Chapter 2: Global Value Chain Embedding and the Double-Edged Sword of Upgrading
Chapter 2: Global Value Chain Embedding and the Two-Sided Logic of Upgrading
2.1 The Houjie Workshop Revisited
Houjie is a town of about 200,000 people in Dongguan, Guangdong, where the Pearl River Delta flattens into a maze of industrial parks. Drive north from Shenzhen along the Guangshen Expressway and you pass exit after exit labeled with the names of towns that between 1995 and 2015 produced a substantial share of the world's furniture, footwear, electronics and Christmas decorations. Houjie specialized in furniture. At its peak, the showrooms along the 107 National Road stretched for kilometers, and buyers from American and European retail chains arrived each spring and autumn to place orders for sofas, beds and office chairs that would be shipped from Yantian port six weeks later.
The workshop I described at the end of Chapter 1 sits in one of these industrial parks. It employs about 400 workers, down from 1,200 in 2012. Its owner, a man in his fifties who started as a carpentry apprentice in the 1980s, now spends more time on compliance documentation than on production. He showed me a binder—the third of four volumes—containing formaldehyde test reports, chain-of-custody certificates for timber, and social audit summaries that his main European customer requires each year. "Ten years ago," he said through an interpreter, "the buyer sent one page of requirements. Now it is a hundred pages, and every year they add something."
The workshop exists because it was embedded in a global value chain. Its growth from a village carpentry shop to an exporter of 40,000 sofas a year was made possible by a foreign buyer's order book, the buyer's design specifications, the buyer's quality standards and, eventually, the buyer's environmental and social requirements. That embedding gave the firm access to markets, capital and knowledge it could not have assembled on its own. It also left the firm dependent on a single customer, locked into a cost structure it did not control, and exposed to compliance demands that originated in regulatory processes it had no voice in shaping.
The tension is not a metaphor I chose for rhetorical convenience. It is the way the literature on GVCs and upgrading has described the relationship between chain participation and firm-level development for at least two decades. Gereffi's early work on buyer-driven commodity chains (1994) argued that US retailers and brand owners shaped overseas production networks in ways that created opportunities for suppliers in developing countries while simultaneously constraining their autonomy. Humphrey and Schmitz (2000, 2002) formalized the distinction between governance in clusters and governance in chains, showing that the same supplier could be upgraded by one buyer and squeezed by another. Kaplinsky (2000) and Kaplinsky and Morris (2001) documented how the distribution of gains from GVC participation varied enormously across firms and regions, with some suppliers capturing rents from process and product upgrading while others remained trapped in low-margin assembly.
That literature was built primarily around economic upgrading: moving from assembly to original equipment manufacturing (OEM), from OEM to original design manufacturing (ODM), from ODM to original brand manufacturing (OBM). Social and environmental upgrading were treated as secondary concerns, if they were treated at all. The evidence from the past decade—including the compliance binders in Houjie—suggests that this separation is no longer tenable. Environmental requirements are now embedded in the commercial terms of GVC participation, and the ability to meet them has become a condition of staying in the chain, not an optional add-on.
Three arguments organize this chapter. GVC embedding has been a powerful engine of economic upgrading for Chinese manufacturers, but the gains have been uneven and the upgrading trajectory has been shaped by the governance structures of the chains in which firms participate. Social and environmental upgrading do not automatically follow from economic upgrading; the relationship among the three is contingent on chain governance, buyer strategy and the institutional context in which suppliers operate. And the two-sided nature of GVC participation has become sharper under carbon constraints, because environmental compliance now carries costs that are not evenly distributed across the chain and because the knowledge required to meet those costs is not evenly accessible to suppliers at different positions in the value chain.
A note on method before proceeding. The evidence I draw on comes from three sources: the published GVC literature, primarily in the global political economy and economic geography traditions; the supply chain management literature on sustainable sourcing and supplier development; and my own field research in Chinese industrial clusters between 2015 and 2023, including interviews with roughly 60 firms in furniture, electronics, textiles and auto parts. The field research was not designed as a representative survey. It was designed to understand how firms at different positions in GVCs experienced the shift from cost-based to compliance-based competition. Where I cite a specific firm or figure, I name it. Where I generalize, I flag the limits of the evidence.
2.2 GVC Participation and Economic Upgrading: What the Evidence Shows
2.2.1 The upgrading lexicon
The GVC literature uses "upgrading" to describe the process by which firms, clusters and regions move from lower-value to higher-value activities within a chain or between chains. Humphrey and Schmitz (2002) distinguished four types: process upgrading (producing more efficiently), product upgrading (producing more sophisticated products), functional upgrading (moving to higher-value functions such as design or marketing), and inter-sectoral upgrading (moving into new sectors). This taxonomy has been widely adopted, and it is useful because it separates different dimensions of improvement that are often conflated in policy discourse.
