Do externalities decompose into information problems and coercion problems, or are they a genuine third category? A sourced review.
Links: Economics, Value and Profit, Market Efficiency and Human Limits, Insurance, Legal Theory, Scope Confusion, The Supply Omission — the same “policy conclusion disguised as an analytical category” move on the supply/demand side; both are “delete the inconvenient half”
The claim: externalities are not a genuine third category of market failure but decompose into two already-understood problems:
This is a recognizable position with identifiable roots in Coase, Dahlman, the Austrian School, and parts of the Chicago/law-and-economics tradition. Below is what the literature actually says.
Ronald Coase’s 1960 paper “The Problem of Social Cost” (Journal of Law and Economics, Vol. 3) made two distinct arguments that are frequently conflated:
Argument A (the “theorem,” zero-transaction-cost case): If transaction costs are zero and property rights are well-defined, private bargaining leads to efficiency regardless of the initial allocation of rights. This was a heuristic device, not a policy prescription.
Argument B (the real-world argument): Transaction costs are never zero, so the initial allocation of rights does matter. Institutions should be designed to minimize transaction costs. The comparison should be between real-world alternatives, not between a flawed market and an idealized government.
Coase understood externalities as reciprocal problems — not “who is causing harm?” but “whose use of a resource should prevail?” He deliberately avoided the term “externality” because he felt it carried a built-in presumption of government intervention.
Coase himself was frustrated by how his work was interpreted — scholars have noted the profession interpreted his article essentially opposite to what he intended, reducing it to a “toy model” of exactly the sort he was criticizing.
Partially accurate, significantly overstated if used to dismiss externalities entirely. Coase identified conditions under which externalities resolve privately, while recognizing those conditions frequently don’t obtain. Using Coase to argue externalities “aren’t real” misrepresents his actual position.
Boudreaux and Meiners (2019), “Externality: Origins and Classifications,” Natural Resources Journal, Vol. 59, No. 1, pp. 1–34. Argue the term has “become nearly meaningless due to its ubiquity” and propose only unexpected spillovers not already reflected in prices should count.
Dahlman (1979), “The Problem of Externality,” Journal of Law and Economics, Vol. 22, No. 1, pp. 141–162. Argued that if bargaining costs exceed expected benefits, the externality is “in fact optimal.” Invoking ubiquitous externalities “simply constitutes a normative judgement about the role of government.”
Demsetz (1969), “Information and Efficiency: Another Viewpoint,” Journal of Law & Economics, 12:1, pp. 1–22. Introduced the “nirvana fallacy” — comparing real markets to an impossible ideal.
Carden and Horwitz (2013): Externality problems are market “failures” only by comparison to the perfectly competitive model’s equilibrium — they fail to live up to a “blackboard ideal.”
Minority position with serious proponents. Mainstream economics (Pigou, Arrow, Stiglitz, Mankiw) still treats externalities as a distinct category. The weaker claim — that many alleged externalities are already priced in or too loosely defined — has broader acceptance. The stronger claim that the category dissolves entirely is not supported by the profession at large.
Rothbard (1982), “Law, Property Rights, and Air Pollution,” Cato Journal, Vol. 2, No. 1. Pollution is a property rights invasion; 19th-century courts deliberately weakened tort protections to favor industrialization.
Block (1998, 2023), including “Addressing Air Pollution Through Property Rights and Nuisance Law” (Cosmos + Taxis, 2023, with Ritter). The state “props up polluters by degrading their nuisance liabilities.”
Epstein has also argued for common-law approaches, with more nuance than the Rothbardian position.
The evidence is largely unfavorable to a pure liability-based approach:
Has some historical basis — legal historians document that 19th-century courts did shift toward balancing pollution harms against economic benefits. But this doesn’t demonstrate that restoring a pre-industrial tort regime would work for modern diffuse, cumulative pollution (greenhouse gases, microplastics, PFAS).
