- Email: [email protected]
An emissions trading scheme design for power industries facing price regulation
Energy Policy 75 (2014) 84–90
Contents lists available at ScienceDirect
Energy Policy journal homepage: www.elsevier.com/locate/enpol
An emissions trading scheme design for power industries facing price regulation Yong-Gun Kim a,n, Jong-Soo Lim b a b
Division of Climate and Air Policy, Korea Environment Institute, 215 Jinheungno, Eunpyeong-Gu, Seoul 122-706, Republic of Korea Department of International Trade, Kwangwoon University, 20 Kwangwoon-Ro, Nowon-Gu, Seoul 139-701, Republic of Korea
H I G H L I G H T S
A rate-based allocation induces power producers to minimize direct emissions. A cap-and-trade on indirect emission induces firms to reduce electricity consumption. These two can jointly achieve market efficiency even in the regulated power market.
art ic l e i nf o
a b s t r a c t
Article history: Received 26 June 2013 Received in revised form 25 June 2014 Accepted 8 July 2014 Available online 31 July 2014
The electricity market, monopolistic in nature, with government price regulation, poses a serious challenge for policy makers with respect to the cost-effectiveness of emissions trading, particularly in Asian countries. This paper argues that a cap-and-trade regulatory system for indirect emissions combined with a rate-based allocation system for direct emissions can achieve market efficiency even in the presence of price and quantity controls in the electricity market. This particular policy mix could provide appropriate incentives for industries to reduce their electricity consumption while inducing power producers to reduce their direct carbon emissions cost-effectively in conditions where there is strict government control of electricity prices. Another advantage of the suggested policy mix is that it allows carbon leakage in cross-border power trades to be effectively eliminated. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Emissions trading Power sector Indirect emissions
1. Introduction Emissions trading schemes (ETS) have been playing a central role in mitigating greenhouse gas (GHG) emissions within many countries, as well as internationally. The EU, Norway, Switzerland, Australia and New Zealand are implementing national ETS, whereas other countries, such as the USA, Canada, and Japan, run ETS at the provincial level. Recently, Korea passed an ETS law setting a goal of implementing an ETS by 2015. China, the largest emitter in the world, also plans to introduce a national ETS starting in 2016. Seven provincial Chinese governments are already in the final preparatory stages for local pilot ETS.1 It is well known that a Pigouvian tax can correct market inefficiency caused by an environmental externality. Within the context of climate change, a fully functional cap-and-trade system n
Corresponding author. Tel.: þ 82 2 380 7637. E-mail addresses: [email protected] (Y.-G. Kim), [email protected] (J.-S. Lim). 1 Refer to International Emissions Trading Association (IETA)’s website for the recent status of emissions trading schemes in the East Asian countries and other regions (www.ieta.org/worldcarbonmarkets). http://dx.doi.org/10.1016/j.enpol.2014.07.011 0301-4215/& 2014 Elsevier Ltd. All rights reserved.
with the carbon price set at the Pigouvian tax rate will also achieve the social optimum provided that market is under free competition.2 The key is to make sure that the price of carbon is passed through to the final consumption stage so that equilibrium consumption will stay at the socially optimal level. In practice, however, market interventions such as government price control causes carbon price to be incompletely conveyed onto the final price, resulting in the allocative inefficiency of the market equilibrium. The electricity industry requires special attention in this regard. Electricity production and use account for a large share of total emissions of greenhouse gases with the largest reduction potential.3 Most countries intervene in their electricity markets through
2 Inefficiencies due to market imperfections and transaction cost were thoroughly investigated in, among others, Sijm et al. (2012), Boom and Dijkstra(2009), and Stavins (1995). 3 According to the International Energy Agency, the electricity and heat production sector was responsible for 41.2% of global CO2 emissions from fossil fuel consumption in 2010. The electricity sector will account for between twothirds and three-quarters of emissions reductions in the next two decades in the USA (Burtraw, 2008).
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
price interventions, resulting in serious market distortions. If the price of electricity cannot be adjusted to reflect the market price of carbon, the lack of the appropriate pass-through of carbon costs to the final consumers of electricity caused by malfunctioning price adjustment mechanisms creates a distortion in energy markets, i.e., non-uniform carbon prices between electric power and other types of energy. The academic community has not reached a consensus on how the pass-through of carbon costs to the consumer price of electricity should be determined and when a certain ETS scheme produces efficient outcomes in the presence of market imperfections. Various ETS schemes have been tried around the world, although the systematic evaluation of their performance must still be conducted. Nelson et al. (2012) reviewed various modeling studies on the impact of carbon prices on electricity markets and found the studies to be entirely inconsistent in their estimation of carbon pass-through; the authors were unable to establish why the estimations vary so significantly.4 Sijm et al. (2012) present a theoretical analysis of the impact of the structure of the power market on the pass-through of carbon costs to electricity prices and conclude that the pass-through rate is significantly affected by market structure, such as the number of firms and government regulations. The success of economic incentive measures crucially depends on the carbon cost pass-through rate because this is the barometer for the efficiency of market mechanisms. This study aims to propose a desirable ETS structure that is suitable for Northeast Asia, including China, Japan and Korea, where electricity markets are strictly controlled by governments. In doing so, we compare various ETS schemes in terms of allocative efficiency and the ratio of carbon price that is passed through onto the final consumption stage under various policy options. Section 2 briefly describes the state of emissions trading schemes in Northeast Asia that involve the electric power industry. The current market situation of the Korean electric power industry is also briefly introduced. Section 3 examines how carbon price pass-through changes according to different choices in emissions trading schemes and provides graphical illustrations with discussions on efficiency implications. Policy recommendations for the design of emissions trading schemes for the electric power industry are provided in this section. Finally, Section 4 presents this study’s conclusions.
2. Emissions trading initiatives in the Northeast Asia and Korean electricity markets 2.1. Current status of ETS in Northeast Asian region Three major Northeast Asian economies, China, Japan, and Korea, plan to introduce ETS as part of their GHG mitigation policies, though at different stages. The electricity industry in all these countries creates serious challenges for the design of ETS systems. Korea enacted a law that mandates the implementation of an ETS in 2015, the first in the region. Korea also implemented the Target Management Scheme (TMS), which is conceptualized as a transitional ETS policy tool, in 2012. Although it does not allow for trading, the TMS shares key elements with ETS: emissions targets for individual sources of emissions and MRV infrastructure. It is noteworthy that under the TMS, both direct emissions from electricity production and indirect emissions from 4 Kim et al. (2010) showed that the pass-through of carbon costs can vary drastically across different systems through a comparative analysis of Australia and Korea.
