external to what?

Every policy brief on climate change, every ESG framework, every carbon market — all rest on one concept: the externality. Costs imposed on third parties not captured in the price. Market failures to be corrected.

The concept is so embedded that questioning it feels like questioning gravity. But the word is younger than most people assume — it entered economics in 1957, attached to the claim that the effects it named were unimportant. Follow it back and a question emerges that unravels the frame:

External to what?

what is an externality?

An externality is a cost or benefit from an economic activity that falls on someone who is not party to the transaction, and that the price does not capture. Air pollution from a factory; the value a neighbor's woodlot adds to your view.

The word comes from Latin externus — "outside, outward, foreign." So it carries a spatial claim before it carries any economics: there is an inside (the relevant system) and an outside (everything else). Whatever gets labeled external is, by construction, not part of the system being analyzed.

That boundary is not found in nature. It is drawn by the analyst, and where it is drawn determines what counts, who pays, and who absorbs.

the intellectual genealogy of the externality

The concept is older than the word, and the gap between them is where the trouble starts.

Malthus (1798) and Mill (1848) had noticed the effects before Sidgwick made them a class. Kapp, writing a decade before Coase, judged economic activity against human needs and ecological indicators rather than market efficiency, and wrote "social costs" because in 1950 "externality" did not exist. The concept had run 150 years without needing a noun. Two entries do most of the work.

Bator coined the word to dismiss what it named. The noun arrives with an MIT economist in "The Simple Analytics of Welfare Maximization" (1957) and "The Anatomy of Market Failure" (1958); Paul Samuelson used it at nearly the same moment, but Bator's use spread. The irony, documented by historian of economic thought Steven Medema: Bator introduced the word while echoing Tibor Scitovsky's view that such effects were unimportant. The term that now carries the weight of environmental economics labeled a category its author considered exceptional and minor. It was never built to hold what we ask it to hold.

Coase is closer to an ally than an opponent. "The Problem of Social Cost" is the most-cited article in legal scholarship by Shapiro and Pearse's all-time ranking, and the strawman version has him believing in frictionless bargaining. He did not — the zero-transaction-cost world was a baseline showing what the real one lacks. Coase never used the term "externality," because he thought it smuggled in the conclusion that state action was required. Kapp and Coase agreed on almost nothing except this: the word gets in the way. His own limit is where this argument turns — zero transaction costs and clean property rights never hold for diffuse, multi-party ecological processes. You cannot negotiate with the atmosphere. The aquifer has no representative at the table.

where the word breaks

Marshall's "external" came with a qualifier — outside the firm, inside the industry — and when the noun was coined seventy years later, the qualifier did not come with it.

Outside the firm became outside the transaction, then outside the economy, then in practice the biosphere. Nobody argued for the enlargement or defended it. That is the etymological payload: a bounded technical term lost its bounds and kept its authority.

the pervasiveness collapse

Pigou's framework works when these effects are rare. Tax the factory. Cap the emissions. Problem solved. But what happens when they are everywhere?

Robert Costanza and colleagues estimated the global value of ecosystem services at $33 trillion per year — a 1997 figure in 1994 dollars, range $16–54 trillion, against a global GNP then of roughly $18 trillion. Their 2014 update revised it to $125–145 trillion per year in 2007 dollars. Treat these as order-of-magnitude. They were contested on publication, and the sharpest objection is real: aggregating marginal values into a planetary total is not a defensible use of the method. Even critics who reject the arithmetic accept the direction — the unpriced share is not small.

"Not a single bee has ever sent you an invoice." — Pavan Sukhdev, on the launch of the TEEB report, October 2010

Bees are the legible case. Gallai and colleagues (2009) put the global value of insect pollination to agriculture at €153 billion in 2005 — usually rounded to "$200 billion" in TEEB-era coverage.

