Chapter 1: From Compliance to Capability: Reframing Green Governance in Global Value Chains
In the summer of 2023, a furniture factory in Foshan received a letter from a European retail buyer. Three paragraphs, polite. Within eighteen months, every container shipped to that buyer would need a product carbon footprint verified by a third party, calculated according to the buyer's own methodology, and accompanied by a five-year reduction plan. The owner had spent two decades building the business on competitive pricing, reliable delivery, and a knack for absorbing whatever specification changes the buyer sent his way. He read the letter twice, then called his accountant with a question that would have been unthinkable a decade earlier: how much would it cost to know, precisely, how much carbon left his factory gates with every chair?
That question, and the organizational changes it set off, is the subject of this book. Variations of the buyer's letter have arrived at thousands of Chinese factories over the past several years, from retailers in the European Union, from brand owners in North America, from domestic platforms and state-owned enterprises responding to their own reporting obligations. What makes the letter worth examining is not its content but its consequences. For the Foshan factory, the carbon footprint demand began as one more item on a list that had been growing since the first ISO 9001 audits of the 1990s. Meeting it forced changes the owner had not anticipated: a data system that revealed energy waste he had never measured, a packaging redesign that cut freight volume by 11 percent, a switch in adhesive suppliers that eliminated a hazardous input, and eventually a line of products marketed under a low-carbon label that sold at a modest premium in the domestic market.
The factory owner did not set out to build a capability. He set out to keep a buyer. But the measurement, the redesign, the supplier conversations changed how he understood his own business. He now talks about carbon the way he once talked about cost: as something to be managed, measured, reduced.
His experience is not typical. It is not even common. But it is real, and it points to a possibility that both policy and academic discourse have been slow to recognize: that the discipline of measuring and managing carbon, water, and materials can generate productive knowledge with value beyond the compliance report.
1.1 The Compliance Trap: Why Regulatory Pressure Alone Fails
The logic of compliance is simple. A regulator, or a buyer acting as a private regulator, sets a standard. The firm measures itself against the standard, identifies gaps, and closes them at the lowest feasible cost. Success is meeting the standard. Failure is not meeting it. The firm's internal organization adapts to this logic: an environment, health, and safety department is created or expanded, reporting lines are drawn, consultants are hired, and the compliance function becomes a specialized silo whose job is to keep the firm out of trouble. The silo talks to the regulator and to the buyer's auditors. It does not talk much to the production department, the design team, or the procurement office, because those functions have their own metrics and their own priorities, and compliance is not one of them.
This pattern is rational in a narrow sense. It minimizes the cost of meeting a given standard. But it has a structural weakness that becomes visible when standards tighten or multiply. Because the compliance function is isolated from the rest of the firm, the knowledge it generates—about where emissions come from, which processes are inefficient, which suppliers are risky—does not flow into decisions that shape the firm's product mix, its process technology, or its market positioning. The firm meets the standard and learns little. When the standard changes, it starts over.
I have seen this pattern repeatedly in the furniture clusters of Guangdong and Zhejiang. A factory receives a buyer's restricted substances list, forwards it to purchasing, and asks suppliers to sign declarations. The declarations come back signed. Nobody in the factory knows whether they are accurate, because nobody has the laboratory capacity or the process knowledge to check. The box is ticked. The underlying risk—that a supplier's formulation has changed, that a batch of adhesive contains a substance that will fail a future audit—remains unaddressed. The firm has complied without building any capacity to know.
Three features interlock to produce this outcome. The first is measurement without understanding. Firms subject to carbon reporting requirements often rely on spend-based estimates or on default emission factors provided by consultants. These estimates satisfy the reporting requirement, but they do not tell the firm where its actual emissions are concentrated. A furniture factory that estimates its footprint from electricity bills and freight invoices will learn that electricity and transport are significant, which it already knew. It will not learn that the drying kilns are running at 15 percent higher energy intensity than a benchmark plant, or that a particular coating line generates disproportionate volatile organic compound emissions because of a poorly maintained ventilation system. Measurement without understanding produces a number, not a lever.
