Chapter 9: Risk, Uncertainty, and the Politics of Low-Carbon Transition: Managing Unintended Consequences

In the autumn of 2021, a furniture factory in Shunde that I had visited twice before shut down its spray-painting line for eleven days. Orders were fine. The workforce was intact. The cause was a sudden tightening of volatile organic compound (VOC) inspections ahead of a provincial environmental review, arriving in the same month that the price of the low-VOC coatings the factory had switched to eighteen months earlier jumped by roughly a third. The owner, a man in his fifties who had built the business from a two-bench workshop, told me on the phone that he had done everything the government asked: installed the new equipment, paid for the certification, trained his workers. "And now," he said, "I am shut down because the paint costs too much and the inspector is nervous." He paused. "Tell me how this is my fault."

That phone call is the seed of this chapter. The previous eight chapters have argued that green governance capability can function as an endogenous driver of industrial upgrading—that firms and clusters which learn to measure, report, and reduce emissions can convert that competence into competitive advantage in global value chains. I still believe that argument. But it holds only under conditions that the transition itself can undermine. A factory that has invested in green capability can be destroyed by a policy shock, a price spike, or a misaligned incentive. A cluster that has begun to climb the green ladder can be pushed off it by a change in enforcement intensity or a shift in the political calendar. The passage from cost-based to green competitive advantage runs through a landscape of risk and uncertainty, populated by winners and losers, mediated by political processes that produce outcomes no one designed.

What kinds of risk does the transition generate, and where do they come from? Who bears them, and how does the political system process the resulting conflicts? Which governance tools exist to manage the unintended consequences, and how well do they perform? The answers are not reassuring in every respect. My assessment, based on the field evidence presented below, is that China's transition governance is stronger on mobilization than on compensation, and stronger on target-setting than on feedback from firms and localities. Those asymmetries are the subject of this chapter.

The analysis draws on three bodies of evidence: the European sociotechnical transitions literature (Geels and Schot 2007; Turnheim and Geels 2012), the empirical literature on Chinese environmental policy and its distributional effects, and my own interviews in Guangdong, Zhejiang, and Jiangsu between 2019 and 2023. I distinguish throughout between risks that are documented and quantified and risks inferred from theory or anecdote. The case material is illustrative rather than representative; the quantitative claims carry the cautions their sources warrant.

9.1 Types and Sources of Risk and Uncertainty: Economic, Social, and Political

9.1.1 The Conceptual Terrain: Risk, Uncertainty, and Transition Pathways

Frank Knight's 1921 distinction remains the cleanest available. Risk describes a situation in which the probabilities of outcomes are known or can be estimated; uncertainty describes a situation in which they cannot. A factory owner who knows that a VOC inspection has a 15 percent chance of finding a violation in any given quarter faces risk. A factory owner who does not know whether the inspection regime itself will be tightened, loosened, or redirected next year faces uncertainty. Risk can be priced, insured against, and managed through diversification. Uncertainty can only be absorbed, hedged through flexibility, or reduced through information gathering.

The low-carbon transition in Chinese manufacturing combines both conditions in an unusual way. The direction of policy is clear: the "dual carbon" goals announced in September 2020 commit China to peaking carbon emissions before 2030 and reaching neutrality before 2060. Carbon market coverage expands on a published schedule. Energy intensity targets in the Five-Year Plans are announced in advance and broken down by province. A firm that pays attention can see the broad contours of the transition five to ten years out. The specific parameters—the pace of enforcement, the allocation of allowances, the timing of price reforms, the design of compensation—are subject to revision, and the revisions are often abrupt. The transition is directionally certain and operationally uncertain. That combination corrodes firm-level planning more than either pure certainty or pure uncertainty would, because it invites investment on the basis of a signal that may later be changed.

Geels and Schot (2007) supply a useful vocabulary. They distinguish transition pathways by the timing and nature of interactions between niche innovations and the regime they seek to displace. In a transformation pathway, regime actors respond to external pressure by reorienting incrementally. In a de-alignment and re-alignment pathway, the regime erodes and multiple niches compete to replace it. In a technological substitution pathway, a niche innovation matures and displaces the regime. Each pathway generates a different risk profile: transformation pathways impose adjustment risks on incumbents but carry low systemic risk; substitution pathways impose high stranded-asset risk on incumbents and high coordination risk on the system; de-alignment pathways generate the highest uncertainty because no one knows which niche will win.

China's manufacturing transition, as I read the evidence, is closest to a transformation pathway in labor-intensive sectors like furniture, where the regime is being reoriented incrementally through regulation, market pressure, and cluster-level experimentation. It is closer to substitution in capital-intensive sectors like steel and cement, where process substitution—electric arc furnaces, alternative clinkers, hydrogen-based reduction—threatens to strand existing assets. Technology-intensive sectors like solar and batteries contain elements of de-alignment, with multiple technological trajectories competing and no clear winner. The three sectoral types therefore carry different risk profiles, and the governance responses need to be differentiated accordingly. Section 9.3 returns to this point.

9.1.2 Economic Risks: Cost, Competitiveness, and the New Arithmetic

The most immediate economic risk facing Chinese manufacturers is the cost of compliance. The compliance cost literature goes back to the 1970s, but the magnitude and distribution of compliance costs in the Chinese low-carbon transition are distinctive in three respects.

The costs are front-loaded and lumpy. A furniture factory that switches from solvent-based to water-based coatings must replace its spray booths, retrain its workers, and absorb a period of lower yields while the new process stabilizes. Capital expenditure concentrates in a single year; the benefits—lower VOC emissions, access to green procurement contracts—accrue over many years. For a small or medium-sized enterprise with limited access to credit, the timing mismatch can be fatal. My Shunde interviewee estimated the total cost of his 2020 conversion at 1.8 million yuan, roughly 40 percent of his annual net profit. He financed it by drawing down reserves and deferring a planned expansion. A larger competitor with access to bank credit at preferential rates would have spread the cost over five years.

