A Method for Estimating Neighborhood-to-Neighborhood Migration Flows from County-to-County Flow Data

Zack W. Almquist, University of Minnesota
Jack DeWaard, University of Minnesota
Susan Brower, State of MInnesota
Liqi Huang, University of Minnesota
Scott Chase, University of Minnesota

Scholars of place-to-place migration flows must balance the need for population-level data and estimates with theoretical and substantive issues that arise from treating administrative boundaries as meaningful existential categories. They must also contend with source data that are incomparable, incomplete, or non-existent. Given these issues, in this paper, we detail a novel approach for deriving population-level estimates of tract-to-tract migration flows to approximate neighborhood-to-neighborhood migration within the United States using county-to-county flow data. Our starting point is with data on annual county-to-county flows provided by the IRS. Leveraging information on county and tract characteristics, we will employ small area estimation methods to impute tract-to-tract migration flows for 1990, 2000, and 2010. After calibrating these estimates, we will then impute tract-to-tract migration flows each year from 1991 to 2010. Relative to other aerial units, tract-to-tract migration more closely approximates existential space, and possesses intuitive linkages to numerous local processes.