Chapter 4: The Dual Carbon Compass: Policy Architecture, Carbon Pricing, and the Incentive Landscape for Manufacturing

On the morning of July 16, 2021, a screen at the Shanghai Environment and Energy Exchange displayed the first trades of China's national emissions trading system. The opening price was 48 yuan per ton of carbon dioxide equivalent. By the close, 4.1 million tons had changed hands, worth about 210 million yuan. Officials from the Ministry of Ecology and Environment (MEE) and the Shanghai municipal government watched from a modest ceremony. For a country producing roughly a quarter of global carbon emissions, the moment carried symbolic weight. The number told a more complicated story. Forty-eight yuan per ton was less than one-seventh the price of a ton of carbon in the European Union's emissions trading system at the time, which hovered around 55 euros. The gap was deliberate. A cap set generously above current emission levels, free allocation for the power sector, and a trading mechanism built to establish a carbon price signal without imposing immediate cost shocks on industry: these were design choices, not accidents.

That gap, and the political economy behind it, is the subject of this chapter. The dual carbon goals—peak carbon emissions before 2030 and carbon neutrality before 2060—set a direction of travel for the world's largest manufacturing economy. A direction is not a policy. Between the announcement of the goals in September 2020 and the actual behavior of firms, there stands an architecture of regulations, pricing instruments, subsidies, standards, and financial incentives. That architecture determines who bears the costs and who gains. It also determines whether green governance capability becomes a genuine competitive advantage for Chinese manufacturers or a compliance exercise with a rising price tag.

The chapter maps this architecture in three parts. The first section describes the policy framework: the institutions, targets, and regulatory structures that constitute the dual carbon goals. The second examines carbon pricing, from the national ETS to pilot programs and the long-debated carbon tax. The third turns to the incentive landscape—the subsidies, standards, and financial instruments that shape firm-level decisions. Throughout, the focus is on manufacturing, because manufacturing is where the dual carbon goals meet the global value chain. The aim is not to catalog every policy document but to explain how the incentive landscape works, and where it fails to work, as a compass for industrial upgrading.

4.1 Policy Architecture of the Dual Carbon Goals

President Xi Jinping announced the dual carbon goals at the United Nations General Assembly on September 22, 2020, in a pre-recorded video address. The phrasing was specific: China would "strive to peak carbon dioxide emissions before 2030 and achieve carbon neutrality before 2060." The first target had been floated in various forms since 2014. The second was a substantial escalation. Carbon neutrality by 2060 meant that a country still building coal-fired power plants and producing more than half the world's steel would need to reduce net emissions to zero within four decades. The announcement took many observers by surprise, including some within China's own policy apparatus. The question that followed was less about the target than about the machinery: what institutional arrangements would translate a stated goal into measurable change?

Over the next two years, the answer took shape as a layered system combining top-level coordination, sectoral targets, and local implementation. Chinese policy documents call it the "1+N" policy system. The "1" is the guiding document, issued by the Central Committee of the Communist Party and the State Council in October 2021, which sets out the overarching philosophy and implementation path. The "N" refers to sectoral and regional implementation plans covering energy, industry, transport, construction, agriculture, and others. By the end of 2022, more than thirty such plans had been issued. The structure resembles the five-year planning system China has used for decades, adapted to a new set of objectives.

The top-level coordination mechanism is the Leading Group on Carbon Peaking and Carbon Neutrality, established in May 2021 and chaired by Vice Premier Han Zheng. The group resolves inter-ministerial conflicts and sets priorities. Below it, responsibilities are distributed across ministries. The National Development and Reform Commission (NDRC) handles overall planning and energy policy. The Ministry of Ecology and Environment manages carbon markets and emissions monitoring. The Ministry of Industry and Information Technology (MIIT) oversees industrial green development. The People's Bank of China (PBOC) coordinates green finance. The National Energy Administration (NEA) manages the power sector. This distribution creates both capacity and friction. Each ministry has its own priorities, its own data systems, and its own relationships with state-owned enterprises and local governments. The Leading Group exists to manage the friction, but friction is inherent in the structure.

The sectoral plans translate the national goals into specific targets. The energy plan, issued by the NDRC and NEA in early 2022, sets a target of 1,200 gigawatts of installed wind and solar capacity by 2030, up from about 635 gigawatts at the end of 2021. The industrial plan, issued by MIIT, calls for a 13.5 percent reduction in energy consumption per unit of industrial value added by 2025 and an 18 percent reduction in carbon emissions per unit of industrial value added over the same period. The construction plan targets green building standards for new construction. The transport plan promotes electric vehicles and rail freight. Each plan includes quantitative targets, implementation measures, and reporting requirements. The quantitative targets are the most visible part of the architecture, and the most easily gamed. A target for energy intensity—energy consumption per unit of GDP—can be met by changing the composition of output rather than by improving efficiency. A target for carbon intensity—emissions per unit of GDP—can be met by the same route. The plans are aware of this problem, which is why they increasingly include absolute targets alongside intensity targets. Absolute targets are harder to enforce, because they require measuring total emissions, not just ratios.

Measurement is where the architecture is weakest. China's emissions accounting system relies on a combination of self-reporting by enterprises, statistical estimates by government agencies, and satellite data. The self-reporting system covers roughly 8,000 enterprises in the power sector and, as of 2024, expanded to include steel, cement, and aluminum. These enterprises report their emissions annually to provincial ecology and environment bureaus, which aggregate the data and submit it to the MEE. The reporting requirements are detailed: enterprises must account for emissions from fuel combustion, industrial processes, and purchased electricity and heat. They must also report production data, which allows the calculation of emissions intensity. Third-party auditors verify the data, and the MEE oversees their qualifications and performance. In principle, the system is rigorous. In practice, it has been plagued by problems of data quality, verification capacity, and incentives to underreport.

Data quality problems are not unique to China. Every emissions trading system faces the challenge of ensuring that reported emissions are accurate. The EU ETS, operating since 2005, has a sophisticated verification system and a track record of penalizing fraud. China's system is newer and less mature. A 2022 investigation by the MEE found that several enterprises in the power sector had underreported their emissions by significant margins, in some cases by more than 20 percent. The investigation led to fines and to revisions in the reporting guidelines. The revisions included stricter requirements for coal sampling and testing, more frequent verification, and a blacklist system for auditors who failed to catch errors. Any system that relies on self-reporting and third-party verification will face pressure from enterprises to minimize reported emissions, especially when those emissions carry a price.

The verification capacity problem is more specific. China has a limited pool of qualified verifiers. The MEE has approved several hundred verification agencies, but their quality varies widely. Many are small consultancies with limited technical capacity. The verification process is often rushed, especially in the months before the compliance deadline. The MEE has tried to address this by training verifiers and conducting spot checks, but the fundamental constraint is the shortage of people with the skills to audit complex industrial processes. Verification agencies are paid by the enterprises they audit, which creates an obvious conflict of interest. The EU ETS addresses this by requiring that verifiers be accredited by national accreditation bodies and by allowing the regulator to suspend verifiers who fail to meet standards. China is moving in this direction, but the transition is incomplete.

Local implementation is perhaps the most important problem. China's climate targets are set nationally, but implementation is local. Provincial and municipal governments enforce regulations, allocate permits, and report data. Their incentives are not always aligned with the national goals. Local officials are evaluated on a combination of economic growth, social stability, and environmental performance. The weight given to environmental performance has increased in recent years, but economic growth remains a priority, especially in provinces that depend on heavy industry for employment and tax revenue. The result is uneven implementation. Some provinces, particularly those in the east with more diversified economies, have moved aggressively on carbon reduction. Others, particularly those in the north and west that depend on coal and heavy industry, have moved more slowly. The central government has tried to address this by including carbon intensity in the cadre evaluation system and by conducting central environmental inspections. The inspections, which began in 2016, have been effective at identifying problems and forcing local governments to act. They are periodic, and their impact depends on the political priority attached to them at any given time.

