Unilateral climate policy design: Efficiency and equity implications of alternative instruments to reduce carbon leakage

Abstract

Global cost-effectiveness of unilateral emission abatement can be seriously hampered by carbon leakage. We assess three widely discussed proposals for leakage reduction: carbon-motivated border tax adjustments, industry exemptions from carbon regulation, and output-based allocation of emission allowances. We find that none of these measures amounts to a “magic bullet” when both efficiency and equity criteria matter. Compared to unilateral emission pricing alone, border carbon adjustments are most effective in leakage reduction and promotion of global cost-effectiveness but can markedly exacerbate regional inequality; exemptions and output-based allocation tend to avoid distributional pitfalls but are less effective in leakage reduction and global cost savings; exemptions may even decrease global cost-effectiveness of unilateral emission abatement.

Introduction

Global cost-effectiveness of unilateral emission abatement can be seriously hampered by emission leakage, i.e., the relocation of emissions to parts of the world economy subject to weaker regulation. There are two main channels through which leakage may occur: as unilaterally abating regions reduce their demand for fossil fuels, international fuel prices fall thereby inducing areas with weaker regulations to increase their fuel demand and emissions; in addition, energy-intensive and trade-exposed (EITE) industries of unilaterally abating countries lose competitiveness on world markets when they face higher cost compared to international rivals which incentivizes the relocation of these industries abroad.

In order to reduce leakage and improve global cost-effectiveness of unilateral action, a number of policy measures has been proposed. Principal among these are border adjustments where emissions embodied in imports from non-regulating regions are taxed at the emission price of the regulating region and emission payments for exports to non-regulating countries are rebated (Hoel, 1991, Markusen, 1975). However, border measures are controversial from the perspective of international trade agreements and their political feasibility is questionable. When border measures are unavailable, differential emission pricing in favor of domestic EITE industries may serve as a substitute (Hoel, 1996). In policy practice, the theoretical argument for differential emission pricing often boils down to complete exemptions of EITE industries as a response to concerns on losses of competitiveness and adverse employment impacts. A third proposal involves the allocation of free emission allowances to EITE industries conditional on production. In contrast to auctioning of emission allowances or unconditional free allowance allocation such an output-based grandfathering system effectively works as a production subsidy to recover (part of) losses in comparative advantage (Böhringer et al., 1998). In the more recent climate policy literature this measure is referred to as output-based allocation (Fischer, 2001).

All of these anti-leakage policy measures – border carbon adjustments, industry exemptions, and output-based allowance allocation – are second-best policy instruments. Thus, they induce distortions of their own which must be weighed against the potential efficiency gains they promise. For example, granting exemptions to EITE industries clearly violates the first-best principle of equating marginal abatement cost across polluters. Thus, the increase in direct abatement cost must be traded off with the indirect economic gains from attenuating leakage.

Theoretical and applied literatures tend to focus on the efficiency effects of alternative anti-leakage policy measures, but their burden-shifting implications are likely to be as or even more important for the role they can play in the international climate policy debate. International price changes that are at the core of the leakage problem also produce terms-of-trade (income) effects. Böhringer and Rutherford (2002) show that these terms-of-trade effects can dominate the overall economic impacts for unilaterally acting countries and likewise induce substantial losses or gains to countries without abatement action. One reason that equity issues seem to take a back seat in the discussion of anti-leakage measures is that unilaterally abating regions are viewed as socially responsible forerunners that are willing to take on costs in order to enhance the prospects of global environmental cooperation in a subsequent step. We show that this perspective must be questioned when we account for the burden-shifting effects of anti-leakage measures.

In our paper we use simulations from a multi-sector, multi-region computable general equilibrium model to investigate the efficiency and equity trade-offs in unilateral climate policy design associated with border carbon adjustments, industry exemptions and output-based allowance allocation. We assess how these additional anti-leakage measures perform with respect to leakage reduction, output changes in industries, global cost savings and distributional impacts compared to a reference unilateral climate policy with uniform emission pricing only. Across all climate policy variants we keep global emissions constant at the level which emerges from uniform emission pricing only. Given the global public good nature of carbon emission abatement and the uncertainty of external cost estimates for carbon emissions, the constant global emission constraint allows for the coherent analysis of efficiency–equity trade-offs. In this framework, leakage reduction implies that unilaterally abating regions must cut back domestic emissions to a lesser extent in order to keep with the global emission constraint.

