Policy Research Working Paper 8909 Implications for Provincial Economies of Meeting China’s NDC through an Emission Trading Scheme A Regional CGE Modeling Analysis Jun Pang Govinda Timilsina Development Economics Development Research Group June 2019 Policy Research Working Paper 8909 Abstract This study analyzes the potential impacts of a national emis- 2030 drops by half, to 15 percent. At the national level, the sion trading scheme on provincial economies in China of emission trading scheme would cause a 1.2 to 1.5 percent meeting China’s emission reduction pledges, the Nation- reduction in gross domestic product from the business as ally Determined Contributions announced under the Paris usual scenario in 2030. If the baseline is corrected, the Agreement. The study developed a multiregional, multisec- impact on gross domestic product drops by two-thirds. toral, recursive-dynamic computable general equilibrium The emission trading scheme would cause some provin- model and calibrated it with the latest provincial-level social cial economies to gain and others to lose. The economic accounting matrices (2012). The study shows that meet- impacts are highly sensitive to the allowance allocation rules. ing China’s Nationally Determined Contributions through Not only the magnitudes, but also the directions of the an emission trading scheme would reduce almost 30 per- economic impacts alter when the allocation rules change. cent of the emission reduction from the business as usual The provinces that rely on coal mining or coal-intensive scenario in 2030. If the baseline is corrected based on infor- manufacturing industries are found to experience relatively mation from a bottom-up energy sector model, TIMES, larger economic losses irrespective of the allowance alloca- the required reduction of emissions from the baseline in tion rules. This paper is a product of the Development Research Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at gtimilsina@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Implications for Provincial Economies of Meeting China’s NDC through an Emission Trading Scheme: A Regional CGE Modeling Analysis 1 Jun Pang and Govinda Timilsina 2 Keywords: emissions trading scheme, renewable energy standards, CGE model; climate change mitigation, China JEL Classifications: D58, Q54 1 The authors would like to thank Martin Raiser, Christophe de Gouvello, Bekele Debele, Dafei Huang, Yanqin Song, Xi Xiangyang, Feng Xiagzhao, Mao Xianqiang, Liu Qiang, Duan Hongbo and Fu Guanjan for their valuable comments and suggestions. The views and interpretations are of authors and should not be attributed to the World Bank Group and the organizations they are affiliated with. We acknowledge World Bank’s Research Support Grant (RSB) for financial support. 2 Jun Pang (pangjun2005@ruc.edu.cn) is a Professor at the School of Environment and Natural Resources, Renmin University , Beijing, China; Govinda Timilsina (gtimilsina@worldbank.org) is a Senior Economist at the Development Research Group, World Bank).   Implications for Provincial Economies of Meeting China’s NDC through an Emission Trading Scheme: A Regional CGE Modeling Analysis 1. Introduction China has made a pledge under the Paris Climate Agreement (‘Paris Agreement’ hereafter) to reduce its emission intensity of the economy by 60-65% in 2030 from 2005 levels (UNFCCC, 2015).  The pledges made by various countries under the Paris Agreement are referred to as their ‘Nationally Determined Contributions’ or ‘NDC’ (UNFCCC, 2016). China is implementing both market and non-market (i.e., regulatory or administrative) policy instruments to achieve its NDCs. For the market instruments, China, to date, has become the leader in implementing the clean development mechanism (CDM). Over the past few years, China experimented with domestic emission trading through seven pilot projects in Beijing, Shanghai, Tianjin, Chongqing, Guangdong, Hubei and Shenzhen (Zhang, 2015). A number of studies have assessed the design, implementation and the performance of these pilots (Zhang et al. 2017; Zhang et al. 2016; Pang and Duan 2016; Munnings et al. 2016; Zhang, 2013). Following the pilot projects, China introduced a national emission trading scheme in 2017. At present, the national ETS is limited to the power sector only. Its extension to other emission- intensive sectors (e.g., iron and steel, cement) is expected. It has always been a great challenge for the Chinese government to narrow the development gap or promote balanced development among different regions (Lo, 2017; Duan and Zhou, 2017). After the nationwide emission trading scheme is introduced, some provinces would be the buyer of the emission allowances and some would be the seller, thereby causing some redistribution of wealth. Also, for some sectors, specifically the energy-intensive and emission-intensive ones, the relative competitiveness in various provinces would probably change, which may lead to the relocation of capital and investment among different regions (Zhu et al. 2017). Since fossil energy accounts for about 90 percent of the total energy 2      consumption in China, a critical question is how the emission trading scheme would help substitute fossil energy sources with non-fossil ones. The study aims to analyze the economy-wide impacts of the national ETS, and especially focuses on the distribution effect among different provinces and economic sectors. It investigates how much each province will reduce its GHG emissions by 2030 to meet the NDC; and how the provinces would trade their emission allowances under alternative schemes for initial allocations of emission allowances. It assesses the economic costs (impacts on GDP), household consumption and welfare, investment, commodity prices and international trade of goods and services. Since the economic impacts of a carbon pricing policy are sensitive to the selection of a baseline, we used information from a bottom-up engineering model, TIMES, that is capable of representing detailed technologies, of correcting the baseline estimated by the CGE model. This is important because CGE models tend to inflate emission baselines. Researchers often adjust the inflation by using an autonomous energy efficiency index (AEEI) parameter. However, the selection of the index is arbitrary. Instead of doing so, we used a correction factor informed by Timilsina et al. (2019), which examines the deviation of the baseline when the CGE model is linked with a bottom-up TIMES model developed for China. There exists a rich literature examining the ETS for controlling environmental pollution. The classic economic theory suggests that an ETS is cost-effective compared to command and control policy for limiting environmental pollution (Coase, 1960; Rose and Tietenberg, 1993). Investigating the ETS for controlling greenhouse gas emissions (GHG) in China, Jotzo and Loschel (2014), Bohringer et al. (2014), Zhang et al. (2014) and Wu et al. (2014) suggest that an ETS could be an efficient policy to reduce GHG emissions in China and thereby could help achieve its climate change mitigation targets. Hübler et al. (2014) assess the economic cost of an ETS in China. 3 Cui et al. (2013, 2014) design an ETS for China based on the carbon                                                              3 It finds that the ETS required to lower China’s emissions intensity per GDP by 45% from 2005 level in 2020 would cause 1% welfare reduction from the business-as-usual (BAU) scenario. 3      abatement cost curves, and then apply it to meet the emission intensity targets set in the twelfth and thirteenth Five-Year Plans. It shows that the cost-saving effect of an ETS depends on the emission reduction target and the regional coverage scope. Goulder et al. (2017) highlights some insights on why China introduced a national ETS instead of a carbon tax. The carbon allowance allocation plays a critical role in the overall design of an ETS. Existing studies, such as Wu et al. (2016), Grimm and Ilieva (2013), Rose et al. (2014), examine the economic and distribution impacts of different initial quota allocation methods, including grandfathering, auctioning and benchmarking. With a laboratory experiment to compare the auctioning and grandfathering approaches for emission cap allocation, Grimm and Ilieva (2013) find that the level of the permit price is sensitive to the allowance allocation mechanism. Rose et al. (2014) compare alternative allocation approaches (free allocation vs. auctioning) for California’s emission trading system and show that free allocation of emission permits benefits more the highest income groups of households thereby deteriorating distributional equity under the emission trading system. Some studies investigate the economic impact of initial quota allocation under ETS in China (Wu et al. 2016; Yuan et al. 2012;  Zhang et al. 2014; Tang et al. 2015). Wu et al. (2016) developed a multi-regional CGE model for China to assess the economic effect of six alternatives for initial quota allocation. Yuan et al. (2012) evaluated the impact of the carbon allowance allocation on China’s regional development. Zhang et al. (2014) assess the economic and distributional impacts of different allocation schemes across the provinces. Tang et al. (2015) use a multi-agent macroeconomic model to analyze impacts of an ETS under two alternative approaches for allowance allocation: grandfathering and benchmarking. The existing studies examining China’s ETS have several limitations. Most of the studies do not examine the impacts of national ETS across the provinces because the models they use do not explicitly represent provincial economies. They analyze the impacts of an ETS across the production sectors, but not across the provinces. Some studies (Yuan et al. 2012, Wu et al. 2016) have developed regional CGE models, however their models are static and are not useful for analyzing the economic impacts of meeting China’s NDC in 2030. Timilsina et al. (2018) analyze the impacts of meeting China’s NDC using a recursive dynamic CGE model, 4      however, the study uses a carbon tax instead of ETS to meet the NDC, and the study does not explicitly represent provincial economics. To the knowledge of the authors, this is the first study that uses a recursive dynamic, multi-regional CGE model with latest available data (2012 provincial SAM) to analyze an ETS to meet China’s NDC under the Paris Agreement. The study finds that meeting China’s NDC through a national emission trading system would cause 0.4% to 1.5% reduction of economic output at the national level (GDP) in 2030 depending upon allowances allocation as well as baseline setting rules. The study is organized as follows. Section 2 presents the details of the multi-region, recursive dynamic CGE model developed for the study, followed by a discussion on data in Section 3. Section 4 presents discussions of results from the model simulations followed by key conclusions drawn in Section 5. 2. Methodology In this section, we present a description of the regional CGE model developed for this study. The Naming rules for variables in the equations: (1) Endogenous variables are named as uppercase letters, and exogenous variables are named as uppercase letters with a cross line above; (2) Variables are generally named as their recognized prefix, as Q is quantity, P is commodity price, W is the factor price, Y is income, E is expenditure; (3) The number of provincial regions in this model is m, and each region has n production sectors; (4) The initial value of all variables in the business-as-usual scenario are used as parameters and are named as original names with 0 added behind, while other parameters are named as lowercase letters. Definition of sets: I,J = {production sectors or commodities}； R,S = {regions}； F = {factors, including capital and labor}； E(  I,J) = {energy sector or commodity}； 5      NE(  I,J) = {non-energy sector or commodity}； The CGE model used in the study is a multi-regional recursive dynamic model to analyze the economic effects of energy and environmental policies in China with distinct description of 31 provinces. It explicitly models the behavior of four economic agents: households, government, enterprises and the rest of the world (ROW). Production sectors are classified into 16 sectors, of which five are energy supply sectors (coal mining, oil and gas extraction, petroleum refinery, gas processing, and electric power generation). Please see Table 1 for the definitions of the sectors. Table 1. Definition of sectors/commodities in the CGE model Sector Name Definition or coverage AGRI Agriculture, Forestry, Animal Husbandry and Fishery COAL Mining and washing of coal OILNG Extraction of petroleum and natural gas MINE Mining and processing of metal and nonmetal FTPMF Food, tobacco, textile, leather, fur, feather, timber, furniture, paper, printing PETRO Processing of petroleum, coking, processing of nuclear fuel CHEMI Manufacture of chemical products NMETA Manufacture of non-metallic mineral products METAL Smelting and processing of metals OTHMF Other manufacture ELECT Production and distribution of electric power and heat power GAS Production and distribution of gas WATER Production and distribution of tap water CONST Construction TRANS Transport, storage and postal services SERVI Other services 2.1 Representing the production sector Figure 1 illustrates the modeling of the behavior of each production sector. We use a six- tier nested constant elasticity of substitution (CES) combination function to represent a production sector. This multi-tier CES representation provides flexibility to the model by allowing different substitution possibilities across the tiers. Like in most CGE model formulations, we assume that the market follows perfect competition and the production process follows constant returns to scale. 