Measuring gentrification: A comparative evaluation of quantitative methods in Washington, DC

Source abstract

Gentrification is a central yet contested topic in urban studies, especially in debates concerning displacement and racial equity. Varied definitions and methods have resulted in inconsistent empirical findings. This study evaluates five widely cited quantitative methods for identifying gentrification at the census tract level in Washington, DC. Using Receiver Operating Characteristic (ROC) curve analysis, we assess the construct validity of each method using change in mortgage lending, professional/managerial/technical employment, and non-Hispanic White population. Results show substantial variation in model performance, with three methods, Ellen & O’Regan, Ding et al. and Mitchell et al. demonstrating the strongest alignment with these empirical indicators. Between 1990 and 2020, our estimates indicate a net loss of more than 62,000 Black residents from tracts classified as gentrifying, accounting for over half of the District’s Black population decline. This study underscores the importance of cross-validation to ensure that gentrification measures meaningfully capture racialized neighborhood change.