The costs are unevenly distributed across the firm size distribution. Smaller firms face higher compliance costs per unit of output because they cannot spread fixed costs over as large a base—a familiar pattern in environmental regulation—and the structure of Chinese industrial clusters amplifies it. In Shunde's furniture district, hundreds of small firms share suppliers, subcontractors, and labor pools. A compliance shock that hits the small firms disproportionately ripples through the entire cluster. The large firms that anchor the cluster's export relationships may absorb the cost; their small suppliers may not. When a small supplier exits, the anchor firm loses a relationship-specific investment and must search for a replacement, which takes time and raises costs. The cluster-level effect can exceed the sum of the firm-level effects.

The costs also interact with existing pressures in ways that are easy to underestimate. The low-carbon transition is occurring simultaneously with rising labor costs, tightening credit conditions for small firms, and the restructuring of global value chains after the pandemic and the trade disputes. A firm facing all four pressures at once has less slack to absorb any one of them. The Shunde factory owner put it plainly: "If it were only the paint, I would be fine. But it is the paint and the wages and the bank and the Americans all at the same time."

Competitiveness risk is the second economic dimension. The concern, familiar from the pollution haven literature, is that stringent environmental regulation will drive production to jurisdictions with laxer standards. The empirical record is mixed and context-dependent. Levinson and Taylor (2008) and Ederington, Levinson, and Minier (2005) found statistically significant but economically modest effects of environmental regulation on trade flows in the United States. Work on China by Cai et al. (2016) and Shi and Xu (2018) finds that the pollution haven effect is real but concentrated in specific sectors and firm types—typically small, low-productivity, export-oriented firms in labor-intensive sectors. The firms most vulnerable to relocation pressure are precisely those least able to invest in green upgrading.

A countervailing dynamic complicates the picture. As I argued in Chapter 4, the relationship between environmental regulation and relocation is not monotonic. Up to a threshold, regulation triggers defensive relocation by marginal firms. Beyond the threshold, firms that have invested in green capability may find their compliance competence becomes a market asset, allowing them to serve customers who demand green products and to command price premiums that offset compliance costs. The threshold is not fixed; it depends on the structure of demand, the availability of finance, and the credibility of enforcement. The policy challenge is to push firms over the threshold rather than into the relocation trap. Whether a given firm crosses that threshold depends on instruments the state controls: credit allocation, phase-in schedules, and the credibility of enforcement.

Stranded assets are the third economic risk, most acute in capital-intensive sectors where long-lived assets—blast furnaces, cement kilns, coal-fired power plants—face premature retirement as carbon prices rise and technology shifts. The stranded asset literature (Caldecott et al. 2016; Ansar, Caldecott, and Tilbury 2013) has focused on fossil fuel reserves and power generation, but the logic extends to any capital stock whose value depends on a carbon-intensive production process. A steel mill built in 2015 with a thirty-year expected life faces a different calculus if carbon prices reach 200 yuan per ton by 2030 than if they stay at 50. That difference can separate a profitable asset from a liability. Chapter 6 takes up stranded asset risk in the capital-intensive sector analysis.

9.1.3 Social Risks: Employment, Community, and the Distribution of Adjustment Costs

The economics literature often treats the social dimension of transition risk as a second-order matter of "adjustment costs" to be handled through retraining and safety nets. That framing understates the political salience of social risk. Employment losses in specific communities, even when small in aggregate, generate disproportionate political resistance because the losses are concentrated and visible while the gains are diffuse. The geography of job losses shapes the politics of transition as much as their aggregate magnitude.

Precise employment estimates are hard to come by, but the available evidence points to three patterns. The net employment effect is likely positive in aggregate, driven by job creation in renewable energy, energy efficiency, and green services. IRENA (2019) counted 3.6 million renewable energy jobs in China in 2018 and projected 5.5 million by 2030; the broader green employment literature (Cai et al. 2014; Mu et al. 2018) suggests that gross job creation in green sectors exceeds gross destruction in brown sectors. The distribution of gains and losses is highly uneven across regions and skill levels. Coal-dependent provinces in Shanxi, Inner Mongolia, and Shaanxi face larger losses than coastal provinces that have already diversified. Low-skilled workers in energy-intensive industries face larger adjustment costs than high-skilled workers in green technology sectors. And the timing is mismatched: job losses in brown sectors tend to occur early in the transition, while job creation in green sectors arrives later, leaving a gap that can last years.

Social risk extends beyond employment to community-level effects. A cluster that loses its anchor industry can enter a spiral of depopulation, declining public services, and social dislocation that outlasts the initial economic shock. Pike et al. (2017) and Houlden (2021) document this dynamic in deindustrializing regions of the UK and US. China has its own version: the resource-dependent cities of the northeast, adjusting to coal and heavy industry decline for two decades, offer a preview of what a faster low-carbon transition might produce elsewhere. Managing the community-level adjustment requires place-based policies that go beyond generic retraining.

There is also a health and environmental justice dimension. The co-benefits of carbon mitigation—reduced air pollution, improved public health—are well documented. Cai et al. (2018), in the Lancet Countdown study, estimated that China's projected carbon dioxide mitigation in the power sector under the Paris Agreement would avoid tens of thousands of premature deaths annually from PM2.5 pollution. These co-benefits are real and substantial, and they strengthen the case for ambitious mitigation. They are also unevenly distributed: the communities that bear the highest compliance costs are not always the communities that receive the highest co-benefits. A coal-mining community that loses jobs when a mine closes may breathe cleaner air, but the health benefit does not compensate the displaced worker for the lost wage. A distributional accounting of transition must include both costs and benefits, and must be honest that they do not always accrue to the same people.

9.1.4 Political Risks: Legitimacy, Enforcement, and the Feedback Problem

The political risks of the transition are the least quantified and the most consequential. They fall into three categories: legitimacy risk, enforcement risk, and feedback risk.

