Why Cities Stopped Growing Up and Started Growing Out

In 1916, New York City passed the first comprehensive zoning ordinance in the United States, partly in response to the Equitable Building, a forty-story tower completed the previous year on Lower Broadway that cast a permanent shadow over nearby properties. The ordinance introduced setback requirements and height limits calibrated by street width. It was a reasonable technical response to a genuine problem. Within two decades, almost every major American city had copied it, often without the genuine problems that motivated the original.

That copying was the beginning of a catastrophe that would take another century to fully register.

American cities are among the least dense wealthy cities on earth. Greater Los Angeles, which most people imagine as sprawling by definition, has a population density of about 7,000 people per square mile. Greater Paris, which most people imagine as a compact European capital, has nearly 21,000. The difference is not geography. The San Fernando Valley is flat. The Seine basin is not dramatically different terrain. The difference is what the two governments decided to permit, and when.

The economic consequences of this difference are substantial. In a 2017 paper, economists Chang-Tai Hsieh and Enrico Moretti calculated that if American cities with high productivity had simply allowed housing supply to grow at the same rate as cities like Atlanta and Houston from 1964 to 2009, US GDP would have been 9.5 percent higher. That is not a rounding error. That is roughly two trillion dollars in annual output that does not exist because San Francisco decided it preferred its Victorian streetscape to new apartments.

The mechanism is straightforward, which makes the political failure harder to excuse. Productive cities pay high wages. High wages attract workers. Workers need housing. When housing supply is constrained, the price of housing rises until it chokes off migration. The workers who would have moved to San Francisco to work in software, or Manhattan to work in finance, or Boston to work in biotech instead stay in places where their skills are less valuable. They are not worse people. They are priced out of the opportunity. The productivity loss is the value of the work they would have done minus the value of the work they actually do, compounded over a career.

This is not a controversial claim among economists. Ed Glaeser at Harvard, Alain Bertaud formerly at the World Bank, Jenny Schuetz at the Brookings Institution, and dozens of others have reached essentially the same conclusion through different methodologies over the past twenty-five years. The argument that zoning restrictions cause housing scarcity, which causes high prices, which reduces economic mobility, is about as settled as empirical economics gets.

What is contested is the politics. And the politics are, at their most basic level, a story about who votes.

Homeowners in desirable American neighborhoods are, as a demographic group, older, wealthier, and more reliably politically active than renters. They attend city council meetings. They respond to neighborhood association surveys. They show up for planning commission hearings. They have, in many cities, effectively captured the local planning apparatus and converted it into an instrument for protecting their property values by restricting competition.

This is not a conspiracy. It does not require bad intent. It requires only that homeowners correctly perceive their interests (rising home values, stable neighborhood character, reduced traffic and parking pressure) and act on them through the political process. The problem is that their interests are directly opposed to the interests of the people who are not yet in the neighborhood: the young workers who cannot afford to move in, the service workers who commute three hours a day because housing near their jobs is unaffordable, the families who cannot access good schools because they cannot access the neighborhoods those schools serve.

The excluded have no vote in local elections because local elections require residency, and residency requires being able to afford to live there.

Houston is the standard counterexample to this story. Texas’s largest city has no traditional Euclidean zoning code. It uses deed restrictions instead, which expire and require collective action to renew. As a result, Houston has built housing at a rate that other large American cities have not matched. From 2010 to 2020, Houston issued more building permits than the entire state of California. Median rent in Houston is roughly half what it is in Los Angeles, despite Houston’s economy growing faster over the same period.

The Houston model is not without problems. The city has serious flood risk issues that permissive development patterns have made worse, as the catastrophic flooding during Hurricane Harvey in 2017 demonstrated. But the core housing affordability achievement is real and replicable. The cities that have not replicated it have made a choice, even if the people making the choice have not framed it that way.

Minneapolis did frame it that way, explicitly, in 2040. The Minneapolis 2040 comprehensive plan eliminated single-family zoning citywide, permitting up to three units on any residential lot. The political process was contentious. Neighborhood associations fought it. The Minnesota Center for Environmental Advocacy filed a lawsuit arguing that the city had not properly analyzed the environmental impact of allowing more housing. The lawsuit delayed implementation for several years.

The environmental objection to urban density is perhaps the most intellectually dishonest argument in American planning debates. Dense cities have dramatically lower per-capita carbon emissions than sprawling ones, because dense cities support walking, cycling, and transit while sprawling ones mandate driving. A resident of Manhattan emits roughly a third of the greenhouse gases of an average American, largely because of transportation patterns. Building apartments in San Francisco and keeping people out of car-dependent suburbs is one of the most effective climate interventions available. Using environmental review requirements to block urban infill housing is, in climate terms, the opposite of environmentalism. It is a legal process that produces environmental harm while wearing environmental language.

The trajectory here is not entirely pessimistic. After a decade of slow reform, several states began preempting local zoning authority in the early 2020s. California passed dozens of housing bills between 2017 and 2024, some of which have meaningfully accelerated construction. Montana eliminated single-family zoning statewide in 2023. Washington and Oregon made similar moves. The federal government has increasingly conditioned transportation and infrastructure funding on local zoning reform, using federal leverage to do what local politics could not.

Whether these reforms are sufficient, and whether they arrive fast enough to matter for the workers already locked out of high-opportunity cities, is genuinely uncertain. The Hsieh-Moretti calculation captures historical losses. The forward-looking question is how much productivity the country will recover, and how quickly.

The answer depends on whether American cities can solve a political problem that their institutional structures make very hard to solve: giving adequate weight to the interests of people who are not yet residents, who do not yet vote in local elections, who need housing that does not yet exist. That is an unusual kind of governance challenge. It requires cities to act against the immediate preferences of their current constituents in order to benefit people who are not yet constituents.

The cities that figure out how to do that will be richer, denser, and more economically dynamic. The cities that do not will become progressively more expensive museums of their own past, inhabited only by the wealthy and the very long-tenured, surrounded by sprawl that serves neither the environment nor the people forced into it.

The 1916 zoning ordinance was a response to a shadow cast by a building. The shadow that bad zoning casts over American cities is harder to see but much larger.