Exploring the Road to Carbon Neutrality

Abstract

How to deal with carbon emissions—a rare externality that spans a large time frame and geographical scope—is a difficult task for the world. This task is particularly challenging for China, mainly in that the country must coordinate dual objectives, including existing economic growth targets and the newly added carbon neutrality goal. Over the past 40 years of reform and opening up, China has been setting economic growth targets and striving to achieve them. In recent years, although growth targets have softened along with the secular decline in potential growth rate, economic growth remains a top priority for China, the world’s largest developing country. We expect China to reach the current standard for a high-income country by the end of the 14th Five-Year Plan period, and to double its GDP or per capita income by 2035. Currently, China is adding a new constraint over the next 40 years. As the world’s largest carbon emitter, China has set out a clear timetable for carbon neutrality—to reduce its carbon emission intensity in 2030 by more than 65% from the 2005 level, and reach the peak of carbon dioxide emissions by 2030 and become carbon neutral by 2060. We note that it will take 71 and 45 years, respectively, for the EU and the US to achieve the carbon neutrality goal from peak carbon emissions (reached by the EU in 1979 and by the US in 2005) to net zero emissions. China’s aggressive timetable to achieve carbon neutrality within 40 years means that the country will face a much steeper slope of carbon emissions than the EU and the US. How will China strike a balance between the objectives of economic growth in the past 40 years and carbon neutrality in the next 40 years? We discuss this issue from an aggregate and a structural point of view. In our aggregate analysis, the most important task is to identify the peak of China’s carbon emissions in 2030. We believe that in order to take economic growth and emission reduction into consideration, it is more appropriate to set the carbon peak target in a range to avoid rigid constraints. From a structural perspective, we discuss how China can achieve its carbon peak and neutrality goals. Under the framework of a “green premium”, we come up with a preliminary idea of “technology + carbon pricing” based on the analysis of eight high-emission industries. We prove that this idea can strike a balance between the constraints of economic growth and carbon neutrality goals through general equilibrium analysis using the computable general equilibrium (CGE) model. Finally, we incorporate social governance into our analysis by discussing the meaning of a negative green premium, and arrive at this formula: the road to carbon neutrality = technology + carbon pricing + social governance.

1.1 Seeking a Peak: 9.9–10.8bn Tonnes of Net Carbon Emissions

We examine China’s carbon emission issue from a historical and a future perspective.¹ As shown in Table 1.1, while China’s annual carbon emissions in 2019 were much larger than other economies, the US and the EU, which started industrialization earlier, had greater cumulative emissions.² Moreover, China’s carbon emissions per capita were 7.1 tonnes, still less than half the US’s 16.1 tonnes (see Table 1.2). Such historical perspective is very important for countries to distinguish their “common but differentiated” responsibilities when making coordinated carbon neutrality goals. However, from a future perspective, these facts are of little significance to China’s achievement of its goals of carbon peak by 2030 and carbon neutrality by 2060.

Table 1.1 Annual and cumulative carbon emissions

Table 1.2 Intensity of carbon emissions and emissions per capita

Although China’s carbon emission intensity has declined the fastest among major economies in the past few decades, its current carbon emission intensity is still about 5 times that of the EU and 3 times that of the US due to large differences in industrial structure. If the road to carbon peak in the next 10 years is understood as seizing the “buffer period” to catch up with the US in terms of cumulative emissions and per capita emissions, the road to carbon neutrality in the subsequent 30 years would be tremendously challenging.

More importantly, carbon peak in the US and the EU is not a target of previously decided emission constraints but a natural result of development, which means it is of little significance for referencing in China’s peak setting. From our point of view, the targeting of a carbon peak by 2030 does not mean that China can emit freely in the next 10 years. Instead, as carbon neutrality has become a new constraint, the target of carbon peaking by 2030 requires that China changes its thinking from now on and take actions immediately to reach the peak as soon as possible and try its best to lower the peak. Additionally, we believe that it is more appropriate to set the target within a range instead of a specific number so as to avoid a rigid constraint in supply side. Such consideration is mainly based on the following three aspects:

First, the statistics of carbon emission are inherently uncertain and are difficult to quantify accurately. Constrained by current technological measures, the amount of greenhouse gas emissions in each country is obtained through aggregating the levels of emission in various activities multiplied by the corresponding emission factors. Such calculation results are uncertain because of inherent methods. To unify the calculation measure for all countries, Intergovernmental Panel on Climate Change (IPCC) has formulated the Guidelines for National Greenhouse Gas Inventories, which provides basic technical regulations and emission factor data applicable to all countries.

Nevertheless, the IPCC also emphasizes and encourages each country to take measurements and parameters that are applicable to the country-specific situation. Since it is difficult to accurately calculate the level of emission activities and emission factors, different institutions have various standards to determine localized parameters. Even different literature in major scientific journals such as Nature would have a difference of 10%–20% in the calculation of carbon emissions in the same country. Due to the inherent uncertainty in carbon emission calculation, it may not conform to the scientific laws to pursue an accurate target of carbon peak.

Secondly, our estimates suggest that China’s net carbon emissions will range from 9.9 to 10.8bn tonnes in 2030. Looking into the future, the peak of China’s carbon emissions depends on two factors. One is the GDP growth trajectory from now to 2030. According to the Outline of the 14th Five-Year Plan (FYP), we assume that GDP will double over the period of 2020–2035 and the GDP growth rate will drop at a constant rate from 6% in 2019.³ Thus we derive the GDP CAGR over 2020–2030 at 5%. The other is the reduction in carbon emission intensity in 2030 from the current level. The authorities have stated that carbon emission intensity in 2030 will drop by more than 65% from the level in 2005. In recent years, the government has announced the reduction in carbon emission intensity each year from the level in 2005 (for example, the intensity in 2019 was 48.1% lower than that in 2005). The 13th and 14th FYPs clearly set a target of reducing carbon emission intensity by 18%. We assume that the 15th FYP will also target an 18% reduction. As such, we estimate that carbon emission intensity will be reduced by an average of 3.9% annually in the next 10 years, and the intensity in 2030 will decline for more than 65% from the level in 2005.