The second feature is the separation of environmental performance from operational decision-making. In the compliance model, environmental data flows upward to regulators and buyers, not sideways to production managers. The production manager who decides how long to run a kiln, or which grade of resin to use, or how to schedule a paint line, does not see the environmental consequences of those decisions in any form that connects to his performance metrics. The metrics that matter—output, yield, on-time delivery, unit cost—are silent on carbon and materials. Operational knowledge and environmental knowledge develop on separate tracks, and the potential for one to inform the other is lost.
The third feature is the treatment of compliance as a boundary condition rather than a design parameter. When a standard is seen as a constraint to be satisfied, the search for solutions is narrow: the firm looks for the cheapest way to meet the constraint, not for ways to redesign the product or process so that the constraint becomes irrelevant or even advantageous. A factory told to reduce VOC emissions from its coating line will typically install abatement equipment—a scrubber, an oxidizer—at the end of the pipe. A factory that treats VOC reduction as a design problem might switch to water-based coatings, which eliminates the problem at the source, reduces fire risk, and opens the possibility of selling into markets with stricter indoor air quality requirements. The first response is compliance. The second is capability. The difference between them is not a difference in regulatory stringency. It is a difference in how the firm frames the problem.
The compliance trap is not a failure of will or intelligence. It is a predictable outcome of an institutional environment in which environmental performance is verified by external actors against external standards, and in which the firm's internal reward systems do not recognize environmental improvement as a source of value. Changing the outcome requires changing the environment: the stringency of the standard matters, but so do the way the standard is implemented, the information that flows with it, and the organizational space firms have to interpret it.
Consider two approaches to carbon disclosure that have emerged in Chinese manufacturing over the past five years. The first is the supplier questionnaire model, in which a brand owner sends a standardized spreadsheet to its suppliers, asks them to fill in their energy consumption and emissions, and aggregates the results for its own reporting. The supplier's incentive is to fill in the spreadsheet quickly and with numbers that will not trigger follow-up questions. Data quality is poor, the supplier learns nothing, and the brand owner's aggregate number is, as one sustainability manager at a Shenzhen-based electronics firm put it, "a work of fiction with a decimal point." The second is the co-investigation model, in which the brand owner sends engineers to work with the supplier on a specific production line, measures energy and materials together, identifies savings, and shares the gains. The supplier learns where its losses are. The brand owner gets better data and a more reliable supplier. Both are forms of private governance. The difference is whether the governance process transfers knowledge or merely extracts data.
The compliance trap also has a macro dimension. At the level of an industrial cluster or a region, a compliance-only approach to environmental regulation tends to select for firms that are good at absorbing compliance costs—typically larger, more formal firms—and to penalize firms that cannot absorb them, which are often smaller and more informal. The result can be a consolidation of production into fewer, larger units, which may be good for average emissions intensity but bad for employment and for the diffuse innovative capacity that clusters provide. Industrial clusters generate competitive advantage through a combination of local knowledge spillovers, specialized labor pools, and dense supplier relationships. A regulatory approach that treats compliance as a firm-level cost and ignores the collective dimension of cluster governance can destroy those advantages without replacing them. The cluster's small firms do not learn to be green; they learn to be gone.
There is a further problem specific to the current phase of global value chain governance. The standards that matter to Chinese exporters are increasingly set not by Chinese regulators but by lead firms and by foreign regulators acting through trade policy. The European Union's Carbon Border Adjustment Mechanism, which began its transitional reporting phase in October 2023, is the most prominent example. CBAM requires importers of steel, aluminum, cement, fertilizer, hydrogen, and electricity to report the embedded emissions of their purchases, and from 2026 to purchase certificates corresponding to those emissions. For a Chinese steel exporter, the mechanism creates a direct financial exposure to its carbon intensity. It also creates an information problem: the exporter must be able to calculate its emissions according to a methodology defined in Brussels, verify them with data that satisfies European auditors, and demonstrate that any reductions are real. A firm that has treated carbon as a compliance box will struggle to produce this information. A firm that has built the capability to measure and manage emissions at the process level will find the reporting requirement a manageable extension of what it already does.
