Chapter 2: Theoretical Foundations: Global Value Chains, Industrial Upgrading and Green Governance

A container ship carrying a Foshan sofa to Rotterdam also carries a question. Who decides what the sofa is made of, how it is finished, and what carbon number is attached to it before it leaves the factory gate? For most of the past three decades, the buyer decided. A European retailer, a North American brand, a Japanese trading house specified the foam density, the fabric composition, the packaging, the delivery window. The Chinese manufacturer competed on price and reliability within those specifications. The specification was the governance mechanism. The price was the outcome.

That arrangement is under strain, and the strain is theoretical before it is practical. The frameworks we use to explain how global value chains work—who holds power, how firms upgrade, what governs the relationship between lead firms and suppliers—were built for a world in which cost was the primary coordination variable and environmental performance was an externality managed through compliance. That world has not disappeared. It has acquired a second coordination variable, and the second variable behaves differently from the first. Cost advantages erode predictably as wages rise. Green advantages compound, because they depend on capabilities that are cumulative, path-dependent, and difficult to imitate. A theory that treats green governance as a subset of compliance will misread what is happening in the factories and clusters where the transition is actually taking place.

This chapter builds the analytical framework the rest of the book uses. It does not attempt a comprehensive survey of three vast literatures. It selects from each the concepts that carry explanatory weight for the question at hand: how does the position of Chinese manufacturing firms in global value chains change when green governance becomes a condition of participation, and under what conditions does that change amount to upgrading rather than compliance? Global value chain theory explains the structure of power and the terms of participation. Industrial upgrading theory explains the sequence and the capabilities through which firms move up the chain. Green governance theory—spanning corporate social responsibility, socio-technical transitions, and the environmental upgrading debate—explains why the environmental dimension can serve as a source of endogenous capability, over and above any additional constraint it imposes.

A note on method before proceeding. The theory reviewed here was developed largely through case research on apparel, footwear, electronics, and automotive value chains. Those sectors remain the empirical backbone of GVC scholarship. But the argument of this book concerns industries with different factor intensities—labor-intensive furniture, technology-intensive electronics, capital-intensive cement and steel—and the theory must be stretched to accommodate them. Where the stretch reveals a weakness in the existing framework, I say so. The reader who wants the short version can read the final section alone; the reader who wants to know why the framework takes the shape it does should read all three.

2.1 Global Value Chain and Production Network Theory: Governance, Lead Firms and Value Chain Climbing

2.1.1 The commodity chain inheritance

The intellectual lineage begins with a simple observation: a product's journey from raw material to final consumer can be decomposed into discrete stages, each of which adds value and each of which may be located in a different country. The observation is old—it animates the eighteenth-century debate over the division of labor—but its systematic application to global production is a product of the late twentieth century. Gereffi's work on buyer-driven commodity chains in the 1990s gave the observation its first influential analytical vocabulary. The insight was that in many consumer goods sectors, the decisive coordination power did not reside with manufacturers but with retailers and brand owners. A US retailer specifying the design, volume, and delivery schedule of apparel produced in East Asia exercised a form of governance that no factory owner could match, because the retailer controlled access to the final market. Gereffi called this "buyer-driven" governance, and contrasted it with "producer-driven" chains in which manufacturers—typically in capital-intensive sectors such as automobiles—held the coordination power.

The distinction matters for the argument of this book because it predicts where green requirements will originate. In buyer-driven chains, the lead firm is a retailer or brand with direct exposure to consumer preferences and reputational risk. Environmental requirements in such chains tend to arrive as product specifications: recycled content, restricted substances, packaging reduction, carbon labeling. In producer-driven chains, the lead firm is a manufacturer with direct exposure to energy costs and regulatory compliance. Environmental requirements there tend to arrive as process standards: emissions limits, energy efficiency benchmarks, carbon market obligations. The Foshan furniture case in Chapter 5 sits squarely in the buyer-driven category. The cement and steel cases in Chapter 7 sit in the producer-driven category, though with a twist: Chinese producers in those sectors are increasingly lead firms in their own right, which changes the governance dynamic entirely.

Useful as it was, the commodity chain framework carried a weakness that later scholarship corrected. It treated governance as a binary—buyer-driven or producer-driven—and it treated the chain as a sequence of stages rather than a network of relationships. Both simplifications obscured the variation that matters most for upgrading. A factory making private-label apparel for a discount retailer and a factory making branded sportswear for a premium brand occupy different positions in the same "buyer-driven" category, but their upgrading prospects differ sharply. The first competes on price in a thin-margin relationship; the second may be drawn into design collaboration, quality upgrading, and—increasingly—joint environmental projects. Governance is a relationship-level attribute, not a sector-level one, and it varies across the firms within a chain.

