Energy Policy 55 (2013) 16–26
Contents lists available at SciVerse ScienceDirect
Energy Policy journal homepage: www.elsevier.com/locate/enpol
Transition towards a low carbon economy: A computable general equilibrium analysis for Poland a,n ¨ Christoph Bohringer , Thomas F. Rutherford b a b
Department of Economics and Center for Transnational Studies, University of Oldenburg, Germany Department of Economics, University of Wisconsin, Madison, USA
H I G H L I G H T S c c c
Economic impact assessment of the EU climate and energy package for Poland. Sensitivity analysis on where-flexibility, revenue recycling and technology choice. Application of a hybrid bottom-up, top-down CGE model.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 November 2011 Accepted 26 November 2012 Available online 22 January 2013
In the transition to sustainable economic structures the European Union assumes a leading role with its climate and energy package which sets ambitious greenhouse gas emission reduction targets by 2020. Among EU Member States, Poland with its heavy energy system reliance on coal is particularly worried on the pending trade-offs between emission regulation and economic growth. In our computable general equilibrium analysis of the EU climate and energy package we show that economic adjustment cost for Poland hinge crucially on restrictions to where-flexibility of emission abatement, revenue recycling, and technological options in the power system. We conclude that more comprehensive flexibility provisions at the EU level and a diligent policy implementation at the national level could achieve the transition towards a low carbon economy at little cost thereby broadening societal support. & 2012 Elsevier Ltd. All rights reserved.
Keywords: Climate policy Impact assessment Computable general equilibrium analysis
1. Introduction Between 1988 and 2005 Poland’s transition to a market economy has been accompanied by a sharp decrease in CO2 emissions along with structural changes towards less energy-intensive production as well as overall energy efficiency improvements. However, a positive correlation between GDP and CO2 emissions has reemerged from 2005 onwards confronting Poland with a potential trade-off between CO2emission reduction and economic growth. While compliance to its reduction target under the Kyoto Protocol at the end of 2012 is ensured, the challenge comes along with Poland’s new obligations under the ambitious EU climate and energy package which imposes an EU-wide emission decrease by 20% in 2020 compared to 2005 emission levels. Poland is among the Top-6 emitters within the European Union accounting for roughly 8% of EU-wide emissions over the last years. The per capita emissions are similar to the EU average, but given its low income level the Polish economy comes out as
n
Corresponding author. Tel.: þ49 441 7984102. ¨ E-mail address:
[email protected]. (C.Bohringer).
0301-4215/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.enpol.2012.11.056
among the most emission-intensive. A distinctive feature of Poland’s composition of CO2 emissions is the dominance of the power sector with an extraordinary dependence on coal. Around 85% of Poland’s CO2 emissions stem from the energy sector, in particular electricity and heat production. More than 90% of electricity is generated by lignite-fired power plants which emit the highest levels of CO2 per unit of electricity across alternative fossil-fuel based power generation technologies—between two to three times as much as gas-fired plants. The heavy reliance of Polish industry and power stations on coal explains concerns in Poland that stringent CO2 emission constraints as put forward by the EU energy and climate package will not only boost domestic electricity prices but also negatively affect competitiveness and overall economic performance. How costly will it be for Poland to move to a lower carbon path? Will Poland be more burdened than the rest of the EU? How will alternative strategies to achieve the EU’s emission reduction targets up to 2020 affect the magnitude and distribution of economic adjustment cost? To gain insights in these questions we make use of a computable general equilibrium (CGE) model that incorporates key determinants of economic impacts triggered by emission regulation. In our numerical simulations we find that compliance to the energy and climate
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
package induces sizeable economic cost for Poland (up to a loss in real income of roughly 1% compared to a business-as-usual situation without emission regulation) that are markedly higher than for the rest of the EU. The adjustment cost for the transition to a lower carbon economy, however, could be reduced substantially through amendments of emission regulation at the superordinate EU level as well as the cost-effective policy implementation at the Member State level. At the superordinate EU level, comprehensive EU-wide emissions trading and in particular the relaxation of supplementarity constraints for the use of the Clean Development Mechanism (CDM) would allow for substantial cost savings. At the Member State level, revenue recycling of regulatory rents through wage subsidies (instead of lump-sum rebates or free allowance allocation to emission-intensive and trade-exposed industries) provides scope for a double dividend, i.e., a reduction of emissions together with reduced unemployment (and a reduction of Poland’s compliance cost by two third). Relaxing expansion constraints on nuclear power is found to cut compliance cost for Poland by about one third. This paper adds a country study for Poland to the applied economic literature on impact assessment of the EU climate and energy package. The specific methodological contribution of our CGE analysis is the focus on economic adjustment of a single EU country—in this case Poland—while accounting for important international spillovers of policy regulation through a multiregion (global) setting.1 Furthermore, our economic impact assessment for Poland exemplifies the critical importance of where-flexibility in emission abatement, revenue recycling, and technology constraints in the electricity sector for the magnitude of economic adjustment cost towards a low carbon economy. The remainder of this article is organized as follows. Section 2 presents the computable general equilibrium model underlying our quantitative analysis of emission regulation in Poland and the EU. Section 3 lays out alternative policy scenarios to meet the emission reduction commitments under the EU climate and energy package. Section 4 presents a discussion of the simulation results. Section 5 summarizes and concludes.
2. Computable general equilibrium model Our quantitative impact assessment of the EU climate and energy package builds on a static multi-sector, multi-region CGE ¨ framework established by Bohringer and Rutherford (2010) for the analysis of greenhouse gas emission control strategies. We extend the generic CGE model with specific features such as labor market rigidities, the bottom-up representation of discrete technologies in electricity production, and alternative revenue recycling mechanisms to reflect central issues in the climate policy debate in Poland and the rest of the EU. 2.1. Model structure For the general reader we restrict the model description to a non-technical summary of key features. Appendix which follows ¨ Bohringer and Rutherford (2010) provides a detailed algebraic description. Our model includes a representative agent for each region who is endowed with three primary factors: labor, capital, and fossilfuel resources (used for the production of fossil fuels). Labor and capital are intersectorally mobile within regions but immobile 1 ¨ Bohringer et al. (2009) compare various impact studies of the EU climate and energy package but none of these studies treat Poland explicitly. Bukowski and Kowal (2010) provide a climate policy impact assessment for Poland based on a dynamic stochastic general equilibrium (DSGE) model but do not account for international feedback and spillover effects as they adopt a single-region framework.