Intellectually serious tradition, but empirically weak as a replacement for regulation. Tort faces structural problems. Most law-and-economics scholars now view tort as a complement to regulation, not a substitute.
Markets self-correct when:
Markets fail to self-correct when:
Not supported by the empirical record for credence goods and latent harms. Markets self-correct well for experience goods but fail systematically for credence goods, latent harms, and information suppression. The examples commonly cited for market self-correction often depend on an underlying regulatory framework.
Recognized minority position, not mainstream. Associated with Austrian School, Mercatus Center, Cato Institute, parts of Chicago School. Public choice theory has achieved mainstream acceptance for its core insight, but the stronger claim that most coordination failures are government-caused is a minority position. Standard economics (Mas-Colell, Varian, Stiglitz) treats externalities, public goods, information asymmetry, and market power as genuine market failures independent of government action.
The fact-check above evaluates the decomposition thesis against the literature and finds it empirically weak in places. But there’s a stronger version of the argument that the decomposition thesis doesn’t quite make — one that reframes what externalities are rather than arguing they decompose into other categories.
All trade has externalities by definition. Causality doesn’t stop at the transaction boundary. Every exchange ripples outward — my lawnmower purchase imposes noise on my neighbor, a trade in Monopoly shifts the competitive landscape for every other player, a factory’s output affects downstream water quality. Due to causality, nothing exists in a vacuum, and any transaction will always impact third-plus parties.
This means “externality” isn’t a special failure category — it’s a description of how causality works across transaction boundaries. Calling externalities a “market failure” is like calling gravity a “flight failure.” It’s not a bug in the system; it’s the medium the system operates in.
Markets price 1st-order consequences because those are what transacting parties care about and have information on. Higher-order effects are diffuse, uncertain, and often unmeasurable at the time of transaction. But they’re not ignored forever — they get internalized over time as they become visible and someone has incentive to price them.
What economists call “market failure” is often just internalization lag — the market hasn’t priced this yet. The lag is a function of two variables:
When both are high (noise from a lawnmower → identifiable neighbor → identifiable source), the externality is trivially managed. When both are low (diffuse pollution → millions of affected parties → cumulative sources), internalization takes longer. The empirical cases where markets “failed” (lead, asbestos, etc.) are cases where measurability was low and producers actively suppressed information, extending the lag artificially.
If externalities can be positive — and obviously they can (planting a flower garden brightens the neighborhood, a beekeeper’s bees pollinate surrounding farms, an educated workforce benefits employers who didn’t pay for the education) — then calling them “market failures” is incoherent on its face. A “failure” that makes everyone better off?
The asymmetry in how the term is applied reveals the bias: economists invoke “externalities” almost exclusively when they want to justify intervention for negative spillovers. Nobody proposes a government program to ensure optimal flower planting. But structurally, positive and negative externalities are identical — a third party affected by a transaction they weren’t part of. The selective application shows that “market failure” is a policy conclusion disguised as an analytical category.
It does: Reframe externalities from a failure category into a feature of causality that markets process over time. This removes the presumption that externalities require intervention and shifts the burden to demonstrating that intervention internalizes faster than the market would — accounting for the state’s own externalities (scope creep, regulatory capture, compliance costs).
It doesn’t: Solve the hard cases. Latent harms with information suppression (asbestos, lead) represent genuine internalization failures where the lag was artificially extended. The reframing doesn’t deny these cases exist — it argues they’re better understood as measurability + property rights problems than as a fundamental category of market failure.
Connection to the vault: The Multiplayer Coalition Problem already uses this framing — “trades have externalities… every trade has parties to the deal and parties affected by the deal.” In Monopoly, nobody calls third-party effects a “game failure.” The affected players respond, adapt, counter-trade. The system processes the externality through subsequent moves. The market works the same way — it just isn’t instantaneous.
The decomposition thesis has identifiable intellectual roots and is not fringe, but: