How Does Racial Segregation Shape Economic Mobility for Black and Low-Income White Children?

This study examines how citywide racial segregation affects intergenerational economic mobility for Black and White children. Using IRS-linked data from the Opportunity Atlas and an instrumental variable strategy based on 19th-century railroad placement, the authors find that a one standard deviation increase in segregation lowers the adult income rank of Black children from the poorest families by 4.4 percentiles (16% of average mobility), and by 3.3 percentiles for low-income White children. Segregation also increases incarceration and teenage birth rates, reduces test scores, and lowers government spending. The findings underscore that segregation significantly restricts upward mobility, especially for disadvantaged groups.

Full Citation

Chyn, E., Haggag, K., & Stuart, B. A. (2025). The effects of racial segregation on intergenerational mobility: Evidence from historical railroad placement. American Economic Journal: Applied Economics. Advance online publication. https://doi.org/10.1257/app.20220371

Extended Summary

Central Research Question

This paper addresses a longstanding but empirically underexplored question: How does citywide racial segregation causally affect intergenerational economic mobility? While prior research has documented that opportunity varies sharply by geography and that Black children consistently experience worse outcomes than their White peers, the mechanisms underlying these disparities remain unclear. This study seeks to identify the causal impact of racial segregation—distinct from compositional or correlated factors—on the long-run economic prospects of children, with particular focus on race and parental income. The authors explore whether greater segregation lowers upward mobility and investigate mechanisms through which segregation may exert its effects.

Previous Literature

The study builds upon several foundational strands of economic and sociological research. Earlier work has highlighted persistent geographic disparities in intergenerational mobility (Chetty et al., 2014), noting that upward mobility for Black children is lower even after controlling for income. Descriptive research from Wilson (1987) and Massey and Denton (1993) argued that racial segregation limits access to opportunity, but causal evidence linking segregation to economic mobility remained limited.

Cutler and Glaeser (1997) and Ananat (2011) used design-based strategies to estimate the effects of segregation on outcomes like poverty and urban inequality, but did not evaluate long-term child outcomes due to data limitations. The current study innovates by using newly available IRS-linked intergenerational data from the Opportunity Atlas and an instrumental variable (IV) approach based on historical railroad placement to identify the causal impact of segregation on mobility. The work also complements studies on neighborhood effects (Chetty and Hendren, 2018) and adds to the growing literature on the long-run impacts of racial inequality and historical institutions.

Data

The analysis draws on a comprehensive set of data sources:

1. Opportunity Atlas: Provides intergenerational mobility measures by race and parent income percentile for children born between 1978 and 1983, tracking adult income in their mid-30s using IRS tax data.

2. Stanford Education Data Archive (SEDA): Offers test score data for grades 3–8 across U.S. school districts, disaggregated by race.

3. Incarceration and Teenage Births: Measures derived from IRS data and Census records to assess downstream behavioral outcomes.

4. Historical Railroad Maps: Used to construct the Railroad Division Index (RDI), capturing how railroad placement in the 19th century physically subdivided cities.

5. Surveys and Fiscal Data: Cooperative Congressional Election Study (CCES), American National Election Studies (ANES), and Census of Governments data were used to analyze public spending and political attitudes.

The study focuses on 121 Northern and Western U.S. metropolitan areas where railroad configuration data were available and links these to contemporary and historical datasets at the metro level.

Methods

The core empirical strategy relies on a two-stage least squares (2SLS) IV approach. The authors instrument for contemporary racial segregation (measured via the dissimilarity index) using the Railroad Division Index (RDI), which captures the extent to which cities were divided into physically separate neighborhoods by historical railroad tracks. The assumption is that railroad-based segmentation in the 19th century led to greater racial sorting during the Great Migration and persisted through the 20th century.

The primary outcome variable is upward mobility, defined as a child’s national income percentile in adulthood conditional on parental income percentile. The regressions are run separately by race and parental income percentile (1st, 25th, 50th, 75th, and 100th), allowing for heterogeneity in the impact of segregation. Additional analyses explore effects on incarceration, teenage births, and test scores.

To assess mechanisms, the authors examine how segregation correlates with public spending, redistributive preferences, and racial attitudes. They also conduct a decomposition to isolate how much of the impact on mobility is driven by “exposure effects” (the cumulative effect of time spent in a place) versus other channels like sorting or place-specific shocks.

Findings/Size Effects

The study’s central finding is that racial segregation significantly reduces intergenerational mobility, particularly for Black and low-income White children.

• For Black children, a one standard deviation (SD) increase in racial segregation leads to a:

  
  

  • 4.4 percentile drop in income rank for children from the 1st percentile of the parental income distribution (16% of baseline upward mobility).

  • 3.9 percentile drop at the 25th percentile, and continued significant declines at the 50th and 75th percentiles.

  
  

• For White children, segregation:

  
  

  • Reduces mobility by 3.3 percentiles for those from the 1st percentile of parental income.

  • Has negative effects at the 25th and 50th percentiles.

  • Has a small positive effect for children from the 100th percentile, implying segregation may preserve privilege at the top.

  
  

Additional outcomes show broad negative effects of segregation:

• Incarceration: For Black boys from the poorest families, a 1 SD increase in segregation raises incarceration by 6.7 percentage points (29%). For White boys, the increase is 1.4 percentage points (22%).

• Teenage childbirth: Segregation increases the probability of a teenage birth by 11 percentage points (22%) for Black girls and 6 percentage points (23%) for White girls from low-income families.

• Test scores: A 1 SD increase in segregation reduces average test scores by 0.13 SD for Black students and 0.07 SD for White students.

Mechanism analysis reveals that segregation affects both the supply of and demand for public goods:

• Public spending: Segregation reduces total per capita government expenditures by 39%, with large cuts to education, health, and public safety.

• Political attitudes: In more segregated metros, White residents show:

  
  

  • 0.45 SD higher opposition to redistributive policies.

  • 0.68 SD more racially conservative attitudes (e.g., opposition to affirmative action, support for aggressive policing).

  • These effects are larger among low-income White respondents, despite segregation harming them economically.

  
  

The decomposition of mobility effects indicates that:

• About 39% of the negative mobility effect for children from the 1st percentile stems from reduced “exposure effects” (i.e., how segregation affects place quality).

• The remainder reflects other mechanisms, such as sorting or direct discrimination.

• Results differ from Chetty and Hendren (2018), who found most of the variation in mobility came from exposure effects alone.

The authors distinguish their findings from prior work on the Great Migration (Derenoncourt, 2022), showing that segregation itself—not just the influx of Black migrants—independently reduces mobility for both Black and White children.

Conclusion

This study provides compelling causal evidence that racial segregation meaningfully reduces intergenerational economic mobility, particularly for children from marginalized racial and economic backgrounds. The effects are economically large, statistically robust, and extend beyond income to include incarceration, teenage pregnancy, and educational achievement. Segregation also undermines public investment and increases racial and class-based resentment, further entrenching inequality.

The findings imply that residential segregation is not merely a historical artifact but a continuing structural barrier to upward mobility. Policymakers seeking to promote equality of opportunity should consider reforms aimed at reducing segregation and mitigating its effects—such as expanding housing access, enforcing integration policies, and ensuring equitable public investment.

More broadly, the paper underscores the need to rethink place-based disadvantage through a racialized lens. As cities across the U.S. grapple with housing affordability, school segregation, and uneven economic development, these findings reinforce the idea that geography and race remain central to understanding and addressing inequality in America.