Conditional: 100% renewable electricity by 2030 Yemen § 14,0% 1,0% 13,0% In energy efficiency standards, energy use regulations and labeling Zambia 25,0% 47,0% In forestry, agriculture and renewable energy and energy efficiency Notes: # The reduction targets should be read as indicative. Data are not comparable due to differences in reference periods, coverage, and methodologies. § Angola, Eritrea, South Sudan and Yemen have signed but not ratified the Paris Agreement. Source: Interim NDC registry at https://www4.unfccc.int/sites/NDCStaging/Pages/All.aspx, and the World Bank NDC Platform at http://spappssecext.worldbank.org/sites/indc/Pages/INDCHome.aspx 12 Thus, the mitigation part of many LDC NDCs is formulated as lists of mitigation projects for which the LDC seeks finance. Zambia’s contribution, for instance, seeks funding within three large areas, sustainable forest management, sustainable agriculture, and renewable energy and energy efficiency, without detailing the design of mitigation projects (ZEMA 2015). Many non-LDC developing countries have made their NDCs partially or wholly conditional on assistance (Pickering et al. 2019). Indonesia, for example, has set an unconditional target of reducing its emissions by 29 percent compared to a business-as-usual scenario by 2030, and a conditional target of up to 41 percent. Pauw et al. (2020) find that conditional support plans will outrun available resources. Third, NDCs may not be directly comparable because LDCs have different capacities to draw up mitigation plans and use different ways of formulating targets. Bhutan, for instance, discusses mitigation generically and sets a simple target to remain carbon neutral rather than an explicit reduction target. Bhutan is one of very few countries that state they currently absorb more greenhouse gases than they emit. Most other LDCs set quantitative targets expressed as reductions from a business-as-usual scenario. Some NDCs include wording on better planning, institution- building, data collection and similar generalized capacity-building, that many LDCs need to better participate in climate mitigation. LDCs may use the Clean Development Mechanism (CDM) to seek foreign aid for mitigation purposes. The CDM was part of the Kyoto Protocol and allows for high-emitting countries to achieve their reduction commitments by offsetting their emissions in ways that contribute to sustainable development in countries hosting offset projects. Thus, the CDM is a market-based mechanism to channel funds for mitigation to developing countries where emissions reductions may often be achieved more cheaply than in industrialized countries, in the process creating opportunities for projects with economic, social, and environmental benefits. To earn credits from an emission project, an investor must follow strict rules. The investor (a firm or a government) reaches an agreement with actors in a developing country to undertake an emissions reduction project. If the investor’s plan follows the rules and the project is implemented correctly, the CDM Executive Board approves it and grants the investor credits for the emission reduction called Certified Emission Reduction units (CERs). These CERs may be traded. The rules dictate that the investor applies methodologies approved by the CDM, including proving that the reduction would not have happened without the project, a process known as ’establishing additionality’. The investor also needs to follow development project good practice rules, such as consulting with local stakeholders and undertaking an environmental impact assessment. Few LDCs have established CDM projects. A paper from 2014 observes that only 19 of the 46 LDCs had registered a CDM project (Michaelowa, Jember, and Diagne 2014). These 19 LDCs had registered a total of 70 project, while 46 were under validation. Uganda was the only LDC with more than 10 projects. Africa as a continent (non-LDCs and LDCs combined) also fares poorly: African countries had registered 191 projects or 2.5 percent of the global total of 7,399 CDM projects by end-2014. African LDCs had registered only 48 projects, representing just 0.6 percent of the total number of projects (Kreibich et al. 2017). Further, few of these CDM projects have led to the issuing of credits, accentuating the difficulties for LDCs (and Africa) in using the CDM. Michaelowa, Jember, and Diagne (2014) find that CERs in LDCs by end-2012 comprised 1.2 percent of the total of 28.2 million, and more than 60 percent of the LDC CERs were from Angolan gas flaring projects, which are atypical for 13 LDCs. Box 1. Regional climate change mitigiation policy in East Africa In 2011, the then member states of the East African Community (Burundi, Kenya, Rwanda, Tanzania, and Uganda12) agreed on a common East African climate change policy. Until now, the EAC have achieved a series of major policy developments at the regional level, including the approval of an EAC climate change policy (EAC Secretariat 2011a) and a master plan for its implementation (EAC Secretariat 2011b), the adoption of modalities for a common EAC climate change fund in 2013, and the issuance of a Declaration on Food Security and Climate Change in 2010. The climate change policy identifies forestry as the sector with most mitigation potential and also emphasizes the potential of energy, industry, waste management, and agriculture. EAC Secretariat (2011a) encourages the EAC’s members to adopt a series of sector-specific actions: - Forestry: to promote alternative energy sources to wood, and to promote reforestation. - Energy: to develop sustainable energy, to devise a precautionary approach to biofuels taking account of food security, and to improve energy efficiency. - Transport: to improve public transport. - Agriculture: to enhance the carbon storage capacity of farming and livestock systems, and to support best agricultural practices. - Waste management: to develop better waste management. Implemention is a joint responsibility between the EAC member states and the EAC secretariat. To date, implementation of the EAC climate change policy has been slow, partly due to funding constraints. The policy primarily relies on development partner funding. The plan foresees – as do individual country plans as expressed in their submissions during the Paris Agreement negotiations – a strong role for international funding mechanisms like the CDM. Sources: EAC Secretariat (2011a, 2011b). Despite the goal of the CDM to offer both offsetting and promote sustainable development, emerging economies, such as, China, India, Brazil, and Mexico, dominate the market for CDM projects. Winkelman (2011) notes that larger economies cause more emissions and thus have more reduction potential, and furthermore rapidly growing economies have greater demands for the products of CDM projects like clean energy. Low levels of economic activity and low institutional and human capacity explain why LDC projects are so few. Yet, African LDCs and Africa as a whole represent around 5 percent and 11 percent of developing country emissions,13 while having registered only 0.6 percent and 2.5 percent of CDM projects (Kreibich et al. 2017). A number of contributions have discussed why LDCs and other poorer countries are only marginally engaged in the CDM.14 Kreibich et al. (2017) argue that emissions in LDCs are dominated by small and dispersed sources, while the fixed costs of setting up CDM projects are too big to target small emission sources. Likewise, Winkelman (2011), in a case study of Niger, 12 South Sudan joined the EAC in April 2016. 13 Here, developing country emissions are emissions from non-Annex 1 countries. 14 A growing literature discusses why poorer countries do not use the CDM, including Arens and Burian 2012; Arens et al. 2011; Byigero, Clancy and Skutsch 2010; Castro and Michaelowa, 2011; Costantini and Sforna 2014; De Lopez et al. 2009; Jung, 2006; Niederberger and Saner 2005; Michaelowa, Jember, and Diagne 2014; Okubo and Michaelowa 2010; Röttgers and Grote 2014; Spalding-Fecher et al. 2012; Winkelman 2011; Winkelman and Moore 2011; and Wood, Sallu and Paavola 2016. 14 finds that the most common type of CDM project, renewable energy, does not reduce emissions enough to justify CDM registration costs. Hence, LDCs are generally much less engaged in mitigation than emerging and industrialized countries. Nevertheless, the NDCs express the desire of LDCs to achieve vital economic growth by using more carbon-efficient technologies than those used by industrialized and emerging economies at a similar stage of development. Some LDCs do have mitigation plans but these tend to be very modest. For example, Malawi has introduced a carbon tax on vehicles to be paid annually: Malawi kwacha (K) 4,000 (US$ 5.44) for engines below 1500CC, to K15,500 (US$ 21.09) above 3,000CC. The tax is so modest it is unlikely to change behavior and thus to have an environmental impact. The Malawian carbon tax only covers vehicles and no other sources of CO2 or other GHGs. Hence, there is little to suggest that current LDC mitigation plans will have significant impacts on their production and trade. Some LDCs participate in regional climate change initiatives, for example, in East Africa, as discussed in Box 1. Due to funding constraints and constraints similar to those experienced by individual LDCs, regional climate change initiatives in which LDCs take part, as exemplified by the EAC climate change policies, are not a major driver for mitigation efforts in LDCs and thus have so far had very limited impact on LDCs’ trading prospects. 4. Carbon border adjustment The problem of uneven carbon regulation The Paris Agreement allows countries to work from highly different starting points and with different ambition levels toward a common goal of zero net emissions. The Agreement aims at global convergence of mitigation efforts but for the first years of implementation, different starting points and ambitions entail heterogeneity and asymmetry in the use of mitigation instruments. Firms in countries with high mitigation ambitions fear this scenario. Steel makers in the EU, for example, must buy emission allowances in the EU’s Emission Trading System and, if the price of these allowances becomes significant, EU steel makers fear that they will be unable to compete with steel makers in countries not subject to such costly carbon regulation. Thus, firms warn against an erosion of competitiveness causing high-emission production to move to low regulation countries. This move could occur when firms in regulated countries close production sites and move them to unregulated countries or, alternatively, when firms in regulated countries lose market shares to firms in unregulated countries. Environmentalists fear this scenario too, with concerns about carbon leakage. Carbon leakage is when carbon emissions rise in countries not under carbon regulation as a result of carbon regulation in pioneering countries. Carbon leakage is therefore linked to the worries of firms fearing loss of competitiveness. In this case, firms in a country lose competitiveness but without an accompanying gain in the fight against climate change because emissions increase elsewhere.15 15 The mechanism of carbon leakage sketched here can be called the competitiveness mechanism. Additional mechanisms of carbon leakage exist. One mechanism works through the price formation of fossil-based fuel. Lowering fuel demand in regulated countries will lower fuel prices, which will in turn stimulate demand in 15 Border adjustment as an answer to uneven carbon regulation The problem of carbon leakage has been known since people started debating carbon regulation. Many contributors – firms, environmentalists, policymakers, and academic economists16 – identify carbon border adjustment as the solution to the problem of carbon leakage. Carbon border adjustment adds a price to imports corresponding to the price of carbon emissions that domestic firms pay. Thus, this adjustment is designed to nullify the cost advantage that firms producing in non-regulated countries enjoy when selling in a regulated country. Border adjustment is often envisioned as a tax levied at the border, but may also occur by other means, for example, in the EU, regulators are currently considering making it mandatory for foreign airlines flying to EU airports to buy carbon emission allowances. This would in theory have the same effect as levying a tax at the border. Despite strong support for border adjustment in countries that have contemplated carbon regulation, few carbon border adjustment schemes have been put in place. Pauer (2018) studies California’s emission trading system, which includes a border carbon adjustment to electricity imports from other US states and argues that this is the only example of a workable carbon border adjustment system. Why are such schemes so little used? Carbon border adjustment can be politically controversial, challenging to reconcile with international trade law, and administratively demanding to implement. Firms and countries likely to be hit are strongly against it – exemplified by the strong resistance the EU met in 2012 when it proposed to include foreign airlines in its ETS. Following resistance from especially the US and China, the EU is so far only including flights within the European Economic Area (the EU, Iceland, Norway, and Switzerland) in its ETS. Border adjustment also meets resistance from those concerned that the scheme may be misused for protectionist purposes (Jensen 2009, Messerlin 2012). The difficulties of reconciling border adjustment with international legal frameworks – rules of the WTO and obligations of climate change agreements – is an issue that has generated a substantial literature trying to guide practitioners on the design of a scheme. Many believe that carbon border adjustment could be made compatible with WTO rules if designed in the right way.17 Any new scheme is, however, very likely to be challenged in the WTO dispute settlement system. The legal unregulated countries, thus cancelling parts of the carbon saving made in countries introducing regulation. Another mechanism works through capital markets where capital seeks unregulated markets due to increased profit opportunities there. We do not consider these alternative mechanisms here because our focus is on the potential impacts of one particular mitigation instrument – border adjustment – on developing countries, and not on carbon leakage per se. Border adjustment measures address carbon leakage through the competitiveness mechanism. Arroyo-Currás et al. (2015) provide an analysis of different mechanisms of carbon leakage in global energy markets. 