Large Lots, Segregated Cities

December 12, 2022

A new Job Market Paper by Tianfang Cui at the University of Pennsylvania studies the impact of exclusionary zoning through one specific and under-examined variable: minimum lot sizes.

Key takeaways:

  1. “From 1940–1970, the rise in central city Black composition in non-Southern central cities modestly accelerated minimum lot size adoption” along with other zoning restrictions.
  2. Meanwhile, migration of lower-income white households to the same cities had no observable relationship with minimum lot size regulations.
  3. The intent to enforce racial exclusion was often explicit: “In states that passed early legislation to desegregate public schools, Black migration had the largest effects on lot size restrictiveness.”

Historians and social scientists have long documented the racist roots of exclusionary zoning in discriminatory mortgage policies and white property owners reacting against influxes of Black and brown migrants. New research by Cui (2022) shows how minimum lot size restrictions formed part of this segregationist effort in response to the postwar Great Migration. 

As industrial jobs attracted Black migrants from Southern states under Jim Crow regimes, those job seekers encountered local governments that were similarly hostile to racial integration.  Cui explains: “For postwar local governments, minimum lot sizes were the main legal tools to control housing development densities.” 

By controlling densities, local governments effectively set a price floor on the local housing stock relative to incomes, virtually guaranteeing that lower-wage Black migrants could not afford to live nearby. Cui finds a strong correlation in national panel data of minimum lot size ordinances from 1945-1970, and his model estimates that “local regulatory responses to Black migration caused at least 600,000 units outside the South to be built less densely than what market forces could have supplied.” Cui also notes: “I find minimal evidence that explanations not dependent on migrants’ racial composition explain the effects.” 

In other words, the data is explained by racism, not just exclusion of the working poor writ large.

Cui’s algorithm takes inputs from data on 86 million homes built in the US that essentially reverse-engineers the likely development patterns that would occur if minimum lot sizes were enacted with the intent of racial exclusion. The algorithm is able to accurately predict both the time when minimum lot sizes were enacted and their relative restrictiveness. Its output effectively confirms Cui’s hypothesis for the empirical data. 

“To build in a neighborhood where the control applied, each housing unit must occupy some minimum square footage of land. A developer who maximizes profits from dense housing may still build units there, but divide land per unit around the smallest allowed lot size. If enough developers reason this way, the minimum lot size has an observable implication: over the distribution of lots across time, development bunch repeatedly on certain lot sizes.” 

In other words, the changes over time in the size and lot coverage of new homes were reliable indicators of minimum lot size restrictions.

Perhaps most significantly, Cui’s study is able to test and reject alternative “color-blind” hypotheses that would suggest racist outcomes were merely some sort of coincidence. One could anticipate critics arguing that “Black migrants may have self-selected to cities with accelerated suburbanization patterns,” introducing a selection bias; or that discrimination against lower incomes was also intended to exclude lower-income white Southern migrants. 

But as Cui shows, these alternatives would not be able to explain empirically observed outcomes. First, Cui adds instrumental variables to the model that “weigh the impact of migration rate shocks in Southern Black counties by cities’ pre-1940 exposure to county migration flows” — which enables him to observe the impacts of Black migration independently of a given region’s suburbanization patterns. 

The findings are startling: each 10 percent of Black population growth in non-suburban cities is associated with a 20 percent increase in restrictive minimum lot sizes and “caused at least 615,000 postwar housing units out of the South to follow binding lot size controls on density.” 

Additionally, while Southern white migrants were poorer than their Northern counterparts, Cui observes “null or negative effects of changes in Southern white composition on suburban lot size outcomes … Because poorer white migrants still earned more than Black migrants, differentiated lot size control responses by migrant race cannot isolate specific race-dependent motives for zoning.”

Rather than income inequality or suburbanization itself, school integration efforts were linked to more restrictive minimum lot sizes. Cui lastly adds to his model “variation in states that, by 1950, adopted a model law that bans discrimination by race in public schools.” The hypothesis that this form of exclusionary zoning “could circumvent de jure limits on school segregation, which could have raised the neighborhood’s amenity value” is thus confirmed by finding “the highest causal effects of Black demographic change on lot size outcomes in states that adopted these laws early.”

Because these findings identify a clear intent to enforce racial segregation through a specific policy intervention, Cui argues that today’s policymakers should take those historical intentions at face value. These findings, he writes, “suggest both equity gains and public good provision gains are possible with targeted legislation overriding local density controls, like California’s Senate Bills 9 and 10 in 2021 or new upzoning regulations in Massachusetts in 2022.”