Media: The Atlantic (September 2022), New York Times (June 2023)
Since 1970, the share of Black individuals living in suburbs of large cities has risen from 16% to 36%. We first show that Black suburbanization has led to major changes in neighborhoods, accounting for a large share of recent increases in both the average Black individual’s neighborhood quality and income segregation within the Black population. We then use an accounting exercise to show that changes in relative suburban amenities and housing prices explain about 60% and 30%, respectively, of Black suburbanization, while regional reallocation, changing educational attainment, and gentrification of Black city neighborhoods play only minor roles.
Neighborhoods are a central concept in urban economics, but the best way to define their boundaries and scale is unclear and depends on the context. Despite this ambiguity, the vast majority of neighborhood research in the U.S. uses the same unit: the census tract. This may limit our understanding of racial and economic segregation, neighborhood effects on individual outcomes, and neighborhood policy. We construct a new, larger, neighborhood geography by applying a community detection algorithm to granular migration data. Motivated by a revealed preference intuition, the algorithm creates groups of five to ten census tracts that see a disproportionate amount of cross-migration. We call these units “districts” and show that they align well with highways, municipal borders, sharp demographic changes, and some city-created neighborhood boundaries. To illustrate potential applications, we show that a traditional tract-level analysis obscures important elements of neighborhood sorting and understates racial disparities in neighborhood quality, while a buffer approach attenuates the estimated relationship between broader neighborhood composition and tract outcomes. We plan to publicly release the district boundaries when the project is complete.
We use administrative data to document a high degree of migration across neighborhoods and neighborhood types defined in terms of poverty rate and median income. Neighborhood quality increases over an individual’s life cycle, and people also move to better neighborhoods in response to earnings improvements. We then develop several implications of these initial facts for high-poverty neighborhoods. First, resident turnover in these areas is rapid. Second, among the people living in a poor neighborhood at a point in time, the distribution of future concentrated poverty exposure is bimodal. Young people, renters, and those with children tend to spend fewer than half of the next ten years in similar neighborhoods, while older people and homeowners are unlikely to exit concentrated poverty. Third, poor neighborhoods tend to remain poor because of a dynamic process in which initial residents experience high earnings growth but disproportionately out-migrate when earnings improve, contrasting with a pure “poverty trap” understanding of persistent concentrated poverty.
Local population decline has spread rapidly since 1970, with half of counties losing population between 2010 and 2020. The workhorse economic models point to net out-migration, likely driven by changing local economies and amenities, as the cause of this trend. However, we show that the share of counties with high net out-migration has not increased. Instead, falling fertility has caused migration rates that used to generate growth to instead result in decline. When we simulate county populations from 1970 to the present holding fertility at its initial level, only 10 percent of counties decline during the 2010s.
Media: Sightline Institute (September 2022)
Local control of land-use regulation creates a not-in-my-backyard (NIMBY) problem that can suppress housing construction, contributing to rising prices and potentially slowing economic growth. I study how increased local control affects housing production by exploiting a common electoral reform—changing from “at-large’’ to “ward” elections for town council. These reforms, which are not typically motivated by housing markets, shrink each representative’s constituency from the entire town to one ward. Results from a variety of difference-in-differences estimators show that this decentralization decreases housing units permitted by 20%, with similar effects on multi- and single-family permits. Effects are larger in whiter and higher-income towns.
Media: Financial Times (September 2023), Crain’s Detroit (August 2019), New York Times (September 2019), Minneapolis Star-Tribune (January 2020), CityLab (June 2019), NPR Rochester (June 2019), CBC Vancouver (June 2019), City Observatory (April 2019), Strong Towns (April 2019), Conversable Economist (July 2019).
I illustrate how new market-rate construction loosens the market for lower-quality housing through a series of moves. First, I use address history data to identify 52,000 residents of new multifamily buildings in large cities, their previous address, the current residents of those addresses, and so on for six rounds. The sequence quickly reaches units in below-median income neighborhoods, which account for nearly 40 percent of the sixth round, and similar patterns appear for neighborhoods in the bottom quintile of income or percent white. Next, I use a simple simulation model to roughly quantify these migratory connections under a range of assumptions. Constructing a new market-rate building that houses 100 people ultimately leads 45 to 70 people to move out of below-median income neighborhoods, with most of the effect occurring within three years. These results suggest that the migration ripple effects of new housing will affect a wide spectrum of neighborhoods and loosen the low-income housing market.
Media: New York Times (February 2020), Los Angeles Times (April 2019), Bloomberg (January 2020), Globe and Mail (January 2020), City Observatory (June 2019), Strong Towns (January 2020), Curbed SF (January 2020).
We study the local effects of new market-rate housing in low-income areas using microdata on large apartment buildings, rents, and migration. New buildings decrease rents in nearby units by about 6 percent relative to units slightly farther away or near sites developed later, and they increase in-migration from low-income areas. We show that new buildings absorb many high-income households and increase the local housing stock substantially. If buildings improve nearby amenities, the effect is not large enough to increase rents. Amenity improvements could be limited because most buildings go into already-changing neighborhoods, or buildings could create disamenities such as congestion.
Media: Bloomberg (February 24, 2020), Newsday (February 11, 2018), Albany Times-Union (July 15, 2017).
I analyze how competition between localities affects tax breaks and firm location decisions. Using data on firm-specific property tax exemptions in New York State, I begin by documenting that spatial competition substantially increases tax breaks in a town. To do so, I exploit variation in the number of counties near a town, which is correlated with the level of competition but uncorrelated with other observable characteristics. I then use this pattern to estimate a model of localities competing for mobile firms by offering tax breaks. In counterfactual exercises, I find that policies that reduce competition between localities, such as restricting which levels of government may offer tax breaks, lower exemptions by up to 30%, but have very little effect on equilibrium firm locations. These findings suggest that local tax break competition primarily lowers the tax rate on mobile firms and is unlikely to substantially improve the efficiency of firm location.
Federal place-based policy could improve efficiency if it targets areas with large amenity or agglomeration externalities. We begin by showing that positive shocks to federal spending in a county and their associated economic stimulus substantially decrease crime, an important amenity. We then employ two machine learning algorithms—causal trees and causal forests—to conduct a data-driven search for heterogeneity in this effect. The effect is larger in below-median income counties, and the difference is economically and statistically significant. This heterogeneity likely improves the efficiency of the many place-based policies that target such areas.