Long-run evidence from Stockholm reveals no link between levels of inequality and elite residential segregation

Rising income inequality has been linked to increasing residential segregation, with profound consequences for social mobility, educational opportunity, and public health. With almost half of the world's population now living in cities this link is of key interest to both researchers and policymakers, yet existing studies examine short periods with monotonically rising inequality, making it impossible to establish whether inequality drives segregation or both respond to common causes. Here we analyse 122 years of geocoded population data from Stockholm (1900–2022) at 100-meter resolution, exploiting Sweden's unique inequality trajectory -- high in the early twentieth century, compressed to historic lows by the 1980s, then rising again -- to test this relationship. We find that elite segregation increased monotonically throughout the period, including during the postwar decades when the top quintile's income advantage over the middle class fell by half. This decoupling demonstrates that elite residential sorting operates independently of income inequality. The poorest quintile followed a contrasting trajectory more consistent with economic models: highly segregated in 1900, integrated by 1985 as incomes compressed and public housing expanded, then re-segregating as inequality rose after 1990. These asymmetric dynamics reveal that fundamentally different mechanisms govern elite enclave formation versus poverty concentration, with important implications for policies aimed at promoting residential integration.