2.1.2 Governance as a continuum: the Gereffi-Humphrey-Sturgeon typology

The corrective came in 2005, when Gereffi, Humphrey, and Sturgeon published a governance typology that remains the most widely used framework in the field. They identified five governance forms, arrayed along a continuum of explicit coordination and power asymmetry:

The typology's analytical power lies in its determinants. Gereffi and colleagues argued that governance form is determined by three variables: the complexity of transactions, the codifiability of information, and the capability of suppliers. When transactions are simple and information is easily codified, market governance suffices. When complexity rises but information remains codifiable, modular governance emerges—suppliers receive detailed specifications and produce to them, but the specification itself carries the coordination. When complexity is high and information resists codification, relational governance based on trust and repeated interaction becomes necessary. When supplier capability is low relative to the complexity of the task, the lead firm either internalizes the activity (hierarchy) or exercises direct control over a dependent supplier (captive).

Consider what this typology implies for green governance. Environmental requirements are, in the first instance, information: a specification for recycled content, a limit on volatile organic compounds, a reporting format for embodied carbon. Whether those requirements can be transmitted through modular governance depends on whether the information is codifiable. A restricted substances list is codifiable; a carbon accounting methodology that requires the supplier to allocate emissions across product lines, verify data against meter readings, and document assumptions is not. The first can be imposed through a market or modular relationship with minimal interaction. The second requires the kind of iterative, trust-dependent exchange that characterizes relational governance. This is the theoretical reason why carbon requirements tend to push supplier relationships toward the relational end of the continuum—and why firms that lack the capability to engage in relational exchange find themselves pushed toward the captive or market end, with correspondingly weaker upgrading prospects.

The typology has been criticized, fairly, for treating governance form as static. In practice, relationships move along the continuum as capabilities change. A supplier that begins in a captive relationship may, over time, develop enough competence to shift toward modular or relational governance. A lead firm that begins with arm's-length market transactions may move toward relational governance when quality problems or reputational risks emerge. The dynamic version of the typology is more useful than the static version, because upgrading is precisely a process of moving from one governance form to another—typically from captive or market toward modular or relational, and occasionally toward a position where the supplier itself exercises governance over its own sub-tier suppliers.

2.1.3 Lead firm power and the terms of participation

The governance typology describes the form of coordination. It does not fully describe the content of power. A lead firm's power over its suppliers rests on several distinct foundations, and the mix matters for how environmental requirements are received. Market access is the first: the lead firm controls the channel to the final consumer, and a supplier excluded from that channel loses revenue. Information is the second: the lead firm holds design specifications, demand forecasts, and consumer preference data that the supplier cannot easily obtain. Standards are the third: the lead firm sets the technical and, increasingly, environmental requirements that define acceptable participation. Finance is the fourth: the lead firm may control payment terms, credit, and access to capital that the supplier needs to invest in capacity.

Environmental requirements activate the standard-setting foundation in a way that interacts with the other three. When a lead firm introduces a carbon labeling requirement, it is exercising standard-setting power. But the requirement only functions if the supplier can generate the underlying data, which depends on information and finance. A supplier that lacks metering equipment, energy management systems, and trained personnel cannot respond to the requirement even if it wants to. The lead firm's standard-setting power thus depends on the supplier's capability, and the capability gap becomes a source of friction. This is the mechanism that Chapter 3 develops in detail: green governance capability is a condition that determines whether the supplier can participate in the relationship at all, and it originates as much in the supplier's own investment as in any directive from the lead firm.

The literature on lead firm power has tended to emphasize the coercive dimension—the lead firm as principal, the supplier as agent, the requirement as a directive. That emphasis is not wrong, but it is incomplete. Lead firms also transfer capabilities. The literature on supplier development, which grew out of the automotive sector, documents the mechanisms through which lead firms invest in supplier competence: technical assistance, joint problem-solving, personnel exchange, and preferential sourcing for suppliers that meet development targets. These mechanisms were originally designed for quality and delivery performance. They are now being repurposed for environmental performance, and the repurposing is uneven. Some lead firms have established supplier energy efficiency programs that provide audits, training, and co-investment. Others have simply added environmental criteria to their scorecards and left suppliers to figure out how to meet them. The difference matters for upgrading outcomes, and it is one of the variables that Chapter 6 examines in the technology-intensive context.