85
electricity consumption are regulated by the authorities. Interestingly, rate-based intensity targets are applied to direct emissions whereas absolute targets are required for indirect emissions. It is not yet clear, however, whether the Korean government will maintain this structure in the forthcoming implementation of an ETS. In China, emissions carbon trading schemes have made significant progress, with five out of the seven provinces having started operation of pilot schemes in 2013. Besides direct emissions, indirect emissions caused by the consumption of outsourced power or heat are covered by the pilot schemes.5 Japan has also studied the possibility of an ETS for the electricity sector. An official document from the Japanese government (Japanese Government, 2010) describes three options for a Japanese capand-trade system. Two of these options propose the regulation of indirect emissions from electricity consumption combined with intensity regulation for the direct emissions of power producers. The emissions trading scheme of the Tokyo Metropolitan Area and other municipalities, which has been running since 2010, regulates indirect emissions from electricity consumption only, not direct emissions from electricity production (Table 1). It is interesting to see that the regulation of indirect emissions is an important part of ETS in these three Northeast Asian countries to circumvent the incomplete pass-through of carbon prices due to government control of electricity prices. Intensity regulations for direct emissions are simultaneously regarded as being complementary in incentivizing power producers to reduce their direct emissions. 2.2. Structure of Korean electricity market Industries emit greenhouse gases in the process of providing goods and services. Although they are not direct emitters, households are responsible for the emissions of industries because their demand is the ultimate cause of those emissions. However, an absolute majority of climate-change-related regulations falls on producers. Provided that price mechanisms properly reveal the cost structure, regulations on producers can alter economic consumption behavior toward the direction intended by policymakers. We are concerned with cases in which the price mechanism does not work properly. When there are price controls, for example, cost hikes due to CO2 regulations on the producers will not be properly conveyed onto the final purchase price. The policy goal of reducing the consumption of targeted products would be under-achieved. The Korean electricity market represents a typical example of this type of case. The Korean electricity industry consists of four sub-industries: electricity generation, transmission, distribution, and sales. The Korean electricity industry was a vertically integrated monopoly until 2001. Since 2001, the upstream power generation monopoly has been divided into six companies. The downstream power transmission, distribution, and final sales supply chains are still monopolized by a public enterprise, KEPCO. The need for market connections between power-generating companies and KEPCO prompted the birth of the Korea Power Exchange (KPX) electricity marketplace. The Korean electricity market is managed through a method known as the ‘Cost-Based Pool’. In this system, the price and quantity of generated electricity are determined. The way in which they are decided is in stark contrast with a typical Walrasian tatonnemont process. The KPX-commissioned ‘Cost Evaluation Committee’ determines the price of electricity based on the cost 5
Refer to Duan et al. (2014) for the detail.
86
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
Table 1 Examples of design options for the electricity sector under ETS. Point of compliance and target
Practicing country or region
Direct emissions
Indirect emissions
Absolute target None Absolute target Relative (intensity) target
None Absolute target Absolute target Absolute target
data submitted by the power-generating companies.6 The powergenerating companies do not bid price, but report their supplying capacities. This system reflects the Korean situation of a fairly monopolistic electricity market structure. The need for monopoly regulation also influences the retail market. The government also determines the retail price of electricity on a ‘gross cost’ basis. The ‘gross cost’ is the variable cost of production plus a ‘reasonable’ return on capital investments. In addition to this ‘average pricing’ practice, various policy considerations, such as energy security, protection of the agrofishery industry, industrial support, general price management, and income redistribution, are given to various end-user groups. As a result, a very complicated final consumer price of electricity has emerged. In this price system, some groups of consumers end up subsidizing other groups by paying more than the average cost of power generation while other groups of end users pay less. The issue of cross-subsidization among different end-user groups caused by this complicated final consumer price system is one of the key issues mentioned at recent debates on reforming the Korean electricity market. In sum, the Korean electricity market has two characteristics that keep traditional ETS from functioning properly. On the one hand, both retail and wholesale electricity prices are tightly controlled by the government. This would cause a hiatus effect that works against the carbon cost pass-through of the electricity market. On the other hand, the two distinctly different principles that guide the setting of the retail and wholesale prices of electricity reduce the possibility that the government will set prices that correctly reflect the cost of carbon. In what follows, we propose an ETS system that overcomes these difficulties while taking the characteristics of the Korean electricity market as given. In so doing, we mimic a free-market carbon cost pass-through mechanism by providing different incentives at different points of compliance. This proposal could provide insights for the design of the Chinese and Japanese ETSs because their electricity markets share similar characteristics with the Korean electricity market.7
3. Evaluation of policy options Let there be two types of emitters in the economy: the ‘electricity industry (E-sector)’ and ‘non-electric sector (NE-sectors),’ represented by the subscripts ‘e’ and ‘ne’, respectively. The NE sector includes household as well as industries other than those included in the E sector. We assume, for now, that both the E 6 The System Marginal Price (SMP, a type of marginal cost) and Capacity Payment (CP, a type of average fixed cost) are the two components of the reimbursement price. 7 According to Baron and Aasrud (2012), the construction and commissioning of large power plants are authorized by the National Development and Reform Commission (NDRC), which, through its price department, determines electricity prices that allow for an adequate financial return. The authors suggested some specific policy changes in the Chinese power sector, including a change to plant dispatch regulations to allow a shift from high- to low-emissions sources because the price of CO2 changes their respective profitabilities.
EU, Australia, New Zealand Tokyo, UK Pilot phase ETS in seven Chinese provinces Korean Target Management Scheme, Japan (under consideration)
and NE sectors are regulated by emissions trading schemes. The total national carbon emissions X, then, are the sum of the carbon emissions of the E sector, X e and those of the NE sector, X ne . X ¼ X e þ X ne Our goal is to control the amount of electricity-related carbon emissions through an emissions trading scheme. When governments design policy mixes for mitigation, two key aspects of regulation must be clearly delineated: The point of compliance, such as upstream versus downstream or producers versus consumers, and the form of compliance target, such as fixed emissions caps (absolute target) or emissions intensity (relative target). In the case of the electricity sector, the most popular and thoroughly discussed point of compliance has been the direct emissions of power producers. The introduction of regulations on the indirect emissions of electricity consumers, however, has recently been gaining momentum. We consider both cases in this article.8 The effect of a policy also depends on the type of compliance target. The urgent need for greenhouse gas mitigation tends to call for absolute emissions targets, whereas concern for the excessive attenuation of economic activity tends to promote relative intensity targets over absolute targets.9 Emissions trading schemes based on absolute (total) emissions targets are known as ‘capand-trade’ programs, which are a common type of ETS used to regulate CO2. Emissions trading schemes based on relative emissions (carbon intensity) targets, known as ‘rate-based emissions trading,’ regulates emissions per unit of output such as tCO2e/ MW h. The Low Carbon Fuel Standard (LCFS) launched in California is a typical rate-based emissions trading scheme.10 The Renewable Portfolio Standard (RPS) and the Renewable Fuel Standard (RFS) are examples of environmental regulation based on the similar concept as rate-based trading. Different combinations of points of compliance and types of targets yield four institutional options, as seen in Table 2.11 Because our focus is on the efficiency consequences of ETS for the electricity sector, we simply assume, without loss of generality, that direct emissions by the NE sector are zero ðX ne ¼ 0Þ. We will analyze the effect of the three options on carbon price passthrough and the possible combinations thereof: Is the passthrough of the carbon price to the consumer price of electricity too low, appropriate or too high? Option A is a cap-and-trade policy on the direct emissions of the regulated sectors. The goal is to reach the emissions target X. 8 Refer to Burtraw (2008) for a detailed discussion on how to regulate CO2 in electricity markets, particularly on the point of compliance, emissions sources versus electricity consumers. 9 Refer to Fischer (2003) for the pros and cons of two emissions trading systems based on absolute and relative emissions targets. The author also evaluates the economic and environmental impacts of linking the two different types of emissions markets and investigates various measures to mitigate the adverse effects of such linkages. 10 Refer to Holland et al. (2009) for a detailed analysis on economic and environmental effects of LCFSs. 11 Option D is not considered here because carbon intensity regulation on indirect emissions is neither effective nor implementable in practice.
