How arbitrary boundaries distort urban segregation measurements in U.S. cities

A new study published in Nature Cities highlights the challenges of measuring residential segregation in American cities. The research shows how arbitrary boundaries in political and statistical geographies can skew segregation estimates. This issue, known as the modifiable areal unit problem (MAUP), has long complicated efforts to assess segregation accurately. Traditional segregation research often relies on fixed geographic units like Census tracts. These predefined areas can introduce bias, as segregation indices are sensitive to where boundaries are drawn. The study used redistricting algorithms to simulate millions of alternative spatial configurations.

In smaller cities, the findings revealed substantial fluctuations in segregation estimates when boundaries changed. However, as city size increased, the impact of redrawing boundaries diminished, leading to more stable results. The study also noted that official Census tract boundaries typically fell near the average of the simulated distributions.

The researchers developed a framework to account for this uncertainty in segregation statistics. By acknowledging the spatial ambiguity in aggregate data, they aim to provide a clearer picture of segregation patterns. The study’s findings have practical implications for researchers, policymakers, and activists working in urban planning, housing, and civil rights. Understanding boundary variability helps refine segregation measurements. This could lead to more informed decisions in addressing residential inequality in cities.