The distinction matters for the argument of this book. Process upgrading—reducing cost, improving quality, meeting delivery schedules—is the form of upgrading most readily available to suppliers in buyer-driven chains. It is also the form most vulnerable to competitive erosion, because the capabilities required are relatively codified and can be replicated by suppliers in lower-cost locations. Product upgrading requires more investment and more absorptive capacity, but it is still typically undertaken within the parameters set by the buyer. Functional upgrading—moving into design, branding or retail—is the most valuable form, and the most contested, because it directly threatens the incumbent lead firms.
Gereffi, Humphrey and Sturgeon (2005) added a governance dimension to this taxonomy. They identified five governance types—market, modular, relational, captive and hierarchy—distinguished by the complexity of transactions, the codifiability of information and the capability of suppliers. In market governance, transactions are relatively simple and suppliers can switch buyers easily. In captive governance, suppliers are dependent on a small number of buyers and face high switching costs. The governance type shapes the upgrading trajectory: suppliers in modular chains have more room to develop independent capabilities than suppliers in captive chains, but they also face more pressure to compete on price.
The governance typology has been criticized for being static and for underplaying the role of supplier agency. Bair and Gereffi (2001), in their study of Torreon's blue jeans industry, showed that local clusters could upgrade even within captive governance structures if local institutions supported collective learning. Bazan and Navas-Alemán (2003), in their comparison of upgrading in global and national value chains in Brazil's Sinos Valley, found that suppliers who participated in both global and domestic chains had more upgrading opportunities than suppliers who participated only in global chains. Governance structure is a constraint, not a determinant. Supplier strategy and local institutional context mediate the relationship between chain position and upgrading outcomes.
For Chinese manufacturers, the governance typology describes a real distribution. Large electronics manufacturers in Shenzhen and Suzhou, such as Foxconn and its competitors, operate in modular or relational governance structures with global brand owners. They have substantial process and product upgrading capabilities, and some have moved into functional upgrading through vertical integration and platform strategies. Furniture exporters in Dongguan and Foshan, by contrast, typically operate in captive governance structures, dependent on a small number of US or EU buyers, with limited room for functional upgrading. The compliance binders in Houjie are a symptom of captive governance: the buyer sets the requirements, the supplier bears the cost of meeting them.
2.2.2 The mechanisms of upgrading
How does GVC participation drive upgrading? Learning by exporting is the most documented mechanism. Suppliers learn about product specifications, quality standards, delivery logistics and, increasingly, environmental and social compliance requirements from their buyers. This learning is often tacit and relationship-specific, and it accumulates over time. The evidence for learning by exporting is strongest for process and product upgrading. A furniture supplier that begins by producing simple bookcases for a US retailer may, over five years, learn to produce upholstered sofas with complex foam and fabric specifications. The learning is real, but it is bounded by the buyer's willingness to share knowledge and by the supplier's absorptive capacity.
Technology transfer, though less common, occurs when buyers provide equipment, software or technical assistance to suppliers, particularly when the supplier's existing capabilities are insufficient to meet the buyer's requirements. Nawrocka (2008) and Nawrocka, Brorson and Lindhqvist (2009) documented how ISO 14001 certification requirements in the electronics sector drove the adoption of environmental management systems among small suppliers in Asia. The transfer was not altruistic—the buyers needed compliant suppliers—but it had the effect of building environmental management capabilities that the suppliers could subsequently use for other purposes.
Market access is the most straightforward mechanism, and the one that policy discourse emphasizes most. GVC participation gives suppliers access to markets they could not reach on their own. But market access is a necessary condition for upgrading, not a sufficient one. Suppliers can have market access for decades without upgrading if the buyer's sourcing strategy does not require or reward it.
Competitive discipline exposes suppliers to competition from other suppliers in the same chain and from suppliers in other chains. This competition can drive efficiency improvements, but it can also drive a race to the bottom in which suppliers compete on cost rather than capability. The outcome depends on the buyer's sourcing strategy: buyers that compete on price alone will select for low-cost suppliers, while buyers that compete on quality or sustainability will select for capable suppliers.
The relative importance of these mechanisms varies across sectors and governance types. In labor-intensive sectors with captive governance, learning by exporting and competitive discipline dominate, and the upgrading trajectory is typically limited to process and product upgrading. In technology-intensive sectors with modular or relational governance, technology transfer and learning by exporting are more important, and functional upgrading is more feasible. In capital-intensive sectors, market access and competitive discipline dominate, and upgrading is often about scale and capital investment rather than capability development.
2.2.3 The evidence on Chinese upgrading
The aggregate evidence on Chinese GVC upgrading is mixed. China's share of global manufacturing value added rose from about 7 percent in 2000 to about 30 percent in 2020, and Chinese firms have moved up the value chain in many sectors. In electronics, Huawei, Xiaomi and BOE have become global leaders in their segments. In autos, Chinese firms have moved from assembly to design and, in the electric vehicle segment, to technology leadership. In furniture, the largest Chinese exporters have moved from OEM to ODM, and some have launched their own brands in domestic and export markets.
The upgrading has been uneven. Many Chinese suppliers remain in captive governance structures, dependent on foreign buyers and confined to low-margin assembly. The compliance binders in Houjie are a case in point: the workshop has upgraded its process capabilities substantially over two decades, but it has not upgraded its functional position in the chain. It still makes sofas to the buyer's design, sells them under the buyer's brand, and bears the compliance costs the buyer specifies.