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
87
Table 2 Institutional options for ETS on emissions from the power sector. Compliance target
Point of compliance Electricity producer (direct emissions) Electricity consumer (indirect emissions)
Total emissions (cap-and-trade)
Carbon intensity (rate-based trading))
Option A Option C
Option B Option D (not considered)
Option 1 can guarantee that the goal X is precisely met.12 Xe ¼ X Under a cap-and-trade scheme, the market price of carbon, P C , passes through onto retail prices provided that no price regulation exists in the market. For the electricity market, the fact that the price shock falls on the power generators implies that the market price of carbon may not be fully reflected in the final consumer price, should there be price controls in the electricity market. In this case, demand for electricity may not properly adjust itself to achieve an efficient allocation of resources. Fig. 1, adapted from Lanz and Rausch (2013), provides efficiency implications of ETS systems in a partial equilibrium framework. In the absence of environmental regulation, initial equilibrium price and quantity of electricity are P0e and Q0e , respectively, determined at the intersection of the market demand curve (D0) and the average cost curve (AC0) of the monopolist power sector. Initial emissions intensity of E sector is the slope of I0e line.13 Resulting direct emissions of E-sector with a given initial emissions intensity I0e is X0e (¼Q0e I0e ). Let’s assume that (Q*e, X*e) yields the socially optimal allocation of resources. When an emissions cap X*e is enforced, i.e., X ¼ X ne , and emissions allowances are auctioned at price P*c, average cost of E sector is increased by P*e I*e to a new AC0 þ P*e I*e, resulting in a new equilibrium (P*e, Q*e). One thing to note is that, in the new equilibrium, the emissions intensity is changed from I0e to I*e according to the cost minimizing behavior of the E sector. If the emissions allowances are distributed for free, a windfall profit of the size P*c X*e would occur. Price regulation such as costof-service regulation would lower the final consumption price of electricity (P*e-Pe0 ) so as to make E sector’s profit to be 0. (Shaded areas in the 1st and 3rd quadrant are equal in size.) At this price, consumption level would increase to Qe0 , and emissions intensity needed to be adjusted to an even lower Ie0 to be able to comply with the emissions cap regulation. Price of carbon would increase to Pc0 under usual assumptions of increasing marginal abatement cost. Carbon price, however, would be incompletely passed through onto the final consumption price (Pe0 o P*e), resulting in an inefficient outcome compared to our benchmark case (Qe0 a Q*e). Option B is a carbon intensity (rate based) regulation on direct emissions, with trading allowed between power producers who over- or under-achieve the intensity targets. Under carbon intensity regulations, the electricity sector must meet its assigned intensities, I e . Therefore, Ie
Xe rI e Qe
12 The other options are also capable of meeting this target, on average, for a certain budget period but, in some cases, it is technically difficult to do this on a yearly basis. Whether to meet the goal on a yearly basis or for a certain budget period would be a non-issue. Considering the climate change and greenhouse gas concentration change mechanisms, it would be desirable to meet the target, on average, on longer budget periods. 13 Carbon intensity, defined as the ratio of ‘emissions due to production' to the ‘produced output', can be defined as, Ie X e =Q e .
Carbon intensity regulations on the electricity industry will have similar consequences to those of cap-and-trade schemes in the sense that the cost of carbon faced by producers is the same. Nonetheless, it is well documented that carbon intensity regulations have an effect similar to that of product subsidies.14 When the power supply increases by ΔQ e , free allowance also increases by ΔQ e I e , leading to a subsidy-like financial gain of ΔQ e I e P c . It follows that the marginal cost of power generation decreases by I e P c , resulting in lower retail prices and higher consumption of electricity compared to those under a cap-andtrade scheme. The unit carbon price that is passed through onto the final retail price would then be P c ðI e =I e Þ P c . We can see from this that carbon price pass-through would be 0 if the E sector’s carbon intensity improves to meet the emissions intensity regulation, i.e., I e ¼ I e . Fig. 2 shows this change. Let’s assume that the government, with the full information on the cost structure of E sector, set the target emissions intensity to be I*e aiming to reach the emissions target X*e. Initial upward movement of the average cost curve from AC0 to AC0 þP*cI*e will move the equilibrium point from (P0e , Q0e ) to (P*e, Q*e). The output subsidy effect of intensity regulation, however, moves average cost curve back to its original position. Equilibrium point will then move back to its original position (P0e , Q0e ). Clearly at this point emissions target X*e implicitly set by the government cannot be achieved unless the target emissions intensity is set at a much lower level IBe which is inferior in terms of cost efficiency compared to the benchmark emissions coefficient I*e. Since the equilibrium price does not reflect the carbon price, windfall profits do not occur here and the government intervention on electricity price such as cost-of-service regulation does not affect the equilibrium under the intensity regulation. Under an absolute target on indirect emissions by NE sector (Option C), an electricity consumers (NE sector) are responsible for their indirect emissions, calculated by the amount of electricity consumed (Qe) multiplied by a pre-determined emissions factor, IIe. Under Option C, carbon costs face by the NE sector would be P c Q e I Ie and the effective carbon price for the direct emissions induced by a unit of electricity consumed is ðI Ie =I e Þ P c . If the applied emissions factor is indeed equal to the actual carbon intensity, the market price of carbon will be fully reflected to the final consumer price of electricity, which, in turn, will adjust the demand for electricity to the most efficient level. However, the electricity consumers, NE sector, can only adjust their electricity consumption to meet the target. The target can neither affect the energy mix decisions of the electricity industry nor select power taps based on the carbon intensity of the supplied power. Therefore, option C fails to provide appropriate incentives for power producers to improve their carbon intensities. Let the emissions cap on the indirect emissions be X ¼ X n . In Fig. 2, the electricity price hike consumers face will shift the demand curve downward by PCc I0e , resulting in a new demand curve D0 PCc I0e . The corresponding level of consumption is QCe
14 Refer to Burtraw et al. (2001) for an in-depth assessment of the output subsidy effect of rate-based allocations.
88
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
Fig. 1. A cap-and-trade scheme with free emissions allowances.
Fig. 2. Intensity target on direct emissions and absolute target on indirect emissions.
without any change in emissions intensity. Clearly, E sector would have no incentive to improve their emissions intensity since they do not have any emissions reduction requirement. Here, the carbon price pass-through is nearly perfect as the point of compliance is at the final consumer. However, this new equilibrium (PCe , QCe ) is sacrificing too much consumption and investing too little on the emissions reduction technology, and thus is inferior to the benchmark case equilibrium (P*e, Q*e). So far we have found the following. (1) Absolute targets on direct emissions with cost-of-service regulation may not yield
efficient outcome due to an inappropriate pass-through of carbon price. (2) Carbon intensity regulations on direct emissions are relatively free of the carbon price pass-through concern but their ‘production subsidy effect’ tends to attenuate incentives to reduce electricity consumption. (3) Emissions caps on indirect emissions work through a demand control mechanism steered by the adding of the cost of carbon to the final consumer price. However, it provides few incentives for power producers to improve their fuel mixes or production efficiency. In short, these 3 options are either incomplete in and of their selves or ineffective in the presence of
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
market intervention by the government. A stronger policy mix is needed. A policy mix that combines options A and C would involve the application of emissions caps on both direct and indirect emissions. A carbon price Pc for the emissions cap on direct emissions will pass through to the final consumption price of electricity, at least in part even under the presence of cost-of-service regulation. On top of that, the carbon cost of ðI Ie =I e Þ P c due to the indirect emissions cap will be added to the final consumption price of electricity, reaching up to P c þ ðI Ie =I e Þ P c per carbon unit. This is nearly double the market price of carbon, raising concern of excessive regulation. Another policy mix that combines options B and C involves the application of carbon intensity regulations on direct emissions and an emissions cap on indirect emissions. Carbon intensity regulations tend to under-shoot the carbon price, P c ðI e =I e Þ P c , due to the subsidy effect, as discussed for option B. Absolute caps on indirect emissions embedded in the electricity consumption of the NE sector, however, raise the carbon price by ðI Ie =I e Þ P c . These two effects can mimic the correct carbon price P if the applied emissions factor ðI Ie Þ and carbon intensity target ðI e Þ are set equal to each other. Therefore, this policy mix conveys the correct price signal to the final consumer while providing power-generating companies with incentives to improve their carbon efficiency. In Fig. 2, the target emissions intensity set at I*e with emissions cap X*e * on the indirect emissions attains a new equilibrium (PBC e , Qe) with * the final cost of electricity consumption Pe. This equilibrium is practically identical to the efficient outcome (benchmark case equilibrium). Note that since the equilibrium price of electricity PBC does not reflect the carbon price Pc, windfall profits do not e occur. Therefore the cost-of-service regulation does not affect the equilibrium price under the policy mix B þC. Table 3 summarizes the carbon price pass-through to the price of electricity under the 5 policy options discussed above. We have shown that the pass-through of the price of carbon functions the best under the policy mix that combines policy options B and C. Under this policy mix, a single carbon price prevailing across the economy is applied to both electricity consumption and fossil fuel combustion, regardless of whether there are interventions in the electricity market. Table 3 also shows that the sum of effective emissions targets for the E and NE sectors under the policy mix B þ C becomes identical to the policy objective if the emissions factor ðI Ie Þ is set equal to the intensity target ðI e Þ. That is, the certainty of absolute emissions control, one of important advantages of cap-and-trade systems for direct emissions (option A), can also be guaranteed under the policy mix that combines relative targets for direct emissions (option B) and absolute targets for indirect emissions (option C).
