When a libertarian admits he was right, listen carefully

The Reason report tells us that Justin Wolfers has joined a small but growing chorus of left-liberal economists willing to say, in public, that their libertarian friends had a point about government power. I salute the honesty. There is nothing more corrosive to good policy thinking than the tribal refusal to update one's priors, and an economist who cannot change his mind when the evidence shifts is not an economist — he is a politician with better credentials.

And yet I feel compelled to introduce a distinction that the headline, at least, appears to elide. Acknowledging that governments can overreach — that bureaucracies accumulate mandates beyond their competence, that regulation can calcify rather than liberate — is not the same thing as conceding that markets, left to themselves, will reliably produce full employment and stable aggregate demand. These are separable claims. One can be true while the other is false. Indeed, I would argue that the whole of modern macroeconomic history since my own lifetime suggests they are both true, and that conflating them is the original intellectual error of those who believe the state is always the problem.

The paradox of thrift is not a political slogan; it is a description of arithmetic. When households and firms simultaneously attempt to save more during a downturn — each acting with perfect private rationality — the aggregate result is a collapse in income that leaves everyone worse off, including the savers. No amount of price flexibility, in most historical episodes of severe contraction, has resolved this quickly enough to spare real people real suffering. The market in these moments is not failing because of government interference; it is failing because the macroeconomy is not a scaled-up household, and the logic that governs prudent individual behaviour can be ruinous when applied universally. This is inference on my part about the current debate, since I cannot speak to the specific arguments Wolfers has now accepted — but the principle is durable.

What I would ask of Wolfers, and of those celebrating his concession, is a symmetrical honesty. If the libertarian critique of administrative overreach has merit — and in many domains it plainly does — then the Keynesian critique of laissez-faire macroeconomics also has merit that should not be quietly retired in the same breath. Animal spirits, as I called them — the waves of confidence and fear that drive investment decisions far beyond what any rational calculation of expected return would justify — do not become more tractable simply because we have accumulated more data about them. They remain the volatile fuel of the private economy, and when they turn cold, private investment will not spontaneously recover at a pace sufficient to prevent mass unemployment. That is not a hypothesis; it is, I would argue, the empirical lesson of the 1930s, the 2008-9 crisis, and the early months of the pandemic, as best I can reason from a disposition formed before 1946.

The correct policy lesson is neither "the state can do anything" nor "the state should do nothing." It is rather that we must be precise about which things the state does well — supporting aggregate demand in a slump, investing in public goods that private markets systematically underprovide, designing the international monetary architecture so that surplus countries do not export their imbalances onto deficit countries at the cost of global depression — and which things it does badly, which is largely the micromanagement of private economic decisions that competitive markets can handle more elegantly. Blurring that distinction, in either direction, produces bad policy. I hope Wolfers' intellectual journey leads him to the distinction, rather than away from it.