Externalities: The Hidden Costs Markets Ignore

In the 1970s, a coal-fired power plant sat on the banks of the Ohio River, burning through trainloads of coal and selling electricity to the surrounding region at a price that comfortably covered its fuel, its turbines, and the wages of everyone who worked there. By the standards of any accountant, the plant was an honest business. It charged what it cost to produce a kilowatt-hour, and customers paid willingly. Yet a few miles downwind, in a school district the plant's owners had never met and would never bill, something else was accumulating: an unusual number of children showing up at the clinic short of breath, the asthma admissions creeping upward season after season.

No invoice connected those two facts. The families breathing the plant's sulfur and particulate matter were never parties to the electricity contract, and the price on the meter said nothing about their hospital visits. That silent wedge, the distance between what the market charged and what the activity actually cost the world, is one of the most consequential ideas in economics. Economists call it an externality, and understanding it explains everything from why we tax carbon to why your neighbor's beehive quietly makes the whole street better off.

What Slips Through the Cracks of a Transaction

An externality exists when a market transaction imposes a cost on, or confers a benefit on, a third party who is not involved in the transaction, and the price the contracting parties face does not reflect that effect. The crucial word is third party. A buyer and a seller meet, they agree on a price, and that price reflects everything the two of them care about. But the consequences of their deal can spill over onto people sitting entirely outside the negotiation, people whose welfare was never priced in because they had no seat at the table.

The coal plant is the textbook case. The buyer wants electricity, the seller wants revenue, and the price they settle on balances those two interests beautifully. What the price ignores is the asthmatic child downwind, because that child neither sold anything nor bought anything. The damage to her lungs is real, it has a genuine economic cost in medical bills and lost school days, and yet it appears nowhere in the transaction that caused it. The market, for all its efficiency at coordinating buyers and sellers, is structurally blind to anyone who is neither.

This blindness is not a moral failing of the people involved. The plant operator is not a villain, and the customer flipping a light switch is not committing a crime. The problem is built into the architecture of the exchange itself. Prices are wonderfully good at carrying information between the two parties who set them, and wonderfully bad at carrying information about everyone else.

The Quiet Gap Between Private and Social Cost

To see why this matters for the economy as a whole, economists distinguish between two kinds of cost. The first is private cost, the expense borne directly by the producer: coal, labor, maintenance, the cost of doing business. The second is social cost, the full cost of the activity to society, which includes the private cost plus whatever damage falls on third parties. When an activity has no spillovers, these two numbers are identical and nothing is amiss. When a third-party cost is present, the social cost lies above the private cost by exactly the amount of the marginal external damage, the extra harm caused by producing one more unit.

Here is where the trouble shows up. A market naturally settles at the quantity where private cost meets price, because that is the point at which it stops being profitable for the producer to make one more unit. But that calculation only weighs the producer's own costs. It never accounts for the harm landing on the school district downwind. Because the producer faces only the private cost, which is lower than the true social cost, the activity looks cheaper than it really is, and so the market produces too much of it.

The result is that the market trades at the wrong quantity. It is not that pollution-causing electricity should never be produced, but that this much of it should not be. Society would be better off with somewhat less coal power and somewhat fewer asthma cases, and the market has no internal mechanism to find that better point on its own. The price signal, the thing markets rely on to guide production, is sending a number that is simply too low.

Two Pictures, Same Shape With the Sign Flipped

The framework becomes genuinely elegant once you notice it runs in both directions. Negative externalities, the coal plant kind, push social cost above private cost, and the consequence is that the market over-produces, churning out more of the harmful thing than is good for anyone. But the same logic, with the sign reversed, describes benefits that spill over rather than costs.

Consider vaccination. When you get vaccinated, you protect yourself, and that private benefit is what motivates your decision. But you also make it marginally harder for a disease to spread to people around you, contributing to herd immunity that protects even those who cannot be vaccinated. That benefit lands on third parties and never enters your personal cost-benefit calculation, so social benefit lies above private benefit. The consequence mirrors the pollution case exactly: the market under-produces. Left to individual choices alone, fewer people get vaccinated than would be ideal for the community, because no one is paid for the protection they confer on their neighbors.

The catalog of examples on each side is long and instructive. On the negative side sit pollution, traffic congestion, antibiotic resistance bred by overuse, and secondhand smoke, all activities whose harms fall on bystanders. On the positive side sit vaccination, beekeeping near orchards (the bees pollinate fruit the beekeeper never harvests), basic scientific research whose discoveries enrich everyone, and herd immunity itself. The framework treats both kinds as the very same divergence between private and social value, with nothing changing but the arithmetic sign.