Legitimacy risk refers to the possibility that the transition loses public support because its costs are perceived as unfair or its benefits as inaccessible. The Chinese political system does not face the same electoral pressures as liberal democracies, but it is not immune to legitimacy concerns. Beeson (2010) and Gilley (2012) have argued that performance legitimacy—the regime's ability to deliver economic growth and public goods—is a key source of political stability. A transition that raises energy prices, displaces workers, and disrupts livelihoods without visible compensating benefits can erode that legitimacy, particularly at the local level where the costs are experienced. The central government's emphasis on "people-centered" development and its recurring concern about "mass incidents" (群体性事件) linked to environmental enforcement suggest that legitimacy is a live concern, not a theoretical one.

Enforcement risk refers to the gap between policy on paper and policy in practice. China's environmental enforcement has strengthened dramatically since the 2014 revision of the Environmental Protection Law and the 2016 central environmental inspection campaign, but enforcement remains uneven across regions, sectors, and time. The campaign-style enforcement (运动式治理) that characterizes Chinese environmental governance has a distinctive risk profile: it produces sharp, episodic increases in enforcement intensity that catch firms unprepared, followed by periods of relative relaxation that lull them into complacency. The Shunde factory's eleven-day shutdown was a product of this dynamic—a sudden tightening in anticipation of a provincial review, not a steady enforcement regime. Firms that invest based on average enforcement intensity find themselves exposed during campaign peaks; firms that invest based on peak intensity over-invest relative to the long-run average. Neither strategy is clearly optimal, which is precisely the problem.

Feedback risk is the most subtle. It refers to the possibility that information flowing from firms and localities to policymakers is distorted in ways that produce bad policy. Chinese environmental governance relies heavily on target-based performance evaluation, which creates strong incentives for local officials to report favorable numbers. Cai (2013) and Wallace (2016) document systematic upward bias in economic growth reporting in China; environmental reporting has the same incentive structure and no stronger independent check. If the emissions data that feed into policy design are distorted, the policies will be mis-calibrated. A carbon market that allocates allowances based on inflated baseline emissions will over-allocate and under-price. An energy intensity target measured against a manipulated baseline will be easier to meet on paper than in reality. All regulatory systems face information asymmetries; the combination of top-down target-setting and limited independent verification makes the problem particularly acute in China.

9.1.5 Uncertainty: Technology, Policy, and the Boundary Conditions of Prediction

Beyond risk, in the Knightian sense, the transition is characterized by genuine uncertainty about technology trajectories, policy parameters, and the behavior of key actors. Three types matter for firm strategy and governance.

Technology uncertainty is most acute where multiple trajectories compete. In steel, the choice among hydrogen-based direct reduced iron (DRI), electric arc furnaces with scrap, and carbon capture and storage on conventional blast furnaces is not settled. Each trajectory has different cost structures, infrastructure requirements, and implications for the spatial organization of the industry. A firm that invests in hydrogen DRI today may find in 2035 that CCS retrofits turned out cheaper, or that a breakthrough in electrolysis made green hydrogen abundant, or that carbon prices rose faster than expected and made all three obsolete relative to a technology that does not yet exist. More research cannot resolve this uncertainty; it is inherent in technological competition. Firms cope through modularity, phased investment, and portfolio strategies. Governments cope through technology-neutral policy—carbon pricing rather than technology mandates—and through public investment in pre-commercial technologies that no single firm can justify.

Policy uncertainty is the second type. The direction of Chinese climate policy is clear, but the instruments, timelines, and stringency levels are subject to revision. The national emissions trading system, launched in 2021 with the power sector, has been gradually expanded, but the pace of expansion, the allocation methodology, and the role of offsets remain unsettled. The carbon price has fluctuated between roughly 40 and 90 yuan per ton since launch, with the variation driven as much by regulatory signals as by fundamentals. For a firm planning a fifteen-year investment, that range is material. Dixit and Pindyck (1994) and Bloom (2009) show that policy uncertainty raises the option value of waiting and delays investment. The logic applies with particular force to green investments, which are often irreversible and have long payback periods. Policy uncertainty can slow the transition even when the direction of policy is favorable.

Behavioral uncertainty—about how suppliers, customers, competitors, and foreign governments will respond—is the third type. The EU's Carbon Border Adjustment Mechanism, phased in from 2023, is the clearest case. Chinese exporters know that CBAM will impose a carbon cost on their exports to the EU, but the precise methodology, the treatment of Chinese carbon prices, and the scope of covered products are still being defined. Firms that export to the EU must decide whether to invest in verified emissions reductions now, which may or may not be recognized under CBAM, or to wait for the rules to clarify, which may leave them exposed when the rules take effect. The uncertainty concerns how carbon will be priced at the border, at what level, and with what recognition of Chinese domestic efforts. Resolving it requires international coordination, and it is one of the most consequential unresolved issues in the political economy of the transition.

Table 9-1 summarizes the risk and uncertainty landscape. The categories overlap and interact; the table is a map, not a taxonomy.

Table 9-1: Risk and Uncertainty in China's Manufacturing Transition

Source: Author's synthesis of literature and field evidence. Severity assessments are qualitative judgments, not the output of a formal risk model.

9.2 The Politics of Transition: Winners, Losers, and Compensation

9.2.1 The Distributional Arithmetic of Decarbonization

Every transition produces winners and losers. The interesting questions are who, how many, how much, and what the political system does about it. In Chinese manufacturing the distributional arithmetic is more complex than the aggregate "net positive" framing suggests.

Return to the furniture cluster. The shift to low-VOC coatings creates winners: coating manufacturers who invested in water-based technology, equipment suppliers who sell the new spray booths, larger firms that can afford conversion and use it to differentiate their products, workers who acquire new skills and command higher wages. It creates losers: small firms that cannot afford conversion, workers whose solvent-based spraying skills become obsolete, distributors whose solvent-based product lines shrink, and the local tax base when firms exit. The net effect on the cluster may be positive—higher value-added, lower emissions, better access to green export markets—but the losers are concentrated and identifiable while the winners are diffuse. That asymmetry is the raw material of political conflict.

