Joining a carbon-policies coalition: Flexible mechanisms, competitiveness and anti-leakage instruments in Europe

The EU countries have among the worlds’ most ambitious policies aimed at combatting greenhouse gas emissions. The EU 2030 climate and energy framework (EC, 2014) includes targets for greenhouse gas emissions for sources embraced by the Emission Trading System (ETS) as well as for those outside of the ETS (NETS). Emissions mitigation efforts can, however, be counteracted by carbon leakage. For this reason, the EU has introduced anti-leakage policy for the most trade-exposed ETS industries. The 2030 climate and energy framework does, in practice, also allow for non-EU associates. Non-member Norway has decided to link its climate policy to the EU framework. This paper takes a look at costs and benefits of such a strategy for a small, open economy. Why does a small country without right to participate in EU decisions lay its fate in the hands of a larger coalition? Are the decisions of the coalition the best options for the small associate? In this context, we also include an analysis of the particular rules designed to limit carbon leakage. We ask whether the instruments chosen by the EU are the best for fighting carbon leakage cost-effectively – for the EU and for the fellow country, Norway. We also question whether anti-leakage instruments are beneficial for the competitiveness of the trade-exposed industries involved, and what are the repercussions for other industries. The main European facilitator for emission cuts is the ETS. The sector regulated by the ETS must limit its emissions by 2030 with 43% compared to its 2005 level. In addition, the non-ETS (NETS) emissions are to be cut by 30% during the same period. The framework in EC (2014) opens for interactions among non-ETS sources across borders and between the ETS and non-ETS sectors, so-called flexible mechanisms. The designs and the coverage of such mechanisms will be important for the costs of the 2030 goals. Norway has committed to climate policy targets for 2030 in line with the EU (Norwegian Ministry of Climate and the Environment, 2015; UNFCCC, 2015). As Norway is part of EU ETS approximately 40% of the Norwegian carbon emissions are already regulated joint with the EU. Recently, the government has also announced its interest to be part of the NETS Effort Sharing Decision. Within the EU 2030 climate and energy framework, existing instruments that are designed to dampen carbon leakage are intended to be prolonged. Potential policy responses to carbon leakage include border carbon adjustments (carbon tariffs and export rebates), allocation of free allowances or other financial compensation schemes (e.g., Hoel,1996; Fischer and Fox, 2012). While border carbon adjustments have been frequently on the agenda, the main compensation arrangement in the EU ETS system has until now, and will probably still be, free allowances. The European Commission (EC) estimates that 43% of the total amount of allowances will not be auctioned, but freely allocated, during 2013-2020, and predicts a similar share for the period 2021-2030. Furthermore. the revised ETS Directive allows for national state aid schemes that compensate the most electro-intensive and trade-exposed industries for increases in electricity costs as a result of the EU ETS. The aid intensity must not exceed 75 % of the eligible costs incurred in 2019 and 2020. We analyse the economic costs of the 2030 emissions caps under different flexibility regimes to identify the impacts of joining a coalition with allowance trading (full or restricted). Besides welfare impacts for both the EU and the small, open economy joining the coalition, we scrutinise the carbon leakage and competitiveness effects of various flexibility designs. Moreover, we identify the impacts of anti-leakage policies, including the intended compensation policies for the years to come. These include the mechanisms already use in the EU ETS: free allowances and financial compensation for indirect costs of CO2-emissions in the electricity market (higher electricity prices) for energy intensive industries. Two aspects are analysed. First, they are compared to other instruments recommended in the literature (carbon tariffs and export rebates; see, e.g., Fischer and Fox, 2012). Second, we scrutinise whether, as common sense seems to suggest, competitiveness and carbon leakage solutions go hand in hand, which need not be true (Böhringer et al., 2015). We use a three-region (Norway, the EU, rest-of-the-world (RoW)), multi-sector CGE model of global trade and energy established for analysing carbon emission control strategies (see, e.g., Böhringer et al., 2010, for a detailed algebraic description). The CGE model is based on the GTAP 8.0 dataset, which includes detailed national accounts on production and consumption (input-output tables) together with bilateral trade flows and CO2 emissions for up to 112 regions, including Norway, and 57 industries (Narayanan et al., 2012). CGE models build on general equilibrium theory that combines equilibrium assumptions with behavioural modelling of rational economic agents. They provide counterfactual ex-ante comparisons, assessing the outcomes with a reform in place to what would have happened had it not been introduced. The modelled behaviour abstracts from heterogeneity within regions by featuring one representative agent in each region that receives income from the three modelled primary factors: labour, capital, and fossil-fuel resources. Labour and capital are mobile across industries within a region, but immobile across regions. Fossil-fuel resources (coal, oil and gas) are specific to the respective extraction industries of each region. Final consumption in each region is determined by the representative household who maximizes welfare subject to its budget constraint with fixed investment (i.e., a given demand for savings) and exogenous government provision of public goods and services. Consumption of the representative agent is given as a CES composite that combines consumption of energy and other consumption goods. Bilateral trade is specified following Armington’s differentiated goods approach, where domestic and foreign goods are distinguished by origin (Armington, 1969). The dataset includes all major primary and secondary energy carriers: coal, crude oil, natural gas, refined oil products, and electricity. In addition, we separate the main emission-intensive and trade-exposed industries: chemical products, non-metallic minerals, iron