Border carbon adjustments (BCAs) are national or possibly multicountry trade measures—typically taxes on imports (and sometimes rebates on exports)—intended to support ambitious national climate mitigation policies. They are meant to address part of the problem that ambitious mitigation policies in one jurisdiction can lead to increased emissions in jurisdictions with less ambitious policies (“leakage”). In particular, they address the portion of leakage associated with energy-intensive production moving from areas with more ambitious policies to those with weaker policies (“competitiveness”). BCAs are being discussed as part of broader carbon pricing policies, like the European Union’s Emissions Trading Scheme (EU ETS), which recently put forward a concrete BCA proposal; they have also been described and modeled alongside a domestic carbon tax. Much has been written about the design of a BCA in this world with what we might call “full” carbon pricing.
Yet, nations’ climate mitigation policies may or may not include carbon pricing, and when they do, the carbon pricing is often not comprehensive. In the United States, for example, carbon pricing has been implemented at the state level (California, Washington State, and the northeastern states’ Regional Greenhouse Gas Initiative) but is currently a lower priority in national policy than incentives and regulatory standards. China has implemented an ETS that allocates free allowances based on performance benchmarks like a firm’s production level of electricity or (in the future) other industrial products. That is, the policy might regulate tons of CO2 per megawatt of electricity, per ton of steel produced, or per ton of cement. This is frequently referred to as a tradable performance standard (TPS; see Pizer and Zhang 2018). Even the EU ETS gives significant free allocation to energy-intensive, trade-exposed industries, thereby blunting some of the ETS effects. This raises the question of how a BCA might work with a “partial-price” or “nonprice” policy.
In this paper, we talk about “partial” price policy as implementing an explicit carbon price that is paid on some, but not all of a firm’s actual emissions. Perhaps there is a free allocation tied, one way or another, to production of a given product. This might be explicit, through a tradable performance standard or output-based allocation, or implicit, through a free allocation that helps address competitiveness effects.
We talk about a “nonprice” policy as regulating emissions through some type of non-tradable technical or performance-based standard; there is no observed price. Although it is possible to estimate an implicit price or marginal cost associated with the most recent (most expensive) ton of carbon dioxide reduced, it is not observed explicitly.
In this short paper we outline basic principles of how such partial-price or nonprice policies might equivalently be applied to imports as a BCA. Full carbon-pricing policies (auctioned ETS credits or a carbon tax) typically put an equivalent price on the carbon content of imports, usually with an adjustment for any carbon pricing in the country of origin. In contrast, partial-price or nonprice policies exempt a portion of the carbon content of imported goods before applying any price. Moreover, the price paid on emissions above the exemption should be based on some notion of marginal cost if a market price or tax is not observed. That is, it should be based on the cost of the last ton abated domestically, not the average cost.
Our lens on this issue is an economic notion of roughly equivalent treatment. That is, are exporters to a regulated market facing the same incentives and charges, on average, as a domestic producer? “On average” is a critical term. Unless there is a transparent, fully national climate policy that is easily replicated on imports, the existence of state- (or even local-) level regulation means different producers will likely face different incentives and costs. Even with national regulation under the Clean Air Act, states may have some discretion in their implementation. Or a national regulation may give some deference to the starting point of individual firms in the application of benchmarks. The choice of how to match a range of observed a range of domestic incentives and charges to BCA parameters has consequences that might motivate matching the high or low end of observed values instead of the average.
We note at the outset that we are also ignoring issues of WTO compatibility. This has been discussed elsewhere at length for full-price policies (Hillman 2013; Howse 2021). Partial-price and especially nonprice policies raise even more issues as the treatment of imports, while attempting to mimic domestic policy incentives and costs, is not the same. There may be no explicit domestic charges even as BCAs are implemented as a charge. We leave this for future work.
BCAs raise myriad other design questions, including treatment of exports, measurement of emissions, scope (e.g., are indirect emissions targeted?), types of imports covered, and use of revenue. There is also the question of BCAs’ fairness with respect to developing and emerging economies. We believe these questions apply regardless of whether there are full-, partial-, or non-price domestic policies and we do not attempt to tackle them here (see, e.g., Marcu, Mehling, and Cosbey 2020). Rather, our plan is, first, to review the costs imposed by full-, partial-, and nonprice policies and the application of BCAs in the context of full-price domestic policies. This frames our economic notion of trying to apply equivalent treatment to imports. We then discuss how BCAs could seek equivalent treatment with a partial-price or nonprice policy similar to the notion applied with full-price policies. Finally, we consider how domestic policies intersect with one other, and how BCAs might account for a trade partner’s similar or different policies.
To continue reading, click "download" above.