16 Support for carbon border adjustment from economists is strong. Many see border adjustment as an appropriate instrument to correct the market failure of too low prices on carbon emissions. In that way, using border adjustment becomes a textbook example of correcting an externality through the use of a Pigouvian tax. A statement of support of carbon taxes, including carbon border adjustment, published in the Wall Street Journal on 17 January 2019 by a large and bi-partisan (in the US political context) group of economists – including 27 Nobel Laureates – exemplifies the support of economists (Economists 2019). 17 The literature on carbon border adjustment and WTO compliance is large. Mehling et al. (2019) and Cosbey et al. (2012) are two examples of major studies that analyze and acknowledge the difficulties of WTO compatibility while providing guidance to policymakers on how to design and implement WTO compatible carbon border adjustment schemes. 16 questions are complex and difficult to solve from the existing rules and case law, thus any scheme will need to be designed carefully and might have to undergo changes before finally accepted in the WTO system. The practical problems of implementing border adjustment are a major hurdle. Authorities have no easy and commonly accepted way to calculate the carbon footprint of a product arriving at the border. Authorities would need to tax this carbon according to the embedded carbon and the taxes (if any) already levied on the product by authorities in the country of production. That calculation is painstakingly difficult to get right. Supply chains are long and complex with sources of carbon emission entering at multiple points and with many possibilities of substitution occurring along the way. Aluminum, for example, has half the carbon footprint if produced from recycled as compared with raw aluminum (Brenton, Edwards-Jones and Jensen 2009) and no customs officer can tell the difference. There is no broad consensus on how to calculate carbon footprints for products and several organizations have developed competing international standards. Will carbon border adjustment affect LDC export opportunities? Before we can answer this question, we must assess how border adjustment may be designed and implemented in practice given that no major economy has yet to establish such a scheme. The political, legal, and practical constraints provide indications of what a system might look like and how it may impact the LDCs. These constraints have been discussed at the political level in the EU and the US and provide ideas of the likely shape of a future scheme, as does the rich academic and policy literature on carbon border adjustment. Below we first review the EU and US political considerations made over the past 15 years. Second, we discuss the nature and implications of the issue of WTO compatibility. Third, we identify the practical problems like methodological difficulties and data limitations associated with designing and running border adjustment schemes. Fourth – drawing on our discussion of political, legal, and practical challenges - we establish a series of principles that policy practitioners will likely follow when designing a carbon border scheme, and, fifth, we finally discuss how this scheme might impact LDC export opportunities. Lessons from the EU and the US The European Commission has announced that a carbon border adjustment mechanism (CBAM) will be part of the ambitious European Green Deal (European Commission 2019, 2020). This scheme will be the first in which major border tax adjustment will be used in climate policy. While the Commission has yet to provide guidance on what the scheme will look like, we discuss the likely underlying principles.18 The Commission has stressed that the scheme will comply with WTO rules and be implemented gradually, starting with selected sectors and gradually extended. In European Commission (2020), the Commission considers a series of policy options: • The policy instrument, with three main options being considered: (i) a carbon tax on selected products – both on imported and domestic products, (ii) a new carbon customs duty or tax on imports, or (iii) the extension of the EU ETS to imports. ▪ The methodology to use to assess the carbon content of imports. The ETS uses EU benchmarks, but if the ETS is extended to cover (some) imports, firms could be allowed to 18 The Commission has stated that full details of the EU scheme will be available in 2021, at the earliest. 17 certify products of lower carbon content than these benchmarks. Alternatively, authorities could also look at ways of encouraging the measurement of product carbon content. Some observers argue that allowing producers to use real data increases the chances that the WTO would find the scheme consistent with WTO rules (Mehling et al. 2019b). ▪ Which sectors to include in the scheme by studying carbon leakage and identifying vulnerable sectors. ▪ Exemption for LDCs or for trading partners with carbon regulation equivalent to the EU’s. In its most recent communication, the Commission does not discuss possibilities for such exemptions, but they have been commonly discussed at the highest policy levels among European leaders in the past.19 ▪ Social and employment impacts. The Commission plans to undertake analysis of such impacts but does not specifically mention impacts on LDCs or other vulnerable countries. ▪ The administrative costs of the CBAM scheme. The Commission notes that “to apply a carbon border adjustment, traders and authorities may need to determine the greenhouse gas emissions associated with the production of an imported product. In order to limit the administrative burden, the options considered should build on existing and agreed methodologies. Ancillary verifications, controls and audits of installations in third countries may be needed. The design of the measure will take into account the need to minimize administrative burden”. There is no specific link between the discussion of administrative costs and LDC concerns. Mehling et al. (2019a) suggest that an EU system is likely to be formulated as an obligation of importers to procure emission allowances through the EU’s ETS rather than as a tax. A tax would require unanimous support in the EU Council – politically difficult to achieve – while the inclusion of importers into the ETS requires only a qualified majority. Bringing importers under the ETS will also allow the EU to use existing structures of administration, including for gathering data. Finally, some categories of developing countries might be exempted from the scheme. Previous proposals from France and Germany have called for exemptions for LDCs specifically, and many observers are convinced that the EU will exempt LDCs (Keane 2012; Mehling et al. 2019b). Many other developing countries could argue that they are also vulnerable. Keane (2012), for instance, argues that members of the Alliance of Small Island States (AOSIS) should qualify for exemption. Yet, WTO rules will make it difficult to exempt groups other than LDCs, which are officially recognized under WTO rules, while respecting the WTO’s national treatment principle.20 Cosbey et al. (2012) argue that not exempting vulnerable developing countries could create a conflict between a scheme and international environmental policy, like the UNFCCC’s principle of common but differentiated responsibility. Using this principle, the Paris Agreement brought developing countries into global climate policy by establishing a regime allowing countries to join from different starting points and with different ambition levels. 19 However, how comparison of the outcomes of different regulatory systems will be undertaken has to be defined and will be demanding. 20 This may become more sensitive as large countries with significant manufacturing exports, such as Bangladesh and Myanmar, graduate from LDC status in the coming decade. 18 Carbon border adjustment has been long in the making, but it looks like the EU will design and implement the first large scheme within the next five years. It is unlikely that LDCs will be much affected initially. The scheme will likely target only high-emitting firms already included in the EU’s ETS. If the price of allowances goes up in the ETS, that could even generate an advantage for the few firms in those industries located in LDCs, like Mozal, the Mozambican aluminum smelter. Other vulnerable countries, like small island states or other categories of countries that do not share the status of the LDCs in the WTO framework, could be subject to border adjustment. This vulnerability could prompt such countries to develop their understanding of emission patterns and any options to develop equivalent carbon regulation domestically to seek exemptions under future carbon border adjustment schemes. The EU, however, plans to expand its carbon regulation into non-ETS sectors, and action in sectors like agriculture and light manufacturing is likely in the longer run. With an established policy for border adjustment, carbon regulation in new sectors will surely lead to calls for additional border adjustment schemes. On the upside, from the viewpoint of LDC export opportunities, the exemption of LDC exports might provide incentives to shift production to LDCs. Keane (2012) expects that LDC goods will be exempt from border adjustment, but not services. Some LDCs and other developing countries have strong tourism industries relying heavily on air transport that may be impacted by current and emerging schemes. There is also the possibility that the EU’s tightening of carbon regulation will lead importers and manufacturers in the EU to put pressure on their overseas suppliers of raw materials and inputs to reduce carbon emissions. LDCs may be affected by this process. It will, however, be demanding for LDCs to undertake carbon accounting and verify the results. Activities such as measurement, certification, inspection, and basic research for carbon accounting are not well developed in LDCs. Nevertheless, it is likely that, even in LDCs, a good understanding of the patterns of carbon emissions throughout the relevant value chains and access to carbon accounting and various services of verification will become sources of competitiveness. Figure 1. Carbon pricing proposals in US Congress since 1990 25 20 15 10 5 0 Note: Years represent the two-year sessions of the US Congress. Source: Price on Carbon (2020). 19 The US Congress has discussed a border carbon adjustment scheme many times. These discussions have been linked to bills on taxing carbon introduced in the Senate, the House, or both. These bills propose to levy taxes or to implement a cap-and-trade system. The website ‘Price on Carbon’ has registered 63 bills introduced in Congress since 1990 (Price on Carbon 2020).21 No bill has passed both chambers. Figure 1 illustrates the evolution of bills. A record number of 20 different bills were discussed during the 111st session in 2009-2010 when the incumbent president Obama strongly promoted carbon regulation. The bill harnessing the most support – the Waxman-Markey cap-and- trade bill entitled ‘the American Clean Energy and Security Act’ – only managed to pass the House narrowly by 219-212 and failed in the Senate.22 WTO compatibility International trade law is often viewed as an obstacle to implementing border carbon adjustment mechanisms, and an extended and extensive debate has been discussed on whether and under which conditions such schemes may be compatible with WTO rules. Here we provide the broad contours of the literature to identify principles that policymakers will use to design WTO-compatible schemes. A carbon border adjustment scheme might breach several WTO disciplines, notably disciplines falling under GATT Article III on national treatment and GATT Article I on most-favored nation treatment.23 Policymakers in the EU and the US have often expressed their desire to design and operate schemes that do not violate WTO rules. To do so, they will either have to look carefully at how to comply with GATT Articles I and III or they must seek an exemption under GATT Article XX, which describes a number of instances where a country can seek exemptions, for example for an environmental measure. Article III:2 specifies that imports cannot be taxed higher than ‘like’ domestic products. Many, but not all, scholars24 believe that two products identical in all attributes except their carbon footprints will be deemed to be ‘like’ products by the WTO Appellate Body (Mehling et al. 2019; Cosbey et al. 2019).25 Thus, according to the national treatment principle enshrined in Article III, an Importer would not be permitted to differentiate between, for example, domestically produced recycled aluminum and imported aluminum produced from bauxite, despite the carbon footprint of recycled aluminum being only half of that produced from the raw material. The scheme could ensure that 21 Some of the 63 bills are linked as a few have been re-introduced after they expired in the previous session of Congress and others build upon each other. 22 In the current 116th Congress that began in 2019, there are eight carbon pricing bills, four of which are bi- partisan and two introduced by Republicans. A trend towards increased bi-partisanship is reflected in the creation of a bi-partisan Climate Solution Caucus in the House in 2016 and a corresponding caucus in the Senate oin 2019. Annex 1 provides a summary of the border carbon measures contained in draft US climate legislation. 