Say “externality” at a dinner party and everyone pictures smoke. The word has been captured almost entirely by its bad half — pollution, traffic, noise. But the model is perfectly symmetric, and its other half is where it earns its keep, because it explains a pattern nobody sees with the naked eye: the world is quietly starved of some of its best things — vaccines, education, research, beautiful buildings — and the same market that over-produces smoke under-produces these. Not because anyone opposes them. Because, just like the smoke, their value leaks out to people who never paid for it. This lesson flips every sign from the last one and watches the distortion run in reverse.
The benefit that spills out
Get a flu shot and you buy yourself a clear private benefit: you, personally, are much less likely to spend a week in bed. That’s the benefit you weigh against the cost and the hassle. But you also quietly protect everyone you would have infected — your kids, your coworkers, the stranger on the bus, the immunocompromised neighbour who can’t be vaccinated at all. That protection is real, it’s valuable, and you can’t bill anyone for it. It spills out, free, to third parties. So the benefit you weigh is smaller than the benefit to the world — and a benefit you can’t capture is a benefit you’ll under-buy.
That’s a positive externality: a benefit from an activity that lands on third parties outside the transaction. And it produces the mirror image of pollution:
- Marginal private benefit (MPB) — the value of one more unit to the decider. This is the ordinary demand curve.
- Marginal social benefit (MSB) — the value of one more unit to everyone: the private benefit plus the external benefit spilled onto others. For a positive-externality good, MSB sits above MPB.
The market clears where supply meets MPB (the only benefit the decider feels). But the socially right quantity is where supply meets MSB — the value the world actually gets. Because MSB is higher, it meets supply later (at a larger quantity). So the market’s quantity lands to the left of the optimum: under-production. Everything is the negative case in a mirror — too little instead of too much.
Private cost vs. social cost
Now push the slider into the blue: a benefit spilled onto others
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 benefit per unit lifts true social value above private. The market makes 7 units but 10 is optimal — it under-produces by 3, leaving $4.5 of welfare (the shaded triangle) on the table.
Drag the slider right, into “external benefit,” and the dashed curve now rises above demand. The red market dot sits to the left of the green optimum — the market stops making units that the world would gladly have, because the people deciding can’t pocket enough of the payoff to bother. That gap is every under-built good thing: the vaccine not taken, the degree not earned, the research not funded.
For a positive-externality good like vaccination, why does the marginal-social-benefit curve sit *above* the demand (marginal-private-benefit) curve?
The same triangle, now made of missed welfare
The deadweight-loss triangle reappears, but its meaning flips in a way worth pausing on. For pollution, the triangle was welfare destroyed by units that shouldn’t have been made. Here it’s welfare never created by units that should have been made and weren’t. Each unit between the market quantity and the optimum is one whose social benefit (the high MSB curve) exceeds its cost (the supply curve) — a genuinely worthwhile unit that simply never happens, because no individual decider captures enough of its benefit to pay for it. The triangle is the value the world would have had, forgone.
Two triangles, one idea
For a negative externality the deadweight loss is welfare destroyed by over-production (units whose social cost beats their benefit, made anyway). For a positive externality it’s welfare forgone by under-production (units whose social benefit beats their cost, never made). Same triangle, opposite sign — both are the market landing on the wrong quantity.
Working the numbers
Let’s price out a year of college education, one student at a time. Each student gets a marginal private benefit (higher lifetime earnings, the value they capture), costs society a marginal cost to educate, and spills a flat $30k external benefit onto everyone else — a more skilled, more productive, more civic society, new ideas, lower crime.
| Student | Marginal private benefit (MPB) | Marginal cost | External benefit | Marginal social benefit (MSB) |
|---|---|---|---|---|
| 1 | $90k | $40k | $30k | $120k |
| 2 | $80k | $50k | $30k | $110k |
| 3 | $70k | $60k | $30k | $100k |
| 4 | $50k | $70k | $30k | $80k |
| 5 | $40k | $80k | $30k | $70k |
What the market does. A student enrols whenever their own benefit covers the cost — whenever MPB ≥ marginal cost. Student 1 (40k): yes. Student 2 (50k): yes. Student 3 (60k): yes. Student 4 (70k): no. The market educates 3 students.