2.1.4 Production networks and the region as unit of analysis

The value chain framework has a spatial blind spot. It traces a product's journey across borders but treats the location of each stage as a point—a factory, a port, a warehouse—rather than as a place with institutions, infrastructure, and inter-firm relationships. The production network literature corrects this by treating the region as the unit of analysis and asking how regional assets—skills, supplier bases, research institutions, infrastructure—are incorporated into global networks. The key insight is that global networks do not simply allocate activities to locations with the lowest costs. They allocate activities to locations where the requisite assets exist, and those assets are often the product of long institutional development.

This matters for green governance because environmental capability is partly a regional asset. A cluster with a shared wastewater treatment facility, a common energy service company, and a local university with materials science expertise offers its firms a different green capability endowment than a cluster without those assets. The furniture clusters of Foshan and the electronics clusters of Shenzhen sit in the same province but draw on different regional endowments, and their green governance pathways diverge accordingly. Chapter 9 develops this argument in detail. For now, the theoretical point is that the value chain framework alone cannot explain why firms in similar chain positions upgrade differently. The regional asset base is part of the explanation.

The production network literature also introduces the concept of the "flagship" firm—the lead firm that orchestrates the network—and distinguishes between networks organized around a single flagship and those organized around multiple, competing flagships. The distinction has consequences for supplier bargaining power. A supplier that serves a single flagship is vulnerable to unilateral changes in requirements. A supplier that serves multiple flagships can play them against each other, and can use the requirements of one as leverage in negotiations with another. The Chinese manufacturers that have navigated the green transition most successfully tend to be those that have diversified their customer base and can therefore treat environmental requirements as a portfolio of demands rather than a single ultimatum.

2.1.5 What the GVC framework explains and what it does not

By this point the reader should have a clear picture of what GVC theory contributes. It explains the architecture of power: who coordinates, through what mechanisms, and on what terms. It explains why governance form varies across relationships and how that variation shapes the space available for supplier capability building. It explains why environmental requirements tend to push relationships toward the relational end of the continuum and why suppliers without relational capability find themselves confined to captive or market positions.

What it does not explain is why some suppliers respond to environmental requirements by upgrading and others respond by exiting or complying minimally. The governance framework predicts the constraints but not the response. To explain the response, we need a theory of upgrading—of how firms build capabilities, in what sequence, and under what conditions the sequence accelerates or stalls. That is the subject of the next section.

2.2 Industrial Upgrading: From Process to Chain Upgrading

2.2.1 The ladder and its limits

The most familiar image in the upgrading literature is the ladder: original equipment assembly (OEA), original equipment manufacturing (OEM), original design manufacturing (ODM), original brand manufacturing (OBM). The ladder describes a sequence in which a firm begins by assembling components to a customer's specification, acquires the capability to manufacture complete products to specification, acquires the capability to design products that the customer then brands, and finally acquires the capability to brand and market its own products. Each rung represents a shift in the locus of value capture, from labor to process engineering to design to brand equity.

The ladder is a useful heuristic and a misleading one. It is useful because it captures a real pattern: many East Asian firms did move through something like this sequence, and the sequence is not arbitrary. It is misleading because it implies a single track, a fixed order, and a destination. In practice, firms skip rungs, move sideways, occupy multiple rungs simultaneously, and sometimes move down. A firm may be an OBM in its domestic market and an OEM for export customers. A firm may develop design capability (ODM) without ever developing brand capability, and may find that position more profitable than OBM. The ladder also obscures the fact that different rungs require different capabilities, and the capability requirements are not simply cumulative. Branding requires marketing and distribution competence that has little to do with the manufacturing competence developed at earlier rungs. A firm that has mastered process engineering may find that its organizational routines are actively hostile to brand building.

Humphrey and Schmitz offered a more analytically precise taxonomy in the early 2000s, and it remains the standard. They distinguished four types of upgrading:

The taxonomy's virtue is that it separates dimensions that the ladder conflates. A firm can upgrade its process without changing its product or its function. It can upgrade its product without changing its function. Functional upgrading—the move from manufacturing to design, or from manufacturing to branding—is the dimension the ladder emphasizes, but it is only one of four, and it is not always the most valuable. Process upgrading in a capital-intensive industry can generate larger absolute returns than functional upgrading in a labor-intensive one, because the volume base is larger.

For the argument of this book, the most important distinction is between process and product upgrading on the one hand and functional upgrading on the other. Green governance capability, as Chapter 3 will argue, is primarily a process and product capability: it involves measuring and managing energy, materials, and emissions (process) and reformulating products to reduce environmental impact (product). But the accumulation of process and product capability can create the conditions for functional upgrading, because a firm that can document its environmental performance can enter new markets, serve new customers, and in some cases build its own brand on a sustainability platform. The relationship is not deterministic. Many firms accumulate green process capability without ever converting it into functional upgrading. But the conversion is possible, and the conditions under which it happens are a central concern of this book.

