When trade policy becomes industrial policy

Carbon borders, chip embargoes, procurement walls, and forced localisation. The regulations reshaping who can build what, where, and for whom.

By VastBlue Editorial · 2026-03-26 · 17 min read

Series: The Chessboard · Episode 5

The customs officer as industrial planner

There was a time — recent enough that many of the people who shaped the policy still hold office — when trade policy and industrial policy were understood as fundamentally different instruments with fundamentally different purposes. Trade policy regulated what crossed borders: tariffs, quotas, anti-dumping duties, rules of origin. Industrial policy determined what got built domestically: subsidies, tax incentives, state-owned enterprises, strategic procurement. The distinction was not merely academic. It was constitutional. The World Trade Organization, established in 1995 as the institutional embodiment of the post-Cold War consensus on economic governance, was explicitly designed to discipline trade barriers while leaving industrial policy — messy, political, domestically contentious — largely outside its jurisdiction. Trade was multilateral. Industry was sovereign. The boundary between them was policed with theological rigour by trade lawyers in Geneva and economics professors in Cambridge.

That boundary has dissolved. Not gradually, through the slow erosion of institutional norms, but rapidly, through a series of deliberate policy decisions made by the world's major economic powers between 2018 and 2025. The United States imposed export controls on semiconductor technology that explicitly aimed to cripple China's ability to manufacture advanced chips — not to protect American industry from unfair competition, but to prevent a geopolitical rival from acquiring a strategic capability. The European Union implemented the Carbon Border Adjustment Mechanism, which applies a carbon cost to imports that reshapes competitive dynamics across entire industrial sectors. China restricted exports of critical minerals that the rest of the world depends on for battery manufacturing, semiconductor production, and defence applications. In each case, the instrument was formally a trade measure — it regulated what crossed a border. But the objective was industrial — it determined who could build what, where, and for whom.

This convergence of trade and industrial policy is the defining regulatory development of the 2020s, and its implications for European companies are profound. Every firm that sources components internationally, sells into regulated markets, bids on public contracts, or handles data across jurisdictions is operating in an environment where the rules of trade are no longer neutral facilitators of commerce but active instruments of industrial strategy. Understanding this shift is not a compliance exercise. It is a survival requirement.

Carbon at the border: CBAM as industrial architecture

The Carbon Border Adjustment Mechanism is, on its surface, an environmental measure. It imposes a carbon cost on imports of carbon-intensive goods — steel, aluminium, cement, fertilisers, electricity, and hydrogen — equivalent to the cost that European producers bear under the EU Emissions Trading System. The stated objective is to prevent carbon leakage: the scenario in which European carbon pricing drives production to jurisdictions with weaker environmental regulation, resulting in no net emissions reduction and a loss of European industrial capacity. The logic is elegant, the environmental case is strong, and the mechanism has been endorsed by climate economists across the political spectrum as a theoretically sound approach to maintaining climate ambition without sacrificing industrial competitiveness.

But CBAM is not merely an environmental measure. It is an industrial policy instrument of extraordinary ambition, and understanding it as such is essential to understanding what it actually does. By imposing a carbon cost on imports, CBAM creates a structural preference for production in jurisdictions with robust carbon pricing systems — which currently means production in Europe, or production in countries that adopt carbon pricing systems compatible with Europe's. The mechanism does not subsidise European industry. It does something potentially more powerful: it imposes a cost on foreign industry that can only be avoided by decarbonising production or paying at the border. The effect is to make the EU's environmental regulatory framework a condition of market access, extending European climate policy beyond European borders through the mechanism of trade.

The industrial implications ripple outward in ways that the mechanism's architects understood but its public presentation often understates. Consider Turkish steel. Turkey is Europe's second-largest source of steel imports, exporting approximately 8 million tonnes of steel products to the EU annually. Turkish steelmaking relies heavily on electric arc furnaces powered by a grid that remains substantially dependent on coal and natural gas. Under CBAM, Turkish steel exporters will face carbon costs that could add 10 to 15 per cent to the price of their product in European markets. Turkey has two options: absorb the cost and accept lower margins, or decarbonise its steel production — which means decarbonising its electricity grid, a multi-decade, multi-billion-euro undertaking. A third option, of course, is to lose the European market and redirect exports elsewhere. Each of these outcomes is, from the perspective of European industrial strategy, desirable. CBAM either generates revenue (from the certificates), drives decarbonisation abroad (benefiting the climate), or reduces foreign competition (benefiting European producers). It is a trade measure that functions as an industrial subsidy, an environmental policy, and a geopolitical lever simultaneously.

