Research Highlights Article
October 8, 2026
Who benefits from upzoning?
A 2016 zoning reform in São Paulo increased the ability of developers to supply housing units by lifting limitations on permitted densification on a block-by-block basis.
Source: cifotart
The share of the world's population living in cities rose from 33 percent in 1960 to 56 percent in 2019 and is projected to reach 68 percent by 2050, with much of that growth concentrated in developing countries. Whether cities can provide sufficient housing for those future residents depends on local land use rules, which determine where and how developers can build.
But measuring the effects of these rules has proven difficult. Data limitations, complexity, and endogeneity have stymied efforts to estimate the value of zoning reforms.
In a paper in the American Economic Journal: Economic Policy, authors Santosh Anagol, Fernando Ferreira, and Jonah Rexer examine a 2016 zoning reform in São Paulo and related detailed administrative data to assess the impact of zoning policies.
"The really hard thing about studying zoning is getting fine-grained information on the actual rules that are enforced on the ground," Anagol told the AEA in an interview.
The central regulatory parameter for land use in São Paulo is the built area ratio (BAR)—floor space divided by lot area. The 2016 zoning reform stripped the city's 32 neighborhood governments of the power to set the maximum BAR, assigning new values block by block with the aim of densifying corridors served by metro, rail, and bus lines.
Across roughly 45,000 blocks, the average maximum BAR rose from 1.55 to 2.09, permitting an average of 36 percent more construction on a given parcel. However, the distribution of reforms was uneven. Forty-five percent of blocks received an increase of 1 BAR or more while nearly half saw no change or a reduction.
Raising the BAR
The chart below shows a map of changes in the maximum allowed BAR from 2004 to 2016. Green indicates an increase in BAR, while purple represents a decrease. Gray indicates no change.
The authors exploited this block-level variation using a boundary discontinuity design. Blocks whose maximum BAR increased were used as treatment groups, and the nearest blocks where the maximum BAR held constant or fell were used as controls.
The authors supplemented the building permit records with property tax records, real estate listings from 2019 to 2022, formal sector wage data, and a 2017 commuting survey of 24,800 households. They then estimated an equilibrium model in which developers choose how much to build, and households choose among 329 commuting zones, based on price, commuting time, access to high-paying jobs, building age, density, and neighborhood characteristics.
According to the authors’ estimates, developers responded quickly to zoning reforms. A 1.4-point increase in maximum BAR raised multifamily permit filings by roughly 66 percent relative to nearby control blocks, with differences emerging about a year after passage of the reform and growing thereafter. Notably, single-family permits were unaffected, consistent with BAR easing constraints primarily on larger structures.
Between 2019 and 2022, listings for sale grew about 10 percent faster in treated blocks. At the commuting-zone level, a 1-point BAR increase was associated with 10.9 percent higher listings growth and a 5.7 percentage-point reduction in prices.
While in the long run it may bring all prices down, the welfare gains, at least initially, may be concentrated amongst higher-income, higher-educated groups.
Santosh Anagol
By simulating the model for ten years following the reform, the authors estimated that the reform added roughly 39,000 units—a 1.6 percent increase in the housing stock—and reduced average prices by 0.4 percent. The authors suggest that this modest average can be explained by BAR rising only 0.54 overall. But neighborhoods receiving the largest increases saw up to 17.5 percent more units and 3 percent lower prices.
"While in the long run it may bring all prices down, the welfare gains, at least initially, may be concentrated amongst higher-income, higher-educated groups," Anagol said.
The reform produced a nominal wealth transfer of 11.33 billion reais, or 1.62 percent of city GDP, from current homeowners and landlords to future real estate owners. The size of the transfer suggests that it may generate political repercussions.
"The entrenched interests lose out from these housing reforms when prices go down, whereas new households benefit," Anagol said.
The work shows that targeted upzoning produces real supply responses and localized affordability gains. But the concentration of losses among existing owners may help explain why reforms of this kind remain difficult to pass.
♦
“Estimating the Economic Value of Zoning Reform” appears in the August 2026 issue of the American Economic Journal: Economic Policy.