Broken Window Fallacy

In plain terms

A shop window gets smashed. A bystander points out that this isn't all bad: the glazier now has work, and will spend the money at the butcher, who spends it at the baker. Money moves, jobs happen. Perhaps the vandal did the town a favour.

He didn't. The shopkeeper was going to spend that money anyway, on shoes, or a sign, or stock. The glazier's gain is the shoemaker's loss, and at the end of it the town has the same money and one fewer window. The activity is visible. The shoes that never got made are not.

The parable is Frédéric Bastiat's, from an 1850 essay whose title is the whole idea: That Which Is Seen, and That Which Is Not Seen.

Why it's fallacious

Every use of a resource has an alternative use. Judging a decision by its visible effects alone, without asking what the same money, time, or labour would have done otherwise, systematically overstates the benefit. The comparison isn't "this activity versus nothing." It's "this activity versus the next best thing."

The error is hard to see precisely because the counterfactual leaves no evidence. Construction crews after a storm are photographable. The extension nobody built, the hire nobody made, the savings nobody kept: those are invisible, and invisible things lose arguments to visible ones.

This puts it in the same family as survivorship bias and cherry picking. In all three, the decisive evidence is missing from view, and in all three the fix is to go looking for the part you can't see.

Canonical example

"The wildfire was devastating, but there's a silver lining for the local economy. Rebuilding will mean years of construction work, contracts for local firms, and hiring."

The construction work is real. So is the hiring. What the argument leaves out is that the money paying for it came from insurers, savings, and public budgets that had other destinations. Those homeowners are not better off for having replaced a house they already owned. The region has spent a great deal to arrive back where it started.

The same shape shows up whenever destruction gets described as stimulus: wars, storms, demolitions, obsolescence. It is a fallacy about accounting, not about sympathy. The rebuilding may well be necessary and good. It just isn't a net gain.

Counter-example (not a fallacy)

"In a deep recession with high unemployment and idle factories, a public spending programme can raise output, because the workers and capital it uses were not producing anything else. The alternative use of those resources was nothing."

This is not the broken window fallacy, and the distinction is the whole substance of a real disagreement among economists rather than a technicality.

Bastiat's parable assumes the glazier was already fully occupied, so employing him on the window means not employing him elsewhere. When resources are genuinely idle, that assumption fails: the counterfactual is not "something else gets built" but "nothing gets built." Whether an economy has that slack at a given moment is an empirical question, and reasonable people read the same data differently.

So the fallacy is committed by anyone who treats destruction as free stimulus without checking the counterfactual. It is not committed merely by arguing that spending can raise output. Invoking "broken window" as a conversation-ender against any spending argument skips exactly the question that matters.

The line: is the alternative use of these resources being examined, or assumed away? Either side can do the assuming.

How to fix it

If you've been linked here, ask what the resources would have done instead, and say so out loud in your argument. "This created jobs" is only half a claim. The full version is "this created jobs that would not otherwise have existed, because those workers and that money had no better use," and that second half is where the actual case lives. If you can defend it, the argument gets much stronger. If you can't, the honest move is to drop the economic-benefit framing and argue for the thing on its own merits. Rebuilding after a disaster does not need to be a net economic win to be worth doing.

If you're on the receiving end, ask the counterfactual question rather than naming the fallacy: "Where would that money have gone otherwise?" That opens the real discussion. Leading with "broken window fallacy" tends to shut it down, and as the counter-example above shows, sometimes the person is right and you would be the one skipping a step.