2.2.2 Capability building: from tacit to codified

The upgrading literature has always been concerned with capabilities, but it has not always been precise about what a capability is. The resource-based view of the firm, developed by Wernerfelt and others in the 1980s, defined capabilities as bundles of resources and routines that are valuable, rare, difficult to imitate, and non-substitutable. The definition is useful for explaining why some firms outperform others, but it is less useful for explaining how capabilities are built, because it treats them as stocks rather than flows.

The literature on technological capability building, developed largely through research on East Asian latecomers, offers a more dynamic account. The key distinction is between tacit and codified knowledge. Tacit knowledge resides in skills, routines, and organizational memory; it is acquired through doing and is difficult to transfer. Codified knowledge resides in documents, procedures, and standards; it can be transferred through instruction. Latecomer firms typically begin with tacit knowledge acquired through production experience—they learn by making things—and gradually codify what they have learned into procedures that can be replicated, audited, and improved. The codification process is what turns experience into capability.

Green governance capability fits this model closely. A factory that has never measured its energy consumption cannot manage it. The first step is metering—installing the equipment and the routines to record what is consumed, where, and when. The second step is analysis—identifying the drivers of consumption and the opportunities for reduction. The third step is codification—writing the procedures that embed the analysis into daily operations. The fourth step is verification—documenting the results in a form that external parties, whether customers or regulators or auditors, will accept. Each step converts tacit experience into codified knowledge, and each step depends on the one before it. A factory that tries to skip from metering to verification without the intervening analysis and codification will produce numbers that do not withstand scrutiny.

This is the capability gap that the Foshan factory in Chapter 1 exemplifies. The factory could produce a sofa; it could not produce a carbon number that a European buyer would accept. The gap was not a lack of effort or a lack of willingness. It was a missing sequence: the factory had never built the routines that convert production experience into verifiable environmental data. The gap is typical of labor-intensive firms in China's export clusters, and it is the reason why the compliance model—impose a standard, expect the supplier to meet it—produces disappointing results in those clusters. The standard assumes a capability that the supplier has not yet built.

2.2.3 The role of buyers in capability building

If capability building is a sequence, who pays for it? The upgrading literature has devoted considerable attention to this question, and the answer is not reassuring for suppliers. Capability building requires investment, and the returns to investment are uncertain and often appropriable by the buyer rather than the supplier. A supplier that invests in energy efficiency reduces its operating costs, which is a private return. But a supplier that invests in carbon accounting reduces its buyers' compliance costs, which is a return that accrues partly to the buyer. The supplier bears the cost; the buyer captures part of the benefit. This is a classic hold-up problem, and it explains why supplier investment in green capability often lags behind the level that would be socially optimal.

Several mechanisms can mitigate the hold-up problem, and the empirical record on each is mixed. Humphrey and Schmitz (2002), studying supplier development in the automotive sector, documented credible commitment through long-term contracts and co-investment: a buyer that signs a multi-year agreement or shares the capital cost of new equipment reduces the supplier's risk and makes investment more attractive. Kaplinsky (2000) analyzed rent sharing, where a buyer pays a premium for verified environmental performance and thereby gives the supplier a direct return on its investment. Gereffi, Humphrey, and Sturgeon (2005) emphasized capability transfer—technical assistance, training, and access to expertise that reduce the cost of capability building—and market access, where verified performance earns additional orders or preferred supplier status. The mechanisms are not equally effective. Credible commitment and rent sharing require the buyer to accept a share of the risk; capability transfer requires the buyer to invest in a relationship that a competitor might later poach. Many lead firms have chosen the cheaper route: add environmental criteria to the scorecard and let the supplier absorb the cost. The difference is visible in the environmental performance of their supplier bases, and it is one of the variables that distinguishes the upgrading trajectories examined in Chapters 5 through 7.

2.2.4 Upgrading in clusters: collective capability

The upgrading literature was originally firm-centric. It asked how an individual firm moves up the ladder. But many Chinese manufacturers are located in clusters—geographic concentrations of firms in the same or related industries—and the cluster context changes the upgrading dynamic. A firm in a cluster can draw on collective resources that an isolated firm cannot: shared suppliers, a local labor pool with industry-specific skills, informal knowledge exchange, and collective institutions such as trade associations and technology centers.

The cluster literature, which developed in parallel with the GVC literature and was brought into dialogue with it by Schmitz and others in the early 2000s, identifies several mechanisms through which clusters support upgrading. Knowledge spillovers are the most documented: firms in a cluster learn from each other, often informally, and the diffusion of new techniques is faster than it would be among dispersed firms. Collective efficiency follows from specialization—firms in a cluster can specialize in different stages of production and achieve economies of scale that no individual firm could achieve. Institutional support comes from collective bodies—training centers, testing laboratories, trade associations—that provide services individual firms could not afford.