Thirdly, the risk of stagflation is looming. As a result, setting the peak target within a range would help improve supply elasticity and balance between economic growth and emission reduction. As the world’s largest developing country, sustainable economic growth remains China’s top priority. Meanwhile, the launch of “carbon peak in 2030, carbon neutral in 2060” suggests that the country is implementing a new constraint to the next 40 years. Before carbon neutral technologies become fully mature, the conflict between the two constraints will persist. In such a scenario, a rigid emission target would be able to help promote emission reduction, but would also easily induce stagflation.

In fact, such risk may have emerged already. The rapid increase in PPI during March and April 2021 may be partially attributed to the post-pandemic demand recovery, but what is more noteworthy is the decrease in supply elasticity under strengthened environmental constraints. This is relevant to the heavy reliance on the rigid administrative policies such as shutdowns or restricted production in certain areas. Although carbon neutrality means cost increases are inevitable, it is not necessary to cope with the transition pressures in short term.

At the current stage, in which the foundation of economic recovery is still unstable, measures with strong supply shocks such as production restriction and shutdown should be implemented with caution. Measures that come with lower social cost and higher efficiency, such as promoting inventory adjustment through incremental reforms, enlarging low-carbon and zero-carbon capacity investment to enhance high-carbon capacity replacement, and strictly controlling new-added high-carbon capacity investments, deserve more consideration. From a medium- and long-term perspective, a more important aspect of overall management is to set the peak emission target within a range to avoid overly rigid supply constraints.

Based on the above-mentioned trajectories of GDP growth and carbon emission intensity reduction in the future, we can discuss in further details the peak of carbon emissions in 2030. First, we need to solve the issue of carbon emission data. The existing carbon emission databases mainly include the China Emission Accounts and Datasets (CEADs), BP database, and official data from the Ministry of Ecology and Environment (MoEE). Since data from the MoEE is not continuous, the time-continuous CEADs and BP databases are more suitable for research (see Fig. 1.1). Carbon emissions according to the three databases (not considering carbon sinks) were 5.4, 6.1, and 5.98bn tonnes in 2005, and 9.44, 9.24, and 10.28bn tonnes in 2014.⁴ Emissions in 2017 according to the CEADs and BP database were 9.34 and 9.3bn tonnes. It can be seen that the numerical difference between the CEADs and BP database has narrowed significantly in recent years. Considering that the CEADs include emission data for 29 industries and is convenient for structural analysis and CGE model simulation, we decided to use the CEADs for analysis.

Fig. 1.1

Source CEADs, BP, MoEE, CICC Research

Carbon emissions according to different databases.

In order to avoid the uncertainty caused by the large differences between early carbon emission data of different databases, we select carbon emissions in 2017 according to the CEADs (9.34bn tonnes) as the calculation benchmark. We calculate peak carbon emissions in 2030 under two scenarios of carbon emission intensity reduction: (1) carbon emission intensity reduced by 46% in 2017 (official data) and by 65% in 2030 (official target) from the level in 2005; and (2) carbon emission intensity falling by an average of 3.9% annually before emissions peak, or carbon emission intensity reduced by 66.5% in 2030 from the level in 2005. Carbon emissions in 2030 will peak at 11.7bn tonnes and 10.8bn tonnes respectively in these two scenarios. Considering that there will be 910mn tonnes of agricultural and forestry carbon sinks in 2030,⁵ net emissions in 2030 will peak at 10.8bn tonnes and 9.9bn tonnes respectively in these two scenarios. Since scenario 2 is based on an additional assumption that the 15th FYP will still target an 18% reduction in carbon emission intensity, which increases the uncertainty of the calculation result, we tend to adopt the calculation result of scenario 1, i.e. a peak of 10.8bn tonnes. In addition, although we use the CEADs as the benchmark database, we also make same calculations based on the BP data in 2019 and the official data in 2014 (see Table 1.3), in order to assess the extent of the calculation uncertainty.

Table 1.3 Calculations of peak carbon emissions (net emissions) in 2030

Discussion on the Uncertainty of Calculations

Our calculations of peak emissions in 2030 based on emission intensity may face uncertainty in three aspects: The assumption on future GDP growth rate, the quality of historical emission data, and the understanding of the emission intensity reduction target itself.

• 1. The assumption on GDP growth rate is mainly based on the proposal to “double China’s economic aggregate or per capita income by 2035”. The uncertainty lies in the assumption that the GDP growth rate will drop at a constant rate from 2019 in the next 15 years.

• 2. As many studies have noticed, the quality of historical data on carbon emissions in early years is not high enough. For example, there is a controversy about the amount of China’s emissions in 2005. We noticed two phenomena in the calculation process. One is that BP data and CEADs data have gradually converged over time. The other is that the officially announced emission intensity in 2017 declined 46% from the level in 2005, and the newer emission data in 2017 is less controversial. Hence, we use 2017, for which data is of higher quality, rather than 2005, for which data is more controversial, as the benchmark for calculations, to avoid the risk of errors stemming from the relatively low quality of early emission data.

• 3. Academically, scholars’ calculations of peak carbon emissions usually refer to net emissions after deduction of carbon sinks. The CEADs and BP databases calculate gross emissions without deduction of carbon sinks. Whether the target of reducing carbon emission intensity in 2030 by 65% from the level in 2005 is based on gross emissions or net emissions does not seem to be specified. In our aforementioned calculation method, we interpret this target from the perspective of gross emissions. We first calculate peak gross emissions in 2030 at 11.7bn tonnes, then deduct carbon sinks in 2030, and derive peak net emissions in 2030 at 10.8bn tonnes, roughly between the peaks calculated by Tsinghua University and the World Resources Institute. If we interpret the target from the perspective of net emissions, we need to first convert the data on gross emissions in the CEADs into data on net emissions based on historical carbon sinks calculated by Tsinghua University and other institutions, and then we arrive at peak net emissions of about 10.9bn tonnes in 2030, close to the calculation result based on gross emissions. Upon consultation with our utilities analysts, we tend to understand the target of emission intensity reduction from the perspective of gross emissions, which means that we estimate peak net emissions in 2030 at about 10.8bn tonnes.