The mechanism's scope, however, reveals its limitations — and the political constraints that shaped it. CBAM currently covers only six product categories, all of them upstream industrial materials. It does not cover manufactured goods. A Chinese-made electric vehicle, whose production involves steel, aluminium, battery chemicals, and electricity — all carbon-intensive inputs — enters the European market with no carbon border adjustment. The aluminium in the vehicle's body might be covered if imported as raw aluminium, but once it is processed into a car door panel and assembled into a finished vehicle, the embedded carbon becomes invisible to the mechanism. This scope limitation is not accidental. Extending CBAM to finished goods would require measuring the embedded carbon content of complex manufactured products — a technical challenge of formidable difficulty — and would provoke trade disputes with virtually every major exporting country simultaneously. The political will for that confrontation does not yet exist.

CBAM is the most sophisticated instrument of trade-as-industrial-policy that any jurisdiction has yet deployed. Its limitation is not its design — it is the political courage required to extend it to the products that actually matter.

Editorial observation

The broader significance of CBAM extends beyond its immediate product scope. It establishes a principle — that environmental regulation can be enforced at the border — that has no natural limiting boundary. Once the mechanism is operational and the administrative infrastructure is built, extending it to additional product categories becomes a technical question rather than a conceptual one. The European Commission has already signalled that it will consider expanding CBAM's scope in the 2026-2028 review period. Organic chemicals, plastics, and downstream steel and aluminium products are all candidates. If CBAM eventually covers the full carbon footprint of manufactured goods, Europe will have constructed, through trade regulation, the most comprehensive system of extraterritorial environmental governance in history. Whether this constitutes trade policy or industrial policy is a distinction without a meaningful difference.

The chip embargo: export controls as economic warfare

On October 7, 2022, the US Bureau of Industry and Security published a set of export control regulations that represented the most significant use of trade restrictions as a tool of industrial destruction since the Cold War. The October 7 controls — as they became known in semiconductor industry shorthand — restricted the export to China of advanced semiconductor manufacturing equipment, advanced chips above specified performance thresholds, and the services of US persons (including US permanent residents) supporting advanced semiconductor manufacturing in China. The controls were subsequently expanded in October 2023 and again in 2024, each iteration closing loopholes and extending coverage to additional technologies, geographies, and activities.

The framing was national security. The objective was industrial. US National Security Advisor Jake Sullivan articulated the logic with unusual candour in a September 2022 speech: the United States would no longer seek to maintain a "relative" advantage over China in critical technologies, accepting that China would advance but more slowly. Instead, the US would seek to maintain "as large a lead as possible." The implication was explicit: export controls would be used not to regulate trade but to arrest the technological development of a strategic competitor. This was not a tariff designed to protect domestic industry from foreign competition. It was a prohibition designed to prevent foreign industry from existing.

Europe's role in this architecture of technological containment is central and uncomfortable. ASML, the Dutch manufacturer of extreme ultraviolet lithography equipment, was the most strategically important company affected by the export controls. Without ASML's EUV machines, advanced chip fabrication below approximately 7 nanometres is physically impossible. The United States pressured the Netherlands — extensively and successfully — to restrict ASML's exports of EUV systems to China. In January 2023, the Dutch government implemented export restrictions that effectively barred ASML from selling its most advanced lithography systems to Chinese customers. Japan followed with parallel restrictions on semiconductor manufacturing equipment from Tokyo Electron, Nikon, and other suppliers. The trilateral US-Netherlands-Japan alignment on semiconductor export controls created a comprehensive embargo on advanced chipmaking technology that China could not circumvent through alternative sourcing.

For Europe, the implications extend far beyond ASML. The export controls establish a precedent that European companies find profoundly discomfiting: that the United States can, through a combination of direct regulation and diplomatic pressure, dictate the terms on which European companies conduct trade with third countries. ASML did not choose to stop selling to China. It was compelled to do so by a regulatory architecture in which American jurisdiction extends, through the "foreign direct product rule," to any technology that incorporates US-origin components, software, or know-how. Since virtually every semiconductor manufacturing tool contains some American-origin technology, the foreign direct product rule gives the US Bureau of Industry and Security effective veto power over equipment sales by companies in any country — including European companies that are not subject to US law but are dependent on US-origin inputs.

