Towards a more effective climate policy on international trade

The study addresses how to assign national responsibility for global greenhouse gas emissions in a way that rewards actions reducing global emissions and penalizes actions that increase them. Traditional methods—production-based accounting (PBA) and consumption-based accounting (CBA)—either attribute emissions to producers (PBA) or final consumers (CBA), but both fail to consistently credit mitigation-enhancing actions or penalize mitigation-worsening actions. Prior adaptations (e.g., technology-adjusted CBA, TCBA/TCBA*) still fall short, particularly by not fully accounting for the benefits of specialization and trade that can reduce global emissions. The authors propose a new framework, Emission Responsibility Allotment (ERA), that links Ricardian trade theory to environmental accounting to create credits and penalties based on the global emission consequences of trade, thereby providing incentives for countries to engage in “green Ricardian trade.”

The paper situates its contribution within debates on producer versus consumer responsibility and hybrid/shared responsibility approaches. It reviews PBA and CBA as dominant methods, notes data and methodological challenges with trade-adjusted accounts, and highlights critiques that neither PBA nor CBA credit countries appropriately for actions that reduce global emissions (Kander et al.). Proposed adaptations such as TCBA and TCBA* (Kander et al.; Domingos/Domingues et al.) are discussed, with examples indicating they may still penalize emission-reducing trade. The paper references broader literature on MRIO-based carbon footprinting, avoided emissions, and drivers of GHG changes, and policy contexts including Kyoto mechanisms and Paris Agreement Article 6 (ITMOs). It also discusses desired properties for responsibility schemes (e.g., sensitivity/certainty, additivity, monotonicity, scale invariance) and convergence/consistency issues across MRIO databases (WIOD, EXIOBASE, EORA, GTAP).

Conceptual framework: The authors adapt classical Ricardian trade theory to environmental accounting, positing that countries should be credited (penalized) when their trade reduces (increases) global emissions relative to a counterfactual. ERA is constructed by starting from CBA responsibilities and adding credits/penalties computed from ex post comparisons of actual trade to a hypothetical no-trade benchmark where imported final goods are assumed produced domestically by the importer. Operationalization: Using a global multi-regional input–output (GMRIO) model (WIOD, 41 countries/regions, 35 industries) linked to environmental satellite accounts for CO₂ (residence principle), the authors compute industry- and country-specific emission multipliers that give global emissions per unit of final demand. For each bilateral pair (exporter R, importer S) and final product, they calculate the global emissions embodied in the actual traded good (produced in R) and compare it with the emissions if the same good were produced domestically in S. The difference represents the global emission change due to that trade flow: negative indicates emission savings (credit), positive indicates extra emissions (penalty). Summing across bilateral flows and products yields each country’s total trade-induced credit or penalty. ERA for a country is then CBA adjusted by these credits/penalties; by construction, the sum of all countries’ credits/penalties is zero (additivity). Properties: The scheme is designed to satisfy key desired properties identified in prior literature: certainty/sensitivity (credits/penalties respond to actions countries can influence—production technology, consumption mix, trade partners/products), additivity (global emissions sum to national responsibilities), and monotonicity (countries that reduce global emissions more than average receive larger credits). The method focuses on final products; the framework can be extended to intermediate products. Analyses are carried out for 1995–2009; additional greenhouse gases (CH₄, N₂O) in CO₂-eq are explored in the Supplementary Information. Bilateral and sectoral decompositions identify where penalties/credits arise (e.g., electrical and optical equipment).