In summary, even with the upper limit of 10.8bn tonnes for peak carbon emission in 2030, our estimate still suggests that China’s emissions can only increase less than 10% in the next 10 years compared with the level in 2020.⁶ At the same time, China’s annualized real GDP growth rate needs to reach 5%. This highlights the difficulty of China’s carbon peaking in the next 10 years. From the perspective of international comparison, China’s peak emissions are much higher than those of the EU and the US, meaning that it is more difficult for China to achieve carbon neutrality after reaching the carbon peak. The EU reached its carbon peak in 1979, with peak emissions at about 4.1bn tonnes and peak emissions per capita at 9.9 tonnes. The US reached its carbon peak in 2005, with peak emissions at about 6.1bn tonnes and peak emissions per capita at 19.6 tonnes. According to our above calculations, China’s peak emissions in 2030 will be 10.8bn tonnes and peak emissions per capita will be 7.4 tonnes. Both the EU⁷ and the US⁸ aim to become carbon neutral by 2050. If the stage from peak to neutrality is defined as the second half of carbon neutrality, the second half will be 71 years for the EU and 45 years for the US, but only 30 years for China. From an aggregate perspective, China will face a much steeper slope of emission reduction than the EU (see Figs. 1.2 and 1.3). How can China achieve this difficult goal? We will discuss the industry-level neutralization path from a structural perspective in the next section.

Fig. 1.2

Source Our World in Data, World Bank, CICC Research. Note tonne = 0 represents the time when carbon peak is reached

Countries’ estimated slope of emission reduction after the peak.Footnote

Since the Our World in Data database does not specify whether a country’s carbon emissions are net emissions or gross emissions, there is a certain degree of uncertainty in the comparison here.

Fig. 1.3

Source Our World in Data, World Bank, CICC Research

Countries’ estimated slope of per capita emission reduction after the peak.Footnote

Since the Our World in Data database does not specify whether a country’s carbon emissions are net emissions or gross emissions, there is a certain degree of uncertainty in the comparison here.

1.2 Structural Path: Analysis Based on Green Premium

After the aggregate target is determined, the next step is to explore how to achieve peak carbon emissions and become carbon neutral. There are two ways to do this: (1) reducing emissions on the demand side, or even directly curbing demand (such as power rationing); and (2) promoting clean energy, improving production technology, and developing carbon capture on the supply side. Restricting demand runs counter to the pursuit of economic growth in the past 40 years, and demand-side management such as power rationing and emission reduction promotion can only realize a (temporary) reduction in emissions but is unable to achieve carbon neutrality. The fundamental solution is to use zero-emission energy and zero-emission production technologies on the supply side. The supply side holds the key to China’s goals to achieve peak emissions at 10.8bn tonnes and become carbon neutral in 30 years.

The key to exploring how to achieve carbon neutrality on the supply side lies in proper handling of emissions of 8 high-emission industries. In the Notice on Key Points for the Launch of the National Carbon Emissions Trading Market issued in 2016, the National Development and Reform Commission listed the petrochemicals, chemicals, building materials, steel, nonferrous metals, paper, power (thermal power), and airlines as high-emission industries. According to the CEADs, national carbon dioxide emissions increased from 3.003bn tonnes in 2000 to 9.339bn tonnes in 2017, and the proportion of these 8 industries¹¹ in total emissions increased from 80% to 88% (see Fig. 1.4), mainly driven by the power and steel industries. Among the 8 industries (see Fig. 1.5), the power industry accounted for the largest proportion of total emissions in 2017 (mainly thermal power, 44%), followed by steel (18%), building materials (13%), transportation (including airlines, 8%), chemicals (3%), petrochemicals (2%), nonferrous metals (1%), and paper (0.3%).

Fig. 1.4

Source CEADs, CICC Research

Carbon emission proportion of 8 high-emission industries combined.

Fig. 1.5

Source CEADs, CICC Research

Carbon emission proportions of 8 high-emission industries.

It should be noted that the CEADs’ statistics on the industrial structure of carbon emissions are mainly based on the net emissions of the production process, and the indirect emissions generated by energy-intensive industries from their electricity consumption are not included in the emissions of the industries. The power industry accounts for nearly half of total emissions because the emissions from power consumption of all sectors are included in the power industry, while the emissions of other high-emission industries such as steel, building materials, nonferrous metals, and chemicals are their direct emissions generated from the production process (such as the emissions generated from blast furnaces’ consumption of iron ore and coke in steelmaking, and from decomposition of limestone and coal-fired kilns in the calcination of cement clinker).

Obviously, direct carbon emissions of the production process are highly related to the production technology, so we can also think about the road to carbon peak and neutrality from the perspective of improving production technology. For example, we can shift from coal power to solar power with zero emissions, and greatly reduce the emissions from steelmaking by using electric arc furnaces. For the emissions from cement clinker calcination that cannot be reduced by changing production technology, we can add carbon capture and storage facilities to reduce emissions or even achieve zero emissions. This idea is reflected as the “green premium”. In this report, we define the green premium, which was coined by Bill Gates, as the percentage increase in the production cost of zero-emission technology vs. current emission-generating technology. In a sense, the key to carbon neutrality lies in reducing the green premium. When the green premium drops to zero, it means that carbon-neutral production technology is very mature and there is no need to use fossil energy.

However, the green premium varies greatly among different industries—the difference between building materials with the highest green premium and nonferrous metals with the lowest green premium is nearly 38 times. What can explain this difference? We divide the 8 high-emission industries into three sectors: power, transportation, and manufacturing (building materials, chemicals, steel, paper, petrochemicals, and nonferrous metals). With different production methods, these three sectors need to be discussed separately. The detailed analysis is as follows.

