Introducing the 2020 to 2010 Census Tract Crosswalk
Although many census tract boundaries do not change every 10 years, in Philadelphia 38 out of 408 (9.3%) tracts had > 2% area change between 2010 and 2020. LTDB allows researchers to include these tracts that changed in their studies.
February 10, 2025
The Urban Health Collaborative (UHC) at Drexel University has released an essential new tool for researchers using census data: the 2020 to 2010 Census Tract Crosswalk. This dataset is a valuable addition to the Longitudinal Tract Database (LTDB), enabling longitudinal comparisons of demographic data despite changes in tract boundaries over time.
The LTDB, which was developed by Dr. John R. Logan and collaborators at Brown University’s American Communities Project, provides a groundbreaking solution for researchers working with historical census data. The LTDB includes tract-level estimates harmonized to 2010 boundaries, covering data from as early as 1970 to the 2020 census. It also offers tools to align non-census data—such as public health, criminal justice, and voting records—with consistent tract boundaries.
Census tract boundaries are modified every decennial census – meaning every ten years the boundaries that define a designated census area change. This complicates longitudinal studies and introduces a common geographic information system (GIS) woe, the Modifiable Areal Unit Problem (MAUP): that how you draw the boundaries of a geographic space can influence the patterns observed in the data, leading to potentially misleading conclusions. To address this and harmonize the altered boundaries, the crosswalk uses two variables in its weighting procedure: areal overlap and total population, both of which are derived from 2020 census block data. These weights are fractional values that can be applied to 2020 tract-level data and indicate the portion of the data that can be applied to each overlapping 2010 tract. Then, this weighted data can be aggregated at the 2010 tract level to obtain final estimates.
This dataset was developed by Stephen Francisco, PSM, and Steve Melly, MS, MA, at the UHC with support from the MESA Neighborhoods & Aging study (R01AG072634, PI: Hirsch).
The crosswalk and supporting code, documentation, and input files that were used to create the crosswalk can be accessed via the LTDB website.