23 Carbon border adjustment that subsidizes exports (to compensate for high domestic carbon prices) could also violate the WTO’s Agreement on Subsidies and Countervailing Measures (see Mehling et al. 2019). In this section, however, we focus on the import side. 24 See Hillman (2013) for an example of an analysis concluding that two products identical apart from their carbon footprints could be considered not to be ’like’ products. 25 If a border adjustment scheme requires an importer to buy emission allowances, Article III:4 might be the relevant discipline because this border adjustment could be viewed as an internal regulation. Here too the GATT does not allow discrimination between ’like’ products, although the wording of Article III:4 is somewhat different from Article III:2. 20 imports are not taxed more than similar domestic products, for example by taxing both domestic and imported aluminum according to the average domestic carbon footprint. What could be problematic is if policymakers seek to tax aluminum of different origins according to their actual carbon footprints. Of course, taxing a product by its carbon footprint is exactly the environmental basis of the border adjustment, thus Article III could push policymakers to compromise on the environmental effectiveness of a carbon border adjustment. Article I’s principle of most-favored nation treatment entails that carbon border adjustment cannot discriminate between different trading partners’ ‘like’ products. This could easily happen, for instance if the policymaker exempts certain countries, for example, countries that have signed specific international environmental agreements. This would prohibit the importing country from distinguishing between an imported product which has different carbon intensities according to the country in which it is produced, for example, aluminum from countries that produce from recycled inputs and those that produce from bauxite. The WTO has a special clause for LDCs though, allowing in principle for differentiation between LDCs and other WTO members. There is, therefore, a legal risk that a border adjustment scheme seeking to discriminate between products based on the production method will not be accepted by the WTO’s dispute settlement system. There is little case law accumulated in the WTO on border adjustment26 and what will be accepted or not will in the end depend on the judgment of the Appellate Body legal experts. Alternatively, a country could seek an exemption under Article XX. Paragraph (b) permits WTO members to adopt a policy inconsistent with GATT disciplines if that policy is necessary to protect human, animal, or plant life or health, and paragraph (g) permits an inconsistent policy relating to the conservation of exhaustible natural resources. In addition to justifying a border adjustment scheme under either paragraph, a country would also have to show consistency with the chapeau of Article XX, which dictates that a carbon border adjustment scheme may not arbitrarily or unjustifiably discriminate between countries and may not be used as a disguised trade restriction. Several scholars see clear options for a border measure implemented to combat climate change to be seen either as a policy protecting life and health, and/or conserving natural resources (Mehling et al. 2019; Hillman 2013). This would apply to a scheme that makes the reduction of carbon leakage the objective, rather than competitiveness concerns. But satisfying the chapeau of Article XX may be more challenging and Hillman (2013) observes that very few measures have so far survived this scrutiny. To increase the likelihood of successfully obtaining an exemption under Article XX, a country could emphasize fairness and due process in the development and implementation of border adjustment as supported by the Appellate Body in the US Shrimp case (Mehling et al. 2019). Several ways exist to enhance fairness and due process, like engaging in good faith negotiations with trading partners, allowing for consultation, providing studies on the risk of carbon leakage, establishing criteria for covered sectors linked to their carbon intensity and their trade exposure, etc. Mehling et al. (2019) suggest strengthening fairness by allowing foreign producers to calculate and submit the real carbon footprint of their products rather than apply benchmarks. Many scholars emphasize that options exist for making border adjustment WTO compatible (although they do not necessarily agree on how). Even if options do exist, the lack of WTO case law 26 But see Mehling et al. (2019) and Pauwelyn (2013) for two recent reviews of the legal debate, which, of course, do not necessarily agree. 21 on border adjustment makes legal challenges very likely. A border adjustment scheme immune to further challenges in the WTO will only arise following a careful design and a period of litigation and adjustments of the scheme establishing the meaning of GATT Articles I, III, and XX and incorporating that understanding into the scheme. This is likely to happen with the EU scheme and if the US proceeds to implement its own carbon border adjustment mechanism. Practical considerations of establishing carbon border adjustment The practical challenges of designing and implementing carbon border adjustment arise from the technical complexity of calculating carbon footprints. Policymakers need such data to set a price on the carbon embedded in imports. The data needs to be accurate and comparable. But carbon footprinting methodologies differ, and data is often a problem. LDCs have a stake in how future border adjustment schemes address practical problems. Common methodologies to calculate carbon content may be more difficult to apply in LDCs. Likewise, LDCs may face more severe data limitations. Plassmann et al. (2010) and Brenton, Edwards-Jones and Jensen (2010), for instance, apply a popular carbon footprinting methodology, the UK PAS 2050 (that has subsequently become the basis for an ISO standard), and show that many subjective choices must be made in the calculations, often product or firm-specific. The standards leave plenty of room to arrive at different estimates by simply varying plausible assumptions and using different yet respectable data sources. Many carbon footprinting analysts apply lifecycle analysis (LCA) to estimate carbon footprints. Competing methodologies include ISO’s carbon footprinting standards entitled ISO 14067 and the Greenhouse Gas Protocol’s standards. Policymakers will need to consider systems’ boundaries (where the analysis should begin and end), the data on which to base carbon content calculations, and benchmarks if data is missing or a producing country is not willing to submit data. Regarding boundaries for the analysis, in carbon reporting, emissions are categorized in three groups: Scope 1, 2, and 3. Scope 1 emissions are those that arise direct from owned or controlled sources, such as fuel combustion for onsite gas boilers and operation of company vehicles. Scope 2 emissions are indirect emissions related to purchased energy, heat, steam, or cooling, for example, those from the operation of machinery using purchased electricity. Scope 3 emissions are all other indirect emissions, including purchased inputs from external suppliers, employee commuting, business travel, waste management, distribution, energy used in consuming the product (for example, electricity used in boiling water for cooking vegetables), and so on. Wide boundaries, for example by including scope 3 emissions, make for a more complete analysis, but at the cost of complicating the calculations and the need for data. Many LDC export commodities are traded in value chains without traceability, making it particularly hard to travel up the value chain to quantify all potential carbon sources. Other LDC exports, like clothing and footwear, are traded through value chains where inputs are shifted frequently and the big brands procuring the exports may often shift suppliers. Cosbey et al. (2012) attempt to identify practical ways to create a carbon border adjustment scheme. They recommend that scope 1 and 2, but not scope 3, emissions should be included to ease the calculations. Emissions from transport to market, and from consumption and disposal of goods, 22 should also be excluded. Proponents of excluding scope 3 emissions often do so because they expect a scheme would cover mainly heavy industrial products, electricity and heating installation, and similar activities in which most of the carbon is emitted at the production site. For other products, like agriculture and light manufacturing, most of the emissions are located upstream in the value chain. Which data and benchmarks to use are subject to debate. The better the data – more precise, from more points in the value chain – the better the estimation will be but the more demanding and costly it will be to do the carbon accounting. Condon and Ignaiuk (2014) mention that one option is to impose the same monitoring and reporting requirements on importers that domestic producers are subject to. This approach would allow the calculation of real emissions made by a firm or product base and levy the corresponding tax. An alternative would be to use industry-wide average emission baselines from the origin country and apply those to imported products (Monjon and Quirion 2011). This would alleviate the difficulties of collecting emissions data in the producing countries. Another approach – discussed by Godard (2007) and Ismer and Neuhoff (2007) – would be to base emissions on the best available technology in use globally. Monjon and Quirion (2011) propose using defined EU product-specific benchmarks rather than difficult-to-determine average best available technology. Such benchmarks were established by the European Commission to set the appropriate amount of free allowances in the EU ETS. These benchmarks are based on the 10 percent best performing installations in the EU. Mehling et al. (2019a) propose that adjustments should be based on benchmarks that reflect average performance, best available technology, or worst available technology in a sector, rather than the actual measured carbon footprint. But foreign producers should be allowed to demonstrate better than benchmark performance, by submitting calculations of the actual carbon footprint of their products. This debate reflects the challenges of data access and methodological difficulties in calculating carbon emissions at the import product level. In general, the more you use benchmarks and readily available data rather than tailor-made product-specific data, the easier and less costly the calculations will be. But ease and low costs may come at the costs of incentives and WTO legality. To give the individual producer and importer an incentive to lower emissions, calculations must reflect the emissions of the individual firm. If the benchmark is crude – like industry averages – the individual firm may face no incentive to lower emissions. With crude data, the incentive is mainly on the producing country, for example, to introduce carbon regulation corresponding to that of the importing country to escape an arbitrary and possibly restrictive border tax adjustment. Some analysts of border adjustment argue that the use of environmental input-output analysis circumvents some of the methodological difficulties inherent in applying LCA. Mehling et al. (2019), in their analysis of the prospects for applying border adjustments, propose that a growing number of multiregional input-output databases contain detailed information on the carbon footprint of traded goods in different sectors and thus help ease data limitations. They somewhat optimistically suggest that the “availability and quality of emissions data, methodologies for measurement emissions, and administrative and technical capacity for implementation [of a border carbon adjustment scheme] have greatly improved” (p. 436). 23 Environmental input-output analysis has some methodological advantages, such as avoiding the need to set boundaries. It is a top-down type of analysis in which detailed footprints are derived from aggregate input-output data. In contrast, LCA is a bottom-up type of analysis where carbon footprints are calculated using finely granulated micro-level data (Kjaer et al. 2015). Input-output analysis will be more relevant for a border adjustment scheme that requires rough sectoral carbon emissions data to levy carbon taxes corresponding to industry averages either in the importing country or in the origin country. But a border adjustment scheme that offers the option to overseas producers to prove the actual carbon footprint of their product may be more effective in achieving its underlying objective of reducing emissions. In which case the scheme will require some form of LCA. Studies of carbon border adjustment schemes rarely take LDC conditions into account and indeed often appear ignorant of them. It is unclear whether input-output analysis really lessens the methodological and data problems or whether input-output analysis has its own methodological and data problems when applied in border adjustment schemes. Environmental input-output analysis has been little applied in LDCs, thus we can expect severe data challenges. The multiregional input- output databases described by Mehling et al. (2019) do not sufficiently cover LDCs. Applying input- output data derived for high and middle-income countries to LDCs may severely misrepresent the nature of technology, production, and production linkages in low-income countries. Finally, there have been some proposals to establish border adjustment without detailed carbon accounting. Some French proposals, referenced in Wiers (2008), propose using very crude data, such as country-level data on overall carbon emissions per capita or per unit of GDP. Stiglitz (2006) argues that unregulated emissions should be seen as subsidies and suggests that WTO members could apply anti-dumping or countervailing duties on countries without sufficient carbon regulation. While that approach appears to situate the problem of unregulated emissions within WTO law, it does not solve the problem of how to measure and determine lower-than-acceptable carbon emission regulation. The determination of anti-dumping cases in the WTO is often technically complicated, as would be the determination of insufficiently regulated emissions. One could imagine a system – not without political and technical challenges – that establishes general rules and where only grave infractions become subject to trade instruments like border taxes. Most disputes could be addressed within WTO committees or through bilateral consultations. Such a system would be in line with Stiglitz (2006) but its effectiveness would depend on a general acceptance of carbon regulation, if the system is not to become overburdened. Only if most WTO members support the general rules, and infractions are rare, will such a system work. How carbon border adjustment might affect LDC export opportunities From this discussion of the political, legal, and practical problems of designing and implementing a carbon border adjustment mechanism, some common principles emerge that are likely to be reflected in actual schemes: • The scope will be limited to carbon-intensive industrial products, most notably cement, aluminum, steel, and similar products, and possibly some chemicals, paper, and others. 