How European Is 'European'? The Industrial Accelerator Act and the Politics of Being a Friend

On March 4, 2026, the European Commission tabled the Industrial Accelerator Act (IAA), the legislative centrepiece of its Clean Industrial Deal. The Act pursues an ambitious target: raising the share of industrial manufacturing in EU GDP from 14.3% in 2024 to 20% by 2035. To get there, it does something the EU has never done at this scale: it builds 'Made in EU' preferences into public procurement and public support schemes for strategic sectors such as steel, cement, aluminium, cars and net-zero technologies. Public money, in short, is to buy European.

But the proposal's most contested element is not the preferences themselves; it is the definition of what counts as 'European'. This question has been debated at length, in the briefing rooms of the Parliament, in the Council's working parties, and in the commentaries that followed the proposal, and it is what divides the three institutions. The question deserves a closer look, because what the debate is really about is whether access to the EU's market is a matter of treaty or a matter of grace.

Being European, by Treaty

Alongside 'content of Union origin' (Article 7), the proposal establishes the concept of 'content equivalent to Union origin' (Article 8). Under the Commission's proposal, content is treated as European not only when it originates in the EU, but also when it originates in third countries with which the EU has concluded an agreement establishing a free trade area or a customs union, or that are parties to the WTO Agreement on Government Procurement (GPA), where relevant Union obligations exist under that agreement. For public support schemes (Article 9), the GPA route is unavailable: only FTA and customs-union partners qualify there.

The consequence of not being 'European' in the IAA's sense is therefore stark, though its reach should not be overstated. The operative provision is Article 11: bidders owned or controlled from third countries without an access agreement are excluded from the procurement procedures in Annexes II and III, which cover specified products (at least 25% low-carbon steel, 5% low-carbon Union-origin concrete and mortar, 25% low-carbon Union-origin aluminium, and the vehicle rules of Annex III) in procedures under the classical procurement directives. The exclusion is not absolute: it falls away where there is a single supplier, where no suitable tender has been offered in the past two years, or where compliance would entail disproportionate cost or technical incompatibility, with a cost difference above 25% presumed disproportionate; for public support schemes, Article 12 allows departures where the requirements raise a product's cost by more than 30%. Still, for a jurisdiction's companies, being on the wrong side of the definition amounts to exclusion, subject to exceptions, from the segments of the EU market through which the EU intends to channel its industrial transition.

Which countries fall on which side? The defensible part of the design is that it is anchored in reciprocity. The EU has no free trade agreement with China, whose companies therefore fall outside the definition across the board. The United States, by contrast, is a GPA party, so its companies would in principle retain access to EU procurement under Article 8, though not to the public support schemes of Article 9. And the category of 'European' suppliers is wide: the EU operates the largest network of trade agreements of any jurisdiction in the world, and although fewer than thirty of those FTAs contain procurement rules, Commission estimates suggest the equivalence clause would still cover some twenty countries, from Canada to Chile to Korea. India, with which the EU is about to conclude an FTA, is not among them. The Commission has defined the concept functionally rather than geographically: 'European' means 'partner'. As a matter of trade policy, that is coherent. As a matter of semantics, it stretches the notion of 'European' well beyond recognition.

The Parliament's Alternative: A Door That Opens and Shuts

The file is not yet in the trilogue, but an agreement is targeted for the end of 2026 under the One Europe, One Market roadmap. Inside the European Parliament, meanwhile, a different philosophy is gaining ground. In their draft report of September 9, 2026, the three rapporteurs, Christophe Grudler (Renew, ITRE), Pierre Jouvet (S&D, IMCO) and Anna Cavazzini (Greens/EFA, INTA), propose abandoning default equivalence for FTA partners. Instead of treating every country with which the EU has concluded an FTA as 'European' by operation of law, the rapporteurs opt for a targeted 'opt-in' system: third-country content would count as Union origin only once a series of conditions are met, including de jure and de facto national treatment, consistency with the EU's economic security, equivalent competitive conditions, and respect for ILO conventions. With more than 5,000 amendments tabled, the committees are not due to vote until December 1, so the shape of Parliament's final position remains uncertain. But the direction of travel is clear.

On this approach, being a friend is not a status. It is a decision, taken by the EU, that can be revisited. And the rapporteurs have settled the institutional design: the Commission would include a third country by delegated act, once the conditions are met, and would have to remove it once they lapse. That gives the Commission the power to open and close access to the EU procurement market for entire jurisdictions, with the Parliament and the Council retaining a right of objection before the act enters into force and a right of revocation afterwards. Before adopting a delegated act, the Commission must consult Member State experts, whose advice it is free to disregard. The choice of instrument is revealing: it keeps the decision in the Commission's hands, with the legislature present only at the veto points. Either way, the equivalence of a third country's content becomes an object of continuous negotiation between the EU institutions, and between the EU and its partners, on terms the EU itself sets. Access is no longer the consequence of a treaty obligation; it becomes a privilege, granted and withdrawn at the EU's discretion.

