I want to be careful here, because this argument is regularly made badly, by people who use anti-technocratic language as a cover for anti-expertise sentiment, or who conflate the legitimate critique of insulated decision-making with the illegitimate claim that experts should simply be ignored. I am not making those arguments. I am making a different and I think more uncomfortable one.

The technocratic turn in liberal governance over the past forty years was not an accident or a conspiracy. It was a response to a real problem: democratic publics are susceptible to demagoguery, short-termism, and motivated reasoning in ways that make them reliably bad at certain kinds of decisions — decisions that require sacrificing present consumption for future benefit, decisions that involve complex technical trade-offs, decisions whose costs are concentrated and whose benefits are diffuse. The solution seemed obvious: take those decisions out of the hands of politicians who respond to electoral incentives and give them to independent institutions staffed by trained experts who can take the long view.

The problem is that this solution generates its own pathology. When consequential decisions about interest rates, drug approvals, pandemic restrictions, and climate targets are made by bodies that are insulated from democratic accountability, and when those decisions impose real costs on real people who had no voice in making them, the resulting resentment does not disappear. It accumulates. And when it eventually finds political expression — which it always does, because democratic energy does not simply vanish when it is excluded from official channels — it tends to find expression in forms that are hostile not just to the specific technocratic institutions that generated the grievance but to expertise itself.

This is not a counsel of despair about democracy. It is an argument that the design of democratic institutions matters enormously, and that the particular design choice of removing contested decisions from democratic deliberation has costs that were not adequately weighed when it was made. The alternative is not to abandon expertise or to pretend that all decisions are equally accessible to lay judgment. It is to build institutions that combine technical competence with genuine accountability — that explain their reasoning in public, that accept challenge, that acknowledge uncertainty, and that treat the people affected by their decisions as participants rather than subjects.

This is harder than either pure technocracy or pure populism. It requires holding two things simultaneously: that expertise is real and matters, and that legitimacy in a democracy ultimately derives from the consent of the governed rather than the credentials of the governors. We have not found a stable way to hold both. Finding it is, I think, the central institutional challenge of the next generation of democratic politics.