In 2025 Pforzheim opened Europe’s first plant for recycling rare-earth magnets; the small but indispensable components inside the electric motors of nearly every EV, every offshore wind turbine, and every drone now in service. Capacity is 100 tons per year, against an EU demand of around 20,000 (Mkango Resources / Investegate₂₀₂₅).

That is 0.5 percent of what Europe needs today. By 2028, if everything goes to plan, the plant will scale to 750 tons. That covers 3.75 percent of today’s demand.

These are not victory numbers. They are the symptoms of a much larger gap between what Europe has decided it wants — strategic autonomy in critical minerals, an industrial base that can supply its own green and military transition — and what it has actually built. The shortfall is not technology, capital or geology. There are two things Europe must now invest in, with the seriousness it has so far reserved for other crises: the speed at which it can break ground on new industrial sites, and the people who still know how to build and run them at scale.

This essay is about what closing the gap actually requires. It can be closed. What closing it takes is more concrete than the current policy debate has yet made it.

China’s position was built on purpose

To understand the size of the task, start with the size of the dependence.

China today extracts roughly 60 percent of the world’s rare-earth ore. But the bottleneck is not in the ground — it is in what happens after. China refines over 90 percent of the world’s rare-earth metal supply, and produces over 90 percent of the permanent magnets that go into wind turbines, electric motors, drones, and precision-guided weapons. This is not a market Europe is behind in. It is a market Europe is barely present in (International Energy Agency₂₀₂₅).

That position did not happen by accident. It is the result of decades of deliberate state investment across the entire value chain — from mining and separation through refining, alloying and component manufacturing — paired with a willingness to absorb short-term commercial losses in exchange for long-term strategic position. China understood, earlier than anyone else, that the country controlling the supply of critical inputs has leverage over every country that depends on them.

Europe and Japan learned exactly what that leverage feels like in 2010, when China cut its rare-earth exports by 40 percent during a diplomatic dispute with Tokyo. The price of neodymium rose by more than 700 percent in a year. Japan’s industrial base – deeply dependent on rare-earth magnets for everything from hybrid vehicles to consumer electronics – found itself with its back against the wall (World Economic Forum₂₀₂₃).

What Japan did next is the most instructive part of the story.

Japan started in 2010. Fifteen years later, it is still not done.

Within months of the 2010 export shock, the Japanese government tabled a supplementary budget of 100 billion yen (roughly one billion euros at the time) for emergency investment in recycling capacity, alternative materials research, and strategic stockpiles. State subsidies followed for companies investing in rare-earth recovery infrastructure. Universities and corporate labs were directed at the problem from the top (World Economic Forum₂₀₂₃).

The engineering wins were real, and they came fast. A research team at Kyoto University developed a process that recovers 96 percent of the neodymium and 91 percent of the dysprosium from used magnets, at extraction efficiencies that until then had been considered industrially infeasibleInteresting Engineering_n.d.. Honda began recycling rare earths from end-of-life hybrid battery modules. When Tokyo hosted the 2020 Olympic Games, every one of the 5,000 medals was made from gold, silver, and copper recycled out of donated consumer electronics — over 80,000 tons of mobile phones and small devices reclaimed through a national collection effort, with none of the medal metal newly mined (World Economic Forum₂₀₂₃).

That is what a serious industrial response looks like. It is also what its limits look like.

Fifteen years on, with all of that investment, all of that research, and a level of state-private coordination Europe has yet to attempt since WWII, Japan today still imports between 63 and 65 percent of its rare earths from China. For the heavy rare earths required in the most advanced magnets — dysprosium, terbium — Japan’s dependence on Chinese supply remains 100 percent (Oxford Institute for Energy Studies₂₀₂₅).

Two lessons follow. First, building alternative supply capacity takes a decade and a half, not a budget cycle. The country that has not started yet is already behind. Second, recycling and urban mining are necessary parts of the answer, but they are not sufficient. To close the loop you also have to break ground — on new mines, new refineries, new separation plants, new magnet manufacturing. It is here that Europe’s specific bottleneck becomes visible.