• 1. Power sector. The green premium in power generation is already negative. Currently hydropower and wind power generation costs are lower than thermal power, and nuclear power and solar power generation costs are slightly higher than thermal power. Based on the current power structure (thermal power 66%, hydropower 19%, wind power 6%, nuclear power 5%, and solar power 4%), the overall clean energy power generation cost is already lower than thermal power. However, the cost for the power grid to realize carbon neutrality is relatively high. Overall, our sector analysts estimate the green premium in the power sector in 2021 at 17%. For the power grid to absorb alternative energy, it is necessary to increase the flexibility of grid resource dispatch in the auxiliary service market, including power adjustment of thermal power units, pumping of pumped storage units, and charging of energy storage power stations. The cost for the power grid increases along with the proportions of wind power and solar power. Currently, thermal power peak shaving is still the lowest-cost method of grid dispatch. Before new technology emerges to greatly reduce the cost for the power grid to absorb alternative energy, thermal power is unlikely to be removed from the power system.

• 2. Transportation sector. The transportation sector includes roads, railways, airlines, and marine transport, which have quite different green premiums. Our sector analysts estimate the green premium in the transportation sector in 2021 at 68%. Under existing technologies, various transportation sub-sectors need to adopt different methods in order to achieve zero emissions. Assuming that passenger vehicles, medium/light-duty trucks and railways use clean power and heavy-duty trucks, airlines and marine transport use hydrogen energy, our sector analysts estimate the green premium at 18% in road passenger transport, 127% in road freight transport, 343% for airlines, 319% in marine transport, and –29% for railways. Since the cost of hydrogen energy is much higher than that of clean power, heavy-duty trucks, airlines, and marine transport have much higher green premiums than road passenger transport and light-duty trucks. As railways become increasingly electrified, their green premium has been negative.

• 3. Manufacturing sector. Under existing technologies, the green premium in the manufacturing sector is related to production technology. High-emission industries such as cement and chemicals need to adopt high-cost carbon capture and other technologies in order to achieve zero emissions. Our sector analysts estimate the green premium in 2021 at 138% in the building materials industry, 53% in the chemical industry, 15% in the steel industry, 11% in the paper industry, 7% in the petrochemical industry, and 4% in the nonferrous metals industry.

The analysis of these three sectors shows that the degree of technology maturity is an important determinant of the green premium. Taking the power sector with the highest emission proportion as an example, zero-carbon electricity has become feasible, as non-fossil energy technologies have made progress and wind and solar power costs have declined significantly in the past 10 years thanks to economies of scale, material replacement and efficiency improvement. Even taking into account grid costs, we expect electricity to achieve a negative green premium within 10 years. In fact, the technological progress achieved by the power sector has made the largest contribution to the continuous decline of the CICC Green Premium Index over the past few years.

Manufacturing industries such as building materials, steel, nonferrous metals, and chemicals have relatively high green premiums. In addition to the cost increase resulting from electrification of the production process, the main reason is that the inevitable use of raw materials and burning of fossil energy in the production of some products makes it necessary to use hydrogen energy and carbon capture technologies in order to achieve carbon neutrality, but these technologies are not yet mature and have high costs. The transportation sector has a high green premium because heavy-duty trucks, airlines and marine transport need to replace existing fuels with hydrogen energy in order to achieve zero emissions, and a future decline in the green premium requires maturity of hydrogen energy technology.

In summary, technological progress plays an important role in lowering the green premium, but it is not the complete answer. We can see from the calculation method that the reduction in green premium can be achieved by relying on technological advancement to reduce the production cost of zero-emission technologies, or by increasing the production cost of emission-generating technologies. In fact, the latter is the most recommended way to achieve carbon neutrality under the framework of mainstream economics. According to mainstream economics, carbon emissions have become a problem because of their negative externalities. From a temporal perspective, it may take decades or even centuries for carbon emissions to have serious consequences by causing climate change, so this externality may have a relatively limited impact on current economic activities and daily life. From a spatial perspective, as carbon emissions enter global atmospheric circulation, emissions in any region will have an impact on the global climate. Since the global ecosystem is the basis for human survival and development, the problem of carbon emissions may have consequences that we cannot afford if it is not properly resolved. The current mainstream belief is that it is necessary to internalize the great externality of carbon emissions in order to solve it. From the perspective of green premium, this is to increase the production cost of existing emission-generating technologies.

The green premium is a more compatible analytical framework for thinking about the road to carbon neutrality that offers two basic paths: (1) the mainstream carbon pricing approach, which internalizes the negative externality of emissions through carbon tax or carbon market pricing mechanisms to increase the production cost of emission-generating technologies; and (2) technological advancement that makes zero-emission technologies more economically mature to reduce their production costs. Which of these two ways is more effective? What do they mean for the output and structure of the economy? How can we achieve the goals of carbon peak by 2030 and carbon neutrality by 2060? We will try to answer these questions in the CGE model analysis in the next section.

1.3 Four Scenarios: General Equilibrium Analysis Under CGE Model

The computable general equilibrium (CGE) model is a quantitative simulation system based on general equilibrium theory, macroeconomic structure relations and national accounting data that describes the operation of the economic system in a panoramic manner. The model is usually used to study the long-term, deterministic behavior of the overall economy and its response to external shocks, and has become a standard tool for global economic and policy analysis. In the model, changes in any part of the structure will have an impact on the entire system, leading to changes in the prices and quantities of goods and factors. When an exogenous shock occurs, the system will transition from one equilibrium state to another, thus fully demonstrating the impact of the shock.

Under constrained conditions, producers minimize costs while consumers maximize utility. The markets for factors of production and goods reach supply and demand balance through production, distribution, exchange, and consumption, thereby forming the equilibrium price. On the basis of the traditional CGE method for studying the environmental economy, we have incorporated endogenous climate-friendly technologies into the framework. Thus, the CGE model in this report comprises a production module, energy module, income and expenditure module, trade module, carbon tax/carbon trading module, dynamic module, climate-friendly technology module, and macro closure module.

