The Rare Earth Problem Washington Doesn't Want to Solve

The United States has a rare earth problem, but it is not the problem Washington is trying to solve (see this and this). The debate in Congress has centered on supply and on the fact that China dominates the processing of rare earth elements and that this dominance creates a vulnerability. The proposed solution, Project Vault, is a $10 billion government stockpile financed through the U.S. Export-Import Bank. This is not a military stockpile. It is a civilian stockpile. And it is the wrong answer to the wrong question.

The right question is not where America will get rare earths. The right question is why America cannot process the rare earths it already has. In almost every case across the critical minerals list, the solution is modest investment in actual production capacity, not a stockpile. A stockpile does not increase total supply. It does not create new production or production capacity. It buffers against an interruption for a defined period, after which time the same dependency reasserts itself exactly as before.

Rare earths are not rare

The term “rare earths” is misleading. Rare earth elements are neither rare nor, technically, earths. They are found in commercially viable concentrations on every inhabited continent. The United States has domestic deposits. Brazil is a producer. Australia has multiple projects in development. The geological assumption that heavy rare earths, which are used in electric vehicles, some military applications, and consumer electronics, were uniquely concentrated in China has turned out to be wrong. Ionic clay deposits of the kind that produce heavy rare earths form wherever granite has weathered in subtropical climates, which describes large parts of the world outside China.

China’s dominance is not a geological accident. It is a processing dominance. The hard part of rare earth supply is not mining; it is separating a mixed concentrate into individual elements. That separation process costs roughly $20 per kilogram of material outside China and requires a capital investment of around a billion dollars for a commercial-scale facility. China built that infrastructure over decades, and now performs roughly 85 percent of global separation capacity.

That is the actual chokepoint, and it has nothing to do with where the ore is in the ground.

The materials are already here, and we cannot touch them

To understand why America’s rare-earth problem is really a regulatory problem, it helps to understand what naturally occurring radioactive materials (NORM) actually are. Every piece of rock on earth contains trace amounts of every element, including radioactive ones. A mineral deposit is simply a place where a particular element is more concentrated than usual. Most of the time, the radioactive trace elements in a deposit stay with the waste rock when the ore is extracted, get dumped back in the hole, and nobody thinks about them again.

Rare earths are different. When you process rare earth ores, the naturally occurring thorium in the rock tends to concentrate alongside the rare earth elements you actually want. The processing step that gives you usable rare earth material also gives you concentrated thorium, which is radioactive. Concentrated radioactivity is, rightly, subject to strict regulation. The question is not whether to regulate it. The question is whether the regulations are sensible.

In the United States, they are not.

After numerous exchanges with Tim Worstall, a former metals trader and commodities analyst who spent decades working in rare earth markets with the enormous patience (thank you, Tim), here’s what I have come to understand. One of the most important rare earth feedstocks in the world is monazite, a mineral that occurs as a byproduct of common mining operations. When companies mine mineral sands for ilmenite and rutile, the feedstocks for titanium production, an industry operating at millions of tons per year, produce monazite as a leftover. Monazite is loaded with rare-earth elements. Because processing involves handling concentrated thorium, and because American NORM regulations make that legally treacherous for any company without specific legacy authorizations that almost no one has, the monazite either gets sold to China for a fraction of its value (because there is no other market) or is simply stockpiled on site indefinitely.

How stupid is that? A mineral containing exactly what America says it urgently needs sits in warehouses and waste piles across the country, made worthless by the regulations.

Nowhere is this problem more vivid than in central Florida, which has been a center of phosphate mining for over a century. That mining produces, as a byproduct, enormous gypsum stacks (waste piles containing monazite, rare earth elements, and other valuable materials). These stacks are environmental liabilities. They sit on the landscape, they require ongoing management, and they represent a decades-long accumulation of material that no one has been able to process commercially. The technology to process them exists. The chemistry is understood. Processing them would accomplish two things simultaneously: it would remediate an environmental problem Florida has been managing for generations, and it would produce domestic rare earth feedstock on a meaningful scale. The uranium, thorium, and radium present in the material would need to be extracted as part of the process, which is not a problem but a feature, since that extraction is precisely what makes the remaining material clean enough to use.

Back when he was in the Senate, Secretary of State Marco Rubio introduced legislation known as the Thorium Act specifically to address this barrier. The premise was straightforward: align American NORM regulations with the reasonable international standards that other countries already use, rather than leaving them subject to legal challenge from environmental groups objecting to low radioactivity levels. Under international standards, handling concentrated thorium from rare earth processing is a routine industrial activity. Under current American law, it is effectively prohibited for anyone to do without legacy authorizations from the nuclear industry that almost no operating company possesses.

Rubio’s legislation did not pass. The feedstock sits in warehouses and on waste piles across the country.