17
between regions. Fossil-fuel resources are specific to fossil fuel production sectors in each region. Production of commodities other than primary fossil fuels and electricity is captured by three-level nested constantelasticity-of-substitution (CES) cost functions that describe the price-dependent use of capital, labor, energy and material in production. At the top level, a CES material composite trades off with an aggregate of capital, labor and energy subject to a constant elasticity of substitution. At the second level, a CES function describes the substitution possibilities between the energy aggregate and the value-added composite of capital and labor. At the third level, capital and labor substitution possibilities within the value-added composite are captured by a CES function and different energy inputs enter the energy composite subject to a constant elasticity of substitution. In the production of fossil fuels, all inputs—except for the sector-specific fossil fuel resource—are aggregated in fixed proportions at the lower nest. At the top level, this non-resource composite trades off with the sector-specific fossil fuel resource at a constant elasticity of substitution. The latter is calibrated in consistency with empirical estimates for the price elasticity of fossil fuel supply. Final consumption demand in each region is determined by the representative agent who maximizes utility subject to a budget constraint with fixed investment (i.e., a given demand for the savings good). Consumption is captured through a CES composite that combines demand for energy and non-energy goods. Substitution patterns across non-energy goods in final consumption are reflected via a CES function; the energy aggregate in final consumption demand consists of the various energy goods trading off at a constant elasticity of substitution. Government provides a public good which is produced with commodities purchased at market prices. These expenditures are financed with tax revenues. The impact assessment of policy interference implicitly involves revenue-neutral tax reforms in order to provide a meaningful welfare comparison without the need to trade off private consumption and government consumption. This is done by keeping the amount of the public good provision fixed and recycling any residual revenue. Bilateral trade is specified following the Armington approach of product heterogeneity where domestic and foreign goods are distinguished by origin (Armington, 1969). All goods used on the domestic market in intermediate and final demand correspond to a CES composite that combines the domestically produced good and the imported good from other regions differentiated by demand category. Domestic production either enters the formation of the Armington good or is exported to satisfy the import demand of other regions. The balance of payment constraint which is warranted through flexible exchange rates incorporates the benchmark trade deficit or surplus for each region. CO2 emissions are linked in fixed proportions to the use of fossil fuels with CO2 coefficients differentiated by the specific carbon content of fuels. CO2 emission abatement can take place via fuel switching (inter-fuel substitution) or energy savings (either by fuel-non-fuel substitution or a scale reduction of production and final demand activities). CO2 abatement requirements are introduced by means of an additional constraint that holds CO2 emissions to a specified limit. Scarcity rents on CO2 emission constraints accrue to the government. Domestic labor markets may exhibit frictions with equilibrium unemployment. To mimic labor market rigidities we adopt a wage-curve relationship. Labor market rigidities are represented at the regional level through the specification of a wage curve (Blanchflower and Oswald, 1995). The wage curve reflects empirical evidence on the inverse relationship between the level of wages and the rate of unemployment which can be derived in
18
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
analytical terms from wage bargaining as well as efficiency of wage mechanisms (Hutton and Ruocco, 1999). Given the paramount importance of the electricity sector with respect to CO2 emission abatement the standard top-down representation of power production through a CES production (cost) function is replaced by a bottom-up activity analysis characterization where several discrete generation technologies compete to supply electricity to regional markets. The price of electricity then is determined by the production cost of the marginal supplier. Power generation technologies respond to changes in electricity prices according to technology-specific supply elasticities (see Rutherford (2002) for details on the calibration technique). In addition, lower and upper bounds on production capacities can provide explicit limits to the decline and the expansions of technologies.
2.2. Data and parameterization The model builds on the GTAP dataset (version 7) with detailed accounts of regional production, regional consumption, bilateral trade flows as well as energy flows and CO2 emissions for the base year 2004 (Badri and Walmsley, 2008). As to the sectoral and regional model resolution the GTAP database is aggregated towards a composite dataset which accounts for the specific requirements of assessing the economic impacts of EU climate policy regulation for Poland. At the sectoral level, the aggregate dataset captures details on differences in factor intensities, degrees of factor substitutability and price elasticities of output demand to trace the structural change in production induced by emission regulation. The energy goods identified are coal, crude oil, natural gas, refined oil products and electricity. The dataset furthermore incorporates energy-intensive and trade-exposed non-energy commodities that are potentially most affected by emission control policies (paper, pulp and print; chemical products; non-metallic minerals; mineral products; non-ferrous metals; iron and steel; air transport; and other transport)—in our analysis these non-energy commodities together with refined oil products are referred to as EITE industries (see Table 1) which might receive preferential regulatory treatment through output-based allocation of emission allowances (see scenario OBA in Table 2). All remaining sectors are summarized through a composite of all other industries and services. At the regional level, the dataset explicitly includes Table 1 Model sectors and regions.
3. Policy scenarios Our primary objective is to quantify the economic impacts of the EU climate and energy package for Poland in relation to the rest of the EU. We furthermore want to gain insights how changes in policy implementation at the superordinate EU level as well as the national (Polish) level affect the magnitude of impacts. To this end, we define a reference scenario that reflects current regulatory provisions and then design alternative scenarios of policy implementation that achieve the same emission reduction target. In the following we lay out the key assumptions for each scenario. 3.1. Reference climate policy for Poland and the rest of the EU
Sectors and commodities
Countries and regions
Energy Coal Crude oil Natural gas Refined oil productsa Electricity
Regions with emission reduction pledges Poland Rest of the EU Remaining industrialized (non-EU) regions
Non-energy Chemical industrya Air transporta Other transporta Non-metallic mineralsa Iron and steel industrya Non-ferrous metalsa Paper–pulp–printa Other manufactures and services
Regions without emission pledges Developing countries
a
Poland as the main country of interest for our analysis and the rest of the European Union. Furthermore, we account for all other major industrialized regions as well as the developing world to capture the international market responses to regional emission constraints and CDM supply options for the developing world. Table 1 summarizes the sectors (commodities) and regions of the composite dataset underlying our impact analysis of alternative CO2 mitigation strategies for Poland. As is customary in applied general equilibrium analysis, base year data together with exogenous elasticities determine the free parameters of the functional forms. Elasticities in international trade and sectoral value-added are based on empirical estimates reported in the GTAP database. Substitution elasticities between production factors capital, labor, energy inputs and non-energy inputs (material) are taken from Okagawa and Ban (2008) who use panel data across 14 OECD regions (mostly EU Member States) and 19 sectors for the period between 1995 and 2004. The elasticities of substitution in fossil fuel sectors are calibrated to match exogenous estimates of fossil-fuel supply elasticities (Graham et al., 1999; Krichene, 2002). The cost of complying with future emission constraints—here the emission reduction targets prescribed by the EU climate and energy package in 2020—are directly linked to the structural characteristics of the economy exhibited in a hypothetical business-as-usual (BAU) situation without emission constraints. The BAU structure of our model regions in 2020 is based on projected energy input demands across sectors, future GDP levels and the international price trajectory for crude oil. The model forward projection employs data from the US Energy Information Agency (International Energy Outlook—EIA, 2008) which is complemented with more detailed information from the European Commission for Poland and the rest of the EU (European Commission, 2008b).
