A carbon price is the instrument economists have recommended for thirty years, on the reasoning that it leaves the choice of how to abate to whoever knows most about the options. The record of implementation is more mixed than the theory. Prices set through political negotiation have usually been too low to change investment decisions, while the sectors that respond most readily to a price signal — electricity generation above all — are also the ones that regulation handles competently. Where a price has clearly done work, it has generally been alongside standards and subsidies rather than instead of them. The instrument is sound; the assumption that adopting it would make the messier alternatives unnecessary has not held.