The unevenness is not random. It correlates with firm size, sector, region and, crucially, with the governance structure of the chain. Large firms in modular or relational chains have upgraded more than small firms in captive chains. Firms in technology-intensive sectors have upgraded more than firms in labor-intensive sectors. Firms in the Pearl River Delta and Yangtze River Delta have upgraded more than firms in inland provinces, though the gap has narrowed as inland provinces have attracted relocated production.
The GVC literature offers a partial explanation for this pattern. Captive governance structures are characterized by high switching costs for suppliers and low switching costs for buyers. The supplier invests in relationship-specific assets—equipment, training, compliance systems—that have little value outside the relationship. The buyer can credibly threaten to switch suppliers, and the supplier cannot credibly threaten to switch buyers. The result is a bargaining asymmetry that limits the supplier's ability to capture the gains from upgrading. The supplier may become more efficient, but the buyer captures most of the efficiency gains through lower prices.
This is the "immiserizing growth" problem that Kaplinsky (2000) and others have documented in GVCs. Suppliers can upgrade their capabilities without capturing the returns, because the governance structure of the chain allows the lead firm to extract the rents. The problem is particularly acute in labor-intensive sectors where the barriers to entry are low and the number of potential suppliers is large.
2.2.4 When upgrading stalls: three cases
Three Chinese firms I interviewed between 2018 and 2021 illustrate different upgrading trajectories and different constraints.
Company A is a furniture manufacturer in Foshan, founded in 1998, employing about 800 workers, exporting about 60 percent of its output to the US and EU. Its main product is outdoor furniture—wicker chairs, tables, sun loungers—sold through large US retailers. The company has upgraded its process capabilities substantially: it has automated its weaving lines, adopted lean manufacturing, and reduced its defect rate from about 8 percent to under 2 percent. It has also upgraded its product capabilities: it now offers a wider range of designs and materials than it did a decade ago. But it has not upgraded its functional position. It still sells under the retailer's brand, at prices the retailer sets, and it has no direct access to end consumers. When I asked the owner why he had not launched his own brand, he laughed. "You need marketing, you need design, you need a warehouse in the US. I don't have those things. And if I try, the retailer will drop me."
Company B is an electronics manufacturer in Shenzhen, founded in 2005, employing about 3,000 workers, producing power supplies and adapters for global brand owners. It has upgraded from OEM to ODM: it now designs its own products, holds patents on several power management technologies, and sells to multiple brand owners. Its governance relationship with buyers is modular rather than captive: the buyers specify the interface standards, but the company controls the design and manufacturing of the internal components. The company has also begun to develop its own brand for the domestic market. The owner told me that the transition from OEM to ODM took about eight years and required substantial investment in R&D and engineering talent. "The buyers helped at first," he said. "They sent engineers to teach us. But after a while, we had to learn on our own."
Company C is a textile and garment manufacturer in Ningbo, founded in 1995, employing about 1,500 workers, producing knitwear for European and Japanese brands. It has upgraded its process capabilities—automated cutting, digital printing, just-in-time delivery—but its functional position has not changed. It remains a contract manufacturer, dependent on a small number of buyers. The owner told me that the company had tried to launch its own brand in Europe in the 2010s, but the attempt failed. "We didn't understand the market," she said. "We spent a lot of money on design and marketing, and we sold almost nothing. Now we focus on what we know."
These three cases are not representative, but they illustrate the range of upgrading outcomes within Chinese manufacturing. Company B upgraded functionally because it operated in a modular governance structure that gave it room to develop independent capabilities. Companies A and C did not, because they operated in captive governance structures that constrained their autonomy. The difference was not primarily about firm-level ambition or managerial quality. It was about the governance structure of the chain and the position of the firm within it.
2.2.5 The limits of the GVC upgrading framework
The GVC upgrading framework has been enormously productive, but it has limits that matter for the argument of this book.
It was built primarily around economic upgrading. Social and environmental upgrading were added later, often as an afterthought. The result is that the literature has a rich account of how firms move from assembly to design, but a thinner account of how they move from compliance to environmental capability. This book aims to fill that gap.
The framework also tends to treat governance structure as exogenous. In reality, governance structures evolve, and suppliers can influence them through collective action, investment in capabilities, and strategic choices about which chains to participate in. The literature on supplier agency (e.g., Bair and Gereffi, 2001; Bazan and Navas-Alemán, 2003) has begun to correct this, but the correction is incomplete.
A third limit is temporal. The framework was built in a period when environmental constraints were less binding than they are now. The carbon border adjustment mechanism (CBAM) in the EU, the strengthening of environmental disclosure requirements in major markets, and the proliferation of corporate net-zero commitments have changed the terms of GVC participation. Environmental compliance is no longer a marginal concern; it is a condition of market access. The GVC upgrading framework needs to be updated to account for this shift.