Regarding leakage reduction and the alleviation of adverse EITE output effects, we find that border carbon adjustments are by far the most effective instrument since they directly level the playing field between regulated domestic EITE production and unregulated EITE production abroad. Border carbon adjustments can provide non-negligible global cost savings. When we consider the case of a small abatement coalition (here: Europe) with ambitious reduction targets (here: a 30% cut from business-as-usual emission levels) – two assumptions that should place anti-leakage measures in a favorable light – the cost savings relative to our reference unilateral policy without anti-leakage measure are close to 20%. Cost savings shrink considerably for output-based allocation (ranging only between 1% and 9% as a function of the coalition size and the reduction target) and exemptions may even increase rather than decrease global economic adjustment cost. The poor efficiency performance of exemptions is due to the sharp trade-off between leakage reduction and the increase in direct abatement cost as cheap abatement options in EITE industries are “exempted”. While border carbon adjustments do have some appeal based on their implications for leakage and global cost-effectiveness, they look less promising when their distributional effects are taken into account. In fact, border carbon adjustments work as a substitute for optimal tariffs shifting a larger part of the economic abatement from abating regions to non-abating regions. Since countries contemplating or currently enacting unilateral climate policies are among the wealthiest nations in the world, border carbon adjustments amplify existing income inequalities (Böhringer et al., 2011a). As a consequence, border carbon adjustments fare poorly when our welfare measures account for even a modest degree of inequality-aversion and there is no transfer mechanism in place to compensate losers under the border-tax-adjustment regime. Output-based allocation and exemptions, on the other hand, have only small additional terms-of-trade effects compared to the reference policy scenario and therefore are preferable to border carbon adjustments as one cares for cost distribution.

There is a large body of numerical studies that quantifies the economic impacts of regulatory measures to reduce leakage and thereby enhance global cost-effectiveness of unilateral emission abatement. Many of these studies build on multi-sector, multi-region general equilibrium models which account for price-dependent market interactions and income effects in a comprehensive economy-wide framework. Border carbon adjustments combining import tariffs with export rebates are found to be effective instruments for leakage reduction with scope for global efficiency gains but also with the potential for adverse distributional impacts on developing countries that are subjected to this measure (e.g., Babiker and Rutherford, 2005, Mattoo et al., 2009, McKibben and Wilcoxen, 2009, Dissou and Eyland, 2011, Winchester et al., 2010, Böhringer et al., 2010a, Böhringer et al., 2011b — see Zhang (forthcoming) for a review of the literature). Output-based allocation can help improve the global cost-effectiveness of unilateral action if leakage rates are substantial (Böhringer et al., 1998, Fischer and Fox, 2009) but they are ranked inferior on global cost-effectiveness grounds compared to border adjustments: Böhringer et al. (2011b) show that – as the size of the climate coalition gets larger – the efficiency cost of implicit output subsidies ultimately outweigh the benefits of reducing emission leakage. As to preferential emission pricing for EITE industries Böhringer et al. (2010b) and Böhringer and Alexeeva-Talebi (2012) find justification for emission price differentiation to deter leakage but caution against substantial excess cost as unilateral reduction targets are more moderate and EITE industries get close to exemptions.¹

Our contribution to the economic literature on unilateral climate policy design is twofold. First, we provide a coherent cross-comparison on the economic implications of border carbon adjustments, output-based allocation and industry exemptions as alternative anti-leakage measures. Second, our cross-comparison does not only include global cost-effectiveness as a common evaluation criterion but also addresses burden shifting implications thereby offering important insights into efficiency and equity trade-offs of unilateral climate policy design.

The remainder of this paper is as follows. In Section 2 we provide a non-technical description of the model structure and its parameterization. In Section 3 we lay out our policy simulations and interpret simulation results. In Section 4 we conclude.