So far in this series, we’ve been quietly making an assumption: that when a buyer and seller agree on a price, that transaction fully captures everything that matters. The buyer gets what they want, the seller gets paid, and the rest of the world is unaffected. Most of the time, that’s a perfectly reasonable assumption. If we buy a sandwich, the sandwich-maker is happy with the money. We are happy with our lunch. Also, nobody three blocks away is affected one way or the other.
But sometimes that assumption breaks down, and it breaks down in ways that matter enormously for how we think about the proper role of government, environmental policy, public health, education funding, and dozens of other live policy debates. This post is about what happens when the effects of a transaction — good or bad — spill over onto people who weren’t part of the deal at all.
What Is an Externality?
An externality occurs when the production or consumption of a good creates a cost or benefit for a third party who didn’t choose to be involved in the transaction, and whose costs or benefits aren’t reflected in the market price.
The key word here is “third party.” If you and a contractor agree on a price to renovate your kitchen, and the contractor does a sloppy job, that’s a problem between the two of you — a private dispute, not an externality. But if that same contractor’s crew leaves a pile of debris on the sidewalk that a neighbor trips over, that’s an externality: a cost imposed on someone who had no say in the original contract and received no compensation in the price.
Externalities come in two flavors, and getting comfortable distinguishing them is essential.
Negative Externalities:
Negative externalities impose costs on third parties. The textbook example is pollution: a factory making steel might pollute the air or a nearby river as a byproduct of production. The factory and its customers benefit from the transaction (steel gets made and sold), but people living downstream or downwind bear a cost. It can be dirtier air, contaminated water, higher rates of certain illnesses without being compensated and without having agreed to bear that cost.
Positive Externalities:
Positive externalities create benefits for third parties. Vaccination is the classic example of this. When we get vaccinated against a contagious disease, we protect ourselves, obviously. But we also reduce the chance of spreading that disease to everyone around us, including people who never paid a cent and never showed up to the doctor’s office. Education is another common example: when you become more educated, you benefit personally (higher future earnings, more opportunities), but society also benefits in ways beyond your personal gain — a more educated population tends to have lower crime rates, more informed civic participation, and faster innovation, benefits that spill over to people who had nothing to do with your individual decision to study.
Why Externalities Cause Markets to “Fail”
Here’s the crucial insight, and it connects directly back to everything we built in our first post on supply and demand. A normal demand curve reflects the private benefit that buyers get from a good. A normal supply curve reflects the private cost that sellers incur to produce it. When the two intersect, we get an equilibrium that perfectly balances private costs and private benefits.
But when there’s an externality, the private costs or benefits no longer match the social (total, including third parties) costs or benefits — and the market equilibrium, optimized only around the private side of the ledger, ends up being wrong from society’s broader point of view.
Negative Externality:
With a negative externality (like pollution), the social cost of production (the steel factory’s own production costs, plus the pollution costs borne by everyone downstream) is higher than the private cost the factory actually pays. Because the factory doesn’t have to pay for the pollution it creates, its supply curve reflects only its own private costs, sitting further to the right (more output, at every price) than the true social-cost supply curve would be. The market produces and consumes more of the polluting good than what is socially optimal. This is sometimes called overproduction relative to the socially optimal level.
Positive Externality:
With a positive externality (like vaccination or education), the social benefit (the individual’s own benefit, plus the spillover benefit to everyone else) is higher than the private benefit that the individual buyer considers when deciding whether to purchase. Because the buyer only weighs their own personal benefit, demand reflects only the private-benefit curve, sitting further to the left (less quantity demanded, at every price) than the true social-benefit demand curve would be. The market produces — and consumes — less of the beneficial good or service than is actually good for society. This is underproduction relative to the socially optimal level.
This is the heart of what economists mean by market failure: a situation where a market, left entirely to itself, fails to produce the quantity that maximizes total benefit to society — not because anyone is behaving irrationally or maliciously, but because the price signal simply isn’t carrying all the relevant information. The factory owner isn’t evil for not voluntarily paying to clean up pollution any more than an individual is selfish for not voluntarily over-buying vaccines for the whole neighborhood — both are responding rationally to the private incentives the market actually presents them with. The problem is structural, baked into how the price mechanism works when third-party effects exist.