Energy-intensive and trade-exposed (EITE) industries.
Within its climate and energy package the EU has already fixed central provisions of emission abatement for EU Member States (European Commission, 2008a). There is a legally binding overall emission reduction target of 20% in 2020 (compared to 1990 emission levels) put into legal force upon mutual agreement between the European Council, the European Parliament, and the European Commission.2 The EU has not only set an EU-wide emission reduction target but also prescribes policy implementation to a larger extent. The aggregate EU emission reduction target is split down into a reduction target of 21% for energyintensive industries and a 10% emission reduction requirement for the remaining sectors of the economy—both targets are taking 2005 as the reference year (note that the cumulative reduction 2 Note that the EU emission reduction target for 2020 compromises all greenhouse gases among which CO2 is by far the most important. In our analysis we focus on CO2 emissions only and adopt the aggregate greenhouse gas reduction target as the CO2 emission reduction target.
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
19
Table 2 Summary of scenario characteristics. Scenario
Assumptions
REF
Default settings for international emission mitigation strategies CO2 emission reduction pledges for Poland, the rest of the EU and other industrialized regions (see Table 3) Segmented CO2emission markets for Poland and the rest of the EU with EU-wide emission trading for energy-intensive industries (EU ETS) and costefficient emission regulation for non-ETS sectors within each Member State; other industrialized regions with cost-efficient regulation of domestic CO2 emissions Limits to CDM offsets for Poland and the rest of the EU as prescribed by the EU climate and energy policy package; other industrialized regions without CDM access; no international emission trading between the EU (incl. Poland) and other industrialized regions Lump-sum recycling of revenues from CO2 emission regulation Bottom-up activity analysis characterization of power supply technologies in Poland and the Rest of the EU with restricted use of nuclear power at BAU capacity level
FLEX_EU FLEX_CDM WSUB OBA
Like REF but with comprehensive EU emission trading Like REF but with comprehensive EU emission trading and relaxed CDM limits Like REF but with revenue recycling via wage subsidies Like REF but with output-based allowance allocation to energy-intensive and trade-exposed (EITE) sectors in Poland and the rest of the EU (free allocation of 70% of EITE’s 2005 reference emission level) Like REF but without nuclear expansion ceiling in Poland Like REF but with ceiling of gas use in Poland’s power generation at BAU level
NUC NOGAS
Table 3 Nominal and effective emission reduction requirements. Source: Own calculations based on EIA (2008) and European Commission (2008b).
Poland (total) ETS Non-ETS EU-26 (total) ETS Non-ETS Other Annex 1
Nominal CO2 reduction pledges (% vis-a -vis 2005)
Effective CO2 reduction pledges (% vis-a -vis 2020)
9.5 21.0 14.0 16.6 21.0 12.5 4.8
23.0 23.7 22.0 18.8 24.3 13.5 16.5
adds up to the overall target of 20% with respect to 1990 emission levels). The emission ceiling for the energy-intensive sectors is implemented centrally through an EU-wide cap and trade system—the so-called EU emission trading system (EU ETS). The revenues from emission trading are shared out to Members State in proportion to their historical ETS emission shares in 2005. The reduction target of 10% for emissions from sectors outside the EU ETS (thereafter referred to as non-ETS sectors) is split across Member States reflecting differences in economic performance (measured in terms of GDP per capita): targets range from a 20% decrease of emissions for high income regions such as Ireland, Luxemburg or Denmark to a 20% increase of emissions for low income regions such as Bulgaria. Under this non-ETS ‘‘burden sharing scheme’’ Poland can increase its non-ETS emissions by 14% by 2020 compared to the reference non-ETS emission level in 2005. The policy regime to achieve the domestic non-ETS targets can be chosen by each Member State. For our simulation analysis we assume that Member States levy a sufficiently high domestic CO2 tax on emissions from non-ETS sectors thereby generating additional revenues for the public sector. While there is no tradability between the ETS and non-ETS segments of the EU economies, smaller fractions of emission reduction obligations in each segment can be achieved outside the EU. According to the EU legislation, the non-ETS sectors are allowed to purchase around a third of their emission reduction requirement via the Clean Development Mechanism (CDM), i.e., emission reductions in non-industrialized countries which are paid for by CDM donors in the EU; the non-ETS sectors can offset up to a fifth of their reduction requirement through CDM imports. As to revenues from auctioned ETS emission allowances and domestic pricing of non-ETS emissions, we assume by default
lump-sum recycling to the representative household in Poland and the rest of the EU while public good provision is kept constant. We adopt the climate policy prescriptions above as our reference scenario REF and subsequently define modifications to investigate how economic impacts for Poland (and the rest of the EU) change depending on alternative assumptions for EU-wide or national implementation rules. 3.2. Alternative climate policy designs for Poland and the rest of the EU Our alternative climate policy variants reflect hypothetical regulatory changes from the REF scenario settings with respect to (i) restrictions on where-flexibility of abatement (EU emission market segmentation and CDM ceilings), (ii) revenue recycling of CO2 rents, and (iii) technological policy constraints in power generation. Scenarios FLEX_EU and FLEX_CDM address the issue of whereflexibility in emission abatement. Scenario FLEX_EU postulates a comprehensive EU emission trading system which covers all sectors across EU Member States. Scenario FLEX_CDM in addition relaxes the supplementarity constraints on CDM—EU Member States can purchase CDM up to a 100% of their respective emission reduction requirement. Scenarios WSUB and OBA consider modifications in the recycling of revenues from emission regulation. Scenario WSUB investigates the scope of a double dividend from environmental regulation. Rents from CO2 regulation are no longer handed back lump-sum to the representative consumer but recycled through revenue-neutral wage subsidies with the objective to reduce unemployment and overall economic adjustment cost to emission constraints. Scenario OBA reflects concerns on the competitiveness of energy-intensive and trade-exposed (EITE) sectors as the EU goes ahead with ambitious emission reduction targets. To ameliorate adverse impacts for EITE industries the EU considers output-based emission allocation, i.e., emission allowances are partially given for free to EITE sectors conditional on their output. Effectively, output-based allocation works as an implicit subsidy to firms’ output. Scenario OBA postulates that 70% of EITE sectors’ emissions in 2005 are handed back for free. Scenarios NUC and NOGAS capture technology-specific policy constraints in power generation for Poland. Scenario NUC allows for the expansion of nuclear power in Poland beyond business-asusual (BAU) levels whereas scenario NOGAS limits the use of gas in Poland’s power system to the BAU level. Both technology scenarios for Poland can be interpreted as a policy move towards reduced energy import dependency.
