Policy options to improve the effectiveness of the EU emissions trading system: A multi-criteria analysis

Energy Policy 57 (2013) 477–490

Contents lists available at SciVerse ScienceDirect

Energy Policy journal homepage: www.elsevier.com/locate/enpol

Policy options to improve the effectiveness of the EU emissions trading system: A multi-criteria analysis$ Stefano Clo b,n, Susan Battles a, Pietro Zoppoli a a b

Italian Ministry of Economy and Finance, Department of the Treasury, Rome, Italy Department of Economics, Business and Statistics, University of Milan, Via Conservatorio 7, 20122 Milano, Italy

H I G H L I G H T S c c c c c

A multi-criteria analysis is adopted to compare policy options to improve the ETS effectiveness. An ETS cap reversible adjustment by a carbon central bank is the first-best option. The establishment of a EU-wide price floor would represent a second-best solution. A national price floor is the worst option but self-interest states may find it optimal. A post-2020 target is not a mutually exclusive option and should be set.

a r t i c l e i n f o

abstract

Article history: Received 28 May 2012 Accepted 9 February 2013 Available online 7 March 2013

This paper considers several policy options which have been proposed to improve the functioning of the ETS. These options require an intervention either on the ETS cap (  30% target, set-aside, carbon central bank, long-term target) or on the carbon price (European and national price floor). We analyse the impact of each policy on the ETS carbon price and emissions. A multi-criteria evaluation method is applied to compare the policy options against a plurality of environmental, economic and procedural criteria. We find that the final ranking depends on the goals to be achieved, i.e., the relative weights attributed to the criteria. When policymakers want mainly to support the carbon price both in the short and long-run, while improving ETS flexibility and harmonization, the CCB and the EU price floor are, respectively ranked as first and second-best options. As the preference for environmental and implementation goals gradually increases, the position of the EU price floor and CCB options tend to invert. The  30% target should be adopted when reducing emissions is the priority goal, while a national price floor is the worst option, in this case. Nevertheless, self-interested States looking for a relatively quick, feasible solution, may find it optimal. & 2013 Elsevier Ltd. All rights reserved.

Keywords: Multi-criteria analysis Carbon central bank Price floor

1. Introduction Since the launching of the ETS in 2005, the carbon price has, on several occasions, dipped beneath the expected level. Following the economic recession, the ETS cap is not stringent anymore and the resulting surplus of allowances will be carried over to future years, lowering the future carbon price. Without a high carbon price and a clear long-term price signal, the ETS may fail to induce emissions abatement by stimulating technological innovation and the adoption of low-carbon technologies (Helm, 2008; European

$ The opinions expressed by the authors in this paper do not necessarily reflect the official position of their respective institutions. We thank two anonymous referees for their useful comments. n Corresponding author. Tel.: þ39 516560014.  E-mail address: [email protected] (S. Clo).

0301-4215/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.enpol.2013.02.015

Commission (EC), 2010b; European Commission (EC), 2011). Against this scenario, several options to support the ETS carbon price have been proposed. They are likely to affect the carbon price and the environmental performance of the ETS in different ways. They have different juridical implications, too, in terms of necessary adjustments to the ETS normative framework and implementation procedures. This paper analyses several policy options that have been proposed to improve the functioning of the ETS: (a) increase in the European emission reduction target ( 30% target); (b) expost ETS cap adjustment (set-aside), that could be permanent (one-off), temporary or progressive; (c) institution of a carbon central bank (CCB); (d) a carbon price floor set either at a national or European level; (e) establishment of binding reduction targets into the future (post-2020). We first analyse the impact of each policy on the ETS carbon price and emissions. Then, building up

S. Clo et al. / Energy Policy 57 (2013) 477–490

478

on Konidari and Mavrakis (2007) (hereafter KM), a multi-criteria evaluation (MCA) method is applied to evaluate and compare the policy options. The MCA is performed in four steps. First step: a comprehensive set of multiple criteria regarding both the economic, environmental and procedural effects of each proposal is selected. Second step: Each of these criteria is weighted. Third step: each policy option is evaluated against the weighted criteria. Fourth step: for any policy option the different grades are summed up, providing a single value which synthesizes its overall performance. The MCA allows for a ranking of the various policy options and makes it possible to identify the most appropriate instrument to improve the effectiveness of the ETS. The paper is organized in the following manner. Section 2 summarizes the inefficiencies that characterize the ETS performance to date. Section 3 introduces the various policy options and analyses their different impact on the carbon price and emissions within the ETS. Section 4 introduces the MCA and describes the criteria selected to evaluate the policy options. Section 5 analyses the techniques we adopt to evaluate and rank the various policy options. Section 6 presents the grades attached to each policy option against the various criteria, while Section 7 describes the final ranking of the policy options and tests its robustness through a sensitivity analysis (SA). Section 8 concludes.

2. State of affairs of the ETS: Problem of price signalling The ETS is characterized by structural rigidity on the supply side, as the ETS cap -the aggregate level of emissions that the ETS sectors can produce- is fixed at a known level, while the carbon market clearing price varies continuously and tends to react strongly to exogenous shocks on the demand side. The economic recession in 2009 caused an unexpected reduction in ETS emissions and, consequently, a structural decrease in the ETS carbon price. Since the ETS cap does not result stringent anymore due to the current surplus of allowances that will be transferred to the third trading period 2013–2020, the future carbon price will be, by all accounts, significantly lower than the expected one1. This event, coupled with the permanence of regulatory uncertainty2 and the lack of a clear long-term price signal, risks negatively affecting ETS installations’ confidence in the ETS and their propensity to adopt low-carbon technology3 (IEA, 2011; European Commission (EC), 2011; Martin et al., 2011). This in turn risks increasing the cost of reaching the European goal of decarbonisation by 2050, since a considerable amount of the capital stock in the energy sector needs to be replaced within the current decade. Moreover, as EU-wide public auctions of allowances are to be held starting in November 2012, the structural reduction in the carbon price is likely to reduce the related

public revenues. Under a political economy perspective, this factor might have also strengthened the regulator’s will to support the carbon price through public intervention on the ETS.

3. Measures for supporting the carbon price and criteria for comparative analysis Several policy proposals to support the ETS carbon price have emerged in the European political climate debate. This section presents these proposals and analyses their impact on the ETS in terms of carbon price and emissions. 3.1. A  30% emissions reduction target The European Commission (EC) has analysed the possibility of increasing the level of ambition of the overall EU emissions target from 20% to 30% by 2020 (European Commission (EC) 2010a,b,c,d) for both ETS sectors and non-ETS sectors4 . Concerning the ETS, a reduction of the cap is expected to increase the average carbon price. This effect is illustrated in Fig. 1, which describes the ETS market in equilibrium given a fixed vertical supply (ETS cap) and a negatively sloped demand, which we assume to be linear. This assumption has been adopted in similar analyses (Sijm et al., 2006; Hepburn et al., 2007; Wirl, 2007) as well as for describing demand in the energy market and in the tradable white certificates market (Sorrell et al., 2009). The results of our analysis do not depend on this assumption. Indeed, we can verify that, while the intensity of the price variation induced by any of the considered policy options depends on the shape of the demand, the ranking in terms of long-term price variation resulting from the comparative analysis among the different policy options does not depend on the shape of the demand. In fact, as the ranking does not depend on the shape of demand, our results are not assumption based. As the linearity of demand is compatible with the comparative results of this paper we can adopt this simplifying assumption. In the short term, when the demand of allowances does not vary, the reduction of the ETS cap (from Cap0 to Cap1) increases the price level and reduces the optimal quantity at equilibrium (from E0 to E1). However, by reducing the ETS cap, the degree of market flexibility remains unchanged compared to the current framework and the carbon price is expected to continue to vary with uncertainty, depending on the variation in the demand of allowances. This effect is shown in Fig. 2: in the long-term, in spite of the lower supply, the uncertain and continuous variation of demand, within the range D2A and D2B, causes a variation of the equilibrium price between p2a and p2b. Thus, this measure is not likely to send a clear price signal. 3.2. Ex-post cap adjustment

