The Effects of Changes in Geographic Coding Methods on Estimated Administrative Record Coverage, Migration Rates, and Estimates

Rodger V. Johnson, U.S. Census Bureau
Esther R. Miller, U.S. Census Bureau
Hyo C. Park, U.S. Census Bureau
Barbara J. van der Vate, U.S. Census Bureau

The U.S. Census Bureau annually estimates domestic migration change among the Nation’s states and counties, as a component to the demographic change model. Exemptions on IRS tax returns are assumed to be representative of the under 65 population. Two consecutive years of matched IRS returns are used to calculate migration rates. A key contributor to accurate migration rates is the quality of the geographic assignment of state and county codes to each matched set of tax returns. The Census Bureau has tested a new geographic coding file to assign more accurate county-level geographic codes from the TIGER System, the Census Bureau’s geographic base used to support censuses and surveys. This paper presents a summary and review of selected findings from testing of the new method upon estimated coverage, migration rates, and estimates. The results demonstrate operational and quality improvements in the resulting data.