We’ve spent three lessons diagnosing the disease: when a cost or benefit lands on third parties, private and social cost come apart, and the market lands on the wrong quantity — too much pollution, too few vaccines, a triangle of welfare gone. This is the lesson where we cure it. And the cure has a single, unifying name that tells you everything: internalizing the externality. Every fix in the kit, however different it looks, does the same one thing — it drags the external cost or benefit back inside the decider’s own accounting, so the private cost they weigh finally equals the social cost the world bears. Do that, and the market — that same faithful machine — re-aims itself at the right target with no further nagging.
The one move every cure makes
The unifying idea
To internalize an externality is to make the decider feel the spillover — to move the external cost or benefit into the price they actually face. Once private cost = social cost, the market’s own equilibrium is the social optimum. Every tool below is a different mechanism for the same move.
You saw this exact move on the chart already, greyed-out and waiting. Let’s switch it on. Drag the slider to a negative externality, then tick the corrective policy — a Pigouvian tax equal to the spillover — and watch what happens.
Private cost vs. social cost
Flip the switch: watch the gap close
Drag the external effect. Negative = a cost dumped on others (the market over-produces); positive = a benefit spilled onto others (it under-produces). Then apply the corrective policy and watch the gap close.
A $3 external cost per unit lifts true social cost above private cost. The market makes 7 units but only 4 is optimal — it over-produces by 3, destroying $4.5 of welfare (the shaded triangle).
Ticking the box doesn’t move the social cost curve — the pollution is just as harmful as before. It moves the private curve up to meet it, by charging the producer exactly the external cost per unit. Now the producer’s own cost finally equals society’s cost, so the quantity they choose is the quantity society wants. The triangle — the dead weight — disappears, because no over-produced units remain. That single animation is the whole lesson; the rest is how you build that switch in the real world, and where each build cracks.
Tool 1 — Pigouvian taxes (and subsidies)
The idea. Named for Arthur Pigou (the man with the railway sparks), a Pigouvian tax is a tax set equal to the external cost of an activity, charged per unit. It’s not designed to raise revenue or to punish — it’s designed to make the price honest. The tax is the missing cost, added back. Its twin for the benefit side is the Pigouvian subsidy: a per-unit payment equal to the external benefit, which lifts private demand up to social benefit so the market makes more of the under-produced good.
In the curves. A Pigouvian tax shifts marginal private cost up by the external cost, so MPC meets MSC and the market quantity falls to the optimum. A Pigouvian subsidy shifts private benefit up by the external benefit, so MPB meets MSB and the quantity rises to the optimum. Exactly the corrective toggle on the chart.
Real examples. Carbon taxes price each tonne of CO₂; congestion charges (London, Singapore, Stockholm) price the delay a driver imposes; tobacco and alcohol “sin” taxes price the health spillovers; bottle deposits price litter. On the subsidy side: free or subsidised vaccines, public funding of education and basic research, tax credits for R&D and rooftop solar.
Where it breaks. The Achilles’ heel is the number: a Pigouvian tax is only right if it equals the true external cost, and that cost is genuinely hard to know — what is a tonne of carbon’s damage? Set it too low and you barely dent the over-production; too high and you choke useful activity into a different distortion. Pigouvian taxes are also politically toxic (nobody cheers a new tax) and can be regressive — a fuel tax bites a delivery driver harder than a banker — unless the revenue is rebated, often as a flat dividend.
A government sets a carbon tax at $40 per tonne, but the best estimate of the true damage from a tonne of CO₂ is about $100. In Pigouvian terms, what's the result?
Tool 2 — Cap-and-trade (the quantity-based cousin)
The idea. Instead of setting the price of the externality and letting quantity adjust (Pigou’s way), you set the quantity and let the price adjust. The regulator caps the total allowed pollution, issues that many permits, and lets firms buy and sell them. Firms that can cut emissions cheaply do so and sell their spare permits; firms for whom cutting is expensive buy permits instead. The cap guarantees the total; the trading ensures the cuts happen wherever they’re cheapest.
Real examples. The U.S. SO₂ “acid rain” program of the 1990s is the celebrated success — it slashed sulphur emissions faster and cheaper than anyone forecast. The EU Emissions Trading System is the largest carbon market today.
Tax vs. cap — the real trade-off. They’re two routes to the same destination, and the choice is about what you’d rather control. A tax fixes the price of pollution and lets the quantity land where it may — good when you care most about cost certainty. A cap fixes the quantity and lets the price float — good when there’s a hard ceiling you must not cross. Under perfect information they’re equivalent; under uncertainty you pick the one whose risk you can live with.
Where it breaks. Set the cap too loose (too many permits) and the price collapses to near zero and bites nothing — a frequent real-world failure. Permits can be gamed, grandfathered to incumbents, or lobbied looser. And like every quantity rule, it needs solid monitoring to verify nobody emits beyond their permits.
They’re the same Pigouvian move approached from opposite ends. A tax sets the price of carbon and lets the market discover the quantity; cap-and-trade sets the quantity and lets the market discover the price. Both internalize the externality by making emitters pay for emitting. With perfect information they reach the identical outcome; the genuine difference is which variable you pin down and which you let float — price certainty (tax) versus quantity certainty (cap). Calling them different philosophies overstates it; they’re two dials on the same machine.
Tool 3 — Regulation (command-and-control)
The idea. The bluntest tool: a central authority simply mandates the outcome. Emission limits per smokestack, catalytic converters required on every car, bans on leaded petrol and CFCs, minimum vaccination for school entry, building codes, compulsory schooling. No price, no market — just a rule, backed by penalties.
