Market Failures III: The Case for (Pigouvian) Taxes, Subsidies, and Incentives
Incentives are better than bans
Everyone wants less traffic, but almost everyone also wants to be one of the cars causing it.
That is the strange honesty of a crowded road. A driver entering at rush hour receives the benefit of the trip, but also imposes delay on everyone else. The private decision is simple: Is this trip worth my time and gas? The social decision is bigger: Is this trip worth my time, my gas, and the delay I impose on hundreds of other people?
A congestion charge does not ban driving. It asks drivers to face more of the cost their trip imposes on everyone else. A ban says, "do not do this." A price says, "do this only if it is worth the full cost."
Internalizing an Externality
The repair is to bring the missing cost or benefit into the decision.
Chapter 18 showed the problem: private decision-makers may ignore costs or benefits that fall on others. This chapter is about repairs. To internalize an externality means to make the decision-maker face the external cost or receive credit for the external benefit.
For a negative externality, the market produces too much because the private cost is too low. The policy goal is to raise the private cost until the decision-maker acts as if they are facing the full social cost. For a positive externality, the market produces too little because the private benefit is too low. The policy goal is to raise the private reward until the decision-maker acts as if they are receiving more of the social benefit.
| Externality | Market problem | Corrective tool | Efficient size |
|---|---|---|---|
| Negative externality | Too much output because MPC is below MSC. | Tax or fee | Tax = marginal external cost at Qs |
| Positive externality | Too little output because MPB is below MSB. | Subsidy | Subsidy = marginal external benefit at Qs |
Quick checkIf a good creates external costs, should policy usually push quantity up or down?▸
Pigouvian Taxes
A well-calibrated tax can make private cost look more like social cost.
A Pigouvian tax is a corrective tax on a negative externality. The point is not "taxes are always good." The point is narrower and more powerful: if an activity imposes an external cost, a tax can make the private decision-maker face that cost.
On the AP graph, the tax shifts MPC/S upward toward MSC. The market moves from Qm, where MPB intersects MPC, to Qs, where MPB intersects MSC. Output falls because some units were worth producing only when the external cost was ignored.
- Marginal social cost is $11.
- Marginal private cost is $8.
- Marginal external cost is MSC - MPC = $11 - $8 = $3.
MisconceptionAny tax on a harmful product is a Pigouvian tax.▸
A true Pigouvian tax is calibrated: it equals the marginal external cost at the efficient quantity, no more and no less. Its job is to make the price tell the truth, then get out of the way. A tax set to raise revenue, or set above the external cost because the activity is unpopular, is not corrective — it overshoots Qs and creates its own deadweight loss. "It taxes a bad thing" does not make a tax Pigouvian; the right size does.
Congestion Pricing
Traffic is a clean everyday example of pricing the external cost.
Congestion pricing charges drivers for scarce road space at crowded times. The idea is not that driving is evil. It is that a road at 8:30 a.m. is not the same economic good as that same road at 2:00 a.m. At rush hour, one more car can slow many other people down.
A congestion price asks: Is this trip worth the full cost, including delay imposed on others? Some drivers still pay because the trip is important. Some shift to transit, carpooling, a different route, remote work, or a less crowded time. That flexibility is why economists often like prices more than commands: people know more about their own alternatives than a central planner does.
There is an equity issue, and it should not be waved away. A congestion charge can burden lower-income drivers if designed badly. Revenue use matters. The money can support transit, reduce other regressive fees, improve roads, or offset burdens on households with fewer options. Good policy design asks not only "Does this improve efficiency?" but also "Who bears the burden, and what do we do with the revenue?"
MisconceptionA price and a ban accomplish the same thing, so it doesn't matter which you use.▸
They reach very different places. A ban treats every use as equally unworthy and forbids it. A price lets the trips, units, or activities that are genuinely worth the full cost continue, while discouraging the ones that are not — and it does this using information only the users have about their own alternatives. Same goal, very different efficiency. That gap is exactly why economists usually reach for the price first.
Retrieval practiceWhy is a congestion charge more flexible than a driving ban?▸
Pigouvian Subsidies
Positive externalities often need encouragement, not punishment.
A Pigouvian subsidy corrects a positive externality. When education, vaccination, or research creates benefits for people beyond the direct decision-maker, the market may produce too little. The private reward is smaller than the social benefit.
On the AP graph, MSB lies above MPB/D. The market quantity is Qm, where MPB intersects MPC. The socially efficient quantity is Qs, where MSB intersects MPC. A subsidy raises the private benefit or lowers the effective price paid by consumers, moving output from Qm to Qs.
- Marginal social benefit is $28.
- Marginal private benefit is $20.
- Marginal external benefit is MSB - MPB = $28 - $20 = $8.