The separation problem

Even if domestic processing of monazite and other feedstocks were unlocked by regulatory reform, the separation problem would remain. Mixed rare earth concentrate still needs to be separated into individual elements, and that separation process is where Chinese dominance is most entrenched, thanks to a 30-year head start.

The up to $20-per-kilogram separation cost outside China is no law of nature, though. It is the cost of current technology, applied under Western regulatory and labor conditions. As Worstall noted to me, and in this piece, the question worth asking is not how to build in the US what China already has, but whether better approaches exist that would change the underlying economics entirely. That is how competitive advantages are achieved in every industry: not by catching up on the incumbent’s terms, but by changing the terms.

Another way that Worstall puts is this:

“No, this is not an argument that we should have subsidized for 40 years to maintain production. It’s going to be vastly cheaper to build new now than it would have been to carry deadbeats for decades. Quite apart from anything else, we’re going to build our new stuff at the edge of the current technological envelope — not just shiny but modern.”

Worstall argues that that’s achieved by funding the basic science that changes what separation costs in the first place. He has identified two approaches worth investigating. The first is vacuum distillation. That’s the same basic principle used in titanium production where the Kroll process achieves separation costs of around 40 cents per kilogram. Whether a variant can be made to work for rare earth elements at economically viable temperatures is an open research question, but Worstall believes the answer is yes and has staked his own research efforts on it. The second is adapting solvent extraction/electro-winning processes developed for uranium processing, technology that was itself adapted to copper processing in the 1980s, to rare earth separation.

Neither approach is guaranteed to work. But the cost of finding out is measured in millions of dollars, not billions. The basic research investment that DARPA or the Department of Energy’s National Laboratories could fund easily would be sufficient to determine whether either is viable. If either works, the cost of rare earth separation could fall by 90 percent, permanently changing the economics of non-Chinese processing everywhere in the world.

This quickly tests the limits of my technical understanding. But what I do know is that whether I like it or not, this is precisely the kind of public-goods investment that some economists identify as a legitimate role for government: basic research the benefits of which cannot be fully captured by any single private investor, and that therefore tends to be underfunded by markets alone. What is not a proper role of government is whatever the administration is concocting with Project Vault. The investment in basic research suggested by Worstall is also extraordinarily cheap relative to the alternative. A few tens of millions of dollars in research funding versus $10 billion in a stockpile managed by an institution with a 90-year record of directing capital toward the politically connected.

More Problems with Project Vault Stockpile:

There are issues with the Project Vault stockpile that go beyond its monetary cost and reliance on Ex-Im Bank. Modest military stockpiling and price floors that keep Western producers afloat through Chinese-induced price troughs are defensible uses of government policy. However, a civilian stockpile does nothing to address China’s processing and separation dominance. A stockpile may buffer against a supply interruption for a period, the same dependency reasserts itself afterward. If the stockpile contains unprocessed rare-earth concentrate, it is even more useless, since the US still lacks the domestic capacity to convert that concentrate into usable materials.

Further, in the case of Project Vault, when the government agrees to purchase at preferential prices from certain suppliers, it removes price-sensitive demand from the market. This lowers the market price everyone else receives, directly deterring investment in developing new sources. The stockpile, designed to create supply diversity, actually works against it by making the economics of new projects worse.

Also worth noting, sophisticated private buyers are already building 6–12 months of safety stock without any government mandate, at no cost to taxpayers.

The bottom line:

The case for government action on rare earths is not obviously wrong. There are genuine national-security considerations, genuine supply-chain vulnerabilities, and real public goods to be achieved, particularly in basic research, that might justify some government response. The question is not whether the government should act but whether it is acting intelligently.

An intelligent response starts with the regulatory changes that cost nothing. Rationalizing NORM regulations to reflect the actual risk profile of naturally occurring radioactive materials, rather than a Cold War-era framework built around weapons-grade nuclear material, would immediately enable domestic feedstock processing that is currently prohibited. Streamlining permitting for rare earth mining and processing facilities would reduce the decade-long timelines that deter private investment. Clarifying the legal pathway for processing monazite and other thorium-containing materials would allow the private sector to develop the supply chains the government is currently trying to build with public money.

A second step could be modest funding for basic research on separation technology. Not a billion-dollar plant. Not a government-owned processing facility. A few research grants to scientists working on alternative separation approaches, with clear milestones and accountability, of the kind DARPA has successfully funded in other domains.

For other critical minerals, the approach is the same: grant the licenses already waiting on government desks, support the modest production investments that private companies are already pursuing, and get out of the way.

What is not needed is a $10 billion stockpile routed through the Export-Import Bank – an institution structured to benefit the likes of Boeing, and a handful of commodity trading firms, and managed by an institution with a terrible track record. The stockpile will not solve the separation problem. It will not rationalize NORM regulation. It will not fund the research that could permanently reduce the cost of non-Chinese processing. It will not change any of the underlying conditions that create the vulnerability it purports to address.