The Return of Industrial Policy

Industrial policy was supposed to be a relic. The consensus that formed after the Cold War held that governments were bad at picking winners, that market allocation of capital was more efficient than state direction, and that comparative advantage would distribute manufacturing globally in ways that benefited all participants. That consensus has fractured, and the fracturing is not primarily ideological — it is empirical.

The semiconductor supply chain crisis of 2021 demonstrated that geographic concentration of production for critical inputs creates strategic vulnerabilities that price efficiency does not account for. Taiwan produces the overwhelming majority of the world's advanced chips. This was a market outcome — TSMC competed effectively and captured the business — but the result is a single-point-of-failure in global electronics manufacturing that no one deliberately designed and no one has a clean plan to address.

The responses have been significant. The US CHIPS Act, EU Chips Act, and parallel initiatives in Japan, India, and South Korea represent the largest coordinated government investment in manufacturing capacity since postwar reconstruction. The stated rationale is security; the underlying logic is that certain industries cannot be allowed to optimize purely for cost because the non-cost factors — resilience, strategic availability, domestic employment — have a value that markets do not price.

Whether this works is a separate question from whether it represents a genuine policy shift. It does represent a genuine shift. The economist who argues in 2026 that industrial policy is categorically inferior to market allocation is arguing against the revealed preferences of every major economy simultaneously. The argument may still be correct in some technical sense; it has lost the political and institutional ground where it used to stand.

The results will take a decade to assess. The investment is already made.