1

At the time when the Climate Package was approved and the ETS cap was fixed for the third trading period (2013–2020), a 30 h/t carbon price was expected (European Commission, 2008), as the European economy was in an upswing, ETS emissions were expected to increase and the ETS cap for the second phase had been reduced by the EC. In particular, the 30 h/t carbon price was used in the impact assessment accompanying the revised ETS Directive. 2 Differently from energy markets, the ETS is a market where firms trade an intangible good generated by the regulator. The ETS performance depends on political decisions that shape its institutional framework. Consequently, uncertainty affecting the carbon price can also be regulatory driven (regulatory uncertainty). Regulatory uncertainty concerning the ETS has arisen in a number of instances (Alberola et al. 2008a,b, Chevalier et al., 2009, Alberola and Chevalier, 2009), impacting negatively on firms’ confidence in this mechanism. 3 In particular, Martin et al. (2011), in a regression analysis of the effect of climate policy on innovation at firms’ level, find a significant and robust positive association between firms’ expectations about the future stringency of their cap and clean innovation.

This option consists in imposing an ex-post adjustment of the ETS cap, without affecting the non-ETS emissions reduction target5 . This proposal can be realized by cancelling part of the amount of allowances to be sold by public auction. We can distinguish between a permanent set-aside (option b.1) where allowances are withdrawn as a one-off measure from the market and then definitively cancelled, and a temporary set-aside (option 4 The effects of this option have been recently assessed for each MS (European Commission (EC), 2012). 5 A general proposal along these lines was included in the legislative report of the Industry, Research and Energy Committee of the European Parliament with reference to the proposed energy efficiency directive, subsequent to its deliberations on 28/02/2012.

S. Clo et al. / Energy Policy 57 (2013) 477–490

479

D2A

p Cap 1

D

Cap 0

p

Cap 1

D

E2A

p2A

E1

p1

p1

D2B

p2B

E0

Cap 0

E1 E2B

p0

E0

p3

E3

p0 q1

q1

q

q0

Fig. 1. Short-term market effect induced by ETS cap reduction (  30% target, permanent and temporary set-aside). Source: Own elaboration.

D2A p D

Cap 1

Cap 0

E2A

p2A D2B

E2B

p2B

E0

p0

q1

q

Fig. 3. Long-term market effect induced by a reintroduction of allowances (temporary set-aside). Source: Own elaboration.

the range over which the carbon price could vary. In the case the entire amount of withdrawn allowances had to be reintroduced in the market before the end of the trading period (from Cap1 to Cap0), the carbon price could even decrease significantly to the point p3 (Fig. 3). 3.3. Carbon central bank

E1

p1

q0

q0

q

Fig. 2. Long term market effect induced by ETS cap reduction (  30% target and permanent set-aside). Source: Own elaboration.

b.2), where the allowances that had been initially withdrawn are subsequently reintroduced over the reference period. In order to minimise the potential disruptive impact of the one-off reduction in the supply of allowances on the carbon price trend and its variability, allowances could also be withdrawn from the market progressively at a constant rate, implying a progressive and linear reduction in the ETS cap (option b.3). In the short-term, both permanent and temporary set-aside measures are expected to increase the ETS carbon price via a reduction in allowances auctioned (Fig. 1). By ensuring a progressive reduction in the cap, the progressive set-aside would favor a gradual transition of the carbon price from p0 to p1 with respect to previous options. In the long-run, by opting for a permanent or progressive set-aside, the degree of ETS flexibility would remain unchanged compared to current legislation. Depending on the fluctuation of demand for allowances, the long-term clearing market price could be higher or lower (p2a and p2b) than the average one (Fig. 2). These options would not be any more successful in sending clear, long-term price signals, as the risk of a future surplus or deficit of allowances is likely persist. A temporary set-aside, which foresees the reintroduction of allowances in the market, could even increase

The current carbon price slump and its predictably high volatility in the future are both due to the combination of uncertain demand and rigidity of supply. Indeed, as the ETS cap is fixed ex-ante, supply cannot adjust to changed circumstances, whatever their nature may be (e.g., exogenous shocks on the demand side). A credible, long-term signalling of the ETS carbon price could instead be favored by introducing a mechanism for a reversible adjustment of the ETS cap based on transparent and pre-determined rules. Such a mechanism could be legally entrusted to an independent central authority. Much like a central bank works to stabilize inflation by controlling the money supply, the hypothesized CCB would control the supply of allowances in order to guarantee the stability of the carbon price within a predetermined fluctuation band. (De Perthuis, 2011). The CCB would increase the quantity of auctioned allowances whenever the clearing price were to be above the desired range and withdraw allowances from the auction with the further possibility of re-introducing them into the market in the case of a bearish market and thus a carbon price below the desired range. This option is expected to support the carbon price via an initial reduction of the allowances auctioned in the ETS, thus increasing its average value (Fig. 1). Moreover, given the possibility to modify the ETS cap in a flexible and predictable way, a CCB would stabilise the ETS carbon price against the risk of unexpected exogenous shocks, sending a certain price signal over the long-run. Moreover, the credibility of this measure requires the institution of an independent authority with the power of intervening in the ETS in a transparent way, by adjusting the supply of allowances according to clear criteria and indexes published in advance, upon which all interventions would be based. The designated authority should thus be insulated from short-term political concerns. Fig. 4 depicts a particular case where the CCB adjusts the ETS cap to maintain the carbon price at a pre-determined fixed level6 6 We assume that the price is fixed at a level that on average ensures the same level of emissions reduction as in the permanent and progressive set aside options.

S. Clo et al. / Energy Policy 57 (2013) 477–490

480

Cap

D2A Cap 2B

p

Cap 1

P

Cap 0

Cap 2A

D D2B

Db

E1

E2B

E2A

p1 Da

E0

p0

q2B

q1

q2A

q0

q

Fig. 4. Short and Long-term market effect induced by a price stabilization mechanism (CCB). Source: Own elaboration.

(from p0 to p1). As the demand of allowances varies with uncertainty (D2A and D2B), the CCB can adjust the supply of allowances accordingly (Cap2A and Cap2B) to maintain the carbon price at a fixed level (p1). Thus, the variation in demand causes a potential variation in emissions produced at equilibrium (q2A and q2B), though the average level of emissions equals that of the permanent and progressive set aside options. While the graphic illustrates the extreme case of a fixed-price, it is more likely that the CCB bank would be committed to allowing the carbon price to vary within a pre-determined corridor. 3.4. A carbon price floor Instead of intervening on the quantity of allowances to be supplied within the ETS, the carbon price could also be supported by introducing a carbon price floor (Helm, 2008). This option could be implemented either at a national level, as in the UK7 (option d.1), or at a EU level (option d.2). The British national price floor consists in an ex-post price adjustment: anytime the carbon price falls below the determined floor, the British power installations have to pay the difference between the carbon price floor and the price paid to purchase the allowances. In this way, firms will end up paying at least the carbon price floor, even if the market carbon price falls beneath it. This increases the cost of emissions in the UK, causing a further reduction in British emissions. However, as the total ETS cap remains unchanged, lower emission reductions would be required on the continent, reducing the demand of allowances and weakening the ETS carbon price. Thus, a national price floor risks lowering the ETS carbon price, without delivering any additional emission reductions across the EU as a whole. The impact of a national price floor is illustrated in Figs. 5 and 6. Fig. 5 describes a market where the overall demand is given by the horizontal aggregation of two separate demands – DA and DB – coming from two separate countries A (UK) and B (rest of Europe). The aggregate demand is a broken line: from (price) A to (price) B, DA is zero and the aggregate demand equals DB, from (price) 0 to (price) A, the aggregate demand is given by the sum of DA and DB. When a national carbon price floor is imposed on country A at a higher level than the previous carbon price (p0), two different 7 The UK will introduce a national price floor from 1 April 2013. The price floor will apply to the power sector and the floor will start at around £16 per ton of carbon dioxide (tCO2) and follow a linear path to target £30/tCO2 in 2020 (both in 2009 prices).

