The Cost of Carbon Leakage: Britain's Carbon Price Support and Cross-border Electricity Trade.

• Author: Guo, Bowei

1. INTRODUCTION

   The legally binding COP-21 Paris Agreement came into force on 4 November 2016. "Its goal is to limit global warming to well below 2, preferably to 1.5 degrees Celsius, compared to pre-industrial levels. To achieve this long-term temperature goal, countries aim to reach global peaking of greenhouse gas emissions as soon as possible to achieve a climate neutral world by mid-century." (1) In response the European Union published its Green Deal with its "ambitious target of a 55% reduction in carbon emissions compared to 1990 levels by 2030, and to become a climate-neutral continent by 2050." (2)

   For economists, the natural policy instrument to reduce C[O.sub.2] emissions is a price on carbon, preferably via a tax rather than a tradable permit, given the persistence of C[O.sub.2] and uncertainties about cost and damage functions (e.g. Nordhaus, 2013; Weitzman, 2015; Andersson, 2019). There are strong arguments for additional performance and emission standards (as distributionally more acceptable, or more acceptable to lobby groups, and as a powerful incentive to develop more efficient and lower emitting technologies, see Stern, 2018). Direct innovation support, or indirect demand-pull through renewables targets also play their part. The EU's Clean Energy Package encourages Member States to support renewable energy at "the lowest possible cost to consumers and taxpayers" using '(M)arket-based mechanisms', such as tendering procedures" (Directive (EU) 2018/2001 [section]19). Mission Innovation and World Bank Group (2019) similarly call for global support for innovation.

   Although a carbon tax may create considerable carbon benefit to the world, its impact can be reduced by leakage through carbon-intensive imports without offsetting measures such as a Border Tax Adjustment on carbon-intensive traded goods (e.g. Babiker, 2005; Elliott et al., 2010; Aichele and Felbermayr, 2015). To address these concerns, the EU has proposed its Carbon Border Adjustment Mechanism as "a climate measure that should prevent the risk of carbon leakage and support the EU's increased ambition on climate mitigation, while ensuring WTO compatibility." (3) Until that has been agreed, regional schemes like the EU Emissions Trading Scheme (ETS) partially mitigate leakage by agreeing a uniform carbon price within the EU for the covered sector (about half total EU's emissions). Initially the EU ETS delivered plausible carbon prices, rising to nearly [euro]30/tonne C[O.sub.2], but with the end of the first trading period in 2007 and no banking, prices fell to zero. The second period started well, but the 2008 financial crisis and increased renewables targets reduced demand for allowances (EUAs), causing prices to fall, reaching their lowest level in 2011.

   The failure of the EU ETS to give adequate, credible and sufficiently durable carbon price signals for long-term investment caused increasing concern. The UK was leading the world in imposing legally-binding emissions targets through the Climate Change Act 2008 (4) and faced an increasingly urgent need for new generation investment. As part of the evolving Electricity Market Reform, in 2011 the UK Government announced plans for a Carbon Price Floor (CPF) from April 2013 to raise the carbon price gradually to [pounds sterling]30/tC[O.sub.2] by 2020 and to [pounds sterling]70/tC[O.sub.2] by 2030, intended to make up for the failure of the EU ETS. The CPF was implemented by publishing a GB (5) Carbon Price Support (CPS) added to the EUA price for generation fuels to increase it to the projected CPF. The CPS grew from [pounds sterling]4.94/tC[O.sub.2] in 2013 to [pounds sterling]9.55/tC[O.sub.2] in 2014, and has been stabilized since 2015 at [pounds sterling]18/tC[O.sub.2].

   Consequently, the total GB carbon cost rose from [pounds sterling]5/tC[O.sub.2] in early 2013 to nearly [pounds sterling]40/tC[O.sub.2] by the end of 2018, and continued to rise once the ETS reforms encouraged the EUA price to increase from 2019. Figure 1 shows the evolution of the (nominal) GB and the EU carbon prices. The two curves start diverging in 2013, with the gap becoming wider in 2014 and 2015. The dashed line represents the GB carbon cost target when the CPF was announced. It was not until late 2018 that the GB carbon cost finally met the initial trajectory, thanks to the reform of the EU ETS, which introduced a Market Stability Reserve that removes excess EUAs and increases its price (Newbery et al., 2019). (6) As the EU's commitment to radical decarbonisation became more credible, the EUA price has continued to rise, exceeding [euro]55/C[O.sub.2] by mid 2021. The UK left the ETS in 2021, but replaced it with its own ETS, trading in mid 2021 at [pounds sterling]50/tC[O.sub.2] (or [euro]59/tC[O.sub.2]), so that the carbon price for GB generation fuel was [pounds sterling]68/tC[O.sub.2] (or [euro]80/tC[O.sub.2]) in mid 2021.