Finally, the framework has relatively little to say about the distribution of upgrading gains within supplier countries. The literature documents that some firms upgrade and others do not, but it does not explain why the gains from upgrading are distributed as they are, or what policy could do to broaden the distribution. This is a question that matters for the green transition, because the costs of environmental compliance are borne disproportionately by smaller, less capable suppliers.
2.3 Social and Environmental Upgrading: Do They Follow Economic Upgrading?
2.3.1 The optimistic view
The optimistic view is that social and environmental upgrading follow naturally from economic upgrading. As firms become more capable and more profitable, they can afford to invest in better working conditions and cleaner production. As they move up the value chain, they face more demanding buyers who require higher standards. The result is a virtuous circle: economic upgrading drives social and environmental upgrading, which in turn supports further economic upgrading.
This view has intuitive appeal, and there is some evidence for it. Firms that have upgraded their process capabilities often have better safety records and lower emissions than firms that have not, because the same investments in equipment and management systems that improve productivity also improve environmental performance. Firms that have upgraded their product capabilities often pay higher wages, because they need more skilled workers. And firms that have upgraded their functional position often have more autonomy to set their own environmental and social standards, because they are less dependent on a single buyer.
The optimistic view is also consistent with the stakeholder theory of the firm (Freeman, 1984; Evan and Freeman, 1998), which argues that firms that manage their relationships with stakeholders—employees, communities, regulators—will outperform firms that do not. The empirical literature on corporate social responsibility (CSR) and firm performance is mixed, but there is a substantial body of evidence that firms with strong environmental and social performance also have strong financial performance (see, e.g., the meta-analyses cited in Garriga and Melé, 2004).
2.3.2 The pessimistic view
The pessimistic view is that social and environmental upgrading do not follow automatically from economic upgrading, and may in some cases be undermined by it. The evidence for this view comes from several sources.
The GVC literature has documented cases where economic upgrading was accompanied by deteriorating working conditions. The classic example is the apparel sector, where suppliers that moved from simple assembly to more complex production sometimes did so by intensifying labor—longer hours, higher quotas, more repetitive tasks—rather than by improving wages or conditions. The "sweatshop" critique of GVCs is, in essence, a critique of the assumption that economic upgrading benefits workers.
The environmental literature has documented cases where economic upgrading increased environmental impact. Firms that move from simple production to more complex production often use more energy, more chemicals and more water. The environmental Kuznets curve—the hypothesis that environmental impact rises and then falls with income—has been widely tested and widely contested. Studies of the curve in China (for example, the panel analyses reviewed by Dinda, 2004, and the more recent provincial-level work by Hao et al., 2016) find that it holds for some pollutants in some contexts but not for others, and that the turning point is not automatic.
The supply chain management literature has documented the limits of buyer-driven social and environmental upgrading. Buyers can impose codes of conduct on their suppliers, but the codes are often weakly enforced, and the costs of compliance are often borne by the suppliers rather than the buyers. Roberts (2003) found that ethical sourcing initiatives had "patchy success" because buyers were unwilling to pay the costs of compliance and suppliers were unwilling to bear them. Pedersen and Andersen (2006) found that codes of conduct were often managed through a "compliance" rather than a "learning" approach, which limited their effectiveness. Lim and Philips (2008) found that the footwear industry's evolving governance structure had created new opportunities for some suppliers but had not improved conditions for workers at the bottom of the chain.
The literature on "greenwashing" has shown that firms can improve their environmental disclosure without improving their environmental performance. The gap between what firms say and what they do is a persistent feature of voluntary environmental governance, and it is particularly pronounced in GVCs where the buyer is far from the production site and the supplier has incentives to report favorably.
2.3.3 The contingent view
The contingent view, which I find most persuasive, is that the relationship between economic upgrading and social/environmental upgrading depends on the governance structure of the chain, the strategy of the lead firm, and the institutional context in which the supplier operates.
Governance structure matters because it shapes the bargaining power of the supplier. In captive governance structures, the supplier has little room to negotiate the terms of compliance, and the costs of compliance are typically passed down to the supplier. In modular or relational governance structures, the supplier has more room to negotiate, and the costs of compliance may be shared. The evidence from my field research supports this: Company B, the Shenzhen electronics manufacturer in a modular governance structure, was able to negotiate a higher price from its buyers to cover the cost of compliance with the EU's Restriction of Hazardous Substances (RoHS) directive. Company A, the Foshan furniture manufacturer in a captive governance structure, absorbed the cost itself.
Lead firm strategy matters because it shapes the demand for social and environmental performance. Buyers that compete on sustainability—Patagonia, Interface, IKEA—create demand for high environmental performance and are willing to pay for it. Buyers that compete on price create demand for low-cost compliance and are not willing to pay for it. The same supplier can have very different compliance experiences depending on which buyer it is supplying.
Institutional context matters because it shapes the enforcement of social and environmental standards. Suppliers in countries with strong environmental regulation and strong labor inspection face higher baseline standards and more credible enforcement than suppliers in countries with weak regulation. The result is that the same buyer code of conduct can have very different effects in different countries. China's environmental regulation has strengthened substantially since 2015, with the introduction of the "ecological civilization" framework, the revised Environmental Protection Law, and the central environmental inspection campaigns. This has raised the baseline for all suppliers, but the enforcement is still uneven across regions and sectors.