P0

Da+Db Qa

Qb

Q

Fig. 5. ETS market equilibrium with aggregated demand (countries A and B). Source: Own elaboration.

equilibria emerge for countries A and B (Fig. 6). In country A, the higher carbon price causes a reduction in produced emissions (DqA), determining a new equilibrium EA1 (qA1, pA1). As a consequence, the aggregate demand is broken in two points. As in the previous case, from A to B, DA is zero and the aggregate demand equals DB. From the price floor level (pA1) to A, both DA and DB are positive and the aggregate demand function is given by the sum of the two national demand functions. From 0 to the price floor (pa1), the aggregate demand function is given again only by demand in country B, as that portion of country A demand that has a willingness to pay below the carbon price floor cannot be satisfied anymore. As a consequence, below the price floor level, aggregate demand rotates from DA þDB pre-floor to DA þDB postfloor. The introduction of a national carbon price floor causes a reduction in the ETS carbon price at equilibrium (from p0 to pb1) and an increase in emissions produced by country B (DqB). Being the cap unchanged, the variation in emissions increase DqB equals, in absolute values, the variation in emissions reduction DqA, determining a leakage of emissions from country A to country B. Instead of setting a national price floor, a central price floor could be imposed within the EU ETS at an EU level (option d.2). There are several modalities for implementing an EU carbon price floor. First, the government may commit to buy back allowances at the floor price, thus reducing the amount of allowances in the market (Hepburn, 2006). However, as allowances have been originally allocated for free, this option could give rise to undesirable redistributive effects or result unfeasible given current budgetary constraints. A second way to introduce a price floor is to establish a reserve price equal to the price floor in public auctions (Grubb and Neuhoff, 2006). The reserve price ensures that any quantity of allowances demanded at a carbon price lower than the reserve price will not enter the market, ensuring an indirect rationing of supply within the ETS. We observe, however, that within the ETS a part of allowances will continue to be allocated for free in the third trading period 2013–2020 and part of the current surplus of unused allowances will be transferred from the second to the third trading period, causing a reduction in the demand of allowances to be acquired via public auction. To overcome this problem, an additional price correcting mechanism in the secondary market would be necessary to align the carbon price in the primary and secondary markets. As provided in the UK, this might consist in a fee equal to the difference between the reserve price and the secondary market clearing price.

S. Clo et al. / Energy Policy 57 (2013) 477–490

p

Cap 0

B

A

481

DB

DA EA1

pA1 +ΔpA p0 -ΔpB pB1

DA+DB Pre-floor

floor EA0

DA+DB post-floor

EB0 E B1

qA0

qA1

qB0

-ΔqA

qB1 +ΔqB

q

Fig. 6. Market effect induced by a national carbon price floor. Source: Own elaboration.

targets up to 2030 and beyond will be needed. This option could reasonably entail resolving issues of supply for Phase IV of the ETS (2020–2030) while declaring the firm intention to set the post 2030 cap by 2020. Given the significant oversupply of allowances that has been created within the ETS, the setting of long-term emission reduction targets is not expected to do much for prices in the short-term. This measure could nevertheless directly support the allowance price in the post-2020 period, as the cap would be reduced. Furthermore, greater certainty regarding longterm reduction targets is expected to strengthen the ETS framework and to better anchor future price expectations, providing greater certainty for long-term investments.

D2A

p

Cap D D2B

p1

E2B

E1

E2A

Floor

E0

p0

q2B

q1

q2A

q0

q

Fig. 7. Short and Long-term market effect induced by a EU price floor. Source: own elaboration.

An EU price floor would send a clearer price signal than the 30% target and the set-aside options, if set over the unconstrained equilibrium price. In this case, when the price floor is set at a level that, on average, ensures the same degree of emissions reduction as foreseen in the previous options, a variation in demand within a certain range (D2A, D2B) causes a variation in emissions produced at equilibrium (q2A, q2B), while price does not vary (Fig. 7). Though the setting of a EU price floor eliminates down-side risks affecting the price level, this option would not ensure price stability if the floor were set at a too low level in the face of a high degree of demand variance. 3.5. Long-term reduction targets This option would involve setting future targets for the post 2020 period. The EU has defined emission reduction goals for ETS sectors up to the year 2020. To reach the more ambitious reduction targets outlined in the EC’s Roadmap to 20508 , binding 8 The revised ETS Directive foresees a reduction of emissions by at least 50% below their 1990 levels by 2050. The European Council has endorsed the more ambitious target of an emissions reduction of 80% by 2050 in its Council Conclusions of 29/30 October 2009.

4. Multi-criteria analysis A multi-criteria evaluation method is applied to compare the policy options against a plurality of environmental, economic and procedural criteria, so as to identify options that would not only be feasible, but also effective in overcoming present shortcomings and avoiding them in the future. This comparative analysis is developed according to the criteria tree introduced by Konidari and Mavrakis (2007). The criteria selected by KM have been commonly used in previous evaluations of climate policies (among others, Kete and Petkova, 2001; Philibert and Pershing, 2001; Sorrell, 2001; Johannsen, 2002; Aldy et al., 2003). They are comprehensive, allowing us to consider the impact of each policy on a plurality of subjects and variables, and they reflect the preferences of various and conflicting stakeholders with different priorities (target groups, decision makers and researchers). Moreover, these criteria are well tailored to our analysis as they have already been adopted to evaluate different policies related to mitigation of climate change, such as the ETS. This tested framework constitutes a solid basis for developing our comparative analysis. The criteria tree is structured on three levels (Fig. 8). The first level concerns the primary objective that instruments should achieve, which in our case is the improvement of the ETS effectiveness in delivering emissions reductions. The second level is composed of three main criteria: environmental performance, political acceptability and feasibility of implementation, each of which is composed of a set of sub-criteria, constituting the third level of the hierarchy tree.

S. Clo et al. / Energy Policy 57 (2013) 477–490

482

Direct contribution to GHG reduction Environmental performance Indirect environmental effects

Static efficiency

Dynamic efficiency

Improve ETS effectiveness in promoting GHG reduction

Competitiveness Political acceptability European harmonization

Flexibility

Stringency for noncompliance Implementation network capacity Feasibility of implementation

Timing of implementation

Impact on public finance

Fig. 8. Multi-criteria tree. Source: own elaboration on Konidari and Mavrakis (2007).