4. Concluding remarks There are three key elements in the design of emissions trading systems: point of compliance, mode of compliance target, and initial distribution of permits. In this paper, we mainly reviewed the carbon price pass-through aspect of the first two elements of ETS systems in the context of the electricity industry. The electricity industry is peculiar in the sense that it suffers from heavy government regulation and market concentration in most cases. We determined that an ETS scheme that combines relative (intensity) targets on direct emissions with absolute emissions caps on indirect emissions performs better than cap-and-trade on direct emissions and any combination thereof considered in this study, in the presence of market imperfection and government price regulations under which the normal carbon price pass-through
89
Table 3 Carbon price pass-through and allowance market under various policy options. Policy option
Carbon price Production (direct emissions)
Allowance market Consumption (indirect Demand emissions) (covered emissions)
Supply (allowance allocation)
E NE sector sector
E sector
A
Pc
Pc
Xe
X
B
Pc
Xe
Ie Qe
C
–
Ie Pc Pc Ie IIe Pc Ie Ie Pc þ Pc Ie Ie Ie Pc Pc þ Pc Ie Ie
AþC
Pc
B þC
Pc
NE sector
–
I Ie Q e –
X
Xe
I Ie
X
Xe
I Ie Q e I e Q e X
Qe X
mechanisms of the competitive market could not function. This ETS scheme provides correct price incentives for consumers to adjust their electricity consumption in line with an efficient allocation of other types of energy. Intensity regulation provides power producers with incentives to reduce their carbon emissions, whereas the output subsidy effect of rate-based updating allocation helps prevent the double burden of carbon costs on electricity consumers. Moreover, this policy mix can be used to expand ETS coverage to include the commercial and public buildings and residential sector, which consume a significant share of energy but usually remain outside ETS schemes due to their heavy reliance on electricity and low levels of direct emissions. The product subsidy effect of intensity regulations reduce the gap between the carbon costs faced by regulated and unregulated sectors, while cap-andtrade on indirect emissions tends to expand the scope of the sectors covered by the policy. These two effects combined can attenuate the leakage caused by regulatory gaps, leading to the second-best solution. Holland (2012) proved that, in the presence of unregulated sectors, a policy mix that combines intensity standards with consumption taxes dominates a policy mix that combines cap-and-trade regulations and consumption taxes. It is interesting to note that the policy mix that combines policy options B and C mimics the former whereas the policy mix that combines policy options A and C mimics the latter. Therefore, we can conclude that a policy mix that combines intensity regulations on direct emissions and emissions caps on indirect emissions can work efficiently even in the presence of incomplete regulation (leakage). Another potential benefit of the intensity regulation of the power sector is that it helps prevent the unintended pass-through of carbon prices to electricity prices in non-ETS sectors, such as the commercial or residential sectors. Under usual cap-and-trade regulations on direct emissions, non-ETS energy consumers face higher electricity prices due to carbon price pass-through from power producers in the ETS sector, while they do not pay for the carbon content of other energy sources. This may cause distortions in the energy market and consumption structure if the government does not implement some countermeasures to mitigate discrepancies in the relative prices of electricity and other types of energy through, for example, carbon taxes on non-electric energy in non-ETS sectors. However, under intensity regulations of the direct emissions of electricity producers, the output subsidy effect impedes the carbon price from passing through onto the final price, thus preventing distortions in energy prices among
90
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
different energy products that may or may not be covered by the ETS. The proposed policy mix is a relatively new concept that may suffer from criticisms of double regulation and higher administrative costs. A mere fact that two regulatory instrument are imposed on a single activity (carbon emissions) raises a concern from industries. They tend to imagine that excessive carbon cost like that of policy mix of option A þC would be imposed on them. Furthermore, administrative cost of suggested policy mix would be inevitably higher than a single policy option cases. However, considering its greenhouse gas reduction potential and its effect on the national economy, the additional complexity and administrative costs of the proposed ETS scheme on the electricity sector might be justified. Furthermore, our analysis demonstrates that the key characteristic of the suggested policy mix is a ‘correct price signal’ to both producers and consumers, showing that the worries of excessive carbon cost burden is a baseless concern. The justification for the suggested policy mix strongly holds especially when any single, simple regulation causes a serious inefficiency in complex situations with multiple policy constraints.
Acknowledgements This research is partially supported by the research project of Korea Environment Institute (“A study on consumption behavior and low carbon society 2013”), Ministry of Environment (“Allowance allocation for power industry”) and 2013 Kwangwoon University research grant.
References Baron, Richard., Aasrud, Andre., 2012. An Emissions Trading System for China’s Power Sector, Electricity in a Climate-constrained World. International Energy Agency. Boom, Jan-Tjeered., Dijkstra, B.R., 2009. Permit trading and credit trading: a comparison of cap-based and rate-based emission trading under perfect and imperfect competition. Environ. Res. Econ. 44, 107–136. Butraw, Dallas., Palmer, K., Bharvirkar, R., Paul, A. 2001. The Effect of Allowance Allocation on the Cost of Carbon Emission Trading, Resources for the Future Discussion Paper 01-03, Resources for the Future, Washington, DC. Burtraw, Dallas., 2008. Regulating CO2 in electricity markets: sources or consumers? Clim. Policy 8 (6), 588–606. Fisher, Carolyn., 2003. Combining rate-based and cap-and-trade emissions policies. Clim. Policy 3 (2), S89–S103. Scheme Options for Japanese Emissions Trading Scheme Based on Cap-and-trade System (Interim Report), Central Environmental Council. Domestic Emissions Trading Subcommittee. Kim, Wook., Chattopadhyay, Deb., Park, Jong-bae., 2010. Impact of carbon cost on wholesale electricity price: a note on price pass-through issues. Energy 35 (8), 3441–3448. Holland, Stephen P., 2012. Emissions taxes versus intensity standards: second-best environmental policies with incomplete regulation. J. Environ. Econ. Manage. 63 (3), 375–387. Holland, Stephen, P., Hughes, Jonathan E., Knittel, Christopher R., 2009. Greenhouse gas reductions under low carbon fuel standards. Am. Econ. J.: Econ. Policy vol. 1 (1), 106–146. Lanz, B., Rausch, S., 2013, Cap-and-Trade Climate Policy, Free Allowances, and Price Regulated Firms, Working Paper 12/178, Economic Working Paper Series, ETH. Nelson, Tim., Kelley, Simon, Orton, Fiona., 2012. A literature review of economic studies on carbon pricing and Australian wholesale electricity markets. Energy Policy 49, 217–224. Duan, Maosheng, Pang, Tao, Zhang, Xiliang, 2014. Review of carbon emissions trading polits in China. Energy Environ. 25 (3), 527–549. Sijm, Jos., Chen, Yihsu, Hobbs, Benjamin F., 2012. The impact of power market structure on CO2 cost pass-through to electricity prices under quantity competition—a theoretical approach. Energy Econ. 34 (4), 1143–1152. Stavins, Robert N., 1995. Transaction costs and tradable permits. J. Environ. Econ. Manage. 29, 133–148.