For green governance, the cluster context is doubly important. On the one hand, clusters accelerate the diffusion of green practices: a firm that adopts a new energy management technique is visible to its neighbors, and the technique spreads. On the other hand, clusters can lock in unsustainable practices: a cluster that has invested heavily in a particular technology may resist the transition to a greener alternative, because the sunk costs are collective as well as individual. The furniture clusters of Foshan illustrate both dynamics. The diffusion of water-based coatings spread rapidly once a few firms demonstrated that the technique was viable. The transition away from solvent-based adhesives has been slower, because the adhesive supply chain is concentrated and the reformulation requires changes that individual firms cannot make alone.

The cluster perspective also complicates the governance story. In a cluster, the lead firm is not the only source of coordination. Local institutions—trade associations, local governments, technology centers—exercise governance over the cluster's firms, and their priorities may diverge from those of the global lead firms. A local government that prioritizes employment may resist environmental requirements that threaten jobs. A trade association that represents incumbent firms may resist standards that favor new entrants. The interaction between global chain governance and local cluster governance is one of the themes that Chapter 9 develops. The theoretical point is that upgrading unfolds in a specific institutional context rather than inside the firm alone, and the context shapes the trajectory.

2.2.5 What upgrading theory adds

The upgrading literature adds three things to the GVC framework. First, it specifies the dimensions along which firms can improve—process, product, function, chain—and thus gives precision to the vague notion of "moving up." Second, it specifies the capability-building sequence through which improvement happens, and thus explains why some firms improve and others do not. Third, it specifies the role of collective actors—buyers, clusters, institutions—in supporting or obstructing capability building.

What it does not yet explain is why the environmental dimension should be treated as a distinct driver of upgrading rather than as one more requirement among many. The upgrading literature has traditionally treated environmental requirements as a cost—something that firms must comply with, and that may or may not be offset by efficiency gains. The green governance literature, to which we now turn, challenges that treatment. It argues that environmental requirements can be a source of capability that firms would not otherwise develop, and that the capability can become a competitive advantage rather than a cost. Whether that argument holds, and under what conditions, is the subject of the next section.

2.3 Green Governance and Sustainability Transition Perspectives

2.3.1 From CSR to green governance

The corporate social responsibility tradition is older than the GVC literature and has its own intellectual genealogy. Its modern form dates to the 1980s, when Freeman's stakeholder theory reframed the firm as an entity accountable to employees, customers, suppliers, communities, and the environment alongside its shareholders. The reframing was normative in origin—it asked what the firm ought to do—but it acquired a strategic dimension as firms discovered that stakeholder management could reduce risk, improve reputation, and open new markets.

The CSR tradition's contribution to the argument of this book is the concept of the business case for environmental performance. The business case literature, which flourished in the 1990s and 2000s, identified several channels through which environmental performance could improve financial performance: cost reduction through energy and materials efficiency, revenue growth through access to environmentally sensitive markets, risk reduction through compliance and reputational management, and intangible benefits through employee recruitment and retention. The literature was empirically contested—the correlation between environmental and financial performance was positive but modest, and causality ran in both directions—but it established the principle that environmental performance could be a source of value rather than only a cost.

The CSR tradition's limitation, for our purposes, is that it treated environmental performance as a voluntary choice. The firm decided whether and how much to invest. The governance of that choice was internal: board committees, sustainability reports, codes of conduct. The external environment—regulators, customers, investors—provided incentives, but the incentives were mediated by managerial discretion. This treatment is inadequate for the current period, in which environmental performance is increasingly a condition of market access rather than a voluntary add-on. A Chinese manufacturer that cannot document its carbon footprint is not making a voluntary choice about whether to invest in sustainability; it is being excluded from a market. The shift from voluntary to mandatory—from CSR to green governance—is the defining feature of the current period, and it requires a theoretical framework that treats environmental requirements as structural rather than discretionary.

The concept of green governance, as used in this book, captures the shift. Governance, in the broad sense, refers to the mechanisms through which behavior is coordinated and constrained. Green governance refers to the mechanisms through which environmental behavior is coordinated and constrained in the value chain: the standards that lead firms impose on suppliers, the metrics that regulators require, the audits that third parties conduct, the reporting formats that investors demand. The mechanisms are diverse, but they share a common feature: they make environmental performance visible, measurable, and consequential. A firm that cannot produce the numbers cannot participate in the governance system, and a firm that cannot participate cannot access the markets that the system governs.

