A Distance Metric for Modeling the Quality of Administrative Records for Use in the 2020 U.S. Census

Abstract

The U.S. Census Bureau is conducting research on using administrative records to reduce the cost while maintaining the quality of the 2020 Census Nonresponse Followup (NRFU). Previous census tests have implemented approaches that use predictive models and optimization procedures to identify vacant and occupied housing units using administrative records. This article details a modification to previous approaches, introducing a simple distance metric to define a quality ranking of housing units to enumerate using administrative records. The distance approach is illustrated, assessed, and compared to a previous approach via a retrospective study of the 2010 U.S. Census.