If Rare Earths Are Not Rare, Why Is There Still a Bottleneck?

Which minerals count as “critical”, and who decides?

Critical minerals are a policy category, not a chemical one. Each country writes its own list, usually applying three tests: the mineral is essential to a strategic industry, its supply is concentrated in a few sources, and it has no near-term substitute.

Typical lists include lithium, nickel, cobalt, copper, graphite, and the rare earth elements. The first five underpin batteries and grids; rare earths go into permanent-magnet motors, sensors, and missile guidance.

Lists change, and the changes are informative in themselves: a material is usually added because some industry has just discovered a dependency it had never accounted for.

Why are rare earths called rare? Are they?

They are not. Rare earth elements are reasonably abundant in the Earth’s crust; the more abundant ones occur at concentrations similar to common metals such as copper and lead. “Rare” refers to how seldom they form concentrated deposits; they are usually dispersed within other minerals.

The difficulty is therefore not finding them but separating them. The seventeen elements are so chemically similar that pulling them apart takes hundreds of solvent-extraction stages, a process that consumes large amounts of energy and acid and produces radioactive waste. The bottleneck is thus not geology but a willingness to bear the environmental cost and capital outlay of that step, which is a political decision rather than a technical one.

Where in the chain is the chokepoint?

The chain has at least four steps: mining, separation and refining, materials manufacture such as permanent magnets, and final assembly. The four differ sharply in how concentrated they are.

Mining is somewhat less concentrated: Australia, the United States, Brazil, and Southeast Asia all produce. The heavier concentration sits in the middle two steps. A country can own the ore and still have to ship it elsewhere for processing, then buy the finished material back.

This is a classic chokepoint, located in a process rather than at sea. It also explains why “we have our own deposits” is no guarantee of security: the ore is yours, but the capability is somebody else’s.

Why haven’t years of subsidies fixed it?

Three constraints bind at once. First, time: a separation plant often needs years to go from launch to steady output, and for a new plant the permitting stage is where the environmental fight takes place. Even reworking an existing plant is slow: MP Materials began upgrading the separation facility at its California mine in 2020, restarted separation in 2023 and does not expect to reach design capacity until the first quarter of 2027.

Second, price. The market is small and highly volatile, and a dominant supplier can let prices fall as new capacity approaches start-up, so the new plant loses viability before it opens. This requires nothing unlawful, only capacity and patience.

Third, process knowledge. Separation parameters, impurity control, and yield rest on decades of accumulated operating experience, which does not ship with the equipment; it is the same difficulty as yield in semiconductors. So at this step friend-shoring is not simply a matter of relocating a plant: somebody has to be willing to fund the unprofitable years.

What should a company actually inventory?

Trace the supply chain to the third and fourth tiers. Critical minerals almost never appear on a first-tier purchase list: you buy motors, sensors, or battery modules, and the minerals sit upstream of them.

Watch the items with a negligible share of cost. The typical single point of failure is a material worth under one per cent of cost with only two suppliers in the world, while procurement systems rank items by spend.

Push the question “do we have a second source?” all the way back to the common upstream supplier. Two suppliers may buy the same magnet material from the same producer, in which case the redundancy exists on the vendor list but not in the physical chain.