ASML reported that in 2026 it sold 'absolutely nothing' of its advanced lithography systems in Europe. The situation follows from a pattern in which the final production stage is routinely moved elsewhere, weakening any national claim to the finished product.
When a complex good is broken into discrete steps, each step can be located where the immediate cost or subsidy advantage is greatest. The firm that performs the earlier step receives payment for its output and has no contractual reason to keep the later step nearby; the firm that performs the later step receives payment for its service and has no reason to locate upstream. The decision to shift a stage is made locally, based on factors such as wage levels, tax incentives, proximity to customers, or access to specialized equipment. Because the stages are loosely coupled — information about where the upstream step occurred is not required for the downstream step to proceed — the chain can drift without any single actor noticing a loss of aggregate value. Over time, the drift can become systematic: the upstream step remains in a high‑cost region because of sunk capital or skilled labor, while the downstream step migrates to a low‑cost region for packaging, testing, or final assembly. The result is a product whose substantive transformation occurs in multiple jurisdictions, yet whose marketing label can be attached to any one of them depending on local rules about what constitutes “substantial transformation.”
The mechanism is not confined to semiconductors. In the medieval wool trade, English producers exported raw wool to Flanders where it was fulled and dyed; English statutes attempted to keep finishing at home, but the higher quality and lower cost of Flemish work drew the process abroad, leaving England with only the low‑value raw material. In the nineteenth‑century patent‑medicine boom, manufacturers sourced active ingredients such as quinine from South America and alcohol from the United States, compounded them in Britain, and sold the product as a “British remedy.” Early food‑labeling laws later emerged because consumers could not discern where the curative substance actually originated. In the twentieth‑century apparel industry, design houses in Milan outsourced fabric weaving to Pakistan, cutting and sewing to Bangladesh, and then marketed the garments as “Italian design,” relying on the fact that the creative stage remained in Europe while the labor‑intensive stages moved to lower‑wage locales. In pharmaceutical manufacturing, the active pharmaceutical ingredient is frequently produced in India or China, while the final formulation, packaging, and labeling occur in Europe or the United States; regulatory frameworks struggle to attribute national origin when the most chemically significant step occurs elsewhere. In software development, a firm may write core algorithms in multiple countries, compile the binary in Ireland to avail of a low corporate‑tax rate, and distribute the product as “Irish software,” even though the majority of code was written outside Ireland.
Each of these cases shares the same causal structure: a production process is divided into stages, each stage is governed by a separate profit motive, the stages are interchangeable enough to be relocated without disrupting the flow of goods, and the information needed to trace the complete transformation is not required for any single stage to operate. The coupling between stages is weak enough that a firm can optimize its own location decision without internalizing the external effect on the ability of any polity to claim the product as its own. The consequence is a systematic erosion of the link between a geographic label and the actual locus of value creation. Policymakers who try to stimulate domestic demand by favoring “Made in X” goods find that the label no longer guarantees that a meaningful share of the product’s value was generated within X; the incentive to invest in local capabilities diminishes because the returns can be captured elsewhere through stage shifting.
The drift also creates a feedback loop that reinforces the movement of stages. As more firms locate a particular stage in a low‑cost jurisdiction, that jurisdiction develops agglomeration economies — specialized suppliers, skilled labor, infrastructure — that further reduce the cost of locating there. The original high‑cost region retains only the stages that are tied to immovable assets such as expensive fabrication plants or proprietary design teams, but even those assets may eventually be relocated if the cost differential becomes large enough or if political risk changes. The United States’ recent decision to reshore advanced semiconductor fabrication, cited in the signal as a reaction to perceived supply‑chain vulnerability from China‑Taiwan tensions, illustrates how policymakers may intervene when the drift threatens a strategic capability they deem essential. Yet such interventions are reactive; they address the symptom of a missing stage rather than the underlying incentive to fragment production.
Because the mechanism depends only on the divisibility of a process into tradable steps and the absence of a requirement to disclose where each step occurs, it can appear in any industry where goods or services are assembled from separable components. The historical record shows that whenever firms can unbundle production, they will do so to capture local advantages, and the resulting geographic diffusion of value makes national attribution increasingly tenuous. The signal from ASML is therefore not an isolated anomaly but a concrete manifestation of a recurring structural tendency: when the final stage of production is routinely detached from the earlier stages, the capacity of any single polity to credibly claim ownership of the finished product erodes, and the policy levers that rely on that claim lose their purchase.