If the trouble with the smoke is that the factory does not pay for it, the most direct remedy is to make it pay — to lay upon each ton of smoke a charge equal to the damage that ton does, so that the cost the factory had shifted onto its neighbors returns to its own ledger. This is the corrective tax, the cure Arthur Pigou proposed, and it works by mending the price itself.
The logic is exact. The factory overproduces because its private cost falls short of the social cost by the amount of the external harm. Impose a tax equal to that harm — a tax per unit of smoke set at the damage per unit — and the factory's private cost rises to meet the social cost precisely. Now, when the factory weighs cost against price, it weighs the true, whole cost, the smoke included; and the quantity it chooses, maximizing its own profit as before, is exactly the quantity that is efficient for society, the one at which the goods are worth what, all costs counted, they truly cost. The tax does not forbid the smoke or cap it; it prices it, and lets the factory decide — but the factory now decides facing the real cost, and so decides rightly. The externality has been internalized: the outside cost brought inside the price.1
There is an elegance to the Pigovian tax that sets it apart from taxes in general, and it is worth dwelling on. Most taxes are a drag on efficiency: tax labor and people work less than they would; tax trade and some good trades do not happen; the tax drives a wedge between what is worth doing and what is done. The Pigovian tax is the rare reverse — a tax that improves efficiency, because the activity it discourages was being overdone to begin with. It corrects rather than distorts. And it offers, some argue, a double dividend: the harm is reduced, and the revenue raised can replace some other, distorting tax, so that society gains twice — cleaner air and a less burdensome way of funding the state. A tax on what we wish less of, in place of a tax on what we wish more of: this is the case for taxing pollution rather than payrolls, carbon rather than capital.2
The Pigovian idea, abstract in 1920, is now written across the statute books of the world, wherever a government has tried to price a harm. The carbon tax is a Pigovian tax on the most consequential externality of all, the warming that the burning of fuel inflicts on every future inhabitant of the earth. The congestion charge is a Pigovian tax on the delay each driver entering a crowded city imposes on all the others. The duties on tobacco and alcohol are, in part, Pigovian — charges meant to make the drinker or smoker bear some of the cost he visits on a health system and a family. In each, the state estimates a harm, attaches a price to it, and lets the market adjust; and in each, the hope is the same as Pigou's: that a price corrected by law will steer private choices toward the social good without the state having to dictate each choice directly.3
The difficulty, and it is a deep one, is that the tax must equal the harm, and the harm is rarely known. What, exactly, is the cost of a ton of carbon — across centuries, across nations, across the uncertain feedbacks of a planet's climate? What is the damage of an hour's smog, the toll of a decibel, the true social cost of a drink? To set the tax rightly, the state must measure what is by its nature diffuse, contested, and uncertain — and a tax set too high discourages too much, a tax set too low too little. The information the Pigovian tax requires is exactly the information that the externality, by scattering its costs across unconsenting multitudes, makes hardest to gather. The tax is right in principle and treacherous in practice, for the number at its heart — the price of the harm — is the very thing no one can quite know.4
Beyond the practical difficulty lies an objection more fundamental, which the next plate will press in full. The Pigovian tax assumes that the factory is the cause of the harm, the one to be charged; it casts the polluter as wrongdoer and the neighbor as victim, and bills the wrongdoer accordingly. But this framing, Ronald Coase would argue, is not obviously right. The smoke harms the laundry only because the laundry is there; the harm is the joint product of two activities placed side by side, and to tax the factory for the laundry's sake is as arbitrary, in principle, as to tax the laundry for the factory's. Pigou's tax presumes an answer — that the smoker should pay — to a question Coase insisted was open: whose right is it, the factory's to smoke or the neighbor's to clean air? The tax mends the price, but on an assumption about who owes whom that the next plate will not let stand.5
The levy is the first and boldest cure: to mend the broken price by force of law, charging the harm by its cost and rewarding the good by its worth, so that the market, its arithmetic corrected, wants the right amount of its own accord. Where the harm can be measured and the right to charge is clear, it is an instrument of rare power — a tax that heals rather than hurts. But it rests on two things the next plates will question: that the harm can be priced by a distant authority, and that the maker of the harm is the one who ought to pay. Coase's reply to the second of these is among the most influential arguments in all of economics, and it begins by denying that the smoke has a villain at all. The next plate takes up the neighbor.