There is no clever way around a shortage

The National Review piece opens with a sentence worth sitting with: there is no clever way around a shortage. That is, of course, precisely right — and it is a sentence that cuts against a great deal of fashionable thinking on both sides of the Atlantic. The market, we are perpetually assured, will find a way; prices will rise, investment will follow, supply will materialise. In the long run, no doubt. But I wrote somewhere that in the long run we are all dead, and the factories that depend on memory chips cannot wait for the long run.

The macroeconomic case here is not complicated, even if the semiconductor engineering is far beyond anything I could have imagined in my own time. Memory chips are what economists would now call a strategic complement — an input without which vast swathes of downstream production simply stop. A shortage is therefore not merely a sectoral inconvenience; it is a potential aggregate demand event. Supply chains seize, investment plans are shelved, hiring freezes follow. The animal spirits of business — already nervous creatures — take fright at uncertainty and withdraw. What begins as a materials problem becomes, with sufficient severity, a macroeconomic one.

The honest objection is this: government has no particular expertise in picking technologies, and the history of industrial policy is littered with expensive failures. I grant it freely. I was never an advocate for the state doing everything; I was an advocate for the state doing what private actors demonstrably will not. The question is not whether Washington possesses superior knowledge of chip architecture — it plainly does not — but whether the private market, left to itself, will invest at the speed and scale that a strategic dependency demands. The evidence from recent years, which I can only observe at one remove, suggests the answer is no.

What I would urge, speaking from disposition rather than from any recollection of events after 1946, is that the relevant frame is public investment as pump primer, not public investment as permanent substitute. The state identifies the gap, commits capital to close it, and then — critically — creates the conditions under which private firms compete to fill it sustainably. The error to avoid is the opposite of intervention: passive confidence that comparative advantage and price signals will, in their own good time, relocate strategically critical production. They may. But the interim cost, measured in lost output and employment, may be very large indeed.

There is also an international monetary dimension worth noting, even if I must tread carefully on modern institutional specifics I cannot claim to know well. A shortage that is geographically concentrated is not merely an engineering problem; it is an architecture problem. The Bretton Woods negotiations I was party to taught me that the world's productive interdependencies require deliberate public design — that leaving the architecture of trade and supply entirely to private convenience produces fragilities that eventually become everyone's crisis. Memory chips, by inference, are the present era's version of that lesson. The nations that understand this will sit down together and build something; the nations that do not will compete in panic when the next disruption arrives.

The practical policy, then, is not tariffs alone, not subsidy alone, and certainly not the comforting fiction that the market will sort it out. It is a sustained public commitment — coordinated where possible with allies — to ensure that no single chokepoint can hold a modern economy to ransom. That is not socialism. It is what I would call, with no apology, the minimum competence of a serious state.