20
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
Fig. 1. BAU technology shares in power generation for 2020. Source: European Commission (2008b)
Across all emission scenarios, climate policy outside Poland and the rest of the EU is captured through an exogenous reduction pledge of the composite industrialized non-EU region (including Canada, USA, Australia, New Zealand, Japan, and Russia). This pledge in 2020 amounts to 4.8% from 2005 emission levels equivalent to the average of official emission reduction targets across individual regions (DCCEE, 2011). We assume that the composite industrialized non-EU region achieves its aggregate emission reduction commitment up to 2020 through cost-efficient domestic action. There is no international trade in emission allowances between the EU (incl. Poland) and other industrialized regions. Furthermore, we assume that other industrialized regions do not make use of CDM offsets so there is no competition with these regions for CDM projects. In the model implementation, CDM supply from developing countries is represented as emission trading with the composite of developing countries DC where the latter adopts its BAU emission level as an upper emission limit. Table 2 summarizes the key settings for scenario REF and the specific differences of alternative emission abatement scenarios. 3.3. Business-as-usual scenario (BAU) Table 3 summarizes how regional emission reduction pledges stated with respect to 2005 emission levels translate into effective emission reduction requirements from the BAU level in 2020 where we adopt business-as-usual projections by the Energy Information Agency (EIA, 2008) and the European Commission (European Commission, 2008b). For Poland and the EU the reduction targets are reported for ETS and non-ETS segments based on the assumption that the aggregate EU-ETS reduction target is uniformly applied to the respective ETS emissions of Member States in 2005. Due to the projected growth of BAU emissions the effective reduction requirements become in general markedly higher. This is in particular the case for the emissions of non-ETS sectors in Poland where the target of 14% (i.e., an increase) with respect to 2005 emission levels turns into a stringent reduction target of 22% with respect to 2020 BAU emission levels. Fig. 1 summarizes the BAU structure of electricity generation in 2020 for Poland and the rest of the EU which are in both regions the major sources for CO2 emissions. Electricity production in Poland is projected to be heavily coal-based (84%) while gasfired power generation (5%) and electricity from renewables (5%) play an inferior role.3 The structure of power generation in the rest of the 3 It should be noted that nuclear power which is not operated in Poland so far amounts to 5% of power generation in the BAU projection for 2020.
EU is much more balanced across coal (20%), gas (30%), nuclear (26%) and renewables (21%). For Poland, power generation is projected to account for roughly half of total CO2 emissions; for the rest of the EU, this share amounts to a third. The heavy reliance of the Polish energy system on coal together with below-EU-average per-capita income implies that the BAU CO2 emission intensity (emissions per GDP) in Poland is more than twice the emission intensity in the (composite) rest of the EU.
4. Simulation results We use our multi-sector, multi-region CGE model to quantify the economic impacts of alternative emission abatement policies for Poland vis-a -vis the rest of the EU. Simulation results are reported as percentage changes in key economic variables from their business-as-usual (BAU) levels except for marginal abatement cost which are given in $US per ton of CO2. Our central indicator for economic adjustment cost at the regional level is the Hicksian equivalent variation (HEV) in income which denotes the amount which is necessary to add to (or deduct from) the benchmark income of the representative consumer so that she enjoys a utility level equal to the one in the counterfactual policy scenario on the basis of ex-ante relative prices. We monitor impacts on real wages and the unemployment rate in Poland and the rest of the EU as an important indicator for the incidence of emission regulation. Furthermore, we report changes in output for energy- and trade-exposed (EITE) industries which are most concerned about the adverse implications of emission regulation. Reflecting the major contribution of the power sector to economy-wide CO2 emissions we provide details on changes in electricity prices, electricity production and shifts in technology shares for Poland and the rest of the EU. Table 4 provides a condensed overview of simulation results. We start with the interpretation of the economic impacts induced by scenario REF which serves as the reference for climate policy implementation in Poland and the rest of the EU. Subsequently, we discuss how the economic impacts change if we consider alternative variants of climate policy design that achieve the identical emission reduction targets. 4.1. Reference climate policy (scenario REF) Implementing the reference settings of the EU climate energy package triggers a welfare loss for Poland of nearly 1% which is
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
21
Table 4 Impacts of alternative emission mitigation scenarios for Poland and the rest of the European Union (own calculations). BAU Welfare (% Hicksian equivalent variation in income) Poland Rest of the EU CO2 values ($US per ton of CO2) Poland ETS Poland non-ETS Rest of the EU ETS Rest of the EU non-ETS Unemployment (% change from BAU) Poland Rest of the EU Real wage (% change from BAU) Poland Rest of the EU
REF
Electricity output (% change from BAU) Poland Rest of the EU Technology shares (in % of total electricity production) Poland Coal 84.1 Gas 5.1 Oil 0.9 Nuclear 5.2 Renewables 4.5 EU
Coal Gas Oil Nuclear Renewables
20.1 29.7 3.6 25.5 21.2
FLEX_CDM
WSUB
OBA
NUC
NOGAS
0.94 0.29
0.86 0.26
0.18 0.03
0.31 0.05
0.96 0.28
0.59 0.28
1.00 0.30
29.7 87.2 29.7 81.9
36.4 36.4 36.4 36.4
7.9 7.9 7.9 7.9
30.1 91.3 30.1 84.0
29.8 87.3 29.8 82.1
26.9 88.3 26.9 81.8
30.1 87.4 30.1 81.9
5.30 1.70
4.14 1.24
1.04 0.30
3.87 0.70
5.18 1.66
3.70 1.64
5.54 1.71
4.05 1.34
3.19 0.98
0.82 0.24
3.21 0.57
3.96 1.31
2.87 1.29
4.22 1.35
2.8 0.8
0.3 0.2
2.0 0.5
1.9 0.7
2.8 0.7
Output of energy-intensive and trade-exposed (EITE) sectors (% from BAU) Poland 2.7 Rest of the EU 0.7 Electricity price (% change from BAU) Poland Rest of the EU
FLEX_EU
2.1 0.6
20.09 9.71
26.21 12.60
5.47 2.58
20.07 9.81
20.27 9.78
9.19 8.68
22.37 9.87
10.82 3.14
13.01 4.26
3.19 0.49
10.24 2.96
10.76 3.12
4.35 2.87
11.87 3.19
73.5 11.8 1.1 5.9 7.8
70.2 14.0 1.1 6.0 8.7
81.9 6.4 1.0 5.4 5.3
75.6 6.2 1.0 5.5 11.7
73.4 11.9 1.1 5.9 7.8
50.7 6.9 1.0 35.5 5.9
66.3 10.0 1.0 15.5 7.2
12.1 31.2 3.7 26.3 26.7
10.6 30.9 3.7 26.6 28.2
18.2 30.2 3.6 25.6 22.4
11.6 24.8 3.4 20.2 40.0
12.1 31.3 3.7 26.3 26.7
12.8 31.2 3.7 26.2 26.1
12.3 31.2 3.7 26.3 26.5
more than three times the adjustment cost of the rest of the EU. Ceteris paribus differences in economic adjustment cost can be traced back to differences in effective emission reduction targets. Poland faces a higher reduction requirement from BAU 2020 emission levels than the rest of the EU—mainly because of the strong business-as-usual emission growth in its non-ETS sectors (see Table 2). Another important cost driver is the ease of carbon substitution in production and final consumption which is implicit to the production technologies and consumer preferences through BAU cost shares and cross-price elasticities of substitution. International spillover effects, i.e., changes in the terms of trade, constitute a further macroeconomic cost determinant which may amplify or attenuate the direct domestic cost of emission abatement; for example, countries which are net importers of fossil fuels will benefit from depressed world market prices as international climate policy action drives down fossil fuel demand (in turn, exporters of fossil fuels will face additional income losses). The price for ETS emission amounts to roughly 30 $US per ton of CO2—much lower than the CO2 prices in the non-ETS sectors for both Poland and the rest of the EU indicating substantial scope for efficiency gains from comprehensive emission trading across all sectors and a relaxation of stringent CDM quotas. Emission reductions, i.e. reduced fossil fuel use, exert a downward pressure on the real wage which leads to an increase in unemployment rates. The latter is more pronounced in the case of Poland than for the rest of the EU.