Contents lists available at ScienceDirect
Energy Policy journal homepage: www.elsevier.com/locate/enpol
An emissions trading scheme design for power industries facing price regulation Yong-Gun Kim a,n, Jong-Soo Lim b a b
Division of Climate and Air Policy, Korea Environment Institute, 215 Jinheungno, Eunpyeong-Gu, Seoul 122-706, Republic of Korea Department of International Trade, Kwangwoon University, 20 Kwangwoon-Ro, Nowon-Gu, Seoul 139-701, Republic of Korea
H I G H L I G H T S
A rate-based allocation induces power producers to minimize direct emissions. A cap-and-trade on indirect emission induces firms to reduce electricity consumption. These two can jointly achieve market efficiency even in the regulated power market.
art ic l e i nf o
a b s t r a c t
Article history: Received 26 June 2013 Received in revised form 25 June 2014 Accepted 8 July 2014 Available online 31 July 2014
The electricity market, monopolistic in nature, with government price regulation, poses a serious challenge for policy makers with respect to the cost-effectiveness of emissions trading, particularly in Asian countries. This paper argues that a cap-and-trade regulatory system for indirect emissions combined with a rate-based allocation system for direct emissions can achieve market efficiency even in the presence of price and quantity controls in the electricity market. This particular policy mix could provide appropriate incentives for industries to reduce their electricity consumption while inducing power producers to reduce their direct carbon emissions cost-effectively in conditions where there is strict government control of electricity prices. Another advantage of the suggested policy mix is that it allows carbon leakage in cross-border power trades to be effectively eliminated. & 2014 Elsevier Ltd. All rights reserved.
Keywords: Emissions trading Power sector Indirect emissions
1. Introduction Emissions trading schemes (ETS) have been playing a central role in mitigating greenhouse gas (GHG) emissions within many countries, as well as internationally. The EU, Norway, Switzerland, Australia and New Zealand are implementing national ETS, whereas other countries, such as the USA, Canada, and Japan, run ETS at the provincial level. Recently, Korea passed an ETS law setting a goal of implementing an ETS by 2015. China, the largest emitter in the world, also plans to introduce a national ETS starting in 2016. Seven provincial Chinese governments are already in the final preparatory stages for local pilot ETS.1 It is well known that a Pigouvian tax can correct market inefficiency caused by an environmental externality. Within the context of climate change, a fully functional cap-and-trade system n
Corresponding author. Tel.: þ 82 2 380 7637. E-mail addresses: [email protected] (Y.-G. Kim), [email protected] (J.-S. Lim). 1 Refer to International Emissions Trading Association (IETA)’s website for the recent status of emissions trading schemes in the East Asian countries and other regions (www.ieta.org/worldcarbonmarkets). http://dx.doi.org/10.1016/j.enpol.2014.07.011 0301-4215/& 2014 Elsevier Ltd. All rights reserved.
with the carbon price set at the Pigouvian tax rate will also achieve the social optimum provided that market is under free competition.2 The key is to make sure that the price of carbon is passed through to the final consumption stage so that equilibrium consumption will stay at the socially optimal level. In practice, however, market interventions such as government price control causes carbon price to be incompletely conveyed onto the final price, resulting in the allocative inefficiency of the market equilibrium. The electricity industry requires special attention in this regard. Electricity production and use account for a large share of total emissions of greenhouse gases with the largest reduction potential.3 Most countries intervene in their electricity markets through
2 Inefficiencies due to market imperfections and transaction cost were thoroughly investigated in, among others, Sijm et al. (2012), Boom and Dijkstra(2009), and Stavins (1995). 3 According to the International Energy Agency, the electricity and heat production sector was responsible for 41.2% of global CO2 emissions from fossil fuel consumption in 2010. The electricity sector will account for between twothirds and three-quarters of emissions reductions in the next two decades in the USA (Burtraw, 2008).
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
price interventions, resulting in serious market distortions. If the price of electricity cannot be adjusted to reflect the market price of carbon, the lack of the appropriate pass-through of carbon costs to the final consumers of electricity caused by malfunctioning price adjustment mechanisms creates a distortion in energy markets, i.e., non-uniform carbon prices between electric power and other types of energy. The academic community has not reached a consensus on how the pass-through of carbon costs to the consumer price of electricity should be determined and when a certain ETS scheme produces efficient outcomes in the presence of market imperfections. Various ETS schemes have been tried around the world, although the systematic evaluation of their performance must still be conducted. Nelson et al. (2012) reviewed various modeling studies on the impact of carbon prices on electricity markets and found the studies to be entirely inconsistent in their estimation of carbon pass-through; the authors were unable to establish why the estimations vary so significantly.4 Sijm et al. (2012) present a theoretical analysis of the impact of the structure of the power market on the pass-through of carbon costs to electricity prices and conclude that the pass-through rate is significantly affected by market structure, such as the number of firms and government regulations. The success of economic incentive measures crucially depends on the carbon cost pass-through rate because this is the barometer for the efficiency of market mechanisms. This study aims to propose a desirable ETS structure that is suitable for Northeast Asia, including China, Japan and Korea, where electricity markets are strictly controlled by governments. In doing so, we compare various ETS schemes in terms of allocative efficiency and the ratio of carbon price that is passed through onto the final consumption stage under various policy options. Section 2 briefly describes the state of emissions trading schemes in Northeast Asia that involve the electric power industry. The current market situation of the Korean electric power industry is also briefly introduced. Section 3 examines how carbon price pass-through changes according to different choices in emissions trading schemes and provides graphical illustrations with discussions on efficiency implications. Policy recommendations for the design of emissions trading schemes for the electric power industry are provided in this section. Finally, Section 4 presents this study’s conclusions.
2. Emissions trading initiatives in the Northeast Asia and Korean electricity markets 2.1. Current status of ETS in Northeast Asian region Three major Northeast Asian economies, China, Japan, and Korea, plan to introduce ETS as part of their GHG mitigation policies, though at different stages. The electricity industry in all these countries creates serious challenges for the design of ETS systems. Korea enacted a law that mandates the implementation of an ETS in 2015, the first in the region. Korea also implemented the Target Management Scheme (TMS), which is conceptualized as a transitional ETS policy tool, in 2012. Although it does not allow for trading, the TMS shares key elements with ETS: emissions targets for individual sources of emissions and MRV infrastructure. It is noteworthy that under the TMS, both direct emissions from electricity production and indirect emissions from 4 Kim et al. (2010) showed that the pass-through of carbon costs can vary drastically across different systems through a comparative analysis of Australia and Korea.