MisconceptionIf something is good, subsidize it as much as possible.▸
The corrective amount is finite and specific: the subsidy equals the marginal external benefit at the efficient quantity. Over-subsidizing pushes output past Qs, which is inefficient in the same way under-providing it was — just in the other direction. Too much of a good thing is still too much. A subsidy that ignores the size of the spillover is no more "right" than a tax that ignores the size of the harm.
Carbon Taxes and Cap-and-Trade
Two different ways to put a price on emissions.
A carbon tax charges per unit of emissions. A firm that emits more pays more. A household or business that finds cheap ways to cut emissions saves money. The policy uses the price system to reward low-cost reductions.
Cap-and-trade works differently. The government sets a total cap on emissions and issues permits. Firms that can reduce emissions cheaply do so and sell extra permits. Firms that face higher reduction costs buy permits. The cap controls the total quantity, while trading lets reductions happen where they cost least.
| Policy | What it fixes | What government sets | Why economists like it |
|---|---|---|---|
| Carbon tax | Emissions price is too low. | Price per unit of emissions. | Simple incentive: pollute less, pay less. |
| Cap-and-trade | Total emissions are too high. | Total emissions cap. | Trading finds lower-cost reductions across firms. |
Quick checkWhich policy sets the quantity directly: carbon tax or cap-and-trade?▸
Incentives vs. Bans
Prices can use local knowledge that commands often miss.
Economists are not allergic to rules. Some harms are extreme enough that bans make sense. Some standards are necessary because measuring every harm perfectly is impossible. But the economist's instinct is to ask whether a price-based tool can achieve the goal with less waste.
Do not command every person to make the same sacrifice. Put a price on the harm and let people find the cheapest ways to reduce it. That is not a slogan against government. It is a way to design government policy that uses the same price signals that make markets powerful in the first place.
| Policy | Strength | Weakness |
|---|---|---|
| Ban | Clear and useful for extreme harms. | Blunt; treats very different cases the same. |
| Regulation | Can target a specific technology, practice, or safety rule. | May miss the cheapest ways to reduce harm. |
| Pigouvian tax | Flexible incentive; makes harmful activity more costly. | Requires measuring and calibrating external cost. |
| Subsidy | Encourages activities with external benefits. | Can be poorly targeted or captured by interest groups. |
| Cap-and-trade | Caps quantity while allowing flexible reductions. | Requires monitoring, enforcement, and permit design. |
The Pigouvian Catch
To price the harm correctly, someone has to know the harm — and want to.
Everything above is real, and where it works, it works beautifully. But the elegant graph hides two problems, and neither is about the theory. Both are about the fixer.
The first is a knowledge problem. The optimal tax equals the marginal external cost at the efficient quantity — a number no one observes directly. What is a ton of carbon's damage worth? An hour of rush-hour delay? A unit of river pollution? These are genuinely hard to measure, contested among honest experts, and they shift with conditions. The graph draws a single clean line for MSC; reality hands a regulator a wide range. Set the tax too high and you destroy valuable activity; too low and the externality survives. A perfectly designed tax requires information that is hard to get.
The second is a public-choice problem. A Pigouvian tax is not handed down by a benevolent planner. It is set by legislators, agencies, and regulators who have their own incentives — re-election, budgets, lobbying, regulatory capture. The same self-interest that let the factory pollute in the first place now operates on the people meant to price the pollution. Affected industries lobby for exemptions, loopholes, and free permits. And a government can grow fond of revenue from a tax that was supposed to shrink the activity — a tax that fully succeeds shrinks its own base, an awkward thing to want.
So the honest conclusion does not retract the chapter; it completes it. The externality diagnosis is settled, and the corrective tax is right in theory and often right in practice. But "market failure" does not prove "government success." A corrective policy has to clear a second bar: can this government measure the harm well enough, and resist capture enough, to set the price near right? Sometimes clearly yes — carbon and congestion are measurable, enforceable, and have worked in real cities and countries. Sometimes no, and a cruder rule, or even tolerating the externality, beats a badly-set tax. The point is not that government cannot price externalities. It is that doing so is a second question with its own answer each time, not an automatic follow-on from "the market failed."
Two traditions sharpen the catch. Public choice, associated with James Buchanan and Gordon Tullock, studies political actors with the same assumption economics uses for everyone else: they respond to incentives. Buchanan called it "politics without romance" — voters, legislators, and regulators are not selfless planners, so the body that corrects a market failure is itself subject to lobbying, capture, and self-interest. Separately, Friedrich Hayek's knowledge problem argues that the information needed to intervene well — here, the true external cost — is dispersed and local, and rarely sits in usable form on a planner's desk.
Mark the weight honestly: these are political economy, not AP supply-and-demand mechanics, and you will not graph "public choice." What is fair game, and what matters, is the concept of government failure — the recognition that the regulator is also self-interested and imperfectly informed, so intervention must be judged, not assumed.
