The Sourcing Story: Industry Rewrote Its Geography Without Rewriting Its Dependence

The shift away from China in apparel assembly was a calculated response to a risk environment that had changed materially since the mid-2010s. Geopolitical tension between the US and China, rising tariffs, and a wave of regulatory measures targeting Chinese supply chains gave brands concrete operational reasons to reduce their exposure. The "China + 1" strategy—sourcing from at least one additional country for every product line sourced from China—moved from boardroom discussion to procurement policy across most of the industry's major players.

The trigger that sharpened this shift most decisively was trade policy. In 2019, the US imposed tariffs of 15% on textile and apparel products from China, subsequently reduced to 7.5%. The introduction of the Uyghur Forced Labour Prevention Act in 2021 raised the stakes further, placing the burden of proof on US importers to demonstrate that goods were not produced with forced labour in Xinjiang. The EU followed with equivalent legislation in 2024, effective from 2027. Together, these measures put compliance costs—blocked shipments, import bans, the burden of proof on importers—squarely on the balance sheet.

Brands also had a second, more structural incentive to diversify: the demands of online retail. The rapid growth of e-commerce, combined with the aggressive expansion of ultra-fast fashion platforms, pushed brands toward sourcing models that could accommodate smaller batches, faster turnarounds, and greater flexibility. A "test and react" approach, in which initial micro-orders are placed in nearshored facilities before successful lines are scaled offshore, became increasingly standard. That model required a geographically dispersed assembly base, with production nodes close to the major consumer markets of Europe and North America.

The beneficiaries of this rebalancing were concentrated in a handful of countries. Bangladesh and Viet Nam absorbed the largest share of assembly displaced from China, with Cambodia also gaining ground. In the Americas, Central America—Honduras, El Salvador, Guatemala—drew investment, supported in part by US government commitments framing regional sourcing as part of a broader migration policy agenda. Between 2021 and 2024, that initiative was associated with USD 950 million in supplier investments and sourcing commitments from US buyers.

Chinese suppliers have played a central role in making this shift possible. Supported by China's outward investment policies since the early 2000s, Chinese firms expanded aggressively into alternative assembly locations—leading investment in Viet Nam, dominating Cambodia's garment sector, and establishing a growing presence in Bangladesh. By some estimates, approximately 65% of Viet Nam's garment production volume is generated by foreign firms, with Chinese investors at the front of that group. In Cambodia, 61% of member companies in the national garment manufacturers' association are Chinese-owned. The diversification of apparel assembly away from China has, to a considerable degree, been financed and operated by Chinese capital.

What made assembly the natural starting point for this diversification was its economics. An apparel factory requires a capital outlay of around USD 1 million—a threshold accessible to a wide range of suppliers, including the large transnational contractors from Hong Kong, South Korea, and Taiwan that had already built production portfolios across multiple countries. Low capital intensity meant low barriers to relocation. It also meant limited path dependency: unlike textile mills or fibre plants, assembly operations are not anchored by decades of fixed investment or cluster infrastructure. They can follow demand, and under sustained pressure from brands, they did.

The result, by 2022, was a meaningfully different map of global apparel assembly. China's share of global apparel exports had fallen from 43% in 2010 to 26%—a substantial shift, driven by deliberate lead firm strategy and enabled by a supplier base with both the capability and the incentive to expand into new locations. For brands managing geopolitical exposure and operational flexibility simultaneously, the diversification of assembly represented a genuine strategic achievement. What it did not represent—as the upstream data makes clear—was a restructuring of the supply chain as a whole.

Assembly may have moved, but the supply chain that makes assembly possible has not. While brands were diversifying their garment production across new geographies, the fabrics, fibres, and accessories feeding those factories remained overwhelmingly sourced from China. This is not a transitional condition. It reflects structural realities that have proved resistant to the same pressures that successfully dispersed assembly, and that show little sign of shifting.

The trade data is unambiguous on fabrics. China's share of global fabric exports rose from 13% in 2000 to 44% in 2022, even as apparel assembly was dispersing. In synthetic and blended fabrics—the dominant categories in modern garment production—China's share reached 49% and 45% respectively. Many of the countries that gained most from assembly diversification remain heavily dependent on Chinese textile imports to keep their factories running. By 2022, fabric and yarn imported from China accounted for more than 60% of total textile imports in five of the leading apparel export countries. Diversified assembly, in other words, is in large part assembled from Chinese inputs.

The economics of fabric production explain why this has been so difficult to change. Establishing a textile mill requires a minimum capital outlay of USD 50 million, and investments frequently exceed USD 200 million. That threshold is an order of magnitude higher than the cost of an apparel factory, and it fundamentally alters the supplier investment calculus. Where assembly relocation offered suppliers a manageable financial commitment backed by credible order volumes, upstream investment in fabric capacity carries risks that most suppliers—operating on thin margins and short-term contracts—are not positioned to absorb. The incentive structure simply does not support it.