24 Limiting the scope to known high-emitting products will make a scheme administratively easier and an exemption under GATT Article XX more likely.27 • LDCs are likely to be exempted. • WTO compatibility is more likely if imported products are taxed no more than domestic products on a volume or per unit basis. Taxing the carbon footprint of imports will likely be challenged and the WTO could find that products distinguishable by their carbon footprints are ’like’ products. • Differentiating imports by country would make compliance with WTO rules more difficult. • Practical concerns of implementation will entail that border adjustment mechanisms be based on readily available benchmarks, such as domestic industry averages, best available technology, or worst available technology, instead of carbon footprints measured at the product level. • Exporters in overseas countries may be given the option of demonstrating the carbon footprint of their products using methodologies defined in relevant standards, but this will require access to extensive data and recognized certification. • The strong political interests in carbon border adjustment and the potential need to seek exemption under GATT Article XX should ensure that there is a fair and transparent process for designing and implementing border adjustment schemes. Many economists28 have expressed support for border adjustment as an appropriate instrument to correct the market failure related to carbon emissions. However, they have been absent from the discussion of the practical problems of designing and operating a border adjustment scheme, even though such issues threaten to undermine the use of carbon pricing as an effective policy instrument to address climate change. Similarly, the practical problems of implementation have tended to be overlooked in the initial EU and US policy proposals on carbon regulation. LDCs have not been involved in discussions of carbon border adjustment. LDCs are poorly placed to participate in policy discussions and negotiations on future border adjustment schemes. Countries implementing border adjustment will likely seek to apply a fair and transparent process, but the capacities of LDCs to participate in such a process are very low. Many LDCs required foreign assistance to develop their NDCs during the Paris Agreement negotiations. Many policy proposals – but not all – plan to exempt LDCs from border adjustment. Making such an exemption WTO compatible should be relatively straightforward. Nevertheless, a number of LDCs29 are in the process of graduating and will likely be subject to future border adjustment schemes. In addition, LDCs may be indirectly impacted, for example, if they export intermediate inputs to 27 Proposals such as the Waxman-Markey 2009 bill mention agro-processed goods and light industrial products but the inclusion of these will add much complexity to the schemes. 28 A statement of support of carbon taxes including carbon border adjustment published in the Wall Street Journal on 17 January 2019 by a large and bi-partisan (in the US political context) group of economists – including 27 Nobel Laureates – exemplifies the support of economists (Economist 2019). 29 Angola is scheduled to graduate in 2021, Bhutan in 2023, and Sao Tome and Principe and the Solomon Islands in 2024. While there are important lessons from those countries that have or are graduating in the next few years, their relevance is somewhat limited for a large country with substantial manufacturing activity such as Myanmar. A number of other countries, all in the Asia-Pacific region, have met the criteria for graduation at least once and are on track to leave in the coming years, including countries that would face similar issues to Myanmar such as Bangladesh, Lao PDR, and Nepal. 25 countries that are subject to a carbon border adjustment tax and they import carbon-intensive products from countries that implement carbon regulation. If LDCs were not exempted, their current exports would be little directly affected by border adjustment plans that are limited to carbon-intensive products. However, inclusion of products from sectors such as mining, agro-processing, and light manufacturing, as in the Waxman-Markey 2009 draft bill, would have significant impacts on LDCs. LDCs with significant exports that are air freighted or with large tourism sectors would be affected if aviation becomes subject to carbon border adjustment. Applying border adjustment to products from LDCs will be problematic if calculations of the actual carbon footprint are required. LCA methodologies and environmental input-output analysis are poorly developed in LDCs, as they have little experience with such tools and are not well placed to participate in the development of internationally accepted methodologies. Likewise, data limitations are very severe. There is extremely limited availability of services, such as relevant consultancy and certification that are needed for firms to calculate, verify, and submit carbon footprints for their exports, and they will likely have to be procured through costly foreign providers. 5. Green liberalization Despite years of negotiations to bring trade barriers on green goods down, goods such as windmills and solar panels are still often taxed at the border. Nearly 20 years ago, the WTO launched negotiations on liberalizing trade in environmental goods and services under the Doha Round. The idea was to lower the costs of acquiring and using environmental technologies to combat a range of environmental problems including climate change. Developing countries could gain better access to markets for green goods and acquire green technology more easily and cheaply. Thus, these negotiations would free trade, promote development, and better preserve the environment causing, among others, de Melo and Solleder (2018) to label the Doha Round as a potential triple win. You might add a fourth possible win, that is the avoidance of a clash between two global legal regimes, the WTO rules and the emerging global rules for environmental protection, including the Sustainable Development Agenda of the United Nations, established in 2015, and the Paris Agreement signed in 2016 to combat climate change. The stalemate of the Doha Round has prevented a global agreement on trade and the environment. A group of mainly high-income WTO members have instead negotiated a plurilateral agreement, the Environmental Goods Agreement (EGA).30 The parties had hoped to finalize negotiations during the last negotiating round, which took place between December 3 and 4, 2016, with a view to finalizing the talks in 2017, but failed to reach agreement mainly because the negotiating parties could not agree on what constitutes a green good. Since then, plurilateral negations have not resumed. Nevertheless, green liberalization is likely to resurface, in a new round of plurilateral negotiations, in regional and bilateral trade agreements. Could developing countries, including LDCs, benefit from some forms of green trade liberalization? A critical issue is how green goods are defined. For example, one commonly used list includes mainly industrial products (the APEC list) while a second 30 The EGA is linked to the WTO framework because, as a plurilateral agreement, the liberalization of trade measures like tariff cuts could be extended to all WTO members if a ’critical mass’ (say 85 -90 percent of world trade in the concerned goods) is reached. 26 list adds ‘environmentally preferable’ products; those produced in ways that are better for the environment, such as organic cotton and other agricultural products produced in environmentally specific ways. In the talks in the WTO Committee on Trade and Environment, Brazil has pushed to categorize biofuels as a green good, but many industrialized countries, such as in the EU, have not wanted to discuss agriculture in the trade and environment negotiations. de Melo and Sollender (2018) find that developing countries have an interest in engaging in green liberalization when the talks include environmentally preferable products, but much less so when green goods are defined by the APEC list. Tariffs are still a significant barrier to trade for environmentally preferable products, while the tariff levels for the goods on the APEC list are only on average 0.5 percent with few tariff peaks. Baltzer and Jensen (2015) found that the focus of the plurilateral EGA on the APEC list of green goods would entail negligible impacts for LDCs. Lower trade barriers on green goods could allow LDCs to increase their exports but the impact would be small since these goods currently account for about 0.5 percent of their total exports, and only a third of these exports are destined for countries negotiating the EGA. Further, LDCs already have tariff-free access to most markets in EGA due to unilateral preferential trade arrangements. If LDCs were to participate in the EGA and reduce their own tariffs on green goods this could provide better access to green technologies. The average LDC tariff on such goods is 8.6 percent, with some tariff peaks up to 40 percent. Yet, other factors are constraining the transfer of technology through imports of green goods. In Myanmar, for example, the study found that adverse government policy and low private investment constrain renewable energy technology transfer. Both of these studies found that LDCs could benefit if green liberalization includes environmentally preferable products. Including agricultural products is, however, politically difficult. Furthermore, trading environmentally preferable products requires agreement on sustainability criteria and standards. Successful implementation is based on traceability in the value chain and the necessary trading infrastructure, such as certification and inspection services to ensure the product is genuinely environmentally preferable. For example, tracing organic cotton to verify it was produced according to relevant standards for organic produce and that it is not mixed together with conventional cotton. Baltzer and Jensen argue that LDCs might want to join the EGA negotiations if they become a forum for developing standards and sustainability criteria. 6. Sustainability standards Governments may use sustainability standards to reduce carbon emissions from particular products. The way such standards are formulated, and the verification mechanisms established to make sure producers follow them will determine market access, including for LDCs. The EU’s Renewable Energy Directive came into force in 2010 and promoted the use of biofuels as a form of renewable energy. Biofuels qualified for tax incentives if they offered a minimum of 35 percent carbon emission saving compared with fossil fuels at the start, before rising to 60 percent in 2018. The directive defines standards for biofuels and introduced a mechanism for reporting the emission savings. Biofuel producers and importers bear the responsibility for demonstrating their compliance with biofuel standards incorporating both environmental and social criteria, while the member states are responsible for verification (Keane 2012). 