At that scale the concept collapses under its own weight. When most of what the economy depends on sits in the external column, the externality is not the exception — the priced transaction is. You do not have a system with occasional externalities. You have an accounting system that captures a fraction of reality and labels the rest "outside."

who draws the boundary — and who pays

The deepest critique is not about magnitude but authorship: who draws the boundary? What counts as internal, what gets banished to external, and whose interests does that line serve?

Jim Lunday puts the structural version in the Cadmus Journal: the concept of externalities "divides the world hierarchically, with the internal dominant and the external subordinate." That division, in his reading, grants "the scientific legitimacy to conquer, manage and use other peoples and the natural world (the external) instrumentally." A power relation wearing a technical term.

On the ground: for more than fifty years, oil operations in Ogoniland, in the Niger Delta, produced costs that appeared in no barrel price. The 2011 UNEP assessment found hydrocarbon contamination in wells across ten communities; at Nisisioken Ogale, families were drinking water with benzene — a known carcinogen — at over 900 times the WHO guideline. UNEP concluded restoration would take 25 to 30 years and recommended an initial $1 billion fund for the first five. Every dollar of that was an "externality" while it accrued. The Ogoni people were the outside. They did not draw the boundary and were not asked about it. Ken Saro-Wiwa and eight others were executed in 1995 for organizing against it.

This is what Joan Martinez-Alier calls the environmentalism of the poor — ecological conflict driven not by post-material preference but by people defending the material basis of their livelihoods. The externality frame is least equipped to see it, because its grammar treats the affected as third parties to somebody else's transaction rather than principals in their own. A concept describing who bears uncounted costs was built almost entirely by economists in Britain and the United States, and the people most often assigned to the external column had no hand in defining it.

the physics, stated carefully

There is a thermodynamic version of this argument. It is a good one, usually stated wrong.

Earth is closed to matter and open to energy. Energy arrives from the sun and leaves as infrared radiation to space, so "nothing ever leaves" is false as stated, and the First Law — energy is conserved — is not the law doing the work.

The Second Law is. Energy and materials degrade in quality as they are used: concentrated becomes diffuse, available becomes unavailable. Nicholas Georgescu-Roegen built The Entropy Law and the Economic Process (1971) on exactly this, arguing that the economic process is irreversibly entropic rather than the closed mechanical loop textbook diagrams imply. For matter the point is mass balance: the atoms in an emission or a tailings pile are still here, moved from a form we could use into one we cannot.

One caveat, because it is the sentence a physicist would go after: cost is not a thermodynamic quantity, and no accounting identity falls out of the Second Law. The physics establishes something narrower but still decisive — the material consequences of economic activity have nowhere to go. The economy is a subsystem of the biosphere, and a subsystem has no outside to put things in.

So when an economist says "externality," the honest translation is: a cost inside the physical system, outside the boundary we drew around our transaction. We declined to assign it to our ledger.

why internalizing externalities fails

The orthodox response is internalization — put a price on it. Carbon markets. Biodiversity credits. Payments for ecosystem services.

One conflation to correct first: a carbon tax is the Pigouvian instrument. Cap-and-trade is not — tradable pollution permits come from the property-rights tradition, formalized by J.H. Dales in Pollution, Property and Prices (1968) on foundations Coase laid. The dominant climate policy of the last thirty years descends from Pigou's critic, not Pigou.

Pricing requires translating ecological function into money, which assumes three things:

• Fungibility — one tonne of CO₂ equals any other, one hectare of forest equals any other. Ecological function is place-specific, relational, and often non-substitutable.
• Marginal analysis applies — small changes, small adjustments. Climate change and biodiversity collapse are non-marginal. Peter Victor's 2020 appraisal identifies precisely this failure: monetary valuation of non-marginal changes is vulnerable to the fallacy of composition. Tipping points do not yield to marginal tax adjustments.
• Nature can be an input — recasting the living world as "natural capital" frames it as a factor of production.