The adverse implications of CO2 emission pricing on EITE industries in Poland and the rest of the EU are rather modest—one reason is that EITE industries face an ETS CO2 price which is just one third of the emission prices to non-ETS sectors; another reason is that competing industries in other industrialized countries are subject to comparable emission constraints. Electricity prices in Poland increase by roughly 20% which is more than double the price increase for the rest of the EU. The reason behind is that power generation in Poland is predominantly based on coal which stands out for the highest specific CO2 content among fossil fuels. CO2 reduction from the power sector takes place through output reduction (in the case of Poland a decline of roughly 10% while only 3% for the rest of the EU), the expansion of renewable power production and through fuel shifting to natural gas.4 4.2. Implications of increased where-flexibility (scenarios FLEX_EU and FLEX_CDM) Within scenario FLEX_EU, Poland and the rest of the EU establish a comprehensive emission trading market with a single EU-wide CO2 emission price of around 36 $US per ton. Equalization of marginal abatement cost across all EU emission sources 4 Note that scenario REF assumes a ceiling of nuclear power in Poland and the rest of the EU at BAU levels.
22
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
reduces compliance cost for Poland and the rest of the EU by roughly 10% compared to the REF cost level.5 The comprehensive EU trading regime puts a higher adjustment pressure on the electricity sector and other ETS sectors as the new CO2 emission price in the combined ETS and non-ETS markets is higher than that for the ETS segment in scenario REF. ETS sectors therefore undertake more emission abatement in scenario FLEX_EU whereas the abatement burden for the non-ETS sectors are lowered. As a consequence, electricity prices increase more, power production decline is stronger and the structural shifts in power generation across technologies become more accentuated. Likewise, the negative implications for EITE output sectors are slightly more pronounced. Scenario FLEX_CDM maintains the assumption of a comprehensive EU emissions trading regime and in addition relaxes the CDM supplementarity constraints. Poland and the rest of the EU are now allowed to import their whole nominal emission reduction requirement stated with respect to 2005 levels (note that this still leaves a substantial gap to the higher effective reduction requirement in 2020). The less stringent supplementarity constraint allows for a drastic decrease in overall economic cost for Poland and the rest of the EU as the CO2 emission price drops to around 8 $US per ton. Poland and the rest of the EU can shift a larger part of abatement to the developing world paying only around 2 $US per ton of CO2 CDM credit: whereas the aggregate EU emission reduction in scenarios REF and FLEX_EU compared to BAU levels amounts to 15% it is only around 3% in scenario FLEX_CDM. The difference between the cost of 2 $US per CDM credit and the EU-internal CO2 price of 8 $US is captured by the shadow price on the CDM quota with quota revenues accruing to the governments in Poland and the rest of the EU (in scenario REF with stricter supplementarity rules the quota rent is even higher and differentiated between ETS and non-ETS segments). Given the low CO2 emission price of 8 $US in scenario FLEX_CDM, adjustment in EITE production and the power sector from the BAU is rather negligible. The assumptions behind FLEX_CDM might be disputed: Is it realistic that the EU allows for very lax supplementarity constraints whereas other industrialized regions abstain from CDM imports at all? To the extent that the other industrialized regions would compete for CDM, the CDM price will go up and the relative attractiveness of CDM offsets to the EU will decline. Another legitimate question is whether developing regions are going to accept that industrialized countries depress international CDM prices through CDM import quotas—would not they rather counteract with rigid export quotas to keep the CDM price up and skim quota rents themselves? The cost savings through CDM might also be overstated in our actual model representation since transaction cost and investments risks are not accounted for (Anger et al., 2007). Finally, opponents to the unrestricted use of CDM highlight the risk of windfall profits without effective emissions reduction since the definition of a hypothetical business-as-usual for CDM exporting regions in the developing world is challenging.
