The Strategic Value of Carbon Tariffs

The full paper was published in the February 2016 issue of American Economic Journal: Economic Policy.

The 21st Conference of the Parties to the United Nations Framework Convention on Climate Change (COP 21), completed at the end of 2015, represents a major event in international climate policy.  The conference concluded with a pledge to pursue policies which will limit global warming “to well below 2 degrees C  … and pursuing efforts to limit the … increase to 1.5 degrees C.”  This climate objective is based on emissions reduction pledges from over 180 countries, and it includes requirements for emissions reporting and pledge updating.

What is less clear from the agreement is how the concrete environmental objectives will be achieved.  Low-carbon technology development, deployment, and finance are all recognized as critical to achieving climate goals, yet many details are left to be sorted out.  Indeed, the Paris agreenment is based on “Intended Nationally Determined Contributions” (INDCs), national emission targets and actions that arise from national policies.  The agreement is mute about the role of trade measures as part of INDCs.

A central preoccupation of the international climate-change debate is the question of when developing nations should accept binding targets on their carbon emissions. Developing countries argue that, in the near term, it is unfair to ask them to cut back on their emissions without compensation for the effect it would have on their prospects for economic growth.  At the same time, the unilateral carbon policies currently being pursued (or contemplated) by developed countries are likely to be highly inefficient due to the fact that these countries have high abatement costs.

Unilateral policies are also subject to carbon leakage – an increase in demand for carbon pollution outside of the regulated jurisdiction due to a policy’s effect on the prices of pollution-intensive goods and fossil fuels- which reduces the global cost effectiveness of subglobal action even further.  In theory, a global emissions cap-and-trade system (or a global carbon price and a set of transfers) could deliver on the demands of developing countries and control world emissions in a cost-effective way. Yet the international policy process remains far from reaching the level of consensus that would be required to implement such a scheme.

Thomas Rutherford
Thomas Rutherford

Against this background, many policy analysts have noted that trade policies could serve as a means to regulate carbon emissions in countries that have ineffective emission controls. The rationale for these measures stems from the observation that developed countries are currently (and increasingly) net importers of embodied carbon emissions from their developing-world trade partners.  In fact, the European Union has already resorted to the use of a “border carbon adjustment” to regulate emissions in international air travel. Hence is appears that policy makers charged with designing INDCs are left to choose one of two types of measures: (i) a “carrot” — climate finance payments funded by developed countries to incentive low carbon development, and (ii) a “stick” — tariffs imposed on the carbon content of goods imported from unconstrained regions.

One way that trade measures could function as an instrument of climate policy is by directly stifling demand for carbon-intensive goods produced in unregulated countries. In this capacity, carbon tariffs – tariffs levied on the direct and indirect carbon emissions embodied in imported goods – have support from the theory of second-best environmental regulation; if governments cannot regulate foreign emissions at the source, tariffs may be justified from a global efficiency perspective. They also have a clear political attraction to climate-concerned countries as a way to protect the competitiveness of energy-intensive, trade-exposed (EITE) industries.

Implementing carbon tariffs, however, could come with substantial costs. Whether or not they could be designed in a manner that is legal under current international trade agreements is an on-going and intensively contested debate. Moreover, the task of calculating tariff rates based on foreign pollution levels is likely to be difficult and contentious. For both of these reasons, there might be a real risk to disrupting the free trade regime that has emerged under the World Trade Organization (WTO) if climate and trade policies are linked via carbon tariffs.

The effectiveness of tariffs would also be limited to the extent that countries can find alternative, unregulated markets in which to sell their carbon-intensive products.  There is a sizable literature employing economic equilibrium models to evaluate the performance of border carbon adjustments. Many of these studies find that these instruments can be effective at curbing carbon leakage and offsetting competitivness losses but have a limited ability to decrease the global efficiency cost of meeting abatement targets. With respect to carbon tariffs, many also identify a strong distributional effect from their use, shifting the burden of climate policy to the countries subjected to them.

