A Dependency Investigation (#07) – May, 2026
About This Series
The Magician’s Hands is a series of dependency investigations. Each report examines a single case in which a structural dependency, between a state and an infrastructure owner, a farmer and a seed company, a continent and an energy supplier, was created, normalised, leveraged, and converted into power. The cases span domains and decades. The grammar beneath them does not change.
The series takes its name from a simple observation: the most consequential things happening in the world are rarely the things that take centre stage. While we watch the visible hands, something else is being built in the structural layer underneath. These reports are an attempt to make that layer legible.
The founding article, “”, sets out the full grammar. Each investigation that follows applies it to a specific case.
Scene-setter
The public story of the European textile industry’s decline is a story about consumers. It is told in the language of appetite: cheap clothing, disposable fashion, the algorithmically optimised desire machine of Shein adding thousands of new items daily at prices below the cost of a meal. The debate that surrounds it is conducted in the registers of ethics and environment, of labour rights abuses in Guangzhou workshops, of polyester microfibres accumulating in ocean sediment, of a generation of young consumers who have been taught to treat a garment as a single-use object. The hands are visible, dramatic, and morally legible. They are also, structurally, beside the point.
What this investigation is looking at is the layer beneath the consumer story: a productive infrastructure transferred, decision by rational decision, from Europe to Asia across four decades, leaving European fashion brands in possession of their names, their design teams, and their retail presence, while surrendering the manufacturing knowledge, the supply chain control, and the production speed that would allow them to compete on any terms other than the ones now set by the system they built and then abandoned. The consumer preference argument, the claim that this is what the market demanded, is not an explanation. It is a rationalisation assembled after the structural outcome had already been determined, and it belongs in the obscuring layer, not in the causal one.
What the five moves will show is this: a manufacturing dependency assembled by European capital through four decades of individually rational sourcing decisions, normalised through the language of free trade, comparative advantage, and consumer benefit, leveraged by a production system that now sets the pace and the price floor the European industry cannot meet, obscured by a moral discourse that keeps the structural question permanently off the table, and converted into a relationship in which the knowledge, the speed, and increasingly the infrastructure itself are owned by the same system that European brands once hired as a subcontractor.
Move 1: Creating the Dependency
The entry point here is creation, not capture. No external actor seized the European textile industry’s productive capacity under conditions of crisis. European capital built the dependency itself, one sourcing decision at a time, across a period long enough that the structural consequence of the aggregate was invisible at the level of any individual decision.
The process began in earnest in the 1970s, when European textile and garment manufacturers, facing rising domestic labour costs and intensifying price competition, began relocating production to Asia. The logic was straightforward and, within its own frame, correct: labour in Hong Kong, Taiwan, South Korea, and subsequently China and Bangladesh was cheaper, the savings were real, and the margin improvement was immediate. Trading companies established offices in Hong Kong to manage the sourcing relationship. As the model proved commercially sound, brands moved to eliminate the intermediary and establish their own buying offices, capturing more of the margin and deepening their operational integration with Asian producers. Each step made sense. None of them was evaluated against the question of what the aggregate, across an entire industry, across three decades, would produce.
The Multi-Fibre Arrangement, signed in 1974 and governing textile trade until its phaseout was completed on 1 January 2005, imposed bilateral quotas on exports from developing countries to major importing markets including the European Community. It is tempting to read the MFA as a structural brake on the dependency, and in a narrow sense it was: it prevented any single country from dominating European import volumes and distributed sourcing across a wider range of low-cost producers, from Morocco and Tunisia in the near-shore to Bangladesh, Sri Lanka, and India further afield. But the MFA did not address the underlying dynamic. It spread the sourcing geography while leaving the core logic intact: European brands were systematically transferring their production requirements to low-wage environments, and the productive infrastructure, the factories, the trained workforces, the fabric mills, the dyehouses and finishing plants, was accumulating in the receiving countries while it was allowed to atrophy in the sending ones.