27 The directive helped form a global demand for biofuels by specifying what the producing firm and country would have to do to get the product recognized as biofuel, and triggered biofuel investments in several countries (Schut et al. 2013). Developing countries with abundant land, suitable rainfall, and cheap labor could be ideal for biofuel production. Mozambique is such a country and in 2012 it enacted the first African biofuel policy framework. This policy was a response to what the government saw as flaws in the EU directive. Both policies attempt to regulate critical issues around biofuel development, including how to undertake carbon accounting to demonstrate the required carbon saving, and meeting social and development criteria, such as avoiding negative effects of large-scale land acquisitions on food security. By December 2008, 17 projects covering 245,404 hectares at a value of US$ 1.3 billion were on the drawing board in Mozambique (Schut et al. 2013). While these plans reflect a true biofuel boom, the Mozambican government identified a total of 7 million hectares as suitable for biofuels (Schut et al. 2013). The Mozambican policy framework, developed with substantial support from foreign researchers, development agencies, and NGOs like WWF, defined national sustainability criteria to regulate biofuel production and to promote its development by branding Mozambique as a responsible producer of biofuel. Schut et al. (2013) observe that the focus on carbon emission reduction and biodiversity in the EU directive did not reflect the focus of the Mozambican government, which was on employment creation, generation of tax revenues, and export earnings. Initially, the government focused on Mozambique’s own need for fuel and hesitated to include criteria for emissions reduction. But during the development of the policy framework the situation changed. Optimism for biofuels waned, the global financial crisis of 2008 caused financial troubles for many investors, and Mozambique discovered both coal and natural gas thereby diverting the attention of the government away from viewing biofuel as a domestic fuel source. Eventually, to facilitate exports to the EU, the framework included an obligation that biofuel producers use an internationally recognized method to calculate carbon emission reductions. Nevertheless, given the broader objectives of the Mozambique government concerning social and economic development the certification programs that would qualify biofuels for the EU market was seen as a necessary yet not sufficient condition for investors to operate in Mozambique. The creation of markets for products characterized by their environmental credentials (environmentally-preferable products) poses the following challenges for LDCs: • knowing the environmental characteristics of their export production vis-a-vis competing supply chains. For climate change, that means knowing how they perform on carbon emissions and where potential reduction possibilities exist. • interacting with policymakers in countries creating sustainability criteria. In the EU biofuel examples, there was a period for comment between publication of the draft and its entry into force, although observers complained that this period was very short (Keane 2012). • representation in databases of carbon emission factors. This is a huge challenge given the dominance of industrialized country concerns in the development of carbon accounting methodologies, the poor capacity of LDCs, and the diverging conditions in LDCs compared with industrialized countries. • development of export products complying with emerging sustainability criteria including climate change-related criteria will be hampered by traditional supply-side constraints in LDC related to the introduction of new products and new technologies. For example, 28 Jatropha, a hardy oilseed crop, was seen in Mozambique as a promising biodiesel source but failed partly due to the poor quality of seeds and poor agricultural extension services that should have supported farmers in cultivating this new crop (Schut et al. 2013). 7. Business carbon management During the past two decades, the public has increasingly demanded that business takes more responsibility in managing environmental problems. Firms in turn have adopted ever more complex policies and codes of corporate social responsibility, including in environmental management. The public is pushing global business on climate change too. Carbon labeling was an early business response to consumer concerns. This is an activity that measures the carbon footprint of a product from the production of raw material and intermediate inputs through the value chain ending with the consumption, disposal, or recycling of the product. The footprint is displayed on individual products, thus allowing consumers choice in reducing their impact on the climate. The UK supermarket Tesco was a frontrunner in the area of carbon labeling. In 2007, its chief executive promised a revolution in green consumption and pledged to label Tesco’s 50,000 own- brand products with the total amount of carbon from production to consumption. Carbon labeling spread quickly and numerous retailers and brands, especially in Northern Europe and the US began counting carbon and displaying their calculations on their products. In 2012, however, Tesco gave up. By then, it had labeled only about 500 products (Financial Times 2012). Many other carbon- labeling initiatives have either been discontinued or have scaled down. Since then, more and more private firms have pledged carbon emissions reduction targets and engage in forms of business carbon management but have abstained from labeling individual products. Tesco’s and others’ carbon-labeling experiences illustrate two issues that continue to pose problems for other types of business carbon management. First, carbon labeling is methodologically complex, often expensive, and difficult to understand for consumers. The data requirements for the detailed carbon accounting necessary are extensive. In principle, carbon footprinting requires measuring emissions at every point along every possible value chain and doing this frequently since suppliers may shift sourcing raw materials to alternative suppliers (each with their own carbon footprint) or adjust production technology. Clearly the calculations have practical limits and the more detailed the more expensive the calculations will be. In Tesco’s case, the costs were too high, and consumers were confused. Tesco displayed the grams of CO2 emitted by individual products. But how do you compare whether to buy a pack of Walkers chips emitting 85 grams or a pint of semi-skimmed milk emitting 800 grams? Tesco did not offer numbers for all their products, so finding directly comparable products was difficult, and even if Tesco had done so would consumers have wanted to do the math? Second, carbon labeling revealed that carbon-footprinting methodologies were predominantly developed by industrialized countries for use in industrialized country contexts (Brenton, Edwards- Jones and Jensen 2009; Plassmann et al. 2010). Thus, using carbon accounting standards risks giving a misleading picture of the carbon footprint of developing country exports. Plassmann et al. (2010) undertook a sensitivity analysis of calculating the carbon footprints of sugar, a typical developing country export, in Zambia and Mauritius. Calculations might be affected by the choice of different methodologies, limited data availability, and uncertainty about key variables. Using one dominant standard – the PAS 2050 developed by the British Standards Institute (BSI) - they found that 29 emissions from changes in land use might dominate the estimated carbon footprint. In developing countries, prior land use is often unknown and, in that case, PAS 2050 prescribes the use of a worst- case benchmark, which in the case studied by Plassmann et al. (2010), increased the calculation by 1900 percent. The problem is acute for developing countries where land with natural vegetation is still being converted and data are often lacking. When land use change is not important, variables such as electricity emission factors, capital inputs, and loss of soil carbon, have significant impacts on the carbon footprint. Plassmann et al. (2010) recommended that standards developers and the users of these standards develop emission factors relevant for tropical countries and bio-regions and encouraged the transparent use of carbon-accounting methodologies, where data sources, uncertainties, and variability are explicitly noted. The costs of carbon accounting for carbon-labeling purposes, the methodological uncertainties, and the difficulties consumers had in using the displayed information led to a global fall in the use of carbon labeling. Few products are labeled today. Yet, calls for carbon-labeling schemes have reappeared in the past couple of years following the rise of climate change concerns. The Danish government in April 2019 shelved plans to develop a government-backed carbon-labeling scheme following protests from retailers and business associations worried about implementation difficulties. Carbon labeling still looms as an idea to activate consumer involvement in the fight against climate change and its resurrection in a different form cannot be excluded. In the past decade, research communities have developed better emissions data that would theoretically enable calculations at lower cost. Different – easier to understand – ways of displaying the information to consumers could be developed, probably in the form of a traffic light system rather than displaying explicit numbers. Business carbon management today has moved from seeking to provide information to, and influence the decisions of, consumers, to being organized at the level of firm management, often having at their disposal a team of experts producing and digesting data from carbon accounting. Today, more firms than ever calculate carbon footprints but instead of using the data to label products, they use it to identify areas in their value chains where they can cut emissions and reduce inputs like heat and electricity. Firms are monitoring their total emissions and assessing the impact of measures to reduce these and to communicate these as total firm savings, often as part of reports on the social and environmental impacts of their business activities. As an example, in November 2019 the Danish retailer Coop launched a new climate strategy in which one of the targets was to reduce indirect emissions from suppliers, transporters, and consumers by one million tons of CO2 by 2030. Coop plans to meet its target by working with “sustainable suppliers” from whom climate-friendly products and production methods will be demanded, and Coop will require all suppliers to develop climate plans. The retailer also plans to support smaller suppliers to convert to more climate-friendly production. Furthermore, Coop will create a “sustainable product catalogue” by creating a climate score card as part of the basis for producing products (Coop 2019). This incorporation of carbon accounting is typical of many, especially European and American, firms. Carbon accounting is the mechanism that allows them to demonstrate that they are ‘green’ and can be trusted at the same time. Two other examples of very large firms are the actions of the leading US retailer Walmart and the consumer goods brand Unilever. Walmart launched its emission reduction project ‘Project Gigaton’ 30 in 2017. The aim is to avoid emissions of one gigaton (one billion tons) of CO2 by 2030 from the global value chains that Walmart controls. To appreciate the challenges facing Walmart and other large firms sourcing from global value chains, it is useful to understand the basics of carbon reporting. As discussed earlier, emissions are categorized in three groups: Scope 1, 2, and 3. Scope 1 and 2 are relatively easy for a firm to report (and easier to work on because they are under the firm’s direct control). Scope 3 emissions are more challenging, both for reporting and achieving emission reductions. The longer and more complex the value chain, the more difficult scope 3 emissions are to deal with. Most carbon emissions that can be traced to Walmart activities are scope 3 emissions. These originate not in Walmart’s distribution centers and stores but during production, transport, and consumption of the many products on Walmart shelves. Working with NGOs, Walmart has identified six areas with the greatest opportunities for reducing emissions: energy use, sustainable agriculture, waste management, deforestation, packaging, and product use. Walmart reports that they have worked with the WWF on the overall concept and design of Project Gigaton, and with other NGOs including the Environmental Defense Fund (EDF), Conservation International (CI), The Nature Conservancy (TNC), the Sustainable Packaging Coalition (SPC), and the CDP, a not-for-profit organisation that runs a leading global disclosure system31, to connect suppliers to measurement methodologies, guidance, and practical tools to help them reduce emissions. Walmart (2109b) has specified accounting methodologies for each area. Connecting with suppliers is key to the success of Project Gigaton in terms of emission reductions. Suppliers set up an account within the project and each year they report their progress according to established measurement methodologies. The suppliers then receive guidance from Walmart and its partners. More than 1,000 suppliers have signed on to Project Gigaton and, in 2018, 380 suppliers reported reducing their emissions by 59 million MT CO2e. In the first two years of the project, suppliers have reported a total reduction of 94 million C02e. Walmart reports that while they plan to enlarge the scope of areas for reduction, they face difficulties. Nearly 80 percent of reported reductions are in energy use, while progress regarding deforestation and sustainable agriculture requires “influencing a disparate set of actors far upstream in the supply chain, addressing interdependencies and barriers in complex social and economic systems, and gaining alignment with others regarding methodologies for measurement and action” (Walmart 2019). Walmart is working with suppliers and NGOs to support the development of tools to enable some improvement in these areas. In its environmental reports, Walmart does not identify the extent to which it is seeking reductions in global supply chains compared to US sources. The emission reduction efforts by consumer goods brand Unilever provide another example of business carbon management and the challenges a large producer faces when seeking to reduce emissions along value chains that often have origins in developing countries. As with Walmart, scope 3 emissions are key to its reduction efforts. Unilever has been calculating its carbon footprint across value chains since 2010 and aims to halve total emissions by 2030.32 It reports that the emissions 31 The CDP was founded as the Carbon Disclosure Project and continues to use its initials but not the full name. 32 https://www.unilever.com/sustainable-living/reducing-environmental-impact/greenhouse-gases/Our- greenhouse-gas-footprint. 