Pigou's Tax and Coase's Bargain

If the problem is a gap between private and social cost, the natural question is how to close it. The first great answer came from the British economist Arthur Pigou, whose book The Economics of Welfare in 1920 proposed a remedy of striking simplicity. Tax a negative externality at the marginal external cost it imposes, Pigou argued, and the producer will suddenly face the full social cost of the activity. Subsidize a positive externality at the marginal external benefit, and the producer will be rewarded for the good they spread around. In either case the wedge closes, the price starts telling the truth, and the market is nudged toward the right quantity. A century later, this Pigouvian tax still anchors the entire enterprise of carbon pricing.

Four decades after Pigou, the economist Ronald Coase offered a provocative challenge. In The Problem of Social Cost in 1960, Coase argued that government intervention might not be necessary at all. When property rights are clearly defined and the cost of bargaining is low, he showed, the private parties involved can negotiate their way to an efficient outcome on their own, without any tax. If the factory has the right to pollute, the affected neighbors can pay it to pollute less. If the neighbors have the right to clean air, the factory can pay them for permission to emit. Either way, as long as bargaining is cheap and rights are clear, the parties will keep trading until they reach the efficient quantity. The insight was deep enough to earn Coase the Nobel Prize in economics in 1991.

These two results divide the policy terrain between them. Pigou says the state should price the externality directly. Coase says that under the right conditions, the state need only assign clear property rights and then step back. The disagreement is less a contradiction than a question of when each approach applies.

Why the Atmosphere Won't Bargain

The honest answer is that Coase's result holds only under conditions that are often absent. His bargaining solution works when transaction costs are low, when property rights are clear, and when the affected parties are few enough to actually sit down and negotiate. Two neighbors and a noisy workshop might genuinely settle things over a fence. But scale the problem up and the conditions collapse.

Atmospheric carbon dioxide violates all three at once. No one holds a clear property right to the atmosphere, so there is no owner to bargain with or pay off. The transaction costs of getting every emitter and every affected person to the table are effectively infinite, because the emitters number in the billions and the affected parties include people not yet born. And the affected parties are not few but the entire human population across generations. There is no fence to negotiate over and no plausible negotiating table large enough. This is precisely why, for the defining environmental problem of our era, policy has returned to Pigou. When bargaining is impossible, a tax on the externality is the remaining tool.

Pricing Carbon, and the Limits of the Idea

That return to Pigou is now visible in policy around the world. Roughly fifty jurisdictions currently run an explicit carbon price, either as a tax or as a tradable permit system. The European Union's Emissions Trading System saw allowances trade between roughly sixty and one hundred euros per tonne of carbon dioxide across 2022 to 2024, while Sweden's carbon tax sits near one hundred thirty dollars per tonne, among the highest in the world.

Sweden's experience is worth dwelling on because it runs long enough to test the idea. In 1991 the country imposed one of the world's first carbon taxes, set then at the equivalent of about thirty dollars per tonne. Over the following decades it ratcheted the rate steadily upward to roughly one hundred thirty dollars per tonne by 2024. The often-cited result is that Swedish emissions fell by about thirty percent against their 1990 level even as the country's GDP roughly doubled, a pairing frequently held up as evidence that decarbonization and growth are not mutually exclusive. The tax did not strangle the economy; the economy grew while the externality shrank.

Yet the framework is not a finished machine, and its critics raise a genuine difficulty. To set a Pigouvian tax correctly, you must first estimate the marginal external cost, and for carbon that estimate, known as the social cost of carbon, is deeply contested. It depends heavily on the discount rate, the rate at which we trade present costs against future harms, and reasonable economists disagree sharply about what that rate should be. A small change in the discount rate can move the estimated social cost of carbon by a large factor, which means the framework shifts the policy debate without eliminating it. Pigou tells us to tax the externality at its marginal cost; he does not tell us, with any certainty, what that number is. The argument moves from whether to price carbon to how much, which is progress, but it remains an argument.

Key Takeaways

An externality is a cost or benefit from a market transaction that falls on a third party and goes unreflected in the price, which is why a coal plant can sell cheap electricity while asthma cases accumulate downwind; because producers face only their private cost rather than the higher social cost that includes external damage, markets over-produce activities with negative externalities (pollution, congestion, antibiotic resistance) and under-produce those with positive ones (vaccination, basic research, herd immunity), the same divergence with the sign flipped. Two intellectual traditions divide the policy response: Arthur Pigou's 1920 proposal to tax harms and subsidize benefits at their marginal external value, and Ronald Coase's 1960 result that clearly defined property rights and low transaction costs let private parties bargain to efficiency without government action. Coase's bargaining solution fails precisely where atmospheric carbon lives (no clear property rights, prohibitive transaction costs, billions of affected parties), which is why modern climate policy has returned to Pigou, with roughly fifty jurisdictions now pricing carbon and Sweden offering decades of evidence that a rising carbon tax can coexist with a growing economy. The framework's lasting limitation is that setting the right tax requires knowing the social cost of carbon, an estimate that hinges on contested discount-rate choices, so externality theory reshapes the policy debate without resolving it.