When China tightened export licences on seven rare-earth elements in April 2025, Western governments responded in the same way as they have to every other critical-minerals shock of the past five years. The European Union accelerated the issuing of extraction permits under the Critical Raw Materials Act. The United States expanded subsidies for domestic refining of the minerals under the Inflation Reduction Act. The Minerals Security Partnership, an initiative launched in 2022 by 14 economies, brought in more producing partner countries.
Each response was based on the same diagnosis: that the mineral supply problem is geological and geographical, and that the solution is to extract more, refine more and source from more places. That diagnosis is only half right. It misses what is now the binding constraint: intellectual property.
How the rush for critical minerals is neglecting human needs
For the most strategically important materials — heavy rare-earth elements, gallium, germanium, silicon, lithium and graphite — the supply chain choke point is no longer at the mine, or even at the refinery. It is in the intellectual property that governs how the raw material is transformed into something useful1.
The raw mineral and the patent for processing it are often owned by different parties, frequently in different jurisdictions and increasingly in nations with trade relations that have moved from competitive to adversarial.
For example, a handful of Chinese, Japanese, South Korean and US firms hold patents for key processes: magnetic separation for rare-earth oxides; processing of graphite for battery anodes; nickel manganese cobalt and lithium iron phosphate cathode chemistries; processes for growing gallium nitride crystals for electronics; and purifying semiconductor-grade silicon.
The result of such a concentrated structure of patent holding is a market failure that geographical diversification cannot fix. Australia can mine all the heavy rare-earth elements it wants, but without access to the magnetic-separation intellectual property, the country cannot produce a usable permanent magnet. And Indonesia can refine as much nickel as it likes, but without the cathode-chemistry patents, it cannot produce a battery cell.
A refinery built without access to intellectual property produces an input that the host country can’t use or sell. This is why the EU’s headline targets under the Critical Raw Materials Act — 10% domestic extraction and 40% domestic processing by 2030 — are necessary, but insufficient. The Inflation Reduction Act; the US-government-led Project Vault, a partnership to set up a strategic reserve of critical minerals; the Minerals Security Partnership; and Japan’s and South Korea’s stockpile programmes have the same flaw.
The political response — to subsidize the parts of the supply chain that are not protected by intellectual property — produces refineries that cannot deliver finished products, but that have a considerable fiscal cost.
The economics were laid out in 1962 by US economist Kenneth Arrow, who showed that technological knowledge is a partial public good that is costly to produce but cheap to copy2. Patent law redresses this imbalance by granting time-limited rights for exclusive use3. The unintended consequence is that when two assets are essential for production, and they are held by different parties under different legal regimes, markets cannot combine them without help4.
Three policy instruments, used together, would address this flaw. They are not theoretical proposals. Each has an institutional precedent and a developed economic logic. What is missing is the political will to combine them.
The patent pool
The first is a multilateral patent pool. The Medicines Patent Pool, established in 2010 with support from global health initiative Unitaid, and based in Geneva, Switzerland, is one precedent5. This organization has expanded access to antiretroviral and tuberculosis therapies across low- and middle-income countries by aggregating patents from multiple pharmaceutical companies into one clearinghouse, rather than leaving agreements over usage to fragmented private negotiations. Its institutional template — voluntary licensing agreements between patent holders and a neutral administrator, with defined geographical and product scope — is transferable to other sectors.
For critical minerals, a similar patent pool could bring together process patents from willing licence holders; to start with, these might include laboratories and firms from geopolitically aligned nations, such as Japan, South Korea, the United States and those in Europe. The pool would make the patents available to refineries outside the dominant jurisdiction under defined terms, including prices and exclusion clauses related to the markets that the final product is destined for.
Climate change is devastating mining of minerals needed to fight it
This would build on bilateral collaborations that are beginning to emerge. For example, in October 2025, the US firm REalloys and the Japan Organization for Metals and Energy Security, a government agency, concluded a memorandum of understanding. It supports the license and transfer of Japanese magnet-fabrication technologies to REalloys facilities in North America, for defence, electric-vehicle and semiconductor applications.
The main objection to patent pools is political rather than economic. Historically, bargaining failures between members have put a stop to such pools. For example, despite broad institutional support, the World Health Organization’s COVID-19 Technology Access Pool6 attracted no contributions of vaccine patents from major pharmaceutical firms after each anticipated that the others would not take part — a classic failure of coordination.
But principles of game theory can be applied to overcome such problems. In 1953, US economist Lloyd Shapley proposed a cooperative-game framework (for which he won a Nobel prize in 2012 with fellow economist Alvin Roth) setting out a mathematically fair way to distribute the total payout or cost of a collaborative effort among participants7,8. Allocations are apportioned according to the value that each individual adds to each possible grouping of players.
Gallium is used to make semiconductors produced at a factory in Austria.Credit: APA-Images/Alamy
The methodology is robust: according to the theory, no subset of pool members can improve their position by defecting, and the more high-value patents that the pool aggregates, the more stable it becomes because the cost of deviating from the arrangement is high. The result is a virtuous dynamic: early participation creates incentives for further participation.
Versions of the Shapley framework have been deployed in the real world: for matching people for kidney exchange programmes, allocating parts of the electromagnetic spectrum for telecommunications, sharing the costs of infrastructure and matching resident physicians with hospitals. It should be applied now to critical-minerals intellectual property, as a solution to the objection that players don’t know how to share the gains.
The legal backstop
The second instrument is compulsory licensing as a credible threat. Article 31 of the World Trade Organization (WTO) Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) permits compulsory licensing of patented technologies under specified conditions, including national emergencies and dependency on imports that are controlled by a foreign jurisdiction, the conduct of which affects the licensing market.
The 2001 Doha Declaration on the TRIPS agreement and public health clarified that public-health emergencies — such as HIV and tuberculosis outbreaks — qualify as the type of crisis that Article 31 was designed to address. For instance, in 2006, Thailand invoked the mechanism to override a patent held by pharmaceutical company Merck for the drug efavirenz, and imported a generic version from India at half the price to treat an extra 20,000 people with HIV9. In response, Merck offered voluntary price reductions.
Metals are key to the global economy — but three challenges threaten supply chains
No comparable clarification exists for industrial supply security, despite the structural parallel. But a coordinated declaration by major consuming economies — analogous to the Doha Declaration in scope, but to address critical-minerals dependency — could establish the basis for compulsory licensing when the patent holder refuses to participate in the pool. As with the Doha clarification, a sovereign state could issue a compulsory licence under domestic law to bypass the patent owner’s consent.
This is not a step to be taken lightly, however. Any clarification would require careful drafting and criteria for invoking it. But its availability would give the voluntary patent pool its negotiating leverage.
Without the backstop, patent holders might decline to join while continuing to extract benefits from the locked-in market. With the backstop, the choice becomes: participate on the terms determined by the Shapley framework, or face compulsory licensing on terms that are determined by national emergency regulation. That asymmetry can turn a stalled negotiation into a productive one.
The demand commitment
The third instrument is for governments to place conditions on procurements, requiring suppliers to work with the patent pool. Governments are among the largest buyers of advanced technologies that depend on critical minerals — through defence, electrification, grid-scale energy storage and industrial-strategy programmes. Ensuring that procurement of materials is managed through the patent pool would help to bring private licence holders to the table voluntarily.