6      Total Regional Output CES Labor-Capital-Energy Non-energy intermediate inputs CES Leontief NE 1 NE 2 … NE 11 NE 12 Labor Capital-Energy CES Capital Energy CES Fossil Fuels Electric Power CES COAL Refined Petroleum & Gas CES Refined Petroleum Gas Figure 1. Structure of the production function in the CGE model 2.1.1 Gross Output At the top of the nesting structure in Figure 1, the aggregation of value-added and intermediate input is expressed as:  iA  iA A , r 1/  i , r QAi , r  α iA, r  [ iA , r  QVAi , r ,r  (1- i A , r )  QINTAi , r ] (1) where QAi ,r is the total production of sector i in region r, QVAi ,r and QINTAi ,r are the input of value-added and intermediate input in sector i in region r respectively, i,r and αi ,r are the A A A share parameter and efficiency parameter; i ,r is the substitution elasticity parameter between value-added and intermediate input, and  i,Ar  1 (1- iA ,r ) ,  iA ,r is the substitution elasticity between value-added and intermediate input. Optimal factor input under total production: 7      A 1 i, A  QINTAi , r  r PVAi ,r  i ,rA   (2) PINTAi , r 1- i ,r  QVA   i ,r  where PVAi ,r and PINTAi ,r are the price of value-added and intermediate input in sector i in region r respectively. Relationship of price of total output: PAi,r  QAi,r  (1 tlgindi,r -tcgindi,r )  PVAi,r  QVAi,r  PINTAi,r  QINTAi,r (3) where PAi ,r is the producer price of sector i in region r, tlgindi,r and tcgindi,r are the rate of indirect tax sector i pays to the local government in region r and the central government respectively. 2.1.2 Non-energy intermediate inputs In the right-hand side of the second tier (from the top) in Figure 1, the aggregate intermediate input of non-energy commodity is given as follows: QINT j ,i ,r  ica j ,i ,r  QINTAi ,r j  NE (4) The price of intermediate input: PINTAi  ica j ,i  PQi j  NE (5) j where QINT j ,i ,r is the quantity of the input of non-energy commodity j as intermediate input of sector i in region r, ica j ,i ,r is the intermediate input coefficient, denoting the proportion of the input of non-energy commodity j in the total intermediate input of sector i in region r. 2.1.3 Labor and capital-energy composite In the left hand side of the same tier (second tier from the top), the composite of labor and capital-energy is derived as: iva iva va ,r 1/  i ,r , r  [ i , r  QLDi , r QVAi ,r   iva va ,r  (1   iva , r )  QKEDi , r ] (6) 8      where QLDi ,r and QKEDi ,r are the input of labor and capital-energy in sector i in region r respectively,  i ,r and i ,r are the share parameter and efficiency parameter;  va va va  i ,r is the , r  1 (1-  i ,r ) ,  i ,r substitution elasticity parameter between labor and capital-energy, and  iva va va is the substitution elasticity between labor and capital-energy. Optimal factor input of value-added: 1 iva  va  QKEDi , r  ,r WLi , r  i , r va   (7) PKEi ,r 1   i , r  QLD   i ,r  Relationship of price of the input of value-added: PVAi ,r  QVAi ,r  WLi ,r  QLDi ,r  PKEi ,r  QKEDi ,r (8) where WLi ,r and PKEi ,r are the price of the input of labor and capital-energy in sector i in region r respectively. 2.1.4 Capital-energy composite In the third tier, the composite of capital and energy is given as follows: ike  ke 1/ ike ,r  [ i , r  QKDi , r QKEDi,r  ike ke ,r  (1  ike ,r )  QEDi , r i ,r ] ,r (9) where QKDi ,r and QEDi ,r are the input of capital and energy in sector i in region r respectively,  i ,r and i ,r are the share parameter and efficiency parameter; ke ke ke  i ,r is the , r  1 (1-  i ,r ) , substitution elasticity parameter between the input of capital and energy, and  ike ke  ke is the substitution elasticity between the input of capital and energy. i ,r Optimal factor input: 1 ike  ke  QEDi ,r  .r WK i ,r  i , r ke    (10) PECi ,r 1   i ,r  QKDi ,r   9      where WKi ,r and PECi ,r are the price of the input of capital and energy in sector i in region r respectively. Relationship of price of the input of capital and energy: PKEi,r  QKEDi,r  WKi,r  QKDi,r  PECi,r  QEDi,r (11) 2.1.5 Composite of electricity and aggregate fossil fuels In the fourth tier of the CES function, the composite of fossil fuel aggregates and electricity is given as: iec iec ec ,r 1/  i ,r , r  [ i , r  QEFi , r QEDi ,r   iec ec ,r  (1   iec , r )  QEEi , r ] (12) where QEFi ,r and QEEi ,r are the input of fossil fuels and electric power in sector i in region r respectively,  i ,r and i ,r are the share parameter and efficiency parameter; ec ec ec  i ,r is the substitution elasticity parameter between the input of fossil fuels and electric power, and , r  1 (1-  ) ,   iec ec i ,r ec is the substitution elasticity between the input of fossil fuels and electric i ,r power. Optimal factor input of the aggregation of energy: 1 iec  ec  QEEi ,r  ,r PEFi , r  i ,r ec    (13) PEEi ,r 1   i , r   QEFi ,r   where PEFi ,r and PEEi ,r are the price of the input of fossil fuels and electric power in sector i in region r respectively. Price relationship of the input of the aggregation of energy: PECi ,r  QEDi ,r  PEFi ,r  QEFi ,r  PEEi ,r  QEEi ,r (14) 2.1.6 Coal and refined petroleum/gas composite 10      In the fifth tier of the CES function, the aggregation of coal and refined petroleum/gas composite is given as: ief ief ef ,r 1/ i ,r , r  [ i , r  QEcoali , r QEFi ,r   ief ef ,r  (1-  ief , r )  QEoilgasi , r ] (15) where QEcoali ,r and QEoilgasi ,r are the input of coal and refined petroleum-gas in sector i in region r respectively,  i ,r and i ,r are the share parameter and efficiency parameter; ef ef ef i ,r is the substitution elasticity parameter between the input of coal and refined petroleum-gas, and , r  1 (1-  ) ,   ief ef i ,r ef is the substitution elasticity between the input of coal and refined i ,r petroleum-gas. First order condition of optimal factor input: 1  ief  ief  QEoilgasi . r  ,r PEcoali , r  ,r   (16) PEoilgasi , r (1   i , r )  ef  QEcoali , r   where PEcoali ,r and PEoilgasi ,r are the price of the input of coal and refined petroleum-gas in sector i in region r respectively. Price relationship of the input of the aggregation of fossil fuels: PEFi ,r  QEFi ,r  PEcoali ,r  QEcoali ,r  PEoilgasi ,r  QEoilgasi ,r (17) 2.1.7 Petroleum products and processed gas composite Finally, at the sixth tier of the CES function, the aggregation of processed gas and refined petroleum is given as: 1/ ipg   i ,pg r   QEoili , r  1   i ,pg pg i , r i ,r pg ,r QEoilgasi ,r   pg r  QEgasi , r (18) i ,r   where QEgasi and QEoili are the input of gas and refined petroleum in sector i in region r respectively,  i pg and  ipg are the share parameter and efficiency parameter; pg i is the 11      substitution elasticity parameter between the input of gas and refined petroleum, and  ipg  1 (1-  pg ) ,  i i pg is the substitution elasticity between the input of gas and refined petroleum. First order condition of optimal factor input: 1 ipg  i pg  QEoili ,r  ,r PEgasi , r  ,r    (19) PEoili , r (1   i ,pg   r )  QEgasi , r  where PEgasi ,r and PEoili ,r are the price of the input of gas and oil in sector i in region r respectively. Price relationship of the input of the aggregation of oil and gas: PEoilgasi ,r  QEoilgasi ,r  PEoili ,r  QEoili ,r  PEgasi ,r  QEgasi ,r (20) 2.1.8 Income and expenditure of firms A production sector’s income comes from capital returns and export tax rebates from the central government. YENTr  shifek WKi,r  QKDi,r  terfr   PEi,r  QEi,r (21) i i where YENTr is the income of enterprises in region r, shifek is the coefficient of the enterprises’ share in capital revenue, terfr is the rate of export tax rebate. The enterprises’ expenditure consists of the enterprise income taxes paid to the local and central government. EXENTr   telr  tecr   WKi,r  QKDi,r (22) i where EXENTr is the expenditure of enterprises in region r, telr and tecr are the rate of income tax the enterprises in region r pay to the local and central government, respectively. 12      Accordingly, enterprise saving is: ESAVr  YENTr  EXENTr (23) where  ESAVr is enterprise saving in region r. 2.2 Modeling inter-provincial and international trade Total output in each region supplies both the domestic and foreign markets. The domestic market is further divided into the local market and markets in other regions in mainland China. Similarly, composite commodities, including commodities produced in the local region, other regions in mainland China and the rest of the world, meet the local demand for commodities and are used for residential and governmental consumption and investment. This paper follows “small-country assumptions”, namely commodity price in the global market is exogenous. China is a price-taker. Changes in the quantity of its imports and exports cannot influence the price. In this model, the imperfect substitutability between commodities supplied in the local region and transferred to other regions in mainland China, and that between commodities supplied in the domestic market and exported to the rest of the world are represented by a CET function with two tiers of nesting. For commodities transferred from other regions in mainland China and imported from the rest of the world, the Armington assumption is followed and a two-tier nested CES function is used to denote the imperfect substitutability between commodities produced in the local region and transferred from other regions in mainland China, and that between commodities produced domestically and imported from the rest of the world. The above description is presented in Figure 2: 13      Total Regional Consumption CES Region … Region Region … Region Import from Domestic produced 31 1 r-1 r+1 Rest of World CES CET Transfers out to Rest Local Produced for Transfers in from Rest of Mainland of China Local Consumption of Mainland of China CES CET Region Region Region Region … … 31 Export to Rest Domestic 1 r-1 r+1 of World Supply CET Total Regional Output Figure 2 Structure of total regional supply and demand 2.2.1 Exports and transfers out The first tier, CET function describing the allocation of supply between domestic market and export: 1/ it,r QAi ,r    it,r  QDAi.r i ,r  (1  it,r )  QEi ,r i ,r  t t   t , it,r  1 (24)  i ,r  where QDAi ,r and QEi ,r are the supply of commodity produced in sector i in region r to domestic market and export respectively, i,r and i ,r are the share parameter and efficiency t t t parameter;  i ,r is the transformation elasticity parameter between domestic market supply and export, and  it,r  1 (  it,r -1) ,  it,r is the transformation elasticity between domestic market supply and export. First order condition: 14      t 1 i,  t   QEi,r  r PDAi, r   i, r t      (25) PEi,r  1     i, r   QDAi,r  where PDAi,r and PEi,r are the domestic price and export price of the commodity produced in sector i in region r. Relationship of price: PAi ,r  QAi ,r  PDAi ,r  QDAi ,r  PEi ,r  QEi ,r (26) Exchange rate conversion between the price of the export commodity in the global market and renminbi: PEi , r  PWE i , r  EXR (27) where PWEi ,r is the international market price of exported commodity i in region r, EXR is the exchange rate. The second tier, CET function describing the allocation of supply between local region and other regions: bl 1/ i, QDAi,r  α  i,bl i,r bl i,r  bl r bl r  QBDi, r  (1   i,bl r )  QLCi, r , i, bl r 1 (28)  i, r  where QBDi ,r and QLCi ,r are the supply of commodity produced in sector i in region r to local market and markets in other regions respectively, i,r and i ,r are the share parameter and bl bl bl efficiency parameter; i ,r is the transformation elasticity parameter between supply in local , r  1 (  i ,r -1) ,  i ,r is the transformation elasticity between region and other regions, and  ibl bl bl supply in the local region and other regions. First order condition: 15      1 ibl   bl   QLCi , r  ,r PBDi , r   i , r bl      (29) PLCi , r  1     i ,r   QBDi , r  where PBDi,r and PLCi,r are the price of commodity produced in sector i in region r in local market and markets in other regions, respectively. Relationship of price: PDAi ,r  QDAi ,r  PBDi ,r  QBDi ,r  PLCi ,r  QLCi ,r (30) The third tier, CET function describing the allocation of commodity for transfers out among other regions: 1 (1 ilc ,r , s ) 1 ilc QLCi ,r   ilc    ,r ,s     ilc 1 (1 ilc  lc ( ilc ,r ,s 1) QRLCi ,r , s  lc   ,r ,s  ,r ,s )  PRLCi ,r , s i ,r ,s  ( s  r ), ilc ,r ,s  1  i ,r ,s    PRLCi ,r , s   s ,r ,s  (31) where QRLCi ,r ,s and PRLCi ,r ,s are the quantity and price of the supply of commodity produced in sector i in region r to region s, i,r.s and i ,r ,s are the share parameter and efficiency lc lc lc parameter;  i ,r , s is the transformation elasticity parameter of commodity for transfers out , r , s  1 (  i ,r ,s -1) ,  i ,r ,s is the transformation elasticity of among other regions, and  ilc lc lc commodity for transfers out among other regions. PLCi ,r  QLCi ,r   PRLCi ,r ,s  QRLCi ,r , s s  r (32) s 2.2.2 Imports and transfers in The first tier, CES function describing the choice between domestic and import commodity: 1/ iq,r QQi ,r     iq i ,r q i , r  q , r  QDCi , r q  (1   iq , r )  QM i , r (33)  i ,r  16      where QQi ,r , QDCi ,r and QM i ,r are the demand for composite commodity i, domestic commodity i and import commodity i in region r respectively, i,r and i ,r are the share q q q parameter and efficiency parameter;  i ,r is the substitution elasticity parameter between domestic and import commodity, and  iq,r  1 (1-  iq,r ) ,  irc ,r is the substitution elasticity between domestic and import commodity. First order condition: 1 iq,r PDCi , r  q   QM i , r    i ,r q      (34) PM i ,r  1     i ,r   QDCi ,r  where PDCi ,r is the price of domestic commodity i in region r, PM i ,r is the price of import commodity i in region r. Composite commodity price is the weighted mean of the price of domestic and import commodity: PQi ,r  QQi ,r  PDCi ,r  QDCi ,r  PM i ,r  QM i ,r (35) where PQi ,r is the price of composite commodity i in region r. Exchange rate conversion between the price of import commodity in the international market and renminbi: PM i ,r  PWM i ,r  (1  tmi )  EXR (36) where PWM i ,r is the international market price of import commodity i in region r, tmi is the import tariff rate of import commodity i. The second tier, CES function describing the choice between commodities produced in local region and transferred from other regions: 17      1/ idr  dr  ,r id,r idr , r   i , r  QBDi , r QDCi ,r   idr r  (1   idr , r )  QLRi , r ,r (37)   where QBDi ,r and QLRi ,r are the demand for commodity i produced in local region and transferred from other regions in region r respectively,  i ,r and i ,r are the share parameter dr dr dr and efficiency parameter;  i ,r is the substitution elasticity parameter between commodity , r  1 (1-  i ,r ) ,  i ,r is the produced in local region and transferred from other regions, and  idr dr dr substitution elasticity between commodity produced in local region and imported from other regions. First order condition: 1 idr   dr   QLRi ,r  ,r PBDi ,r   i , r dr      (38) PLRi ,r  1     i ,r   QBDi ,r  where PBDi ,r is the price of locally produced commodity i in region r, PLRi ,r is the price of commodity i transferred from other regions in region r. Domestic commodity price is the weighted mean of the price of commodity produced in local region and transferred from other regions: PDCi ,r  QDCi ,r  PBDi ,r  QBDi ,r  PLRi ,r  QLRi ,r (39) The third tier, CES function describing the choice of commodity for transfers in among other regions: 1 (1  ilr , s ,r ) 1  ilr QLR   ilr    , s ,r     ilr ( s  r ) (40) 1 (1  ilr , s ,r )  lr (  ilr , s ,r 1) QRLRi , s , r  lr i , r   ,s ,r   PRLRi , s , r i ,s ,r   i , s ,r  PRLR i ,s, r    s ,s ,r  where QRLRi , s ,r and PRLRi ,s,r are the quantity and price of commodity i transferred from region s in region r, i,s.r and i,s,r are the share parameter and efficiency parameter; lr lr lr  i , s ,r is the substitution elasticity parameter of commodity for transfers in among other regions, and 18      , s , r  1 (1-   ilr lr i , s ,r ) ,  lr is the substitution elasticity of commodity for transfers in among other i , s ,r regions. PLRi ,r  QLRi ,r   PRLRi ,s ,r  QRLRi , s ,r s  r (41) s 2.3 Modeling household behavior In this model, the households’ income is composed of labor payments, capital revenue and transfer payments from local government. YH r  WLi ,r  QLDi ,r +shifhk WKi ,r  QKDi ,r  TSLGTOH r (42) i i where YH r is the income of households in region r, shifhk is the coefficient of the households’ share in capital revenue, TSLGTOH r is transfer payments from local government. The consumption function of households is assumed as a Cobb-Douglas utility function in this model, which can derive the final consumption of households as the following equation: PQi , r  QH i , r  shrhi , r  mpcr  (1  thlr  thcr )  YH r (43) where QHi ,r is the consumption of commodity i of households in region r, thlr and thcr are the rate of income tax the households in region r pay to the local and central government respectively, mpcr is the marginal propensity to consume of the households in region r, shrhi ,r is the share of the consumption of commodity i in the expenditure of households in region r. The households’ expenditure contains final consumption and income taxes paid to the local and central governments: EH r   PQi , r  QH i ,r   thlr  thcr   YH r (44) i where EH r is the expenditure of households in region r. Accordingly, household saving is: 19      HSAVr  YHr  EHr (45) where HSAVr is household saving in region r. 2.4 Modeling Central and Provincial Governments 2.4.1 Central government The central government’s income in region r is composed of indirect tax, tariff, income taxes paid by households and enterprises, and revenues turned over by the local government. YCGr  tcgindi,r  PAi,r  QAi,r  tmi  QMi,r  PMW i,r  EXR  thcr  YHr  tecr  WKi,r  QKDi,r  TRANSRTC r (46) i i i where YCGr is the income of the central government in region r, TRANSRTC r is revenues turned over by the local government in region r. The central government’s expenditure in region r includes commodity consumption, export tax rebates to enterprises and transfer payments to the local government. ECGr   PQi , r  QCGi , r  terf r   PEi ,r  QEi ,r  TRANSCTR r (47) i i where ECGr is the central government’s expenditure in region r, TRANSCTR r is the central government’s transfer payments to the local government in region r. In the equation above, consumption demand of the central government in region r is: PQi ,r  QCGi ,r  shrcgi ,r  mpccg r  YCGr (48) where shrcgi ,r is the spending share of the central government’s consumption of commodity i in region r, mpccgr is the central government’s marginal propensity to consume in region r. Accordingly, central government saving resulting from region r is: CGSAVr  YCGr  ECGr (49) 20      where CGSAVr is central government saving resulting from region r. 2.4.2 Local government The local government’s income consists of indirect tax, income taxes paid by households and enterprises, and transfer payments from the central government. YLGr  tlgindi,r  PAi,r  QAi,r  thlr  YHr +telr  WKi,r  QKDi,r +TRANSCTRr (50) i i where YLGr is the income of the local government in region r, TRANSCTR r is transfer payments from the central government. The local government’s expenditure contains commodity consumption, transfer payments to households and revenues turned over to the central government. ELGr   PQi ,r  QLGi ,r +TSLGTOH r +TRANSRTC r (51) i where ELGr is the local government’s expenditure in region r. In the equation above, consumption demand of the local government is: PQi,r  QLGi ,r  shrlgi ,r  mpclg r  YLGr (52) where shrlgi,r is the spending share of the local government’s consumption of commodity i, mpclgr is the local government’s marginal propensity to consume. Accordingly, local government saving is: LGSAVr  YLGr  ELGr (53) where LGSAVr is the local government saving in region r. 21      2.5 CO2 emissions and Emission Trading 2.5.1 Carbon emission Carbon emission coefficient (ton CO2/10,000 yuan) of three kinds of fossil fuel inputs (coal, refined petroleum, gas) of each industry in this model can be obtained from the data of the base year, which can derive the calculation of carbon emission as the following equations: QEMISi ,r  coef coal  QEcoali ,r  coef oil  QEoili ,r +coef gas  QEgasi ,r (54) where coef coal , coef oil , and coef gas are carbon emission coefficient of coal, crude oil & natural gas, refined petroleum, and gas, respectively, QEMISi ,r is the CO2 emissions of sector i in region r. QTEMIS  QEMISi,r (55) r i where QTEMIS is total amount of national carbon emissions. 2.5.2 Emission allowance or permit price CTRi  pcarbont  QEMISi (56) TOCTR   CTRi (57) i where pcarbont is the permit price at time period t, CTR i is the total money value of CO2 emitted from sector i. The total amount of national carbon allowance can be obtained as follows: QTPermitt  (1  mtratet )  QTEMISD 0 t (58) Where QTEMISD 0 t is total national CO2 emission under the dynamic baseline run, QTPermitt is the total amount of national carbon emission permits at time period t, mtratet is the annual 22      mitigation rate of national CO2 emission at time period t. The total carbon emission permits will be allocated to different ETS participants in two ways: full free allocation and full auction. 2.5.3 Modeling of ETS with a grandfathering allocation scheme If we take full free allocation, the emission-based grandfathering rule had been adopted here to allocate the total carbon emission permits to each industry: QEMISD0i ,r ,t QFPermiti ,r ,t   QTPermitt (59) QEMISD0t where QFPermiti ,r ,t is free carbon emission permits of industry i in region r, QEMISD0i ,r ,t is the CO2 emission of industry i in region r under the dynamic baseline run at time period t. We have the following equation to define the CO2 emission of each industry: QEMISi.r ,t  QFPermiti ,r ,t  TradingPermiti ,r ,t (60) Where QEMISi.r ,t is CO2 emission of industry i in region r, TradingPermiti ,r ,t is carbon emission permits sold or purchased of industry i of region r in ETS market, the latter will be positive when purchasing additional permits in ETS market and negative when selling redundant permits for profits in the ETS market. At the national ETS market, the sum of purchased trading permits should equal to the sum of sold trading permits at each time period, it means the total sum of trading permits of all industries at each time period should be zero.   TradingPermit r i i , r ,t 0 (61) The expenditure or income of industry i of region r in the ETS market at time period t can be calculated by use the following equation: carbon expi ,r ,t  pcarbont  TradingPermiti , r ,t (62) 23      Where carbon expr ,i ,t is the expenditure (positive) or income (negative) of industry i of region r in the ETS market, pcarbont is the carbon price of time period t in the national ETS market. Under the grandfathering allocation scheme, the cost equation at the first tier of the production function needs to subtract the value of free permits which should not be included in the expenditure of industry i. PAi,r ,t  QAi,r ,t (1 tcgindi.r  t lg indi,r )  PVAi,r ,t  QVAi,r ,t  PINTAi,r ,t  QINTAi,r ,t  pcarbont  QFPermiti,r ,t (3’) The first order condition of the fifth tier of the production function should be adjusted as: 1  ief （1  pca rbon  coef coal )  PEcoali , r  ief  QEoilgasi .r  ,r  ,r   (16’) PEoilgasi , r (1   i , r )  ef  QEco ali , r   The price relationship of the input of the aggregation of fossil fuels should be adjusted as: PEFi ,r  QEFi ,r  （ 1  pcarbont  coefcoal )  PEcoali ,r  QEcoali ,r  PEoilgasi ,r  QEoilgasi ,r (17’) Similarly, the first order condition of the sixth tier of the production function should be adjusted as: 1 ipg （1  pcarbont  coef gas )  PEgasi ,r  i ,pg  QEoili ,r  ,r  r   (19’) （1  pcarbont  coef oil )  PEoili ,r (1   i ,r )  pg   QEgasi ,r  The price relationship of the input of the aggregation of refined petroleum and gas should be adjusted as: PEoilgasi,r  QEoilgasi,r  （ 1 pcarbont  coefoil )  PEoili,r  QEoili,r  （ 1 pcarbont  coefgas )  PEgasi,r  QEgasi,r (20’) 2.5.4 Modeling of ETS with an auctioning allocation scheme If we allocate the carbon emission permits by full auction, the total auction permits should equal the total carbon mission permits at each time period. 