4.3. Implications of alternative revenue recycling(scenarios WSUB and OBA)
deliberate revenue recycling (Goulder, 1995). Scenario WSUB supports this weak double dividend hypothesis as revenues from emission regulation are not returned lump-sum but used to reduce labor cost. In this case, the downward pressure of emission pricing on wages can be more than offset through wage subsidies such that real wages increase and unemployment falls. Revenue recycling through labor cost cuts hardly affects the marginal cost of abatement in ETS or non-ETS sectors but generates substantial economy-wide cost savings via the implicit relaxation of labor market rigidities. Stringent emission constraints raise concerns on the competitiveness of energy-intensive and trade-exposed (EITE) industries in the EU compared to rivals abroad which may benefit from more lenient domestic climate policies. As a response, the EU climate and energy package foresees the possibility of output-based emission allowance allocation to EITE sectors. An important feature of output-based rebating is that the allocations are updated (unlike with ordinary grandfathering) based on recent measures of economic activity, namely production. As a consequence, output-based rebating functions as a production subsidy. Scenario OBA quantifies the economic implications of output-based allocation to EITE sectors in the EU ETS where 70% of the 2005 carbon emissions in EITE sectors are handed back as an implicit production subsidy. Output-based allocation slightly reduces output losses of EITE sectors. As EITE sectors only account for a small share in overall emissions and production activity the welfare losses of distortionary production subsides under OBA are very small. 4.4. Implications of power technology regulations(scenarios NUC and NOGAS) Coal-based power production accounts for more than 50% of Poland’s total CO2 emissions and therefore is at the core of any emission reduction strategy. Apart from the costs and potentials of renewable energies in power production, the ease of substituting away from coal in Poland’s electricity system hinges to a large extent on policy constraints for nuclear expansion and the willingness to increase dependency on foreign gas imports. In the REF scenario, nuclear power is constrained at the BAU capacity level whereas gas generation responds elastically to changes in electricity prices and input cost. Scenario NUC allows for nuclear expansion in Poland which cuts overall compliance cost for Poland by a third. The increase of electricity prices and in turn the decline of electricity output is roughly halved as the share in nuclear power generation goes up from around 6% in scenario REF to more than 35% in scenario NUC. As a consequence of the massive nuclear power expansion, the shares of coal, gas and renewables decline markedly compared to the REF scenario. Scenario NOGAS keeps the nuclear ceiling as in REF but restricts in addition the use of gas-fired power plants in Poland at the BAU level thereby mimicking energy security concerns of Poland about gas import dependency from abroad. Our simulation results suggest that limitation of gas inputs to power generation in Poland at the BAU level causes only modest additional cost.
5. Conclusions In the reference scenario the scarcity rents from emission caps in the ETS and the non-ETS segments are recycled lump-sum to the representative consumer in each region. However, initial tax distortions and labor market imperfections open up for attenuating the negative impacts of emission regulation through more 5 The excess cost of CO2 market segmentation under REF must be considered rather as a lower bound estimate in our simulation analysis since we assume costeffective emission reduction across all non-ETS segments in the rest of the EU.
In the battle against climate change the EU has taken a proactive stance with stringent emission reduction targets for 2020. Among EU Member States, Poland with its heavy energy system reliance on coal is particularly concerned on the pending trade-offs between CO2 emission constraints and economic growth. Our impact assessment of climate policy for Poland and the rest of the EU supports these concerns as Poland suffers from a sizeable loss in real income (around 1% compared to a business-as-usual situation
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
without emission regulation) which is two times higher than for the rest of the EU. However, our simulation results also indicate that increased where-flexibility of emission abatement could substantially reduce the cost of decarbonization. Deliberate revenue recycling provides another option for larger alleviations of the economic burden triggered by carbon emission constraint: if CO2 emission rents are recycled through cuts in labor cost Poland and the rest of the EU may even reap a double dividend, i.e., a reduction of emissions together with a reduction of unemployment. Climate policies that renounce on ‘‘smart’’ recycling as they stick to lump-sum transfers or free allowance allocation can turn out quite costly. Beyond whereflexibility and revenue recycling technological options in electricity generation which accounts for most of the CO2 emissions particularly in Poland but also in the rest of the EU are of critical importance for the adjustment cost to emission constraints. Restrictions to nuclear power or gas imports can induce significant additional cost of emission reduction which must be traded off against other policy objectives or societal preferences such as technology-specific risk aversion or energy security. We conclude that more comprehensive flexibility provisions at the superordinate EU level and a diligent policy implementation at the national level could achieve Poland’s transition towards a low carbon economy at relatively modest economic cost thereby broadening societal support for ambitious EU climate policy targets.
23
Notations for variables and parameters are listed in Tables A1–A6. Figs. A1–A4 provide a graphical exposition of the cost (production) and expenditure (utility) functions. (1) Zero-profit conditions Production of goods except fossil fuels (geFF): At the lower level, a value-added composite trades off with an energy composite subject to a constant elasticity of substitution (CES). At the upper level, a CES function describes the substitution possibilities between a material composite
Table A1 Indices (sets). g
Sectors and commodities (g ¼i), final consumption composite (g ¼ C), investment composite (g ¼ I), public good composite (g ¼G) i Sectors and commodities t Electricity generation technologies r (alias s) Regions EG Energy goods: coal, crude oil, refined oil, gas and electricity FF Fossil fuels: coal, crude oil and gas
Table A2 Activity variables.
Acknowledgments We are indebted to two anonymous referees for helpful comments. The authors would like to thank Olga Kiuila, Erika Jorgenson and Leszek Kasek for research cooperation within the World Bank Project ‘‘Transition to a Low Carbon Economy in Poland’’. Support from the German Research Foundation (BO 1713/5-1) is gratefully acknowledged. The ideas expressed here remain those of the authors who remain solely responsible for errors and omissions.
Ygr Ytr Mgr Egr KLgr Aigr IMir ur
Production of item g in region r Electricity generation by technology t in region r Material composite for item g in region r Energy composite for item g in region r Value-added composite for item g in region r Armington aggregate of commodity i for demand category (item) g in region r Aggregate imports of commodity i in region r Unemployment rate (rationing of labor supply)
Table A3 Price variables.
Appendix. Algebraic model summary Our central model consists of two classes of conditions that characterize a competitive market equilibrium: (i) zero-profit conditions for constant-returns-to-scale (CRTS) production and (ii) market-clearance conditions for all goods and factors. In equilibrium, zero-profit conditions determine activity levels and market-clearance conditions determine price levels. Market distortions can be captured through price and quantity constraints which are associated with endogenous endowment rationing and endogenous taxes or subsidies. The incorporation of involuntary unemployment through a wage-curve relationship provides an example for a price constraint which involves rationing of (labor) endowment (see (9) and (23) below). We use the notation Pzir to denote the unit-profit function (calculated as the difference between unit revenue and unit cost) for CRTS production of sector i in region r where z is the name assigned to the associated production activity. Exploiting Hotelling’s Lemma we can differentiate the unit-profit function with respect to input and output prices to obtain compensated demand and supply coefficients which are used subsequently in the market-clearance conditions. We employ g as an index for all sectors/commodities i (g¼i), the final consumption composite (g¼C), the public good composite (g¼G), and for aggregate investment (g¼I). The index r (aliased with s) denotes regions. The index EG represents the subset of all energy goods (here: coal, oil, gas, electricity) and the label FF denotes the subset of fossil fuels (here: coal, oil, gas).