2.3.4 Empirical evidence from Chinese firms
The empirical evidence from Chinese firms is consistent with the contingent view. Several studies have found that GVC participation is associated with better environmental performance, but the association is mediated by the governance structure and the buyer's strategy.
Nawrocka (2008) studied small electronics companies in Asia and found that ISO 14001 adoption was driven primarily by buyer requirements, not by domestic regulation. The companies that adopted ISO 14001 improved their environmental management systems, but the improvements were often limited to documentation and procedures rather than substantive changes in production processes. Nawrocka, Brorson and Lindhqvist (2009) found similar results in a study of environmental supply chain practices: the buyers' requirements drove the adoption of environmental management systems, but the systems were often "ceremonial" rather than substantive.
My own field research suggests that the picture is more nuanced. The furniture manufacturers I interviewed in Dongguan and Foshan had made substantial environmental improvements since 2015, driven by a combination of buyer requirements and domestic regulation. The improvements included switching to water-based adhesives, installing dust collection systems, and adopting chain-of-custody certification for timber. These were not merely ceremonial: the changes in production processes were real, and they reduced emissions and worker exposure to hazardous chemicals. But the improvements were costly, and the costs were borne primarily by the suppliers. The buyers did not pay higher prices; they simply required compliance as a condition of continued business.
The distributional question is important. The environmental improvements in the furniture cluster benefited workers (through reduced exposure to chemicals and dust) and surrounding communities (through reduced air pollution). But they also benefited the buyers, who could claim credit for greening their supply chains without bearing the cost. And they disadvantaged the smaller suppliers, who could not afford the investments and were forced to exit the market. The consolidation of the cluster—from thousands of small workshops in the 2000s to a few hundred larger firms in the 2020s—was driven partly by environmental compliance costs.
2.3.5 The "compliance to capability" transition
The most interesting finding from my field research is that some firms have made a transition from compliance to capability. They began by meeting buyer requirements, but over time they developed environmental capabilities that went beyond compliance and became a source of competitive advantage.
The Houjie workshop is an example. When the European buyer first required formaldehyde testing in 2012, the owner treated it as a cost of doing business. He hired a consultant, installed a testing laboratory, and switched to low-formaldehyde adhesives. The cost was significant—about 3 percent of revenue—and the buyer did not pay more. But over time, the owner discovered that the low-formaldehyde adhesives were also easier to work with, and that the testing laboratory could be used to develop new products. By 2018, the workshop had developed a line of children's furniture that met the strictest European standards for chemical emissions, and it was selling this line to a different set of buyers at a higher margin. The compliance capability had become a product capability.
This transition is not automatic, and it is not universal. It depends on the firm's absorptive capacity, the nature of the environmental requirement, and the market opportunities available. But it is the mechanism through which environmental compliance can become a source of competitive advantage rather than a cost. The rest of this book is, in a sense, an exploration of the conditions under which this transition occurs.
Van Weele and Vivanco (2014) describe a similar transition in their work on CSR in value chain relationships. They argue that firms move from "compliance" to "value creation" when they treat CSR requirements as opportunities to innovate rather than as costs to minimize. The key mechanism is what they call "solution space reduction": when a buyer imposes a CSR requirement, it reduces the supplier's product and process solution space, which forces the supplier to innovate. The innovation can be defensive (finding a cheaper way to comply) or offensive (finding a new product or process that turns the requirement into an advantage). The offensive path is more likely when the supplier has strong absorptive capacity and when the buyer is willing to share the gains from innovation.
Kibbeling (2010), in a dissertation on value creation in supply chains, makes a related argument. She shows that CSR requirements can drive innovation in supply chain relationships because they force suppliers to develop new products and processes that they would not have developed otherwise. The innovation is not guaranteed—some suppliers respond to CSR requirements by exiting the market or by shirking—but it is a real possibility when the conditions are right.
2.3.6 What the book does not argue
I should be clear about what I am not arguing in this section. I am not arguing that GVC participation automatically leads to social and environmental upgrading. The evidence does not support that claim. I am not arguing that buyers are the driving force behind environmental improvement in Chinese manufacturing. Domestic regulation has been at least as important, particularly since 2015. And I am not arguing that the compliance-to-capability transition is easy or common. It is difficult and relatively rare, and the firms that achieve it are not representative of the broader population.
What I am arguing is that the transition is possible, that it is more likely under certain conditions (modular or relational governance, buyers that compete on sustainability, strong firm-level absorptive capacity), and that understanding those conditions is important for policy and for firm strategy. The rest of the book develops this argument in detail, sector by sector.