The compliance-only model is a growing liability in a trading environment where carbon is becoming a product attribute that must be documented, verified, and priced. The firms that escape it will be the ones that treat green governance as a capability. The ones that do not will find themselves paying for their own ignorance, either through the price of CBAM certificates or through the loss of buyers who need reliable carbon data and cannot get it.
1.2 Green Governance as Capability: A New Conceptual Framework
What does it mean to treat green governance as a capability? The term has a specific meaning in the strategy and innovation literature that goes back to Edith Penrose's work on the growth of the firm and, more directly, to the resource-based view articulated by Birger Wernerfelt and Jay Barney in the 1980s and 1990s. In that tradition, a capability is a firm's capacity to perform a particular activity reliably and at scale, built through learning and investment, and difficult for competitors to imitate because it is embedded in organizational routines rather than in a single machine or a single person. A capability is not a resource in the sense of a stock of money or a piece of land. It is a way of doing things that the firm has learned.
Green governance capability, as I use the term, has four dimensions. Each is a distinct organizational capacity, and each can be built or neglected independently.
The first is measurement capability: the ability to produce reliable, granular data on material and energy flows at the level of the product, the process, and the facility. This is harder than it sounds. It requires meters, sensors, and data infrastructure, but it also requires a chart of accounts that maps energy and materials to products, and a culture in which measurement is treated as a management tool rather than a reporting obligation. A firm that can measure its emissions at the process level knows something about its production that a firm relying on spend-based estimates does not.
The second dimension is process capability: the ability to redesign products and processes to reduce environmental impact without sacrificing cost, quality, or delivery performance. This is where green governance connects most directly to the innovation literature. Eco-innovation is not a separate category of innovation. It is the application of the firm's existing design and engineering capabilities to a new set of constraints and objectives. The furniture factory that switched to water-based coatings was not doing something fundamentally different from what it did when it redesigned a chair for flat-pack shipping. It was applying the same design capability to a different problem. The difference is that the environmental problem requires a particular kind of knowledge—about chemistry, about energy systems, about life-cycle impacts—that the firm may not have in-house and must learn to access.
The third dimension is supply chain capability: the ability to extend green governance beyond the firm's own gates to its suppliers, and to do so in a way that improves rather than merely audits their performance. This is the dimension that connects most directly to the GVC literature. Humphrey and Schmitz, in their work on chain governance and upgrading, distinguished between different modes of governance—market, network, hierarchy, and quasi-hierarchy—and argued that the mode of governance shapes the upgrading opportunities available to suppliers. A supplier in a quasi-hierarchical relationship with a lead firm receives detailed specifications and is audited against them; its upgrading path is constrained by the lead firm's requirements. A supplier in a network relationship has more room to innovate, but also more uncertainty about whether its innovations will be valued. Green governance capability at the supply chain level means being able to work with suppliers to identify and realize environmental improvements, rather than simply passing down requirements and checking compliance. It requires a relationship that supports joint problem-solving.
The fourth dimension is institutional capability: the ability to navigate and shape the regulatory and market environment in which the firm operates. This includes the ability to anticipate regulatory change, to participate in standard-setting processes, to access green finance, and to communicate credibly with buyers, regulators, and the public about environmental performance. A firm with strong institutional capability does not simply react to carbon pricing or disclosure requirements. It positions itself to benefit from them.