The shift from CSR to green governance has a specific institutional expression in China's dual carbon goals, announced in 2020. The goals—peak carbon by 2030, carbon neutrality by 2060—converted environmental performance from a voluntary corporate choice into a national policy commitment with binding implications for firms. The implications are not uniform across sectors: a cement producer faces direct emissions limits, a furniture exporter faces product-level carbon requirements from export markets, an electronics firm faces supply chain reporting obligations from its customers. But the direction is the same. Environmental performance is becoming a condition of operation, and the capability to measure and manage it is becoming a condition of competitiveness.

2.3.2 The multi-level perspective on socio-technical transitions

The CSR tradition explains why firms might choose to improve environmental performance. It does not explain how entire production systems—sectors, clusters, value chains—transition from one technological regime to another. For that, the sustainability transitions literature offers a framework, and the most influential version is the multi-level perspective (MLP) developed by Geels and colleagues.

The MLP distinguishes three levels of analysis. The niche level is where radical innovations emerge and develop protected from the selection pressures of the mainstream market. The regime level is the dominant socio-technical configuration: the technologies, institutions, business models, and user practices that constitute "how things are done" in a sector. The landscape level is the broader context—macroeconomic conditions, cultural values, political coalitions—that shapes the regime and the niches but is relatively slow to change.

The framework's central claim is that transitions occur through the interaction of these levels. Niches develop innovations that are initially uncompetitive. Landscape pressures—climate policy, resource scarcity, shifting consumer preferences—destabilize the regime. The destabilization creates windows of opportunity through which niche innovations can enter the mainstream. The regime does not collapse; it reconfigures, incorporating elements of the niche while preserving much of its existing structure.

For the green governance of Chinese manufacturing, the MLP offers several insights. Green technologies often begin as niche innovations—a new coating, a new kiln design, a new energy management system—that are uncompetitive at prevailing prices. Their development depends on protection: research funding, demonstration projects, early-adopter customers. Regime incumbents resist transitions, not because they are irrational but because their investments, competencies, and business models are adapted to the existing regime. A cement producer that has invested in a particular kiln technology will resist a technology that renders the investment obsolete. Landscape pressures—in this case, the dual carbon goals and the export market's carbon requirements—are the proximate cause of regime destabilization. Without those pressures, the niches would remain niches.

The MLP has been criticized for being too structural, for underplaying the role of agency, and for treating the regime as more coherent than it often is. Those criticisms are fair, and the framework should be used with them in mind. But the framework's core insight—that transitions involve the interaction of protected innovation, regime resistance, and landscape pressure—is essential for understanding why the green transition of Chinese manufacturing is uneven. Some sectors have moved rapidly because the landscape pressure was intense and the niche innovations were mature. Others have moved slowly because the regime was entrenched and the niches were immature. The variation is not random; it follows the MLP's logic.

2.3.3 Environmental upgrading: the debate

The concept of environmental upgrading sits at the intersection of the GVC and transitions literatures, and it is the most directly relevant concept for this book. Environmental upgrading refers to an improvement in the environmental performance of a firm's activities—reduced emissions, reduced resource use, reduced waste—that is achieved through its participation in a value chain. The concept is analogous to economic upgrading, but it measures a different dimension of improvement, and the relationship between the two is contested.

The debate has two positions. The optimistic position, associated with researchers who study lead firm-driven sustainability initiatives, argues that environmental requirements from buyers can drive upgrading. The mechanism is straightforward: the buyer sets a standard, the supplier invests to meet it, the investment generates efficiency gains and capability that the supplier would not otherwise have developed, and the supplier emerges more competitive. The pessimistic position, associated with researchers who study the political economy of value chains, argues that environmental requirements often amount to a cost transfer. The buyer captures the reputational benefit; the supplier bears the compliance cost; and the cost is passed on to workers through lower wages or to the environment through displaced production. The supplier may comply, but it does not upgrade.

Both positions have empirical support, which suggests that the relationship is conditional. The conditions, as the literature has gradually clarified, include the following. First, the nature of the requirement: requirements that are simple and codifiable can be met through compliance, while requirements that require learning and investment can trigger capability building. Second, the supplier's existing capability: a supplier with strong process engineering can absorb a new requirement and convert it into an improvement, while a supplier without that capability can only comply. Third, the buyer's commitment: a buyer that co-invests, shares rents, and provides technical assistance is more likely to generate upgrading than a buyer that simply audits and sanctions. Fourth, the market structure: a supplier with diversified customers has more bargaining power and can capture more of the return on its investment than a supplier dependent on a single buyer.