In the context of carbon neutrality, the energy module is the key part of the model. The energy module consists of two parts: fossil energy (coal, oil, natural gas) and electricity. Thermal power and renewable energy make up the electricity system. On this basis, the energy structure is generated endogenously by the model and changes over time. Energy and non-energy factors are intermediate inputs, which together with the added value of capital and labor constitute total input. Essentially, we take fossil energy as a factor of production in the model and assign a price to it through carbon tax/carbon trading to correct the externality of carbon emissions. In this sense, the model has a similar core as the classical growth model of Nobel Prize laureate William Nordhaus, which includes carbon cycle. In the absence of climate-friendly technological advancements, companies will make a trade-off between the benefits of using one more unit of fossil energy and the additional costs paid given the carbon costs. The economy will eventually converge to a new equilibrium growth path, which often corresponds to lower economic output. The gap between the old and new equilibrium output can be understood as the social cost of emission reduction. Introduction of climate-friendly technologies can help reduce the social cost of emission reduction, thereby lifting the equilibrium growth path. Climate-friendly technologies are driven by investment, which is composed of corporate investment and carbon fee (from carbon tax or carbon trading) reutilization. More investment in climate-friendly technologies can help companies reduce the additional carbon cost to a greater extent, minimizing the negative impact of carbon tax/carbon trading on output.

1.3.1 Model Data

We choose 2017 as the base year. The main data sources of the model include the following: (1) China’s social accounting matrix (SAM). According to China’s 2017 input-output tables of 149 sectors, we obtain the energy input-output tables of 29 sectors, including coal, oil, natural gas, thermal power, and renewable energy through merge and split. The fiscal and tax data in the SAM are derived from the 2018 China Taxation Yearbook and 2018 Finance Yearbook of China. (2) Exogenous elasticity of substitution, such as the elasticity of substitution between inputs in the production function, the elasticity of substitution between imported and domestic products in the CES function of the foreign trade module, etc. The data makes reference to the Global Trade Analysis Project (GTAP) database. (3) Carbon dioxide emission coefficients derived from calculating sectors’ carbon dioxide emissions and energy product consumption. The carbon dioxide data come from the CEADs.

1.3.2 Scenario Assumptions

In order to quantify the different effects of the aforementioned two paths in terms of carbon neutrality and economic development, we examine the following four scenarios:

• 1. Business-as-usual scenario (BAU): No constraint is imposed on carbon emissions. The economic growth rate refers to the forecast made by the CICC macro team based on the proposal to double China’s economic aggregate or per capita income by 2035.

• 2. Carbon tax scenario (M1): A unified carbon tax is imposed on carbon-emission industries so that China can strive to achieve the goals of carbon peaking by 2030 and neutrality by 2060, and keep peak emissions as close to 10.8bn tonnes as possible.

• 3. Carbon trading scenario (M2): Carbon trading is introduced for 8 high-emission industries. Free carbon allowances are set based on the green premium of industries—free allowances should be lower for industries with higher green premiums to strengthen their incentives to cut emissions. Thus, China can strive to achieve the goals of carbon peaking by 2030 and neutrality by 2060, and keep peak emissions as close to 10.8bn tonnes as possible.

• 4. Carbon tax + carbon trading + technological progress scenario (M3): A three-pronged approach is adopted to achieve the goals of carbon peaking (at 10.8bn tonnes) by 2030 and neutrality by 2060, while maintaining economic growth.

1.3.3 BAU Scenario

In this scenario, there is no constraint on carbon emissions and the only goal is to double China’s economic aggregate or per capita income by 2035. The figures below depict the corresponding economic growth and carbon emission trajectory simulations: GDP will grow at an average annual rate of 5.2% from 2021 to 2030¹²; carbon emissions will be 15.8bn tonnes in 2030, and after deducting 900mn tonnes of carbon sequestration, net emissions will be 14.9bn tonnes.

1.3.4 Carbon Tax Scenario (M1)

In theory, as long as China keeps raising carbon tax, the country can achieve the goals of carbon peaking by 2030 and neutrality by 2060, and keep peak emissions close to 10.8bn tonnes. However, a relentless increase in carbon tax could paralyze the economic system—in that case, the CGE model is unable to converge to an equilibrium path. This means that the setting of carbon tax needs to take into account both economic development and carbon neutrality goals. Based on the green premium calculated by our sector analysts, we derive the parity carbon cost¹³ of the whole society at about Rmb377/tonne. Countries that have levied a carbon tax (mostly developed countries) set the tax at Rmb80-800/tonne. Our basic materials analysts estimate that for industries to remain profitable, the maximum affordable carbon tax is Rmb100/tonne for steel and cement industries and Rmb60/tonne for the aluminum industry. Meanwhile, our calculations show that the CGE model cannot converge to an equilibrium path when carbon tax is higher than Rmb150/tonne.¹⁴ After considering all these factors, we set carbon tax at Rmb100/tonne. Later we will conduct a sensitivity analysis of carbon tax in a bid to identify what would be an extreme level of carbon tax and the time path of carbon tax rate change.

In this scenario, carbon emissions continue to increase and cannot be peaked, let alone achievement of carbon neutrality. Specifically, net emissions will reach 12.3bn tonnes in 2030¹⁵ and 33.4bn tonnes in 2060 (see Fig. 1.6), but still 2.6 and 16.6bn tonnes lower than the BAU scenario. GDP will decline by 0.6% in 2030 and 0.9% in 2060 compared to the BAU scenario. Prices are under greater upward pressure than in the BAU scenario, especially in the first few years. Wholesale & retail, restaurant & hotel, transportation, and information & financial service industries will face higher price pressures in the next few years (see Fig. 1.7). The imposition of a carbon tax will also bring about changes in the economic structure. The output of the power sector excluding thermal power in 2030 will be more than 2% higher than in the BAU scenario, while the output of coal, thermal power, natural gas extraction, and building materials industries will be under pressure (the output of coal processing and mining industries will be over 7% lower than in the BAU scenario).

Fig. 1.6

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

GDP loss in the carbon tax scenario compared to the BAU scenario.

Fig. 1.7

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

Industries with higher cost pressures in the carbon tax scenario.