These four dimensions are not a checklist. They are an analytical framework for understanding why some firms and clusters make the transition from compliance to capability and others do not. The framework implies that green governance capability is path-dependent: firms build it incrementally, through specific projects and specific problem-solving episodes, and the capability accumulates in organizational routines and in the skills of particular people. It also implies that the capability is relational: it depends on the firm's relationships with suppliers, buyers, regulators, and knowledge providers, and it can be strengthened or weakened by the governance of those relationships.
The distinction between compliance and capability is not a distinction between good firms and bad firms. It is a distinction between two ways of organizing the relationship between environmental performance and business strategy. In the compliance model, environmental performance is a constraint imposed from outside. In the capability model, it is a source of information and a target for innovation. The same firm can operate in both modes at different times, or in different parts of its business. The question for research and policy is what shifts the balance.
Treating green governance as a capability does not mean that regulation is unnecessary or that firms will green themselves if left alone. The evidence points the other way. The firms that have built the strongest green governance capabilities are, almost without exception, firms that faced demanding external requirements—from regulators, from buyers, or from both—and had the organizational resources to respond to them. The capability is a response to pressure, not a substitute for it. What matters is the form of the pressure and the space it creates for learning. A regulatory regime that specifies outcomes and allows firms to find their own ways of meeting them is more likely to build capability than a regime that specifies technologies. A buyer relationship that involves joint problem-solving is more likely to build capability than one that relies on audits and declarations. The design of the pressure matters as much as its intensity.
This point has a corollary relevant to the Chinese context. China's dual carbon goals—peak emissions before 2030 and carbon neutrality before 2060—have created a policy environment in which carbon performance is becoming a mainstream business concern rather than a niche compliance issue. The goals have been translated into a growing array of sectoral targets, carbon market rules, green finance standards, and local government initiatives. For manufacturing firms, the effect is not a single new regulation but a dense and shifting set of expectations that touch energy procurement, product design, supplier management, and export strategy. Firms that treat these expectations as a series of compliance boxes will find themselves perpetually behind. Firms that build the capability to measure, manage, and improve environmental performance will find that the same capability serves them across multiple requirements. The dual carbon goals are not just a constraint. They are a demand signal for capability.
The framework also has implications for how we think about industrial upgrading. The GVC literature has traditionally defined upgrading in terms of the functions that firms perform: process upgrading (doing the same thing more efficiently), product upgrading (making better products), functional upgrading (moving into higher-value functions such as design or marketing), and inter-sectoral upgrading (moving into new sectors). Green governance capability cuts across these categories. It can support process upgrading by reducing energy and material costs. It can support product upgrading by enabling products that meet environmental requirements or carry environmental attributes. It can support functional upgrading by giving firms the knowledge and credibility to offer sustainability services to customers. And it can support inter-sectoral upgrading by creating capabilities that are valuable in new sectors—battery manufacturing, for example, or green building materials.
The relationship between green governance capability and upgrading is not automatic. A firm can build measurement capability and use it only for reporting. A firm can build process capability and use it to reduce costs without changing its position in the value chain. The capability is a necessary condition for green upgrading, not a sufficient one. What converts capability into upgrading is the firm's strategic intent and the market's willingness to reward environmental performance. Both are variable. The rest of this book examines the conditions under which they align.
I use the term "green governance" rather than the more familiar "environmental management" or "sustainability" for a reason. Governance signals that the capability is not confined to the firm's internal operations. It extends to the relationships through which the firm's environmental performance is shaped: relationships with suppliers, with buyers, with regulators, with financiers, and with the local institutions of the clusters in which the firm is embedded. Governance, in this usage, is the exercise of authority and coordination across organizational boundaries. A firm that can govern its supply chain's environmental performance has a capability that a firm that can only manage its own operations does not. The term also connects to the GVC literature's use of "chain governance" to describe the coordination of activities across borders. Green governance capability is the firm-level counterpart to chain governance: it is what a firm needs to participate effectively in a chain that is increasingly organized around environmental requirements.