85
electricity consumption are regulated by the authorities. Interestingly, rate-based intensity targets are applied to direct emissions whereas absolute targets are required for indirect emissions. It is not yet clear, however, whether the Korean government will maintain this structure in the forthcoming implementation of an ETS. In China, emissions carbon trading schemes have made significant progress, with five out of the seven provinces having started operation of pilot schemes in 2013. Besides direct emissions, indirect emissions caused by the consumption of outsourced power or heat are covered by the pilot schemes.5 Japan has also studied the possibility of an ETS for the electricity sector. An official document from the Japanese government (Japanese Government, 2010) describes three options for a Japanese capand-trade system. Two of these options propose the regulation of indirect emissions from electricity consumption combined with intensity regulation for the direct emissions of power producers. The emissions trading scheme of the Tokyo Metropolitan Area and other municipalities, which has been running since 2010, regulates indirect emissions from electricity consumption only, not direct emissions from electricity production (Table 1). It is interesting to see that the regulation of indirect emissions is an important part of ETS in these three Northeast Asian countries to circumvent the incomplete pass-through of carbon prices due to government control of electricity prices. Intensity regulations for direct emissions are simultaneously regarded as being complementary in incentivizing power producers to reduce their direct emissions. 2.2. Structure of Korean electricity market Industries emit greenhouse gases in the process of providing goods and services. Although they are not direct emitters, households are responsible for the emissions of industries because their demand is the ultimate cause of those emissions. However, an absolute majority of climate-change-related regulations falls on producers. Provided that price mechanisms properly reveal the cost structure, regulations on producers can alter economic consumption behavior toward the direction intended by policymakers. We are concerned with cases in which the price mechanism does not work properly. When there are price controls, for example, cost hikes due to CO2 regulations on the producers will not be properly conveyed onto the final purchase price. The policy goal of reducing the consumption of targeted products would be under-achieved. The Korean electricity market represents a typical example of this type of case. The Korean electricity industry consists of four sub-industries: electricity generation, transmission, distribution, and sales. The Korean electricity industry was a vertically integrated monopoly until 2001. Since 2001, the upstream power generation monopoly has been divided into six companies. The downstream power transmission, distribution, and final sales supply chains are still monopolized by a public enterprise, KEPCO. The need for market connections between power-generating companies and KEPCO prompted the birth of the Korea Power Exchange (KPX) electricity marketplace. The Korean electricity market is managed through a method known as the ‘Cost-Based Pool’. In this system, the price and quantity of generated electricity are determined. The way in which they are decided is in stark contrast with a typical Walrasian tatonnemont process. The KPX-commissioned ‘Cost Evaluation Committee’ determines the price of electricity based on the cost 5
Refer to Duan et al. (2014) for the detail.
86
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
Table 1 Examples of design options for the electricity sector under ETS. Point of compliance and target
Practicing country or region
Direct emissions
Indirect emissions
Absolute target None Absolute target Relative (intensity) target
None Absolute target Absolute target Absolute target
data submitted by the power-generating companies.6 The powergenerating companies do not bid price, but report their supplying capacities. This system reflects the Korean situation of a fairly monopolistic electricity market structure. The need for monopoly regulation also influences the retail market. The government also determines the retail price of electricity on a ‘gross cost’ basis. The ‘gross cost’ is the variable cost of production plus a ‘reasonable’ return on capital investments. In addition to this ‘average pricing’ practice, various policy considerations, such as energy security, protection of the agrofishery industry, industrial support, general price management, and income redistribution, are given to various end-user groups. As a result, a very complicated final consumer price of electricity has emerged. In this price system, some groups of consumers end up subsidizing other groups by paying more than the average cost of power generation while other groups of end users pay less. The issue of cross-subsidization among different end-user groups caused by this complicated final consumer price system is one of the key issues mentioned at recent debates on reforming the Korean electricity market. In sum, the Korean electricity market has two characteristics that keep traditional ETS from functioning properly. On the one hand, both retail and wholesale electricity prices are tightly controlled by the government. This would cause a hiatus effect that works against the carbon cost pass-through of the electricity market. On the other hand, the two distinctly different principles that guide the setting of the retail and wholesale prices of electricity reduce the possibility that the government will set prices that correctly reflect the cost of carbon. In what follows, we propose an ETS system that overcomes these difficulties while taking the characteristics of the Korean electricity market as given. In so doing, we mimic a free-market carbon cost pass-through mechanism by providing different incentives at different points of compliance. This proposal could provide insights for the design of the Chinese and Japanese ETSs because their electricity markets share similar characteristics with the Korean electricity market.7
3. Evaluation of policy options Let there be two types of emitters in the economy: the ‘electricity industry (E-sector)’ and ‘non-electric sector (NE-sectors),’ represented by the subscripts ‘e’ and ‘ne’, respectively. The NE sector includes household as well as industries other than those included in the E sector. We assume, for now, that both the E 6 The System Marginal Price (SMP, a type of marginal cost) and Capacity Payment (CP, a type of average fixed cost) are the two components of the reimbursement price. 7 According to Baron and Aasrud (2012), the construction and commissioning of large power plants are authorized by the National Development and Reform Commission (NDRC), which, through its price department, determines electricity prices that allow for an adequate financial return. The authors suggested some specific policy changes in the Chinese power sector, including a change to plant dispatch regulations to allow a shift from high- to low-emissions sources because the price of CO2 changes their respective profitabilities.
EU, Australia, New Zealand Tokyo, UK Pilot phase ETS in seven Chinese provinces Korean Target Management Scheme, Japan (under consideration)
and NE sectors are regulated by emissions trading schemes. The total national carbon emissions X, then, are the sum of the carbon emissions of the E sector, X e and those of the NE sector, X ne . X ¼ X e þ X ne Our goal is to control the amount of electricity-related carbon emissions through an emissions trading scheme. When governments design policy mixes for mitigation, two key aspects of regulation must be clearly delineated: The point of compliance, such as upstream versus downstream or producers versus consumers, and the form of compliance target, such as fixed emissions caps (absolute target) or emissions intensity (relative target). In the case of the electricity sector, the most popular and thoroughly discussed point of compliance has been the direct emissions of power producers. The introduction of regulations on the indirect emissions of electricity consumers, however, has recently been gaining momentum. We consider both cases in this article.8 The effect of a policy also depends on the type of compliance target. The urgent need for greenhouse gas mitigation tends to call for absolute emissions targets, whereas concern for the excessive attenuation of economic activity tends to promote relative intensity targets over absolute targets.9 Emissions trading schemes based on absolute (total) emissions targets are known as ‘capand-trade’ programs, which are a common type of ETS used to regulate CO2. Emissions trading schemes based on relative emissions (carbon intensity) targets, known as ‘rate-based emissions trading,’ regulates emissions per unit of output such as tCO2e/ MW h. The Low Carbon Fuel Standard (LCFS) launched in California is a typical rate-based emissions trading scheme.10 The Renewable Portfolio Standard (RPS) and the Renewable Fuel Standard (RFS) are examples of environmental regulation based on the similar concept as rate-based trading. Different combinations of points of compliance and types of targets yield four institutional options, as seen in Table 2.11 Because our focus is on the efficiency consequences of ETS for the electricity sector, we simply assume, without loss of generality, that direct emissions by the NE sector are zero ðX ne ¼ 0Þ. We will analyze the effect of the three options on carbon price passthrough and the possible combinations thereof: Is the passthrough of the carbon price to the consumer price of electricity too low, appropriate or too high? Option A is a cap-and-trade policy on the direct emissions of the regulated sectors. The goal is to reach the emissions target X. 8 Refer to Burtraw (2008) for a detailed discussion on how to regulate CO2 in electricity markets, particularly on the point of compliance, emissions sources versus electricity consumers. 9 Refer to Fischer (2003) for the pros and cons of two emissions trading systems based on absolute and relative emissions targets. The author also evaluates the economic and environmental impacts of linking the two different types of emissions markets and investigates various measures to mitigate the adverse effects of such linkages. 10 Refer to Holland et al. (2009) for a detailed analysis on economic and environmental effects of LCFSs. 11 Option D is not considered here because carbon intensity regulation on indirect emissions is neither effective nor implementable in practice.