Three frameworks from the political economy literature help organize the analysis. The compensation principle (Kaldor 1939; Hicks 1939) holds that a policy change is efficient if the winners could in principle compensate the losers and still be better off; it is silent on whether compensation actually occurs, which is a political question. The just transition literature (ILO 2015; UNFCCC 2015; Heffron and McCauley 2018) develops the normative case for compensation and the institutional mechanisms for delivering it. The political economy of reform literature (Rodrik 1996; Fernandez and Rodrik 1991) shows that reforms are more likely to succeed when losers are compensated and when uncertainty about who wins and loses is reduced. The Chinese case lets us examine how these principles operate in a system without electoral competition that is nonetheless sensitive to distributional conflict.

The arithmetic has a spatial dimension that is often underweighted. Costs concentrate in regions specializing in carbon-intensive industry: Shanxi and Inner Mongolia in coal, Hebei in steel, Liaoning in heavy industry. Benefits concentrate in regions specializing in green technology: Jiangsu and Zhejiang in solar and batteries, Guangdong in electric vehicles and green electronics. The interregional transfers implied by the transition are substantial. A coal-mining province that loses tax revenue and employment when mines close does not automatically receive the green investment that replaces it; the solar factory is more likely to be built in a coastal province with better infrastructure, a more skilled workforce, and closer proximity to export markets. The result can be widening regional disparities, which is politically awkward in a system that has made regional balance a priority since the 2000s.

The 2021 "dual control" (双控) campaign on energy consumption and energy intensity illustrates the dynamic. The campaign aimed to reduce energy intensity and cap total energy consumption, and it was implemented with particular stringency in provinces that had lagged on their targets. Zhejiang and Guangdong, among others, imposed rolling power cuts on industrial users that autumn. The cuts were disruptive—factories lost production days, supply chains were interrupted—and they fell disproportionately on energy-intensive SMEs that had not yet upgraded their equipment. The central government responded by clarifying that the campaign should not be implemented in a "campaign-style" (运动式) manner, and the State Council issued guidance emphasizing energy security and "orderly" implementation. The episode shows both the distributional conflict and the feedback mechanism—central correction of local over-implementation—that characterize the Chinese political system.

9.2.2 Winners: The Political Coalition for Green Upgrading

Four clusters of beneficiaries can be identified, each with different interests and different capacities for political influence.

The green technology sector itself—solar manufacturers, wind turbine producers, battery makers, electric vehicle manufacturers, and their supply chains—has grown rapidly. By 2023 China accounted for roughly 80 percent of global solar module production and 60 percent of global wind turbine production. The sector's interests are straightforward: continued demand for green products, protection from foreign trade barriers, access to finance and land. Its political influence operates through industrial policy channels—the Ministry of Industry and Information Technology, the National Development and Reform Commission, local development zones—rather than through electoral politics. Its growth has been a major source of employment and export revenue, which gives it a strong claim on government attention.

A second cluster is the segment of incumbent industry that has invested in green capability and stands to benefit from higher environmental standards—the "green first-mover" constituency discussed in Chapter 8. It includes large steel producers that have invested in ultra-low emission technologies, cement producers that have invested in alternative fuels, and furniture manufacturers that have converted to water-based coatings. These firms have a competitive interest in stringent enforcement, because enforcement raises rivals' costs and rewards their own investments. Their influence runs through industry associations, standard-setting processes, and informal channels to local governments. It is strongest in clusters where the first movers are large enough to anchor the local economy.

The financial sector, or at least the segment committed to green finance, forms a third cluster. Chinese banks have issued substantial volumes of green bonds, and the People's Bank of China has promoted green finance through its macroprudential framework. The PBOC has been a leading international advocate of green finance, and green lending is a growth market at home. The sector's influence is exercised through the financial regulatory system and through its role in allocating credit, which gives it leverage over the pace of industrial upgrading.

The fourth cluster is the segment of the public that benefits from improved environmental quality. This constituency is diffuse, but it has become more politically salient since the air pollution crises of the 2010s. The "not in my backyard" dynamic that once impeded environmental enforcement has reversed in many cities: residents now demand stricter enforcement and protest against polluting facilities. Their influence is exercised through complaints, media coverage, and the central environmental inspection system, which has used public complaints as a source of information about local enforcement failures.

These clusters do not always agree. The green technology sector wants aggressive deployment; the green first-mover incumbents want stringent enforcement; the financial sector wants profitable lending opportunities; the public wants clean air. They form a broad coalition in favor of the transition's direction even when they disagree about its pace and instruments. The political economy of the transition is therefore not a simple contest between green and brown coalitions. It is a contest among green constituencies about the distribution of transition costs and benefits, overlaid on a contest between green and brown interests.

9.2.3 Losers: The Political Economy of Resistance and Adjustment

The losers from the transition are, politically, a more challenging group. They are concentrated in specific regions and sectors, they have identifiable grievances, and they have channels through which to express them. Mildenberger (2020), Stokes (2020), and Colgan, Green, and Hale (2021) have emphasized the role of loser mobilization in slowing or reversing climate policy in democracies. China has no electoral mechanism through which losers can punish incumbents, but the political system is not insulated from distributional conflict. The mechanisms differ; the dynamics are not absent.

The most visible losers are workers and communities in coal-dependent regions. China's coal industry employed roughly 3.5 million workers at its peak in the early 2010s; the number has declined substantially as the industry consolidated, and further decline is inevitable as coal's share of the energy mix falls. Closing a mine can eliminate hundreds or thousands of jobs in a single community with limited alternative employment. The government has responded with early retirement schemes, retraining programs, and regional diversification policies, but the scale of the adjustment is large and the resources devoted to it have been modest relative to the challenge. Zhang et al. (2021) and Wang and Lo (2021) find that compensation has been partial and uneven, and that affected communities have borne a disproportionate share of adjustment costs.