1. Environmental performance (EP): contribution of the instrument to achieving the environmental goal of emissions reduction. This criterion is divided into two sub-criteria: 1.1. Direct contribution to reduction of GHG emissions: amount of emissions reduction achieved under the policy option. 1.2. Indirect environmental effects: ancillary environmental outcomes attributed to each policy option. 2. Political acceptability (PA): extent to which the EU Member States (MS) and the EU ETS stakeholders are able to welcome and accept the proposed instrument. Political acceptability depends on several factors which are captured by the following six sub-criteria: 2.1 Static efficiency: the capacity of the instrument to achieve the desired goal in an effective way. We measure this criterion by considering to which extent the policy options reviewed manage to support the carbon price by increasing its average value in the short-term. 2.2. Dynamic efficiency: ‘‘the property of the instrument to create, offer or allow compliance options that support research projects, incremental and radical pioneer technologies and techniques, and institutional or organizational innovations’’ (Konidari and Mavrakis (2007), p. 6239). This criterion is measured by considering the capacity of each policy both to support the carbon price and to send a clear and long-term carbon price signal, thus reducing the range within which the carbon price can vary and increasing firms’ confidence in the ETS.

2.3. Competitiveness: the extent to which a policy option is likely to increase the cost of production in the ETS sectors over the short-run. To assess the impact of policy options on ETS firms’ competitiveness, we consider that the higher the carbon price, the higher their compliance cost is. 2.4. European harmonization9 : the contribution of each policy option to increasing harmonization within the ETS, between different MS and between climate policies on ETS and non-ETS sectors. 2.5. Flexibility: the capacity of the instrument to increase the options available to regulated agents by which they can comply with the imposed regulation. Flexibility within the ETS can be improved through options like the Clean Development or market-linking mechanisms that increase abatement opportunities. Moreover, flexibility can be improved through instruments that reduce rigidity on the supply side and allow adjustments to the cap in the case of exogenous shocks on the demand side. 2.6. Stringency for non-compliance: the capacity to monitor the level of compliance and to impose sanctions on those who fail to comply. 3. Feasibility of Implementation (FI): the ‘‘aggregate applicability of the instrument linked with national infrastructural (y) and legal framework’’ (Konidari and Mavrakis (2007), p. 6240).

9 It corresponds to the ‘‘equity’’ criterion adopted by KM: the fairness of the instrument in allocating costs and benefits among the regulated subjects.

S. Clo et al. / Energy Policy 57 (2013) 477–490

Assessment of the policy options against this criterion is based on the following three sub-criteria. 3.1 Implementation network capacity: the effective capacity of the institutions to design, support and implement the chosen instrument. 3.2. Timing of implementation10 : the amount of time roughly required for the necessary political debate on any proposed instrument and, subsequently, for its effective design, implementation and enforcement. Indeed, the time element is crucial when considering the feasibility of introducing any policy option within the ETS legal framework. 3.3. Impact on public finance: how public revenues from auctioning are likely to vary under each of the proposed measures. The size and direction of variation in public revenues depends on the slope of the demand curve11 . The quantitative analysis performed by Capros et al. (2011) in the impact assessment of the  30% target proposal12 concludes that by reducing the ETS cap, public revenues from auctioning are likely to rise, meaning that the increase in public revenues induced by a higher carbon price (price effect) is likely to exceed the reduction in public revenues caused by a decrease in the total amount of allowances sold via public auction (quantity effect)13 . Moreover, it has been asserted that ‘‘Under a 20% scenario, the current allocation methods will generate public sector revenues in Phase III of the EU ETS in the order of h150–190bn. (y) Should the EU move to a 30% commitment, this rises to h200–310bn’’ (p. 22 Grubb and Cooper, 2011). We take this conclusion as a benchmark to assess the impact of each policy option on public finances. Moreover, when the policy option would involve changes not only for the ETS but also for non-ETS sectors, we also consider the effect of changes in the non ETS target on public finances, as well.

5. Ranking techniques We adopt two alternative approaches to evaluate each policy option and determine a final ranking among them. Under the Method 1 we equally weigh the selected criteria (equal weight method), limiting our assessment of each policy option to whether it has a positive ( þ1), a neutral (0), or a negative ( 1) impact in terms of each criterion (Oikonomou and Jepma, 2008). By summing up all the values so obtained, we determine a final grade which synthetizes the overall performance of each option. This simple method allows us to evaluate the performance of the different policy options on an aggregate basis and to cardinally rank them. Moreover, the equal weight method requires a minimal knowledge about preferences and has been popularly applied in energy decision making issues (Wang et al., 2009). On the other hand, this methodology does not allow us to distinguish between policy options in terms of the different intensity of their impact and, by assuming equal weights, it cannot take into account different preferences that policymakers and ETS’ stakeholders might have regarding the various criteria. These drawbacks make it difficult to determine an unequivocal ranking among the policy instruments. 10

It corresponds to the ‘‘administrative feasibility’’ criterion adopted by KM. 11 Generally speaking, when the slope of the demand curve is relatively flat the quantity effect will be greater than the price effect, causing a reduction of public revenues. Vice-versa, when the slope of the demand curve is relatively steep, then the price effect is greater than the quantity effect, increasing public revenues. 12 This assessment is based on the PRIMES and GAINS models. 13 ‘‘The carbon price increases from h16 in the reference case to h30, allowing total revenues from auctioning in 2020 to increase from h21 billion in reference to h29 billion with the auctioning set-aside’’ (p. 47 EC 2010a).

483

Table 1 Criteria and weight coefficients. Source: Konidari and Mavrakis (2007). Criteria

Weight coefficients

1. Environmental performance 1.1 Direct contribution to reduction of GHG emissions 1.2 Indirect environmental effects 2. Political acceptability 2.1 Static efficiency 2.2 Dynamic efficiency 2.3 Competitiveness 2.4 European harmonization 2.5 Flexibility 2.6 Stringency for non-compliance 3. Feasibility of implementation 3.1 Implementation network capacity 3.2 Timing of implementation 3.3 Impact on public finance

0.168 0.833 0.167 0.738 0.474 0.183 0.085 0.175 0.051 0.032 0.094 0.309 0.581 0.11

Table 2 Scale of grades to assess the instruments’ performance. Source: Konidari and Mavrakis (2007). Grade with method 2 Null 0 Slightly more than null, less than 1 very bad Verybad 2 Bad 3 More than bad less than moderate 4 Moderate 5 More than moderate less than good 6 Good 7 More than good, less than very good 8 Verygood 9 Excellent 10

Grade with method 1 

0 þ

Therefore, we apply a second method to compare policy options according to a more comprehensive technique. Method 2 entails the adoption of the weighted criteria estimated by Konidari and Mavrakis (2007) on the basis of the expressed preferences of three main stakeholder groups involved in climate policy issues (policy makers, researchers and target groups)14 (Table 1). The performance of the policy options against each weighted criterion is graded using the scale [0,10], where the significance of each number within the scale is reported in Table 2 (Konidari and Mavrakis, 2007). A final value which synthetizes the option’s overall performance is determined using the weighted sum method15. Finally, we control the robustness of this ranking by means of a Sensitivity Analysis (SA). This approach is commonly used in energy system decision making processes (Wang et al., 2009; Afgan et al., 2007; Afgan and Carvalho, 2008; Diakoulaki et al., 2007; Dinca et al., 2007; Pilavachi et al., 2009).

14 These weighted coefficients, estimated using the Analytical Hierarchy Process technique, have been tested with the Saaty and Pelaez J.I. and Lamata M.T. consistency indexes and can be adopted to evaluate different climate policy options. For an exhaustive analysis the reader should refer to Konidari and Mavrakis, 2007. 15 For any given policy option, the grades given on each criterion are weighted according to the specific criterion’s weight. The weighted grades are then summed up, resulting in a unique final value for each option. Moreover, to express differences in performance of the evaluated policy options in relative terms, direct grading can be normalized (see KM for details). We have verified that our final normalized and non-normalized ranking coincide, therefore we abstain from reporting the normalization process.

