The Greedy City

The bigger the city, the saying goes, the more people have and the less it satisfies them. A house wider than the neighbor's; the nagging need to at least keep pace. We tend to give this a moral name. Greed. The city is an arena where the grasping climb over one another, and its hardness is, in the end, a vice of the individual heart.

Two accounts usually prop up this conventional wisdom. One is that the city stains the people in it: where advertising and shop windows and other people's eyes are everywhere, the mind soaks in consumerism. The other runs the opposite way—the greedy and the ambitious were drawn to the big city to begin with. The city makes people greedy, or the city gathers greedy people. Either way, greed stays a problem of the individual.

But what if greed is something the city manufactures? And manufactures with the very apparatus that makes the city rich? If the engine that makes the city wealthy and the engine that makes its people feel small are in fact one engine, then what we have been calling an individual vice turns out to be something else.

The City Runs Two Ledgers of Opposite Sign

Start with how the city grows its wealth. It is well established that the more populous a city, the higher its wages per person and its patents per person. The usual explanation is "economies of scale." Just as a bigger household runs leaner, a bigger city runs more efficiently. Half true. But in the other half, the city does something no living organism ever does.

The rule for how much some quantity grows as size grows is called scaling. When population doubles, a quantity that grows exactly twofold is linear; one that grows less than twofold is sublinear (economies of scale, thrift); one that grows more is superlinear (amplification). Compress that slope into a single number and you get the exponent. Below one, sublinear; above one, superlinear.

Life is a sublinear world. The larger the body, the less energy each single cell spends. Basal metabolic rate is roughly proportional to body mass to the three-quarters power—an exponent below one, sublinear (the exact value is an approximation). It works this way because a central pump, the heart, delivers resources through vessels to every last cell. As scale grows, this delivery network bottlenecks, and the cost of keeping one unit alive falls.

A city's physical infrastructure behaves the same way up to this point. Roads, gas stations, power lines are sublinear: double the population and you do not need to double them. But turn to what people make, and the sign flips. Wages, patents, total output—when population doubles, output per person, far from shrinking through economies of scale, rises. In one body, a thrift ledger (infrastructure) and an amplification ledger (economic output) run at once, with opposite signs.

What flips the sign is not size but connection. In an organism, resources travel one way, from the heart to passive cells. A city has no such central heart. Its cells—people—do not wait for a delivery; they meet, trade, and pass ideas back and forth. Less a parcel service than a peer-to-peer marketplace or neighbors trading favors. As population grows, these chances to exchange—the connection density—grow faster than population itself. In one city's telecommunications network, the sheer volume of interactions between people was observed to rise superlinearly, as roughly the 1.17 power of population. The force with which a city grows its wealth superlinearly comes not from size but from this closeness.

That is the skeleton of the original story. The city grows wealth superlinearly through the density of mutual exchange. But a question remains. If the city grows wealth like that, what happens to the mind that feels that wealth? Wealth is never felt as an absolute sum. It is always felt against someone else, and who that someone is changes in the city.

When a Richer Neighbor Is Right in Front of You

Say you earn $60,000 a year. In one neighborhood that is a comfortable living; in the heart of Manhattan it is a meager one. Same number, different feel. That is because people register their own condition not against an absolute level but against some baseline. That baseline is the reference point, and the tendency for judgment to cling to it rather than to absolute magnitude is reference dependence, a concept set down half a century ago.

Reference point · reference dependence · concepts set down in Daniel Kahneman and Amos Tversky's prospect theory (1979).

What happens when the reference point rises? One more thing compounds this. For a loss of the same size, people feel the sting more sharply than the joy of an equivalent gain—roughly twice as much, a tendency called loss aversion. As the reference point rises, yesterday's identical absolute income slides toward the loss region, and though nothing has been taken away, it comes to feel like a shortfall.

That "twice" was measured with one's own current state as the reference point, weighing gains against losses. Carrying it over to social comparison, where the neighbor is the reference point, is this article's analogy—not a claim that the same experiment measured exactly twice as much in social comparison too.