The power sector is the center of the architecture. Electricity generation accounts for roughly 27 percent of global greenhouse gas emissions, making it the single largest emitting sector. In China, the share is higher—about 40 percent of national emissions—because the power sector is dominated by coal. The dual carbon goals therefore hinge on decarbonizing electricity. The strategy has three components: replacing coal-fired generation with renewables, improving the efficiency of existing coal plants, and building a grid that can absorb variable renewable energy. The first component is proceeding rapidly. China installed more solar and wind capacity in 2023 than the United States has installed in its entire history. The second component is more contentious. Coal plants are being retrofitted for flexibility, which allows them to ramp up and down more quickly to balance the grid. Retrofitting does not reduce emissions; it makes coal plants more compatible with renewables. The third component—grid modernization—is the most difficult. China's grid is divided into regional grids that are not fully interconnected. Moving electricity from sunny and windy regions in the west to load centers in the east requires new transmission lines, which take years to build and face local opposition.

Manufacturing accounts for about 30 percent of China's emissions, with steel, cement, and aluminum accounting for the bulk of industrial emissions. The dual carbon goals require these industries to reduce emissions intensity and, eventually, absolute emissions. The policy tools include energy efficiency standards, fuel switching, process substitution, and carbon pricing. The effectiveness of these tools varies by industry. Steel can be decarbonized through electric arc furnaces powered by green electricity or through hydrogen-based direct reduced iron. Cement can be decarbonized through clinker substitution and carbon capture. Aluminum can be decarbonized through inert anode technology and green electricity. Each pathway requires different investments, different technologies, and different policy support. The industrial plans recognize this diversity, but implementation is uneven. Some industries have clear pathways and strong policy support; others do not.

The architecture also includes cross-cutting policies that affect all sectors. The most important is the national carbon market, discussed in the next section. Others include the energy consumption dual control system, which caps total energy consumption and energy intensity; the green finance system, which provides subsidized loans for green projects; and the green certificate system, which allows renewable energy producers to sell certificates to consumers who want to claim green attributes. These policies interact in complex ways. The dual control system can constrain the growth of energy-intensive industries, but it can also penalize firms that are growing rapidly and therefore need more energy. The green finance system can lower the cost of capital for green projects, but it can also create bubbles if the standards for what counts as green are too loose. The green certificate system can provide additional revenue for renewable energy producers, but it can also fragment the market if different certificate systems are not harmonized.

The architecture has evolved significantly since the dual carbon goals were announced. The initial version, outlined in the 2021 guiding document, was relatively general. Subsequent documents have added detail and specificity. The 2022 implementation plan for energy conservation and emission reduction set specific targets for each province. The 2023 revision of the emissions trading regulations strengthened penalties for non-compliance. The 2024 expansion of the carbon market to steel, cement, and aluminum brought new sectors into the pricing system. Each revision has been driven by a combination of learning from experience and responding to external pressure. The learning is evident in the technical details: the reporting guidelines have become more precise, the verification requirements more stringent, the allocation methods more sophisticated. The external pressure comes from international partners, who want China to strengthen its climate commitments, and from domestic constituents, who want cleaner air and water.

The architecture's effectiveness is difficult to assess because the goals are long-term and the data is incomplete. Some trends are clear. China's carbon intensity—emissions per unit of GDP—has declined steadily since 2005, when the government first set a carbon intensity target. The decline has been driven by structural change (a shift from industry to services), energy efficiency improvements, and fuel switching. The absolute level of emissions, however, has continued to rise, albeit more slowly in recent years. The peak is expected to occur sometime between 2025 and 2030, depending on the pace of economic growth and the effectiveness of policy. The neutrality target for 2060 is far more ambitious and will require emissions reductions plus carbon removal from the atmosphere through afforestation, soil carbon sequestration, and carbon capture and storage. The technologies for removal are not yet deployed at scale, and their costs are uncertain. The architecture includes provisions for research and development, but the gap between current policy and the 2060 target is large.

The policy architecture's most important feature, for the purposes of this book, is its effect on firm-level incentives. The architecture does not directly tell firms what to do. It creates a set of prices, standards, and subsidies that shape the calculus of investment. A firm that faces a carbon price will compare the cost of paying the price with the cost of reducing emissions. A firm that faces an energy efficiency standard will compare the cost of compliance with the cost of non-compliance. A firm that faces a subsidy for green technology will compare the benefit of the subsidy with the cost of adopting the technology. The architecture's effectiveness depends on whether these comparisons lead firms to invest in green governance capability. That is the question to which the rest of this chapter turns.

4.1.1 The Institutional Map: Who Decides What

The institutional map of China's climate policy is more complex than the "1+N" label suggests. At the apex is the Central Committee's Leading Group on Carbon Peaking and Carbon Neutrality, which meets irregularly and sets broad direction. Below it, the State Council's Carbon Peak and Carbon Neutrality Leading Group, chaired by the Vice Premier, coordinates implementation. The NDRC serves as the secretariat and the lead agency for overall planning. The MEE is responsible for carbon markets, emissions monitoring, and climate change adaptation. The MIIT is responsible for industrial green development. The NEA is responsible for the power sector. The PBOC is responsible for green finance. The Ministry of Finance (MOF) is responsible for fiscal incentives. The Ministry of Science and Technology (MOST) is responsible for research and development. The National Bureau of Statistics (NBS) is responsible for data.

This distribution creates a coordination problem. Each ministry has its own budget, its own staff, and its own relationships with local governments and state-owned enterprises. The NDRC can set targets, but the MEE enforces them. The MEE can allocate permits, but the NEA determines how much coal is burned. The MOF can provide subsidies, but the PBOC determines whether green loans are available. The coordination problem is managed through a series of inter-ministerial committees and working groups. The Leading Group is the most important, but it meets only a few times a year. Day-to-day coordination is handled by the NDRC's Department of Resource Conservation and Environmental Protection, which has a staff of several dozen people. That staff is responsible for drafting plans, monitoring progress, and resolving disputes. It is a small team for a large task.

Each province has a Development and Reform Commission, an Ecology and Environment Bureau, an Industry and Information Technology Bureau, and an Energy Bureau. These agencies mirror the national ministries, but their authority and capacity vary widely. In wealthy provinces like Guangdong and Zhejiang, the agencies are well-staffed and technically competent. In poorer provinces like Shanxi and Inner Mongolia, they are understaffed and often lack the technical skills to monitor emissions or enforce regulations. The variation tracks resources, and it also tracks political priorities. Some provincial leaders have made carbon reduction a signature issue. Others have treated it as a compliance exercise. The central government has tried to address this by including carbon intensity in the cadre evaluation system and by conducting central environmental inspections. The inspections, which began in 2016, have been effective at identifying problems and forcing local governments to act. They are periodic, and their impact depends on the political priority attached to them at any given time.

The institutional map also includes a set of quasi-governmental organizations that play important roles. The China Emissions Exchange, based in Beijing, and the Shanghai Environment and Energy Exchange are the two main trading platforms for carbon allowances. The China Carbon Registry, based in Wuhan, records ownership of allowances. The National Center for Climate Change Strategy and International Cooperation (NCSC), a think tank affiliated with the MEE, provides technical support for policy design. The Chinese Academy of Environmental Planning (CAEP), another MEE-affiliated think tank, conducts research on emissions accounting and policy evaluation. These organizations are not decision-makers, but they shape the technical details of policy, and their expertise is often decisive in determining what is feasible.

State-owned enterprises complicate the map further. China's largest emitters are SOEs in the power, steel, cement, and aluminum sectors. These enterprises have their own internal planning processes, their own relationships with ministries, and their own access to finance. They are subject to the same regulations as private firms, but their size and political connections give them influence over policy design. The SOE sector is also where the government has the most direct leverage. The State-owned Assets Supervision and Administration Commission (SASAC) can set performance targets for SOEs, including carbon reduction targets. SASAC has used this authority to push SOEs to adopt green technologies and to report their emissions more accurately. The effectiveness of this pressure depends on the priorities of SASAC's leadership and on the willingness of SOE managers to comply. In some cases, SOEs have been leaders in adopting green technologies; in others, they have resisted changes that would reduce their profitability.

The institutional map is not a hierarchy in the strict sense. It is a network of organizations with overlapping mandates, competing priorities, and uneven capacity. The network is held together by the political priority attached to the dual carbon goals and by the personal authority of the leaders who champion them. When that priority is high, the network functions reasonably well. When it is low, coordination breaks down, and policies are implemented unevenly. The architecture's effectiveness therefore depends on its design and on the political will behind it.

4.1.2 Targets, Indicators, and the Problem of Measurement

The dual carbon goals are expressed as targets: peak before 2030, neutrality before 2060. Targets are useful for mobilizing action, and they are also problematic. A target is a number, and a number can be met in many ways. The question is whether the ways that are chosen actually reduce emissions or simply shift them around.

The measurement problem begins with the choice of indicator. China's climate targets are expressed in two main forms: carbon intensity (emissions per unit of GDP) and absolute emissions. The intensity target is easier to meet because GDP growth can dilute the effect of emissions growth. If emissions rise by 5 percent and GDP rises by 6 percent, carbon intensity falls by about 1 percent. The absolute target is harder because it requires emissions to peak and then decline. China's current targets include both: a carbon intensity reduction target for 2025 and a peaking target for 2030. The intensity target is binding; the peaking target is aspirational. The distinction matters because it determines how much pressure firms face. An intensity target allows emissions to rise as long as they rise more slowly than GDP. A peaking target requires emissions to stop rising altogether.

The measurement problem also involves the choice of accounting method. China's official emissions inventory is based on a combination of bottom-up reporting by enterprises and top-down estimates by government agencies. The bottom-up data is more accurate for individual facilities but is incomplete because many small emitters are not required to report. The top-down data is more comprehensive but is based on statistical estimates that can be inaccurate. The two methods are reconciled through a process of adjustment that is not always transparent. Different agencies can produce different estimates of the same emissions. The discrepancies are not large enough to undermine the overall trend, but they are large enough to complicate policy evaluation.

Emissions can be counted at the point of production (territorial emissions) or at the point of consumption (consumption-based emissions). China's targets are based on territorial emissions, which means that emissions from producing goods for export are counted in China, not in the importing country. This accounting choice has significant implications for trade policy. If the EU imposes a carbon border adjustment mechanism (CBAM) on imports from China, it will effectively charge Chinese exporters for emissions that are already counted in China's territorial inventory. The double counting is a source of friction in international climate negotiations. China has argued that consumption-based accounting would be fairer, because it would assign responsibility to the countries that consume the goods. The EU has argued that territorial accounting is simpler and more consistent with the UNFCCC framework. The dispute is unresolved, and it is likely to become more contentious as carbon prices rise.

The measurement problem extends to the firm level. A firm's carbon footprint depends on the boundary of its operations. If a firm owns its power plant, the emissions from that plant are counted in the firm's footprint. If the firm buys electricity from the grid, the emissions from the grid are counted in the firm's footprint only if the firm uses a location-based method. If the firm uses a market-based method, it can claim zero emissions if it purchases renewable energy certificates. The choice of method can change a firm's reported emissions by a factor of two or more. The GHG Protocol, the international standard for corporate emissions accounting, allows both methods but requires firms to report which one they use. In practice, many firms report only the method that makes them look better. The result is a landscape of corporate emissions data that is difficult to compare across firms and over time.

The measurement problem is not just technical. It is political. The choice of what to measure, how to measure it, and who gets to verify the measurement determines who bears the cost of carbon reduction. A firm that can define its boundary narrowly will have lower reported emissions and therefore lower costs. A firm that is required to report comprehensively will have higher reported emissions and higher costs. The politics of measurement is therefore the politics of cost allocation. The Chinese government has tried to address this by standardizing reporting requirements and by increasing the frequency of verification. Standardization is a slow process, and verification is expensive. The result is a system that is more rigorous than it was a decade ago but still far from perfect.

4.1.3 The Power Sector as the Pivot

The power sector is the pivot of China's climate policy because it is the largest source of emissions and the key to decarbonizing other sectors. If electricity is clean, then electric vehicles, electric heating, and electric industrial processes can be clean too. If electricity is dirty, then electrification simply shifts emissions from tailpipes and smokestacks to power plants. The dual carbon goals therefore depend on a rapid transition in the power sector.

The transition has three components: generation, grid, and markets. The generation component is the most visible. China has installed more renewable capacity than any other country. By the end of 2023, installed wind and solar capacity exceeded 1,000 gigawatts, up from about 100 gigawatts in 2015. The growth has been driven by a combination of government targets, subsidies, and falling costs. The subsidies, which were financed by a surcharge on electricity bills, were phased out for new onshore wind and solar projects in 2021. The phase-out was possible because the costs of renewable generation had fallen below the costs of new coal generation in many parts of the country. The falling costs, in turn, were driven by learning-by-doing in manufacturing. Chinese manufacturers now produce most of the world's solar panels and a large share of its wind turbines. The scale of production has driven down costs and improved quality.

The grid component is more difficult. China's renewable resources are concentrated in the west and north, while its load centers are in the east and south. Moving electricity from west to east requires long-distance transmission lines, which are expensive and take years to build. China has built several ultra-high-voltage (UHV) transmission lines, which can carry electricity over thousands of kilometers with low losses. The UHV network is not yet sufficient to absorb the full potential of renewable generation. The result is curtailment: renewable generators are sometimes forced to reduce output because the grid cannot absorb their electricity. Curtailment rates have fallen in recent years, but they remain significant in some regions. The grid also faces a stability problem. Renewable generation is variable, which means that the grid needs flexible resources to balance supply and demand. Coal plants can provide flexibility, but they are also the largest source of emissions. Battery storage can provide flexibility, but it is expensive. Pumped hydro can provide flexibility, but it is limited by geography. The grid of the future will need a mix of all three, plus demand response and interconnections with neighboring grids.

The market component is the least developed. China's electricity market is still in transition from a centrally planned system to a market-based system. The government sets wholesale prices, and generators are dispatched according to a merit order that prioritizes coal plants. The merit order is designed to ensure that the cheapest generation is dispatched first, but it does not account for the environmental costs of coal. The result is that coal plants are often dispatched ahead of renewable generators, even when renewable generation is available. The government has tried to address this by introducing a "green dispatch" mechanism, which prioritizes renewable generation. The mechanism is not fully implemented, and its effectiveness varies by province. The government has also tried to introduce a national carbon market, which would raise the cost of coal generation and make renewable generation more competitive. The carbon market's coverage is limited to the power sector, and its prices are low. The combination of a low carbon price and a coal-friendly dispatch order means that the market signal for decarbonization is weak.