24      QTAPermitt  QTPermitt (63) Where QTAPermitt is the total auction permits at the national level at time period t. The total auction permits equal the sum of permits purchased at the auction market of each industry. QTAPermitt  QAPermiti ,r ,t (64) r i Where QAPermiti ,r ,t is the permits of industry i of region r purchased at the auction market. For each industry, their CO2 emission should equal their permits purchased at the auction market. QEMISi,r ,t  QAPermiti,r ,t (65) Where QEMIS r ,i ,t is the CO2 emissions of sector i of region r at time period t. In this CGE model, we assume the auction price of carbon emission permits is equal to the trading price of carbon emissions rights in equilibrium state. Therefore, the expenditure of industry i of region r in the ETS market can be calculated by using the following equation: carbon expi ,r ,t  pcarbont  QAPermiti ,r ,t (66) Under the auctioning allocation scheme, all revenue from the permits auction is allocated to the central government. The transfer between the central government and local government follows the same rule under the baseline. Under the auctioning allocation scheme, the income of permits auctioning should be added to the revenue of the central government as follows. YCGr   tcgind i , r  PAi , r  QAi , r   tmi  QM i , r  PMW i , r  EXR  thcr  YH r i i tecr   WK i ,r  QKDi ,r  TRANSRTC r   pcarbont  QAPermiti ,r ,t (47’) i r i 25      Meanwhile, the first order condition of the fifth tier of the production function should be adjusted as: 1  ief （1  pca rbon  coef coal )  PEcoali , r  ief  QEoilgasi .r  ,r  ,r   (16’) PEoilgasi , r (1   i , r )  ef  QEco ali , r   The price relationship of the input of the aggregation of fossil fuels should be adjusted as: PEFi ,r  QEFi ,r  （ 1  pcarbont  coef coal )  PEcoali ,r  QEcoali ,r  PEoilgasi ,r  QEoilgasi ,r (17’) Similarly, the first order condition of the sixth tier of the production function should be adjusted as: 1 ipg （1  pcarbont  coef gas )  PEgasi ,r  i ,pg  QEoili ,r  ,r  r   (19’) （1  pcarbont  coef oil )  PEoili ,r (1   i ,r )  pg   QEgasi ,r  The price relationship of the input of the aggregation of refined petroleum and gas should be adjusted as: PEoilgasi,r  QEoilgasi,r  （ 1 pcarbont  coefoil )  PEoili,r  QEoili,r  （ 1 pcarbont  coefgas )  PEgasi,r  QEgasi,r (20’) 2.6 Market clearing 2.6.1 Commodity market clearing For the non-energy commodities as intermediate inputs, we have: QQi,r  icai, j,r  QINTAj,r +QHi,r  QCGi,r  QLGi,r  QLCi,r  QINVi,r  STOCKi,r , i  NE (67) j where QINVi ,r is the demands for commodities i used as investment, QSTOCK i ,r is the demands for commodity i used as stock. For energy commodities, we have： 26      QQi ,r   QE j ,i ,r +QH i ,r  QCGi ,r  QLGi ,r  QLCi ,r  QINVi ,r  STOCK i ,r , iE (68) j Where QE j ,i ,r is the different energy commodity inputs of every industry i, here j refers to four kinds of energy commodities inputs of industry j, i.e., QEE j , QEcoal j , QEoil j , QEgas j . 2.6.2 Labor flow and labor market clearing For labor market clearing, this model assumes labor flow among regions is not completely free and inter-regional and inter-industry wage differences are the main factors of labor flow. This model first sets an exogenous total labor force in the whole country, equal to the sum of the labor force in each region, which is the sum of labor force in the region’s various industries. The incomplete free flow of labor among different regions and industries is resulted from the existing wage differences (represented by wage distortion coefficient). Average wage in each region is equal to national average wage times the wage distortion coefficient in that region. Average wage in each sector in the same region is equal to the regional average wage times the wage distortion coefficient in that sector. TTQL   TQLr (67) r WAGEr  distortr  TWAGE (68) TQLr   QLDi ,r (69) i WLi ,r  distli ,r WAGEr (70) where TTQL is total national labor supply, TQLr is total labor supply in region r, TWAGE is national average wage, WAGEr is average wage in region r, distortr is wage distortion coefficient in region r, distli ,r is wage distortion coefficient of sector i in region r. 27      2.6.3 Investment and saving balance Total investment is decided by total savings： TOTINV + PQi ,r  QSTOCK i ,r    HSAVr  ESAVr +CGSAVr  LGSAVr  RSAVr  EXR  FSAVr   WALRAS (71) r i r PQi ,r  QINVi , r  sha reinvi , r  TOTINV (72) where TOTINV is total investment, FSAVr is the foreign savings in region r, WALRAS is a dummy variable. shareinvi is the share of commodity i used as investment in the total investment. 2.6.4 Income and expenditure balance of other regions The difference value of income and expenditure of other regions is the savings of other regions. RSAVr =  PRLRi ,s,r QRLRi , s ,r   PRLCi ,r , s QRLCi ,r , s （73） sr i sr i 2.6.5 Foreign income and expenditure balance The difference value of income and expenditure of foreign countries is foreign savings.  PWM i i ,r  QM i ,r   PWEi ,r  QEi , r  FSAVr  EXR i (74) 2.6.6 Macroeconomic closure The “neoclassical closure” rule is adopted in this model. In this model, all the savings are transformed into investment, and the total investment equals total savings endogenously, all factors are fully employed in the whole economy. The nominal GDP can be calculated from the following equation: GDPVAr   PQi,r  (QHi,r  QCGi,r  QLGi,r  QINVi,r  QSTOCK i ,r )    PRLCi ,r ,s  QRLCi,r ,s i i s              PRLR i s i ,s ,r  QRLRi , s ,r   PEi , r  QEi , r   PM i , r  QM i , r    i i 28      +  tmi  QM i , r  PMW i , r  EXR                 (75) i where GDPVAr is the nominal GDP of region r. The real GDP can be calculated as follows: i  GDPr   QHi ,r  QCGi ,r +QLGi ,r  QINVi ,r  QSTOCK i ,r + QRLCi ,r ,s  s -  QRLRi , s ,r + QEi ,r   QM i , r +  tmi  QM i ,r  PMW i ,r  EXR      (76)  s i i i where GDPr  is the real GDP of region r. Therefore, the GDP index can be obtained by the following equation: GDPr PGDPr  (77) GDPVAr where GDPr  is the real GDP of region r. Meanwhile, the CPI can be obtained as follows:  PQ i,r  QH 0i,r CPI = r i (78)  PQ0 r i i,r  QH 0i,r Where CPI is the consumer price index. In the model, household welfare variation is measured by using the Hicksian equivalent variation (EV). 2.7 Model dynamics The model is made dynamic through the population growth rate (i.e., labor supply growth rate) and investment. Demand for the total capital of the current time period is determined by the previous time period’s capital stock, depreciation and new-added investment. In this model, the new capital stock resulting from previous investment is allocated across sectors in proportion to each sector’s share in aggregate capital income, and these proportions are adjusted by the ratio of each sector’s profit rate to the average profit rate for the whole 29      economy. This is similar to the method used by James Thurlow (2004). In addition, Autonomous Energy Efficiency Improvement (AEEI) in the CGE model is considered in this study, and is assumed to be 1% per year following the common assumptions in CGE models (Wing, I. S., 2008).  Since the available social accounting matrix (SAM) is for 2012, our base year is 2012. If the model has to adopt the projected growth rate of GDP (e.g., projected by the government), it is done through adjustments in total factor productivities (TFPs). 2.7.1 Labor supply growth Labor supply in different time periods is described as: TTQL t 1  (1  lgow)TTQL t (79) where lgow is the annual growth rate of labor supply. 2.7.2 TFP increase In the model, TFP increase is represented by the change of the technology parameter in the second tier of the CES production function. αiva t1  (1  tgrow)  αiva t (80) Where tgrow is annual growth rate of TFP. 2.7.3 Capital Accumulation We adopted the method used by James Thurlow (2004) to describe the capital accumulation in different time periods. In the model, total capital supply is endogenous in a given time period and the total available capital is determined by the previous time period’s capital stock and new investment. In this model, the new capital stock resulting from previous investment is allocated across sectors in proportion to each sector’s share in aggregate capital income, and these proportions are adjusted by the ratio of each sector’s profit rate to the average profit rate for the whole economy. 30      QKDi , r ,t WKAr ,t    WK i ,r ,t (81) i  QKD i i , r ,t Where WKAr ,t is average capital rental rate of region r at time period t. QKDi ,r ,t  WKi ,r ,t  sharenki ,r ,t   1  i ,r  (  1)  (82)  QKD i i , r ,t   WKAr ,t   Where sharenki ,r ,t is the share of the new capital investment of industry i at time time period t, i ,r is the inter-sectoral mobility coefficient of investment in region r. The value of i ,r can be chosen from 0 to 1, i ,r is 0 means there is no inter-sectoral mobility of investment, whereas i ,r is 1 means there is full inter-sectoral mobility of investment. QINVi , r ,t PK r ,t   PQi , r .t  (83) i  QINV i i , r ,t Where PK r ,t is the price of capital in region r at time period t.  PQ i , r ,t  QINVi ,r ,t QINDi ,r ,t  sharenki ,r ,t  i (84) PK r ,t Where QINDi ,r ,t is the new-added capital of industry i in region r at time time period t. TOTINDt    PK r .t  QINDi , r ,t r i Where TOTINDt is the total new-added capital at time period t. QINDi ,r,t TQKAi ,r,t 1  TQKAi ,r,t  (1   depri ,r ) (85) TQKAi ,r,t Where TQKAi ,r,t 1 is the capital stock of industry i in region r at time period t+1, depri , r is the depreciation rate of industry i in region r. 31      TQKt 1   (TQKAi ,r ,t  (1  depri ,r )  QINDi ,r ,t ) (86) r i Where TQK t 1 is the new aggregate quantity of capital at time period t+1. 2.8 Using information from a bottom-up model to correct the baseline Economic impacts of a policy (ETS here) are measured by comparing the economic variables under the policy with the corresponding variables in the baseline. Thus, the magnitudes of economic impacts are sensitive to the baseline. Timilsina et al. (2019) finds that economic impacts of meeting China’ NDC estimated by a CGE model drops by two-thirds if the CGE model is linked with a bottom-up engineering model, TIMES, that is capable of representing detailed technologies. The same could be true for the results estimated by our regional CGE model developed for this study. However, the TIMES model used in Timilsina et al. (2019) is a national model, and developing a regional TIMES model for this study is constrained by lack of data. Therefore, we used the baseline correction factor estimated by Timilsina et al. (2019) in this study. Doing so could compromise the preciseness of the baseline correction, but the correction factor is unlikely to change significantly across the provinces. This is because the physical characteristics of a technology used in a production process (e.g., arc furnace for steel making, gas turbine for power generation) are unlikely to change across the provinces. Even if they change because of different mix of technologies in different provinces for an industry, the tremendous efforts needed to collect the data would not allow to estimate the baseline correction factors by province. When we apply the baseline correction factor, the equations in Section 2.5 above will get modified to reflect the smaller gap between the revised baseline and the NDC targets. The correction of the baseline is equivalent to using the autonomous energy efficiency index (AEEI), which is a common practice in studies that employ dynamic CGE models as the main analytical tools. The advantage here is that the correction factor is estimated through a rigorous process of linking a CGE model with a bottom-up TIMES model for China (see 32      Timilsina et al. 2019). In the existing CGE models, the AEEI are selected arbitrarily to adjust the baseline to satisfy the modeler’s judgement or policy makers’ expectation. 3. Data and Parameters In this paper, the benchmark year for the CGE model is 2012, and 31 SAMs of 31 provinces that are suitable to the structure of the multi-regional CGE model have been constructed as the database. Firstly, we constructed a 2012 multi-regional input-output (MRIO) table covering 31 provinces of China based on the published 2012 input-output tables of the 31 provinces (NBS,2016). Secondly, 31 macro SAMs of the 31 provinces were compiled to provide a consistent macroeconomic data framework. Thirdly, based on the 31 macro SAMs, 31 disaggregated SAMs are developed for the multi-regional CGE model. The provincial SAMs are available at World Bank’s Development data hive4. In addition to the SAM, there are two types of parameters that should be estimated before the CGE model can be used for policy analysis: one is a share parameter, such as consumer expenditure share, and the other is an elasticity parameter, such as the elasticity of substitution in the production function. The share parameters can be estimated from the SAM using a calibration method that assumes the benchmark year represented by the SAM is an equilibrium solution for the CGE model, but the elasticity parameters need to be obtained based on the relevant literatures (Bao, Q., et al, 2013; Zhai Fan and Hertel, T.W, 2006; Zhang, Z., 1998). The CO2 emission factors from fossil fuels are calculated based on IEA statistical data (2014). The BAU scenario is constructed using assumptions on labor supply growth and projected real GDP growth, as shown in Table 2. Table 2 The Labor supply growth rate and GDP growth rate from 2012 to 2030 Year Labor supply growth rate Average annual GDP growth rate 2012-2015 2.70% 8.51% 2015-2020 4.60% 6.56% 2020-2025 0 5.49% 2025-2030 0 5.47% Note: Values for 2012-2015 are actual or observed values. The projected GDP is set at a moderate level based on a review of relevant literature(Li Shan Tong, 2010; Johansson A., et al.,2013; OECD,2012) .Labor supply                                                              4 https://datacatalog.worldbank.org/ 33      growth rate are taken from China Statistic Year book 2017 and UNDESA(2015) with very small modification based on our understanding about labor quality development tendency of China in the future. 