The debate's resolution, if there is one, is that environmental upgrading is possible but not automatic. It depends on the interaction of requirement characteristics, supplier capability, buyer commitment, and market structure. The framework this book develops is essentially a specification of those conditions for the Chinese manufacturing context. The chapters that follow examine how the conditions vary across factor intensities and how they shape the trajectory from compliance to capability.

2.3.4 The compliance-capability distinction

The most important conceptual distinction in this book is between compliance and capability. Compliance is the ability to meet a specified requirement. Capability is the ability to generate, evaluate, and improve environmental performance beyond the specified requirement. The distinction is not original—it appears in the CSR literature as the difference between "doing less harm" and "creating shared value," and in the innovation literature as the difference between adaptive and generative learning—but its application to green governance in value chains is the book's analytical core.

The distinction matters because compliance and capability have different determinants and different consequences. Compliance depends on the clarity of the requirement, the availability of the technology, and the supplier's ability to absorb the cost. Capability depends on the supplier's measurement infrastructure, management systems, technical expertise, and organizational routines. A supplier can comply without building capability: it can purchase a technology, follow a procedure, and pass an audit without understanding why the procedure works or how to improve it. A supplier that builds capability can do something that a compliant supplier cannot: it can anticipate requirements, identify opportunities, and improve performance without being told to.

The distinction also has a temporal dimension. Compliance is episodic: a supplier complies when a requirement is imposed and reverts to previous practice when the requirement is removed. Capability is cumulative: a supplier that has built measurement infrastructure can use it for multiple purposes, and each use improves the infrastructure. The cumulative nature of capability is what makes it a source of competitive advantage. A competitor can copy a technology, but it cannot copy the routines and knowledge that surround the technology without investing in the same learning process.

The compliance-capability distinction is the hinge on which the book's argument turns. The claim is that the Chinese manufacturers that will climb the value chain in the coming decade are not the ones that comply most efficiently with the current set of requirements. They are the ones that build the capability to generate and improve environmental performance, and that use the capability to enter new markets, serve new customers, and capture new rents. The claim is conditional—capability building requires investment, and the returns are uncertain—but it is the claim that the evidence in the following chapters supports.

2.3.5 Green governance as a value chain capability

The synthesis of the three literatures yields the book's central theoretical proposition: green governance is a value chain capability, not a regulatory constraint. The proposition has three components.

First, green governance is a capability in the technical sense: it is a bundle of routines, skills, and infrastructure that enables a firm to measure, manage, and improve its environmental performance. The capability is distinct from any particular technology or standard, because it is the ability to respond to new technologies and standards as they emerge. A firm with the capability can adapt; a firm without it can only comply.

Second, green governance is a value chain capability in the sense that it is exercised in the context of relationships with buyers, suppliers, and regulators. The capability is not exercised in isolation; it is exercised in response to the requirements and opportunities that the value chain presents. A firm that builds the capability without access to a value chain that rewards it will not capture the returns. A firm that has access to a demanding value chain but lacks the capability will not be able to respond.

Third, green governance is a source of value chain power in the sense that it changes the terms of participation. A firm that can document its environmental performance can negotiate with buyers from a position of strength, because it can demonstrate compliance with requirements that other suppliers cannot meet. A firm that can improve its performance can offer buyers something that other suppliers cannot offer: a path to the buyer's own environmental targets. The capability thus shifts the relationship from one of dependence to one of mutual dependence, and mutual dependence is the foundation of the relational governance that supports further upgrading.

The proposition is not a claim that green governance always leads to upgrading. It is a claim about the conditions under which it can. The conditions are the subject of the rest of the book. Chapter 3 develops the mechanism in detail, specifying how the capability is built and how it translates into value chain position. Chapter 4 examines the measurement and standards infrastructure that the capability depends on. Chapters 5 through 7 examine the differentiated pathways across factor intensities. Chapter 8 examines the specific impact of the dual carbon goals on export competitiveness. Chapter 9 examines the spatial and cluster dimensions. Chapter 10 draws the policy implications.

2.4 Assembling the Framework

The three literatures reviewed in this chapter each contribute a piece of the analytical framework. GVC theory contributes the governance architecture: the forms of coordination, the foundations of lead firm power, and the space that governance form leaves for supplier capability building. Upgrading theory contributes the capability sequence: the distinction between process, product, functional, and chain upgrading, the tacit-to-codified learning path, and the collective dimension of cluster-based capability. Green governance theory contributes the environmental dimension: the shift from voluntary CSR to structural green governance, the multi-level dynamics of socio-technical transitions, and the contested but conditionally positive relationship between environmental requirements and upgrading.