The power sector's transition is also complicated by the role of coal. Coal is a fuel, and it is also a source of employment, tax revenue, and political power in provinces like Shanxi and Inner Mongolia. The coal industry employs hundreds of thousands of people and generates significant revenue for local governments. Closing coal mines and coal plants therefore has political consequences. The central government has tried to manage these consequences by providing transition support to coal-dependent regions and by retraining workers. The support is limited, and the transition is slow. The result is a tension between the national goal of decarbonization and the local goal of economic stability. The tension is managed through a combination of targets, subsidies, and political pressure. It is not resolved.

4.1.4 Industrial Policy and the Greening of Manufacturing

Manufacturing is where the dual carbon goals meet the global value chain. China's manufacturers produce goods for domestic consumption and for export. Their emissions are counted in China's territorial inventory, but their products are sold worldwide. The greening of manufacturing therefore has implications for China's climate targets, for its trade relationships, and for its position in global value chains.

The industrial policy response to the dual carbon goals has three main components: efficiency standards, process substitution, and circular economy. The efficiency standards are the most mature. China has a comprehensive system of energy efficiency standards for industrial equipment, buildings, and appliances. The standards are updated periodically, and they have been effective at reducing energy consumption per unit of output. The standards are enforced through a combination of mandatory labeling, market surveillance, and penalties for non-compliance. Enforcement is not perfect, but it has improved over time. The standards have also had a spillover effect on exports: Chinese manufacturers that meet domestic standards are often able to meet international standards, which gives them access to markets that require high efficiency.

Process substitution is more ambitious. It involves replacing carbon-intensive processes with low-carbon alternatives. In steel, the main alternative is electric arc furnace (EAF) steelmaking, which uses scrap steel and electricity instead of iron ore and coal. EAF steel has lower emissions than blast furnace steel, but its emissions depend on the carbon intensity of the electricity used. If the electricity is clean, EAF steel can be nearly zero-carbon. If the electricity is dirty, EAF steel can be as emissions-intensive as blast furnace steel. China's EAF share is about 10 percent of total steel production, compared with about 70 percent in the United States and 40 percent in the EU. The government has set a target of 15 percent EAF share by 2025 and 20 percent by 2030. The target is ambitious but not impossible. The main constraints are the availability of scrap steel, the cost of electricity, and the resistance of incumbent steelmakers who have invested in blast furnaces.

In cement, the main alternative is clinker substitution. Clinker is the key ingredient in cement, and its production is emissions-intensive because it requires heating limestone to high temperatures. Substituting clinker with fly ash, slag, or other materials can reduce emissions by 20 to 30 percent. China's cement industry has already adopted some substitution, but the potential is far from exhausted. The main constraints are the availability of substitute materials, the quality requirements of construction standards, and the resistance of cement producers who have invested in clinker capacity.

In aluminum, the main alternative is inert anode technology, which replaces carbon anodes with materials that do not emit CO2 during electrolysis. The technology is still in the pilot stage, and its costs are uncertain. The more immediate option is to power aluminum smelters with renewable electricity. China's aluminum industry is concentrated in provinces with abundant coal, such as Xinjiang and Inner Mongolia, because coal is cheap. Moving smelters to provinces with abundant renewables, such as Yunnan and Sichuan, would reduce emissions but would require new investment and would disrupt local economies. Some smelters have already moved to Yunnan, where hydropower is abundant during the wet season. The move has reduced emissions but has also created new problems, because hydropower is scarce during the dry season, and smelters must either curtail production or buy coal-fired electricity.

The circular economy is the third component. It involves reducing waste, reusing materials, and recycling products. China has made significant progress in recycling, particularly in metals and plastics. The recycling industry is large but fragmented, and its environmental performance varies widely. The government has tried to formalize the industry by licensing recyclers and by setting standards for recycling processes. The formalization has improved environmental performance but has also raised costs, which has pushed some recycling activity into the informal sector. The circular economy is important for decarbonization because it reduces the demand for primary materials, which are emissions-intensive to produce. It is not a substitute for decarbonizing primary production. It is a complement.

The industrial policy response also includes a set of financial incentives. The government provides subsidies for energy efficiency upgrades, for renewable energy projects, and for green technology research and development. The subsidies are administered through a variety of programs, including the central budget, local budgets, and state-owned banks. The subsidies are often conditional on performance: a firm that achieves a certain level of energy savings receives a payment, while a firm that fails to achieve the level receives nothing. The performance condition is intended to ensure that subsidies are effective, but it also creates a measurement problem, because the savings must be verified. The verification is often done by third-party auditors, whose quality varies. The result is a subsidy system that is more effective than a blank check but less effective than a well-designed carbon price.

4.1.5 The Limits of the Architecture

The policy architecture has achieved a great deal. It has mobilized attention, created institutions, set targets, and begun to change investment patterns. It also has limits, and those limits are important for understanding the incentive landscape that firms face.

The first limit is the gap between targets and implementation. Targets are set at the national level, but implementation is local. Local governments have discretion in how they implement policies, and their incentives are not always aligned with national goals. The result is uneven implementation: some regions move quickly, others slowly. The central government has tried to address this through inspections and cadre evaluations, but the problem persists. The gap is not unique to China; federal systems everywhere face similar challenges. The gap is larger in China because the central government has fewer direct tools for enforcing local compliance.

The second limit is the weakness of the carbon price. The national ETS is the centerpiece of the pricing architecture, but its prices are low and its coverage is limited. The low price reflects a generous cap and free allocation. The limited coverage reflects the difficulty of extending the system to sectors with complex production processes and heterogeneous products. The result is that the carbon price signal is weak, and firms have little incentive to reduce emissions beyond what is required by regulations. The weakness is not accidental; it reflects a political choice to prioritize economic stability over rapid decarbonization. The choice has consequences: it slows the transition and makes it more expensive in the long run.

The third limit is the measurement problem. Emissions accounting is the foundation of any climate policy, and it is also the weakest link. The data is incomplete, the verification is uneven, and the incentives to underreport are strong. The problem is not unique to China, but it is more severe in China because the system is newer and the capacity is thinner. The measurement problem undermines the effectiveness of every other policy: a carbon price that is based on inaccurate data will not produce the intended behavior; a subsidy that is based on unverified savings will not produce the intended investment; a standard that is not enforced will not produce the intended reduction.

The fourth limit is the tension between decarbonization and economic growth. China's economy is still growing, and growth requires energy. The dual carbon goals require that energy growth be decoupled from emissions growth. The decoupling is happening, but slowly. The pace of decoupling depends on the composition of growth: if growth is driven by services and high-tech manufacturing, emissions can decline; if growth is driven by construction and heavy industry, emissions will rise. The government has tried to shift the composition of growth toward services and high-tech manufacturing, but the shift is gradual. The tension between decarbonization and growth is managed through a combination of targets, subsidies, and political pressure. It is not resolved.

These limits do not mean that the architecture is failing. They mean that the architecture is incomplete. It has created a framework for action, but it has not yet created a set of incentives strong enough to drive the scale of investment that the dual carbon goals require. The next section examines the pricing instruments that are supposed to provide those incentives.