So what sets the reference point? People compare themselves to others in order to evaluate themselves, and above all to others like them (social comparison). Beyond that, some values come not from absolute magnitude but from relative rank. It is zero-sum: your rise is my fall. At a concert, the person in front stands up to see better, so the person behind stands too, and in the end everyone is on their feet, no one sees any better, and only their legs ache. This is an example of the positional good. A race for rank cancels itself out this way and wears everyone down.

Social comparison · Leon Festinger (1954).

Positional good · named by Fred Hirsch (1976); example from Robert Frank.

So far this is a story of the laboratory and of theory. These experiments and theories were not about cities. What they uncovered is the wiring of the mind, and laying that wiring against the density of the city is this article's interpretation. But there is one field study that laid the chain directly onto the space of the city. Working from US household data, one economist showed that the higher the income of neighbors living in the same area, the lower a person's self-reported happiness—even after controlling for that person's own income. The effect is not small. A rise in neighbors' earnings cut happiness by about as much as an equally sized fall in one's own income. And the effect was large among people who actually socialize with their neighbors, and weak among those who socialize only with friends outside the area. The reference point is the neighbor right in front of you.

This field study · economist Luttmer's (2005) work on neighbors' income and happiness.

One seam is worth pausing on. You might object that the window of comparison today is not over the fence but the screen in your hand, and that the real reference point is set by media, not by the city. There is something to that—but the distinction that field study's data draws is exactly where the two part ways. What cut into happiness was not the celebrity on the screen but the neighbor people actually socialize with. Physical proximity narrows the reference group to people who are like you and seem within reach rather than strangers on a screen, and because they seem within reach they hurt more. What the city does is less to invent a new reference point than to pull the reference point toward this near and narrow side. Still, one caveat still stands. What that field study showed establishes that neighbors' income matters—not the finer claim that density scales the effect proportionally. The direction is clear, but the dose-response of "how much" is still only thinly drawn.

Now the chain closes. The denser the city, the closer and more often a richer neighbor comes into view (social comparison). The more that happens, the higher the reference point rises (reference dependence); the higher it rises, the further the same absolute income slides into the loss region (loss aversion); and the deficit grows accordingly. What is worth noting is that this deficit is not a side effect spinning off separately from the wealth engine. The closeness in which people meet and measure themselves is the very variable that grew wealth superlinearly, and comparison is output that same engine turned out alongside it. The closeness with which the city grows wealth superlinearly is at the same time what makes that wealth feel small.

Let me be clear about one thing. This article does not claim, as an aggregate, that "greed grows as some power of population." The bridge between the city and the mind is not an exponent but a causal chain. The city as a variable (density) pushes up an input to the individual mind (the reference point), and individuals so altered add up into the texture of a society. This is a different order of thing from a story that extrapolates experiments in individual psychology straight onto city statistics.

One Engine, Two Faces

Here is where this article plants its feet. When a richer neighbor is in view, I do two things at once. I feel smaller (relative deprivation), and I strive harder (ambition).

Here the article takes one step further and proposes something. That striving to overtake the reference point may itself be fuel that feeds the density of interaction—working longer hours, taking on more risk, gathering into ever closer places. But this is not a fact the scaling literature has established; it is this article's hypothesis. The superlinearity of innovation can turn independently of any striving for status, on pure knowledge spillover alone: pack people close together and ideas spread on their own. So this fuel hypothesis can be set aside.

Set it aside and the argument still stands, because it stands on a claim narrower in scope than the hypothesis but harder in evidence. The closeness that makes the city rich is a matter of proximity and visibility, of people near one another and often in view. That same closeness pushes up the reference point behind that deficit. That the variable which grew wealth and the variable which grew deficit are one and the same—this weaker reading alone is enough to make "the same density" hold. Reference dependence is the point where that density touches the human mind, not a flaw wrongly lodged in the engine. The ambition that makes the city rich and the greed that makes its people feel small are, in this way, two faces sharing the same density-driven disposition.