pgr pM gr
Price of item g in region r Price of material composite for item g in region r
pEgr
Price of energy composite for item g in region r
pKL gr
Price of value-added composite for item g in region r
pAigr
Price of Armington good i for demand category (item) g in region r
pIM ir wr vr aKLE jr
Price of import composite for good i in region r
xtr
Technology-specific capacity rent in region r Carbon value in region r
2 pCO r
Price of labor (wage rate) in region r Price of capital services (rental rate) in region r Rent to fossil fuel resources in region r (iAFF)
Table A4 Endowments and emission coefficients. Lr
Aggregate labor endowment for region r
K ir
Capital endowment of sector i in region r
X tr
Capacity limit of electricity generation technology t in region r
Q ir
Endowment of fossil fuel resource i for region r (iAFF)
Br
Initial balance of payment deficit or surplus in region r (note:
P
B r ¼ 0)
r
CO 2r Endowment of carbon emission rights in region r 2 Carbon emission coefficient for fossil fuel i in demand category g of aCO igr region r (iAFF) er Wage curve coefficient in region r
24
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
Table A5 Cost shares.
yM gr
Cost share of the material composite in production of item g in region r
yEgr
Cost share of the energy composite in the aggregate of energy and valueadded of item g in region r Cost share of the material input i in the material composite of item g in region r Cost share of the energy input i in the energy composite of item g in region r Cost share of capital within the value-added of item g in region r
yMN igr yEN igr yKgr yQgr
Cost share of fossil fuel resource in fossil fuel production (gAFF) of region r Cost share of labor in non-resource inputs to fossil fuel production (gAFF) of region ra Cost share of capital in non-resource inputs to fossil fuel production (gAFF) of region ra Cost share of good i in non-resource inputs to fossil fuel production (gAFF) of region r Cost share of domestic output i within the Armington item g of region ra
yLgr yKgr yFF igr yAigr yM isr
Cost share of exports of good i from region s in the import composite of good i in region r
a For the activity analysis representation of electricity production the set L K A of value shares expands to ytr , ytr , yitr :
Table A6 Elasticities. KLEM gr
s
sKLE gr sM gr sKL gr sEgr yELE gr yOGC gr yOG gr
sQgr sXtr
sAir sIM ir gr a b c
Substitution between the material composite and the energy-valueadded aggregate in the production of item g in region ra Substitution between energy and the value-added nest of production of item g in region ra Substitution between material inputs within the energy composite in the production of item g in region ra Substitution between capital and labor within the value-added composite in the production of item g in region ra Substitution between energy inputs within the energy composite in the production of item g in region r (by default: 0.5) Substitution between electricity and the other oil–gas–coal composite in the production of item g in region r Substitution between the oil–gas composite and coal in the production of item g in region r Substitution between oil and gas in the production of item g in region r Substitution between natural resource input and the Leontief composite of all other inputs in fossil fuel production (gAFF) of region r (calibrated consistently to exogenous supply elasticities)b Substitution between technology-specific resource (capacity) input and the Leontief composite of all other inputs in electricity generation by technology t in region r (calibrated consistently to exogenous supply elasticities) Substitution between the import composite and the domestic input to Armington production of good i in region rc Substitution between imports from different regions within the import composite for good i in region rb Wage curve elasticities
see Okagawa and Ban (2008). see Graham et al. (1999) and Krichene (2002). see Badri and Walmsley (2008).
and the CES aggregate of value-added and energy.6 YY KLEM KLE M Mð1sgr Þ M E Eð1sgr Þ ¼ pgr ½ygr pgr þ 1ygr ½ygr pgr gr KL 1sKLE KLEM ð gr Þ ð1sKLEM E Þ=ð1sKLE Þ r0 gr gr Þ 1=ð1-sgr þ 1ygr pgr
6
Note that the input to the value-added composite for final demand components (g ¼C, G, I) is zero.
(2) Sector-specific material aggregate: Non-energy commodities enter into the material aggregate subject to a constant elasticity of substitution. 2 31=ð1sMgr Þ YM X MN Að1sMgr Þ M 4 5 ¼p y p r0 gr
gr
igr
igr
i= 2EG
(3) Sector-specific energy aggregate: Energy goods enter into the energy aggregate subject to a constant elasticity of substitution. Carbon emissions are linked to (fossil) fuel inputs in fixed proportions. E " # 1sEgr 1=ð1sgr Þ X EN YE E A CO2 CO2 ¼ p y p þ p a r0 igr gr r igr igr gr i A EG
(4) Sector-specific value-added aggregate: Capital and labor trade off at a constant elasticity of substitution. h i1=ð1sKL YKL gr Þ K ð1sgrKL Þ þ 1yK wr ð1sKL gr Þ ¼ pKL r0 gr ygr vr gr gr (5) Production of fossil fuels (gAFF): A fuel-specific resource trades off with all other inputs (which are combined in fixed proportions) at a constant elasticity of substitution. The latter can be calibrated to be consistent with exogenous estimates of fossil fuel supply elasticities. YY gr
2
Q
1sQ gr
¼ pgr 4ygr qgr
1sQgr X A Q L K þ 1ygr ygr wr þ ygr vr þ yigr pAigr
#1=ð1sQgr Þ r 0
i
(6) Armington aggregate: The domestically produced commodity is combined with a composite of imported commodities (of the same variety) at a constant elasticity of substitution. 1=ð1sA Þ r 0 YA A ir 1sA A 1sir A ir A IM p ¼ p y p þ 1 y igr igr ir igr ir igr (7) Aggregate imports across import regions: Within the import composite imported goods (of the same variety) from different regions trade off at a constant elasticity of substitution. 2 #1=ð1sIM Þ r 0 ir YIM X IM 1sIM IM 4 ir ¼ p y pis ir
isr
ir
s
(8) Activity analysis representation of electricity production: The top-down representation of electricity production through nested CES functions (see (1)) can be replaced by means of bottom-up activity analysis with discrete power production technologies producing the same (homogenous) output. For each power production technology a technologyspecific resource (capacity) is combined with a Leontief composite of all other inputs subject to a constant elasticity of substitution. The latter can be calibrated to be consistent with exogenous estimates of supply elasticities. h YY X 1sX X L K ¼ pELE,r ytr xtr tr þ 1ytr ytr wr þ ytr vr tr X # 1sXtr 1=ð1str Þ r 0 X A yitr pAi,ELE,r þ i
(9) Market-clearance conditions Labor: Fixed labor endowment adjusted by involuntary unemployment matches aggregate labor demand across all production sectors.