- ERAs are generally very close to CBAs for most countries and years (1995–2009), implying that trade-related credits/penalties are typically small relative to overall footprints. - Distributional pattern: Richer countries tend, on average, to receive small credits, while poorer countries tend to pay small penalties when comparing ERA to CBA; however, differences are modest and dominated by a few large cases. - Largest penalties (2009): China (+129.17 Mt CO₂) and the USA (+71.58 Mt CO₂) exhibit the largest positive ERA–CBA differences (penalties), largely driven by bilateral trade between them. - Notable credits (examples from 2009): Luxembourg (−3.68 Mt; −46% of its CBA), Denmark (−8.35 Mt; −13%), Ireland (−9.44 Mt; −16%); several smaller EU economies collectively show credits, partly due to intra-EU trade. - Bilateral decomposition (2009): The USA–China trade relationship accounts for an estimated penalty of about 61 Mt CO₂, with major contributions from electrical and optical equipment (c14; ~36 Mt), textiles (c4; ~6 Mt), machinery (c13; ~5 Mt), and manufacturing nec (c16; ~6 Mt). Flows from China to USA and EU are large in ERA terms, with China a net exporter of embodied emissions and the USA/EU net importers. - Sectoral highlights: For China, electrical and optical equipment (c14) imposes a large penalty (~68 Mt CO₂); additional penalties arise in textiles (c4), basic/fabricated metals, and manufacturing nec. For the USA, penalties are notable in electrical and optical equipment (~37 Mt) and manufacturing nec (~13 Mt). India shows a penalty in manufacturing nec (~14 Mt), while RoW shows a penalty in manufacturing nec (~12 Mt) but a credit in water transport (c24; ~29 Mt). - Time pattern: In 1995, most countries exhibited penalties; by 2009, the pattern reversed with more countries receiving credits than penalties, indicating changes in trade patterns and technologies over time. - Policy implication: Because ERA–CBA gaps are generally small, most countries’ trade patterns neither substantially increase nor decrease global emissions relative to the average; this suggests significant potential for targeted “green Ricardian” trade strategies to realize larger credits if environmental considerations were prioritized in trade decisions.

The findings show ERA closely tracks CBA, indicating that current trade patterns do not markedly improve or worsen global emissions relative to average practices. However, ERA adds crucial information: it quantifies the global emission consequences of specific bilateral trade flows and sectors, enabling targeted incentives and policy design. ERA aligns with Paris Agreement Article 6 objectives to ensure overall mitigation in international carbon markets (ITMOs) by providing a robust, additive, and monotonic ex post measure of trade-induced mitigation (or deterioration). ERA can be used to evaluate the mitigation impact of market mechanisms, inform border carbon adjustment discussions, and support global stocktakes by complementing PBA-based inventories with trade-sensitive responsibility measures. The sectoral/bilateral resolution highlights where changes in trading partners, supply chains, or product mixes could reduce global emissions, guiding ex ante strategy. ERA inherits advantages of CBA (equity, consumer responsibility perspective) and its practical challenges, but benefits from improving MRIO data convergence and transparency. Overall, ERA provides a pragmatic bridge between accounting and incentive design, complementing considerations of equity, capacity, and historical responsibility in negotiations.

The paper introduces Emission Responsibility Allotment (ERA), an accounting framework that augments CBA with credits and penalties derived from the global emission effects of bilateral trade, grounded in Ricardian specialization. ERA satisfies key desired properties (certainty/sensitivity, additivity, monotonicity) and can evaluate both ex post outcomes and inform ex ante policy choices. Empirical application to WIOD (1995–2009) shows ERA values are generally close to CBA, with notable penalties for the USA and China and modest credits for many high-income, trade-intensive smaller economies. ERA is well suited to assess and track the overall mitigation impact of Article 6 market mechanisms (ITMOs) and to guide policies aiming at greener trade patterns. Future research should extend ERA to intermediate goods and services, explore shared responsibility variants, incorporate capital stock dynamics and MNE affiliates, and leverage converging MRIO datasets for more granular sectoral and regional analysis to support periodic Paris Agreement stocktakes.

- Data and model constraints: ERA relies on MRIO tables and environmental satellite accounts; results are sensitive to sectoral/regional aggregation, homogeneity assumptions, and the residence vs territorial principle. Typical IO limitations apply (fixed coefficients, lack of substitution, product aggregation bias). - Scope: The main implementation focuses on final goods trade; services and intermediate inputs are discussed but not the primary focus in the main text. - Database variability: Differences between MRIO databases (WIOD, EXIOBASE, EORA, GTAP) can introduce variability, though convergence is improving; choice of database and environmental stressor definitions can affect outcomes. - Measurement noise and text inconsistencies: Some reported figures in narrative sections show inconsistencies typical of large MRIO exercises and editorial artifacts; robust interpretation should rely on provided supplementary data and code. - Policy uncertainty: ERA evaluates ex post impacts and can inform ex ante scenarios, but real-world behavioral responses, policy interactions (e.g., border adjustments), and future technological change introduce uncertainty. - Scale invariance debates: Discussion notes theoretical concerns about scale invariance and aggregation effects in IO-based measures, which may affect strict satisfaction of all proposed properties under certain aggregations.