2.4 The Risks That Accompany the Benefits
2.4.1 The benefits of GVC participation
The benefits of GVC participation are well documented and should not be understated. GVC embedding gave Chinese manufacturers access to markets, technology, capital and networks. The furniture cluster in Dongguan, for example, grew from a few dozen workshops in the early 1990s to thousands of firms by the mid-2000s, almost entirely on the back of export orders from US and EU retailers. Access to those markets allowed Chinese firms to achieve scale that would have been impossible in the domestic market alone. Buyers transferred design specifications, quality standards, production techniques and, more recently, environmental and social compliance requirements. The transfer was not free—suppliers paid for it through lower prices and higher compliance costs—but it was real. GVC participation also made it easier for Chinese firms to access trade finance, foreign direct investment and, in some cases, equity investment from global buyers, and it connected them to global networks of suppliers, logistics providers, certification bodies and industry associations that provided information, contacts and legitimacy difficult to assemble independently.
The result has been a substantial improvement in the capabilities of Chinese manufacturers. The firms I interviewed in 2020 were, on average, much more capable than the firms I interviewed in 2010. They had better equipment, better management systems, better environmental and social compliance, and better access to global markets. The improvement was not uniform, but it was real.
2.4.2 The risks of GVC participation
The risks are equally real, and they have become more salient as the green transition has accelerated.
Dependency is the most obvious risk. Suppliers in captive governance structures are dependent on a small number of buyers, and the dependency is self-reinforcing: the more the supplier invests in relationship-specific assets, the more costly it is to switch buyers. The dependency gives the buyer bargaining power, which the buyer uses to extract price concessions and to shift compliance costs to the supplier. The supplier is caught in a trap: it cannot afford to lose the buyer, and it cannot afford to keep the buyer.
Lock-in follows from the same logic but operates through a different channel. Suppliers that have invested heavily in a particular product or process may find it difficult to switch when the market changes. The lock-in is partly physical (equipment, infrastructure) and partly cognitive (managerial routines, workforce skills). The furniture manufacturers in Dongguan that invested in solid wood production in the 2000s found it difficult to switch to panel-based production when the market shifted in the 2010s. The ones that made the switch survived; the ones that did not, exited.
Cost pass-through is the third risk, and it is the one that shows up most directly in margins. Compliance costs are typically borne by the supplier, not the buyer. The supplier may be able to pass some of the cost on to workers (through lower wages) or to the environment (through externalities), but the scope for pass-through is limited. The result is a squeeze on margins, which is particularly severe for smaller suppliers. The consolidation of the furniture cluster in Dongguan is partly a result of this squeeze: the smaller suppliers could not afford the compliance costs and exited the market.
Regulatory exposure is a fourth risk, and it has grown as exporting destinations have added environmental and labor provisions to their trade regimes. Suppliers are exposed to regulatory changes in their export markets, over which they have no influence. The EU's CBAM, the EU's deforestation regulation, the US's Uyghur Forced Labor Prevention Act, and the proliferation of product-specific environmental standards have all created new compliance requirements for Chinese exporters. The requirements are often complex, and the timelines are often short. Suppliers that cannot adapt quickly lose market access.
Reputational risk is the fifth, and it is the least controllable. Suppliers are exposed to reputational risks from their buyers' activities and from the activities of other suppliers in the same chain. If a buyer is accused of labor abuses in its supply chain, all of its suppliers may face scrutiny, even if they are not implicated. The reputational risk is particularly acute for suppliers in labor-intensive sectors, where the risk of labor abuses is higher and the scrutiny is more intense.
2.4.3 The green transition sharpens the risks
The green transition sharpens each of these risks, for three reasons.
Environmental compliance is more costly than previous forms of compliance. The cost of meeting formaldehyde standards in furniture is relatively modest—about 3 percent of revenue in the Houjie case. The cost of decarbonizing production is much larger. A furniture manufacturer that switches from coal-fired boilers to natural gas or electric heating faces a capital cost of several million yuan and an operating cost increase of 20 to 40 percent. A textile manufacturer that switches from conventional dyeing to waterless dyeing faces a capital cost of tens of millions of yuan. These costs are difficult to absorb, particularly for smaller suppliers.
Environmental compliance also requires capabilities that many suppliers do not have. Meeting a formaldehyde standard requires a testing laboratory and a supplier of low-formaldehyde adhesives. Decarbonizing production requires energy management, carbon accounting, and access to renewable energy. These capabilities are not widely available in Chinese industrial clusters, and the firms that have them are disproportionately large and well-resourced. The result is that the green transition is likely to accelerate the consolidation of clusters, with smaller and less capable firms exiting the market.
Uncertainty compounds the other two. The regulatory landscape is evolving rapidly, and the requirements are often unclear or contested. The EU's CBAM, for example, is being phased in over a period of years, and the detailed rules for calculating embedded emissions are still being developed. Suppliers that invest in compliance may find that the rules change before they can recover their investment. The uncertainty is a particular burden for smaller firms, which have less capacity to absorb regulatory risk.
2.4.4 Case evidence: the furniture cluster under carbon constraints
The furniture cluster in Dongguan provides a concrete illustration of how the green transition sharpens these risks. I conducted fieldwork in the cluster in 2019 and 2021, interviewing 25 firms and several industry association representatives.