Fixing Negative Externalities: Taxes, Caps, and Regulation
If the core problem with a negative externality is that producers don’t bear the full social cost of their actions, then the most direct economic solution is to make them bear it. This policy tool is named after the economist Arthur Pigou: a Pigouvian tax. It is a tax set equal to the size of the external cost. It is designed to push the private cost facing the producer up to match the true social cost.
Pigouvian Tax:
A Pigouvian tax shifts the factory’s effective supply curve left. It raises the market price and reduces the quantity produced. This brings the market closer to the socially efficient outcome. The efficient outcome is the quantity society would have chosen if all costs had been fully accounted for in the first place. Carbon taxes are a real-world example of this exact logic applied to climate change. By attaching a price to carbon emissions, the tax forces producers and consumers to internalise a cost (long-term climate damage) that the market would otherwise let them ignore entirely.
Cap and Trade System:
A closely related approach is a cap-and-trade system, used for things like sulfur dioxide emissions (which cause acid rain) and, in various forms, carbon emissions. Instead of directly setting a tax dollar amount, the government sets an overall cap on total allowable pollution, issues a corresponding number of tradable permits, and lets firms buy and sell permits among themselves. Firms that can reduce pollution cheaply will do so and sell their spare permits; firms for whom reduction is very expensive will buy extra permits instead. The appeal of this system is that it achieves a specific, guaranteed total pollution reduction (the cap), while still letting the market figure out the most cost-efficient way to distribute who does the actual reducing — generally a more flexible and often cheaper approach than a one-size-fits-all regulation telling every single firm exactly how much to cut.
Beyond taxes and tradable permits, governments also frequently use direct regulation — simply mandating maximum pollution levels, required equipment (catalytic converters, scrubbers on smokestacks), or outright bans on certain especially harmful activities. Regulation can be more straightforward to enforce and easier for the public to understand than a tax, though economists often point out it can be less efficient, since it typically doesn’t let firms find their own cheapest path to compliance the way a tax or tradable-permit system does.
Fixing Positive Externalities: Subsidies and Public Provision
The mirror-image solution applies to positive externalities. If the core problem is that buyers don’t capture the full social benefit of their choices, the most direct fix is a subsidy — essentially a negative tax, paid by the government to lower the effective price the buyer faces, nudging consumption up toward the socially optimal quantity.
This is why many governments heavily subsidize vaccines, sometimes making them entirely free at the point of use, even though producing and distributing a vaccine isn’t actually free — the government effectively absorbs the gap between the low private benefit an individual might perceive (a mild illness avoided) and the much larger social benefit of reduced disease transmission across the entire population.
Most countries subsidise education in a particular way. They fund Public schools through general taxation. Subsidised public universities and government-backed student loan programs generate substantial positive spillovers beyond what any individual student personally captures in higher wages.
Research and development is another classic case: a private company that invests in basic scientific research often can’t capture all the value that research eventually generates, since competitors and society at large frequently benefit from the resulting knowledge too (a phenomenon sometimes called the “free rider” effect on ideas), which tends to make private firms underinvest in basic research relative to what’s socially optimal. This is the standard economic justification for government funding of basic scientific research through public agencies and grants — not because government scientists are necessarily better at research than private ones, but because the spillover benefits of basic research are so large and so broadly diffused that no single private firm has sufficient incentive to fund it at the socially optimal level on its own.
The Coase Theorem: A Famous Counterpoint
Before we move on, it’s worth knowing about one of the most influential and counterintuitive ideas in this entire area: the Coase theorem, named after economist Ronald Coase. Coase argued that, under certain idealized conditions — clearly defined property rights, and sufficiently low costs of negotiating between the parties involved — private parties can actually negotiate their way to an efficient outcome on their own, without any need for government taxes or regulation at all, regardless of who initially holds the legal right.
Picture a simple example: a factory that pollutes. If property rights are clear (say, the factory has the legal right to operate) and negotiation between just these two specific parties is cheap and easy, Coase argued the laundromat owner could simply pay the factory to reduce its pollution, up to the point where the cost of doing so for the factory exceeds the benefit to the laundromat — arriving, through pure private negotiation, at the same efficient quantity of pollution reduction that a government tax might have aimed for in the first place.