ð1ur ÞL r Z
X g
Y KL gr
@
QKL gr
@ wr
þ
X t
Y tr
Q @ Ytr : @wr
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
25
(10) Capital: Fixed capital endowment matches aggregate capital demand across all production sectors. Q Q X KL @ KL X @ Ytr gr Kr ¼ Y gr þ Y tr : @vr @vr g t
(17) Armington composite: Supply of the good- and sector-specific Armington composite meets its intermediate demand. Q Q @ Ygr X @ Ytr Aigr ¼ Y gr þ Y : tr @pAigr @pAi,ELE,r t
(11) Fossil fuel resources (gAFF): The fixed endowment with a specific resource meets the demand for the production of each fossil fuel. Q @ Ygr Q gr Z Y gr : @ qgr
(18) Commodities (g¼i): Supply of each commodity meets its demand across Armington production activities as well as export demand in the import composites of all other regions. Q Q X @ Aigr X @ IM is Y ir Z Aigr þ IM is : @pir @pir g sar
(12) Technology-specific resources: The fixed endowment with a specific resource (capacity) meets its demand in power production for each technology. Q @ Ytr X tr Z Y tr : @ xtr (13) Material composite: Supply of the sector-specific material composite meets its intermediate demand. Q @ Ygr M gr Z Y gr : @pM gr (14) Energy composite: Supply of the sector-specific energy composite meets its intermediate demand. Q @ Ygr : Egr ZY gr @pEgr (15) Value-added composite: Supply of the value-added composite meets its intermediate demand. Q @ Ygr : KLgr Z Y gr @ pKL gr (16) Import composite: Supply of the sector-specific import composite meets its intermediate demand. Q X @ Aigr Aigr : IMir Z @ pIM g ir
(19) Private consumption (g¼C): Supply of the composite consumption good equals demand which is determined through the disposal income of the representative agent. X X X 2 Y Cr pCr Z wr L r þ vgr K gr þ qir Q ir þ xtr X tr þ pCO CO 2r þ B r : r g
t
i A FF
(20) Public consumption (g¼G): Supply of the composite public good meets the exogenous demand for public good provision. Y Gr Z G r : (21) Investment (g ¼I): Supply of the investment good meets the exogenous demand for investments (savings). Y Ir ZI r : (22) Carbon emissions: The economy-wide budget of CO2 emission allowances provides an upper bound on the permissible CO2 emissions associated with fossil fuel combustion in production and consumption. CO 2r Z
XX g
@PE gr 2 aCO Egr igr : A 2 CO2 aigr @ pigr þpCO i A FF r
(23) Price constraint Wage curve: We introduce unemployment through the specification of a wage curve which postulates an inverse relationship between the real wage and the rate of unemployment. wr log ¼ er þ gr logður Þ: pCr
Domestic output
CES KLEM CES ( σgr )
Capital-Labour-Energy (KLE)
Material CES composite (M)
( σgrM )
KLE CES ( σgr )
Capital-Labor (KL) CES ( Capital (K)
KL σgr
Energy (E) CES ( σgr
ELE
) Oil-Gas-Coal
Labor (L)
Electricity
CES ( σgr Coal
OGC
Oil-Gas
OG CES ( σgr )
Oil
Gas
Fig. A1. Nesting in production (except for fossil fuels).
)
)
26
C. B¨ ohringer, T.F. Rutherford / Energy Policy 55 (2013) 16–26
Domestic output
CES (σ Qgr )
Fuel-specific resource
Other resource inputs Leontief
Intermediate inputs
Labour
Capital
Fig. A2. Nesting in fossil fuel production.
Domestic output
CES (σ Xtr )
Other resource inputs
Technology-specific capacity
Leontief Intermediate inputs
Labour
Capital
Fig. A3. Nesting in technology-specific electricity production.
Armington good CES ( σ ir ) A
Domestic output
CES import composite from other regions
( σ irIM)
Fig. A4. Nesting in Armington production.
References ¨ Anger, N., Bohringer, C., Moslener, U., 2007. Macroeconomics impacts of the CDM: the role of investment barriers and regulations. Climate Policy 7 (2007), 500–517. Armington, P.S., 1969. A theory of demand for producers distinguished by place of production. IMF Staff Papers 16 (1), 159–178. Badri, N.G., Walmsley, T.L. (Eds.), 2008. Purdue University. Blanchflower, D.G., Oswald, A.J.,, 1995. An introduction to the wage curve. Journal of Economic Perspectives 9, 153–167. ¨ Bohringer, C., Rutherford, T.F., 2010. The cost of compliance: a CGE assessment of Canada’s policy options under the Kyoto Protocol. World Economy 33 (2), 177–211. ¨ Bohringer, C., Tol, R.S.J., Rutherford, T.F., 2009. The EU 20/20/2020 targets: an overview of the EMF22 assessment. Energy Economics 31 (2), 268–273. Bukowski, M., Kowal, P., 2010. Large Scale, Multi-sector DSGE Model as a Climate Policy Assessment Tool—Macroeconomic Mitigation Options (MEMO) Model for Poland, IBS Working Paper #03/2010. Institute for Structural Research (IBS), Warsaw. DCCEE, Department of Climate Change and Energy Efficiency, Australia, 2011. Status of global mitigation action—current targets and policies in key countries. Available from: /http://www.climatechange.gov.au/government/ initiatives/multi-party-committee/ /media/publications/mpccc/status-globalmitigation-action.pdfS.
European Commission, 2008a. The climate action and renewable energy package, Europe’s climate change opportunity. Available from: /http://ec.europa.eu/ environment/climat/climate_action.htmS. European Commission, 2008b. European energy and transport—trends to 2030 (2007 update). Available from: /http://ec.europa.eu/dgs/energy_transport/figures/tren ds_2030_update_2007/energy_transport_trends_2030_update_2007_en.pdfS. EIA, 2008. International Energy Outlook’, Energy Information Administration, US Department of Energy. Available from: /http://www.eia.doe.gov/oiaf/ieo/S. Graham, P., Thorpe, S., Hogan, L., 1999. Non-competitive market behavior in the international coking coal market. Energy Economics 21, 195–212. Goulder, L.H., 1995. Environmental taxation and the double dividend: a reader’s guide. International Tax and Public Finance 2, 157–185. Hutton, J., Ruocco, A., 1999. Tax reform and employment in Europe. International Tax and Public Finance 6 (3), 263–287. Krichene, N., 2002. World crude oil and natural gas: a demand and supply model. Energy Economics 24, 557–576. Okagawa, A., Ban, K., 2008. Estimation of Substitution Elasticities for CGE Models. Mimeo, Discussion Paper 08–16, Osaka University, Osaka, Japan. Available from: /http://www.esri.go.jp/jp/workshop/080225/03_report5_Okagawa.pdfS. Rutherford, T., 2002. Lecture Notes on Constant Elasticity Functions. University of Colorado.