and steel products, and non-ferrous metals, as they will be the most affected by emission control policies and the prime candidates for output-based rebates. A balance of payment constraint incorporates the base-year trade deficit or surplus for each region. CO2-emissions are linked in fixed proportions to the use of fossil fuels, with CO2-coefficients differentiated by the specific carbon content of fuels. The model also includes process emissions linked directly to output. Our Baseline scenario for 2030 is based on continuing the energy and climate policy as before the 2030 framework was launched. We base the inputs on (EC, 2016), which projects that about 80% of EU’s committed CO2 abatement from 2005 for 2030 is realised under the continued, old, policy regime. We then look at three Framework scenarios, F1, F2 and F3. In all the Framework scenarios, the EU needs to cut its 2030 emissions from the Baseline with 10% to meet its commitments. This is true for both the ETS and the NETS sector. F1 assumes full cross-border and cross-industry flexibility within the two sectors, but no flexibility across EU ETS and NETS. This results in different CO2-prices in the markets for ETS and NETS. F2 simulates the extreme flexibility regime with full allowance trading among countries and across ETS and NETS sectors, which will result in a common carbon price for all EU emissions sources. F3 still allows for full trading in the ETS, but no trading across countries in NETS nor across the two sectors. The last alternative gives a common price in ETS, but country specific prices in NETS as we assume cost-effective NETS policy within countries. In all scenarios, Norway is modelled as part of the EU. F3 implies a 40% reduction target in NETS for Norway, in accordance with the current proposal of EC. We then look at the impacts of the various components of the carbon policies and how they interact. This is done by introducing into the Framework scenarios EU’s anti-leakage and competitiveness policies successively; first the free allowances in ETS, then the electricity CO2-compensation schemes in the different flexibility mechanisms alternatives. By that, we are able to examine whether the flexibility matters for the effects of leakage and competitiveness policies. We also examine border carbon adjustments as an alternative instrument for reducing leakage. We do sensitivity analyses of the remaining cap from benchmark (increased from 10% to 20%) and for less compensation in the anti-leakage mechanisms. In addition, in order to benchmark the linking up of Norway’s policies to that of the EU, we also simulate a Norwegian act-alone scenario where the 2030 targets are met without linking policies to the EU policy. Our preliminary results indicate that using flexibility mechanisms in NETS will significantly reduce the abatement costs for the coalition as a whole and for Norway, in particular. Abatement costs do, for instance, more than halve from the non-flexible scenario F3 to F1 and are cut significantly more to the fully flexible F2. Even if flexibility is encouraged and some mechanisms will be provided for the period 2021-2030, each country is free to use them or not. Political signals both in Norway and other European countries indicate that several countries intend to heavily rely on unilateral abatement within NETS. The proposed reduction target for Norway is at the maximum (40%) and virtually no mitigation is expected in the Baseline scenario. Therefore, a 40% commitment from the Baseline remains in our Framework scenarios. Moreover, as abatement options are relatively expensive Norway is much better off within a coalition that practices flexibility. The expensive options are due to already very small gas consumption in households and a scattered population. Even though Norway acting totally alone, i.e., dropping out of the ETS, is currently unrealistic, the comparison between the unilateral and collaborating regimes is interesting both as a benchmarking for the linking strategy, and as lessons for other countries not yet determined to collaborate in a coalition or not. The comparison shows that joining the coalition is costly for Norway, if that implies a separate ETS and NETS target. The reason is a large abatement cost wedge between the two sectors. It is, therefore, cost-effective for Norway to abate the lion’s share within industries that are ETS-regulated today. We also find that EU’s anti-leakage policies seem to be relatively costly for the economy as a whole, but mostly, but not always, benefit the individual industries involved. Thus, a welcome auxiliary benefit of reducing carbon leakage is that the policies favour certain industries. In the wake of the Paris agreement, carbon leakage will be far less topical, as a much larger share of global emissions will be subject to some sorts of caps. Expectedly, motivating policy measures by their anti-leakage effect will be less legitimate, yet still tempting, as different (shadow) prices of the caps in different regions will mean loss of competitiveness for firms in ambitious countries like the European. Nevertheless, there is not necessarily a correlation between carbon leakage and competitiveness losses. For Norway acting alone, anti-leakage policies worsen competitiveness of some domestic trade-exposed industries. We find quite heterogenous effects from industry to industry, depending on their electricity intensity, embodied emissions in imports and export shares. Least profitable for the exporting industries are border carbon adjustments, though they more effectively alleviate carbon leakage than the options used in the EU today and suggested for the forthcoming period. An interesting effect of accounting for process emissions is also identified. Anti-leakage policies tend to shift abatement from reducing the output in emission-intensive industries to reducing their energy input. However, when process emissions are accounted for, the emissions from process industries will respond less and abatement costs amplified. Even if carbon leakage is smaller for the coalition than for Norway acting alone, the leakage effects of joining the EU coalition also vary a lot among industries. Not only the total cap, but also the allocation of emissions matters for leakage effects. Norwegian industries with high abatement costs will not abate to the same extent in a coalition. Some of these industries have relatively low leakage effects to the rest of the world, and the reallocation of abatement contribute to increasing leakage.