1.3.5 Carbon Trading Scenario (M2)

We assume that only the petrochemical, chemical, building materials, steel, nonferrous metals, paper, power, and airline industries will be engaged in carbon trading in the unified carbon market, and we determine the free emission allowance for each industry based on the green premium calculated by our sector analysts and Guangdong province’s 2020 Carbon Emission Allowance Allocation Implementation Plan. Free allowances should be lower for industries with higher green premiums to increase their incentives to switch to zero-emission technologies. The nonferrous metals, petrochemical, and paper industries with lower green premiums have a free allowance of 90%; the steel, power, chemical, and airline industries have a free allowance of 80%; and the building materials industry with the highest green premium has a free allowance of 70%.¹⁶

In this scenario, carbon emissions would peak in 2042 at 17.1bn tonnes, before declining to 13bn tonnes in 2060 (see Fig. 1.8). Although this scenario still does not result in carbon neutrality, it represents a considerable improvement compared to the carbon tax scenario. GDP performance in this scenario is also better than the carbon tax scenario. Compared to the BAU scenario, GDP would only decrease by an average of 0.15% per year in the next few decades. Similar to carbon tax, carbon trading would also bring inflationary pressure, especially in the first few years. The thermal power, building materials, chemical, steel, and airline industries would face greater price pressures (Fig. 1.9). In terms of industry structure, the proportion of the power sector excluding thermal power in 2030 would be much higher than in the BAU scenario, while of coal processing and mining industries would be 10% and 5.5% lower than in the BAU scenario.

Fig. 1.8

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

GDP loss in the carbon trading scenario compared to the BAU scenario.

Fig. 1.9

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

Industries with higher cost pressures in the carbon trading scenario.

In the computable general equilibrium model, factor prices are jointly determined by the forces of supply and demand in the market, and carbon trading prices are no exception. The parity carbon cost implied by the green premium measures the relative cost of emission-generating technology to zero-emission technology. When the carbon trading price is close to this cost, theoretically effective emission reduction can be achieved. According to the analysis in the M2 scenario, assuming that a carbon trading mechanism is introduced in 2021, the carbon price would rise from Rmb31/tonne at first to Rmb650/tonne in 2060. As mentioned above, the overall parity carbon cost implied by the green premium is Rmb377/tonne. Simulation results show that the carbon trading price in the next 20 years will be much lower than the parity carbon cost, which may be an important reason why emissions cannot be peaked in time. When the carbon price approaches or even exceeds the parity carbon cost, emissions are significantly reduced. This means that raising the carbon price, especially when it is raised to near the parity carbon cost, can effectively reduce emissions.

1.3.6 Carbon Tax + Carbon Trading + Technological Progress Scenario (M3)

As mentioned above, while carbon pricing (in the form of a carbon tax or a carbon market) is the mainstream way to solve the externality of emissions, carbon pricing alone cannot achieve a balance between GDP growth and carbon neutrality. In other words, if the GDP growth target is not abandoned, it will be difficult to achieve carbon neutrality through carbon pricing alone. Therefore, we need to shift from the mainstream path of carbon pricing to the green premium analysis framework. It is necessary to not only increase the production cost of emission-generating technologies through carbon pricing, but also reduce the production cost of zero-emission technologies through technological progress.

For the analysis of technological progress, we adopt the “S-curve” theory of technological innovation. Freeman and Louçã (2001) proposed a long wave (or Kondratieff Wave) theory of technological revolution from a historical perspective. The life cycle of any widely used technology consists of six phases: (1) the laboratory/invention phase, (2) decisive demonstration, (3) explosive growth, (4) continued high growth, (5) slowdown, and (6) maturity. The whole process forms an “S-shaped curve”, which may be driven by the investment path, according to Köhler (2005). This S-shaped curve is widely accepted as a way to describe technological change. In our model, this S-shaped curve in a reduced form encompasses all climate-friendly technologies: energy efficiency technologies, emission reduction technologies, carbon sequestration, carbon capture and storage, etc.

Take the power industry that accounts for the bulk of carbon emissions as an example. Using clean energy (such as solar energy) for power generation on the supply side and increasing the electrification rate on the demand side is the best technological path to achieve carbon neutrality. Solar power generation technology is mature and has been successfully commercialized. On the other hand, grid absorption still faces challenges given that energy storage has not yet reached parity. For power transmission, it is necessary to establish distributed smart grids and change the power structure. Breakthroughs must be made in infrastructure technologies that enable the transition from a centralized grid to a smart grid and the realization of full electrification. On the application side, the fundamental solution is to comprehensively increase the electrification rate (electric vehicles, electric arc furnaces for steelmaking, etc.). In Fig. 1.10, the vertical axis represents the extent of application or penetration rate of technology over time, and the S1 curve depicts the technology path of the power industry towards carbon neutrality. We are now at a stage where the technology curve gradually begins to steepen.

Fig. 1.10

Source Centre for Economic Systems Simulation Research, CICC Research

Climate-friendly technology advancement curve.

Compared to the technology path represented by the S1 curve, the S2 curve represents a new generation of nuclear power technology, hydrogen energy technology, and carbon capture technology. These technologies are still in the laboratory or demonstration phase and are far from successful commercialization, meaning great uncertainty. We combine the S1 and S2 curves to derive the S curve, which represents all climate-friendly technologies. We define it as a share of non-carbon technology (SNCT) curve. The faster the curve begins to steepen, the faster the technology is commercialized; the steeper the curve is, the faster the large-scale application of the technology will be.

The SNCT curve measures the extent of decline in the carbon emission coefficient of the economic system with the advancement of climate-friendly technologies. Here, we use a curve to simulate the entire life cycle of climate-friendly technologies. This curve is determined by five exogenous parameters (to construct the curve shape and simulate technology life cycle, growth rate, etc.) plus an endogenous variable (climate-friendly technology investment). We assume that taking actions to address climate change will lead to a transfer of investment from traditional energy technologies to climate-friendly technologies, and this can effectively increase the share of climate-friendly technologies. In this sense, this curve in a reduced form summarizes a series of investment-driven technological advancements from improving energy efficiency to decarbonizing and reducing carbon (such as carbon capture and carbon fixation). Therefore, this curve explains how to lower the carbon emission coefficient and thus reduce emissions.

In this scenario, we adopt the same carbon tax of Rmb100/tonne as in the M1 scenario and the same carbon allowances for the 8 high-emission industries as in the M2 scenario. Through simulation, we found that the combination of carbon tax, carbon trading, and technological progress can achieve the dual goals of economic growth and carbon neutrality. The GDP loss compared to the BAU scenario is almost negligible. After 2040, GDP even exceeds that in the BAU scenario, indicating that technological progress offsets or even more than offsets the negative impact of carbon pricing on economic growth. In this scenario, carbon emissions peak in 2030 at 10.9bn tonnes, before decreasing year by year towards the goal of net zero emissions (see Fig. 1.11).