The Council's Counter-Draft: Back to Agreements

The Parliament's opt-in system, however, is not the only alternative on the table. The Irish Presidency's second compromise text of 30 July 2026 renames the concept altogether: where the Commission speaks of content 'equivalent to Union origin', the Council introduces 'partner origin' (new Article 7a), a term that candidly concedes the point that this is not about Europe. Eligibility still flows from the agreements, GPA parties and FTA or customs-union partners, calibrated product by product, but the Commission polices the friendships by implementing acts under the examination procedure, removing a country that fails to deliver promised national treatment, creates high-risk dependencies, or falls under an exception in the applicable agreement. Where the rapporteurs want to unmoor access from treaties, the Council keeps agreements as the anchor: the Commission still proposes and adopts the implementing acts, and the Member States' role under the examination procedure is confined to blocking them by qualified majority, a high threshold that in practice leaves the valve in the Commission's hands.

It is worth being precise about what the Commission's own proposal does, because the differences between the three institutions are smaller than the rhetoric suggests. Even under the Commission's text, equivalence is not a fixed status: Articles 8(2) and 9(2) already allow the Commission to exclude a country by delegated act if it fails to grant national treatment, or on other grounds. The real difference is the default. The Commission starts from default-in with an exclusion power; the Parliament starts from default-out with an inclusion power; the Council reverts to agreements, with the Commission policing the lists under comitology. In all three versions, the EU itself decides, and continues to decide, when access is granted and when it is taken away. The only question is which institution pulls the lever, and who can hold it back. Member States, moreover, have not been uniformly comfortable with the file's direction: in its BNC-fiche on the proposal, the Netherlands welcomed the aim of strengthening economic security and open strategic autonomy, but insisted that the measures remain targeted, proportionate and WTO-consistent, and cautioned against letting an industrial instrument blur into an economic-security tool with an outsized role for the Commission.

Reciprocity, Retrenchment, or Something Else?

What does the shift mean in international legal terms? General WTO law offers the EU considerable latitude: government procurement is carved out of the national treatment obligations of both the GATT (Article III:8(a)) and the GATS (Article XIII), and whether the MFN obligation of GATT Article I reaches procurement remains debated. What remains are the instruments that do cover procurement, the GPA and the procurement chapters of the EU's FTAs, and it is against those instruments that a system which opens the door to friends but reserves the right to shut it at any moment must be measured. The public support schemes of Article 9 of the Commission's proposal raise an additional question: linking consumption subsidies to Union-origin conditions is not sheltered by the procurement carve-out at all, and some experts see a serious risk of a WTO challenge there by countries that are not party to an FTA with the EU.

There is also a reciprocity problem. A system in which access is always conditional sits uneasily next to third countries that grant access on the basis of fixed, treaty-anchored commitments. Canada is the obvious comparison, though not an uncomplicated one: its Buy Canadian policy is agreement-based, but it has itself drawn WTO concerns, including from the EU and the US, that it flouts Canada's procurement commitments, and Canada has removed the Canadian International Trade Tribunal's jurisdiction over such challenges (SOR/2025-247). The deeper point survives the caveats: one partner treats the agreement as the deal; the other treats it as the opening bid.

More fundamentally, all three variants are manifestations of the EU's drive for strategic autonomy: the EU decides, and continues to decide, when access is granted, and it can take access away whenever it chooses. The question is whether this is what the EU needs in a world in which it has enemies, to be sure, but in which it also has friends. A system in which even friends must continuously earn their 'European' status, at the EU's discretion, is likely to alienate precisely the partners the EU is courting for its own geopolitical hedging, from Canada to Japan. It converts partnership into permanent probation. What presents itself as strategic autonomy begins to look like strategic retrenchment, and retrenchment of this kind is difficult to distinguish from protectionism.

And there is a domestic cost. Narrowing the pool of qualifying bidders means fewer candidates for EU tenders. Fewer candidates mean less competition, and less competition puts upward pressure on the prices of public projects. The Commission's own impact assessment expects net benefits of around €8 billion by 2030 under its preferred option, but that calculation assumes its own, default-in design; the more restrictive variants on the table would shrink the field further, and the cost-difference and disproportionate-cost escape valves can only compensate so far. Every public euro spent on an overpriced tender is a euro not spent on the next one. The IAA's stated objective is to increase decarbonised and resilient industrial production in the EU, in service of Europe's competitiveness, economic security and sustainable economic growth; a discretionary friends-list that narrows the field and risks inflating the bill works against all three. Europe risks ending up neither more autonomous nor more competitive, merely more closed.

The institutions have given themselves until the end of the year to reach an agreement. One hopes the negotiators remember that in a world where the EU's competitors use procurement to reward loyalty and punish enemies, the EU's comparative advantage was never that it could play that game better. It was that it did not play it at all.

Next
Next

Will Canada Become the EU’s First ‘Associate Member’?