Is the economy zero-sum?

A suspicion worth taking seriously

The thought usually arrives with a number attached. A pay statement sits next to a headline about a record executive package, or a rent increase lands in the same week as a landlord’s third property purchase, and the arithmetic suggests itself: if they have more, it came from somewhere, and the somewhere looks like me. Held one way, this is the oldest intuition in economic life. Held another way, it is a precise empirical claim about where fortunes come from, and it can be tested.

The claim deserves better than the two reflexes it usually meets. One reflex dismisses it: wealth is created, trade benefits both sides, envy explains the rest. The other reflex embraces it: every fortune is extraction wearing a suit. Both reflexes fail for the same reason. The economy is enormous, and different parts of it work differently. Some of its margins create value for both parties in every transaction. Others move a fixed prize from one set of hands to another, and no amount of growth changes that. An honest answer to the chapter’s question has to say which margins are which, because the reader lives on several of them at once.

What an exchange creates

Start with the mechanism that makes the dismissive reflex partly right. A voluntary trade happens because the two sides value the same things differently. A commuter values the fourth coffee of the week at more than its price; the café values the price at more than the coffee. Both walk away holding something they prefer to what they gave up. Nothing was conjured from air, yet the trade left more satisfied preferences in the world than it found. Economists call the gap between what a buyer would have paid and what they did pay consumer surplus, and its twin on the seller’s side producer surplus. The surplus is the created value, and it exists only because the exchange happened.

The accounting carries one condition: it counts the people at the table. When a trade pushes costs onto someone who never agreed to it, a neighborhood breathing a factory’s smoke, a river fished empty downstream, buyer and seller can both gain while the full ledger goes negative. Absent uncompensated effects on third parties, voluntary exchange creates value; where those effects are large, even a willing trade can subtract from the total.

The mechanism compounds through specialization. Because exchange exists, a person can spend the whole week doing the one thing they do best and trade for everything else, and output rises for reasons that have nothing to do with anyone working harder. The surgeon does not grow her own food, and the farmer does not remove his own appendix; both are richer for the arrangement, and the added output belongs to no one’s loss column. Most of what separates a modern income from a medieval one is this compounding, iterated across billions of specializations.

The same logic scales up. It is the reason trade exists between towns and between countries: differences in costs, endowments, and wants mean that the act of exchanging leaves both places better supplied than self-sufficiency would. When later chapters ask what a trade deficit means or who won from globalization, the mechanism under examination is this one, operating at continental size (the accounting lives with Volume III: Countries, in the chapters on external accounts and on regions that lost their industries).

The pie, measured over a millennium

If the economy were zero-sum in aggregate, the total would be roughly fixed: enrichment here would require impoverishment there, and the world average could not move much. The world average has moved. Reconstructions of output per person, assembled by the Maddison Project from tax rolls, harvest records, and national accounts, show a line that barely lifts for eight centuries and then climbs almost vertically after 1820.

A fixed pie cannot do this. Whatever else is true about distribution, the total that gets distributed has grown by an order of magnitude, and most people alive today are net beneficiaries of that growth rather than victims of it. This is the strongest single fact against the universal form of the zero-sum claim, and it is worth staring at before turning to the margins where the claim survives.

The growth is not an artifact of long horizons; it is visible inside a single reading lifetime. By the World Bank’s extreme poverty line, more than a third of humanity lived below it in 1990 (38%) and under a tenth did by 2019 (8.5%; World Bank). Whoever was getting rich across those three decades, the poorest were not the ones paying for it in aggregate. The word matters: in aggregate. Aggregates are where the zero-sum claim dies, and the reader’s daily life is not lived in aggregate, which is why the next section is the important one.

The margins that really are zero-sum

The intuition survives the aggregate evidence because parts of daily economic life do work the way the intuition says. Three margins matter most, and the reader meets all three between waking and sleeping.

Located land. The number of plots within reach of a city’s good jobs is nearly fixed. When the city prospers, the prosperity is bid into the price of those plots, and every dollar of the increase is paid by someone who does not own one to someone who does. Growth does not add supply; it adds price. This is the engine behind the rent question, and its full treatment, including the evidence from cities that permit building and cities that do not, lives in Volume III: Countries, in the chapter on housing in rich cities. What matters here is its classification: the land margin is genuinely zero-sum in the short run, and the reader’s sense of paying tribute on it is not arithmetic confusion.

Position and status. Some goods are valuable only because others lack them: the corner office, the admission letter from a college that rejects most applicants, the address that signals arrival. Fred Hirsch, in Social Limits to Growth (1976), named these positional goods and observed that growth cannot expand them, because their supply is defined by rank. A society can double its output and mint no additional top decile. Competition for positional goods absorbs real resources and returns no aggregate gain; it is zero-sum by construction.

The split of the surplus. Even a value-creating exchange contains a distributive fight inside it. The surplus from the coffee is real, but where the price lands between the buyer’s maximum and the seller’s minimum decides who keeps how much of it. Scale this to a workplace: employer and employee jointly create value, and the wage decides the split. The split moves with bargaining power, and bargaining power is exactly the kind of thing one side can gain only as the other loses. Why pay tracks leverage, and what happens when leverage collapses, is the opening question of Volume IV: Work.

The billionaire ledger

The chapter’s question usually arrives pointed at the very top, so the top deserves its own accounting. Fortunes are built on identifiably different foundations, and the zero-sum verdict differs by foundation.

Fortunes built on innovation sit mostly on the positive-sum side, and the striking finding is how small the builder’s cut tends to be. William Nordhaus examined American nonfarm business over the postwar decades and asked what share of the value created by innovation was retained by the innovators as profit. His estimate: about 2.2%, with the remainder passing to consumers through falling prices and improving quality (“Schumpeterian Profits in the American Economy,” NBER Working Paper 10433, 2004). A fortune can be a rounding error on the value that produced it. That is what the study shows, and all it shows: it prices the surplus from technological advance across one economy’s postwar run. It does not sample billionaires, and it cannot say how much of any particular fortune traces to innovation rather than to the foundations described next.

Fortunes built on enclosure sit on the other side. Wealth assembled from monopoly position, from regulatory protection against competitors, from privatized public assets, or from the appreciation of fixed-supply holdings is wealth whose counterpart really is a loss distributed across everyone else, in overcharges, blocked alternatives, or rents. The mechanisms that make such positions durable, scale, networks, switching costs, and captured regulators, are the subject of Chapter 5, and the most extreme national cases, where a ruling clique’s fortune is the treasury, appear with the resource curse in Volume III.

The two foundations even leave different fingerprints, which gives the reader a usable test. A fortune built on creation tends to sit beside falling prices, expanding output, or new capabilities in whatever the fortune’s owner sells; the wealth and the customer’s gain moved together. A fortune built on enclosure tends to sit beside rising prices for an unchanged thing: the same spectrum, the same plots of land, the same drug decades after its discovery. The test is rough, and many fortunes mix the two, but it redirects the argument from the size of the number to the source of it, which is where the argument belongs.

So the ledger requires reading each fortune before ruling on it, which is precisely what the universal claim refuses to do.

Where the argument goes next

The zero-sum question is this volume’s foundation because every later chapter stands on one side of it or the other. Systems that organize exchange are machines for finding positive-sum trades; systems that organize privilege are machines for defending zero-sum positions; most real economies, as the next chapter shows, are both machines running at once. And the reader’s original question had a second half: when the zero-sum margins grow too large, when too much of what people pay flows to positions rather than production, what breaks first has a historical answer, recorded in Volume V: Pathologies, in the chapter on when inequality breaks a regime.