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
87
Table 2 Institutional options for ETS on emissions from the power sector. Compliance target
Point of compliance Electricity producer (direct emissions) Electricity consumer (indirect emissions)
Total emissions (cap-and-trade)
Carbon intensity (rate-based trading))
Option A Option C
Option B Option D (not considered)
Option 1 can guarantee that the goal X is precisely met.12 Xe ¼ X Under a cap-and-trade scheme, the market price of carbon, P C , passes through onto retail prices provided that no price regulation exists in the market. For the electricity market, the fact that the price shock falls on the power generators implies that the market price of carbon may not be fully reflected in the final consumer price, should there be price controls in the electricity market. In this case, demand for electricity may not properly adjust itself to achieve an efficient allocation of resources. Fig. 1, adapted from Lanz and Rausch (2013), provides efficiency implications of ETS systems in a partial equilibrium framework. In the absence of environmental regulation, initial equilibrium price and quantity of electricity are P0e and Q0e , respectively, determined at the intersection of the market demand curve (D0) and the average cost curve (AC0) of the monopolist power sector. Initial emissions intensity of E sector is the slope of I0e line.13 Resulting direct emissions of E-sector with a given initial emissions intensity I0e is X0e (¼Q0e I0e ). Let’s assume that (Q*e, X*e) yields the socially optimal allocation of resources. When an emissions cap X*e is enforced, i.e., X ¼ X ne , and emissions allowances are auctioned at price P*c, average cost of E sector is increased by P*e I*e to a new AC0 þ P*e I*e, resulting in a new equilibrium (P*e, Q*e). One thing to note is that, in the new equilibrium, the emissions intensity is changed from I0e to I*e according to the cost minimizing behavior of the E sector. If the emissions allowances are distributed for free, a windfall profit of the size P*c X*e would occur. Price regulation such as costof-service regulation would lower the final consumption price of electricity (P*e-Pe0 ) so as to make E sector’s profit to be 0. (Shaded areas in the 1st and 3rd quadrant are equal in size.) At this price, consumption level would increase to Qe0 , and emissions intensity needed to be adjusted to an even lower Ie0 to be able to comply with the emissions cap regulation. Price of carbon would increase to Pc0 under usual assumptions of increasing marginal abatement cost. Carbon price, however, would be incompletely passed through onto the final consumption price (Pe0 o P*e), resulting in an inefficient outcome compared to our benchmark case (Qe0 a Q*e). Option B is a carbon intensity (rate based) regulation on direct emissions, with trading allowed between power producers who over- or under-achieve the intensity targets. Under carbon intensity regulations, the electricity sector must meet its assigned intensities, I e . Therefore, Ie
Xe rI e Qe
12 The other options are also capable of meeting this target, on average, for a certain budget period but, in some cases, it is technically difficult to do this on a yearly basis. Whether to meet the goal on a yearly basis or for a certain budget period would be a non-issue. Considering the climate change and greenhouse gas concentration change mechanisms, it would be desirable to meet the target, on average, on longer budget periods. 13 Carbon intensity, defined as the ratio of ‘emissions due to production' to the ‘produced output', can be defined as, Ie X e =Q e .
Carbon intensity regulations on the electricity industry will have similar consequences to those of cap-and-trade schemes in the sense that the cost of carbon faced by producers is the same. Nonetheless, it is well documented that carbon intensity regulations have an effect similar to that of product subsidies.14 When the power supply increases by ΔQ e , free allowance also increases by ΔQ e I e , leading to a subsidy-like financial gain of ΔQ e I e P c . It follows that the marginal cost of power generation decreases by I e P c , resulting in lower retail prices and higher consumption of electricity compared to those under a cap-andtrade scheme. The unit carbon price that is passed through onto the final retail price would then be P c ðI e =I e Þ P c . We can see from this that carbon price pass-through would be 0 if the E sector’s carbon intensity improves to meet the emissions intensity regulation, i.e., I e ¼ I e . Fig. 2 shows this change. Let’s assume that the government, with the full information on the cost structure of E sector, set the target emissions intensity to be I*e aiming to reach the emissions target X*e. Initial upward movement of the average cost curve from AC0 to AC0 þP*cI*e will move the equilibrium point from (P0e , Q0e ) to (P*e, Q*e). The output subsidy effect of intensity regulation, however, moves average cost curve back to its original position. Equilibrium point will then move back to its original position (P0e , Q0e ). Clearly at this point emissions target X*e implicitly set by the government cannot be achieved unless the target emissions intensity is set at a much lower level IBe which is inferior in terms of cost efficiency compared to the benchmark emissions coefficient I*e. Since the equilibrium price does not reflect the carbon price, windfall profits do not occur here and the government intervention on electricity price such as cost-of-service regulation does not affect the equilibrium under the intensity regulation. Under an absolute target on indirect emissions by NE sector (Option C), an electricity consumers (NE sector) are responsible for their indirect emissions, calculated by the amount of electricity consumed (Qe) multiplied by a pre-determined emissions factor, IIe. Under Option C, carbon costs face by the NE sector would be P c Q e I Ie and the effective carbon price for the direct emissions induced by a unit of electricity consumed is ðI Ie =I e Þ P c . If the applied emissions factor is indeed equal to the actual carbon intensity, the market price of carbon will be fully reflected to the final consumer price of electricity, which, in turn, will adjust the demand for electricity to the most efficient level. However, the electricity consumers, NE sector, can only adjust their electricity consumption to meet the target. The target can neither affect the energy mix decisions of the electricity industry nor select power taps based on the carbon intensity of the supplied power. Therefore, option C fails to provide appropriate incentives for power producers to improve their carbon intensities. Let the emissions cap on the indirect emissions be X ¼ X n . In Fig. 2, the electricity price hike consumers face will shift the demand curve downward by PCc I0e , resulting in a new demand curve D0 PCc I0e . The corresponding level of consumption is QCe
14 Refer to Burtraw et al. (2001) for an in-depth assessment of the output subsidy effect of rate-based allocations.
88
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
Fig. 1. A cap-and-trade scheme with free emissions allowances.
Fig. 2. Intensity target on direct emissions and absolute target on indirect emissions.
without any change in emissions intensity. Clearly, E sector would have no incentive to improve their emissions intensity since they do not have any emissions reduction requirement. Here, the carbon price pass-through is nearly perfect as the point of compliance is at the final consumer. However, this new equilibrium (PCe , QCe ) is sacrificing too much consumption and investing too little on the emissions reduction technology, and thus is inferior to the benchmark case equilibrium (P*e, Q*e). So far we have found the following. (1) Absolute targets on direct emissions with cost-of-service regulation may not yield
efficient outcome due to an inappropriate pass-through of carbon price. (2) Carbon intensity regulations on direct emissions are relatively free of the carbon price pass-through concern but their ‘production subsidy effect’ tends to attenuate incentives to reduce electricity consumption. (3) Emissions caps on indirect emissions work through a demand control mechanism steered by the adding of the cost of carbon to the final consumer price. However, it provides few incentives for power producers to improve their fuel mixes or production efficiency. In short, these 3 options are either incomplete in and of their selves or ineffective in the presence of
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
market intervention by the government. A stronger policy mix is needed. A policy mix that combines options A and C would involve the application of emissions caps on both direct and indirect emissions. A carbon price Pc for the emissions cap on direct emissions will pass through to the final consumption price of electricity, at least in part even under the presence of cost-of-service regulation. On top of that, the carbon cost of ðI Ie =I e Þ P c due to the indirect emissions cap will be added to the final consumption price of electricity, reaching up to P c þ ðI Ie =I e Þ P c per carbon unit. This is nearly double the market price of carbon, raising concern of excessive regulation. Another policy mix that combines options B and C involves the application of carbon intensity regulations on direct emissions and an emissions cap on indirect emissions. Carbon intensity regulations tend to under-shoot the carbon price, P c ðI e =I e Þ P c , due to the subsidy effect, as discussed for option B. Absolute caps on indirect emissions embedded in the electricity consumption of the NE sector, however, raise the carbon price by ðI Ie =I e Þ P c . These two effects can mimic the correct carbon price P if the applied emissions factor ðI Ie Þ and carbon intensity target ðI e Þ are set equal to each other. Therefore, this policy mix conveys the correct price signal to the final consumer while providing power-generating companies with incentives to improve their carbon efficiency. In Fig. 2, the target emissions intensity set at I*e with emissions cap X*e * on the indirect emissions attains a new equilibrium (PBC e , Qe) with * the final cost of electricity consumption Pe. This equilibrium is practically identical to the efficient outcome (benchmark case equilibrium). Note that since the equilibrium price of electricity PBC does not reflect the carbon price Pc, windfall profits do not e occur. Therefore the cost-of-service regulation does not affect the equilibrium price under the policy mix B þC. Table 3 summarizes the carbon price pass-through to the price of electricity under the 5 policy options discussed above. We have shown that the pass-through of the price of carbon functions the best under the policy mix that combines policy options B and C. Under this policy mix, a single carbon price prevailing across the economy is applied to both electricity consumption and fossil fuel combustion, regardless of whether there are interventions in the electricity market. Table 3 also shows that the sum of effective emissions targets for the E and NE sectors under the policy mix B þ C becomes identical to the policy objective if the emissions factor ðI Ie Þ is set equal to the intensity target ðI e Þ. That is, the certainty of absolute emissions control, one of important advantages of cap-and-trade systems for direct emissions (option A), can also be guaranteed under the policy mix that combines relative targets for direct emissions (option B) and absolute targets for indirect emissions (option C).