484

S. Clo et al. / Energy Policy 57 (2013) 477–490

6. Grading of each instrument’s performance In this section each policy option is evaluated according to the selected criteria in terms of its impact on overall EU performance. We recall that any grade between 0 and 4 under Method 2 corresponds to a negative (  ) grade under Method 1, a moderate grade 5 corresponds to a neutral grade (0), while any grade between 6 and 10 under Method 2 corresponds to a positive grade (þ) under Method 1 (Table 2). Assessments are mainly based on the economic analysis of the impact of each policy proposal on the ETS’ emissions and price (as illustrated in Figs. 2–7 in Section 3). 6.1. Grading environmental performance The imposition of a  30% target would induce the highest (i) direct reduction of EU emissions, as both ETS and non-ETS emissions would be reduced (assessment: very good). The setting of long-term targets would imply greater EU emission reductions over the long-run, without promoting any additional reductions in the short-run. For this reason it is evaluated slightly less than the  30% target reduction option (more than good less than very good). The permanent and progressive set-aside options are graded as good because over the 2013–2020 period they would induce an emissions reduction only in the ETS (Fig. 2). The CCB and an EU price floor are also graded as good because they would reduce emissions to a level that would be, on average, equal to the one induced by the permanent and progressive set-aside (Figs. 4 and 7). Finally, both the temporary set-aside and the national price floor options have a neutral impact on emissions, as the overall ETS cap would remain unchanged and no further reduction would be induced compared to the current scenario (moderate). Like KM, we assume that (ii) indirect environmental effects are proportional to the direct contribution of these options in reducing emissions. The same grades as obtained with the previous sub-criterion are thus assigned. 6.2. Grading political acceptability The first sub criteria used is (i) static efficiency. In the shortterm, the 30% target, the permanent and temporary set-aside all increase the carbon price to the same level (Fig. 1), resulting in an improved effectiveness of the ETS in inducing emission reductions (between good and very good). The CCB and the EU price floor are equally effective in improving the ETS because in the short-term they support the carbon price at the same level than the previous options (Figs. 4 and 7), and they are evaluated with the same grade (between good and very good). The progressive set-aside induces a gradual price adjustment, thus in the short-term the carbon price is supported at a lower level than with previous options and consequently it is graded slightly less (good). Setting long-term targets does not modify the ETS framework in the short-term, thus this option is considered to be moderate in improving ETS effectiveness. Finally, a national carbon price floor is the worst option as it would have a bad impact on the EU carbon price, worsening the overall ETS effectiveness. The institution of a CCB is considered the best option to improve the ETS in terms of (ii) dynamic efficiency (excellent). This option would result in a clear long-term price signal, as a reversible intervention on the ETS cap would allow the carbon price to be kept within a pre-defined range, ensuring price stability (Fig. 4). An EU price floor is likely to reduce the range over which where the carbon price can vary and is thus evaluated between good and very good, as firms can rely on a minimum EU based carbon price to plan their investments (Fig. 7). The definition of post-2020 targets would also improve the long-term price

signal by increasing the underlying regulatory certainty (between good and very good). The long-term effects induced by a reduction of the ETS cap remain uncertain. The  30% target and the progressive and permanent set-aside options are not likely to improve or worsen the ETS dynamic efficiency (moderate) as the degree of ETS flexibility would remain unchanged compared to the current legislation and the price is expected to continue to vary with uncertainty depending on the evolution of demand (Fig. 2). A temporary set-aside is expected to have a very bad impact as in the long-term the carbon price can be expected to vary within an even wider range (Fig. 3). Finally, a national price floor risks creating ambiguous price signals, possibly leading to normative arbitrage among countries (bad). The third sub-criteria is (iii) competitiveness. The  30% target, the permanent set-aside and the EU price floor are likely to worsen the ETS installations’ competitiveness, compared to the current situation, due to the expected increase in the carbon price (bad). The progressive set-aside has also a negative impact on competitiveness. However, the slower price adjustment is likely to reduce the risk of a bullish price shock, giving ETS firms more time to adjust their investment plans. Therefore, this option is judged to be more than bad less than moderate. With a CCB, firms’ competitiveness could be worsened by the short-term reduction of the cap (Fig. 4) while, on the other hand, the risk of increases in the cost of compliance over the longer term would be lower given the possibility to limit the carbon price increase (moderate). Likewise the impact of the adoption of long-term targets is evaluated as being moderate as it is not expected to impact significantly on the carbon price trend in the 2013–2020 period. A grade between moderate and good is assigned to the national price floor because, in spite of the EU intra-leakage effect, the asymmetric price trend between the country that adopts the price floor and the rest of the EU ETS is expected to slightly reduce the ETS average carbon price (Fig. 6). Finally, a grade between moderate and good is also assigned to the temporary set-aside because the reintroduction of allowances over the third trading period is expected to reduce the average carbon price (Fig. 3). (iv) European harmonization. The three variants of the set-aside option are European-based and would impact on the ETS in a uniform way. As such, the application of any one of the three would not have distributional implications within the ETS or between MS (good). The same judgement holds for long-term reduction targets, as measures to define post 2020 emission targets would be decided upon and undertaken within the EU. The EU price floor would also ensure a good level of harmonization within the ETS. The 30% target is European-based and is the only option which would reduce both ETS and non-ETS targets, thus improving the harmonization between the related sectors. However, this option would affect non-ETS national targets differently, having different distributional effects among MS (European Commission (EC), 2012). Given these counterbalancing effects we grade this option as moderate. Instead, a national carbon price is likely to be very bad as it risks to distort the ETS, limiting integration of the internal market. Finally, the CCB is likely to have an impact in terms of harmonization and equity that is slightly better than other options (between good and very good). Not only is this option European-based by nature; by keeping price oscillations circumscribed within a pre-defined range, the CCB would permit a greater harmonization between the level of the ETS carbon price and that of a potential carbon tax on non-ETS sectors in a future revised EU directive on energy taxation. None of the proposed options would increase and/or diversify the means by which regulated agents can comply with the imposed regulation. As such, they have only a moderate impact in terms of (v) flexibility, as no further abatement opportunities