What’s socially optimal. Society should educate a student whenever the full benefit covers the cost — MSB ≥ marginal cost. Student 4: MSB 70k — worth it! Student 5: MSB 80k — not worth it. The optimum is 4 students.
The loss. The market under-produces by 1 student (the 4th), whose forgone welfare is MSB 70k = **50k, sensibly walked away from a $70k cost. The spillover they couldn’t capture was exactly what tipped it.
The market still educates 3 (its private numbers are unchanged). But MSB is now MPB + 50k + 100k against a 40k + 90k against an $80k cost (also worth it now). The optimum jumps to 5 students, so the market under-produces by 2. A bigger positive spillover means a bigger gap between what’s privately and socially worthwhile — and a bigger pile of forgone welfare. Same lesson as the negative side: the size of the spillover is the size of the problem.
A pharmaceutical executive says: 'We can't justify the research budget for a new antibiotic — even if it works, we'd sell very little of it, because doctors will hoard it for emergencies.' How does the externality lens read this?
The gallery of positive externalities
Wherever an activity throws off benefits its decider can’t bill for, you’ll find under-production. The recurring cast:
- Vaccination & public health — you buy your own protection and give away everyone-else’s-protection (herd immunity) for free.
- Education — the graduate captures higher wages, but a more skilled, informed, innovative society spills onto everyone, across generations.
- Basic research & R&D — knowledge is the purest positive externality: once discovered, anyone can build on it. The discoverer captures a sliver; the world keeps the rest forever.
- Beautiful or restored buildings — the owner pays; the whole street enjoys the view (the façade case from lesson 2).
- Beekeeping & orchards — the beekeeper sells honey; the neighbouring orchards get pollination free. The textbook example, and a real one.
- Network goods & getting vaccinated against misinformation — anything where your participation makes the thing better for others.
The diagnostic question simply flips: who gets a benefit from this that they didn’t pay for? If the answer is “people outside the deal,” it’s a positive externality, and you can predict — sight unseen — that the market makes too little of it.
The pitfall: forgetting the benefit half entirely
The most common externality mistake isn’t getting the analysis wrong — it’s never running it, because positive externalities are invisible by construction. An over-produced bad (smog, noise) announces itself; an under-produced good is a non-event — the vaccine not taken, the research never funded, the building never restored leaves no smoke to notice. Train yourself to ask not just “what’s being over-done here?” but “what worthwhile thing is quietly not happening — and is a spillover the reason?”
Recap
- A positive externality is a benefit that spills onto third parties, so marginal social benefit (MSB) = MPB + external benefit sits above the private demand curve.
- The market clears where supply meets MPB; the optimum is where supply meets MSB. Because MSB is higher, it meets supply later — so the market lands to its left and under-produces.
- The deadweight-loss triangle reappears as welfare forgone — worthwhile units (social benefit > cost) that never get made because no decider can capture enough of the payoff. In our worked example: market 3, optimum 4, $10k of value lost.
- The cast: vaccines, education, research, restored buildings, beekeeping — all under-produced for the same mirror-image reason pollution is over-produced.
- The deepest pitfall is not seeing it at all: an over-produced bad is loud, an under-produced good is a silent non-event. Always ask what worthwhile thing isn’t happening.
Positive externalities — check
A town is deciding how much to spend on a free vaccination campaign. Left purely to individuals, what does the externality lens predict about how many people get vaccinated?
Check your answer to continue.
Next up: the repair kit. In Internalizing the Externality we close both gaps — Pigouvian taxes and subsidies, cap-and-trade, regulation, and the surprising Coase theorem — and flip on the chart’s corrective toggle to watch the market dot snap back onto the optimum.