The cluster's environmental challenges are substantial. Furniture manufacturing involves several energy-intensive processes: drying of timber, curing of finishes, and heating of adhesives. The cluster's firms historically relied on coal-fired boilers, which were a major source of air pollution. The Dongguan municipal government began phasing out coal-fired boilers in 2015, requiring firms to switch to natural gas or electric heating. The switch was costly: a typical firm faced a capital cost of 2 to 5 million yuan and an operating cost increase of 25 to 40 percent.
The firms' responses varied. The larger firms, with more than 500 employees, generally complied and absorbed the cost. Some of them used the switch as an opportunity to upgrade their production processes, investing in more efficient drying equipment and in heat recovery systems. The smaller firms, with fewer than 200 employees, often could not afford the switch. Some of them relocated to inland provinces where the enforcement was less strict. Others exited the market. The cluster's firm count fell from about 3,000 in 2015 to about 1,800 in 2021.
The buyers' responses were mixed. Some buyers—particularly European buyers with strong sustainability commitments—supported the switch by paying slightly higher prices and by providing technical assistance. Others did not. A manager at a mid-sized firm told me: "Our main customer is a US retailer. When we told them about the boiler switch, they said, 'That's your problem, not ours.' We asked for a 5 percent price increase to cover the cost, and they said no. We had to absorb it."
The distributional outcome is clear: the costs of the green transition in the furniture cluster were borne primarily by the suppliers, and within the supplier population, primarily by the smaller firms. The buyers captured the reputational benefits of a greener supply chain without bearing the costs. The workers and communities captured the environmental benefits (cleaner air, lower emissions) but also bore the costs of job losses as smaller firms exited.
This is the two-sided logic in its sharpest form. GVC participation gave the cluster access to markets and knowledge that drove decades of growth. It also exposed the cluster to compliance requirements that it could not meet without significant cost, and it distributed the costs in a way that accelerated consolidation and dislocation.
2.4.5 Conditions for managing the risks
The risks are not inevitable. There are conditions under which GVC participation can drive upgrading without exposing suppliers to unacceptable costs. I identify four.
Governance structure matters most. Suppliers in modular or relational governance structures have more room to negotiate the terms of compliance and to capture some of the gains from upgrading. The Shenzhen electronics manufacturer (Company B) was able to negotiate a higher price to cover RoHS compliance because its buyers needed its design capabilities and could not easily switch to another supplier. Suppliers in captive governance structures, like the Foshan furniture manufacturer, have less room to negotiate. The policy implication is that suppliers should seek to move from captive to modular or relational governance where possible, and that buyers and governments should support this transition.
Buyer strategy is equally important. Buyers that compete on sustainability are more likely to share the costs of compliance and to provide technical assistance. The European buyers in the furniture cluster who supported the boiler switch are an example. Buyers that compete on price are more likely to shift costs to suppliers. The policy implication is that suppliers should diversify their buyer base and seek out buyers with strong sustainability commitments. This is easier said than done, particularly for small suppliers, but it is a real strategic option.
Firm-level absorptive capacity determines whether a compliance requirement becomes an innovation opportunity or a pure cost. The Houjie workshop's transition from compliance to capability was possible because the owner had the technical knowledge to recognize that low-formaldehyde adhesives could be used to develop a new product line. Firms without this capacity are more likely to experience compliance as a pure cost. The policy implication is that investment in firm-level capabilities—through training, technical assistance andaccess to finance—can increase the likelihood that compliance drives upgrading.
Institutional support is the fourth condition, and it is the one most amenable to policy. Governments can reduce the cost of compliance by providing subsidies, technical assistance and infrastructure. The Dongguan municipal government's support for the boiler switch—which included a subsidy of up to 30 percent of the capital cost—reduced the burden on firms. Industry associations can provide collective services—testing laboratories, training programs, advocacy—that individual firms cannot afford. The policy implication is that public and collective institutions have a role to play in managing the risks of GVC participation.
2.4.6 The limits of the GVC framework under carbon constraints
The GVC framework was built to explain the organization of production in a world where environmental constraints were relatively weak. It needs to be updated for a world where carbon constraints are binding.
The most important update is the recognition that carbon is a factor of production, and that its price varies across jurisdictions. The EU's emissions trading system (ETS) prices carbon at around 80 to 100 euros per tonne (as of 2024), while China's national ETS prices it at around 60 to 80 yuan per tonne (about 8 to 11 euros). The US has no national carbon price. This variation creates arbitrage opportunities: production can relocate to jurisdictions with lower carbon prices, and carbon-intensive goods can be imported from those jurisdictions. The EU's CBAM is designed to close this arbitrage by imposing a carbon price on imports equivalent to the domestic price.
The implication for GVCs is significant. In a world with carbon border adjustments, the location of production is no longer determined solely by labor costs, logistics costs and agglomeration economies. It is also determined by the carbon intensity of production and the carbon price in the exporting jurisdiction. Chinese manufacturers that produce with coal-fired electricity face a carbon cost disadvantage in EU markets that they did not face before. The disadvantage is particularly severe for carbon-intensive products—steel, aluminum, cement, fertilizer—but it also affects downstream products that embed carbon, such as furniture and electronics.