S. Clo et al. / Energy Policy 57 (2013) 477–490

would be introduced and the current functioning of the ETS would remain largely unchanged. Only the CCB would improve the flexibility of the ETS, as the cap could be rapidly modified to balance market demand in the case of unexpected exogenous shocks (between good and very good). (vi) Stringency for noncompliance is the same for all the policy options as the rules for non-compliance would not be modified. The level of stringency is very good. 6.3. Grading feasibility of implementation In terms of (i) implementation network capacity, most of the policy options can be designed and implemented within the existing and experimented institutions and networks; they are given a grade between moderate and good. The CCB is the only option that requires an extension of the existing network capacity through the creation of an independent central authority. This option is graded between bad and moderate. (ii) Timing of implementation in the case of a CCB is judged to be slightly more than null, less than very bad. In fact, this option would require a deep modification of the ETS framework. As the establishment of the common auctioning platform required an impact assessment, a public consultation, a proper regulation and a public procurement process, we would expect that similar steps would need to be undertaken to set up an independent CCB, implying a relatively long timeframe for implementation. A bad grade is attached to the  30% option where timing represents a critical issue since not only the ETS Directive, but also the non-ETS Decision No 406/2009/EC, would need to be amended through a procedure involving both the European Council and the Parliament. This would require considerable negotiation between MS and European institutions; a significant amount of time could be required for a sufficient consensus to be reached. The negotiation of post-2020 binding targets is a complex issue, linked not only to the current economic juncture but also to future progress in international negotiations and, like the 30% target, it would require an amendment to both the ETS and non-ETS Directives (bad). Time frameworks for implementation of the three set-aside options might differ, as these options have different requirements. In particular, a temporary set-aside would not modify the overall amount of allowances auctioned in the 2013–2020 period, making it presumably easier to form a political consensus on its adoption within a limited lapse of time (between moderate and good). Conversely, by modifying the overall ETS cap, the permanent and progressive set-aside options would most likely require both a longer political debate and adjustments in the relevant legislation, including the revision of the EU ETS Directive itself. These options are judged to be more than bad less than moderate. Similarly, the introduction of an EU carbon price floor would require MS to modify the ETS Directive and framework. As many countries may oppose this option, the time needed for implementing this option risks to be lengthy (more than bad less than moderate). On the contrary, a national carbon price floor could be adopted within a limited time period by governments intervening directly on the national legislation and without requiring any change to the ETS framework (between good and very good). Concerning the last sub-criterion, (iii) impact on public finance, both a permanent and a progressive reduction of the ETS cap would be expected to increase public revenues from auctioned allowances compared to the current situation, as asserted by Capros et al. (2011) and Cooper and Grubb (2012) (good). The CCB and the EU price floor options are also likely to have a good impact on public revenues as the carbon price would be maintained at a higher level than the current one. The definition of long-term targets is not expected to impact on public finance, as this option is not expected to modify the EU carbon price over the period

485

2013–2020 (moderate). The temporary set-aside is expected to have an impact on public finances between bad and moderate, as the reintroduction of previously back-loaded allowances, coupled with an uncertain variance in demand, is expected to reduce the average carbon price (Fig. 3). The impact on public finances of the  30% target is also expected to be lower than the permanent and progressive set-aside options because the strengthening of the non-ETS target is likely to pose an extra burden on public budgets, increasing the cost of compliance, unless measures were introduced on the non-ETS sectors to raise new funds (between bad and moderate). Finally, the national floor is graded as bad as it risks to worsen public finances in all countries except those imposing it, since the ETS carbon price would likely be lowered.

7. Sensitivity analysis of the results According to the ranking determined with Method 1 (equally weighted coefficients), both the CCB and the EU price floor constitute first-best options. Both the progressive set-aside, the permanent set-aside and long-term targets represent second-best choices, while the temporary set-aside covers the third position. The  30% target is classified as fourth, while the setting of a national price floor represents the worst option (Table 3). This first method does not allow us to determine an unequivocal ranking among the considered policy options. To overcome this problem, we look at the final ranking determined with Method 2. In this case, the CCB constitutes the first-best option while the national price floor continues to be the worst one, as determined with Method 1. However, differently from Method 1, we can now rely on a clear ranking, where the EU price floor covers the second position; the permanent set-aside and the  30% target cover, respectively, the third and fourth positions; the progressive set-aside is classified as fifth; long-term targets, as sixth. The temporary set-aside is the next to last option (Table 4). This final ranking depends both on the assigned grades and on the coefficients’ weights estimated by Konidari and Mavrakis, 2007. These reveal a high preference for political acceptability (PA 0.738), compared to environmental performance (EP 0.168) and the feasibility of implementation (FI 0.094). Since we cannot exclude a priori that the current political preferences might differ from the ones expressed by the KM coefficients, a sensitivity analysis, where the coefficients’ weights are gradually modified to reflect other possible sets of preferences,16 has been performed to test the robustness of this ranking. Three different cases have been developed.

Case 1. This case highlights that when there is a higher preference toward the instrument’s EP compared to its PA, the 30% target and the long-term target options improve their ranking position against the other policy options. In detail, by keeping FI constant, reducing the PA’s coefficient by a certain percentage and by increasing the weight of EP by the same amount, the ranking remains unchanged up to a 19% gradual reduction of the PA coefficient. A 20% variation reverses the position between the 30% target option and the EU price floor, which is downgraded to the third position. After a 30% reduction of the PA criterion, the 30% target 16 First, out of three weight coefficients, one is increased gradually by a certain percentage, the second is lowered by the same percentage and the third one is adjusted accordingly. Second, one weight coefficient is increased gradually, while the second one is reduced by the same amount and the third one is kept constant (Konidari and Mavrakis (2007)).

S. Clo et al. / Energy Policy 57 (2013) 477–490

486

Table 3 Grades and ranking of the policy options (method 1). Source: Own elaboration. Criteria

 30 (%)

Environmental performance Direct contribution to reduction of GHG emissions Indirectenvironmentaleffects Politicalacceptability Static efficiency Dynamic efficiency Competitiveness EU harmonization Flexibility Stringency for non-compliance Feasibility of implementation Implementation network capacity Timing of implementation Impact on public finance TOTAL Final ranking

Permanent setaside

Temporary setaside

Progressive setaside

CCB Nationalpricefloor EU pricefloor

Long-term target

1

1

0

1

1

0

1

1

1

1

0

1

1

0

1

1

1 0 1 0 0 1

1 0 1 1 0 1

1 1 1 1 0 1

1 0 1 1 0 1

1 1 0 1 1 1

1 1 1 1 0 1

1 1 1 1 0 1

0 1 0 1 0 1

1 1 1

1 1 1

1 1 1

1 -1 1

1 1 1

1 1 1

1 1 1

1 1 0

2 41

5 21

4 31

5 21

6 11

0 51

6 11

5 21

Table 4 Grades and ranking of the policy options (method 2). Source: Own elaboration. Criteria

Weightcoefficients  30 (%)

Environmental performance Direct reduction of emissions Indirectenvironmentaleffects Politicalacceptability Static efficiency Dynamic efficiency Competitiveness EU harmonization Flexibility Stringency for noncompliance Feasibility of implementation Impl. network capacity Timing of implementation Impact on public finance

0.168 0.833 0.167 0.738 0.474 0.183 0.085 0.175 0.051 0.032

TOTAL weighted Ranking weighted

0.094 0.309 0.581 0.11

Permanent setaside

Temporary setaside

Progressive set- CCB aside

Nationalpricefloor EU pricefloor

Long-term target

9 9

7 7

5 5

7 7

7 7

5 5

7 7

8 8

8 5 3 5 5 9

8 5 3 7 5 9

8 2 6 7 5 9

7 5 4 7 5 9

8 10 5 8 8 9

3 3 6 2 5 9

8 8 3 7 5 9

5 8 5 7 5 9

6 3 4

6 4 7

6 6 4

6 4 7

4 1 7

6 8 3

6 4 7

6 3 5

6.60 41

6.61 31

6.13 71

becomes the first option while the CCB is downgraded to the second position. By further reducing the weight of PA, the long-term target option gradually improves; after a 55% variation, the long-term target reaches the second position, downgrading the CCB to the third one (Table 5 and Fig. 9).17 Case 2. The CCB’s ranking position worsens also when more weight in preferences is given to the FI criterion compared to PA. In detail, by gradually reducing the PA coefficient, increasing the EP coefficient by the same percentage and adjusting FI accordingly, the ranking remains unchanged up to a 20%

17 A similar trend can be observed when the weight of PA is gradually reduced by a certain percentage, the FI weight is increased by the same percentage and the EP coefficient is adjusted accordingly, the ranking remains unchanged up to a 22% variation. At a 23% variation of PA and FI, the  30% target becomes the secondbest option, downgrading the EU price floor to the third position. The CCB remains the first-best option up to a 30% variation of the FI and PA coefficients; with a 31% variation the  30% target becomes the first-best option. By further increasing the percentage variation, the long-term target option gradually improves its position and after a 57% variation it becomes the second option.