Fig. 1.11

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

GDP in the M3 scenario compared to the BAU scenario.

The simulation results also show that carbon tax and carbon trading still put upward pressure on costs in the economy in the early stages, possibly because the effect of technological progress has not fully kicked in. However, the cost pressure is milder than in the scenarios without technological progress (see Fig. 1.12 for industries with higher cost pressures in the M3 scenario). We analyze changes in the industrial structure from two perspectives: (1) the change in industry output compared to the BAU scenario, and (2) the change in the proportion of industry output over time. First, the output of the solar power, hydropower, wind power, and nuclear power industries in 2030 all increase more than 3% compared to the BAU scenario, while the output of coal processing and mining industries shrink considerably. Second, the proportions of output of agriculture, public services, and light industries show more obvious declines over time, while the proportions of output of equipment manufacturing, real estate & leasing, and information & financial services industries increase year by year.

Fig. 1.12

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

Industries with higher cost pressures in the M3 scenario.

The carbon trading price will rise gradually from Rmb10/tonne in 2021 to Rmb140/tonne before 2030. At a time when climate-friendly technologies have not been adopted on a large scale, rising carbon prices will help peak emissions in a timely manner and at around 10.9bn tonnes. When green technologies are widely used, carbon prices will drop. As an endogenous variable of the S-shaped technology curve, investment is the key to driving technological progress. The model estimates that R&D investment in green technologies will account for about 2% of GDP, and it will increase gradually in the first few years to help bring technologies from the laboratory phase into commercial use. In the M3 scenario, the climate-friendly technology curve SNCT simulates the shape of the S curve. The SNCT will reach 0.32 in 2030 and 0.99 in 2060, meaning that the carbon emission coefficient¹⁷ in 2030 and 2060 will be 68% and 1% respectively of that in 2020 (see Fig. 1.13). As such, China will be able to strike a balance between the existing economic growth objective and the new carbon neutrality goal. The energy structure will also be optimized gradually, with clean energy accounting for 44% in 2030 and 96% in 2060.

Fig. 1.13

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

Carbon trading price and Climate-friendly technology advancement curve in the M3 scenario.

1.3.7 Sensitivity Analysis of Technology Curve

A comparative analysis of the above-mentioned four scenarios shows that technological progress is essential to balance economic growth and carbon neutrality goals. However, since technological innovation is highly uncertain, we need to examine the impact of an optimistic/pessimistic scenarios for technology on GDP and carbon emissions. We assume that the climate-friendly technology curve in the M3 scenario is the baseline scenario of technological progress, and the pessimistic scenario is a situation in which technological development is disappointing or commercialization is difficult. Intuitively, the technology curve in the pessimistic scenario should be flatter than in the baseline scenario: the laboratory phase may be longer, large-scale application of technology may face greater resistance, or the penetration rate of technology may be lower. In reality, the pessimistic scenario may correspond to technologies that are not yet mature and quite uncertain, such as hydrogen energy and carbon capture. The opposite is true for the optimistic scenario.

Consistent with intuition, GDP is considerably affected in the pessimistic technological scenario (Fig. 1.14), and the carbon emission curve is significantly elevated—emissions do not peak until 2034 and the peak is as high as 13.4bn tonnes (Fig. 1.15). The simulated technology curve is also in line with our thinking about the technology path in the pessimistic scenario. Due to disappointing technological progress, the share of clean energy is relatively low and the carbon trading price is much higher after carbon emission peaks. In addition, compared with baseline and optimistic scenarios, the carbon trading market lasts longer in the pessimistic scenario—the price mechanism needs to play a greater role in both price and time dimensions.

Fig. 1.14

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

GDP changes in different technological scenarios compared to the BAU scenario.

Fig. 1.15

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

Carbon emission simulation.

1.3.8 Sensitivity Analysis of Carbon Tax

According to the simulation of the M1 scenario, it is difficult to peak carbon emissions at a carbon tax that can achieve model convergence. As mentioned earlier, as long as the carbon tax rate is high enough (regardless of economic growth), it can correct the negative externality of carbon emissions and turn the emission curve down at a certain point. In the M1 scenario, emissions fail to peak before 2060, illustrating two points: (1) the calculation time span is not long enough; or (2) the carbon tax rate used in the model is too low. Here, we set a goal of peaking carbon emissions by 2060¹⁸ and calculate how much carbon tax is required. After simulation calculations, we derive a carbon price index (see Fig. 1.16). This curve roughly follows an exponential increase, from Rmb46/tonne in 2021 to Rmb140/tonne in 2030, and then to Rmb3,300/tonne in 2060. With such a high carbon tax, emissions would enter a plateau phase in 2040 and begin to slowly decline in 2052. Economic growth is also under great pressure, with a GDP loss of 3%–5% in the final 10 years compared to the BAU scenario (Fig. 1.17). This simulation further confirms that carbon tax alone cannot enable China to achieve the goals of carbon peak by 2030 and neutrality by 2060.

Fig. 1.16

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

Extreme carbon tax rate and carbon emission simulation.

Fig. 1.17

Source CEADs, Centre for Economic Systems Simulation Research, CICC Research

GDP change compared to the BAU scenario.

1.4 Road to Carbon Neutrality = Technology + Carbon Pricing + Social Governance

The aforementioned general equilibrium analysis based on the CGE model shows that internalization of the negative externality of carbon emissions alone could not solve the conflict between economic growth and carbon neutrality goals, and only a combination of technology and carbon pricing under the green premium framework could help strike a balance between the two goals. Reaching such a conclusion does not mean the end of the research in achieving carbon neutrality. Instead, it marks the starting point of exploring China’s road to carbon neutrality. For example, given that technology is so important to achieve carbon neutrality, what carbon neutral technologies are worth nurturing? What is the current development status of these technologies? We will discuss carbon neutral technologies in more detail in the following chapters.