This is a genuinely important and elegant insight. But its real-world applicability is sharply limited by a key qualifier baked right into the theorem i.e., low negotiating (transaction) costs. Climate change is the polar opposite case. It involves literally billions of affected parties (everyone on Earth, including future generations not yet born) and millions of polluting sources, making this kind of direct private negotiation between all the relevant parties essentially impossible. This is exactly why the Coase theorem is often invoked in economics courses as a genuinely brilliant theoretical benchmark. While government intervention (taxes, cap-and-trade, regulation) remains the dominant real-world tool for handling externalities at any significant scale, like pollution affecting an entire region or planet.
Public Goods: A Special and Important Case
Closely related to externalities — and worth its own dedicated discussion — is the concept of a public good. Economists define goods along two key dimensions:
Excludability: can you prevent people who haven’t paid from using the good? Rivalry: does one person’s use of the good reduce the amount available for others?
Most goods you buy every day are both excludable and rival — a sandwich, for instance. The shop can prevent you from eating it if you don’t pay (excludable), and once you eat it, nobody else can (rival). Economists call these private goods.
A public good is defined as both non-excludable (you can’t realistically stop people from benefiting, whether they pay or not) and non-rival (one person’s enjoyment doesn’t diminish anyone else’s). National defense is the textbook example: once a country has a functioning military protecting its borders and airspace, it’s protecting literally every resident, taxpayer or not, and your safety doesn’t subtract from your neighbor’s safety. Other classic examples include public street lighting, basic scientific knowledge once published and freely available, and clean air in a broad regional sense (closely related to the externality discussion above).
There’s also a useful intermediate category worth knowing: common resources, which are non-excludable but rival — think ocean fish stocks, or a public pasture used for grazing. Anyone can use them (non-excludable). But one person’s use does reduce what’s left for everyone else.
The Free-Rider Problem
Public goods create a particularly thorny challenge for private markets, known as the free-rider problem. Because a public good is non-excludable, individuals have a strong rational incentive to avoid paying for it while still enjoying its benefits, hoping that others will pay instead. If everyone reasons this way — and a perfectly rational, narrowly self-interested person genuinely would — the good ends up chronically underfunded or never produced at all, even though society as a whole would clearly benefit from having it.
Imagine trying to fund national defense through purely voluntary private donations. Most individuals would correctly reason that my personal contribution is a tiny drop in the bucket. If most people reason this way, total voluntary contributions fall drastically short of what’s needed, even though virtually everyone genuinely values having a national defense.
This free-rider logic is the standard core economic justification for funding certain goods through mandatory taxation rather than relying on voluntary private markets. By making payment compulsory for everyone through the tax system, governments sidestep the free-rider problem entirely. They ensure the public good gets funded at something closer to society’s true valuation of it.
Bringing It All Together
It’s worth stepping back to notice the common thread running through this entire post: in every single case we’ve discussed — pollution, vaccination, education, basic research, national defense — the fundamental problem is the same. The price signal that guides ordinary market transactions, the very mechanism we celebrated in our first two posts for elegantly coordinating supply and demand, simply doesn’t capture certain costs or benefits that spill over onto people outside the original transaction. Left entirely alone, markets handle private goods exceptionally well; they’re far less reliable when it comes to genuine externalities and public goods.
This is also exactly why “the market versus the government” framing that so often dominates everyday political debate is, from a careful economic standpoint, somewhat oversimplified. Most professional economists — across a genuinely wide range of political viewpoints — broadly agree that markets are remarkably effective at allocating most private goods, and that well-designed government intervention has a clear, well-justified role to play specifically in cases of externalities and public goods. The real, substantive disagreements tend to be about magnitude and method: how large is a particular externality, exactly, and is a tax, a cap-and-trade system, or direct regulation the most efficient tool for addressing it in a given case? These are genuinely difficult empirical and political questions. We will find serious economists landing in different places on many of them. But the underlying theoretical framework explaining why government intervention can improve on an unregulated market outcome in these specific cases is widely shared common ground.
What’s Next
We’ve spent this entire post, and the three before it, deep inside microeconomics — the study of individual markets, individual firms, and individual decisions, one transaction and one industry at a time. Starting with our next post, we’re going to zoom dramatically out and shift into macroeconomics — the study of entire economies, examined as whole, interconnected systems. Our first stop will be the most famous economic statistic in the world: GDP, or Gross Domestic Product. We’ll dig into exactly what it measures, what it deliberately leaves out, and why a single number gets so much attention from economists, journalists, investors, and politicians alike, every single quarter.