Victor's conclusion, after reviewing the concepts, methods, and data behind monetary valuation: such valuations "should be used with great caution or not at all" in public policy with environmental consequences. His sharpest point should worry anyone in this business: conceptualizing nature as capital invites exploitation of the rest of nature. Once a forest is denominated in dollars, it can be compared to a shopping mall — and outbid.

does ensurance price nature?

Yes, and the critique above applies to us. RealValue computes ecosystem service value in dollars per acre per year across a matrix of ecosystem stocks and service flows. That is monetary valuation. The position, and its limit:

• The number is instrumental, not ontological. A dollar figure on a wetland says what it would cost to replace or maintain a function, not what the wetland is worth. Treat those as the same and Victor's critique lands on us.
• The price is a bridge, not a verdict. Its job is to make a dependency legible to capital that would otherwise never see it, so protection can be funded. Naturalizing finance, not financializing nature.
• We size premiums, not exchange rates. RealValue answers "what does it cost to keep this functioning, and who depends on it?" — not "what may this place be traded for?" A certificate is tied to a specific asset for that reason: it is deliberately bad at fungibility.
• The limit is real. Any number that can fund a forest can also price one for sale. Good intentions do not constrain that; structure does — asset-specific instruments, no offset accounting, transparent methodology.

Anyone who claims to have dissolved this tension is selling something. We have chosen which side to err on and built the instruments accordingly.

what works instead of internalization

None of these alternatives dissolve boundaries. They relocate them, legitimize them, or make them explicit rather than leaving them to the analyst.

One correction to a claim often made in our own direction: Georgescu-Roegen did not ground value in exergy. He coined "energetic dogma" as a rebuke to precisely that position. Exergy accounting belongs to Robert Ayres, material throughput to Herman Daly; his contribution was the entropic framing, and he was scathing about energy theories of value built on it.

ensurance's move

Ensurance starts from a premise: ecological function is internal to all value. Not external to the economy. Not a market failure to correct. The foundation the economy runs on.

On one real parcel: Lookout Mountain is 1,729 acres of forest, riparian corridor, and wetlands in the Blue Mountains, draining to the Grande Ronde and on to the Snake and Columbia. Assessed condition 8 of 10. Its measured service flows come to $15.0 million per year against an asset cost of $5.35 million — a natural cap rate of 281%. Clean water alone accounts for $1.5 million per year; erosion control, $1.5 million; pollination, $3.7 million. The premium to move the whole parcel to ENSURED is $247,500 — roughly 1.6% of one year of the flows it produces.

That is the argument in one line item. No hypothetical harm to cost out, no counterparty to find, no credit issued that lets someone pollute elsewhere. The question is narrower and answerable: who depends on this watershed, and what does it cost to keep it working?

An ensurance certificate is not an offset or a credit. It does not price a harm. It funds the maintenance of a dependency — a specific place, a specific function, backed by an assessment you can inspect — and routes value to the agent account that holds responsibility for that place.

the linguistic opportunity

"Externality" is a word worth teaching against. Every time someone uses it, there is an opening:

External to what?

The answer almost always reveals the assumption: the economy is the inside, nature is the outside. Ensurance inverts it. The living world is the inside. The economy is a pattern that runs on top of it.

When you stop calling ecological costs "external," you stop needing to "internalize" them. They were never outside. They were unfunded. Unaccounted. Deferred.

If you're putting this to work Monday: run the substitution in your own documents. Anywhere a memo, model, or disclosure says "externality," write "cost we assigned to someone else" and see whether the sentence still holds. Where it doesn't, you've found a dependency worth pricing properly — start with the stocks and flows behind a real parcel, or talk to someone who does this work.

the series

This is part of a series on the words we use to avoid funding what matters.

• there are no externalities — the unified argument
• there is no risk transfer — insurance's founding euphemism
• the appreciation protocol — depreciation and its inverse
• $1 now or $4 later — deferred maintenance and nature's backlog
• price is what we pay — price, cost, and value as three separate decisions
• external to what? — the boundary buried in a word (you are here)