Small and medium-sized enterprises in energy-intensive sectors that cannot afford compliance form a second group. They are not organized politically the way large incumbents are; they lack the resources to lobby effectively and the visibility to attract public attention. But they are numerous, and their exit can have significant effects on local employment and tax revenue. The political system's response has been ambivalent. Programs to support SME upgrading exist—green credit, technical assistance, cluster-level shared facilities—but the enforcement regime does not distinguish sharply between firms that are trying and firms that are not. The Shunde owner's complaint—"I did everything they asked, and I am still shut down"—captures the sense of unfairness that can erode support for the transition among precisely the firms most willing to comply.

Local governments in carbon-intensive regions form a third group, their fiscal capacity dependent on the tax revenue and employment generated by carbon-intensive industry. China's fiscal system gives local governments strong incentives to protect their tax bases, which can translate into lax or selective enforcement. The central environmental inspection system was designed in part to counteract these incentives by auditing local enforcement and holding officials accountable. The underlying fiscal incentive remains, and it shapes the political economy of enforcement at the local level.

Distributional conflicts are processed through several channels. The most formal is the system of compensation and adjustment programs administered by the National Development and Reform Commission and the Ministry of Human Resources and Social Security—real but under-resourced relative to the scale of the challenge. A second channel is the cadre evaluation system, which rewards local officials for meeting environmental targets and penalizes them for social instability; the balance between these objectives is negotiated through target-setting and informal guidance from higher levels of government. A third is the petition system (信访), through which affected individuals and communities can appeal to higher levels of government when local channels fail. The petition system is slow, biased toward those with the resources to navigate it, and sometimes suppressed, but it provides a feedback mechanism absent in a purely top-down system.

The state is not a unitary actor; it is a coalition of agencies and levels of government with different interests and constituencies. The market is not a level playing field; it is a structure of firms with different capacities to adapt. The transition is mediated by the interaction of these actors, and the outcomes are not determined in advance. The governance challenge is to manage distributional conflict in ways that sustain the coalition for the transition while providing meaningful compensation to the losers.

9.2.4 The Compensation Question: Mechanisms, Adequacy, and Legitimacy

If the losers from decarbonization are not compensated, they will resist—through political channels, through non-compliance, or through exit. If they are compensated, the transition can proceed with broader support, but compensation is costly and raises its own distributional questions: who pays, and how much is enough. The Chinese approach has mixed ex ante design, with compensation built into policy from the start, and ex post adjustment, with compensation added in response to political pressure. The mix has worked better in some sectors than others.

The most developed mechanisms are in coal. The 2016 supply-side structural reform (供给侧结构性改革) that accelerated mine closures included a 100 billion yuan fund for worker resettlement, early retirement provisions for older workers, retraining for younger workers, and fiscal transfers to coal-dependent regions. The program was substantial by international standards, but evaluations by Zhang et al. (2021) suggest that it reached only a fraction of affected workers and that retraining had limited success in placing workers in new jobs. The gap between the scale of the problem and the scale of the response is a recurring theme in Chinese just transition policy.

Compensation mechanisms in manufacturing are less developed. The phase-in of carbon pricing, the allocation of free allowances, and revenue recycling are all potential channels, but their design is still being worked out. The national ETS allocates allowances freely to covered entities, which functions as implicit compensation to incumbents; auctioning, which would generate revenue for compensation, has been gradual. The use of carbon market revenue for just transition purposes is not institutionalized, and the debate about allocation is unresolved.

Compensation is not only about money. It is also about voice and recognition. Workers and communities that feel consulted, whose concerns have been heard, are more likely to accept adjustment than those who feel decisions have been imposed on them. The Chinese political system provides limited formal channels for stakeholder consultation in policy design, but informal mechanisms—industry associations, local people's congresses, expert advisory committees—can serve this function, with effectiveness varying widely across regions and sectors. Schlosberg (2007) and Walker (2012) argue that recognition and participation are as important as material compensation in shaping perceptions of fairness. Chinese transition policy has not fully absorbed that insight; it tends to emphasize material compensation over procedural inclusion.

The legitimacy of compensation depends on its perceived fairness, which depends on the visibility of costs and benefits. A compensation program invisible to the people it is meant to compensate will not generate political support, however generous. A program that is visible but perceived as inadequate will generate resentment. The government has invested in communicating the benefits of the transition—cleaner air, green jobs, technological leadership—and has been less effective at communicating the costs and the compensation. The result is a gap between the aggregate narrative and the lived experience of affected firms and workers. That gap is a political risk in its own right.

9.3 Managing Unintended Consequences: Policy Tools and Governance Mechanisms for Risk Mitigation

9.3.1 The Governance Toolbox: A Framework for Assessment

What tools does the Chinese governance system have for managing the risks and unintended consequences of the low-carbon transition, and how well do they work? Four categories organize the analysis: economic instruments, regulatory instruments, informational instruments, and institutional mechanisms. For each I assess the design logic, the evidence on effectiveness, and the failure modes the Chinese context reveals.

"Unintended consequences" covers benign effects as well as corrosive ones. The co-benefits of air quality improvement, the technological spillovers from solar manufacturing, the cluster-level learning that occurs when firms collaborate on compliance—these are unintended and desirable. The governance challenge is not to eliminate unintended consequences but to distinguish the tolerable from the corrosive, and to build feedback mechanisms that allow the system to correct course when corrosive ones appear. That is a cybernetic problem as much as a political one: does the Chinese system have the sensors and the correction mechanisms to perform it?

9.3.2 Economic Instruments: Pricing, Subsidies, and the Problem of Volatility

The first line of defense is economic: price carbon, subsidize green investment, use fiscal transfers to compensate losers. The theoretical case for carbon pricing as the least-cost instrument is well established (Nordhaus 2019; Stern and Stiglitz 2021). The Chinese case is instructive because it combines an ambitious carbon market with an extensive system of subsidies and administrative targets, and the interaction effects are not always benign.