MisconceptionShowing a market fails proves the government should step in.▸
A market failure is a necessary condition for intervention, not a sufficient one. It establishes that the unregulated outcome is flawed — but the proposed fix is run by people who are also imperfectly informed and self-interested, so it might improve on the market or might not. "The market failed" and "this government policy will help" are two separate claims, each needing its own evidence. Skipping the second is the most common way good economics turns into bad policy.
You now hold the whole unit in your hands: the failures — escapability, missing institutions, the commons, externalities, hidden information — and the repairs, each with its own catch. The last chapter puts them all in one room, on one person. Healthcare is the market where nearly every failure in this unit fires at once: inelastic demand when you cannot walk away, public-good spillovers, a commons in antibiotic resistance, externalities in contagion, and information so lopsided that the patient often cannot judge what they are buying. It is the capstone because it is where the entire unit comes due. Chapter 20
Correct the externality
The moves you can now make: size a corrective tax or subsidy to the external cost or benefit, move the market from Qm to Qs without confusing the two graphs, and judge when a price beats a ban — and when the fixer's limits matter.
Corrective tax
A factory's marginal private cost is $8 and its marginal social cost is $11 at the efficient quantity. What Pigouvian tax would internalize the externality?
Reveal answer▸
$3 per unit. MEC = MSC - MPC = $11 - $8.
Corrective subsidy
A vaccination creates $20 of private benefit and $28 of social benefit at the efficient quantity. What Pigouvian subsidy would internalize the externality?
Reveal answer▸
$8 per unit. MEB = MSB - MPB = $28 - $20.
Policy comparison
Why might economists prefer cap-and-trade to a rule requiring every firm to cut emissions by the same percentage?
Reveal answer▸
Different firms have different costs of reducing emissions. Trading lets firms with low reduction costs cut more and sell permits, while firms with high reduction costs buy permits. The same total cap can be achieved at lower cost.
Correcting pollution and education spillovers
A competitive market for a product creates pollution. In a separate market, job training creates benefits for firms and workers beyond the person receiving the training.
- 1 pt Identify the type of externality in the pollution market.
- 2 pts Draw a correctly labeled negative externality graph with MPC, MSC, MPB/D, Qm, and Qs.
- 1 pt Explain how a Pigouvian tax changes output.
- 1 pt Calculate the optimal tax if MSC is $14 and MPC is $9 at Qs.
- 1 pt Identify the type of externality in the job-training market.
- 2 pts Draw or describe the positive externality correction with MSB above MPB/D and a subsidy moving Qm toward Qs.
- 1 pt Explain why a price-based policy may be more flexible than a ban.
Reveal scoring guide▸
- 1 ptPollution is a negative production externality.
- 1 ptCorrectly labels MPC/S and MPB/D.
- 1 ptDraws MSC above MPC and labels Qm and Qs with Qm > Qs.
- 1 ptExplains a tax raises private cost and reduces output toward Qs.
- 1 ptOptimal tax = $14 - $9 = $5.
- 1 ptJob training is a positive externality.
- 1 ptCorrectly states MSB is above MPB/D.
- 1 ptCorrectly states subsidy increases quantity from Qm toward Qs.
- 1 ptExplains prices allow people/firms to choose lower-cost adjustments.
Reveal common mistakes▸
- Setting the tax equal to the total external cost rather than the marginal external cost at the efficient quantity.
- Mixing up the corrections: taxing a positive externality or subsidizing a negative one.
- Drawing MSB below MPB for a positive externality (it sits above) or MSC below MPC for a negative one (it sits above).
- Assuming any corrective policy automatically improves welfare, without checking whether the harm can be measured and the rule enforced.
When prices omit external costs or benefits, well-designed taxes, subsidies, and tradable permits can move private incentives toward social efficiency — but only if someone can measure the harm and set the price honestly.
they would usually prefer incentives to bans — use prices before prohibitions, put a price on the harm and let people find the cheapest ways to reduce it — while admitting that a price set by the wrong number, or the wrong motives, can do its own damage.
- A calibrated Pigouvian tax reduces a negative externality toward Qs; a subsidy raises a positive one toward Qs.
- The right correction equals the external cost or benefit at the efficient quantity — not more.
- Tradable permits can hit a cap while routing reductions to whoever can make them cheapest.
- Can this government measure the harm and resist capture well enough to set the price near right?
- How should the revenue be used, especially when the burden falls unevenly?
- When is the harm too serious or too hard to price, so that a rule beats a price?
When a policy "puts a price on" something you dislike, ask two questions: is the price the size of the actual harm, and who set it? A corrective tax and a money grab can look identical from the outside.
To change behavior, price the outcome you care about and let people find their own cheapest path — but expect the people being priced to lobby the number. Design as if the rule will be gamed, because it will.