4. Results from Model Simulations In this section, we discuss the allocation of initial allowances, and results of our CGE model on the impacts of ETS to meet China’s NDC. Although we present the detailed results in several tables, the focus of discussions would be on intuition and implications of the results rather than quantitative information provided by the results. 4.1 Emission Allowances, Trade and Reduction 4.1.1 Allocation of allowances There are many possibilities for allocating the initial allowances across the sectors or emitters, including free allowances based on historical emissions, outputs, emission intensity, or auctioning of emission allowances (Cheng, 2018; Pang and Duan, 2016; Wu et al. 2016). For example, Wu et al. (2016) consider five alternative approaches (GDP based, population- based, emission based, based on ability to pay and based on government’s five-year plan) to freely allocate national emissions to provinces. In this paper, we do not allocate the emission allowances to the provinces. Instead we directly allocate allowances to emitters (here production sectors) in each province. There are 496 emitters (16 sectors in 31 provinces) participating in the national ETS. We considered two approaches for allocation or distribution of allowances: free allocation (or grandfathering) and auction. Under the free allocation, we considered two criteria for the allocation of the national emission allowances in a given year based on: (i) their baseline emissions, (ii) their baseline outputs. Two cases are also considered under auctioning based on schemes used to recycle the auction revenues to the economy. These are: (i) use of the auction revenue by the government for public consumption or public investment, and (ii) recycling the revenue to households as a lump-sum rebate. The national emission allowance in 2030 is set in a way that it meets China’s NDC, meaning that China’s emission intensity (CO2 emission divided by GDP) would be 65% below its 2005 level. CO2 emissions and GDP are projected by the CGE model in the baseline. The 34      total allowances in 2030 are estimated to be 8,413 million (one allowance is equal to one ton of CO2 emissions). The total number of emission allowances does not change between the with and without linking the CGE model to information obtained from the bottom-up TIMES model. The information from the TIMES model reduces the baseline emissions and the gap between the emissions under the baseline and the NDC. Under the ETS, a total number of allowances in 2030 would be 70.3% of the baseline emissions in 2030 if the information obtained from the bottom-up model is not taken into account. This means there would be 28.8% reduction of national emission from the baseline under the ETS in 2030 under the no-linking case. Since the linking of CGE with the information from the bottom-up model reduces 15.5% of baseline emissions, reduction of emissions under the linked case would be 13.3% (28.8% - 15.5%) in 2030 under the linked case. The targeted level of allowed emissions under the NDC in 2030 remains the same under both cases. We assumed that the NDC target would be met by reducing emissions from the production sectors only. Emissions from the final consumption sectors (i.e., households, governments, capital goods) will not be capped, and they do not participate in the national ETS. However, these sectors also experience emission reductions due to indirect effects (because of increased consumer prices under the ETS). It also excludes emissions from private transportation because it is accounted for under the household sector in the CGE model, and the household sector does not participate in the national ETS. Figure 3a and 3b present the allocation of allowances across the sectors and also across the provinces, respectively. Emission allowance received by a sector is highly sensitive to the criteria used for the allocation. If the emission is used as an allocation criterion, sectors with higher CO2 emissions, such as electricity, non-metallic minerals, metals, coal mining, would receive the relatively higher number of emission allowances. On the other hand, if the sectoral outputs are used as an allocation criterion, sectors with larger outputs, such as the service sector, the construction sector, the other manufacturing sector receive a relatively higher number of emission allowances. While the emission based allocation rule is preferable to emission- intensive sectors, the output-based allocation rule is preferable to less emission-intensive sectors. Higher allowance received means a higher opportunity to benefit from the emission trading. It is, therefore, the government’s priority how the emission allowances are distributed. 35      If the economy is based on fossil fuel industry and the government prefers not to penalize too much the fossil fuel industries, it may employ the emission-based allocation rule. On the other hand, if the government wants to reward sectors that contribute to the economy with a lower carbon footprint, it might prefer the output-based allocation rule. It is also possible to combine both approaches by using emission intensity (emission divided by sectoral output) as an allocation criterion. The allocations of emission allowances to a province depend on the total emissions generated from its economic activities. Provinces like Shandong get highest allowances under either criterion (emission based or output based) because of the size of the economy in terms of CO2 emissions and economic outputs. That is true for Jiangsu, Hebei, Guangdong and Beijing. If a province has a large number of industries that produce a relatively higher amount of emissions such as coal-fired power plants, cement plants, it gets more allowances under the emission-based allocation (see for example, Hebei, Inner Mongolia, Shanxi and Henan). On the other hand, provinces having more output producing industries with relatively lower emission intensities would get a higher number of allowances under the output-based allocation criterion. Of the 31 provinces, 13 provinces will get a higher number of emission allowances under the emission based free allocation rule, whereas 15 provinces will get a higher number of emission allowances under the output based free allocation of allowances. Four provinces will get about the same number of emission allowances no matter whether the allocation rule is based on emission or outputs. Note that the free allocation of emission allowances is independent of whether or not the CGE model is linked with the information from the bottom- up TIMES model. 36      Figure 3. Emission allowances under grandfathering to meet China’s NDC in 2030 (a) Free allocation of emission allowances across the sectors in 2030 (Million) SERVI TRANS CONST WATER GAS ELECT OTHMF METAL NMETA CHEMI PETRO FTPMF MINE OILNG COAL AGRI 0 300 600 900 1200 1500 1800 2100 FA-Output FA-Emission                 37        (b) Free allocation of emission allowances across the provinces in 2030 (Million) Zhejiang Yunnan Xizang Xinjiang Tianjin Sichuan Shanxi Shanghai Shandong Shaanxi Qinghai Ningxia Inner Mongolia Liaoning Jilin Jiangxi Jiangsu Hunan Hubei Henan Heilongjiang Hebei Hainan Guizhou Guangxi Guangdong Gansu Fujian Chongqing Beijing Anhui 0 100 200 300 400 500 600 700 800 900 1000 1100 FA-Output FA-Emission If the emission allowances are distributed through auctioning, the sectors which have relatively lower marginal costs of reducing their CO2 emissions are expected to get (i.e., purchase) a higher number of allowances. Figure 4 illustrates the distribution of allowances across the sectors and provinces in 2030 under the auctioning rule. Note that in our study it is not a province rather sectors operating under its jurisdiction that participate under the auctioning. It is also possible that the central government would allocate allowances to provinces using some criteria and allow the provincial governments to do the auctioning. 38      However, the rule to allocate across the provinces is not necessarily the most efficient one. Moreover, the same firm (e.g., cement company or steel company) might have plants in different provinces. Therefore, it would be more efficient to have a national auction to allocate emission allowances across the emitters. Moreover, allowances are normally distributed for a period rather than for a single year so that emitters have flexibilities to meet the targets at the end of the period rather than meeting each year. As can be seen from right hand panel of Figure 4, the electricity sector purchases the highest amounts of emission allowances under auctioning followed by other emission intensive sectors, non-metallic minerals, metal sector, petroleum refinery sector and the coal sector. These sectors need more allowances to keep their business operating. If they do not buy higher numbers of allowances they have to curtail their operations (i.e., reduce their outputs), which would be more expensive to the economy. Moreover, these sectors have relatively lower marginal costs of reducing their emissions. Which province purchases the higher number of allowances depends on how many sectors with low marginal costs of CO2 abatement it has. 39      Figure 4. Distribution of emission allowances under auctioning to meet China’s NDC in 2030 Shandong 816 ELECT 2,011 Jiangsu 770 Hebei 652 NMETA 1,307 Henan 540 Guangdong 465 METAL 1,172 Inner Mongolia 418 Zhejiang 401 PETRO 658 Hunan 354 Anhui 330 COAL 655 Sichuan 320 Liaoning 309 CHEMI 651 Shanxi 284 Shaanxi 267 CONST 513 Shanghai 267 Beijing 228 TRANS 455 Heilongjiang 211 Tianjin 202 SERVI 293 Hubei 197 Jiangxi 192 OTHMF 211 Jilin 190 Guizhou 168 FTPMF 172 Yunnan 125 Guangxi 119 MINE 144 Fujian 117 Xinjiang GAS 107 108 Chongqing 102 AGRI 55 Gansu 101 Ningxia 83 OILNG 7 Qinghai 49 Hainan 26 WATER 2 Xizang 1 When emission allowances are distributed through auctioning, the central government first decides the total or national cap corresponding to the NDC target. A sector or emitter buys emission allowances from the government in a similar manner as if it is ready to pay a carbon tax to meet the national CO2 mitigation target. The underlying assumption here is that the auction price and permit price are equal, meaning that if an emitter purchases a higher amount of permits than it needs it can sell in the market at the same price it purchased. While this approach provides incentives to an emitter which has a lower marginal cost of its reduction, it may not be easy to implement in practice because the auction price and permit price may be different thereby creating a possibility of arbitrage. 