4.2 Carbon Pricing: ETS, Carbon Tax, and Market Mechanisms

Carbon pricing is the economists' preferred tool for reducing emissions. The logic is straightforward: if emissions carry a price, firms and consumers will reduce them. The price can be set directly, through a tax, or indirectly, through a cap-and-trade system. Both approaches have strengths and weaknesses. A tax provides price certainty but not quantity certainty: firms know what they will pay per ton, but the government does not know how much emissions will be reduced. A cap-and-trade system provides quantity certainty but not price certainty: the government knows how much emissions will be allowed, but firms do not know what they will pay per ton. The choice between the two depends on which uncertainty is more costly. If the cost of exceeding a temperature threshold is high, quantity certainty is more important, and cap-and-trade is preferable. If the cost of abatement is uncertain, price certainty is more important, and a tax is preferable.

China has chosen cap-and-trade for its national carbon market, supplemented by pilot programs and a set of complementary policies. The choice reflects a combination of factors: the influence of international models, particularly the EU ETS; the desire to maintain control over the quantity of emissions; and the political difficulty of introducing a new tax. The choice also reflects a learning process: China's pilot carbon markets, which began in 2013 in seven provinces and cities, provided experience with cap-and-trade that informed the design of the national system.

4.2.1 The National ETS: Design and Performance

The national ETS was launched in July 2021 after years of preparation. Its design reflects the lessons of the pilots and the EU ETS, adapted to China's institutional context. The system covers the power sector, which accounts for about 40 percent of national emissions. It includes about 2,200 enterprises, each of which emits more than 26,000 tons of CO2 per year. The cap is set by the MEE based on a combination of historical emissions and production data. Allowances are allocated for free, based on benchmarks that vary by technology type. The benchmarks are set to reward efficient plants and penalize inefficient ones. The system includes a trading platform, a registry, and a verification process. Compliance is enforced through fines and through the threat of reduced allocations in future years.

The system's performance has been mixed. On the positive side, it has established a carbon price where none existed before. The price has ranged from about 40 to 90 yuan per ton since launch, with an average of about 55 yuan. The price signal has encouraged some firms to invest in efficiency improvements and fuel switching. The system has also created a market infrastructure—trading platforms, registries, verifiers—that can be expanded to other sectors. The expansion to steel, cement, and aluminum in 2024 was a significant step, bringing additional emissions into the pricing system and creating a more comprehensive carbon market.

The system's weaknesses are equally significant. The cap is set generously, which means that allowances are abundant and prices are low. The free allocation method, based on benchmarks, rewards efficient plants but does not create a strong incentive for inefficient plants to shut down, because they still receive free allowances. The coverage is limited to the power sector, which means that emissions from other sectors are not priced. The verification process is uneven, which means that reported emissions may be inaccurate. The market is also thin: trading volumes are low, and many enterprises hold allowances rather than trade them. The thinness of the market means that prices are volatile and may not reflect the true marginal cost of abatement.

The low price is the system's most important weakness. At 55 yuan per ton, the carbon price is far below the social cost of carbon, which the World Bank's High-Level Commission on Carbon Prices estimated at $40–80 per ton by 2020 and rising thereafter—a range that translates to several hundred yuan at current exchange rates. The low price means that firms have little incentive to reduce emissions beyond what is required by regulations. A firm that faces a carbon price of 55 yuan per ton will compare that price with the cost of reducing emissions. If the cost of reduction is higher than 55 yuan per ton, the firm will pay the price rather than reduce emissions. For many firms, the cost of reduction is higher, especially for deep decarbonization investments such as hydrogen-based steelmaking or carbon capture and storage. The result is that the carbon price is not driving the scale of investment that the dual carbon goals require.

The government is aware of this problem and has taken steps to address it. The 2024 expansion of the market to steel, cement, and aluminum will increase demand for allowances and may raise prices. The revision of the allocation method, which introduces auctioning for some allowances, will reduce the supply of free allowances and may also raise prices. The introduction of a market stability reserve, modeled on the EU's mechanism, will adjust the supply of allowances in response to price movements. These reforms are incremental, but they signal a direction of travel toward a more binding carbon price. The pace of reform will depend on the political priority attached to the carbon market and on the willingness of covered enterprises to accept higher costs.

The national ETS also faces a structural problem: the power sector is not fully marketized. Electricity prices are regulated, which means that generators cannot pass the carbon cost through to consumers. If a generator faces a carbon price but cannot raise its electricity price, the carbon cost comes out of its margin. The result is that generators have an incentive to reduce emissions, but they also have an incentive to resist the carbon price. The government has tried to address this by allowing some pass-through of carbon costs in wholesale electricity prices, but the pass-through is partial and uneven. The structural problem is likely to persist until electricity market reform is complete.

4.2.2 The Pilot Markets: Lessons from Local Experience

Before the national ETS, China operated seven pilot carbon markets in Beijing, Tianjin, Shanghai, Chongqing, Hubei, Guangdong, and Shenzhen. The pilots began in 2013 and covered a range of sectors, including power, steel, cement, chemicals, and aviation. They were designed as experiments, and they produced a rich set of lessons that informed the design of the national system.

The pilots' most important lesson was that carbon markets can work in China, but they require strong institutions. The pilots that performed best—Shenzhen, Guangdong, and Hubei—had well-staffed regulators, clear rules, and effective enforcement. The pilots that performed worst had weak institutions, unclear rules, and lax enforcement. The variation in performance was not just a matter of design; it was a matter of capacity. A carbon market is a complex institution that requires sophisticated data systems, trained staff, and credible enforcement. Building these takes time and resources.

The pilots also demonstrated the importance of coverage and stringency. The pilots that covered more sectors and set tighter caps produced higher prices and more abatement. The pilots that covered fewer sectors and set looser caps produced lower prices and less abatement. The Shenzhen pilot, which covered a wide range of sectors and set a tight cap, had the highest prices and the most trading activity. The Chongqing pilot, which covered fewer sectors and set a looser cap, had the lowest prices and the least trading activity. A carbon market's effectiveness depends on its ambition.

Several pilots found that enterprises had underreported their emissions, sometimes by large margins. The underreporting was driven by a combination of factors: unclear reporting guidelines, weak verification, and incentives to minimize costs. The pilots responded by strengthening verification and by imposing penalties for misreporting. The national system has adopted some of these measures, but the problem persists.

The pilots' experience with market stability was also instructive. Several pilots experienced price crashes when the supply of allowances exceeded demand. The crashes reduced the carbon price to near zero and undermined the market's credibility. The pilots responded by introducing mechanisms to adjust supply, such as allowance reserves and price floors. The national system has adopted a market stability reserve, which is designed to prevent similar crashes. The reserve's effectiveness will depend on its design and on the political will to use it.

The pilots also demonstrated the value of experimentation. The pilots tested different allocation methods, different coverage rules, and different enforcement mechanisms. The lessons from these experiments informed the design of the national system. The pilots continue to operate, and they continue to provide lessons. Some pilots have expanded their coverage to new sectors; others have introduced new financial products. The pilots are a laboratory for carbon market design, and their experience is valuable for the national system's evolution.

4.2.3 The Carbon Tax Debate

China does not have a carbon tax, and the debate about whether to introduce one has been ongoing for more than a decade. The debate matters because a carbon tax would complement the ETS and would cover sectors that are not included in the ETS. A tax would also provide price certainty, which the ETS does not. A tax faces political obstacles that the ETS does not.

The case for a carbon tax is straightforward. A tax would set a clear price on carbon, which would encourage firms and consumers to reduce emissions. A tax would be simpler to administer than an ETS, because it would not require a cap, an allocation method, or a trading platform. A tax would generate revenue that could be used to reduce other taxes, to fund green investment, or to compensate low-income households. A tax would also cover sectors that are not included in the ETS, such as transport and buildings, which are significant sources of emissions. The economists' case for a tax is strong.