ASML's 2024 annual report disclosed that China had been its largest single market by revenue in the preceding year — a fact that made the commercial cost of the export controls unmistakable. The Dutch company was being asked to forgo its most profitable market to serve a strategic objective defined by a foreign government. The diplomatic framing emphasised shared values and mutual security interests. The economic reality was that a European company was bearing a disproportionate share of the cost of an American strategy. This asymmetry — where the United States defines the strategy and Europe absorbs the cost — has become a recurring feature of the new trade-as-industrial-policy landscape.

Procurement walls and the return of economic nationalism

If CBAM uses environmental regulation as an industrial instrument and export controls use security regulation as an industrial instrument, procurement rules represent the most direct fusion of trade and industrial policy: the government as customer, using its purchasing power to determine which companies, from which countries, using which technologies, are permitted to participate in public markets. Public procurement accounts for approximately 14 per cent of European GDP — over €2 trillion annually across the EU member states. The rules governing who can bid on that spending are, in effect, the largest industrial policy lever available to any European government.

For decades, the European approach to public procurement was governed by a principle of open competition. The EU's procurement directives required member states to award public contracts on the basis of objective criteria — price, quality, technical capability — without discrimination based on the nationality of the bidder. This openness was consistent with the EU's broader commitment to free trade and was enshrined in the Government Procurement Agreement (GPA) at the WTO, to which the EU is a signatory. European public contracts were, in principle, open to bidders from any GPA signatory country.

That principle is being systematically qualified, narrowed, and in some domains abandoned. The drivers are multiple and reinforcing. The COVID-19 pandemic revealed that European public health systems were dependent on imported PPE, pharmaceuticals, and medical devices that became unavailable when every country simultaneously attempted to secure domestic supply. The semiconductor shortage demonstrated that European automotive and defence industries relied on chips manufactured in geopolitically vulnerable geographies. Russia's invasion of Ukraine exposed Europe's dependence on Russian energy and prompted a wholesale reassessment of supply chain security across all critical sectors.

The European Commission's International Procurement Instrument (IPI), which entered into force in August 2022, marked the formal shift. The IPI gives the Commission the power to restrict access to EU public procurement markets for companies from countries that do not offer reciprocal access to their own procurement markets. The target, though diplomatically unnamed, is primarily China — which maintains extensive restrictions on foreign participation in its public procurement while Chinese companies bid aggressively on European contracts, particularly in infrastructure, telecommunications, and renewable energy. The IPI allows the Commission to impose score adjustments of up to 100 per cent on bids from non-reciprocal countries, effectively pricing them out of European public markets.

But the IPI is only the most visible element of a broader procurement transformation. Across member states, national security and strategic autonomy considerations are being embedded into procurement criteria in ways that reshape competitive dynamics without triggering formal trade disputes. France's 2024 defence procurement reforms introduced "European preference" criteria for military contracts, explicitly favouring bidders with European-headquartered parent companies and European-located production facilities. Germany's revised procurement rules for critical infrastructure — including telecommunications, energy, and transport — now require security assessments of vendors from "non-allied" countries, a category broad enough to exclude Chinese and, in some interpretations, even some American technology suppliers. Several member states have implemented or are considering "strategic autonomy" criteria that weight bids higher when the technology and intellectual property are European-controlled.

The cumulative effect is the construction of a procurement architecture that, while never explicitly protectionist in its language, operates as a systematic preference for European suppliers in European public markets. The legal framing emphasises security, resilience, and reciprocity rather than industrial advantage. But the industrial effect is identical: European companies that manufacture in Europe, employ European workers, and maintain European-controlled supply chains receive structural advantages in the largest single market for public spending on earth.

No European politician will say the words "Buy European." But across procurement law, security screening, and strategic autonomy criteria, that is precisely what is being built — one contract clause at a time.