40      4.1.2 Trading of Emissions Allowances The volume of emissions that an emitter (here sector) trades depends on the level of its initial allowance and the amounts of emissions it would like a release to keep its desired level of outputs. Since the level of emission allowances a sector receives depends on the allocation rule, the volume of trade is highly sensitive to the rule for allowance allocation. Since the amount of emission trading is also sensitive to the gap between the baseline emissions and emissions under the ETS (i.e., desired level of emission reduction of an emitter), baseline setting is a critical determinant of the volume of traded emissions and market clearing emission permit prices. Therefore, the volume of emissions traded and prices of emission permits are sensitive whether or not the CGE model is linked with the information from the bottom-up model, TIMES. Note that the TIMES model significantly dampens the baseline (it reduces baseline emissions by 15.5% in 2030). Table 3(a) and (b) presents the volume of emission trading by sector and by provinces. As illustrated in these tables, there are huge differences in the volume of emission traded and prices of emission permits between the allowance allocation rules. If the CGE and TIMES models are not linked, the total volume of permits traded under the output-based allocation rule is seven times as high as that under the emission-based allocation rule. This is because the output-based rule offers much more allowances to the sectors which do not need (they produce more output not necessarily the emissions), so they have a large number of allowances to sell. On the other hand, sectors that produce more emissions receive a lower number of allowances, so they need to buy a large number of allowances. Thus, the size of the emission market is much larger under the output-based allocation rule as compared to the emission-based allocation rule. Larger the market would mean more flexibility for trade, thereby lowering the prices of permits. The market clearing permit price under the output-based allocation rule would be (¥406/tCO2). On the other hand, the total supply of emission allowances under the emission-based allocation rule are much smaller as compared to that under the output-based allocation rule, the market clearing prices of emission allowances would be very high, ¥706/tCO2. When we linked the CGE and TIMES model the total (national level), baseline emissions would be smaller, and so does the amounts of emissions to be reduced to meet the NDC targets. 41      Under the linked case, the total volume of emission trade would be almost half of that under the not-linked case if allowances are allocated based on emissions. The price of permits drops by 63% from ¥706/tCO2 under the unlinked case to ¥263/tCO2. This substantial drop is caused by the fact the average marginal abatement cost curve is quadric in nature in China; if emission mitigation drops by a unit, the price drops by its square. This is explained in Timilsina et al. (2019). When allowances are allocated based on outputs, permit prices under the linked case would be by 40% (¥164/tCO2) of the unlinked case (¥406/tCO2). Table 3. Trading of emission permits under ETS in 2030 (a) By Sector Sector Emission trading under when allowances are freely allocated through the grandfathering rules Emission based Output-based No-Linkage Linkage No-Linkage Linkage AGRI 5.2 2.4 -120.8 -125.5 COAL -189.5 -97.2 516.6 646.3 OILNG 0.7 0.4 -42.3 -42.9 MINE 12.5 5.8 -38.4 -47.8 FTPMF -4.2 -1.2 -375.3 -380.9 PETRO -3.1 1.6 490.8 519.4 CHEMI 3.5 3.0 198.6 180.5 NMETA -16.0 -6.1 801.5 784.0 METAL 15.1 9.1 273.9 229.9 OTHMF 7.1 3.7 -1,430.4 -1,447.1 ELECT 37.2 17.8 1,722.4 1,785.1 GAS -9.7 -4.3 87.7 99.6 WATER 0.1 0.1 -10.5 -10.7 CONST 25.1 12.2 -1,157.1 -1,193.1 TRANS 81.1 36.1 156.0 105.6 SERVI 34.9 16.5 -1,072.6 -1,102.5 Note: +ve sign refers to purchasing of allowances and whereas -ve sign refers to the selling of allowances. Total trade is the sum of permits sold and bought at the national level. (b) By Province Province Emission trading under when allowances are freely allocated through the grandfathering rules Emission based Output-based No-Linkage Linkage No-Linkage Linkage Anhui 8.8 6.1 88.9 94.2 Beijing 10.9 5.4 -136.0 -138.8 Chongqing 9.1 4.7 -96.5 -105.9 Fujian -1.0 -1.2 -49.6 -54.2 Gansu 2.2 1.3 27.6 25.4 Guangdong -20.6 -13.9 -29.9 -17.6 Guangxi 5.8 3.0 -30.0 -38.3 42      Guizhou -8.9 -4.8 92.9 108.7 Hainan 2.5 1.2 -19.4 -22.2 Hebei -9.8 -4.1 213.3 223.0 Heilongjiang 0.2 0.1 40.2 41.7 Henan -7.4 -6.7 141.3 169.1 Hubei -10.1 -6.2 -13.0 -10.5 Hunan 3.3 3.0 95.2 103.5 Jiangsu 33.7 19.9 -203.4 -238.2 Jiangxi 12.4 7.3 -67.3 -82.6 Jilin 6.6 3.9 -22.1 -28.0 Liaoning 15.7 7.1 -97.5 -118.6 Inner Mongolia -64.4 -36.4 171.0 219.7 Ningxia 0.3 0.2 41.1 43.0 Qinghai 2.7 1.6 -31.9 -44.7 Shaanxi -3.5 0.0 57.3 52.1 Shandong 18.4 12.3 -271.7 -297.8 Shanghai 16.2 7.8 -62.8 -70.1 Shanxi -28.2 -16.1 155.3 195.8 Sichuan -7.0 -2.7 60.9 59.9 Tianjin -9.7 -4.9 -65.1 -58.1 Xinjiang 8.4 4.4 34.6 27.7 Xizang 0.2 0.1 -6.8 -6.9 Yunnan 5.3 3.1 -6.7 -11.4 Zhejiang 8.0 4.5 -10.0 -20.0 Note: +ve sign refers to purchasing of allowances and whereas -ve sign refers to the selling of allowances. Total trade is the sum of permits sold and bought at the national level. Under the auctioning, emitters have to purchase allowances, through auctioning, from the government to cover every single ton of CO2 they emit. Governments use revenues from the allowance auction for various purposes, like in the case of carbon tax revenue. In our study, we considered two schemes of revenue recycling for demonstration. In the first scheme, auction revenue is utilized by the provincial governments for public consumption and public investment. As expected, allowance prices under auctioning would be smallest. Under the linked case, it would be ¥374/tCO2. It decreases by 60% to ¥149/tCO2 under the linked case. Please note that the size of the emission allowance price is determined the size of the emission gap between the baseline and the NDC target. Since the gap reduces under the linked case, so does the market clearing allowance price. These results indicate the efficiency of auctioning over the free allocation of emission allowances based grandfathering approaches. The volumes of emission trade by the province under the grandfathering are provided in table 3b. The direction and magnitude of emissions trade of a province are highly sensitive 43      to the rule of allocation of an emission allowance. The direction of emission trade can be divided into four groups. The first group of provinces is those where the direction of emission trade remains the same under both allocation rules. Nine provinces fall under this group. Five provinces (Anhui, Heilongjiang, Hunan, Ningxia, and Xinjiang) would be the buyers of allowances, whereas other four provinces (Fujian, Guangdong, Hubei and Tianjin) would be the seller of allowances irrespective of allocation rules. The second group of provinces is those which switchover their positions when allocation rule is changed. Eighteen provinces fall under this group. Eight provinces (Beijing, Chongqing, Guangxi, Hebei, Henan, Inner Mongolia, Shanxi, Sichuan) which would be on seller’s position under the emission-based allocation rule would turn into buyer’s position when allowance allocation rule gets changed to output-based. The opposite happens to 10 provinces (Hainan, Jiangsu, Jiangxi, Jilin, Liaoning, Qinghai, Shandong, Shanghai, Xizang and Zhejiang). These results indicate the importance of selecting the rule for allowance allocation. While an allocation rule would be good at the national level does not mean that it would be better for a province. At the national level, the total value of emission trade in 2030 ranges from 51 billion yuan to 3,155 billion yuan depending on the allowance allocation or distribution rules and also depending on baseline emissions (see Table 4). Table 4. Volume and value of allowance trade and allowance price in 2030 Free Allowances (Grandfathering) Auctioning Emission- based Output-based Not linked Linked Not linked Linked Not linked Linked Total volume of trade 341 194 2,439 2,728 8,437 8,502 (million) Allowance price (¥/tCO2) 706 263 406 164 374 149 Total value of trade 241 51 990 447 3,155 1,267 (Billion Yuan) 4.1.3 Emissions Reductions Meeting China’s NDC through ETS would result in about 29% reduction of national CO2 emissions in 2030 from the baseline. The levels of emission reductions vary significantly across the sectors as shown in Table 5a, and across the provinces as shown in Table 5b. 44      Auctioning causes larger reduction of emissions than grandfathering in fossil fuels and electricity generation sectors. The reverse is true in the remaining sectors. For example, auctioning causes higher level of emission reduction than does the grandfathering in 10 provinces; the reverse is true in the rest. Five provinces (Shanxi, Inner Mongolia, Guizhou, Tianjin, Guangdong) would experience more than 30% of their CO2 reduction in 2030 from the baseline under both schemes of allowances allocation. This is because the economy in these provinces relies on coal mining or coal-intensive industries; meeting ETS means a reduction of more emissions from the coal sector and coal-intensive sectors. When the CGE model uses a lower baseline consistent with the bottom-up TIMES model, emission reductions in all provinces drops by 50% under both free allocation and auctioning rules for allocation of emission allowances. The lower requirement of emission reduction from the baseline when the baseline itself shrinks would mean a lower level of market clearing allowance price. Table 5. Emission reductions under ETS in 2030 (a) By Sector Sector Free Allowances (Grandfathering) Auctioning Without linkage With linkage Without linkage With linkage Emission Output- Emission Output- Revenue Revenue Revenue Revenue based based based based GOV HH GOV HH AGRI -20.8 -17.3 -8.9 -7.5 -16.7 -16.3 -7.2 -7.1 COAL -43.7 -47.3 -22.5 -24.2 -47.8 -48.1 -24.3 -24.4 OILNG -21.0 -20.4 -8.0 -8.5 -23.0 -22.6 -9.8 -9.7 MINE -21.8 -19.2 -9.4 -8.4 -20.3 -19.9 -8.9 -8.7 FTPMF -30.5 -25.0 -13.9 -11.3 -24.4 -24.3 -11.0 -11.0 PETRO -29.0 -32.8 -13.1 -15.5 -32.5 -32.7 -15.3 -15.4 CHEMI -28.3 -24.2 -12.9 -11.0 -24.8 -24.6 -11.3 -11.2 NMETA -29.6 -25.8 -13.7 -11.8 -26.5 -26.3 -12.1 -12.0 METAL -27.7 -22.9 -12.6 -10.4 -23.8 -23.5 -10.8 -10.6 OTHMF -26.1 -21.6 -11.6 -9.7 -22.3 -22.0 -10.0 -9.8 ELECT -27.4 -31.1 -12.6 -14.5 -31.0 -31.0 -14.4 -14.4 GAS -34.1 -40.4 -16.2 -20.2 -40.0 -39.9 -19.9 -19.9 WATER -22.2 -19.8 -9.8 -8.9 -19.8 -19.4 -8.9 -8.7 CONST -24.8 -19.8 -11.0 -8.7 -20.4 -20.0 -9.0 -8.8 TRANS -13.1 -11.3 -5.0 -4.6 -11.7 -12.3 -4.8 -5.1 SERVI -18.7 -15.8 -7.6 -6.7 -14.2 -16.1 -6.0 -6.9 45      (b) By Province Province Free Allowances (Grandfathering) Auctioning Without linkage With linkage Without linkage With linkage Emissio Output Emissio Output Revenue Revenue Revenue Revenue n based -based n based -based GOV HH GOV HH Anhui -26.8 -29.2 -11.7 -13.2 -29.8 -29.7 -13.5 -13.5 Beijing -25.5 -27.3 -11.4 -12.9 -32.3 -32.4 -15.4 -15.4 Chongqing -21.1 -14.7 -8.6 -6.0 -14.9 -14.4 -6.0 -5.8 Fujian -29.3 -23.5 -14.2 -10.9 -23.3 -23.1 -10.8 -10.7 Gansu -27.1 -24.7 -12.2 -11.1 -24.5 -24.3 -11.0 -10.9 Guangdong -31.7 -32.2 -15.7 -15.9 -32.0 -32.1 -15.7 -15.8 Guangxi -24.7 -18.1 -10.8 -7.7 -18.2 -17.9 -7.8 -7.6 Guizhou -32.0 -38.6 -15.4 -18.9 -37.5 -37.5 -18.2 -18.2 Hainan -20.4 -12.5 -8.6 -5.1 -13.6 -13.3 -5.7 -5.6 Hebei -29.8 -29.6 -13.8 -13.8 -29.7 -29.8 -13.9 -13.9 Heilongjiang -28.6 -29.0 -13.3 -13.6 -28.2 -28.3 -13.1 -13.2 Henan -29.6 -34.8 -14.3 -17.2 -33.9 -33.8 -16.6 -16.5 Hubei -32.3 -29.7 -16.0 -14.0 -28.5 -28.8 -13.3 -13.4 Hunan -28.0 -30.5 -12.6 -14.1 -29.6 -29.5 -13.6 -13.5 Jiangsu -25.3 -21.6 -10.9 -9.3 -21.6 -21.6 -9.3 -9.3 Jiangxi -23.4 -16.4 -9.5 -6.6 -18.5 -18.1 -7.6 -7.4 Jilin -26.1 -23.9 -11.4 -10.8 -23.7 -23.5 -10.6 -10.5 Liaoning -24.6 -19.1 -11.1 -8.6 -20.0 -19.7 -9.1 -8.9 Inner Mongolia -37.8 -42.0 -19.5 -21.8 -41.0 -40.9 -20.9 -21.0 Ningxia -28.4 -30.6 -13.1 -14.4 -30.7 -30.6 -14.4 -14.3 Qinghai -24.2 12.7 -10.0 11.0 -16.7 -17.9 -4.1 -4.6 Shaanxi -29.7 -24.5 -13.3 -10.6 -24.1 -24.0 -10.4 -10.3 Shandong -27.0 -23.8 -12.0 -10.5 -25.7 -25.7 -11.6 -11.5 Shanghai -24.2 -24.4 -10.7 -10.9 -25.6 -25.9 -11.6 -11.7 Shanxi -34.4 -43.4 -17.2 -21.8 -41.8 -42.1 -20.7 -20.9 Sichuan -30.3 -27.3 -14.1 -12.4 -26.5 -26.4 -12.0 -12.0 Tianjin -31.8 -32.9 -15.2 -16.1 -34.6 -34.7 -17.0 -17.1 Xinjiang -22.4 -18.2 -9.3 -7.5 -17.9 -17.9 -7.5 -7.4 Xizang -15.8 -9.7 -6.9 -4.2 -11.0 -10.4 -4.8 -4.5 Yunnan -25.4 -22.5 -11.0 -9.8 -22.4 -22.2 -9.8 -9.7 Zhejiang -27.2 -24.6 -12.3 -11.1 -24.6 -24.6 -11.1 -11.1 China -28.7 -28.2 -13.3 -13.2 -28.5 -28.5 -13.3 -13.3 The national level emission reductions are not much sensitive across the allocations rules by design because the emission caps and corresponding emission reductions are fixed or exogenous. Provincial level emission reductions vary across the allowance allocation rules. The percentage emission reductions from a province are not different significantly between the revenue recycling schemes under the auctioning. 