As the fight against climate change intensifies, governments are exploring all available policy tools, including trade policy. But does it actually help? This column examines the case of developing countries fulfilling their voluntary emission reduction pledges under the Paris Agreement. Our findings suggest a limited role for trade policy as a climate strategy. Tariff reforms to reduce trade in dirty goods and border adjustments to mitigate carbon tax leakage often come with high trade costs, but have a minimal impact on welfare or on achieving voluntary emission reductions.
Trade-related climate measures (TrCMs) have surged in the last decade (Xu and Monteiro 2022). With a global carbon tax off the table, this comes as no surprise. Unilateral climate policies have left governments concerned about their free riding and competition consequences (pollution heavens) and are increasingly seeing trade policy as a solution for these ‘carbon leakages’.
Most of the research on TrCMs has been focused on the world’s largest economies, which have led the way in adopting these measures and have been fearing the leakages the most. Results are often mixed and, by in large, suggest a modest impact on global emissions (Copeland et al. 2022). However, there’s a significant gap in research regarding the potential merits of TrCMs for smaller, developing economies.
These economies face distinct policy dilemmas. Because they are small, they have fewer reasons to worry about the second-order effects of carbon leakage, such as lower oil prices. Furthermore, since their share of global emissions is modest and trade is typically a large part of their GDP, the welfare benefits of TrCMs from lower emissions are typically outweighed by trade-related losses.
To better understand these trade-offs, we develop a new general equilibrium trade model (Caliendo et al. 2024). It builds on a well-established framework (Caliendo and Parro 2015) that considers multiple countries, multiple sectors, intermediate goods, input-output linkages, and trade policy. We extend this framework in three important ways:
- We model extractive sectors with producers varying in productivity, using labour, dirty resources, and intermediate goods.
- We consider the differential carbon emissions across sectors and countries, with households incurring disutility from these (global) emissions, treated as an externality following Shapiro (2021).
- We introduce a broad set of taxes on production, consumption, and labour, which allows us to examine trade, emission, and welfare effects of different options of carbon taxes and tariffs on emissions.
We focus on a representative group of 48 medium-sized and small developing countries from Latin America and the Caribbean (LAC), Africa, and Eurasia (63% of all developing countries’ emissions). We examine the role of two frequently discussed TrCMs in meeting their ‘net-zero emissions’ VER targets by 2050 or later: (1) a tariff reform to eliminate potential incentives for trade in dirty goods; and (2) a border adjustment tariff (carbon tariff) in combination with a carbon tax. The model was calibrated using data from 2014 (GTAP-10), covering 104 countries (including the 48 in our sample) across three extractive sectors and 31 non-extractive sectors.
Stand-alone and climate-cum-trade actions
The analysis is organised around two sets of counterfactuals. In the first set, we explore how far trade policy can go alone on mitigation. It is inspired by Shapiro’s (2021) findings that tariff scalation in most countries provides a subsidy to emissions as it favours trade in intermediate, emission-intensive goods. We study two types of tariff harmonisation: (1) targeting the countries’ bilateral mean; and (2) targeting the most-favoured nation (MFN) average applied by OECD countries. This second scenario reflects two practical concerns: existing trade agreements might limit tariff adjustments, and harmonisation by the bilateral means might imply costly tariffs for intermediate goods.
In the second set, we examine trade policy as a supporting role to climate actions designed to achieve net-zero targets. Here, we analyse the impact of a production carbon tax, with and without a carbon tariff to mitigate leakages. These surcharges are levied according to the country-sector direct emission intensity (CO2e tonnes per dollar of output), using a global social cost of carbon of US$ 163 per tonne of emission (EPA 2023). We further explore variations in:
- sector incidence: energy-intensive-trade-exposed (e.g. steel, aluminium, cement) and electricity (EITE-E) sectors versus all sectors;
- type of gas: CO2 versus all gases; and
- international environment: scenarios with and without a large country ‘climate club’.
Acting alone
Figures 1 and 2 present the emissions and welfare impacts of the two sets of counterfactuals, assuming developing countries are acting alone, that is, no carbon taxes or tariffs abroad. The first takeaway on emissions is clear: stand-alone trade policies, as shown in the first two columns, have a limited impact, with tariff reforms yielding very mild effects on emissions. No country comes even close to achieving its net-zero goal.