The framework that results can be stated compactly. A firm's position in a global value chain is determined by the interaction of three variables: the governance form of its relationships (which determines the requirements it faces and the support it receives), its capability endowment (which determines how it can respond), and the institutional context (which determines the resources and constraints that surround it). Green governance enters the framework as a capability that is partly determined by the governance form and the institutional context, and partly autonomous—a firm can invest in it, and the investment can change its position in the chain. The relationship is recursive: capability shapes position, and position shapes the requirements that drive further capability building. The recursion operates onlywhen the firm can convert each round of requirements into codified routines that survive the departure of the engineer who built them. Where that conversion fails—where the knowledge stays in one person's head, or where the metering equipment is installed but never read—the loop stalls at the first turn.

Consider what the framework implies for the Foshan factory. The factory occupies a captive or market position in a buyer-driven chain. Its capability endowment is thin: it can manufacture to specification but cannot measure or document its environmental performance. Its institutional context includes a cluster with some collective assets—a local testing center, a trade association—but the assets were built for quality and cost, not for carbon. Under the compliance model, the factory faces a requirement it cannot meet and either exits the export market or pays a consultant to produce a number that may not withstand audit. Under the capability model, the factory invests in metering, trains a manager, codifies procedures, and gradually builds the infrastructure to generate its own numbers. The second path is slower and more expensive in the short run. It is also the only path that leads anywhere.

The framework also implies a research agenda, which Chapter 10 develops. The agenda is not simply to document more cases of green upgrading. It is to specify the conditions under which the recursion runs in a virtuous direction: which governance forms, which capability endowments, and which institutional contexts generate the momentum that carries a firm from compliance to capability to position. The chapters that follow take up that agenda sector by sector, and they return at the end to the factory that could not produce a carbon number. The capability to produce the number is the point.

2.4.1 Three qualifications

The framework is deliberately modest, and three qualifications should be stated before the empirical chapters proceed.

The first concerns causality. The claim that green governance capability drives value chain climbing is a claim about a mechanism, not a claim about a universal law. The mechanism operates under specified conditions, and where the conditions are absent, the mechanism does not operate. The empirical chapters identify cases where the mechanism is visible and cases where it is not, and the variation is as informative as the central tendency. A theory that predicted upgrading everywhere would be a poor theory, because upgrading is not universal.

The second concerns measurement. The framework treats capability as a latent variable—something that is inferred from observable behavior rather than directly measured. The inference is necessarily imperfect, and the empirical chapters rely on proxies: the presence of metering infrastructure, the existence of environmental management systems, the proportion of revenue from customers with environmental requirements, the ability to pass third-party audits. Each proxy captures part of the capability and misses part. The reader should treat the measures as indicative rather than definitive, and should be alert to the possibility that the proxies are correlated with the capability for reasons other than the capability itself—firm size, for example, or export experience.

The third concerns generalization. The framework was developed for Chinese manufacturing firms in export-oriented value chains. It may apply to firms in other latecomer economies, and it may apply to firms in domestic value chains, but the application is not automatic. The Chinese context has specific features—the scale of the manufacturing base, the role of the state in setting environmental targets, the density of industrial clusters—that shape how the mechanism operates. The framework should be transported with those features in mind.

2.4.2 The road to Chapter 3

The reader who has followed the argument to this point should have a clear picture of the theoretical terrain. GVC theory explains the architecture of power and the terms of participation. Upgrading theory explains the capability sequence and the conditions under which firms move along it. Green governance theory explains why the environmental dimension is a potential source of capability rather than only a constraint. The synthesis—green governance as a value chain capability—is the book's central proposition.

Chapter 3 takes the proposition apart and puts it back together with more precision. It specifies the components of green governance capability—measurement, management, verification, and innovation—and traces how each component is built and how the components interact. It examines the mechanisms through which capability translates into value chain position: the reduction of information asymmetry, the shift in bargaining power, the access to new markets and customers, and the capture of rents that would otherwise accrue to the lead firm. And it examines the conditions under which the translation fails: the capability is built but not rewarded, the investment is made but not recouped, the position improves but not enough to escape the captive relationship.

The chapter also returns to a theme that this chapter has only touched on: the difference between green governance as a set of requirements imposed on a firm and green governance as a set of routines that the firm itself exercises. A firm that experiences green governance as an external imposition will comply and wait for the imposition to pass. A firm that exercises green governance as an internal routine will improve and look for the next opportunity. Moving a firm from the first stance to the second requires the measurement infrastructure, the trained personnel, and the codified procedures described above—and it requires a buyer or a cluster institution willing to reward the investment before the returns are visible. Chapter 3 examines how that move is made, and why it fails more often than it succeeds.