Where it shines. When the right quantity of the externality is essentially zero (lead in petrol, CFCs in the ozone layer, asbestos), a flat ban is simpler and surer than fiddling with a tax. Regulation is also the tool of choice when harm is catastrophic or irreversible, when the activity is hard to meter for a tax, or when a clear, enforceable standard beats a price nobody can compute.
Where it breaks. Command-and-control is inefficient in the economist’s exact sense: a uniform rule (“every plant cuts 30%”) forces expensive cutters and cheap cutters to do the same thing, when a tax or a permit market would let the cheap cutters do more and save everyone money. It’s informationally demanding (the regulator must know the right standard), rigid (slow to update as technology changes), and prone to capture — written by the very industry it polices.
Tool 4 — The Coase theorem (maybe you don’t need the government at all)
The idea. Here’s the surprise that won Ronald Coase a Nobel Prize and upended the field. Pigou assumed externalities require an outside authority — a tax, a cap, a rule. Coase showed that’s not always so. His theorem: if property rights are clearly defined and people can bargain at low cost, the parties will negotiate their own way to the efficient outcome on their own — regardless of who initially holds the right.
The intuition. Recall Pigou’s railway throwing sparks onto a farmer’s wheat. Pigou says: tax the railway. Coase says: wait. If the right to the airspace is clear — say the farmer has the right to spark-free fields — and the two can talk cheaply, then if running an extra train is worth more to the railway than the burnt wheat costs the farmer, the railway will simply pay the farmer for permission, and the train runs. If it’s not worth that much, it won’t. Either way you land on the efficient amount of train-running — and remarkably, the same efficient amount results even if you’d assigned the right the other way (the railway free to spark, the farmer paying it to stop). Clear rights plus cheap bargaining internalize the externality privately.
Real examples. Two neighbours settling a noise or tree-shade dispute over the fence; a downstream and upstream firm negotiating water use; carbon-offset deals and conservation easements where a polluter pays a landowner to preserve forest. Small-number, well-defined cases.
Where it breaks — and why Pigou usually still wins. Coase’s own point was that his ideal conditions are often absent, and the gaps tell you when to reach for a tax instead. It fails when transaction costs are high (you cannot assemble eight billion people to bargain over the climate), when property rights are unclear or unassignable (who owns the atmosphere?), when there are too many parties or a holdout can extort the deal, and when information is poor. Notice these are exactly the conditions of every big externality — which is why Coase explains the back-fence settlement but not climate change, and why governments still levy Pigouvian taxes for the large, many-party, ill-defined-rights cases.
Coase's real lesson
Coase isn’t “markets fix everything, government go home.” His deeper point is that the choice between private bargaining and government action turns on transaction costs. Low costs and clear rights → let the parties bargain. High costs, fuzzy rights, a cast of millions → you need a tax, a cap, or a rule. The theorem is a diagnostic, not a libertarian slogan.
Two adjacent factory owners can cheaply negotiate, and the law clearly says the downstream one owns the right to clean water. By the Coase theorem, what happens to the upstream factory's pollution?
Choosing the tool — match the fix to the externality
There’s no master key, only a diagnosis. Read the externality — how many parties, how clear the rights, how measurable the spillover, how catastrophic the harm — and match the remedy.
| Tool | Best when… | Breaks when… | Real example |
|---|---|---|---|
| Pigouvian tax / subsidy | The spillover is roughly knowable and meterable per unit; many diffuse parties | The right number is unknowable; politically toxic or regressive | Carbon tax, congestion charge, vaccine subsidy |
| Cap-and-trade | You need a hard quantity ceiling and want cuts made cheaply | The cap is set too loose; permits gamed or grandfathered | US SO₂ program, EU ETS |
| Regulation / bans | The right amount is ~zero, or harm is catastrophic/irreversible | Uniform rule wastes money; rigid; captured | Leaded-petrol ban, CFC ban, school vaccine mandates |
| Coase / private bargaining | Few parties, clear property rights, low transaction costs | Many parties, fuzzy rights, high transaction costs | Neighbour disputes, conservation easements |
The pitfall is the one-size ideologue — “tax everything,” “deregulate everything,” “just let people bargain.” A tax is useless where you need a flat ban on a poison; bargaining is useless across a billion strangers; a ban wastes money where a price would do. The grown-up move is diagnostic: often you blend them — a carbon tax plus efficiency regulations plus R&D subsidies, each aimed at the part of the problem it fits.
Pick a term, then click its definition.
Recap
- Every cure internalizes the externality — it drags the external cost or benefit into the decider’s own price, so private cost = social cost and the market’s equilibrium becomes the social optimum. (That’s the chart’s corrective toggle.)
- Pigouvian taxes/subsidies price the spillover back in per unit — powerful and precise if you know the right number; limited by that number’s difficulty, politics, and regressivity.
- Cap-and-trade fixes the quantity and lets the price float — the mirror of a tax; great for a hard ceiling, vulnerable to a loose cap and gamed permits.
- Regulation/bans mandate the outcome — best when the right amount is ~zero or harm is catastrophic; inefficient and rigid as a general tool.
- The Coase theorem — clear rights + cheap bargaining → the parties fix it themselves, no government needed; but it fails for the many-party, fuzzy-rights, high-transaction-cost cases that describe every big externality. No panacea: diagnose the externality and match — or blend — the tool.
Internalizing externalities — final check
A city wants to cut downtown traffic. It introduces a charge to drive into the centre at peak hours, set roughly equal to the delay each car imposes on others. What kind of tool is this, and why should it work?
Check your answer to continue.
That’s the repair kit, and the end of the teaching. You can now diagnose an externality, draw its two curves, measure the welfare it costs, and reach for the right tool — or blend — to close the gap. The last step is to prove it sticks: head to the Final Exam and put the whole course to the test.