The national ETS, launched in 2021, is the centerpiece of China's carbon pricing architecture. Its design has been deliberately gradual: it covers the power sector first, allocates allowances freely based on historical intensity, and has a soft compliance mechanism. The carbon price has ranged between roughly 40 and 90 yuan per ton, far below the levels most models suggest are needed to meet the 2060 target (World Bank 2023; Duan et al. 2022). The gradual design is defensible on political economy grounds—it gives covered entities time to adjust and avoids a shock to electricity prices—but the price signal is too weak to drive the scale of investment the transition requires. The gap between the price that exists and the price that would be needed is deferred risk. The adjustment will happen eventually, and the longer it is deferred, the sharper it will be.

Volatility is a second concern. The price responds to regulatory signals, allowance allocation decisions, and the availability of offsets, and it can move sharply on policy announcements. For a firm making a long-lived investment, that volatility is a source of uncertainty that can delay or distort decisions. The European experience with the EU ETS—where the price collapsed in 2007 and again in 2013 before the market stability reserve stabilized it—suggests that carbon markets require active management to function as a reliable investment signal. China's ETS has not yet developed the institutional mechanisms—auction reserves, price floors, automatic adjustment—that would smooth the price path. Fell (2016) and Quemin and Trotignon (2021) offer a menu of design options, but the political economy of adopting them is difficult: a price floor raises costs for covered entities, and a price ceiling undermines environmental integrity. The Chinese system has erred on the side of softness, which is understandable but not sustainable indefinitely.

Subsidies are the second economic instrument, deployed at scale. Renewable energy has benefited from feed-in tariffs, tax incentives, and below-market credit; electric vehicles from purchase subsidies, license plate quotas, and charging infrastructure investment. The renewable energy subsidy alone has run into the hundreds of billions of yuan, and its effectiveness in driving deployment is well documented. Subsidies have their own risk profile. They can create dependency, distort investment decisions, and generate rent-seeking. The phase-out of renewable energy subsidies in 2021—the "grid parity" policy—was a deliberate attempt to wean the sector off support. It has been broadly successful, but it exposed developers to market risk they had not previously faced. The transition from subsidy-driven to market-driven deployment is a risky moment in any sector's development, and the governance challenge is to manage it without destroying the industry's momentum.

The fiscal dimension of compensation is the third economic instrument and the least developed. Revenue from carbon pricing—from auctions or a carbon tax—could fund compensation, retraining, and regional diversification. The Chinese ETS currently generates negligible revenue because allowances are allocated freely, so the fiscal capacity for compensation is limited. The broader fiscal system provides transfers to lagging regions, but these are not explicitly linked to the transition and are not calibrated to its distributional effects. Robins, Brunsting, and Wood (2019) and Mendez (2021) argue that dedicated transition funds, capitalized from carbon revenue and directed to affected communities, can be an effective compensation mechanism. China has not established such a fund at the national level, though some provinces have experimented with local versions. The absence of a dedicated fund is a gap in the governance architecture, and it will become more visible asthe transition accelerates.

9.3.3 Regulatory Instruments: Enforcement, Standards, and the Campaign Problem

The second line of defense is regulatory: set standards, enforce them, adjust them as technology improves. China's environmental regulatory architecture has been substantially strengthened since 2014, but it retains features that generate risk for regulated firms.

The most important is campaign-style enforcement (运动式治理). Campaigns concentrate attention and resources on a problem for a defined period, often in response to a crisis or a political priority, and they can produce rapid results. The 2016 central environmental inspection campaign, which sent teams to every province and held thousands of officials accountable, was effective in raising the profile of enforcement. Campaigns have a characteristic risk profile. They are episodic rather than continuous, so firms face sharp swings in enforcement intensity. They are often driven by political calendars rather than by firm-level compliance trajectories, so a firm that has invested in compliance can still be caught by a campaign that targets its sector or region. And they generate perverse incentives: a firm that knows a campaign is temporary may delay compliance until it passes, or over-comply during the campaign to avoid being targeted and then relax afterward.

The "dual control" campaign of 2021 is the clearest recent example. It aimed to reduce energy intensity and cap total energy consumption, and it was implemented with particular stringency in provinces that had lagged on their targets. The rolling power cuts that resulted fell disproportionately on energy-intensive SMEs. The central government's subsequent correction—emphasizing "orderly" implementation and energy security—was a feedback response, but it came after the disruption. The episode shows both the strengths and the weaknesses of campaign governance: it mobilizes action quickly, and it produces collateral damage that is hard to anticipate and costly to reverse.

The alternative is a predictable, rules-based enforcement regime with clear standards, published enforcement schedules, and consistent penalties. Coglianese (2017) and Wiener (2016) argue that predictability is valuable because it allows firms to plan compliance investments. China has moved in this direction—the 2014 Environmental Protection Law introduced daily fines, the 2018 Environmental Protection Tax Law created a more predictable fee structure, and the national ETS publishes allocation rules in advance—but the campaign dynamic persists, particularly at the local level where enforcement is decentralized and political pressures are intense.

The standards themselves are a second regulatory risk. Chinese environmental standards have been tightened repeatedly over the past decade, and tightening will continue as technology improves and the 2060 target approaches. A standard met today may be tightened tomorrow, requiring additional investment. That is a normal feature of an evolving regulatory regime; the issue is the pace and predictability of the tightening. Standards tightened on a published schedule let firms plan. Standards tightened abruptly catch them unprepared. The ultra-low emission standard for coal-fired power plants, introduced in 2014 and tightened since, was tightened on a relatively predictable schedule with a phase-in period that allowed firms to adjust. The VOC standards for furniture coatings have been tightened more abruptly and with less warning, which is part of why the Shunde factory owner felt ambushed.

