The geography of critical minerals is not the geography of industrial power. While critical minerals are naturally geographically dispersed, the capabilities required to process and refine them remain highly concentrated in a small number of industrial ecosystems.
For six key energy minerals, the average share of the world's three largest refining countries increased from roughly 82 percent in 2020 to 86 percent in 2024, even as governments increasingly sought to diversify their supply chains. The concentration is particularly striking because it occurs not at the point of extraction, but at the stage where raw materials are transformed into strategically usable inputs.
This highlights a defining industrial chokepoint of contemporary geoeconomics: processing. But the strategic importance of processing cannot be understood through capacity alone. Processing sits within industrial ecosystems built through accumulated expertise, tacit knowledge, specialized suppliers, infrastructure, technology, downstream demand, and sustained investment. The fundamental unit of modern industrial policy has therefore increasingly become the industrial ecosystem itself rather than the mine, refinery, or subsidy alone.
Processing Has Become a Defining Chokepoint
Rare earths make the distinction clear. China accounted for around 60 percent of global mining of magnet rare earths in 2024, but 91 percent of refined output and 94 percent of permanent-magnet production. Natural resources create strategic potential, but processing transforms that potential into industrial capability. Ownership alone is therefore not enough. Those who control processing wield leverage even over resources they do not themselves possess.
That distinction matters because critical-mineral processing feeds technologies central to energy systems, advanced manufacturing, semiconductors, aerospace and defense, and the digital economy. Minerals acquire much of their strategic value only after successive stages of separation, purification, refining, metallization, or other forms of processing turn them into specification-grade inputs. A new mine may diversify where material originates without meaningfully diversifying where it becomes useful.
China's 2025 export controls illustrate how industrial concentration can be transformed into an instrument of economic statecraft without requiring a permanent embargo. Restrictions imposed upstream rapidly propagated throughout global manufacturing. Among the many effects these controls had on industry were forcing some automakers to reduce production or even halt operations altogether, jeopardizing an estimated $6.5 trillion in annual downstream production outside China.
The strategic importance of these markets can be underestimated because many are relatively small in economic value. Yet the International Energy Agency (IEA) notes that disruptions to even small critical-mineral markets can have disproportionately large consequences for downstream industries. This vulnerability is amplified by the breadth of China's processing position: it is the leading refiner of 19 of 20 energy-related strategic minerals, averaging roughly 70 percent market share across each.
The broader vulnerability becomes clearer when assessed through an N-1 stress test, which, in this case, asks whether global supply could meet demand if the largest supplier and its domestic demand were removed from the market. Despite adequate aggregate supply, the answer for 2035 is striking: remaining supply would cover only 35–40% of remaining demand for graphite and rare earths, less than 55% for nickel, and around 65% for lithium and cobalt. Concentration therefore creates a geopolitical vulnerability that headline measures of global supply obscure.
Processing Power Persists
If the concentration were simply a matter of constructing additional refineries, the policy attention devoted to critical minerals thus far should already be producing a more geographically diversified processing base. Instead, refining has become more concentrated. Nearly 90 percent of refining supply growth between 2020 and 2024 came from the existing largest supplier for nearly all critical minerals.
The explanation lies partly in capability accumulation. Industrial capabilities are acquired cumulatively through investment, production experience, technological learning, organizational knowledge, and skill development. Industrialization involves more than purchasing equipment. Firms and economies must develop the capabilities required to use, adapt, and improve technology.
Yet not all of these capabilities can simply be purchased or transferred. Much of the knowledge embedded within mature industrial ecosystems is tacit. It is accumulated through production, organizational routines, specialized labor, supplier relationships, and repeated problem-solving. This makes industrial capability partly local and experiential rather than fully codifiable and transferable.
These capabilities are also path-dependent. Industrial development builds upon what already exists. Evolutionary economic geography has long emphasized how historical processes and accumulated knowledge shape where economic activity develops and persists. Once specialized labor, suppliers, engineering expertise, infrastructure, customers, and technical knowledge cluster around an industry and area, each can reinforce the others and produce cumulative advantages that become increasingly difficult for new entrants to replicate.
Processing is better understood as an ecosystem than as infrastructure. A government can finance a refinery relatively quickly. Reproducing the network of capabilities surrounding an incumbent processor is significantly harder. This helps explain why new capacity can exist on paper yet struggle to compete commercially and why moving extraction or final assembly does not eliminate the chokepoints between them.
Industrial Policy Catches Up
Previous phases of critical-mineral policy treated dependence as a sourcing problem: find another deposit, secure another supplier, finance another mine, or build another processing facility. Such narrow measures diversify individual links in a supply chain, but they do not reproduce the capabilities that make the full chain function. A mine whose output must still be refined abroad, or a refinery dependent on foreign technology, expertise, intermediate inputs, and downstream customers, can relocate activity without eliminating the underlying chokepoint.
Industrial policy appears to be catching up with this reality.
The European Union's Critical Raw Materials Act seeks, by 2030, extraction capacity equivalent to at least 10 percent of annual consumption, processing capacity of at least 40 percent, and recycling capacity of at least 25 percent. It also aims to prevent more than 65 percent of annual needs for a strategic raw material at any relevant processing stage from originating in a single third country.
The United States is similarly targeting capabilities across the production chain. Department of Energy initiatives now encompass rare-earth extraction and separation, gallium and germanium refining and alloying, lithium separation and processing, recovery from waste streams, and integrated rare-earth demonstration facilities.
Increasingly, however, industrial policy also reaches beyond financing physical capacity. The 2025 partnership between the U.S. Department of Defense and MP Materials paired public support with long-term demand commitments intended to support an integrated domestic chain extending into rare-earth magnet manufacturing. Similar logic is apparent in corporate long-term offtake agreements, recycling partnerships, strategic stockpiles, and allied efforts to coordinate investment.
These interventions reflect an important evolution in the industrial-policy problem. Building a plant is insufficient if it lacks skilled workers, reliable feedstock, processing know-how, downstream customers, supporting suppliers, and the commercial resilience to survive price competition from established producers. Effective diversification requires policymakers to think beyond capacity toward capability, and beyond individual facilities toward the industrial ecosystems that allow those capabilities to become commercially self-sustaining.
From Capacity to Capability
The next phase of the critical-minerals contest is unlikely to overturn the geography of processing power overnight. Despite proliferating subsidies, partnerships, and project announcements, incumbent ecosystems retain advantages accumulated over decades. New mines will open sooner than new processing networks can mature. Stockpiling and friend-shoring can mitigate vulnerabilities in the meantime.
The more meaningful measure of industrial resilience will not be how many projects governments announce, but whether alternative ecosystems can be built, maintained, and sustained, as evidenced by whether they can repeatedly transform raw materials into commercially viable industrial products at scale.
Here lies the deeper lesson of today’s processing chokepoints. Resources have always mattered, but possessing them is not the same as possessing the capabilities that make them useful and unlock their strategic value. As governments rediscover industrial policy, their success will increasingly depend on recognizing that industrial power resides not in any single mine, refinery, or subsidy, but in the ecosystems that connect and sustain them.
This creates a consequential transition period for the near-term future. Even if today's diversification strategies ultimately succeed, states attempting to build alternative ecosystems will remain dependent on many of the processors they are now trying to reduce their dependence on. Incumbent suppliers will continue to retain considerable leverage even as their customers undertake the slow work of reproducing the capabilities that created that leverage in the first place.
For the foreseeable future, the central contest in industrial policy will not simply be over access to critical resources, but rather the capabilities that transform those resources into enduring geopolitical and economic power.