   While the EU ETS harmonizes carbon prices and thus reduces distortions within the EU, it is still prone to leakage to the rest of the world. The main industries affected by carbon leakage are carbon-intensive traded goods such as steel, aluminium and cement (Fowlie et al., 2016). The electricity sector is, however, considerably more carbon intensive and in the EU-28 accounted for just over 20% of total greenhouse gas (GHG) emissions in 2018, with very little decrease since 1990. Figure 2 shows considerable fluctuations for the UK, remaining higher than the EU until the recent sharp decrease as coal has been driven out of the system by the CPS.

   The electricity sector is therefore of central importance when studying the impact of differential carbon prices. It has the added advantage that electricity is not widely traded outside the boundary of the EU, but within the EU, Great Britain (GB) faces potentially a 13% import share (and an actual share of 6.4% in 2018). (7) A study of differential carbon prices within EU's Integrated Electricity Market (8) therefore isolates the impact, and allows us to ignore the rest of the world, except, crucially, for the impact on global emissions. (9)

   This article develops a cost-benefit methodology for quantifying the impact of an asymmetric carbon tax on electricity trade within a closed region such as the EU or North America, illustrated using the GB carbon tax, the CPS. While it is relatively simple to characterize the qualitative impact--an increase in domestic and foreign wholesale electricity prices, an increase in imports, etc., any serious policy analysis also needs to quantify these impacts, to judge whether they are sufficiently large to justify policy action, and that is the purpose of this article.

   We assume that the CPS has a first order impact on global emissions through its impact on electricity prices and generation fuel mix, but we ignore second order effects via possible consequential changes in the prices of other goods. If W is global welfare, then [DELTA]W is the change in global welfare that increases from a fall in total emissions. If the economic cost of carbon (SCC) is C, and deadweight loss is L (whose measurement is described below), then,

   [DELTA]W = ([DELTA]E + [??])x C - L, (1)

   where [DELTA]E denotes the emissions reduction due to changes in GB's fuel mix (holding imports fixed), and e denotes the emissions reduction (or increases if negative) due to GB's increased imports from interconnected countries due to the GB-only carbon tax.

   This article quantifies the costs and benefits of cross-border electricity trade between interconnected countries in the presence of asymmetric carbon taxes during 2014-2020, the entire and complete period when GB participated the EU Integrated Electricity Market. While cross-border trade can deliver appreciable benefits if prices are efficient in both countries, distorted prices in one country can reduce and could even reverse these benefits. It is clearly important to establish whether this is the case and that requires quantifying the impact of the asymmetry in carbon prices. It takes GB as a case study and quantifies the impact of the CPS on electricity prices, interconnector flows, congestion income (from buying low and selling high), and the economic value from trade. It also estimates the resulting deadweight loss and carbon leakage. This has implications for the design and ideally harmonization of EU and UK carbon prices and taxes to improve the efficiency of electricity trade.

   One obvious criticism of the ETS is that any carbon reductions within the covered sector will be completely offset by extra emissions in other sectors or countries, as the ETS sets an overall cap on total EU emissions. A carbon tax without an emissions cap would avoid this waterbed effect. In this article we treat both the CPS and all EU emission allowances, EUAs, as carbon taxes, for several reasons. First, both carbon taxes and emission allowances provide emitters with financial incentives to reduce C[O.sub.2] emissions, or put another way, internalise the externality of C[O.sub.2] emissions. Second, policies introduced after setting the last price cap that subsequently (and unexpectedly) reduced emissions (like the EU Renewables targets) put pressure on the EU to tighten future caps, or to cancel excess EUAs, as with the Market Stability Reserve. In addition, policies that have lasting effects on emissions, such as investment in zero carbon generation that displaces fossil fuels, are included in the trajectory to net-zero by 2050 and will enjoy the rapid increase in EUA prices that reflect that commitment. In this article we therefore treat EUAs as carbon taxes, particularly given the workings of the Market Stability Reserve.

   We estimate that over 2015-2020 when the CPS stabilized at [pounds sterling]18 ([euro]20) /tC[O.sub.2], the CPS increased global welfare by [euro]2.9[+ or -]0.1 billion/yr (mainly through displacing GB coal), but the asymmetric carbon taxes created deadweight losses of...