6.32 51

7.43 3.97 11 81

7.01 21

6.18 61

variation of the coefficients. At a 21% variation, the CCB reverses its position with the EU price floor, which remains the first best up to an 84% variation. At an 85% variation, where the PA is the least weighted criterion, while the FI becomes the most relevant criterion, the national price floor becomes the first-best option, since it has the highest performance under the feasibility criterion. We also observe that after a 62% variation, the CCB becomes the worst option (Table 6 and Fig. 10).18 Case 3. Finally, when the preference for the PA criterion is reinforced, then the CCB and the EU price floor options maintain their positions. Case 3a By reducing gradually the

18 A similar trend can be observed by keeping EP constant, gradually reducing the PA coefficient by a certain percentage and by adjusting the FI by the same amount. In this case, we find that the ranking remains unchanged until a 17% variation of the PA and FI’s coefficients. After an 18% variation, the CCB reverses its position with the EU price floor, which then covers the first position up to a 71% variation. At a 72% variation in coefficients, the national price floor reaches the first position. We also observe that after a 53% variation the CCB becomes the last option.

S. Clo et al. / Energy Policy 57 (2013) 477–490

487

Several measures aimed at improving the effectiveness of the ETS have been proposed within the European climate policy debate. This paper has analysed the economic and environmental impact of these proposals on the ETS, both on the carbon price and on the level of emissions. Moreover, these options have been compared against the KM multi-criteria tree using two alternative ranking techniques. First, we have equally weighted the selected criteria and assessed if each policy option has a positive (þ1), neutral (0), a negative (1) impact against each criterion (Method 1). Subsequently, the criteria have been weighted to take into account

stakeholders’ preferences and performance of the policy options has been graded using the scale [0,10] in order to consider the intensity of their impact against the several criteria (Method 2). Both approaches confirm the CCB as the first-best option and the national price floor as the worst, but only the second method allows us to determine an unequivocal ranking among the policy instruments where the EU price floor covers the second position. This ranking depends both on the assigned grades and on the criteria’s weight coefficients estimated by KM. Since we cannot exclude a priori that the current political preferences might differ from the ones expressed by the KM coefficients, a sensitivity analysis has been developed to test the robustness of this ranking. The SA has revealed that in the best scenario the final ranking remains substantially unchanged (case 3b), while in the worst case scenario it remains unchanged for a variation up to 19% of the weight coefficients (Case 1), while further variations in the coefficients’ weights modify the ranking. The SA has also confirmed that the first-best option highly depends on political preferences. In particular, when policymakers have as a primary goal to increase the ETS’ economic effectiveness and political acceptability, then the CCB is the first-best option, while the EU price floor covers the second position (SA case 3). Indeed, by allowing for a reversible adjustment of the ETS cap according to clear, pre-defined rules, the CCB would increase flexibility in the ETS and improve both its static and dynamic efficiency, by ensuring support and stability to the carbon price. Moreover, the CCB would allow for a higher degree of harmonization within the EU and among ETS and non-ETS sectors. Against these benefits, this option is likely to bring a lower

Table 5 Weight coefficients with percentage reduction of PA, FI constant and EP adjustment (Case 1). Source: Own elaboration.

Table 6 Weight coefficients with a percentage reduction of PA, percentage increase of EP and FI adjustment (Case 2). Source: Own elaboration.

EP coefficient, increasing the FI weight by the same percentage and by adjusting PA, the PA’s coefficient gradually increases and the ranking of the first three options always remains unchanged. As the weight of FI increases, the temporary setaside option gradually improves its position, and after a 56% variation becomes the fourth option, while the position of the  30% target option worsens due to the lower weight of EP; after a 67% variation it becomes the fifth option (Fig. 11). Case 3b By keeping EP constant, gradually reducing FI and increasing PA by the same amount, the ranking remains substantially unchanged (only the  30% target and the permanent set-aside slightly switch their positions after a 15% variation of the coefficients) (Fig. 12).

8. Summary and conclusions

EP (adjsument) PA (% reduction) FI (fixed)

0 (%)

20 (%)

30 (%)

55 (%)

0.168 0.738 0.094

0.315 0.590 0.094

0.389 0.516 0.094

0.574 0.332 0.094

EP (% increase) PA (% reduction) FI (adjustment)

0 (%)

21 (%)

62 (%)

85 (%)

0.168 0.738 0.094

0.203 0.583 0.214

0.272 0.280 0.447

0.311 0.111 0.579

8 7.8 7.6 7.4 7.2 7 6.8 6.6 6.4 6.2

0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 22% 24% 26% 28% 30% 32% 34% 36% 38% 40% 42% 44% 46% 48% 50% 52% 54% 56% 58% 60% 62% 64% 66% 68% 70% 72% 74% 76% 78% 80% 82% 84% 86% 88% 90% 92% 94% 96% 98% 100%

6

-30%

permanent set-aside CCB

temporary set-aside

EU price floor

progressive set-aside

Long-term target

Fig. 9. Ranking variation with percentage reduction of PA, FI constant and EP adjustment (Case 1). Source: Own elaboration.

S. Clo et al. / Energy Policy 57 (2013) 477–490

488

7.5 7 6.5 6 5.5 5 4.5

0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 22% 24% 26% 28% 30% 32% 34% 36% 38% 40% 42% 44% 46% 48% 50% 52% 54% 56% 58% 60% 62% 64% 66% 68% 70% 72% 74% 76% 78% 80% 82% 84% 86% 88% 90% 92% 94% 96% 98% 100%

4 -30% CCB

permanent set-aside national price floor

temporary set-aside EU price floor

progressive set-aside Long-term target

Fig. 10. Ranking variation with a percentage reduction of PA, percentage increase of EP and FI adjustment (Case 2). Source: own elaboration.

8

7.5

7

6.5

5.5

0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 22% 24% 26% 28% 30% 32% 34% 36% 38% 40% 42% 44% 46% 48% 50% 52% 54% 56% 58% 60% 62% 64% 66% 68% 70% 72% 74% 76% 78% 80% 82% 84% 86% 88% 90% 92% 94% 96% 98% 100%

6

-30% progressive set-aside

permanent set-aside CCB

temporary set-aside EU price floor

Long-term target

Fig. 11. Ranking variation under a percentage variation of EP (  ) and FI (þ) with PA adjustment (Case 3a). Source: own elaboration.

emissions reduction than other options like the 30% target. Moreover, it is likely to require a significant degree of intervention upon the current ETS framework that would imply a relatively long time-frame for implementation, as the institution of an independent CCB and the definition of clear rules for adjusting the ETS cap require a long legal procedure as well as a determined political will to do so. Given these shortcomings, when the relevance of the environment performance and implementation feasibility goals gradually increases against the political acceptability criterion, the EU price floor gains in position against the CCB. Thus, when the policymakers have a preference for an instrument that is both environmentally and economically effective and that can be adopted in the short-term with a low cost of implementation, the

EU price floor constitutes the most tailored option (SA case 2). The CCB is not the most suitable policy option when the stakeholders want to mainly reach an environmental goal. In this case, the  30% target becomes the policy option they should adopt (SA case 1). Given its asymmetric impact between the country which adopts it and the rest of the ETS, a national price floor is generally classified as the worst option. Nevertheless, self-interested States looking for a feasible solution may find it to be the optimal solution. In fact, we find that if the feasibility of implementation is the main priority of policymakers, then the national price floor can even become the first-best option (SA case 2). Interestingly, the set-aside option that currently appears to be the main point of reference in the EU political agenda, does not cover the first position in any of the analyzed scenarios. Finally,

S. Clo et al. / Energy Policy 57 (2013) 477–490

489

8.5 8 7.5 7 6.5 6 5.5 5 -30%

permanent set-aside temporary set-aside progressive set-aside CCB EU price floor Long-term target

Fig. 12. ranking variation under a percentage reduction of FI, EP fixed and PA adjusted (Case 3b). Source: Own elaboration.

in spite of not being a first-best option, the setting of long-term targets could be applied in conjunction with practically any of the options discussed and would have the benefit of supporting the current ETS as its stringency would be guaranteed into the future. (Tables 5 and 6).