The case against a carbon tax is also strong. A tax would raise energy prices, which would hurt low-income households and energy-intensive industries. A tax would be visible, which would make it politically unpopular. A tax would require legislation, which would be difficult to pass in China's political system. A tax would also overlap with the ETS, which would create double pricing and administrative complexity. The government has preferred the ETS because it is less visible, more flexible, and easier to implement through administrative measures. The preference reflects a political judgment about what is feasible, not a judgment about what is economically optimal.

The debate has evolved over time. In the early 2010s, the Ministry of Finance and some academic institutions advocated for a carbon tax. The proposal was to introduce a tax on fossil fuels, starting at a low rate and increasing over time. The proposal was studied and debated but never implemented. The main obstacle was opposition from energy-intensive industries and from local governments that depend on those industries for revenue. The ETS emerged as an alternative that could achieve similar goals without the political cost of a new tax.

In recent years, the debate has revived. Some scholars have argued that a carbon tax is necessary to complement the ETS, because the ETS covers only a subset of emissions and its prices are too low. Others have argued that a tax would be redundant, because the ETS can be strengthened to cover more sectors and to set higher prices. The government has not taken a position, but the 2021 guiding document mentions "improving the carbon pricing mechanism" without specifying whether a tax is part of the plan. The debate is likely to continue, but a carbon tax is unlikely to be introduced in the near term. The political obstacles are too great, and the ETS is the preferred instrument.

4.2.4 Voluntary Markets and Carbon Offsets

Voluntary carbon markets are a small but growing part of China's carbon pricing landscape. These markets allow firms and individuals to purchase carbon credits from projects that reduce or remove emissions, such as reforestation, renewable energy, or methane capture. The credits are used to offset the buyer's emissions, either for compliance purposes or for voluntary commitments. China has a domestic voluntary market, the China Certified Emission Reduction (CCER) scheme, which was launched in 2012 and relaunched in 2023 after a period of suspension. The scheme allows projects to earn credits that can be used to offset a portion of emissions under the ETS or sold to voluntary buyers.

The voluntary market's role in China's climate policy is limited but not negligible. The CCER scheme provides a source of finance for projects that would not otherwise be viable, such as small-scale renewable energy or forestry projects. The scheme also provides a way for firms to meet voluntary commitments, such as those made under the Science Based Targets initiative (SBTi) or the RE100. The scheme has been criticized for problems of additionality, double counting, and quality. Additionality is the question of whether a project would have happened without the credit revenue. If the answer is yes, the credit does not represent a real reduction. Double counting is the question of whether the same reduction is claimed by two parties. Quality is the question of whether the reduction is permanent and verifiable. These problems are not unique to China; they affect voluntary markets everywhere. They are particularly acute in China, where the regulatory infrastructure is still developing.

The government has tried to address these problems by strengthening the CCER scheme's rules. The relaunched scheme includes stricter additionality tests, clearer methodologies, and a registry that tracks credits from issuance to retirement. The scheme also excludes certain project types that were controversial, such as large hydropower. The reforms are intended to improve quality and to restore confidence in the market. Their effectiveness will depend on the rigor of implementation.

4.2.5 The Price Signal and Firm Behavior

The effectiveness of carbon pricing depends on how firms respond to the price signal. The evidence from China is mixed. Some firms have responded by investing in efficiency improvements, fuel switching, and process substitution. Others have responded by paying the price and continuing with business as usual. The difference depends on the size of the price, the cost of abatement, and the firm's capacity to invest.

The price is low, which means that the incentive to abate is weak. A firm that faces a price of 55 yuan per ton will only invest in abatement if the cost of abatement is lower than 55 yuan per ton. For many firms, the cost of abatement is higher, especially for deep decarbonization. The result is that the price signal is not driving the scale of investment that the dual carbon goals require. The government has tried to address this by providing subsidies for green investment, which effectively lower the cost of abatement. Subsidies are not a substitute for a carbon price; they are a complement. A well-designed policy would combine a higher carbon price with targeted subsidies for technologies that are not yet cost-competitive.

The cost of abatement varies widely across firms and sectors. For some firms, the cost is low: a cement plant can reduce emissions by substituting clinker with fly ash at a cost of a few yuan per ton. For other firms, the cost is high: a steel plant can reduce emissions by switching to hydrogen-based direct reduced iron at a cost of hundreds of yuan per ton. The variation means that a uniform carbon price will produce uneven results. A low price will induce the cheap abatement but not the expensive abatement. A high price will induce both but will also impose large costs on firms that have no cheap abatement options. The design of carbon pricing therefore involves a trade-off between effectiveness and cost.

The firm's capacity to invest is also important. Large firms with access to capital can invest in abatement; small firms with limited capital cannot. The result is that carbon pricing may favor large firms over small firms, which could accelerate consolidation in energy-intensive industries. The consolidation may be good for efficiency, but it may also have social consequences, such as job losses in regions that depend on small firms. The government has tried to address this by providing transition support, but the support is limited.

The firm's response also depends on its expectations about future prices. If a firm expects the carbon price to rise, it may invest in abatement now to avoid higher costs later. If a firm expects the price to remain low, it may delay investment. The government's commitment to raising the price over time is therefore crucial. The 2024 reforms signal a direction of travel toward higher prices, but the pace is uncertain. The uncertainty may discourage investment, because firms do not know what price to plan for. A credible long-term price trajectory would reduce uncertainty and encourage investment. The government has not yet published such a trajectory, but the reforms are a step in that direction.

4.3 Incentive Landscape for Manufacturing: Subsidies, Standards, and Green Finance

Carbon pricing is only one part of the incentive landscape. Firms also face subsidies, standards, and green finance, each of which shapes their decisions about green investment. The interaction of these instruments determines whether firms invest in green governance capability or simply comply with regulations. This section examines each instrument in turn and then considers how they interact.

4.3.1 Subsidies and Fiscal Incentives

China's government provides a wide range of subsidies for green investment. The subsidies are administered through national programs, provincial programs, and local programs. They cover energy efficiency upgrades, renewable energy projects, electric vehicles, green buildings, and research and development. The total amount of subsidies is difficult to estimate because they are spread across many programs and levels of government. The scale is large: the central government alone spends hundreds of billions of yuan per year on energy conservation and environmental protection.

The subsidies are designed to address the gap between the private cost and the social benefit of green investment. A firm that invests in energy efficiency may not capture the full benefit of its investment, because some of the benefit accrues to society through reduced emissions. A subsidy can close the gap and make the investment worthwhile for the firm. The subsidies are also designed to address the learning-by-doing problem: early-stage green technologies are expensive, but their costs fall as they are deployed. A subsidy can support early deployment and accelerate the learning curve.

The subsidies' effectiveness varies. Some subsidies have been highly effective: the subsidy for solar panels, for example, helped China build a world-class solar industry and drove down costs by more than 90 percent. Other subsidies have been less effective: the subsidy for energy efficiency upgrades in heavy industry, for example, has been criticized for being too small and too fragmented to induce significant investment. The difference depends on the design of the subsidy, the size of the incentive, and the capacity of firms to respond.

The subsidies also have unintended consequences. The solar subsidy, for example, led to overcapacity and a boom-bust cycle that bankrupted many firms. The electric vehicle subsidy, similarly, led to overcapacity and to a proliferation of low-quality producers. The overcapacity was partly a result of the subsidy design, which rewarded production rather than performance. The government has tried to address this by phasing out subsidies and by introducing performance standards. The boom-bust cycle is a recurring feature of subsidy programs, and it is likely to recur.