The moment of irreversibility arrived not at a single decision point but through the compounding of structural facts. European textile manufacturing capacity declined as investment followed the sourcing relationships eastward. The skilled workforce, the pattern-cutters, dyers, finishers, and quality controllers whose tacit knowledge takes a generation to build and cannot be reconstructed from a manual, aged and retired without being replaced. Vocational training systems that had sustained the industry contracted as the industry itself contracted. By the time the MFA phaseout removed the last institutional brake, the question of whether to source in Asia had effectively ceased to be a choice for most European fashion businesses. The alternative had been quietly dismantled. When quotas were abolished on 1 January 2005, Chinese textile and clothing exports to the EU nearly doubled in value in the first three quarters of the year. The structural outcome that four decades of individual decisions had produced arrived in a matter of months. What OAPEC found in 1973 when it looked at Western energy infrastructure, a dependency already built, already load-bearing, already undefended, Asian manufacturing found in 2005 when it looked at the European fashion industry. The dependency was not created by the phaseout. The phaseout simply removed the last thing standing between the dependency and its full expression.
Move 2: Normalising the Dependency
The normalisation of the European fashion industry’s dependency on Asian manufacturing operated through two mechanisms running simultaneously, and their interaction is what made the structural outcome so completely invisible for so long.
The narrative normalisation was carried by a cluster of stories that reinforced each other without requiring coordination. Free trade was the primary register: the dismantling of the MFA was presented as a development gain, a long-overdue correction that would allow low-income countries to benefit from their comparative advantage in labour-intensive production. This was not entirely false. Textile manufacturing employment in Bangladesh, Cambodia, and Sri Lanka did expand significantly as European sourcing moved through those markets. But the development framing successfully displaced the structural question of what Europe was surrendering in exchange for the efficiency gains. Comparative advantage supplied the economic orthodoxy: Europe should concentrate on design, branding, and retail while production migrated to where it could be done most cheaply. This argument was presented as settled science rather than as a policy choice with distributional and strategic consequences. The efficiency gains were real and immediate; the structural costs were diffuse and deferred. Consumer benefit was the third register: cheaper clothing was an unambiguous good for household budgets, particularly for lower-income European households. This argument was irrefutable on its own terms and effectively closed the political space for any counter-argument that might have sounded like defending expensive domestic production at the consumer’s expense. By the time the consumer preference argument was added, the normalisation architecture was complete. What had begun as a sourcing strategy had been converted into a description of the natural order.
The operational normalisation required no narrative. It worked through use, through the daily accumulation of structural fact. As Asian sourcing became dominant, the comparative infrastructure of European textile production deteriorated: machinery aged without replacement, factories closed, suppliers of specialist inputs disappeared, and the ecosystem that sustained domestic production shrank below the threshold at which it could regenerate itself. European fashion businesses that might have retained some domestic or near-shore production found that the minimum order quantities, the lead times, and above all the price points available from European manufacturers had become structurally incompatible with the commercial model the industry had built around Asian supply. The switching cost rose every year, not because anyone was raising it deliberately, but because the alternative was quietly ceasing to exist. By the late 1990s, Asian sourcing was not a strategy European fashion businesses were choosing. It was simply how the industry worked, invisible as infrastructure, assumed as a given, as natural and as unexamined as the electricity supply.
Move 3: Leveraging the Dependency
The leverage in this case does not arrive through a single actor making a single decision. It arrives through the structural arithmetic of what four decades of sourcing choices have built. The European fashion industry has spent those decades optimising for a supply chain it does not own. The productive infrastructure, the factories, the integrated fabric and dyehouse networks, the finishing capacity, the logistical systems built around volume and speed, exists in Asia. European brands retain the design brief, the brand identity, and the retail margin. The knowledge of how to turn the brief into a garment at speed and at scale belongs to someone else.
Shein is not the cause of this condition. It is the condition made explicit and weaponised. Founded in its current fast fashion form in 2015 and headquartered in Singapore with its manufacturing operations centred on Guangzhou, Shein is a producer-integrated retailer that controls the full value chain from real-time consumer data collection through algorithmic trend identification, design, manufacture using its dense network of Guangzhou-area suppliers, and direct-to-consumer logistics without a European brand intermediary at any point. It adds approximately 2,000 to 6,000 new items to its store each day, with design-to-production cycles running between three and fourteen days depending on the product and whether it represents a new design or a reorder of a proven style. It sells complete outfits at price points that European brands cannot approach, not because European designers lack talent or European consumers lack taste, but because Europe no longer possesses the production infrastructure that would allow it to compete on either speed or cost. Shein did not create this situation. It found an industry that had transferred its productive capacity to the system Shein operates within, and it competed on the terms that system makes possible. The leverage is structural, not conspiratorial.