References Afgan, NH, Pilavachi, PA, Carvalho, MG., 2007. Multi-criteria evaluation of natural gas resources. Energy Policy 35 (2007), 704–713. Afgan, NH, Carvalho, MG., 2008. Sustainability assessment of a hybrid energy system. Energy Policy 36 (2008), 2903–2910. Alberola, E., Chevalier, J., Che ze, B., 2008a. Price drivers and structural breaks in European carbon prices 2005–2007. Energy Policy 36 (2), 787–797. Alberola, E., Chevalier, J., Che ze, B., 2008b. The EU emissions trading scheme: the effects of industrial production and CO2 emissions on european carbon prices. International Economics 116 93–26. Alberola, E., Chevalier, J., 2009. European carbon prices and banking restrictions: evidence from phase I (2005–2007). The Energy Journal 30 (3), 51–80. Aldy, J.E., Barrett, S., Stavins, N.R., 2003. Thirteen plus one: a comparison of global climate policy architectures. Climate Policy 3, 373–397. Capros P. Mantzos L., Parousos L., Tasios N., Klaassen G., Van Ierland T., 2011, Analysis of the EU policy package on climate change and renewables, 39, Issue 3, March 2011, Pages 1476–1485. Chevalier, J., 2009. Carbon futures and macroeconomic risk factors: a view from the EU ETS. Energy Economics 31 (4), 614–625. De Perthuis, C., 2011. Carbon Market Regulation: The Case for a CO2 Central Bank. Paris Dauphin University Climate Economic Chair Information and Debates Series. Diakoulaki, D, Karangelis, 2007. F. Multi-criteria decision analysis and cost-benefit analysis of alternative scenarios for the power generation sector in Greece. Renewable and Sustainable Energy Reviews 11 (2007), 716–727. Dinca, C, Badea, A, Rousseaux, P, Apostol, T., 2007. A multi-criteria approach to evaluate the natural gas energy systems. Energy Policy 35 (2007), 5754–5765. European Commission (EC) (2008), Commission Staff Working Document Accompanying Document to the Proposal for a Directive of the European Parliament and of the Council Amending Directive 2003/87/EC so as to Improve and Extend the EU Greenhouse Gas Emission Allowance Trading System, Brussels. European Commission (EC) (2010a) SEC(2010) 650 Commission Staff Working Document Accompanying The Communication From The Commission to The European Parliament, The Council, The European Economic And Social Committee And The Committee Of The Regions—Analysis of Options to Move Beyond 20% Greenhouse Gas Emission Reductions and Assessing the Risk of Carbon Leakage Background Information and Analysis, Brussels. European Commission (EC) (2010b), Commission Regulation on the Timing, Administration and Other Aspects of Auctioning of Greenhouse Gas Emission Allowances Pursuant to Directive 2003/87/EC of the European Parliament and

the Council Establishing a Scheme for Greenhouse Gas Emission Allowances Trading within the Community, Brussels. European Commission (EC) (2010c), EU Energy Trends to 2010—Update 2009, European Commission, Directorate-General for Energy. European Commission (EC) (2010d) 677 final, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, Energy Infrastructure Priorities for 2020 and beyond—A Blueprint for an Integrated European Energy Network, Brussels. European Commission (EC) (2011) 288 Final, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, A Roadmap for Moving to a Competitive Low Carbon Economy in 2050, Brussels. European Commission (EC) (2012) Staff Working Paper SWD 5 final, Analysis of Options beyond 20% GHG Emission Reductions: Member State Results, Brussels. Grubb, M., Neuhoff, K., 2006. Allocation and competitiveness in the EU emissions trading scheme: policy overview. Climate Policy 6, 7–30. Grubb M., Cooper S. (2011), Revenue dimensions of the EU ETS phase III, climate strategies working paper. Available at /http://www.climatestrategies.org/ research/our-reports/category/61/313.htmlS. Grubb M. (2012). Strengthening the EU ETS: Creating a stable platform for EU energy sector investment, Climate Strategies Report. Helm D. (2008). Caps and Floors for the EU ETS: A Practical Carbon Price Version for Website, 13 October. Hepburn, C., 2006. Regulation by prices, quantities, or both: a review of instrument choice. Oxford Review of Economic Policy 22 (2), 226–247. Hepburn C., Quah J.K.-H., Ritz R.A. (2007) Emissions Trading and Profit-Neutral Grandfathering. Johannsen, K.S., 2002. Combining voluntary agreements and taxes-an evaluation of the Danish agreement scheme on energy efficiency in industry. Journal of Cleaner Production 10, 129–141. Kete, N., Petkova, E. (2001). Assessing good practices in policies and measures to mitigate climate change in Central and Eastern Europe. In: Workshop on Good Practices in Policies and Measures, 8–10 October 2001, Copenhagen. Konidari, P., Mavrakis, 2007. A multi-criteria evaluation method for climate change mitigation policy instruments. Energy Policy vol. 35, 6235–6257. IEA 2011. World Energy Outlook 2010, Paris. Martin R., Muuls M., & Wagner U. (2011), Climate Change, Investment and Carbon Markets and Prices—Evidence from Manager Interviews. Oikonomou, V., Jepma, C.J., 2008. A framework on interactions of climate and energy policy instruments. Mitigation and Adaptation Strategies for Global Change 13 (2008), 131–156. Philibert, C., Pershing, J., 2001. Considering the options: climate targets for all countries. Climate Policy 1 (2), 211–227. Pilavachi, PA, Stephanidis, SD, Pappas, VA, Afgan, NH., 2009. Multi-criteria evaluation of hydrogen and natural gas fuelled power plant technologies. Applied Thermal Engineering 29 (2009), 2228–2234. Sijm, J., Neuhoff, K., Chen, Y., 2006. CO2 cost pass-through and windfall profits in the power sector. Climate Policy 6 (2006), 49–72.

490

S. Clo et al. / Energy Policy 57 (2013) 477–490

Sorrell, S., Harrison, D., Radov, D., Klevnas, P., Foss, A., 2009. White certificate schemes: economic analysis and interactions with the EU ETS. Energy Policy 37 (2009), 29–42. Sorrell, S. (2001), Policy Design and Policy Interaction: Literature Review and Methodological Issues, for the INTERACT Program.

Wang, J., Jing, Y., Zhang, C., Zhao, J., 2009. Review on multi-criteria decision analysis aid in sustainable energy decision-making. Renewable and Sustainable Energy Reviews 13, 2263–2278. Wirl F., 2007, Oligopoly meets oligopsony: The case of permits 58, 3, November 2009, Pages 329–337.