Editorial observation

Forced localisation: when data and production cannot leave

The final element in the convergence of trade and industrial policy is forced localisation — the requirement that certain activities, data, or production processes remain within a specific jurisdiction as a condition of market access. Localisation requirements have existed in various forms for decades: local content rules in automotive manufacturing, domestic broadcasting quotas in media regulation, technology transfer requirements in developing-country investment regimes. What is new is their proliferation, their extension to digital services and data, and their increasingly explicit connection to industrial strategy rather than mere regulatory sovereignty.

Data localisation is the most rapidly expanding frontier. The EU's General Data Protection Regulation, while not formally a localisation mandate, creates conditions that strongly incentivise data processing within European jurisdiction. The invalidation of the EU-US Privacy Shield by the Court of Justice of the European Union in the 2020 Schrems II decision — on the grounds that US surveillance law did not provide adequate protection for European personal data — cast a legal shadow over transatlantic data transfers that persists despite the adoption of the EU-US Data Privacy Framework in 2023. For companies processing sensitive data — healthcare, financial services, critical infrastructure — the legal uncertainty of transatlantic transfers has made European data localisation the path of least regulatory risk.

The industrial effect is substantial. When data must be processed locally, the infrastructure to process it must also be local. Data centres must be built. Cloud services must be operated from European soil. Engineering teams must be located within European jurisdiction. The requirement cascades through the technology stack, creating demand for European data centre construction, European cloud services, European AI training infrastructure. This is precisely the industrial outcome that European policymakers desire — not through subsidy, but through regulation that makes non-European alternatives legally precarious.

Beyond data, physical production localisation requirements are proliferating in the sectors that define the energy transition. The EU Battery Regulation, which entered into force in August 2023, imposes lifecycle requirements — carbon footprint declarations, recycled content minimums, due diligence obligations, and digital battery passports — that strongly favour batteries manufactured in jurisdictions with transparent supply chains and robust regulatory compliance infrastructure. While the regulation does not explicitly require batteries to be manufactured in Europe, its compliance requirements are designed in a manner that European manufacturers can meet more readily than competitors operating in less regulated environments. The effect, whether intended or emergent, is a localisation bias embedded in technical regulation.

France has been the most aggressive European proponent of localisation as industrial strategy. The French government's "sovereign cloud" doctrine requires that classified and sensitive government data be processed only on cloud infrastructure certified under the SecNumCloud standard — a certification that, by design, requires the cloud provider to be majority-owned by European entities and to operate without any extraterritorial legal obligations. The standard effectively excludes AWS, Azure, and Google Cloud from the most sensitive segments of the French public cloud market unless they restructure their European operations into legally independent entities — which Microsoft and Google have begun doing through joint ventures with French partners Orange and Thales, respectively. The industrial logic is transparent: France is using security certification as a mechanism to force the creation of European-controlled cloud infrastructure that would not exist if market forces alone determined the outcome.

Germany's approach has been more cautious but directionally similar. The German Federal Office for Information Security (BSI) has proposed technical sovereignty requirements for critical infrastructure operators that would mandate European-controlled technology stacks for certain applications. The Gaia-X initiative, co-founded by France and Germany, was explicitly designed to create a European data infrastructure ecosystem as an alternative to American hyperscalers. While Gaia-X has struggled with implementation — its governance has been criticised as overly complex, and its technical deliverables have lagged expectations — the strategic intent is clear: to create the conditions under which European data processing becomes not just an option but a requirement.

The new operating environment

What emerges from these parallel developments — carbon border adjustments, export controls, procurement preferences, and forced localisation — is not a coherent grand strategy but something more consequential: a new operating environment in which the rules of international trade have been repurposed as instruments of industrial competition. The rules are not being broken. They are being rewritten, reinterpreted, and extended in ways that the institutional architecture of the post-1995 trading system was not designed to accommodate.

For European companies, this new environment presents both structural risks and structural opportunities, and distinguishing between them requires a granularity of analysis that most strategic planning processes do not yet incorporate. The risks are concentrated in two areas. First, regulatory fragmentation: a company that manufactures in Europe, sells globally, and sources components from multiple jurisdictions must now navigate an expanding matrix of carbon border costs, export control compliance, procurement eligibility requirements, and data localisation mandates that vary by destination market, by product category, and increasingly by the nationality of the company's ultimate parent. Compliance costs are rising. Lead times for market access are lengthening. The friction that trade policy was originally designed to reduce is being reintroduced, deliberately, as a tool of strategic competition.

