The gentrification intensity draws on ~800 metropolitan areas. The indicators displayed in the panel consistently linked to class upgrading were selected through principal component analysis of ~400 variables.

Factor analysis combines these into a class status score and a class change score, both benchmarked against regional averages so each tract is evaluated relative to its own metropolitan context.

A tract is classified as gentrified if it started below its regional average, experienced above-average upgrading, and ended above the regional average.

The map covers ~56,000 census tracts, built from Decennial Census and American Community Survey data via the National Historical GIS and IPUMS API.

Two formats are available: The primary version maps gentrification from 1990 to 2020, crosswalked to 2020 tract boundaries using the NHGIS Geographic Crosswalk methodology.

The historical version extends coverage to 1970, crosswalked to 2010 boundaries using both NHGIS and Longitudinal Tract Database methods, covering ~49,000 tracts. This version introduces more spatial noise and is best suited for historical research only.

Replication data and code are available on Harvard Dataverse (Lauermann et al. 2026).

On the repository, R and Python scripts can be run to reproduce the full dataset. Download the folder structure, add your free IPUMS API key, and execute from any directory.

Finished data tables and boundary layers are also available for direct download without running the code.

Lauermann, John, Viggiani, Alice, Wu, Yuanhao, & Smash, Nathan (2026) National Gentrification Intensity Map: Mapping gentrification across US communities, 1970 to 2020, The Professional Geographer, https://doi.org/10.1080/00330124.2026.2625975

Lauermann, John, Alexander, Zoe, & Wang, Ziqi (2025) Mapping super-gentrification in large US cities, 1990-2020, Urban Geography, https://doi.org/10.1080/02723638.2025.2528418