Beyond the investment barrier, China's fabric and fibre ecosystem creates a form of industrial lock-in that is difficult to replicate elsewhere. Its dominance is not merely a function of scale, though scale matters. It is also a function of variety. China is the primary global source for mass-market synthetic fabrics, novelty blends, cotton-cellulose, and cotton-spandex materials—product categories for which no alternative supplier base offers comparable breadth. Lead firm fabric managers have noted that only China can reliably meet rapidly changing design demands, supplying novel yarn blends and fabric types at short notice. The UFLPA complicates this further: 85% of China's cotton comes from Xinjiang, and China accounts for 20% of global cotton supply—yet China's dominance in synthetic textiles, produced largely outside Xinjiang, remains unaffected by the legislation.

South Korean and Taiwanese mills fill specific niches in technical and performance fabrics, and Bangladesh and Viet Nam have developed some synthetic fabric capacity. But these are narrow alternatives, not systemic substitutes.

Fibre production follows the same pattern. Synthetic fibres—polyester, nylon, and related materials—account for around two-thirds of the global textile fibre market, and China dominates their production. In 2022, China held 17.5% of global synthetic fibre exports, with most domestic production feeding its own yarn-spinning industry. In accessories—trims, buttons, zippers, hangers, elastic—China's position is equally dominant, holding more than one-third of export market share across key categories. No other single country exceeds 10% in any accessory category.

State support, which helped drive assembly diversification in several locations, has largely failed to materialise for upstream segments. Viet Nam's EU free trade agreement created a meaningful incentive for textile investment through its fabric-forward rules of origin, and some capacity has developed as a result. But this remains the exception. Across Central America, North Africa, and South Asia, the targeted subsidies and policy frameworks needed to catalyse textile mill investment have not emerged at the scale required to challenge China's position. The supplier investment case is weak, path dependency is strong, and state action has not bridged the gap between them.

What the industry has built, through a decade of deliberate sourcing adjustment, is a supply chain that is more distributed without being any less dependent on China. The risk map looks different from 2015. But the industrial core—the fabric mills, fibre producers, and accessories clusters that give assembly its raw material—remains centred in China, and that gap between surface diversification and structural continuity is what the research identifies as the defining feature of the current moment.

Lead firms have, in effect, constructed a multi-tiered sourcing architecture that manages this tension rather than resolves it. Regional assembly hubs serve the speed and flexibility requirements of online retail—Türkiye and North Africa for Europe, Central America for the US. China continues to supply the domestic Chinese consumer market, which remains significant for brands including Nike, Adidas, and Fast Retailing, each of whom derived more than 15% of their revenues from China by 2019. And China supplies the inputs—the variety, the scale, the price points—that the rest of the sourcing network depends on. It hedges geopolitical exposure at the assembly layer while keeping Chinese industrial depth intact at every layer beneath it.

This architecture functions, but it does not constitute decoupling. The McKinsey Chief Procurement Officers surveys found that 70% of executives had considered moving parts of their supply chains closer to end markets—and most have not followed through. Yet nearshoring and verticality strategies, which would require co-locating fabric and fibre production with assembly, have shown limited materialisation. Between March 2020 and October 2024, industry tracking identified 34 announced investments in regional verticality and 23 in nearshoring, alongside just eight in onshoring—a modest number relative to the scale of the global supply base, and concentrated heavily in Central America and the USA. The gap between executive intention and investment reality points to the same structural constraints that have kept upstream production anchored in China.

The framing of this situation as "resilience" deserves scrutiny. Resilience, in supply chain terms, implies a capacity to absorb disruption and recover. What brands have built is better described as risk distribution within a system that remains structurally dependent. Assembly can flex—it can shift between countries, respond to tariff changes, adjust to regulatory pressure. But if the fabric supply is disrupted, or if Chinese accessories become inaccessible, the flexibility of assembly counts for little. The system's resilience is real at one layer and largely notional at the others.

The question this raises is one of durability. The current configuration was calibrated to the risk environment of the mid-2010s to mid-2020s—elevated geopolitical tension, yes, but not outright decoupling. Escalating trade measures, including the US government's move in early 2025 to end duty-free treatment for low-value shipments from China, suggest that the pressure on sourcing structures is continuing to intensify. The USFIA benchmarking data consistently shows that most sourcing officers want to reduce China dependence further—but wanting and executing are different problems when the upstream infrastructure has no credible alternative at scale.

What the research makes plain is that the geography of garment production is shifting at the edges while remaining stable at the core. That stability has served brands well as a short-term hedge. Whether it holds under sustained pressure, or simply defers a harder reckoning, will depend on how far the next round of trade disruption actually goes.

The Reckoning Still Ahead

The reorganisation of garment supply chains suggests movement, but not rupture. Assembly migrates while the system's industrial spine holds. Brands have managed the pressure intelligently, distributing risk at the layer that was easiest to move while preserving the efficiencies that China's upstream infrastructure provides. That calculation has held. As geopolitical pressure builds, the question is a practical one: how much more strain can a system carry when its industrial core was never really moved?