31 from its own factories amount to less than 5 percent of the total carbon footprint of its products. Consumer use accounts for more than 60 percent, and those associated with raw materials for about 25 percent. Unilever is measuring emissions across the lifecycle of a group of goods that are representative of the company’s 12 product categories.33 These footprint calculations focus on 14 key countries, namely Brazil, China, France, Germany, India, Indonesia, Italy, Mexico, the Netherlands, Russia, South Africa, Turkey, UK, and the US, which represent around 60–70 percent of total global sales volume. An early evaluation of Unilever’s methodology was generally positive but criticized for the omission of land use changes (Unilever 2011). Incorporating land use changes is challenging and particularly difficult to get right in developing countries. With inclusion, some LDCs might lose access to Unilever-controlled value chains because dominating carbon-accounting standards force the use of very high emission benchmarks in cases where previous land use changes cannot be documented - a common occurrence in LDCs as noted earlier (Plassmann et al. 2010). The experience of firms such as Walmart and Unilever suggest two important issues that need to be addressed: first, to understand emissions along complex value chains and implement appropriate solutions; and second, to convince the public that they are implementing credible measures to solve environmental problems. To address these challenges firms have made joint efforts, sometimes with NGOs, to pool resources and allow independent bodies to contribute to the design and implementation of instruments, thereby hoping to convince the public of the sincerity of the effort. Such joint efforts are known by names such as ‘multi-stakeholder initiatives’, ‘collective governance’ and ‘sustainability roundtables’. Some examples include the Forest Stewardship Council, the Marine Stewardship Council, the Roundtable on Sustainable Palm Oil, the Roundtable on Responsible Soy, the Roundtable on the Better Cotton Initiative, the Rainforest Alliance, and the Extractive Industries Transparency Initiative.34 Sometimes such multi-stakeholder initiatives include public sector bodies as well as international agencies.35 Multi-stakeholder initiatives are voluntary in nature (as opposed to government-enforced regulation) but may establish de facto mandatory market requirements because of the influence of large retailers or brands. Thus, when such initiatives adopt standards and codes, these often end up dominating particular market segments and developing country suppliers may only gain access to these segments by complying with the requirements of the initiative. Businesses often work with initiatives led by NGOs so as to draw on more resources and commit to reduction emissions in a credible way by using internationally accepted measurement standards and methodologies and allowing independent bodies to verify their reductions. The Science-Based Target initiative (SBTi) is one such initiative in which both Walmart and Unilever participate. This is a partnership between three international NGOs – the CDP, the World Resources Institute and the 33 The 12 product categories are: beverages, deodorants, dressings, hair care, ice cream, home & hygiene, fabric solutions, fabric sensations, oral care, savory, skin care, and skin cleansing. 34 Many more exist. A recent paper by Bakker, Rasche & Ponte (2019) mentions 45 different initiatives involving over 9,000 firms in more than 170 countries. 35 The World Bank, for example, seeks to influence a number of multi-stakeholder initiatives to help them learn what works and what does not and to assist them to become more effective. It adopted its first guidance note of engaging with multi-stakeholder initiatives in 2009 (World Bank 2009). 32 WWF – and the UN-linked corporate responsibility initiative UN Global Compact. SBTi assists firms to set and verify carbon reduction targets. SBTi maintains says that 17% of Fortune Global 500 firms have committed to set science-based emissions reduction targets (SBTi 2018a). The targets may be set at different levels of ambition and scope depending on the nature and capacities of the participating firms. When scope 3 emissions cover more than 40 percent of total emissions, firms are obliged to include an “ambitious and measurable” scope 3 target. Clothing brand H&M, for example, aims to be “climate positive” in its entire supply chain by 2040 (reducing emissions by more than their value chains emit), reducing scope 1 and 2 emissions by 40 percent before 2030 and scope 3 emissions from purchased raw materials, fabric production, and garments by 59 percent before 2030.36 H&M has committed to these targets through the SBTi. The difference between carbon-labeling initiatives and the initiatives within business carbon management is the scope of the initiatives and the way the results of carbon accounting are used. Carbon labeling requires an analysis of all carbon emissions and a display of the results on individual products. Business carbon management rarely includes a full overview of a firm’s carbon emissions but focuses on selected segments for which the firm sets emission targets. The analytical results are used to verify progress in complying with reduction targets and are displayed in corporate material. Both approaches rely on globally accepted measurement standards and methodologies. Walmart’s supply chain work, for instance, is based on identifying hotspots, especially energy use, and reducing emissions therein, rather than product carbon footprinting. Carbon accounting is used as a research tool to identify where savings can be made. How firms in industrialized countries address reducing scope 3 emissions may determine future market access for LDCs and other developing countries. There is very little empirical evidence of the consequences of industrialized country firms such as retailers and big global brands exercising pressure on their suppliers to reduce carbon emissions. But it is very likely that the resulting impact on suppliers, including LDC producers, will be significant. The ability to reduce carbon emissions in a verifiable way may become a major source of competitiveness in market segments catering to climate-conscious buyers. The SBTi views the setting of reduction targets for scope 3 emissions as best practice and encourages firms participating in the initiative to follow this approach: “Pressure on companies from investors, customers, peers, suppliers and civil society to fully measure, manage and reduce their impact on the climate continues to increase. Consequently, reporting and reducing scope 3 emissions has become an integral aspect of reporting frameworks such as the CDP climate change questionnaire, the recommendations of the Task Force on Climate-related Financial Disclosures (TCFD), and initiatives to drive ambitious corporate action like the Science Based Targets initiative and WWF’s Climate Savers program.” (SBTi 2018b:10) The initiative offers flexibility in that a firm may commit to influencing only a certain set of actors in their value chain or only certain activities. The general approach is to work on what is feasible. The SBTi recommends that firms set scope 3 targets by screening their emissions along their value chains to gather knowledge on emission hotspots and subsequently work on reductions of emissions from these hotspots rather than the whole value chain. More than 2,800 firms reported measurements of 36 https://hmgroup.com/sustainability/Planet/climate.html. 33 scope 3 emissions to the CDP in 2017, yet only 27 percent of those firms calculated emissions for all categories they considered relevant. Of the firms measuring scope 3 emissions, 368 firms had set scope 3 emission reduction targets. During the process of verification applied by the SBTi some 150 firms had their targets approved as ‘science-based’ (SBTi 2018b). For carbon footprinting, SBTi relies on the standards and methodologies of the Greenhouse Gas Protocol (GHG Protocol), an NGO established by the World Resources Institute and the World Business Council for Sustainable Development. The GHG Protocol has produced one of the three dominant global standards for product-level carbon footprinting. The other two are the PAS 2050, developed by British Standards (BSI), and ISO 14067:2018, developed by the International Organization for Standardization (ISO). All three standards are based on the principles of LCA as expressed in the global LCA standards ISO 14040 and ISO 14044. The GHG Protocol has also produced a standard for scope 3 emission accounting (GHG Protocol 2011) and a tool, the Scope 3 Evaluator37 software, guiding firms through the steps of evaluating the value chain. The standards and methodologies of the GHG Protocol have been developed by actors primarily from industrialized countries and working for firms in these countries. Even the development of the scope 3 standard has been heavily influenced by industrialized countries despite the fact that value chains frequently reach into developing countries and the cheapest options for reducing carbon emissions could reside in the developing country segments of these chains. Complex value chains in developing countries, and especially those in LDCs, often lack traceability and relevant data. Some firms in industrial countries involved in these value chains have recognized this situation and have been involved in an initiative to address these issues and to incentivize firms to reduce emissions in such value chains. A coalition of NGOs (Climate KIC, CDP, Livelihoods Funds, WWF, and the WRI) and the firms Danone and Mars has initiated a program called ‘Value Change’ (Value Change 2018a, 2018b). A number of firms including Barry Callebaut, Ben & Jerry’s, Cargill, General Mills, L’Oréal, PepsiCo and Target, as well as Mars and Danone, are piloting implementation of the program. The Value Change program will test sector-specific applications like textiles.38 To support its work, Value Change has produced a general guide (Value Change 2018a) and will produce a series of sectoral guides. The first of these is a guide on carbon soil sequestration (Value Change 2018b) and work is ongoing on textiles. The objective of the scheme is to allow firms to get credit for emission reductions even if, because of poor traceability, conventional methods for carbon footprinting cannot precisely demonstrate that these interventions are reducing the carbon footprint of the products being purchased through the value chain (Value Change 2018a). For example, a firm procuring coffee from a province in Burundi provides assistance to farmers to reduce their emissions but is not able to maintain traceability to all of the hundreds or even thousands of small farmers supplying coffee. Thus, they can document this reduction and receive credit for it even though they are unable to verify that all the farmers receiving assistance sell their coffee to that particular firm. While the reduction will have taken place, it is not recorded as part of the firm’s carbon footprint based on traditional carbon footprinting (Value Change 2018a). This approach to overcoming the problems of traceability and poor data can be seen as a hybrid of carbon footprinting and a carbon offsetting project. 37 See https://ghgprotocol.org/scope-3-evaluator 38 https://www.goldstandard.org/articles/value-change 34 Danone provides a case study of the application of the Value Change idea: “Danone’s supply chain is complex and global: its milk is sourced directly from 120,000 suppliers, many of which are subsistence farms with fewer than ten cows. […] To generate emissions estimates from its global suppliers, the company developed a framework of representative farms across the globe that are monitored daily and extrapolates data from these farms to similar ones. While recognizing the importance of collecting high-quality data, Danone also sees the importance of identifying which farming practices drive the most improvement in soil carbon sequestration. Instead of striving for exhaustive data collection, which can be costly and overwhelming, Danone’s proxy assessment is more feasible and cost-efficient.” (SBTi 2018b:18) The issue of soil carbon sequestration is a hot topic in the carbon footprinting environment. Danone would like to get credit for savings in soil sequestration, but current standards including the GHG Protocol standards do not yet allow it (Plassmann and Norton 2017). The development of business carbon management focuses exclusively (and not surprisingly) on business issues and on emissions reduction but does not include a development perspective. Documents describing methodologies only refer to development issues in generic terms, as is illustrated by this quote from the Value Change initiative: ”It is beyond the scope of this Guidance to provide specific information on these points. Instead it is recommended to follow guidance such as UNDP Social and Environmental Standards or best practice approaches such as Gold Standard for the Global Goals. This is particularly important for interventions in supply chains in developing countries or in vulnerable communities.” (Value Change 2018a:17). Key issues in developing countries, such as land use change and carbon sequestration, can be critical sources of efficient emissions reduction. Recognizing the developmental dimension should be a fundamental part of understanding the nature of business operations in LDCs and thus in making the right business and climate change mitigation decisions. 