To date, research has focused almost entirely on the effectiveness of carbon tariffs as a form of direct regulation. However, trade measures could also work to control pollution indirectly – as an environmentally sanctioned punishment that speeds the adoption of emission controls in unregulated countries. The burden-shifting effect identified by the quantitative literature suggests that carbon tariffs have the potential to confer substantial trade gains to countries that use them and trade losses to those subjected to them. Thus the threat of carbon tariffs – rather than their actual application – could lead to more effective carbon abatement policy if unregulated countries prefer to adopt domestic emission controls than to face tariffs. It is this role for carbon tariffs that we explore here.

The Montreal Protocol, the international agreement responsible for regulating the use of ozone-depleting substances, is a prominent example of this idea put into practice. While the trade sanctions it specifies have never been envoked, the credible threat of their use is thought to be important in sustaining the widespread participation it has produced. It has argued that this approach may be applicable to international climate negotiations.

On the other hand, countries subjected to carbon tariffs may prefer to adopt countervailing tariffs of their own rather than suffer the cost of emissions regulation, a response that could significantly increase costs of combating global warming. Many policymakers have expressed concern that the specter of the tariffs could disrupt on-going international climate policy negotiations or trade relations.

In this paper we ask: which of these regimes is likely to arise from the self-interested policy choices of countries and what does it mean for the prospect of designing effective international responses to climate change?  To answer these questions we use a computable general equilibrium (CGE) model of the world economy and carbon emissions to generate the payoffs of a policy game.Policy Game Diagram

In the game, we assume that a coalition of Annex-I countries (those countries that agreed to take on abatement responsibilities under the Kyoto Protocol) is commited to reducing global emissions below business-asusual levels. To achieve this goal, the coalition regulates its domestic emissions. Coalition members also choose whether or not to deploy carbon tariffs against non-coalition countries with unregulated emissions. Noncoalition countries may respond by adopting emission regulations of their own, retaliating against the carbon tariffs with countervailing measures, or by taking no action – leaving their emissions unregulated (and suffering the consequences of the carbon tariffs should they be employed.) We consider Nash equilibria of a simultaneous-move game in which no coalition member wishes to change its policy with regard to the use of carbon tariffs and no non-coalition region wishes to change how it responds to the coalition given the policies of other model regions.

In our baseline policy experiments, there is a unique Nash equilibrium prediction. We find that the threat to use carbon tariffs is credible for all coalition regions. In response, China and Russia – two major polluters outside the coalition – respond by adopting binding abatement targets.  All other non-coalition countries prefer to retaliate.

Cooperation by China and Russia lowers the global welfare cost of achieving a 10% reduction in global emissions by one half relative to the case where coalition countries undertake all of this abatement on their own.  China and Russia are motivated to cooperate for two main reasons.  First, they avoid the punishment of carbon tariffs by doing so. Second, they are dependent on the economic performance of coalition economies – as a destination market for their exports and as the origin of imports.  When China and Russia take on some of the abatement responsibility, less is required of coalition countries to meet the assumed global reduction target. In addition, the overall efficiency of the global economy improves when these countries take on more of the global abatement burden because they are the source of low-cost abatement opportunities. Thus, the global pattern of abatement effort moves closer to a first-best allocation.  Both of these effects benefit China and Russia.

These findings are robust to a number of key assumptions regarding the structure of the policy game, the ambition of the coalition in reducing global emissions and key parameter values in the calibration of the economic equilibrium model used to generate the game payoffs.  Our analysis contributes to a large literature on quantitative assessment of border carbon adjustments, as it is the first to consider the role of these instruments as an inducement to cooperation rather than a direct source of environmental improvement. It is also connected to the gametheoretic literature examining whether or not linking trade and environmental policy negotiations can promote more effective international environmental agreements; none of the experiments in that literature attempt to quantify these outcomes or develop applications to the problem of climate change policy.