The direct leverage is visible in the commercial pressure on European fashion businesses: compressed margins, accelerated trend cycles they cannot match, and a consumer price expectation set at a floor European production cannot reach. European fast fashion incumbents have responded by deepening their own Asian sourcing commitments, which is the dependency grammar’s most characteristic response to leverage: compliance that deepens the exposure rather than addressing it.
The option-space leverage is subtler and more durable. European brands do not need to be told that rebuilding domestic or near-shore manufacturing capacity is not viable. They already know it, because the ecosystem that would make it viable, the suppliers, the skilled workforce, the training infrastructure, the minimum order economics, no longer exists at the required scale. The option was not closed by a decision. It was closed by the atrophy of everything that would have made it real.
The Xcelerator programme, through which Shein has begun offering its supply chain infrastructure as a service to smaller European brands, is the next move in the sequence and the most structurally revealing. Launched in 2025, it gives participating brands access to factories with five-to-seven-day turnaround times, in exchange for opening a storefront on Shein’s marketplace. A brand that adopts Xcelerator gains production speed and reduced minimum order quantities. It also embeds itself operationally in a competitor’s infrastructure, transfers its production data, its trend signals, its volume patterns, and its customer behaviour into a system whose operator has every incentive to use that data to compete more effectively against the brand that supplied it. This is dependency creation at the point of leverage. The industry that built the original infrastructure and then surrendered it is now being offered access to what it surrendered, as a service, on terms set by the party that holds it.
Move 4: Obscuring the Mechanism
The central obscuring mechanism in this case is moral displacement, and it operates with considerable sophistication because the moral concerns it foregrounds are entirely legitimate. The debate about the European textile industry’s dependence on Asian production is conducted almost exclusively in the registers of labour rights, environmental impact, and consumer ethics. Shein’s workers in Guangzhou work punishing hours for piece rates. The industry’s carbon footprint is substantial. Fast fashion’s waste volumes are genuinely alarming. These are real problems. They are also the visible hands.
What the ethical framing successfully keeps off the table is the structural question: how did the European fashion industry arrive at a position in which it cannot compete with a company doing what European manufacturers were once paid to do, on European soil, with European workers? That question requires examining forty years of sourcing decisions, the commercial logic that drove them, the institutional frameworks that enabled and rewarded them, and the policy choices that cleared the path. The ethical frame redirects that examination toward the symptoms and away from the architecture that produced them. The European Commission’s primary policy response to the structural situation is organised around precisely this redirection. The EU Strategy for Sustainable and Circular Textiles, published in March 2022 and progressively implemented through the Ecodesign for Sustainable Products Regulation (which entered into force in July 2024), mandatory Extended Producer Responsibility schemes, and proposed Digital Product Passports, addresses labour standards, chemical safety, circularity, and end-of-life waste management. These instruments address the conditions of production in the existing supply chain. They do not address the question of why the supply chain is where it is, who put it there, and what it would take to change that.
The temporal displacement mechanism runs in close parallel. The executives who led the offshoring decisions of the 1980s and 1990s retired with the margin improvements those decisions generated already booked. The institutional memory of what European textile manufacturing at scale actually required, the supplier ecosystems, the vocational training pipelines, the minimum viable production volumes, was not recorded in any form that survived the departure of the people who held it. The industry leaders and policymakers who now face Shein’s structural challenge inherited a supply chain architecture they did not choose and, in most cases, a knowledge base that no longer includes what rebuilding would require. The temporal gap between the decisions and the consequences is precisely long enough to make accountability structurally impossible. There is no one to call. The decisions were legal, were rewarded by financial markets, and were made in good faith within the commercial logic of their moment. The consequence is the current situation, and the current situation has no author.
The consumer preference argument deserves separate attention because it is the obscuring mechanism that feels most like an explanation. The claim that European brands moved production to Asia because consumers demanded cheap clothing reverses the actual sequence. The price points that trained European consumers to expect cheap clothing were made possible by the offshoring decisions that preceded and enabled them. Consumer expectation is the normalised outcome of the structural decision, not its cause. Treating consumer preference as the explanatory variable is precisely the kind of rationalisation that makes the grammar invisible: it locates agency in the diffuse, unorganisable mass of individual consumer choices rather than in the specific, traceable, institutionally embedded decisions of brand executives, retail buyers, and financial capital.