The third regulatory risk is the interaction between environmental standards and other regulatory objectives. A firm compliant with environmental standards may still be shut down for violating safety standards, land-use rules, or labor regulations. Compliance is a multi-dimensional problem, and a firm that is green may still be vulnerable to enforcement action on other grounds. Black (2008) and Gunningham and Grabosky (1998) emphasize that firms operate in a complex regulatory environment and that interactions between regimes produce unintended consequences. In China the campaign dynamic often cuts across regulatory domains—a single campaign may target environmental, safety, and land-use violations simultaneously—which increases the risk that a compliant firm will be caught up in enforcement action not aimed at it.

9.3.4 Informational Instruments: MRV, Disclosure, and the Feedback Problem

The third line of defense is informational: measure emissions, report them, verify them, and use the information to adjust policy. The MRV infrastructure discussed in Chapter 3 is the foundation, and its quality determines the effectiveness of both carbon pricing and regulatory enforcement.

The feedback problem identified earlier is the central challenge. If the emissions data that feed into policy are distorted, the policies will be mis-calibrated. China's MRV system has improved substantially—the national ETS requires covered entities to report emissions using standardized methodologies, and third-party verification is mandatory—but verification quality remains uneven. The verification market is young, verifiers are not always independent of the entities they verify, and penalties for misreporting are modest relative to the gains from under-reporting. Zhang (2020) and Duan et al. (2022) document these weaknesses, and the international experience with emissions trading—the EU ETS, the California cap-and-trade program—suggests that verification quality is a persistent challenge requiring continuous investment and independent oversight.

Disclosure is a second informational instrument. China has made progress in requiring listed companies to disclose environmental information, and the 2021 Environmental Information Disclosure Reform expanded the scope of mandatory disclosure. Quality is variable and enforcement of disclosure requirements is weak. Wang et al. (2020) and Lu and Abeysekera (2021) find that disclosure is often selective: firms disclose favorable information and withhold unfavorable information. The information environment is better than it was a decade ago and still far from the transparency effective governance requires.

The third informational instrument is the use of data for policy learning. The system collects vast amounts of environmental data—from monitoring stations, firm reporting, satellite observation—but the capacity to analyze and act on it is uneven. The central government has invested in environmental big data platforms, and some provinces have developed sophisticated monitoring systems, but integration across agencies and levels of government remains incomplete. Liu et al. (2021) and Zhang and Wen (2022) suggest that the potential is large but the institutional barriers—data silos, incompatible formats, limited analytical capacity—are substantial. The feedback loop is only as good as the weakest link in the data chain.

9.3.5 Institutional Mechanisms: Coordination, Experimentation, and the Learning System

The fourth line of defense is institutional: the mechanisms through which the system coordinates across agencies and levels of government, experiments with policy designs, and learns from experience. China's governance system has distinctive strengths here—large-scale experimentation, rapid resource mobilization, vertical integration that transmits central priorities to local levels—and distinctive weaknesses: limited horizontal coordination, weak independent oversight, and top-down target-setting that can suppress local information.

Policy experimentation is among the most studied features of Chinese governance. The pilot (试点) model, in which a policy is tested in a designated jurisdiction before scaling up, has been used extensively in environmental governance: the carbon trading pilots that preceded the national ETS, the low-carbon city pilots, the green finance pilots in Zhejiang and Guangdong. Heilmann (2008) and Zhu and Zhao (2021) find that the model can generate information about what works, with limitations: pilots are often selected for political reasons rather than representativeness, lessons are not always systematically evaluated, and scaling up can lose the contextual knowledge that made the pilot successful. The carbon trading pilots generated valuable experience with allowance allocation and MRV, but the national ETS that emerged is a different design, not a scaled-up pilot; it reflects national-level political compromises.

Coordination is a second institutional feature. The transition cuts across the NDRC, MIIT, MEE, MOF, PBOC, and others, and coordination among them is a persistent challenge. The "dual carbon" leading group established under the State Council in 2021 was an attempt to improve coordination at the highest level, followed by sectoral working groups and the integration of carbon targets into the Five-Year Plan process. Pollitt (2003) and Christensen and Lægreid (2007) show that coordination is always difficult and that vertical integration can be an advantage in transmitting priorities and a disadvantage in integrating diverse perspectives. Coordination at the top can produce a unified target that is not well-calibrated to heterogeneous conditions on the ground.

The learning mechanism is the third institutional feature and the one most directly relevant to managing unintended consequences. A system that can detect unintended consequences and correct course is more resilient than one that cannot. China's feedback mechanisms—the petition system, central environmental inspection, the cadre evaluation system, the policy research community—vary in effectiveness. Central environmental inspection has functioned as a powerful feedback mechanism: it has uncovered local enforcement failures, held officials accountable, and generated information that informed central policy. But it is periodic, not continuous, and its findings are filtered through the political process before reaching the public. The policy research community—universities, think tanks, international organizations—provides a degree of independent analysis, but its influence on policy is variable and its access to data is often constrained.

The learning system is real but imperfect. It detects some unintended consequences and corrects some course deviations, but it is better at detecting failures of implementation than failures of design, better at responding to crises than at anticipating them, and better at learning within the system than at learning from outside it. The European Union's governance of the energy transition has its own weaknesses, including slow decision-making and fragmented authority; no system is fully equipped to manage the complexity of the transition. The question is whether a system is capable of adaptive improvement. On that criterion China's record is mixed: it has corrected course when problems became visible—the dual control correction, the subsidy phase-out, the ETS design adjustments—and it has been slower to anticipate problems before they become visible.

9.3.6 Managing the Relocation-Upgrade Nexus: A Sectoral Application

Return to the sector that opened the chapter: furniture, and more broadly the labor-intensive sectors that face the most acute relocation-upgrade nexus. The governance challenges here are distinctive because the firms are small, the clusters fragmented, and the resources for risk management limited. The general tools described above—carbon pricing, standards, MRV, coordination—operate at a scale often above the firm level. The question is how to make them work for the firms most vulnerable.