The term also carries a normative charge. Governance implies legitimacy and accountability, not just efficiency. A firm that governs its environmental performance is accountable for it, to regulators, to buyers, to communities, and to its own workers. The capability is not just a technical capacity. It is a capacity for responsible action. Firms build green governance capability for instrumental reasons—to keep buyers, to avoid penalties, to reduce costs—and the normative dimension is often secondary. But the capability, once built, creates the possibility of accountability that did not exist before. A firm that can measure its emissions can be held responsible for them. A firm that cannot measure them cannot. The measurement infrastructure is also an accountability infrastructure.
The development of green governance capability is not a purely firm-level process. It depends on the infrastructure of measurement, verification, and reporting that surrounds the firm—what Chapter 3 calls the MRV system (monitoring, reporting, and verification). It depends on the policy architecture that prices carbon and sets standards. It depends on the collective institutions of clusters. And it depends on the governance of global value chains, which determines what information and what requirements flow to suppliers. The capability is firm-level, but the conditions for building it are systemic.
1.3 Research Questions, Methodology, and Book Structure
Three questions drive the book. The first is descriptive: how are Chinese manufacturing firms actually governing their environmental performance as they embed themselves in global value chains? The second is analytical: under what conditions does green governance become a capability that drives upgrading, rather than a compliance cost that constrains it? The third is prescriptive: what policies, institutions, and firm strategies would shift the balance toward capability-building?
The descriptive question is harder to answer than it might seem, because the data on firm-level environmental governance in China is uneven. Official statistics report aggregate energy consumption and emissions by sector and by region, but they do not reveal how firms measure, manage, and improve their performance. The carbon market's MRV system produces data on covered firms, but coverage is limited and the data is not publicly available at the firm level. Buyer audits produce data, but it is proprietary. Our knowledge of what firms actually do is therefore based on case studies, surveys, and interviews rather than on comprehensive administrative data. This book draws on all three.
The analytical question requires a framework that connects firm-level capabilities to chain-level governance and to the broader institutional environment. I develop that framework in Chapter 2, drawing on the GVC literature, the innovation literature, and the growing body of work on green industrial policy. The framework's key move is to treat green governance capability as an intermediate variable between external pressure and upgrading outcomes. External pressure—from regulators, buyers, and markets—does not automatically produce upgrading. It produces upgrading when it induces firms to build the capability to measure, manage, and improve environmental performance, and when the firms' strategic intent and market conditions allow them to convert that capability into a better position in the value chain.
The prescriptive question is where the book's argument has the most at stake. If green governance is a capability, then policies that treat it as a compliance cost are missing the point. The book argues for a policy approach that combines demanding outcome-based standards with support for capability-building—measurement infrastructure, technical assistance, cluster-level coordination, and access to green finance. It also argues that the governance of global value chains matters: buyer practices that extract data without transferring knowledge are less likely to produce upgrading than practices that involve joint problem-solving.
The methodology is comparative and case-based. I examine three groups of industries, distinguished by factor intensity: labor-intensive industries (furniture, textiles, and apparel), technology-intensive industries (electronics and electrical equipment), and capital-intensive industries (steel, cement, and chemicals). The factor-intensity distinction matters because the pathways to green governance capability differ by industry. Labor-intensive industries face cost pressure that makes environmental investment difficult, but they also have relatively simple production processes that can be reorganized with modest investment. Technology-intensive industries have the R&D capacity to develop eco-innovations, but they face rapid technological change that can strand environmental investments. Capital-intensive industries have the scale to invest in process substitution, but they also have long-lived assets that can become stranded in a low-carbon transition. Chapters 5, 6, and 7 examine each group in turn.