Move 5: Converting the Value
The conversion in this case has occurred at several levels simultaneously, and the most consequential of them is the one least amenable to financial measurement.
The commercial conversion is visible and ongoing. Productive capacity, the ability to manufacture garments at speed, at scale, and at a price point the market will bear, has migrated from Europe to Asia. The value that European textile manufacturers once captured through production, wages, supplier margins, and the full vertical chain of textile manufacturing, now accrues to Asian producers and to the retailers who source from them at prices European manufacturers cannot meet. The European fashion brand retains the brand premium. It does not retain the productive infrastructure through which that premium was once generated. It has become, structurally, a design and marketing operation dependent on a production system it neither owns nor controls.
The knowledge conversion is more durable and more difficult to reverse. Tacit manufacturing knowledge, the kind embedded in skilled hands and accumulated through the repetition of complex physical processes, does not transfer cleanly in either direction. When European textile manufacturing expertise migrated through the movement of sourcing relationships, through the closure of factories, the retirement of skilled workers, and the contraction of vocational training systems, it did not simply relocate. It atrophied in Europe and compounded in Asia. The receiving countries built on the knowledge transferred to them. Their workforces became more skilled, their production systems more sophisticated, their supplier ecosystems more integrated. The gap between European and Asian production capability, which was initially a labour cost gap, has become a knowledge and infrastructure gap that labour cost alone can no longer explain or close.
The normative conversion is the dimension least visible from inside Europe. The production system that now supplies the European fashion market sets the standards, the pace, the minimum order economics, and the price expectations against which every other production option is measured. European near-shoring advocates, when they make their case for restoring domestic or regional production capacity, are measured against a benchmark that the existing system, the dependency they are proposing to exit, has established. The option they are proposing is evaluated on the terms of the arrangement they are proposing to leave, which is a reliable way to ensure that the alternative never competes on a level surface.
The counter-move that was available and was not taken deserves to close this move, because it is the most structurally revealing detail in the case. Digital textile printing technology, developed from the 1990s onward, offered European producers a route to near-shore competitiveness that did not require matching Asian volume or labour costs. Short-run, on-demand, just-in-time production of printed textiles, enabled by digital printing systems, was technically viable and economically rational for certain market segments. It offered speed and flexibility where Asian producers offered scale and cost. It was a different competition, on different terms, in a space the existing dependency could not easily occupy. The technology was adopted. The commercial model was not restructured around it. European print service providers and fashion suppliers acquired the equipment and continued to operate within a business logic built for a different production architecture. The dependency grammar absorbed the counter-move, not by defeating it but by leaving its institutional context unchanged. The tool arrived. The structural conditions that would have made the tool transformative did not.
Closing
What this case reveals about the grammar is something the OPEC investigation began to surface and the Monsanto case confirmed: the most complete dependencies are the ones built by the dependent party itself. OAPEC found a dependency that postwar Western policy had constructed over thirty years. Monsanto found a seed supply system that agricultural modernisation had made structurally amenable to consolidation. The European fashion industry did not have its productive infrastructure taken. It transferred it, in exchange for margin improvements that were real, immediate, and individually rational, across a period long enough that no single actor ever faced the full structural consequence of the aggregate. By the time the consequence became visible, the decision-makers had dispersed, the institutional memory had evaporated, and the infrastructure that might have sustained a different outcome had ceased to exist.
The grammar’s most durable feature is not leverage. Leverage is visible and can, in principle, be resisted. The most durable feature is the point at which the dependency stops being a relationship and becomes the architecture. When that transition occurs, the dependent party is not choosing to remain dependent each day. It is simply operating within conditions it did not choose and can no longer clearly see. The switching cost is no longer a calculation. It is a fact about the world.
The question the reader should carry forward is not how to avoid being Shein’s competitor. It is the earlier question, the one that was never asked in any European boardroom or trade ministry in the 1980s or 1990s with sufficient seriousness to alter what was being built: what are we actually transferring, and to whom, and what will it cost to get it back?
Comments are closed