Evidence from Shunde and from comparable clusters in Zhejiang and Jiangsu suggests three lessons. The timing of enforcement matters as much as its stringency. Firms can adapt to stringent standards if they have time to plan and invest; they cannot adapt to abrupt changes that arrive without warning. The Shunde owner's complaint was not that the standards were too strict but that the timing was unpredictable. A predictable enforcement schedule—published in advance, with phase-in periods calibrated to firm size—would let firms plan and would reduce the risk of abrupt disruption. The compliance assistance model, in which regulators work with firms to identify compliance pathways before imposing penalties, has been used in some Chinese localities and has shown promise in reducing disruption.

The cluster level is the right level for risk management in fragmented sectors. Individual SMEs cannot afford the fixed costs of compliance infrastructure—shared wastewater treatment, shared coating facilities, shared MRV systems—but clusters can. The shared facility model, in which a cluster-level entity invests in compliance infrastructure and charges user fees, has been used in some Chinese industrial parks and has reduced the cost of compliance for small firms. The governance challenge is to design the facility so that it is financially sustainable, equitably accessible, and accountable to its users. Ostrom's (1990) work on club goods and common-pool resources offers useful principles, but the Chinese context adds complications: the shared facility is often built by a local government or a state-owned enterprise, and its governance may not reflect the interests of the small firms that use it.

Compensation for firms that cannot adapt must be real. Some firms will exit, and their workers will need support. The governance system has programs for worker retraining and early retirement, but their reach is limited and their effectiveness uneven. The Shunde owner, if he exits, will receive little beyond the salvage value of his equipment; his workers will receive unemployment insurance for a limited period and whatever retraining the local government provides. The gap between the rhetoric of just transition and the reality of adjustment assistance is a source of legitimacy risk, and it will become more visible as the transition accelerates.

Furniture is not the most carbon-intensive sector in Chinese manufacturing, and its transition is not the most dramatic. It is representative of a large class of sectors—textiles, toys, electronics assembly, food processing—where firms are small, clusters fragmented, and the transition experienced as a series of discrete shocks rather than a smooth gradient. The governance challenge in these sectors is to manage the adjustment in ways that preserve the cluster's productive capacity while reducing its environmental footprint, which requires predictable enforcement, cluster-level infrastructure, and real compensation. China has made progress on the first, uneven progress on the second, and limited progress on the third. On current evidence, the binding constraint is the third: until adjustment assistance reaches the small firms and workers who bear the concentrated costs, each enforcement campaign will generate a new cohort of the disgruntled, and the coalition for the transition will narrow.

9.4 Conclusion: Governance in the Face of Uncertainty

The low-carbon transition in Chinese manufacturing generates economic, social, and political risks that are unevenly distributed across firms, sectors, and regions. These risks interact with genuine uncertainty about technology, policy, and international responses, and the interaction is more corrosive to firm-level planning than either risk or uncertainty alone. The political system processes the resulting distributional conflicts through a mix of compensation, cadre evaluation, and informal feedback. Compensation is partial, feedback is imperfect, and the legitimacy of the transition depends on perceived fairness in ways the current policy mix does not fully address.

The governance tools available—economic, regulatory, informational, institutional—are real but uneven in effectiveness. Carbon pricing is too soft and too volatile to drive the scale of investment required. Enforcement is too campaign-driven to provide the predictability firms need. MRV is improving but remains vulnerable to distortion. Coordination is strong vertically and weak horizontally. The learning system corrects course when problems become visible but is slower to anticipate them.

None of this argues against the transition. The case for decarbonization is overwhelming, and the Chinese manufacturing sector's capacity to adapt has been demonstrated repeatedly over four decades. The argument is that the transition's success depends on governance choices that are not automatic: the design of compensation, the predictability of enforcement, the quality of information, the responsiveness of the learning system. These choices can be made well or badly. The evidence in this chapter suggests the Chinese system is making some of them well and others badly, and that the weakest link is compensation for the small firms and workers who bear the concentrated costs of adjustment.

The chapter's contribution to the book's overall argument is to qualify it. The claim that green governance capability can be an endogenous driver of upgrading holds for firms and clusters with the resources and institutional support to develop that capability. It does not hold for firms caught between stringent standards and inadequate compensation, or for clusters exposed to enforcement campaigns they cannot anticipate. The transition's benefits are real and unevenly distributed, and the governance challenge is to manage the distribution in ways that sustain the coalition for the transition. That challenge is political as much as technical, and better data alone will not resolve it.

The next chapter takes up the policy mix and strategic outlook. It asks how the tools analyzed here—carbon pricing, standards, MRV, compensation, coordination—can be combined into a coherent package that supports the transition from cost-based to green competitive advantage. The analysis in this chapter suggests the package must include instruments for compensation and learning as well as instruments for pricing and regulation. A policy mix that prices carbon without compensating losers, enforces standards without providing predictability, or measures emissions without verifying them will generate the unintended consequences described here. The design of that mix is the central governance question of the transition, and it is where the next chapter begins.

Note on sources: This chapter draws on the sociotechnical transitions literature (Geels and Schot 2007; Kemp, Schot, and Hoogma 1998; Turnheim and Geels 2012), the risk and uncertainty literature (Knight 1921; Dixit and Pindyck 1994; Bloom 2009), the just transition literature (ILO 2015; Heffron and McCauley 2018), and the political economy of decarbonization literature (Mildenberger 2020; Stokes 2020; Colgan, Green, and Hale 2021). Evidence on Chinese environmental enforcement draws on the cadre evaluation literature, the campaign governance literature, and the author's field interviews in Shunde, Foshan, Shaoxing, and Shenzhen between 2019 and 2023. Carbon market data are from the World Bank's State and Trends of Carbon Pricing (2023) and Duan et al. (2022). Employment estimates draw on IRENA (2019) and Cai et al. (2014, 2018). The Shunde factory case is based on the author's interviews; details have been altered to protect the identity of the interviewee. The risk severity assessments in Table 9-1 are the author's qualitative judgments and should not be read as the output of a formal risk model.