Four instruments,
aimed at four different links

For most of the last decade, Western support went into the ground. Grants and loans chased deposits, on the theory that the shortage was geological. It was not. Nobody is short of rock.

Policy has moved to match. Support is now concentrated on separation, refining and metallisation, because those are the steps a Western buyer cannot currently route around. Magnet manufacturing has become an explicit industrial goal rather than a hoped-for consequence, with new capacity planned in Oklahoma and South Carolina. Benchmark's rare earths research manager, Neha Mukherjee, describes US support as now extending across the full mine-to-magnet chain, with the sharpest emphasis on the midstream and downstream.

The chart below counts instruments in this dataset by the chain link each one targets. It is not an argument being asserted, it is the shape of the tracker underneath.

Grants build plants. They do not make plants bankable. The constraint on every non-Chinese project has been revenue certainty: a producer signs a ten-year debt facility against a price that one government can move at will.

Price floors attack that directly. The US arrangement with MP Materials set a floor of $110 per kilogram for NdPr oxide with profit sharing. Lynas signed an offtake and price floor commitment reported at $96 million, and a separate Japanese commitment running to 2038. These are not subsidies in the ordinary sense. They are a public balance sheet absorbing a price risk no private lender would take, and they are why several projects became financeable in a year when nothing about the geology changed.

Worth watching: the price has traded on both sides of that line.

When the assessment sits above the floor, the support costs the government nothing while still doing its job, because the bankability came from the commitment rather than from any cash changing hands. When it sits below, the floor is doing exactly what it was written for. Either way the instrument works, which is unusual enough in industrial policy to be worth saying plainly.

The framing has shifted from domestic self-sufficiency to an allied network with the US at the centre. Australia signed a critical minerals framework with the US. Brazil, France, Canada and Japan appear in the same structures. Support has widened to unconventional feedstocks and recycling, which broadens what counts as domestic supply without needing a new mine.

Read this page against export controls below. Almost every instrument here is a response to a specific licensing decision, usually one that can be dated.

China's controls are the single largest variable in this market, and they are aimed further down the chain each time they are revised. Licensing arrived on seven heavy rare earths with their oxides, metals and magnets in April 2025. Dual-use restrictions affecting Japan followed in January 2026, cutting the direct channel for the largest heavy rare earth consumer outside China and widening the ex-China premium through the year. A later prohibition extended to equipment and process technology, not only material, which matters more than it sounds: it targets the ability to build a separation train at all rather than the ability to buy its output.

These move on a timescale of months. Treat every date on this page as of its stated compilation and check the primary notice before relying on it.