A 27-month target meets a 10-to-20 year reality

In 2024 the EU’s Critical Raw Materials Act came into force. It sets ambitious targets: by 2030, Europe is to extract at least 10 percent of its own demand for strategic raw materials, process at least 40 percent and recycle at least 25 percent. To make this possible the law caps the permitting timeline for designated “strategic projects” at 27 months for mining and 15 months for processing and recycling (European Parliament and Council₂₀₂₄).

27 months. It is an ambitious number – until it meets the real one.

The typical mining project in Europe today takes 10 to 20 years to develop. Environmental impact assessments alone can run for four years. Appeals against already-issued permits add further delays that the new legislation does not directly address (Oxford Institute for Energy Studies₂₀₂₅). The 47 strategic projects designated by the European Commission in March 2025, under the CRMA’s accelerated pathway, are still in early-stage development (European Commission₂₀₂₅).

Now compare that with China. In 2025 the South China Morning Post reported on a Chinese critical-minerals processing facility that went from project decision to full operation in 10 months – through a level of coordination between provincial government, state finance and industrial operators that observers described as without modern precedent. Integrated permitting tracks. State capital available immediately. Political authority used to remove obstacles in real time (South China Morning Post₂₀₂₅).

Ten months versus ten to twenty years. That is not a difference in bureaucratic efficiency. It is a structural difference in whether you treat critical minerals as defence infrastructure or as a question of municipal land-use planning.

This is the first half of Europe’s bottleneck. The one that gets discussed. The second half almost never does.

What permits cannot do

Skills are the second bottleneck, and they show up at Pforzheim itself.

The plant runs on a magnet-recycling process called HPMS — hydrogen processing of magnet scrap — developed at the University of Birmingham. The technology is licensed in. Pforzheim is in Germany. The know-how is from the United Kingdom. It tells you that even Europe’s first piece of broken ground in rare-earth recycling depended on a research base that was built somewhere else, because the equivalent rare-earth metallurgy capacity no longer exists at the scale the continent needs (Mkango Resources / Investegate₂₀₂₅).

The same hollowing is visible across European heavy industry. Since the Antwerp summit in 2024, the European chemical industry alone has lost 101 industrial plants, 75,000 manufacturing jobs, and 70 billion euros of production capacity (data flagged by Velina Tchakarova). Every one of those job losses is a senior process engineer who does not get replaced, a line operator pipeline that thins out, an apprenticeship programme that stops running. INEOS chairman Sir Jim Ratcliffe described the wider picture without sentiment: Europe is systematically eroding its industrial foundations while other regions accelerate.

Skills cannot be legislated into existence. The 27-month permitting target in the CRMA does not, by itself, make a metallurgist. The 47 strategic projects do not, on their own, train a single processing operator. Refineries, separation plants, and magnet facilities all require workforces of senior chemists, material engineers, and process technicians who develop over decades – and who, once gone, take a generation to rebuild. Europe shed that pipeline over twenty years of de-industrialisation. Replacing it is the part of the work that no permitting reform can solve, and it is the part Europe has so far barely started.

You can permit a refinery in 18 months. You cannot, in 18 months, conjure the people who know how to run it at scale.

What it already looks like, where Europe is already building

Finland is showing one answer.

In Espoo, just outside Helsinki, a new chip ecosystem is being built. The complex represents over half a billion euros of investment, with capacity for millions of components per year – including circuits for artificial intelligence and quantum computing (Ilta-Sanomat_n.d.). The Finnish anti-drone company Sensofusion is preparing to launch a Finnish-built satellite in 2027, designed to detect attack drones from orbit (Ilta-Sanomat_n.d.). These are not strategies on paper. They are concrete capital commitments with concrete timelines (flagged by Sari Arho Havren).

