On July 1, the European Union quietly crossed a Rubicon in trade policy. The new steel instrument – Regulation 2026/1384 – entered into application.

It is not merely a technical successor to the safeguard measures that expired the day before, after eight years and the exhaustion of their legal life under WTO rules. It is something more consequential: the first EU trade defence tool explicitly designed around the concept of global overcapacity rather than injury from a surge in imports, which is embedded in safeguard measures.

Beyond the rhetorical change, this new overcapacity instrument has basically halved tariff-free quotas for imported steel relative to the 2024 baseline. Furthermore, the out-of-quota import duty has been raised to a prohibitive 50%. Finally, there is much less room for circumvention of such import duties through third countries, thanks to “melt and pour” traceability requirements.

This change in Europe’s trade policy matters far beyond steel. For the first time, the EU has legislated on the premise that structural overcapacity – overwhelmingly, though not exclusively, Chinese – constitutes a trade distortion warranting a standing defensive architecture, without the straitjacket of the WTO Safeguards Agreement.

The instrument is, in effect, a test run. If it works – if it stabilises prices, deters trade deflection from the newly protected American market into the EU, and survives legal challenge – the pressure to replicate it for other sectors will be enormous. Chinese overcapacity is not a steel story; it is a systemic feature of an economy that persistently suppresses household consumption and channels savings into industrial investment. Excess capacity in electric vehicles, batteries, solar panels, wind turbines, legacy semiconductors, basic chemicals and, increasingly, robotics and machine tools all follow the same logic.

Here, however, is the uncomfortable truth Brussels must confront: the steel model cannot be generalised – for three key reasons.

First, the arithmetic of retaliation. Steel is one sector; a wall of steel-style instruments across a dozen sectors would invite a comprehensive Chinese response at a moment when European exporters – from luxury goods to aircraft to pharmaceuticals – remain deeply exposed to the Chinese market, and when Beijing has demonstrated, through its export controls on rare earths and its new supply chain security regulations, that it can weaponise Europe’s upstream dependencies with precision.

Second, the downstream cost. Every tariff-rate quota protects a producer by taxing a user. Steel’s users — automotive, construction, machinery — are already complaining, and steel is a relatively simple case where European production capacity largely exists. In solar panels, that capacity is gone; protection would raise the cost of the energy transition without much chance of reviving a European industry. In batteries and EVs, the calculus is genuinely contested. Multiplying such instruments means multiplying these trade-offs, and Europe’s fiscal and industrial capacity to compensate downstream sectors is finite.

Third, the WTO question. The steel instrument has been carefully drafted as a standalone regulation under the common commercial policy, precisely to escape the constraints of the safeguards framework. Doing this once, for a sector where virtually every major economy has already erected barriers, may be tolerable to the trading system. Doing it serially would confirm that the EU has abandoned the rules-based order it claims to defend — a reputational asset that still underpins its FTA agenda, from India to Mercosur.

If the EU cannot protect everything, it must choose. And this is where the real danger lies. The political economy of trade defence in Europe has always favoured sectors with concentrated production, strong federations and sympathetic member states. This is clearly the case of the steel industry, with its 300,000 direct jobs and one of the most effective lobbies in Brussels.

And yet the sectors that will define European competitiveness over the next two decades – power electronics, electrolysers, advanced batteries, industrial robotics, the equipment stack underpinning AI and the energy transition – are more fragmented, newer, and far less audible in the Berlaymont.

The criterion for protection should therefore be forward-looking and ruthlessly strategic: Where does defending or building European capacity today generate the technologies, skills and supply chain positions Europe will need tomorrow? Where would surrender create irreversible dependence on a Chinese-controlled stack?

By that test, some legacy sectors will qualify – steel itself has a genuine claim through defence and infrastructure, which is partly why the July 1 instrument is defensible. But steel’s privileged treatment owes at least as much to lobbying power as to strategic foresight, and that is exactly the precedent the EU must not institutionalise.

If overcapacity policy becomes a queue in which the loudest incumbents are served first, Europe will spend its limited protective capital preserving the industries of the twentieth century while ceding those of the twenty-first.

The six-month review clause built into the steel regulation, and the Commission’s promise to assess extensions of product scope, make this an immediate question rather than a theoretical one.

But the test is not whether the instrument protects steel; it is whether the EU can build, around it, a disciplined doctrine for choosing what to protect — before the lobbies choose first.