Residential-commercial zoning is a defining feature of urban planning in the United States, with large areas of most American cities reserved primarily for housing and commercial activity concentrated in separate districts. The economic costs of this separation are rarely examined directly. The problem is not just that neighborhoods are segregated, but that the segregation is often backwards: land zoned for housing in many American cities would generate higher economic returns as commercial space, while commercial districts occupy areas where residents would be willing to pay a premium to live. Zoning locks land into its second-best use.
In this paper, the authors study how residential-commercial zoning affects the allocation of urban space, and whether those effects are for better or worse. They use neighborhood-level data on residents, workers, and floor space to infer the price gap between commercial and residential space in 34 large U.S. metropolitan areas, taking that gap as a measure of how much zoning distorts land use relative to what markets alone would produce. They compare US cities to Taipei, where zoning places few restrictions on the mixing of residential and commercial uses, and find the following:
- American neighborhoods are far more specialized than Taipei’s. US neighborhoods tend to be either predominantly residential or predominantly commercial, with relatively little mixing of the two. Neighborhoods in Taipei, by contrast, are substantially more mixed. The average aggregate specialization index across U.S. cities is 0.58, compared to 0.49 in Taipei—lower than the least specialized American city (Austin, at 0.53).
- Zoning shapes how space is allocated across neighborhoods far more than market forces alone would. Using neighborhood-level data on residents, workers, and floor space, the authors estimate how much of the difference in land use across neighborhoods reflects zoning rather than underlying characteristics like productivity or residential amenities. That zoning-driven variation is substantially larger in American cities than in Taipei, where land use instead reflects market forces like productivity or residential amenities.
- This zoning-driven variation reduces welfare. In a counterfactual where the degree of zoning-driven variation in each American city is reduced to Taipei’s level, neighborhoods become substantially more mixed and the welfare of the average resident increases by 9.3%. The gains range widely. Chicago has the most to gain (30.3%), while San Jose has the least (2.6%).
- Zoning does not address the issues it aims to. The traditional rationale for separating residential and commercial uses is that it reduces negative spillovers like congestion, noise, and pollution. An alternative view, associated with Jane Jacobs, holds that mixing generates positive spillovers through local services and neighborhood vitality. The authors find that where zoning is systematic, it is more consistent with the Jacobs view. Neighborhoods that market forces would push toward commercial use tend to receive more residential zoning, suggesting planners are trying to promote mixing. But this accounts for only a small fraction of actual zoning decisions, between 5% and 31% across cities. Most zoning bears no clear relationship to any neighborhood characteristic that would justify intervention.
- This poor targeting is costly. Even if all the mixing-oriented zoning were kept in place and only the portion unrelated to any neighborhood characteristic were removed, resident welfare in the average American city would still increase by 2.8%.
The paper’s central finding is a paradox. Although the systematic intent of zoning in American cities appears to promote mixed-use development, the large unexplained component swamps this signal, and the net effect of zoning is to increase the segregation of residential and commercial activity across neighborhoods. Understanding why zoning is so poorly targeted to neighborhood characteristics—and what political and institutional forces shape its allocation across space—remains an important open question.