Second, retaliatory escalation. CBAM has already provoked formal objections from India, Brazil, South Africa, and Turkey — all of which argue that the mechanism is a disguised trade barrier that discriminates against developing countries whose decarbonisation is less advanced than Europe's. China has responded to semiconductor export controls with critical mineral export restrictions that directly affect European battery and electronics manufacturers. The United States' use of the foreign direct product rule to control European companies' trade with China has generated quiet but genuine resentment in European capitals and corporate boardrooms. Each instrument of trade-as-industrial-policy invites a response, and the cumulative effect of these responses is a trading system that is becoming progressively more adversarial, more fragmented, and more expensive to navigate.

The opportunities are less obvious but potentially more durable. European companies that are headquartered in Europe, manufacture in Europe, and maintain European-controlled supply chains are increasingly advantaged in the European market — not because they are more competitive in a market sense, but because the regulatory architecture is being constructed around them. A European battery manufacturer that complies with the Battery Regulation, sources materials through transparent supply chains, and processes data within European jurisdiction faces lower regulatory friction than a Chinese competitor whose supply chain transparency is harder to demonstrate and whose data processing practices may not satisfy European requirements. This is not competitiveness earned through innovation or efficiency. It is competitiveness conferred through regulatory design. Whether this is desirable policy is debatable. That it is the emerging reality is not.

• CBAM imposes carbon costs on imports — creating structural advantage for decarbonised European production in covered sectors.
• Semiconductor export controls give the US effective veto power over European companies' trade with China — reshaping supply chain strategy across the electronics industry.
• The International Procurement Instrument restricts access to €2 trillion in European public spending — favouring companies from countries offering reciprocal market access.
• Data localisation and sovereign cloud requirements force technology infrastructure investment into European jurisdiction — creating demand for European cloud, data centre, and AI capability.
• Battery Regulation compliance requirements favour manufacturers with transparent, European-aligned supply chains — raising the bar for non-European competitors in the EV value chain.

The deeper transformation is conceptual. For three decades, the dominant framework for understanding international trade was that regulation should be minimised, that markets should be opened, and that the role of government was to set neutral rules and then step back. That framework has been replaced — in Washington, in Beijing, in Brussels — by a framework in which regulation is an active tool of economic strategy, markets are instruments to be shaped rather than forces to be respected, and the role of government is to determine outcomes rather than merely to facilitate them. Trade policy has become industrial policy. The customs officer has become the industrial planner. And the companies that understand this shift earliest will be the ones that navigate it most successfully.

The question is no longer whether governments will use trade regulation as industrial policy. They already do, on every continent, in every major economy. The question is whether European companies recognise that the rules of the game have changed — and whether they are building their strategies for the game that is being played, not the one that used to be.

Editorial observation

Sources

1. European Commission — Carbon Border Adjustment Mechanism — https://taxation-customs.ec.europa.eu/carbon-border-adjustment-mechanism_en
2. US Bureau of Industry and Security — October 7, 2022 Export Controls — https://www.bis.doc.gov/index.php/documents/about-bis/newsroom/press-releases/3158-2022-10-07-bis-press-release-advanced-computing-and-semiconductor-manufacturing-controls-final/file
3. European Commission — International Procurement Instrument — https://policy.trade.ec.europa.eu/enforcement-and-protection/international-procurement-instrument_en
4. ASML Annual Report 2024 — https://www.asml.com/en/investors/annual-report
5. EU Battery Regulation — Regulation (EU) 2023/1542 — https://eur-lex.europa.eu/eli/reg/2023/1542/oj
6. Jake Sullivan — Remarks on the Biden-Harris Administration's National Security Strategy — https://www.whitehouse.gov/briefing-room/speeches-remarks/2022/09/16/remarks-by-national-security-advisor-jake-sullivan-at-the-special-competitive-studies-project-global-emerging-technologies-summit/
7. OECD Digital Services Trade Restrictiveness Index — https://www.oecd.org/trade/topics/digital-trade/
8. ANSSI — SecNumCloud Qualification — https://www.ssi.gouv.fr/administration/qualifications/prestataires-de-services-de-confiance-qualifies/prestataires-de-cloud-secnumcloud/