By contrast, a carbon tax targeting EITE-E sectors, along the lines of current EU policies, has a significantly larger impact on emissions, albeit with variation across regions (third column). Africa shows the lowest effect due to the relatively small size of its EITE-E sectors. When the carbon tax is applied to all sectors and gases (fifth column), LAC experiences the largest reduction, reflecting the greater weight of its agriculture-related emissions. Only Chile and Costa Rica manage to meet their net-zero targets in both scenarios.
As the fourth and sixth columns show, adding a carbon tariff does not significantly affect emissions. This is unsurprising since the tariff objective is not to cut domestic emissions, but to mitigate carbon leakages. When a carbon tax is applied to EITE-E sectors, these leakages (the percentage of emissions cut at home that are offset abroad) vary from 12.8% in LAC, 15.5% in Eurasia and 13.8% in Africa – within the lower end of estimates for developed economies (Bohringer et al. 2022). When a carbon tariff is imposed, these leakages are largely eliminated, however, they account for a tiny share of domestic and global emissions.
Figure 1 Emission effects of developing countries acting alone, by region
Note: Each counterfactual (CTF) is evaluated individually for each developing country in our sample. A few outliers were not displayed: Chile CTF3 and CTF4: -193%, CTF5: -418%, CTF6 -427% ; Costa Rica CTF5: -228% CTF6: -235%.
On welfare, the first two columns of Figure 2 show that tariff reforms have a relatively modest impact. By contrast, a carbon tax on EITE-E sectors (third column) results in overall losses, which are larger when the tax is applied to all sectors and gases (fifth column). This suggests most countries face a trade-off between welfare and emissions. When carbon tariffs are added (fourth and sixth columns), the median marginal welfare effects are small and very heterogeneous, with quite a few countries experiencing losses, particularly in Africa. In addition, even in those countries that stand to gain, the benefits are so small that they are most likely to be offset by the carbon tariff’s prohibitive administrative costs.
Figure 2 Welfare effects of developing countries acting alone, by region
Note: Each counterfactual is evaluated individually for each developing country in our sample.
Large country climate club
What happens if the US, EU, and China form a ‘climate club’, with a carbon tax on EITE-E goods and an EU-like carbon tariff on non-members? Three responses by developing countries are considered: (1) business as usual; (2) each country simultaneously responds with a carbon tax; and (3) each country simultaneously adds carbon taxes and carbon tariffs on all partners.
Figure 3 shows that in the first scenario (column 1), the coalition has little to no impact on developing countries’ emissions, without major carbon leakages. However, when countries respond with some combination of carbon tax and tariffs (columns 2 and 3), they see substantial emission reductions, with similar impacts across regions. As before, carbon tariffs barely affect the results, but this time no country achieves their net-zero targets under any policy response. This stems from the fact that a carbon tax in large economies, everything else constant, fosters trade and production in developing countries, mitigating the impact of the carbon tax.
Figure 3 Emission effects with a large country climate club, by region
Note: Counterfactuals in columns 2 and 3 are evaluated individually for each developing country in our sample.
The welfare results (Figure 4, column 1) show that while the club leads to an overall reduction in real income for nearly all countries, this is, in most cases, outweighed by the welfare gains from a reduction in global emissions. When countries do respond, welfare effects are more heterogenous, with winners concentrated in LAC. Countries in Eurasia and Africa are more exposed to the EITE-E sectors from the coalition. As with emissions, adding a carbon tariff has minimal effect, though it slightly improves welfare in Eurasia.
Figure 4 Welfare effects with a large country climate club, by region
Note: Counterfactuals in columns 2 and 3 are evaluated individually for each developing countries in our sample.
Wrapping up
To achieve their VER pledges, governments in small and medium developing countries should prioritise policies that directly target emissions. TrCMs offer a limited path forward. While they may seem attractive, they often come with trade and welfare drawbacks while delivering negligible emission reductions. Even tariff harmonisation, which minimises these drawbacks, has minimal impact on emissions. Carbon tariffs, with or without a large country climate club, provide little additional benefit to carbon taxes – the only measure that brings these countries closer to their emission targets.
Source : VOXeu