The GVC framework also needs to account for the role of scope 3 emissions. Large multinational buyers are increasingly required to report and reduce their scope 3 emissions—the emissions embedded in their supply chains. For many buyers, scope 3 emissions are the largest part of their carbon footprint, and they are under pressure from investors and regulators to reduce them. The result is that buyers are imposing carbon requirements on their suppliers, often without providing the resources to meet them. The requirements are a new form of compliance demand, and they are likely to become more stringent over time.
The framework also needs to account for the emergence of green industrial policy in major markets. The US Inflation Reduction Act, the EU's Green Deal Industrial Plan, and Japan's Green Transformation program all provide subsidies and incentives for domestic green production. These policies create a competitive disadvantage for Chinese exporters, who do not have access to the same subsidies. The disadvantage is particularly acute in sectors such as electric vehicles, batteries and solar panels, where the subsidies are large and the competition is intense.
These changes do not invalidate the GVC framework, but they require it to be extended. The framework's core insight—that GVC participation shapes upgrading trajectories through governance structures and buyer strategies—remains valid. What needs to be added is an account of how carbon constraints reshape governance structures and buyer strategies, and how they distribute the costs and benefits of upgrading.
2.4.7 Should Chinese manufacturers withdraw from GVCs?
If GVC participation carries these risks, should Chinese manufacturers reduce their participation? Should they turn inward, focus on the domestic market, and reduce their dependence on foreign buyers?
Withdrawal is not the answer. The domestic market cannot absorb the export volume in many sectors. The furniture cluster in Dongguan exports about 60 percent of its output; the domestic market cannot absorb that volume at current prices. Moreover, the capabilities that firms have built through GVC participation—design, quality management, environmental compliance—are valuable in the domestic market as well. Firms that withdraw from export markets may lose those capabilities over time. And the green transition is global, not national. Chinese firms that want to compete in the global market will have to meet global environmental standards, whether they participate in GVCs or not. Withdrawing from GVCs does not avoid the standards; it only avoids the support that GVC participation can provide.
The better response is not to reduce GVC participation but to change its terms. Suppliers can diversify their buyer base, move from captive to modular governance, invest in capabilities that reduce their dependence on any single buyer, and organize collectively to negotiate better terms. Governments can support these strategies through industrial policy, trade policy and institutional development. Buyers can be pressured—by regulators, investors and consumers—to share the costs of compliance and to support supplier upgrading. None of these responses is easy, and none is guaranteed to succeed. But they are more promising than withdrawal.
2.5 Conclusion: From Two-Sided Logic to Deliberate Upgrading
The Houjie workshop, where this chapter began, is a small example of a large phenomenon. The workshop's owner built a business by embedding himself in a global value chain. He learned from his buyers, invested in compliance, and gradually developed capabilities he did not have before. The compliance binder that he showed me is a record of that learning, and also a record of its limits. The workshop has upgraded its process and product capabilities, but it has not upgraded its functional position. It remains dependent on a small number of buyers, and it bears the costs of compliance without capturing the gains.
The evidence reviewed in this chapter suggests that this pattern is common but not universal. GVC participation drives economic upgrading through learning, technology transfer, market access and competitive discipline. It drives social and environmental upgrading when governance structures give suppliers room to negotiate, when buyers compete on sustainability, when firms have strong absorptive capacity and when institutions provide support. When these conditions are absent, GVC participation can trap suppliers in low-margin, high-compliance positions that limit their ability to capture the gains from upgrading.
The green transition sharpens the two-sided logic. Environmental compliance is more costly, more capability-intensive and more uncertain than previous forms of compliance. The firms that can afford it will consolidate their positions; the firms that cannot will exit or relocate. The distributional consequences—for firms, workers, communities and regions—will be significant, and they will not be evenly distributed.
The argument of this book is that these consequences are not inevitable. Firms can build green governance capability deliberately, and the firms that build it early can turn compliance into competitive advantage. The Houjie workshop's transition from compliance to capability—from formaldehyde testing to a children's furniture line—is a small example. The rest of the book examines how this transition can be scaled up, sector by sector, and what policies and strategies can support it.
The evidence I have reviewed is drawn primarily from firm-level case studies, industry association reports and the published GVC literature. It is not a representative survey, and it does not support strong causal claims. The claim I am making is modest: that the conditions under which GVC participation drives upgrading are identifiable, that they vary systematically across sectors and governance structures, and that they can be influenced by firm strategy and public policy. The chapters that follow test this claim against the evidence from three sectors—furniture, electronics and auto parts—and against the emerging evidence on the dual carbon goals. The test is not definitive, but it is the best we can do with the evidence available.
The workshop in Houjie is still running. On the morning I left, the owner was walking a new European buyer through the testing laboratory he had installed in 2012, pointing out the formaldehyde chamber and the chain-of-custody files. The buyer was taking notes. The workshop had not escaped the terms of GVC participation. It had learned to work within them, and in a few product lines, to turn them to its advantage. That is the narrow path this book tries to map.