The Density Dividend

Geoffrey West spent years measuring cities. His finding is simple: when a city doubles in population, its economic output increases by roughly 115 percent. Wages, patents, GDP — all follow the same curve, exponent of about 1.15.¹ A city of 10 million is measurably, predictably more productive per person than a city of 5 million.

Meanwhile, the infrastructure to support that city — roads, cables, gas stations — scales at about 0.85. You need less than double the infrastructure for double the people. Cities get more productive and more efficient as they grow. The regularity holds across cultures, continents, and centuries.

When I first encountered West's work, I recognized something I had been seeing on the ground across 30+ markets but did not have the language for. The reason the first 72 hours in a new city matter so much is that you are trying to read whether a city has crossed the density threshold where this superlinear math kicks in — or whether it is still too thin for a marketplace to compound.

Why density is the whole game

Edward Glaeser quantified the mechanism. Workers in US metros above 1 million people earn roughly 30 percent more than workers in smaller metros, controlling for education and experience.² Three sources: labor market pooling, input sharing, knowledge spillovers. Dense cities match the right worker faster, the right supplier cheaper, the right idea sooner.

Glaeser also found a threshold. No country has reached middle-income status — roughly $10,000 GDP per capita — without urbanizing past 50 percent.³ Urbanization precedes development, not the other way around.

Africa is urbanizing faster than any continent in history. The question is whether the density within those cities produces agglomeration gains — or whether it is just people stacking up without the connective tissue that turns proximity into productivity.

The Jacobs conditions

Jane Jacobs answered that in 1961 without any data science. She watched cities. Four conditions determine whether density generates diversity or just congestion.

Mixed primary uses — a district must serve more than one function so people show up at different times for different reasons. Short blocks — frequent turns multiply chance encounters and exchange. Buildings of varying age — old buildings provide the cheap space new ideas need. Sufficient concentration of people.

Her sharper argument came in The Economy of Cities. Cities grow not by exporting more but by learning to produce locally what they previously imported. Each round of import replacement generates new skills, new suppliers, new adjacent possibilities. The compound engine runs on density.

Walk through Plateau in Abidjan and you see the Jacobs conditions at work. Mixed commercial and residential. Short walking distances between clusters of activity. Old buildings housing small businesses alongside newer construction. Walk through parts of Kinshasa and you see density without those conditions — people present, but the connective tissue missing. Both cities have the population. Only one has the structure that converts population into the superlinear output West measured.

What this means for marketplaces

A marketplace business is a density accelerator. Every matched trip, every completed delivery increases the interaction rate in a city. West found that people in bigger cities walk faster — social metabolism accelerates with scale. A ride-hail platform does the same thing artificially. It compresses the distance between someone who needs to move and a driver who can move them. It speeds up the transaction clock of a city.

This is why city-level thinking matters operationally. You do not build density at the country level. You build it within a 15-kilometer radius. Whoever achieves density first in that radius wins — and the superlinear math means the returns to the first mover are not linear advantages but exponential ones. The second ride-hail player in a city is competing against the compounding effects of the leader's density, not for half the market. Stories of the second player overtaking the first almost do not exist.

I have watched this dynamic in city after city — Abidjan, Dakar, Lusaka. The first platform to reach critical mass wins the scaling law itself. Didi proved it against Uber in China. No amount of capital overcomes a density lead once the superlinear math is compounding for the incumbent.

Where the dividend is uncollected

The implication of West's math is that the unrealized productivity gains in dense emerging market cities are enormous and mathematically predictable. Lagos has the population density. It lacks the infrastructure efficiency and connective tissue to capture the full superlinear dividend. Nairobi, Dar es Salaam, Kinshasa — same story. The population is there. The agglomeration gains are sitting on the table uncollected.

Glaeser put it directly: the developing world will add roughly 2.5 billion urban residents by 2050. Getting urban governance and infrastructure right in those cities matters more than almost any aid program.

For a marketplace operator, the question is sharper. Which cities have the population density and the Jacobs conditions but lack the digital connective tissue? Those are the cities where a platform captures the most value — because it is not just matching supply and demand. It is unlocking the superlinear gains that the city's physical density already makes possible but that bad infrastructure, bad transport, and bad information have been suppressing.

In Abidjan, optimized routing returned hundreds of thousands of hours to commuters in a single year. Those hours went back into the economy as transactions, as work, as productive time that bad routing had been stealing. The density dividend, collected.

The financing loop compounds this further. Each vehicle added to the platform increases the city's transaction density. Each transaction generates data that makes the next vehicle financeable. The loop is a mechanical implementation of West's superlinear curve — each unit added produces more than one unit of additional output. The math says this should work. Thirty markets say it does.

Jacobs had a name for what happens when you get this right. She called it "new work" — economic activities that emerge not from planning but from the recombination of existing skills and needs in a dense environment. The brassiere manufacturer emerging from a dressmaker's knowledge. The fintech product emerging from ride-hail data. The boring business emerging from an institutional vacuum. None of it is planned. All of it requires density.