The original story already held the seed of this symmetry—another of the paper's findings, the one set aside earlier. If the city grows ideas and wages superlinearly through closeness, then epidemics and overload should ride that same closeness and grow too. This is the prediction that the creativity machine is also an amplifier of pathology, a dark symmetry. That symmetry now replays inside the human mind. The city gnaws at satisfaction to the same degree that it grows wealth superlinearly. As income rises the reference point rises with it, so that however fast you run on the treadmill you stay in place. This running in place is called the hedonic treadmill. The old observation that even lottery winners who came into a windfall turn out, before long, no happier than the neighbors who did not win points to this running in place.

The exact exponent on the pathology side turns out to be contentious under more rigorous statistics.

This lottery-winner observation rests on so small a sample that it is not strong evidence, to be taken only as an image.

This argument comes with a few boundaries worth marking in advance. First, the article puts no exponent on greed. It is the same honesty with which the original story withheld a settled exponent from the superlinearity on the pathology side. What the article asserts is not a value but only the direction of the engine.

One passage calls for particular care. The point is not to say "city people are objectively unhappier." There is strong evidence that higher income generally brings higher happiness. One study showed that both within nations and across them, absolute income correlates clearly with happiness, and that no satiation point appears at which happiness stops rising (plateaus) as people grow richer. This may look at odds with the treadmill above, but the two measure different axes. The treadmill and the lottery winners are a story on the vertical axis, that of a single individual adapting to change over time; that study is a story on the horizontal axis, of many nations laid side by side at one moment, income weighed against happiness. That absolute income raises welfare and that the mind soon adapts to the income it has gained do not collide. What is contentious comes next: whether relative income offsets and cuts into that welfare. That aggregate question has not been settled either way. What this article stands on is not that aggregate conclusion but the facts that reference dependence and social comparison are robust in the laboratory and the field, and the direction that density raises the visibility of the reference point.

This study · Stevenson and Wolfers (2008).

On this aggregate question, Easterlin's long-standing paradox and its rebuttals remain evenly matched.

What about selection? The objection that ambitious people flock to the big city in the first place. Selection is real. But the weight of the case rests not on the covariation but on what remains after controls. In large-scale Spanish administrative data, even after stripping out each individual worker's hidden ability, roughly half of the wage effect that city size confers does not vanish; it stays. That neighbor effect, too, is a figure net of the person's own income and individual traits. Which means density does its own work, apart from ability or disposition. This does not erase selection, of course—but trying to explain even this residual by selection alone gets strained. The observation that good things (ambition, innovation) and bad things (greed, deficit) move in the same direction cannot, on its own, nail down a common engine (since a single status-seeking disposition tilting toward the city would produce the same covariation). That observation is only corroborating; the weight is carried by density's share, which survives the controls.

Nor does this engine run evenly across the city. Who reaps the gains of closeness (higher wages) and who bears its shadow (a heavier deficit) splits along class and neighborhood. A scaling curve that lumps the city into a single point cannot show the distribution inside it.

The City That Manufactures Greed

So the phrase "the greedy city" comes to mean something other than it did at the start. It is not a city where greedy people have gathered but a city that manufactures greed. The sense of deficit we had been calling an individual vice was a structural shadow—one left by the very connection density that makes the city rich, cast on a mind that leans on a reference point. The density that grew innovation grew this along with it.

Two things follow. First, moralizing that blames the individual has the wrong address. Individual differences are real, of course, and it is true that the temperate are less easily swayed. Granting the direction does not make personal responsibility disappear. But reference dependence is not willpower; it is the mind's default wiring, and density pushes up the baseline that wiring stands on across an entire population. This is not an opponent you can defeat by preaching contentment. Second, you cannot take only the good part. As long as the density that grew innovation grows deficit along with it, a prescription that keeps only innovation and cuts out only deficit ends up fighting the engine itself.

The lever, then, is not restraint of the mind but the structure of connection. It is a question of who is made to compare what with whom, how the visibility of the reference point is arranged, and how the gains and shadows that closeness produces are shared. The problem is not individual greed. It is the geometry of density. And that is where the repair belongs too.

The original story said the city is not a large organism but stands as its mirror image. Where a body that delivers resources from the center economizes sublinearly, a city of mutual exchange amplifies superlinearly. Now one more thing can be added. The city does not only write the rules of life backward. It writes the rules of satisfaction backward too.