The subsidies are also vulnerable to fraud. Firms have been known to claim subsidies for projects that were not completed, to inflate the reported savings, or to double-count the same project under multiple programs. The government has tried to address this by strengthening verification and by imposing penalties for fraud. Verification is expensive, and the penalties are often too small to deter fraud. The result is a subsidy system that is less effective than it could be.

The fiscal incentives also include tax breaks. Firms that invest in energy-saving equipment can claim accelerated depreciation. Firms that produce renewable energy can claim VAT exemptions. Firms that invest in research and development can claim super-deductions for R&D expenses. The tax breaks are less visible than direct subsidies, but they can be significant. The accelerated depreciation, for example, allows firms to recover their investment more quickly, which improves the cash flow and the return on investment. The R&D super-deduction, which allows firms to deduct 175 percent of their R&D expenses from taxable income, is one of the most generous in the world. The tax breaks are administered by the tax authorities, and their effectiveness depends on the firms' awareness and on the complexity of the rules.

4.3.2 Standards and Regulatory Requirements

Standards are a powerful but often overlooked instrument of green governance. A standard sets a minimum requirement for a product or process, and firms that fail to meet it face penalties or exclusion from the market. Standards can be more effective than prices in some contexts, because they directly require the desired behavior rather than relying on firms to respond to a price signal. Standards can also be more politically feasible than prices, because they are less visible and because they can be framed as protecting consumers or the environment.

China has a comprehensive system of green standards. The system includes energy efficiency standards for appliances, buildings, and industrial equipment; emissions standards for vehicles and power plants; water use standards for industry; and green product standards for a wide range of consumer goods. The standards are set by a combination of ministries, including the Standardization Administration of China (SAC), the MEE, the MIIT, and the NEA. The standards are updated periodically, and they have become progressively stricter over time.

The standards' effectiveness depends on enforcement. China's enforcement record is mixed. On the positive side, the government has invested heavily in monitoring and inspection, particularly for air and water pollution. The central environmental inspections, which began in 2016, have been effective at identifying violations and forcing local governments to act. On the negative side, enforcement is uneven across regions and sectors. Some local governments lack the capacity or the will to enforce standards; others are captured by the industries they regulate. The result is that standards are strictly enforced in some places and loosely enforced in others. The variation creates an uneven playing field, which disadvantages firms that comply and rewards firms that do not.

The standards also have a trade dimension. China's standards are increasingly aligned with international standards, which makes it easier for Chinese firms to export. The alignment is not complete, and some Chinese standards are stricter than international standards, which can create barriers to imports. The EU's CBAM, which will impose a carbon price on imports, is likely to interact with China's standards in complex ways. Firms that meet China's standards may still face CBAM charges if their emissions are higher than the EU's benchmarks. The interaction is a source of uncertainty for Chinese exporters, and it is likely to become more important as CBAM is phased in.

The standards also have an innovation dimension. Strict standards can drive innovation by forcing firms to develop new technologies and processes. The solar industry is an example: China's renewable portfolio standards and grid connection standards forced solar manufacturers to improve efficiency and reduce costs. The electric vehicle industry is another example: China's fuel economy standards and new energy vehicle mandates forced automakers to develop electric models. The innovation effect depends on the stringency of the standard and on the capacity of firms to respond. A standard that is too lax will not drive innovation; a standard that is too strict will drive firms out of business.

4.3.3 Green Finance: Channels and Constraints

Green finance is the third pillar of the incentive landscape. It includes green bonds, green loans, green funds, and green insurance. The purpose of green finance is to channel capital toward green projects by reducing the cost of capital or by increasing the availability of funding. China has become a global leader in green finance: its green bond market is one of the largest in the world, and its green loan market is the largest. The scale of green finance reflects a combination of government policy, market demand, and international pressure.

The government's role in green finance is both direct and indirect. Directly, the government has created a policy framework for green finance, including the Green Bond Endorsed Project Catalogue, which defines what counts as green; the Green Credit Guidelines, which encourage banks to lend to green projects; and the green finance pilot zones, which test new financial instruments. Indirectly, the government has used moral suasion and window guidance to encourage financial institutions to support green projects. The combination of direct and indirect measures has been effective at mobilizing capital, but it has also created problems of quality and accountability.

The quality problem is the most important. The green bond market has been criticized for "greenwashing," or the practice of labeling a bond as green when the underlying project is not genuinely green. The problem is not unique to China, but it is particularly acute in China because the definition of green is broad and the verification is weak. The Green Bond Endorsed Project Catalogue includes projects that would not be considered green in international markets, such as clean coal and energy efficiency improvements in coal plants. The inclusion reflects China's energy reality—coal is still the dominant source of electricity—but it creates confusion for international investors who use stricter definitions. The government has tried to address this by aligning the catalogue with international standards, but the alignment is incomplete.

The accountability problem is also important. Green finance is supposed to support projects that would not otherwise be financed, but it is difficult to verify additionality. A firm that issues a green bond may have funded the project anyway, in which case the bond does not represent an additional green investment. The problem is compounded by the fact that green bonds often carry a "greenium," or a lower yield than conventional bonds, which means that issuers can save money by labeling their bonds green. The greenium creates an incentive to greenwash, because the benefit of the label exceeds the cost of compliance. The government has tried to address this by requiring third-party verification and by publishing the list of approved verifiers. The verification is often superficial, and the penalties for mislabeling are weak.

The green finance market also faces a maturity mismatch. Green projects often have long payback periods, while green loans and bonds often have short maturities. The mismatch creates refinancing risk, which can discourage investment. The government has tried to address this by encouraging the development of long-term green bonds and by providing guarantees for green loans. The mismatch persists, particularly for small firms that lack access to long-term capital.

The green finance market is also concentrated. Most green finance flows to large firms and to projects in wealthy provinces. Small firms and projects in poor provinces have limited access. The concentration reflects the risk preferences of financial institutions, which prefer large, creditworthy borrowers. The government has tried to address this by creating green finance pilot zones in less developed regions and by providing subsidized loans for small firms. The concentration persists, and it is likely to remain a constraint on the green transition in less developed regions.

4.3.4 The Interaction of Instruments

The incentive landscape is not a set of independent instruments; it is a system in which the instruments interact. The interaction can be complementary or contradictory, and the net effect depends on the design of the instruments and on the context in which they are applied.

The interaction between carbon pricing and subsidies is complementary in some cases and contradictory in others. A carbon price raises the cost of emissions, which makes green investment more attractive. A subsidy lowers the cost of green investment, which makes it more affordable. The combination can be more effective than either instrument alone, because the carbon price provides a sustained incentive while the subsidy addresses the upfront cost. The combination can also be redundant: if the carbon price is high enough, the subsidy is unnecessary; if the subsidy is generous enough, the carbon price is irrelevant. The redundancy is not necessarily a problem, because the instruments serve different purposes, but it can lead to overinvestment or to windfall profits for firms that would have invested anyway.

The interaction between standards and carbon pricing is also complex. A standard requires a minimum level of performance, while a carbon price allows firms to choose between reducing emissions and paying the price. The combination can be more effective than either instrument alone, because the standard ensures that all firms meet a minimum level while the carbon price encourages firms to go beyond the minimum. The combination can also be contradictory: if the standard is strict enough, the carbon price is redundant for covered firms; if the carbon price is high enough, the standard is redundant. The contradiction is managed by setting the standard at a level that is achievable and the carbon price at a level that is meaningful. The calibration is difficult, and it is likely to require adjustment over time.