In terms of carbon pricing, the establishment of a unified national carbon market is accelerating, and the 8 high-emission industries may gradually be included in the unified carbon market during the 14th FYP period. As China promotes carbon neutrality, the implementation of a unified carbon price for different industries through the establishment of a unified market seems to have become a consensus in the country. However, our analysis shows that the situation may not be that simple. For example, when internalizing the negative externality of emissions, it may not be appropriate to calculate the discounting of social costs according to mainstream practices, and a differentiated pricing perspective based on net social costs should be adopted instead; the analysis of green premium shows that it may not be appropriate to choose a carbon pricing method based solely on the proportion of emissions; also, a unified carbon market may exacerbate pollution problems in the Beijing-Tianjin-Hebei region. For detailed analysis on these issues, please refer to a later chapter that directly addresses carbon market, carbon tax and other carbon pricing methods.

More importantly, technology and carbon pricing are just two basic means to reducing the green premium, and they could not represent a comprehensive solution to carbon neutrality under this framework. According to the calculation of our sector analysts, the green premium of the power industry will be −3.9% in 2030. A negative green premium means that in terms of economic performance, zero-emission technologies would be decisively superior to emission-generating technologies. However, due to the path dependence of economic development, or a lack of sufficient awareness of the importance of carbon neutrality among decision makers or the general public, industries may not shift to more economical zero-emission production technologies, and may instead stick with high-cost emission-generating technologies. Such problems cannot be resolved by relying solely on “technology + carbon pricing”. It also requires the intervention of social governance or public policy.

From the perspective of social governance, there are mainly three types of policies, i.e., price-based, command-based and publicity-based policies. In fact, technology, carbon pricing, and social governance, which together constitute the road to carbon neutrality, are closely correlated with each other. For example, technological progress and improvement in production processes cannot be achieved without support from public policies such as fiscal subsidies. Carbon tax and auctions in the carbon market are important sources of such funds. Carbon pricing itself is part of public policies, and publicity-based policies could help raise people’s awareness of carbon neutrality, and thus reduce friction costs in the implementation of carbon pricing policies. In view of the green premium analysis framework, the “technology + carbon pricing + social governance” approach is important to the road to carbon neutrality in three ways.

First, the approach addresses issues related to innovation and fairness on the road to carbon neutrality. The green premium only measures the difference in production cost between zero-emission technologies and emission-generating technologies, while it does not measure the total cost borne by the society for carbon neutrality. The simulation calculations based on the CGE model show that the output of coal, building materials, and chemical industries may shrink considerably, while the proportions of output of information & financial services and real estate & leasing industries may expand. Such structural changes may cause a series of social problems. For example, the large-scale launch of zero-emission power generation capacity will inevitably lead to large-scale idling of thermal power assets, as well as subsequent layoffs at coal mines and thermal power plants, and an increase in non-performing loans at banks. Transformation will be required for coal-producing provinces. In a more general sense, the reliance on innovation to achieve carbon neutrality may lead to unfair development between high-carbon industries, workers and regions, and low-carbon industries, workers, and regions. Such problems may not be resolved by relying solely on the “technology + carbon pricing” approach, and coordinated arrangements and early planning from the perspective of social governance are needed. Whether the innovation- and fairness-related issues can be well addressed determines whether the entire society can continue to move towards carbon neutrality.

Second, the “technology + carbon pricing + social governance” approach is vital to coping with risks of rebound in the green premium. The large-scale adoption of zero-emission technologies may result in shrinking demand for emission-generating technologies. For example, the large-scale launch of clean energy power generation capacity will affect the demand for thermal power generation, thereby reducing fossil energy prices and production costs of emission-generating technologies. Such a phenomenon may lead to a rebound in green premiums. If the demand for fossil energy continues to shrink, the decline in prices will likely also lead to spontaneous supply-side contraction. Therefore, with large-scale promotion of zero-emission technologies, fossil energy prices may face both upward and downward pressures. It is difficult to predict which factor will dominate, but one thing is certain: the change in the green premium will be a dynamic process, and the green premium may rebound in the downward trend, which may even weaken the momentum around reaching carbon neutrality in society as a whole. Therefore, it is necessary to restrict the re-expansion of fossil energy-based power generation capacity through command-based policies when it is economically feasible to adopt zero-emission technologies on a large scale.

Third, “technology + carbon pricing + social governance” approach puts emphasis on the role of green finance. As we mentioned earlier, the estimated green premium of the power industry is negative in 2030. This means that even with carbon pricing and technological progress, the economically advantageous clean power generation technologies will not fully replace thermal power generation technologies in the next 10 years. This may be because technologies still need to be improved or carbon prices are not high enough. However, it also reveals a social governance problem, i.e., how to formulate reasonable public policies to promote the formation of economically beneficial zero-emission production capacity as soon as possible so as to accelerate the replacement of existing emission-generating assets. We believe green finance will play an important role in this regard. In fact, the significance of green finance to carbon neutrality lies not only in accelerating the promotion of mature technologies, but also in the highly uncertain R&D phase. For example, carbon neutral technologies that are still in the laboratory phase such as hydrogen energy and carbon capture still require a large amount of R&D funding. Do carbon neutral technologies require different forms of financial support in the two different stages of development? What are the implications of green finance for carbon neutrality? We will examine these topics in greater detail in Chapter 4.

The above thoughts are based on the assumption of a closed economy. As China emphasizes the new dual-circulation pattern with domestic and foreign markets reinforcing each other, it is also necessary to take an international perspective when dealing with carbon emissions, an externality that requires long-term and comprehensive efforts. On March 10, 2021, the European Parliament passed a resolution on the EU’s Carbon Border Adjustment Mechanism (CBAM). According to the mechanism, if economies that have trade relations with the EU do not comply with the EU’s carbon emission regulations, the EU can impose carbon tariffs on goods imported from them. This can be regarded as the first step towards the implementation of Nordhaus’ international “Climate Club” proposal. It is also a direct application of the carbon tax-based pricing mechanism at the governance level of the international community. What are the implications of the “technology + carbon pricing + social governance” approach from an international perspective and how will it play its role? We will analyze these in the chapter that focuses on international cooperation.