What is striking about Finland is not the scale — it is the posture. The decisions are being taken as if industrial capacity in critical sectors were a strategic question, not a planning question. The result is that things are getting built.

That is what changing posture looks like in practice, and it has a recognisable shape.

It means treating strategic projects as defence infrastructure: dedicated permitting tracks, environmental review run in parallel rather than sequentially, appeals processes that are expedited rather than open-ended, and investment-backed targets for recycling and urban mining at the level of individual sites with individual deadlines. Pforzheim is one plant, doing 0.5 percent of what is required. Closing the loop properly (across magnets, batteries and electronics) needs the same political weight that the CRMA was supposed to provide, but applied site by site, with somebody politically accountable for each one.

It means creating a parallel skills programme with metallurgists, chemists, process engineers and line technicians. The kind of vocational and university pipeline that produces these workers in the numbers a continental industrial base actually needs, not the numbers a contracting one can absorb. This is the part of the agenda that requires national governments and member-state education systems to move at the same time as the EU does – and it is the part that, if it does not happen, will quietly defeat every permit reform that does.

It means recognising that the political weather has changed. The April 2026 US-EU critical minerals deal – explicitly designed to weaken China’s dominance over the supply of strategic materials – is the clearest signal yet that the political momentum to act is shifting ()Bloomberg₂₀₂₆). But signals are not facilities. Treaties are not refineries. Meetings are not magnets. The work that turns commitments into broken ground still has to be done.

Two arguments against, and why they do not hold

Two predictable counterarguments arise here.

The first is that this is the kind of state-led industrial planning that has failed in Europe before. Concorde. The European chip champions of the 1980s. Various Common Agricultural Policy detours. The track record, the argument goes, is not encouraging. Better to let markets allocate.

This argument misunderstands the alternative. Europe is already pursuing an industrial policy. It is the policy of remaining structurally dependent on Chinese rare-earth processing, on Chinese magnet manufacturing, and on a supply chain that can be tightened or cut at the political discretion of a strategic competitor. That, too, is a planning failure. It is just a planning failure that nobody has to take responsibility for, because no one ever decided it on the record. Choosing to build is the more honest version of the same decision.

The second argument is that even if everything were accelerated tomorrow, Europe would, like Japan, still be heavily dependent on China a decade from now. So why bother?

This is not a reason to delay. It is the strongest possible reason to start now and at scale. Japan started in 2010 and is not done in 2026. If Europe starts in 2026, it is not done in 2042. If Europe starts in 2032, it is not done in 2048. The cost of delay is not measured in budget cycles but in the strategic position of the European industrial base across the next two decades. The compound interest of inaction runs the wrong way (as Sander Tordoir has consistently argued).

What it takes

Let me return to where this began.

Pforzheim recycles 100 tons of rare-earth magnets a year. By 2028, plans are to increase the number to 750 tons per annum. EU demand today is 20,000 tons per annum. By any honest accounting, the plant covers 0.5 percent of what is needed today and 3.75 percent of what is needed in three years.

This is not a solution. It is a proof of concept with a single piece of broken ground that demonstrates the path is real. The technology to recover rare earths from scrap exists. The capital is available. The know-how, even if some of it had to be imported from Birmingham, is reachable. What is missing is the political infrastructure that translates strategic necessity into permitted sites and trained workforces, on the timescale that the situation demands.

Japan showed that it is possible to respond seriously to a critical-minerals shock — and that even with a serious response, building real capacity takes more than a decade. Europe under Covid showed that it can mobilise on a defence-policy timescale when it decides to. Finland is showing what that decision looks like when it is made early. The capability is not missing. The decision is.

That decision can still be made. The question for European policy is not whether it has the means to break ground and close the loop. It manifestly does. The question is whether it has the political clarity to use those means at the speed the situation requires, and the willingness to rebuild the human capacity, not just the legal framework, that closing the gap will take.

Pforzheim is one piece of broken ground. Europe needs the rest of the ground broken (and the loop closed) and it needs the people to do both. That is what it takes.