4. Concluding remarks There are three key elements in the design of emissions trading systems: point of compliance, mode of compliance target, and initial distribution of permits. In this paper, we mainly reviewed the carbon price pass-through aspect of the first two elements of ETS systems in the context of the electricity industry. The electricity industry is peculiar in the sense that it suffers from heavy government regulation and market concentration in most cases. We determined that an ETS scheme that combines relative (intensity) targets on direct emissions with absolute emissions caps on indirect emissions performs better than cap-and-trade on direct emissions and any combination thereof considered in this study, in the presence of market imperfection and government price regulations under which the normal carbon price pass-through
89
Table 3 Carbon price pass-through and allowance market under various policy options. Policy option
Carbon price Production (direct emissions)
Allowance market Consumption (indirect Demand emissions) (covered emissions)
Supply (allowance allocation)
E NE sector sector
E sector
A
Pc
Pc
Xe
X
B
Pc
Xe
Ie Qe
C
–
Ie Pc Pc Ie IIe Pc Ie Ie Pc þ Pc Ie Ie Ie Pc Pc þ Pc Ie Ie
AþC
Pc
B þC
Pc
NE sector
–
I Ie Q e –
X
Xe
I Ie
X
Xe
I Ie Q e I e Q e X
Qe X
mechanisms of the competitive market could not function. This ETS scheme provides correct price incentives for consumers to adjust their electricity consumption in line with an efficient allocation of other types of energy. Intensity regulation provides power producers with incentives to reduce their carbon emissions, whereas the output subsidy effect of rate-based updating allocation helps prevent the double burden of carbon costs on electricity consumers. Moreover, this policy mix can be used to expand ETS coverage to include the commercial and public buildings and residential sector, which consume a significant share of energy but usually remain outside ETS schemes due to their heavy reliance on electricity and low levels of direct emissions. The product subsidy effect of intensity regulations reduce the gap between the carbon costs faced by regulated and unregulated sectors, while cap-andtrade on indirect emissions tends to expand the scope of the sectors covered by the policy. These two effects combined can attenuate the leakage caused by regulatory gaps, leading to the second-best solution. Holland (2012) proved that, in the presence of unregulated sectors, a policy mix that combines intensity standards with consumption taxes dominates a policy mix that combines cap-and-trade regulations and consumption taxes. It is interesting to note that the policy mix that combines policy options B and C mimics the former whereas the policy mix that combines policy options A and C mimics the latter. Therefore, we can conclude that a policy mix that combines intensity regulations on direct emissions and emissions caps on indirect emissions can work efficiently even in the presence of incomplete regulation (leakage). Another potential benefit of the intensity regulation of the power sector is that it helps prevent the unintended pass-through of carbon prices to electricity prices in non-ETS sectors, such as the commercial or residential sectors. Under usual cap-and-trade regulations on direct emissions, non-ETS energy consumers face higher electricity prices due to carbon price pass-through from power producers in the ETS sector, while they do not pay for the carbon content of other energy sources. This may cause distortions in the energy market and consumption structure if the government does not implement some countermeasures to mitigate discrepancies in the relative prices of electricity and other types of energy through, for example, carbon taxes on non-electric energy in non-ETS sectors. However, under intensity regulations of the direct emissions of electricity producers, the output subsidy effect impedes the carbon price from passing through onto the final price, thus preventing distortions in energy prices among
90
Y.-G. Kim, J.-S. Lim / Energy Policy 75 (2014) 84–90
different energy products that may or may not be covered by the ETS. The proposed policy mix is a relatively new concept that may suffer from criticisms of double regulation and higher administrative costs. A mere fact that two regulatory instrument are imposed on a single activity (carbon emissions) raises a concern from industries. They tend to imagine that excessive carbon cost like that of policy mix of option A þC would be imposed on them. Furthermore, administrative cost of suggested policy mix would be inevitably higher than a single policy option cases. However, considering its greenhouse gas reduction potential and its effect on the national economy, the additional complexity and administrative costs of the proposed ETS scheme on the electricity sector might be justified. Furthermore, our analysis demonstrates that the key characteristic of the suggested policy mix is a ‘correct price signal’ to both producers and consumers, showing that the worries of excessive carbon cost burden is a baseless concern. The justification for the suggested policy mix strongly holds especially when any single, simple regulation causes a serious inefficiency in complex situations with multiple policy constraints.
Acknowledgements This research is partially supported by the research project of Korea Environment Institute (“A study on consumption behavior and low carbon society 2013”), Ministry of Environment (“Allowance allocation for power industry”) and 2013 Kwangwoon University research grant.
References Baron, Richard., Aasrud, Andre., 2012. An Emissions Trading System for China’s Power Sector, Electricity in a Climate-constrained World. International Energy Agency. Boom, Jan-Tjeered., Dijkstra, B.R., 2009. Permit trading and credit trading: a comparison of cap-based and rate-based emission trading under perfect and imperfect competition. Environ. Res. Econ. 44, 107–136. Butraw, Dallas., Palmer, K., Bharvirkar, R., Paul, A. 2001. The Effect of Allowance Allocation on the Cost of Carbon Emission Trading, Resources for the Future Discussion Paper 01-03, Resources for the Future, Washington, DC. Burtraw, Dallas., 2008. Regulating CO2 in electricity markets: sources or consumers? Clim. Policy 8 (6), 588–606. Fisher, Carolyn., 2003. Combining rate-based and cap-and-trade emissions policies. Clim. Policy 3 (2), S89–S103. Scheme Options for Japanese Emissions Trading Scheme Based on Cap-and-trade System (Interim Report), Central Environmental Council. Domestic Emissions Trading Subcommittee. Kim, Wook., Chattopadhyay, Deb., Park, Jong-bae., 2010. Impact of carbon cost on wholesale electricity price: a note on price pass-through issues. Energy 35 (8), 3441–3448. Holland, Stephen P., 2012. Emissions taxes versus intensity standards: second-best environmental policies with incomplete regulation. J. Environ. Econ. Manage. 63 (3), 375–387. Holland, Stephen, P., Hughes, Jonathan E., Knittel, Christopher R., 2009. Greenhouse gas reductions under low carbon fuel standards. Am. Econ. J.: Econ. Policy vol. 1 (1), 106–146. Lanz, B., Rausch, S., 2013, Cap-and-Trade Climate Policy, Free Allowances, and Price Regulated Firms, Working Paper 12/178, Economic Working Paper Series, ETH. Nelson, Tim., Kelley, Simon, Orton, Fiona., 2012. A literature review of economic studies on carbon pricing and Australian wholesale electricity markets. Energy Policy 49, 217–224. Duan, Maosheng, Pang, Tao, Zhang, Xiliang, 2014. Review of carbon emissions trading polits in China. Energy Environ. 25 (3), 527–549. Sijm, Jos., Chen, Yihsu, Hobbs, Benjamin F., 2012. The impact of power market structure on CO2 cost pass-through to electricity prices under quantity competition—a theoretical approach. Energy Econ. 34 (4), 1143–1152. Stavins, Robert N., 1995. Transaction costs and tradable permits. J. Environ. Econ. Manage. 29, 133–148.