46      4.2 Economic Impacts of ETS 4.2.1. Impacts on Economic Outputs (GDP) Table 6(a) and 6(b) presents the impacts of ETS on GDP. The impacts vary widely depending on the baseline projection (i.e., whether or not information from the bottom-up TIMES model is incorporated in the CGE model) and the allowance allocation rules. While the emission trading would cause a slight loss of GDP at the national level in 2030 from the baseline (0.9% to 1.5% in the absence of linkage and 0.3% to 0.5% with linkage), there are large variations on the GDP loss across the provinces. Note that ETS does not cause economic losses to all provinces; it increases economic outputs in some provinces. Whether a province gains or losses its economic outputs depend on the allowance allocation rule as different allocation rules affect the provincial economies, their energy supply structures, and emissions differently. When allowances are freely allocated based on baseline emissions, six provinces (Fujian, Guangdong, Guangxi, Hebei, Hubei and Liaoning) would experience gains in their GDP. If allowances are freely allocated based on baseline outputs, 10 provinces (Beijing, Chongqing, Fujian, Guangxi, Hainan, Jiangsu, Liaoning, Qinghai, Xizang and Yunnan) will experience gains in the GDP. The same happens in 11 provinces (Chongqing, Fujian, Guangdong, Guangxi, Hainan, Hubei, Jiangsu, Liaoning, Xizang, Yunnan and Zhejiang) when allowances are distributed through auctioning. Four provinces (Inner Mongolia, Ningxia, Shanxi and Shaanxi) suffer the most due to the ETS irrespective of emission allocation rules because of their reliance on coal mining and other coal-based industries. However, the magnitude of their economic loss decreases substantially if their baselines emissions decrease (linked case). Therefore, it should not be concluded that the economic losses in these four industries would be as high as indicated by our model results under the unlinked case. Nevertheless, policy makers should consider compensation measures in these economies as any sort of climate change mitigation policies (perhaps except carbon capture and storage) are expected to hit hard these economies. The rest of the provinces (Anhui, Gansu, Guizhou, Heilongjiang, Hunan, Jiangxi, Jilin, Shandong, Shanghai, Sichuan, Tianjin and Xinjiang) also face a varying degree of economic losses under all schemes of allowances allocation 47      considered. While the sale of emission allowances could help to some extent to increase the GDP, but other factors such as labor mobility, capital or investment mobility, and changes in consumption by households and provincial governments and changes in international and inter- provincial trades due to emission trading are more dominant to affect the GDP. Changes in final consumption, investments and trade will be discussed later.   Table 6. Impacts on GDP (by province) and gross output (by sector) due to ETS in 2030 (a) Impacts on provincial and national GDP in 2030 (% change from the baseline) Province Free Allowances (Grandfathering) Auctioning Without linkage With linkage Without linkage With linkage Emissio Output Emissio Output- Revenue Revenue Revenue Revenue n based -based n based based GOV HH GOV HH Anhui -0.7 -3.1 0.1 -1.1 -3.3 -3.1 -1.2 -1.1 Beijing -7.3 0.8 -3.5 -0.3 -2.8 -2.8 -1.9 -1.9 Chongqing -0.9 1.2 -0.3 0.5 1.2 1.9 0.5 0.7 Fujian 0.8 1.4 0.3 0.6 1.9 2.2 0.8 0.9 Gansu -0.1 -1.2 0.1 -0.4 -1.1 -0.8 -0.3 -0.2 Guangdong 0.9 -0.3 0.5 -0.1 0.9 0.3 0.4 0.2 Guangxi 1.0 1.6 0.5 0.7 1.9 2.5 0.8 1.0 Guizhou -1.6 -7.1 -0.5 -3.0 -6.0 -5.8 -2.5 -2.4 Hainan -3.4 1.0 -2.1 0.1 4.6 8.1 1.1 1.8 Hebei 1.0 -2.7 0.8 -0.9 -2.0 -2.3 -0.6 -0.7 Heilongjiang -2.1 -2.9 -0.6 -1.2 -1.5 -1.8 -0.5 -0.6 Henan -0.4 -1.1 -0.1 -0.5 -0.7 -0.4 -0.3 -0.1 Hubei 0.6 -0.5 0.4 -0.2 0.7 0.8 0.3 0.4 Hunan -0.6 -3.4 -0.1 -1.3 -2.5 -2.3 -0.9 -0.8 Jiangsu -1.1 0.2 -0.3 0.1 1.5 0.5 0.6 0.3 Jiangxi -2.8 0.0 -0.9 0.0 -1.6 -1.0 -0.6 -0.3 Jilin -0.9 -0.9 -0.2 -0.3 -1.0 -0.5 -0.3 -0.1 Liaoning 1.0 2.4 0.5 1.0 1.9 2.4 0.9 1.0 Inner Mongolia -2.4 -5.9 -1.2 -2.8 -5.3 -4.3 -2.5 -2.1 Ningxia -2.3 -6.4 -0.8 -2.6 -6.5 -6.3 -2.6 -2.5 Qinghai -7.6 8.7 -3.3 5.6 -4.4 -4.8 -1.3 -1.6 Shaanxi -4.2 -4.2 -1.6 -1.6 -3.7 -3.6 -1.4 -1.3 Shandong -2.3 -1.4 -0.6 -0.5 -2.3 -2.7 -0.8 -1.0 Shanghai -0.6 -0.7 0.1 -0.2 -1.1 -1.6 -0.3 -0.6 Shanxi -4.0 -5.3 -1.4 -4.6 -6.8 -7.2 -3.4 -4.0 Sichuan -0.5 -1.4 -0.1 -0.5 -0.6 -0.5 -0.2 -0.1 Tianjin -0.7 -0.5 -0.2 -0.3 -0.5 -0.8 -0.2 -0.3 Xinjiang -2.1 -1.3 -0.6 -0.3 -1.0 -0.8 -0.2 -0.1 Xizang -0.1 1.5 -0.1 0.6 0.8 1.8 0.3 0.7 Yunnan -0.1 0.1 0.1 0.1 0.1 0.6 0.1 0.3 Zhejiang -0.2 -0.4 0.1 -0.1 0.2 -0.3 0.1 -0.1 China -1.5 -0.9 -0.5 -0.3 -1.2 -1.3 -0.4 -0.5   48      (b) Impacts on sectoral outputs in 2030 (% change from the baseline) Province Free Allowances (Grandfathering) Auctioning Without linkage With linkage Without linkage With linkage Emission Output- Emission Output- Revenue Revenue Revenue Revenue based based based based GOV HH GOV HH AGRI 0.0 -0.1 0.0 -0.1 0.3 1.0 0.1 0.4 COAL -28.6 -32.0 -13.3 -15.0 -35.5 -35.5 -16.7 -16.7 OILNG -9.9 -6.3 -3.8 -2.3 -8.4 -7.1 -3.4 -3.1 MINE -1.5 -1.7 -0.4 -0.6 -3.9 -3.0 -1.5 -1.1 FTPMF -1.3 -0.5 -0.4 -0.2 -1.3 -0.5 -0.5 -0.2 PETRO -9.7 -8.7 -3.3 -3.3 -10.9 -10.7 -4.3 -4.3 CHEMI -2.1 -2.3 -0.6 -0.9 -4.2 -3.6 -1.7 -1.4 NMETA -2.9 0.1 -1.1 0.6 -4.5 -3.8 -1.5 -1.2 METAL -1.7 -1.3 -0.5 -0.5 -3.6 -2.7 -1.4 -1.0 OTHMF -1.7 -0.5 -0.6 -0.2 -2.4 -1.6 -1.0 -0.6 ELECT -4.3 -1.6 -1.5 -0.6 -6.3 -5.7 -2.7 -2.4 GAS -17.9 -22.9 -7.7 -10.6 -23.4 -22.9 -10.7 -10.6 WATER -1.5 -2.0 -0.5 -0.8 -2.9 -2.3 -1.2 -0.9 CONST -1.1 -0.8 -0.3 -0.3 -2.3 -1.4 -0.9 -0.5 TRANS -1.8 -2.0 -0.6 -0.8 -2.1 -2.6 -0.8 -1.0 SERVI -1.2 -0.5 -0.4 -0.2 1.5 -0.5 0.6 -0.2 Since different allocation rules are favorable for different provinces as they either increase the GDP or minimize the reduction if the ETS causes a GDP loss, a single criterion for emission allocation would not be preferred. A mixed allocation scheme would be desirable. There could be many possibilities for mixing the allocation rules; finding the best mix is beyond the scope of this study. It could be part of the future extension of the study. Table 6(b) shows that sectoral outputs of each sector drop under all schemes of allocating/distributing emission allowances except the agriculture and service sector under auctioning. Auctioning and output based free allocation cause relatively higher drops of sectoral outputs of the fossil fuel industries (i.e., coal, petroleum refinery, gas processing) and emission-intensive industries (e.g., chemicals, electricity, non-metallic minerals, metals). On the other hand, auctioning and output based free allocation of allowances are favorable to less emission-intensive industries (e.g., service, agriculture). Under the auctioning, national level GDP impact is slightly worse when auction revenues are recycled to households than when it is used by the governments for public consumption and public investment. This is because the average ratio (across provinces) of government 49      savings to total government revenyes is higher than the average ratio of household savings (across provinces) to total the disposable income. 4.2.2 Impacts on Household Income due to ETS Table 7 illustrates the impact of ETS on household income under linkage cases, which helps explain the impacts on GDP discussed earlier. Emission trading causes a loss in household income in most provinces because meeting NDC causes 10 to 40 percent of emission reduction from the baseline. These levels of emission reduction cause a significant increase in prices of commodities, produced particularly from energy industries and energy-intensive industries, which would, in turn, decrease the demand for those commodities and outputs produced from those industries. This would lead drops in wage rates and capital rents and ultimately cause reductions in household income in these provinces. While the provision of emission trading helps lower these losses, it cannot completely offset the losses. At the national level, the negative impacts on household income under the output-based grandfathering scheme will be less than that under the emission-based grandfathering scheme, also allocating the revenue of auctioning to the household will have smaller negative impacts on household income than allocating to the local government. Therefore, at the national level, for grandfathering rules, the output-based scheme will have lower negative impacts on household income than emission-based scheme. In the case of auctioning, allocating the revenue to household will have less negative impacts on household income than allocating the revenue to local government. At the provincial level, all provinces will have negative impacts on household income under the emission-based grandfathering scheme and auctioning revenue allocated to local government scheme, but several provinces will have positive impacts on household income under other two schemes, for example, Beijing, Liaoning and Qinghai under output-based grandfathering scheme, and Beijing, Henan and Jiangxi under auctioning revenue allocated to household scheme. In the latter cases, the losses of household income in these provinces will be offset by the provision of emission trading. 50      Table 7. Impacts on household income (% change from the baseline in 2030) Free Allowances Auctioning Province (Grandfathering) With Linkage With Linkage Emission To Local To Output based based Government Household Anhui -0.7 -1.1 -1.8 -0.4 Beijing -0.9 0.3 -0.4 0.0 Chongqing -0.8 0.0 -0.8 -0.1 Fujian -0.5 -0.2 -1.0 -0.3 Gansu -0.6 -0.7 -1.4 -0.3 Guangdong -0.3 -0.3 -0.7 -0.1 Guangxi -0.5 -0.1 -0.9 -0.1 Guizhou -0.9 -2.4 -2.7 -1.1 Hainan -0.8 -0.1 -0.9 -0.2 Hebei -0.4 -1.2 -1.6 -0.4 Heilongjiang -0.8 -1.2 -1.5 -0.5 Henan -0.6 -0.8 -1.3 0.0 Hubei -0.5 -0.5 -1.0 -0.3 Hunan -0.6 -1.3 -1.7 -0.6 Jiangsu -0.8 -0.4 -0.7 -0.2 Jiangxi -1.0 -0.2 -1.1 0.1 Jilin -0.7 -0.8 -1.5 -0.4 Liaoning -0.5 0.1 -0.9 0.0 Inner Mongolia -1.1 -1.7 -2.3 -0.3 Ningxia -1.2 -2.3 -2.8 -1.0 Qinghai -2.3 3.7 -1.6 -0.9 Shaanxi -1.2 -1.0 -1.7 -0.6 Shandong -1.0 -0.6 -1.2 -0.3 Shanghai -0.6 -0.4 -1.1 -0.4 Shanxi -1.4 -3.0 -2.9 -1.2 Sichuan -0.7 -0.8 -1.3 -0.3 Tianjin -1.0 -0.6 -1.4 -0.5 Xinjiang -0.6 -0.6 -1.2 -0.3 Xizang -0.7 0.0 -1.1 -0.7 Yunnan -0.6 -0.4 -1.1 -0.3 Zhejiang -0.5 -0.4 -0.9 -0.2 China -0.7 -0.6 -1.2 -0.3 4.2.3 Impacts on International Trade due to ETS Tables 8(a) and 8(b) present impacts of ETS on imports, and Tables 9(a) and 9(b) present impacts of ETS on exports. Except exports under the output-based grandfathering scheme, meeting China’s NDC would negatively impact China’s international trade even if the ETS, a flexible market-based mechanism, is used to reduce CO2 emissions. The impact on China’s exports will be positive if the output-based grandfathering scheme is adopted. 51      Imports of tradable commodities in all provinces would drop when emissions allowances are distributed through auctioning, no matter allocating the revenue to local government or household, and also the negative impacts on imports of all provinces when emissions allowances are distributed through emission-based grandfathering. However, imports of tradable commodities in several provinces (e.g., Beijing, Gansu, Hunan, Qinghai) would increase slightly when emission allowances are allocated through output-based grandfathering. When considering the impacts on import by sector, every sector will decrease their imports when emission allowances are distributed through emission-based grandfathering, but some sectors (for example, non-metal sector, metal sector, electric power sector) will increase their imports under two kinds of auctioning schemes and output-based grandfathering scheme. In most of the provinces, negative impacts on imports are higher under auctioning as compared to grandfathering, and also negative impacts on imports are higher under emission-based grandfathering as compared to output-based grandfathering. Table 8. Impacts on Import (% change from the baseline in 2030) (a) By Sector Free Allowances Auctioning Sector (Grandfathering) With Linkage With Linkage Emission Revenue Revenue Output-based based To Local Government To Household AGRI -1.3 -1.0 -1.9 -1.7 COAL -18.4 -12.3 -13.6 -13.5 OILNG -1.8 -0.7 -2.1 -2.0 MINE -0.1 0.0 -1.0 -0.6 FTPMF -0.8 -0.3 -1.2 -0.9 PETRO -3.0 -0.5 -1.4 -1.4 CHEMI -0.2 0.7 0.1 0.3 NMETA 0.0 4.0 3.0 3.5 METAL 0.0 1.1 0.4 0.8 OTHMF -0.4 -0.4 -1.0 -0.6 ELECT -0.8 5.8 4.1 4.5 GAS -5.7 1.2 -2.3 -2.0 WATER -1.4 1.5 -1.1 -0.7 CONST -0.4 -0.2 -0.9 -0.5 TRANS -1.1 -0.5 -1.3 -1.4 SERVI -1.0 -0.3 -0.6 -1.4 52      (b) By Province Free Allowances (Grandfathering) Auctioning Province With Linkage With Linkage Revenue Revenue Emission based Output-based To Local To Household Government Anhui -0.4 -0.9 -1.5 -1.1 Beijing -1.5 0.1 -0.6 -0.7 Chongqing -0.6 -0.2 -1.1 -0.6 Fujian -0.5 -0.2 -0.7 -0.4 Gansu -0.3 0.5 -0.6 -0.2 Guangdong -0.7 -0.5 -1.0 -0.8 Guangxi -1.3 -0.7 -1.5 -1.1 Guizhou -0.3 -2.7 -3.3 -2.8 Hainan -1.1 -0.1 -1.1 -0.8 Hebei -0.2 -0.7 -1.0 -0.9 Heilongjiang -0.8 -1.3 -1.8 -1.6 Henan -0.5 -0.4 -1.0 -0.7 Hubei -1.0 -0.3 -0.8 -0.4 Hunan -0.4 0.2 -0.5 0.0 Jiangsu -0.5 -0.1 -0.6 -0.3 Jiangxi -0.2 -0.1 -0.6 -0.2 Jilin -0.5 -0.8 -1.4 -1.0 Liaoning -0.5 0.0 -0.9 -0.5 Inner Mongolia -1.2 -1.7 -1.0 -1.6 Ningxia -0.4 -1.4 -2.1 -1.5 Qinghai -1.9 2.9 -1.6 -1.5 Shaanxi -1.0 -0.6 -1.2 -0.8 Shandong -0.9 -0.6 -1.3 -1.2 Shanghai -0.7 -0.5 -1.2 -1.1 Shanxi -0.7 -1.4 -1.9 -1.5 Sichuan -0.4 -0.2 -0.9 -0.5 Tianjin -0.8 -0.7 -1.2 -1.1 Xinjiang -0.5 -0.6 -1.5 -1.1 Xizang -0.6 0.0 -0.7 -0.2 Yunnan -0.8 -0.6 -1.5 -1.0 Zhejiang -0.6 -0.4 -0.9 -0.7 China -0.8 -0.3 -1.0 -0.8 Unlike the impacts on imports of different provinces, no matter emissions allowances are distributed through auctioning or grandfathering, although most of the provinces would decrease their exports, there are several provinces would increase their exports by different extent, for example, Fujian would increase the exports under all four different emission allowance distribution options. When considering the impacts on exports by sector, most of the 53      sectors would decrease their exports no matter auctioning and grandfathering, but there are several sectors would increase exports under every emissions allowance allocation option, for example, the agriculture sector would increase the exports under all four options. Table 9. Impacts on Export (% change from the baseline in 2030) (a) By Sector Free Allowances Auctioning Sector (Grandfathering) With Linkage With Linkage Emission Revenue Revenue Output-based based To Local Government To Household AGRI 1.3 1.8 3.1 3.3 COAL -12.7 -21.6 -24.8 -25.1 OILNG -37.3 -4.9 12.3 21.6 MINE -4.4 3.2 -0.5 -0.2 FTPMF -0.5 1.2 0.4 0.6 PETRO -4.1 -2.6 -6.0 -5.9 CHEMI -2.1 -4.0 -6.4 -6.2 NMETA -7.8 12.2 -3.5 -4.1 METAL -1.2 -4.7 -6.6 -6.4 OTHMF -0.6 1.0 -0.4 -0.1 ELECT -3.4 -12.4 -15.4 -15.6 GAS -8.6 -6.7 -12.4 -12.3 WATER 3.4 3.8 7.0 7.1 CONST 0.1 0.3 -0.3 0.1 TRANS -0.7 -2.5 -2.9 -3.1 SERVI 0.0 0.8 2.8 1.8 (b) By Province Free Allowances (Grandfathering) Auctioning Province With Linkage With Linkage Revenue Revenue Emission based Output-based To Local To Household Government Anhui -1.2 -6.1 -6.5 -6.6 Beijing -6.0 -7.6 -10.0 -10.0 Chongqing -0.4 1.9 2.1 2.3 Fujian 0.3 0.5 0.8 0.9 Gansu -0.5 -5.9 -5.7 -5.7 Guangdong 0.0 -0.4 -0.7 -0.7 Guangxi -0.3 0.4 0.2 0.4 Guizhou -0.5 -5.9 -5.6 -5.6 Hainan -2.0 -9.3 4.9 4.9 Hebei -0.8 -6.3 -6.8 -6.7 Heilongjiang -1.2 -4.9 -4.5 -4.5 Henan -0.6 -1.9 -1.9 -1.7 54      Hubei -0.4 -2.8 -1.8 -1.7 Hunan -0.5 -5.3 -4.7 -4.5 Jiangsu -1.1 -0.2 -0.9 -0.5 Jiangxi -3.1 -2.8 -4.8 -4.5 Jilin -1.2 -2.1 -2.2 -2.0 Liaoning -0.6 0.4 -0.6 -0.4 Inner Mongolia -0.3 -4.7 -3.2 -3.7 Ningxia -0.5 -9.4 -10.0 -10.0 Qinghai -4.4 7.7 -2.5 -3.2 Shaanxi -0.1 -1.2 -1.7 -1.6 Shandong -3.0 -3.9 -7.0 -6.9 Shanghai -0.8 -1.9 -3.5 -3.4 Shanxi -1.8 -2.6 -7.3 -7.3 Sichuan -0.6 -2.3 -2.0 -1.8 Tianjin -1.0 -1.4 -2.2 -2.1 Xinjiang -0.2 -2.7 -2.4 -2.4 Xizang -0.1 1.7 1.5 1.8 Yunnan -0.4 -2.1 -1.8 -1.7 Zhejiang -0.6 -0.6 -1.4 -1.4 China -3.5 1.5 -3.6 -3.6 5. Conclusions In this paper, we analyze the potential impacts on the provincial economies of meeting China’s emission reduction pledges made under the Paris Agreement using a national emission trading scheme. For the analysis, we developed a multi-regional, multi-sectoral, recursive dynamic CGE model and calibrated it with the data from the latest (2012) provincial-level social accounting matrices. We consider two cases for allocation of emission allowances: grandfathering and auctioning. Grandfathering allocation is carried out under two approaches: emission-based and output-based. The revenue collected through allowance auctioning is recycled to the local economy through two schemes: public consumption/investment and lump- sum transfer to households. The national emission trading is assumed to start in 2020 and national caps on emissions will be tightened gradually over the years and fixed at the level in 2030 to meet the emission intensity targets under the Paris Agreement. The study also considers an alternative baseline, which incorporates detailed technological information by using the 55      information from Timilsina et al. (2019) that links top-down CGE models to the bottom-up TIMES model and estimates the baseline correction factors. The study shows that China would be required to reduce almost 30% of its CO2 emissions from the business as the usual situation in 2030 to meet its nationally determined contribution (NDC) targets or pledges under the Paris Agreement. Note that required emission reductions from the baseline are highly sensitive to the projection of the baseline. When the information from the bottom-up TIMES model is used, the baseline emission drops by about half. Consequently, the required emission reductions to meet the NDC targets also drop, and so do the corresponding economic impacts. The study finds that meeting China’s nationally determined contribution (NDC) or pledges under the Paris Agreement would cause 0.4% to 1.5% reduction of national economic output (GDP) in 2030 depending on (a) allocation of distribution of emission allowances and (b) baseline setting (i.e., utilizing or not the baseline correction factor derived from the linking of the CGE model with the bottom-up model). For some provinces, the ETS provides an opportunity to increase their economic outputs (e.g., Fujian, Guangdong, Guangxi, Liaoning) under all schemes of allowances allocation/distribution. Provincial economies of Inner Mongolia, Ningxia, Shanxi and Shaanxi that highly rely on the coal mining sector and coal- intensive manufacturing sectors (electricity, non-metallic minerals, metals) would face relatively higher economic loss under each scheme of allowances allocation considered. Provincial GDP impacts widely vary depending upon economic structures and energy supply systems. Allowances allocation rules change not only the magnitude of the impacts, but also the direction of the impacts. The magnitudes of economic impacts of the ETS are highly sensitive to the baseline emissions. Our study finds that if the baseline is corrected by using a factor derived by Timilsina et al. (2019) through the linking of a top-down CGE model and bottom-up TIMES model, the magnitudes of economic impacts drop by almost two-thirds. Correction of the baseline, however, does not alter the direction of the impacts. The sensitivity of economic impacts of an ETS to the allowance allocation rules and also to the baseline setting suggests that design of ETS is critical for fair and equitable distribution 56      of the impacts across the provincial economies and sectoral outputs. Policy makers should consider compensation measures in those economies and sectors which face deep adverse economic impacts from an ETS. What are the appropriate compensation measures for an equitable and fair distribution of the impacts of an ETS in China and beyond is an important research question for the future expansion of this study.   References Bohringer, Christoph, Bouwe Dijkstra and Knut Einar Rosendahl (2014). Sectoral and Regional Expansion of Emissions Trading, Resource and Energy Economics, 37, pp. 201-25 Coase, Ronald H. (1960). The Problem of Social Cost. Journal of Law and Economics, Vol. 3, No. 1, pp. 1–44. Cui, Lian-Biao, Ying Fan, Lei Zhu, Qing-Hua Bi and Yi Zhang. 2013. The Cost Saving Effect of Carbon Markets in China for Achieving the Reduction Targets in the “12th Five-Year Plan”. Chinese Journal of Management Science, 21(1): 37-46. (in Chinese) Cui, Lian-Biao, Ying Fan, Lei Zhu, Qing-Hua Bi. 2014. How Will the Emissions Trading Scheme Save Cost for Achieving China’s 2020 Carbon Intensity Reduction Target?. Applied Energy, 136(12): 1043-1052. Duan, Maosheng and Li Zhou (2017). Key issues in designing China’s national carbon emissions trading system. Economics of Energy & Environmental Policy, Vol. 6, No. 2. Pp. 55-72. Goulder, Lawrence H., Richard D. Morgenstern, Clayton Munnings and Jeremy Schreifels (2017). China’s national carbon dioxide emission trading system: An introduction. Economics of Energy & Environmental Policy, Vol. 6, No. 2. pp 1-18. Grimm, Veronika and Lyuba Ilieva (2013). An Experiment on Emissions Trading: The Effect of Different Allocation Mechanisms. Journal of Regulatory Economics, Vol. 44, No. 3, pp. 308-38. 57      Hubler, Michael, Sebastian Voigt and Andreas Loschel (2014). Designing an Emissions Trading Scheme for China--An Up-to-Date Climate Policy Assessment, Energy Policy, Vol. 75, pp. 57-72. Johansson A, Guillemette Y, Murtin F et al. 2013. Long-term growth scenarios. OECD economics department working papers, Paris: OECD Publishing. Jotzo, Frank and Andreas Loschel (2014). Emissions Trading in China: Emerging Experiences and International Lessons: Guest Editorial. Energy Policy, Vol. 75, pp. 3- 8. Li Shan Tong. 2010. China's economic prospect for the 12th five-year plan period and 2030. China J Rev Econ Res;43:2–27. Lo, Alex Y. (2016). Challenges to the Development of Carbon Markets in China. Climate Policy 16 (1): 109–24. Munnings, Clayton, Richard D. Morgenstern, Zhongmin Wang and Xu Liu (2016). Assessing the Design of Three Carbon Trading Pilot Programs in China. Energy Policy 96: 688– 99. National Bureau of Statistics (NBS). 2016. China provincial input–output tables 2012. China Statistics Press. (in Chinese). OECD. 2012. Looking to 2060: Long-term global growth prospects. OECD economic policy papers. Paris: OECD Publishing. Pang, Tao and Maosheng Duan (2016). Cap Setting and Allowance Allocation in China’s Emissions Trading Pilot Programmes: Special Issues and Innovative Solutions. Climate Policy 16 (7): 815–35. Rose Adam and Tom Tietenberg (1993). An International System of Tradable CO2 Entitlements: Implications for Economic Development. Journal of Environment and Development, Vol. 2, pp. 1-36. 58      Rose, Adam, Dan Wei and Fynnwin Prager (2014). Distributional Impacts of Greenhouse Gas Emissions Trading: Alternative Allocation and Recycling Strategies in California. Contemporary Economic Policy, Vol. 30, iss. 4, pp. 603-17. Tang, Ling, Jiaqian Wu, Lean Yu, Qin Bao (2015). Carbon emissions trading scheme exploration in China: A multi-agent-based model. Energy Policy, Vol. 81, pp. 152- 169. Timilsina, Govinda R., Jun Pang and Xi Yang (2019). Linking Top-Down and Bottom-UP models for Climate Policy Analysis: The Case of China. World Bank Policy Research Working Paper #8905. World bank, Washington, DC. United Nations Department of Economic and Social Affairs (UNDESA) (2015). World population prospects. 2015. Population Division, UNDESA, United Nations, New York. United Nations Framework Convention on Climate Change (UNFCCC) (2015). Report of the Conference of the Parties on its nineteenth session, held in Paris from 30 November to 11 December 2015, Addendum Part Two: Action taken by the Conference of the Parties at its twenty one session, UNFCCC, Bonn, Germany. United Nations Framework Convention on Climate Change (UNFCCC) (2016). Aggregate effect of the intended nationally determined contributions: an update, Synthesis report by the secretariat. http://unfccc.int/resource/docs/2016/cop22/eng/02.pdf Wang, Jianzhou, Yao Dong, Jie Wu, Ren Mu, He Jiang (2011). Coal production forecast and low carbon policies in China. Energy Policy, Vol. 39, pp. 5970-5979. Wing, I. S., 2008. The synthesis of bottom-up and top-down approaches to climate policy modeling: electric power technology detail in a social accounting framework. Energy Economics, 30(2), 547-573. Wu, Jianxin, Zhi-Yong Kang, Ning Zhang (2017). Carbon emission reduction potentials under different polices in Chinese cities: A scenario-based analysis. Journal of Cleaner Production, Vol. 161, pp. 1226-1236. 59      Wu, Jie; Fan, Ying; Xia, Yan (2016). The Economic Effects of Initial Quota Allocations on Carbon Emissions Trading in China, Energy Journal, Special Issue 1, Vol 16, pp. 129- 51. Wu, Libo, Haoqi Qian and Jin Li (2014). Advancing the Experiment to Reality: Perspectives on Shanghai Pilot Carbon Emissions Trading Scheme. Energy Policy, Vol. 75, pp. 22- 30. Yuan, Y.N., M.J. Shi, N. Li, and S.L. Zhou. (2012). Intensity Allocation Criteria of Carbon Emissions Permits and Regional Economic Development in China—Based on a 30- Province/Autonomous Region Computable General Equilibrium Model. Advances in Climate Change Research 3(3): 154–62. Zhang, D., M. Springmann and V.J. Karplus (2016). Equity and emissions trading in China. Climatic Change, 134: 131–146. Zhang, Da; Valerie J. Karplus, Cyril Cassisa and Xiliang Zhang (2014). Emissions Trading in China: Progress and Prospects. Energy Policy, Vol. 75, pp. 9-16. Zhang, Hao (2013). Design Elements of Emissions Trading Regulation in China’s Pilot Programs: Regulatory Challenges and Prospects. Environmental and Planning Law Journal 30 (4): 342–356. Zhang, Junjie, Zhenxuan Wang, and Xinming Du (2017). Lessons Learned from China’s Regional Carbon Market Pilots. Economics of Energy & Environmental Policy 6 (2): 19–38. Zhang, Z. X. (2015). Carbon Emissions Trading in China: The Evolution from Pilots to a Nationwide Scheme, CCEP Working Paper 1503, April 2015. School of Economics, Fudan University, Shanghai, China. Zhu, Lei; Zhang, Xiao-Bing; Li, Yuan; Wang, Xu and Guo, Jianxin (2017). Can an Emission Trading Scheme Promote the Withdrawal of Outdated Capacity in Energy-Intensive Sectors? A Case Study on China's Iron and Steel Industry. Energy Economics, v. 63, pp. 332-47. 60