Green finance provides the capital that firms need to respond to carbon prices and standards. Without access to capital, firms cannot invest in green technology, even if the incentives are strong. The enabling role is particularly important for small firms, which often lack the internal resources to finance green investment. The government has tried to address this by creating dedicated green finance channels for small firms, but the channels are limited.

The interaction of instruments also creates a risk of policy fragmentation. Each instrument is administered by a different agency, with different objectives, different data systems, and different timelines. The fragmentation can lead to inconsistency: a firm may face a strict standard from one agency and a lax enforcement from another; a subsidy from one program may be offset by a tax from another. The fragmentation is a source of uncertainty for firms, and it can undermine the effectiveness of the overall policy. The government has tried to address this by creating coordination mechanisms, such as the Leading Group and the inter-ministerial committees. Coordination is difficult, and fragmentation persists.

4.3.5 Firm-Level Responses: Evidence from the Field

The incentive landscape shapes firm behavior, but the shaping is not deterministic. Firms respond differently to the same incentives, depending on their size, their sector, their ownership, and their management capacity. This section draws on field evidence to describe how firms actually respond.

Large state-owned enterprises are the most responsive to the incentive landscape, for several reasons. They have the resources to invest in green technology, the technical capacity to measure and report emissions, and the political connections to access subsidies and green finance. They also face direct pressure from SASAC, which sets performance targets that include carbon reduction. The combination of resources, capacity, and pressure makes SOEs the leaders in green investment in many sectors. The leadership is uneven: some SOEs have made genuine progress, while others have complied superficially, by reporting emissions reductions that are not real or by investing in projects that are labeled green but are not.

Private large firms are also responsive, but for different reasons. They face pressure from international customers, who require them to meet environmental standards and to report emissions. They also face pressure from investors, who are increasingly attentive to environmental, social, and governance (ESG) performance. The pressure from customers and investors is often more important than the pressure from domestic regulations, because it affects their access to markets and capital. The result is that private large firms in export-oriented sectors are often more advanced in green governance than their domestic-oriented peers.

Small and medium-sized enterprises are the least responsive, for reasons that are structural. They lack the resources to invest in green technology, the capacity to measure and report emissions, and the connections to access subsidies and green finance. They also face competitive pressure that makes it difficult to pass on the cost of green investment to customers. The result is that SMEs are often laggards in green governance, even when they face the same regulations as larger firms. The lag is a problem for the overall transition, because SMEs account for a large share of industrial output and employment.

The variation in firm responses is not just a matter of size. It is also a matter of sector. Firms in sectors with clear decarbonization pathways—such as power generation, cement, and aluminum—are more responsive than firms in sectors with uncertain pathways, such as chemicals and ceramics. The clarity of the pathway matters because it determines whether firms can plan their investments. A firm that knows how to decarbonize will invest; a firm that does not know will wait.

The variation is also a matter of ownership. Private firms are more responsive to market signals, while SOEs are more responsive to political signals. The difference means that the same policy instrument can have different effects on different types of firms. A carbon price, for example, will have a stronger effect on private firms, which are more sensitive to costs, than on SOEs, which can absorb costs through their balance sheets. A political target, conversely, will have a stronger effect on SOEs than on private firms. The design of policy therefore needs to account for the ownership structure of the sectors it targets.

The field evidence also shows that firms respond to the incentive landscape as a whole, not to individual instruments. A firm that faces a carbon price, a standard, and a subsidy will consider all three when making investment decisions. The interaction of the instruments determines the net incentive. If the instruments are complementary, the net incentive is strong; if they are contradictory, the net incentive is weak. The implication is that policy design should focus on the coherence of the overall landscape, not on the strength of individual instruments.

4.3.6 The Compass and the Map

The dual carbon goals are a compass: they point in a direction, but they do not specify the route. The policy architecture, the carbon price, and the incentive landscape are the map: they describe the terrain, the obstacles, and the available paths. The map is incomplete, and some of its features are misleading. It is the best map available, and firms must use it to navigate.

The map's most important feature is the direction of travel. The carbon price is rising, the standards are tightening, and the subsidies are shifting from deployment to innovation. The direction is clear even if the pace is uncertain. Firms that plan for a higher carbon price and stricter standards will be better positioned than firms that assume the status quo will continue. The planning horizon matters: a firm that invests in a coal plant with a 40-year lifetime is making a bet that the carbon price will remain low for four decades. That bet is unlikely to pay off.

The map's second feature is the variation across sectors. Some sectors have clear decarbonization pathways and strong policy support; others do not. The variation creates opportunities and risks. A firm in a sector with a clear pathway can invest with confidence; a firm in a sector with an uncertain pathway faces a harder choice. The choice depends on the firm's assessment of the technology and the policy. A firm that believes the technology will improve and the policy will tighten may invest early; a firm that believes the opposite may wait.

The map's third feature is the importance of measurement. Every instrument in the landscape depends on measurement: the carbon price depends on reported emissions, the standards depend on tested performance, the subsidies depend on verified savings, and the green finance depends on certified projects. Firms that can measure accurately will be able to navigate the landscape; firms that cannot will be at a disadvantage. Measurement capability is a competitive asset, and it is also a compliance requirement.

The map's fourth feature is the role of the global value chain. The incentive landscape is not purely domestic. Firms that export to Europe face the CBAM, which will impose a carbon price on their products. Firms that supply global brands face customer requirements for environmental performance. Firms that compete for international capital face investor pressure for ESG performance. The global dimension of the incentive landscape is becoming more important, and it is likely to become more important still as carbon pricing spreads to other countries.

The chapter's argument is that the incentive landscape is a necessary but not sufficient condition for green governance capability. The landscape creates pressure and opportunity, but firms must respond. The response depends on their capacity to measure, to invest, and to learn. The chapters that follow examine how firms in different factor-intensity categories have responded, and what has happened when they have not. The labor-intensive industries, examined in Chapter 5, face the sharpest trade-off between the cost of measurement and the margin available to absorb it. The technology-intensive industries, examined in Chapter 6, have the most to gain from measurement, because their products are increasingly evaluated on environmental performance. The capital-intensive industries, examined in Chapter 7, face the most difficult transition, because their assets are long-lived and their processes are hard to change. The furniture case, examined in Chapter 9, illustrates how these dynamics play out in a traditional advantageous industry. The clusters, examined in Chapter 8, show how collective capability can compensate for individual weakness.

Before turning to those chapters, it is worth pausing on a question that the incentive landscape raises but does not answer. The question is whether the landscape is designed to produce green governance capability or simply to produce compliance. A compliance-oriented landscape sets minimum requirements and penalizes firms that fail to meet them. A capability-oriented landscape creates incentives for firms to invest in measurement, learning, and innovation. The two are not the same. A firm can comply with a standard without developing the capability to go beyond it. A firm can pay a carbon price without changing its production process. The distinction matters because the dual carbon goals require capability, not just compliance. The landscape's design will determine which one emerges.

The evidence from the first years of the dual carbon goals suggests that the landscape is producing a mix of both. Some firms have developed genuine capability: they have built measurement systems, invested in green technology, and integrated environmental performance into their strategy. Others have complied superficially: they have reported emissions, paid the price, and continued with business as usual. The mix is not surprising, given the weakness of the carbon price and the unevenness of enforcement. The question is whether the mix will shift toward capability as the landscape tightens. The answer depends on the political will to raise the price, tighten the standards, and enforce the rules. That will is the ultimate determinant of whether the compass points toward a genuine transition or toward a compliance exercise with a rising price tag.

The next chapter begins the industry-level analysis with the labor-intensive sectors, where the trade-off between measurement cost and margin is sharpest.