8. Concluding discussion Climate change mitigation instruments will create challenges and provide new opportunities for LDC exports. Many LDCs remain dependent on a few exports, mainly commodities and light manufacturing. The industrialized countries, and especially the EU and US, have been major markets, yet, in recent years LDCs have shifted to trade more with other developing countries, in particular, China. A little more than a decade ago, half of LDC exports went to other developing countries, today it is two thirds. Thus, while mitigation policies in the EU are being developed faster, it is those in emerging markets that will likely have the biggest influence on LDC exports (Keane 2013). Instruments that regulate carbon emissions upstream in the value chains may, however, also affect LDC exports to emerging economies that are inputs into products ultimately being exported to the EU and other industrialized countries. In summary, a very significant part of LDC exports may be affected by climate mitigation instruments that influence market access and the competitiveness of participants in global value chains. Three sets of issues arise for LDCs. First, a set of traditional trade policy issues: 35 ▪ Ensuring LDC exemption from the burden of compliance with carbon border adjustment mechanisms. ▪ Including products of importance to LDC exports in green liberalization, such as, environmentally preferable products. ▪ Seeking preferential access for LDC exports of environmentally preferable products in emerging markets, such as China (few emerging countries have participated in the EGA negotiations). ▪ Going beyond tariffs, and addressing non-tariff barriers, such as restrictive regulations and onerous conformity assessments in importing countries that undermine LDC export potential in environmentally preferable products. ▪ Providing trade-related technical assistance to support LDCs in addressing supply-side and behind the border constraints so as to be able to benefit from liberalization of environmentally preferable products. A second set of issues goes beyond trade policy and deeper into the agenda of regulatory development and convergence: ▪ Building capacity of LDCs to participate in policy development in the nexus between trade and the environment. For example, contributing to the development of sustainability criteria of the EU renewable energy directive. ▪ Enhancing the ability of LDCs to participate in the development and use of carbon- accounting standards and methodologies that reflect LDC conditions, especially in sectors such as agriculture and light manufacturing. In particular, there is a need for LDCs to better contribute to the development of the mainly private GHG Protocol standards and methodologies, develop capacity to apply such standards and methodologies, and enhance basic carbon emissions research and relevant services, such as certification. The third set of issues relates to private business carbon management practices: ▪ Allowing LDCs’ interests to be reflected in the development of carbon-accounting standards and methodologies. Initiatives including the Value Change program suggest that these are emerging faster from the private sector, while publicly developed standards, ISO and BSI, are more static. ▪ Ensuring that the realities of conditions in LDCs are reflected in existing databases used with emission reduction initiatives. ▪ Enhancing interaction with the firms, NGOs, and multi-stakeholder initiatives involved in developing and implementing business carbon management. Development agencies, including the World Bank, have a role to play in all of these challenges. The development of multi-stakeholder initiatives, for example, has so far taken place with an absence of development voices. The NGOs involved have had a strong environmental focus. The World Bank has adopted a policy of engaging with critical non-state actors (World Bank 2009) and has had a history of engaging with multi-stakeholder initiatives in palm oil, sustainable fisheries, and other environmental issues, but has not contributed to the development of multi-stakeholder initiatives in business carbon management. It could contribute with convening power, analytics, and policy and investment support. For example, assisting national efforts to establish certification systems and 36 implementing changes in production methods to mitigate climate change in ways that the producer can get credit for and use as a competitive advantage when accessing global value chains. There is a general lack of knowledge of how LDC value chains perform in terms of carbon emissions and how LDCs will fit into future value chains under business carbon management. LDCs need to understand their ‘carbon competitiveness’ in key exports, including the pattern of their emissions compared with those of other countries, and the ability of their producers to benefit from ‘carbon efficiencies’, through instruments such as sustainability criteria, standards, and certification. Further research on this emerging policy issue is required, building on the available evidence of carbon emissions and reduction opportunities in value chains. This evidence 39 is far from comprehensive and more illustrative than representative. This needs to be combined with the much richer literature on standards and developing country exports to address the question of how climate-friendly global value chains can be inclusive global value chains? 39 See, for example, studies for coffee (Maina et al. 2014, Rikxoort 2015, Haggar & Schepp 2012, Sevenster & Verhagen 2010, Killian 2013), cocoa (Ortiz-Rodríguez 2016), fresh produce, textiles and clothing (Bevilacqua et al. 2011, EEA 2019), biofuels (UNCTAD 2014), transport (Monkelbaan 2010, Keane 2012), tourism (Lenzen), and South-South trade generally (Meng et al. 2018). 37 References Arens, Cristof, and Martin Burian. 2012. Integrating Africa’s Least Developed Countries into the Global Carbon Market. Summary of key outcomes. Final project report. Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU), Division E II 7 – European Climate and Energy Policy, New Market Mechanisms. Arens, Cristof; Hanna Wang-Helmreich, Glenns S. Hodes, and Martin Burian. 2011. Assessing Support Activities by International Donors for CDM Development in Sub-Saharan Africa with Focus on Selected Least Developed Countries. Wuppertal Institute and GFA Envest. Baltzer, Kenneth and Michael Friis Jensen. 2016. Study of the Impacts of Green Trade Liberalisation on Least Developed Countries. Department of Food and Resource Economics, University of Copenhagen Bartels, Lorand. 2012. The Inclusion of Aviation in the EUs ETS: WTO Law Considerations. ICTSD Global Platform on Climate Change, Issue Paper No. 6. Brenton, Paul, Gareth Edwards-Jones and Michael Friis Jensen. 2009. Carbon Labelling and Low- income Country Exports: A Review of the Development Issues. Development Policy Review, 2009, 27 (3): 243-267 Brenton, Paul, Gareth Edwards-Jones and Michael Friis Jensen. 2010. Carbon Footprints and Food Systems: Do Current Accounting Methodologies Disadvantage Developing Countries? World Bank, Washington. Byigero, Alfred. D., Joy Clancy, and M. M. Skutsch. 2010. CDM in sub-Saharan Africa and the prospects of the Nairobi Framework Initiative. Climate Policy 10 (2), 181–189. Castro, Paolo, and Axel Michaelowa. 2011. Would preferential access measures be sufficient to overcome current barriers to CDM projects in least developed countries? Climate and Development 3(2), 123–142. Condon, Madison & Ada Ignaciuk. 2014. Border Carbon Adjustment and International Trade: A Literature Review. COM/TAD/ENV/JWPTE(2014)51/ FINAL. Paris: Organisation for Economic Co- operation and Development (OECD). Coop. 2019. ‘Klima’. https://ansvarlighed.coop.dk/hjertesager/klima. Accessed 25 January 2020. Cosbey Aaron; Susanne Droege; Carolyn Fischer; and Clayton Munnings. 2019. Developing Guidance for Implementing Border Carbon Adjustments: Lessons, Cautions, and Research Needs from the Literature. Review of Environmental Economics and Policy Volume 13, Issue 1, Winter 2019, pages 3-22. Cosbey, Aaron; Susanne Droege; Carolyn Fischer; Julia Reinaud; John Stephenson; Lutz Weischer, and Peter Wooders. 2012. A Guide for the Concerned: Guidance on the elaboration and implementation of border carbon adjustment. Policy Report No. 03, November 2012. Stockholm: Enviroment and Trade in a World of Interdependence (ENTWINED). Costantini, V., and Giorgia Sforna. 2014. Do bilateral trade relationships influence the distribution of CDM projects? Climate Policy 14 (5), 559–580. De Lopez, Thanakvaro; Tin Ponlok; Keisuke lyadomi; Sergio Santos, and Bridget McIntosh. 2009. Clean Development Mechanism and Least Developed Countries: Changing the Rules for Greater Participation. Journal of Environment & Development, 18(4), 436-452. 38 de Melo J., Solleder J-M. 2018. Barriers to trade in environmental goods: How Important they are and what should developing countries expect from their removal. FERDI Working paper P235. Deutch, Theodore Eliot. 2019. Energy Innovation and Carbon Dividend Act of 2019. Proposal introduced on 24 January 2019 in the House of Representatives, 116th Congress, 1st Session, H. R. 763. EAC Secretariat. 2011a. EAC Climate Change Policy. EAC Secretariat, Arusha. EAC Secretariat. 2011b. East African Community Climate Change Master Plan 2011-2031. September 2011. EAC Secretariat, Arusha. EPA. 2009. The Effects of H.R. 2454 on International Competitiveness and Emission Leakage in Energy-Intensive Trade-Exposed Industries An Interagency Report Responding to a Request from Senators Bayh, Specter, Stabenow, McCaskill, and Brown. 2 December 2009. Washington, DC: Environmental Protection Agency (EPA). https://www.epa.gov/sites/production/files/2016- 07/documents/interagencyreportcompetitiveness-emissionleakage.pdf. European Commission. 2020. Inception Impact Assessment – Carbon Border Adjustment Mechanism. Ref. Ares(2020)1350037 - 04/03/2020. Brussels: European Commission. European Commission. 2019. The European Green Deal. Communication from the Commission to the European Parliament, the European Council, the Council, The European Economic and Social Committee and the Committee of the Regions. COM(2019) 640 final. 11.12.2019. Brussels: European Commission. Financial Times. 2012. Tesco steps back on carbon footprint labelling. Financial Times January 31 2012. Godard, Olivier. 2007. Unilateral European post-Kyoto climate policy and economic adjustment at EU borders. EDF-École Polytechnique, Cahier No. DDX 07-15. Good, Philip. 2010. Tourism, Aviation and Emissions Trading, presentation made at Caribbean Tourism Summit, 14 March 2010, Brussels, DG Climate Action, European Commission. http://www.onecaribbean. org/content/files/Philip-Good-European-commission.pdf Halim, Ronald A.; Tristan Smith, and Dominik Englert. 2019. Understanding the Economic Impacts of Greenhouse Gas Mitigation Policies on Shipping – What is the State of the Art of Current Modeling Approachers? World Bank Policy Research Working Paper #8695. Hedley, Adam; Nicholas Rock, and Peter Zaman. 2016. First ever global regime for aviation emissions: ICAO adopts Global Market-Based Measure to combat aircraft CO2 emissions. Reed Smith Client Alerts 10 October 2016. https://www.reedsmith.com/en/perspectives/2016/10/first-ever-global-regime-for-aviation- emissions-ic Hillman, Jennifer. 2013. Changing Climate for Carbon Taxes – Who’s Afraid of the WTO? Climate & Energy Policy Paper Series, German Marshall Fund of the United States. Hulac, Benjamin. 2015. Pollution from Planes and Ships Left Out of Paris Agreement. Scientific American published 15 December 2015. https://www.scientificamerican.com/article/pollution- from-planes-and-ships-left-out-of-paris-agreement Ismer, Roland and Karsten Neuhoff. 2007. Border tax adjustment: a feasible way to support stringent emission trading. European Journal of Law and Economics, Vol. 24, No. 2, pp. 137–164. 39 Jensen, Michael Friis. 2009. Leveling or Mining the Playing Field? Implementation Problems of Carbon-Motivated Border Adjustment Taxes. PREM notes Trade December 2009 Number 146. Washington DC: World Bank. Jung, Martina. 2006. Host country attractiveness for CDM non-sink projects. Energy Policy, 34 (15), 2173–2184. Kaufman, Noah. 2018. How the Bipartisan Energy Innovation and Carbon Dividend Act Compares to Other Carbon Tax Proposals. Working Paper, November 2018. Center of Global Energy Policy, Columbia University. Keane, Jodie. 2012. The aviation industry, the EU’s Emissions Trading Scheme and Small and Vulnerable Economies: development-friendly frameworks. ODI Project Briefing No. 76, August 2012. Overseas Development Institute, London. Killian, Bernard, Lloyd Rivera, Melissa Soto, and David Navichoc. 2013. Carbon Footprint across the Coffee Supply Chain: The Case of Costa Rican Coffee. Journal of Agricultural Science and Technology B 3 (2013): 151-170. Kjaer, Louise Laumann, Niels Karim Høst-Madsen, Jannick H. Schmidt, and Tim C. McAloone. 2015. Application of Environmental Input-Output Analysis for Corporate and Product Environmental Footprints – Learnings from Three Cases. Sustainability 7, pp. 11438-11461. Kreibich, Nicolas, Lukas Hermwille, Carsten Warnecke, and Christof Arens. 2017. An update on the Clean Development Mechanism in Africa in times of market crisis, Climate and Development, 9:2, 178-190. Kusova, Beata, and Julien Dufour. 2019. An overview of CORSIA for African airlines. Africa Wings 46 October 2019, pp. 32-38. Larsson, Jörgen, Anna Elofsson, Thomas Sterner, and Jonas Åkerman. 2019. International and national climate policies for aviation: a review, Climate Policy, 19:6, 787-799. Mattoo, Aaditya, and Arvind Subramanian. 2013. Greenprint - A New Approach to Cooperation on Climate Change. Washington, DC: Center for Global Development. Mehling, Michael A., Harro van Asselt, Kasturi Das, Susanne Droege, and Cleo Verkujil. 2019. Designing Border Carbon Adjustments for Enhanced Climate Action. American Journal of International Law, 113 (3) pp. 433-481. Messerlin, Patrick. 2012. Climate and trade policies: from mutual destruction to mutual support. World Trade Review 11 (1), pp.53-80. Michaelowa, Axel, Gebru Jember, and El Hadji Mbaye Diagne. 2014. Lessons from the CDM in LDCs, for the design of NMM and FVA, updated January 2014. LDC Paper Series. Publications and Policy Analysis Unit , European Capacity Building Initiative. Monjon, Stéphanie and Philippe Quirion. 2011. Addressing leakage in the EU ETS: border adjustment or output-based allocation? Ecological Economics Vol. 70, No. 11, pp. 1957–1971. Murphy, Andrew. 2018. Planes and ships can’t escape Paris climate commitments. Transport & Environment 4 May 2018. https://www.transportenvironment.org/newsroom/blog/planes-and- ships-cant-escape-paris-climate-commitments Niederberger, Anne Arquit, and Raymond Saner. 