Imagine you run a restaurant chain and your biggest competitor owns some of the farms that supply your ingredients. You could force them to sell those farms — but if they also own the only industrial kitchen big enough to process the raw harvest into usable product, you haven't solved your supply problem. You've just made the kitchen owner's life slightly harder while the food still flows through them. That's the core mechanism in this paper about lithium. Kei Irizawa builds a structural model of the global lithium supply chain — mines on one side, refineries on the other — where cross-border ownership changes the game in three specific ways: it lowers the effective ore cost for vertically integrated refiners, it incentivizes mines to restrict spot-market sales, and it softens competitive pressure. The model is estimated on global data covering both mines and refineries, and then used to simulate two national-security-motivated policies: forced divestiture of Chinese ownership stakes in overseas lithium producers, and an outright ban on ore exports to China. The headline finding is brutally asymmetric. Divestiture — the policy Western governments are most actively pursuing — increases spot ore supply (because mines lose their incentive to favor the integrated owner) but leaves total lithium output nearly unchanged. The ore just reshuffles among buyers. Chinese refineries still purchase on the open market, so continued trade preserves Chinese access to the mineral. The ownership transfer does not 'reserve' ore for domestic production in any meaningful sense. The export ban scenario is where real damage shows up. Cutting off Chinese refineries from ore supply reduces global lithium output because existing refining capacity outside China simply cannot absorb the displaced volume. The bottleneck isn't in the ground — it's in the processing. This is the paper's sharpest structural insight: the binding constraint in the lithium chain is refinery capacity, not mine ownership. Policies targeting ownership misidentify the chokepoint. Methodologically, this is an industrial organization paper with a well-specified structural model — demand and supply sides, vertical relationships, and ownership as a parameter affecting transaction costs. The counterfactual simulations are model-driven, not reduced-form, which means the results depend on the model's assumptions about how ownership affects costs and competition. The author is transparent about this, but it also means the quantitative magnitudes carry structural model uncertainty on top of data uncertainty. The broader implication cuts against the grain of current critical-minerals policy in the US, EU, and Australia. If forced divestiture doesn't secure supply and export bans actively shrink it, the policy logic needs to shift from 'who owns the mines' to 'who has the refining capacity.' Building domestic or allied refining capacity — a much slower, more capital-intensive project — is the actual path to supply resilience, and this paper provides formal economic grounding for that argument. What makes this paper worth reading isn't a surprising technical method but a clean structural argument deployed at exactly the moment policymakers need it. The lithium chain is the test case, but the logic generalizes to any critical mineral where refining capacity is concentrated and ownership policy is being used as a substitute for industrial capacity policy.