Within each group, I use a combination of firm-level case studies and cluster-level analysis. The case studies are based on factory visits, interviews with managers and engineers, and reviews of firm documents. The cluster-level analysis draws on my earlier research on industrial clusters and on a new round of fieldwork conducted in the furniture clusters of Guangdong, the electronics clusters of the Yangtze River Delta, and the steel and chemical clusters of Hebei and Shandong. The cluster is the right unit of analysis for several reasons. Clusters are where the collective infrastructure of green governance—shared testing facilities, common effluent treatment, energy service companies, industry associations—is located. They are also where the informal institutions that shape firm behavior—norms of cooperation, expectations about regulatory enforcement, patterns of knowledge sharing—are reproduced. And they are where the tension between compliance and capability is most visible, because cluster-level environmental regulation often takes the form of collective targets and collective enforcement.
Chapter 9 presents an extended case study of the furniture industry, which serves as a lens for examining the book's argument in a traditional advantageous industry facing the dual pressures of the dual carbon goals and changing export markets. The furniture case is instructive because it is a labor-intensive industry with low barriers to entry, intense cost competition, and high exposure to global value chain governance. It is also an industry where the environmental issues are tangible—VOCs from coatings, formaldehyde from adhesives, energy from drying, waste from offcuts—and where the solutions are within reach of firms with modest technical capacity. If green governance capability can drive upgrading in furniture, it can drive it almost anywhere. If it cannot, the limits of the capability argument will be visible.
Chapter 2 develops the theoretical foundations, situating the argument in the GVC literature, the innovation literature, and the literature on green industrial policy. It introduces the concept of the green capability turn and explains how it differs from earlier approaches to environmental governance in value chains. Chapter 3 examines the measurement infrastructure—carbon accounting, MRV systems, and the boundaries of evidence—that underpins any claim about green governance. Without reliable measurement, the capability argument is empty. Chapter 4 analyzes the policy architecture of China's dual carbon goals, including carbon pricing, green finance, and sectoral targets, and assesses how the incentive landscape shapes firm behavior. Chapters 5, 6, and 7 present the industry studies. Chapter 8 examines clusters as arenas of green governance, focusing on collective capability, informal institutions, and local coordination. Chapter 9 presents the furniture case in detail. Chapter 10 concludes by restating the argument, assessing its limits, and identifying the policy and research implications.
My previous research examined the internationalization of industrial clusters—how clusters of small and medium-sized firms in China inserted themselves into global value chains and upgraded over time. That work emphasized the role of local institutions, supplier relationships, and learning by exporting in the upgrading process. This book extends that analysis into the environmental domain. The question is whether the same mechanisms that drove upgrading in the past—learning through buyer relationships, collective action within clusters, institutional support from local governments—can drive green upgrading in the present. The answer, developed across the chapters, is that they can, but only if the governance of the value chain and the design of environmental policy create the space for learning rather than simply demanding compliance.
The book does not argue that Chinese manufacturing is on an inevitable path to green upgrading. The evidence does not support that claim. Many firms remain trapped in compliance, and some are not even compliant. It does not argue that the dual carbon goals will automatically produce capability-building. The goals create pressure, but pressure can produce defensive responses as easily as it can produce learning. It does not argue that global value chain governance is benign. Lead firms' sustainability requirements can be extractive, and the data they demand can be used to shift costs and risks onto suppliers without sharing the benefits. Nor does it argue that environmental performance always pays. In many market segments, it does not, at least not in the short run. The book's argument is conditional: under specified conditions, green governance capability becomes a driver of upgrading. Identifying those conditions, and understanding why they are met in some firms and clusters and not others, is the task of the chapters that follow.
The Foshan factory owner, for his part, did not set out to build a capability. He set out to keep a buyer. But the process of meeting the buyer's requirement—the measurement, the redesign, the supplier conversations—changed how he understood his own business. He now talks about carbon the way he once talked about cost: as something to be managed, measured, and reduced. That shift in understanding, from compliance to capability, is the subject of this book. It is happening in thousands of factories across China, unevenly, imperfectly, and often for reasons that have little to do with environmental conviction. But it is happening.