2005. Exploring the relationship between FDI flows 40 and CDM potential. Transnational Corporations 14 (1), 1–40. Okubo, Yuri, and Axel Michaelowa. 2010. Effectiveness of subsidies for the Clean Development Mechanism: Past experiences with capacity building in Africa and LDCs. Climate and Development 2 (1), 30–49. Pauer, S U. 2018. Including electricity imports in California’s cap-and-trade program: A case study of a border carbon adjustment in practice, Electricity Journal 31(10): 39–45. Pauw, W. P., P. Castro, J. Pickering, and S. Bhasin. 2020. Conditional nationally determined contributions in the Paris Agreement: foothold for equity or Achilles heel? Climate Policy 20:4, 468-484. Pauwelyn, Joost. 2013. Carbon leakage measures and border tax adjustment under WTO law. Chapter 15, pp. 448-506 in Geert Van Calster & Denise Prévost, eds.: Research Handbook on Environment, Health and the WTO. Edward Elgar Publishing. Pickering et al. 2019. https://www.devpolicy.org/conditions-and-risks-attached-unpacking- developing-countries-conditional-contributions-to-the-paris-agreement-20190828. Plassmann, Katharina & Andrew Norton. 2017. Recognizing the benefits of above-ground carbon sequestration in the carbon footprint of products derived from woody perennial systems. Carbon Management, Volume 8, Issue 4. Plassmann, K, Norton A, Attarzadeh N, Jensen M.F, Brenton P and Edwards-Jones, G. 2010. Methodological complexities of product carbon footprinting: a sensitivity analysis of key variables in a developing country context, Environmental Science and Policy, 13, 393-404 Price on Carbon. 2020. “Know the Legislation.” Price on Carbon.org. Last modified March 10, 2020. https://priceoncarbon.org/business-society/history-of-federal-legislation-2 Röttgers, Dirk, and Ulrike Grote. 2014. Africa and the Clean Development Mechanism: What Determines Project Investments? World Development 62, 201–212. Sakai, Marco & Barrett, John. 2016. Border carbon adjustments: Addressing emissions embodied in trade. Energy Policy 92, 102-110. Sao Tome & Principe. 2015. Intended Nationally Determined Contribution, September 2015. Democratic Republic of Sao Tome & Principe. SBTI. 2018a. Surge in global business embracing climate science to navigate low-carbon transition. Press release of 13 September 2018. Science Based Targets Initiative. Accessed 9 October 2020. Available at https://sciencebasedtargets.org/2018/09/13/surge-in-global-business-embracing- climate-science-to-navigate-low-carbon-transition. SBTi. 2018b. Value Change in the Value Chain: Best Practices in Scope 3 Greenhouse Gas Management. Version 3.0 November 2018. Science Based Targets initiative (SBTi). Schut, Marc; Nuria Cunha Soares; Maja Slingerland and Gerrie van de Ven. 2013. Mozambique’s policy framework for sustainable biofuels : A reflection on the development of the first African policy framework for sustainable biofuels. Paper presented at ‘Tracing biofuel transitions: policies, practices and phase-outs’, Eindhoven, the Netherlands, Spalding-Fecher, Randall, Amrita Narayan Achanta, Peter Erickson, Erik Haites, Michael Lazarus, Neha Pahuja, Nimisha Pandey, Stephen Seres, and Ritika Tewari. 2012. Assessing the impact of 41 the Clean Development Mechanism. Report commissioned by the high-level panel on the CDM policy dialogue. Stoefs, Wijnand. 2020. High time the maritime sector sailed into the EU carbon market. Carbon Market Watch 27 February 2020. https://carbonmarketwatch.org/2020/02/27/high-time-the- maritime-sector-sailed-into-the-eu-carbon-market Stiglitz, Joseph E. 2006. A New Agenda for Global Warming. Economists’ Voice, July 2006. Unilever. 2011. Peer Review Panel’s comments on Unilever’s ‘Less Environmental Impact’ (LEI) Programme. Available at https://www.unilever.com/Images/peerreviewpanelscommentsonunileverslessenvironmentalim pactprogrammetcm244-409762en.pdf. Value Change. 2018a. Value Chain (Scope 3) Interventions – Greenhouse Gas Accounting & Reporting Guidance. Draft V6 – for testing. September 2018. Available at https://www.goldstandard.org/sites/default/files/documents/201809scope3guidancetestingdraf tv1pdf.pdf. Value Change. 2018b. Value Chain (Scope 3) Interventions – Guidance for Soil Organic Carbon. Draft Version for Consultation, November 2018. Available at https://www.goldstandard.org/sites/default/files/documents/201810valuechaininterventionssoc guidelinesdraft.pdf. Walmart. 2019. 2019 Walmart Environmental, Social & Governance Report. Walmart. Wiers, Jochem. 2008. EU Unilateral Action? Presentation made at the International Workshop on Post 2012 Climate and Trade Policies , Geneva, September 2008. Winkelman, Andrew Gene. 2011. The Clean Development Mechanism and Least Developed Countries: Key Determinants for Enhancing Climate Change Mitigative Capacity with Case Study Analysis of Niger. Ph.D. dissertation. University of Michigan. Winkelman, Andrew Gene, and Michael R. Moore. 2011. Explaining the differential distribution of Clean Development Mechanism projects across host countries. Energy Policy 39 (3), 1132–1143. Wood, Benjamin T., Susannah M. Sallu, and Jouni Paavola. 2016. Can CDM finance energy access in Least Developed Countries? Evidence from Tanzania. Climate Policy 16 (4), 456-473. World Bank. 2009. Guidance note on multi-stakeholder engagement. Washington DC: World Bank. ZEMA. 2015. Zambia’s intended nationally determined contribution (INDC) to the 2015 agreement on climate change. Zambia Environmental Management Agency (ZEMA). https://www4.unfccc.int/sites/NDCStaging/Pages/All.aspx 42 Annex 1: Approaches to Border Carbon Adjustment in Draft US Climate Legislation Two draft bills exemplify the different approaches to border carbon adjustment discussed in the US Congress: the 2009 American Clean Energy and Security Act by Waxman-Markey and the 2019 Energy Innovation and Carbon Dividend Act,40 the first draft climate legislation in a decade with some bi-partisan support. The centerpiece of the American Clean Energy and Security Act would have been a cap-and-trade scheme (like the EU’s ETS) to gradually reduce US total emissions from 2012 to 2050. The bill also contained a series of additional instruments including requiring utilities to procure 20 percent of their electricity from renewable sources and the subsidization of new technologies, such as carbon capture, energy efficiency instruments, renewable energy, and electric vehicles. Mechanisms to address carbon leakage and loss of competitiveness were among the additional instruments. The bill contained two mechanisms to address carbon leakage and competitiveness issues. An ’Allowance Rebate Program’ would provide temporary rebates (guaranteed through 2025 and phased out by 2035) for energy-intensive and trade-exposed industries. The bill defines criteria for energy intensity and trade exposure and, based on these, the Environmental Protection Agency (EPA) identified 46 sectors41 that would be eligible for rebates (EPA 2009). These 46 sectors were spread out across quite diverse industries, many of which were heavy industry, such as cement, iron and steel, and aluminum. But primary production including the mining of iron, copper, and nickel were also identified, as was some agro-processing (malt manufacturing and corn milling). Consumer goods could even be affected as the EPA found that ceramics and yarn spinning would be eligible. The Allowance Rebate Program would have delivered a similar outcome to the practice of supplying free allowances for specified industries in the EU’s ETS. A border adjustment scheme was the bill’s alternative mechanism. This scheme was known as the ’International Reserve Allowance Program’. It would require importers in the same energy-intensive and trade-exposed industries to purchase emission allowances. Naturally, this requirement would raise the price of imports and the program would therefore work as a border adjustment scheme. This scheme would start in 2020 (eight years after the planned start of the cap-and-trade system). It would only be applied if negotiations with trade partners failed and these countries did not introduce equivalent legislation, and if rebates on allowances failed to fully compensate affected industries. The bill only provides a rough sketch of its border adjustment mechanism. For example, it does not specify whether an importer must purchase allowances equivalent to the actual carbon content of imports or the carbon content in comparable domestic production (Mattoo and Subramanian 2013). The bill does not discuss or take into account eventual methodological problems of implementing border adjustment or address the to-be-expected data limitations that many developing countries might face. Likely, the intent of the bill is to postpone decisions on such issues to the implementation stage and the implementing authorities decide on such issues. The Waxman-Markey bill would have exempted LDCs from border adjustment as well as countries that met US requirements for having carbon regulation equivalent to that of the US and ‘countries of 40 This bill was originally introduced by Rep. Ted Deutch with bi-partisan support (two democrat and three republican) co-sponsors (Kaufman 2018). The original 2018 bill expired when the new Congress began in January 2019 after which Rep. Deutch re-introduced the bill as the 2019 Energy Innovation and Carbon Dividend Act. Currently, this bill has 80 co-sponsors only one of which is republican. The 2018 bill had a companion bill in the Senate, introduced by the senators Chris Coons (democrat) and Jeff Flake (republican). The bill is based on a policy proposal elaborated by the Citizens’ Climate Lobby, a climate policy -focused NGO that continues to support the bill. 41 Sectors were defined at the six-digit NAICS level. 43 minimal importance’.42 The bill does not specify in any detail what the requirements would be for a country to be considered as having equivalent carbon regulation. It appears that such requirements could be vary by industry, thus opening up the possibility of not implementing border adjustment for specific industries in a country. The proposed bill would establish a national carbon tax to incentivize business and consumers toward green innovation and consumption and return the proceeds of the tax to consumers in the form of monthly rebate checks. The bill would tax CO2 at the source. The coalmine, the natural gas processing plant, the petroleum refinery, and importers of foreign fuels would all pay the CO2 tax and pass this on to downstream consumers. It would cover nearly all CO2 emissions from the US economy but would not regulate methane emissions or the emissions of most greenhouse gases that are not CO2 from industrial processes. The bill places a separate fee on hydrofluorocarbons (HFCs) emissions. It exempts CO2 emissions from agriculture and the armed forces. Sources of greenhouse gases not covered by the bill, such as motor vehicle emissions, would be subject to other legislation, such as existing EPA regulations. The carbon tax would initially be set at a low level of $15/ton but would increase rapidly to nearly $100/ton in 2030 and could go even higher if total emissions remain above the targets in the bill. These carbon taxes are much higher than the taxes in other proposed climate policy bills recently discussed in Congress (Kaufman 2018). The Deutch bill proposes a border adjustment in the form of a tax on carbon-intensive imports and a rebate to exporters of the same products. Three other carbon tax proposals currently being discussed in the US contain similar border adjustments (Kaufman 2018). The bill stipulates that imports must be taxed by an amount equal to what a producer would have paid had the product been produced in the US, minus any foreign tax on carbon. To make that calculation, the border official would need to know (a) the life cycle emissions of covered fuels (the embedded carbon) of the imported product; (b) the relevant US carbon tax for that fuel; and (c) the foreign (if any) tax on carbon (be it from a carbon tax or a cap-and-trade system) and levy a tax equal to (a) x (b) – (c). Similarly, US carbon-intensive exports would get a rebate of an amount equal to the extra cost they pay for carbon relative to their foreign competitor. The draft bill provides very little guidance on how to make these calculations. The bill only includes a one-page section on methodology. This states that authorities shall use “ reliable methodologies” and, if required data are not available, apply “the best available methodologies for interpolating data gaps”. This is to be done while ensuring that border adjustments are “consistent with international treaties and agreements” (Deutch 2019: 26-27). The Deutch bill does not foresee exemptions for LDCs. The bill does stipulate that a US trading partner may be exempted if it introduces a certain climate change policy. High-emitting countries must have policies to reduce their greenhouse gas emissions by at least as much as the US. Low- emitting countries, such as LDCs, must prevent their greenhouse gas emissions from increasing. This condition will be nearly impossible for LDCs to fulfill as emissions increase from poverty-reducing economic growth. Thus, this is inconsistent with the Paris Agreement, which provides for LDCs to grow their economies in a less carbon-intensive manner than in the past but not necessarily reducing their overall emissions. The draft bill leaves many implementation details to be defined. Most of the text of the bill is devoted to issues other than border adjustment. The bill does not elaborate on: which imported products would be deemed as carbon intensive and thus be subject to a border adjustment tax; which methodology would be used to calculate embedded carbon for imports; how to establish carbon taxes on imports, for example, it could be on a per country, per industry, per product basis; or how to define high- and low-emitting countries. 42 A country responsible for less than 0.5% of global emissions and less than 5% of US imports of an industry. 44