Katie Genadek, University of Colorado
John Sullivan, Census Bureau
Carlos Bacerra, U.S. Census Bureau
Place of birth information from surveys and administrative records is often used as a measure of childhood exposure to environment and policy conditions that follow geographic boundaries. However, survey and administrative records sources may provide inaccurate or misleading measures of place of birth and in order to use birthplace as a measure of exposure, researchers should understand the childhood migration in that time period. In this paper, we use data from the Social Security Administration’s Numerical Identification file, the 2000 and 2010 censuses, the American Community Survey, and state vital records, to assess the quality of place of birth information and to examine its utility as a measure of childhood residential location. We use string reports of place of birth from applications for Social Security Numbers coded to standard numeric state and county codes to investigate multiple sources of potential bias in the use of place of birth as a measure of childhood residence, including discrepancies between reported place of birth and actual residence and selective early childhood migration away from place of birth. We describe the development of these data and document the differences in birthplace and residence and migration away from residence at birth.
No extended abstract or paper available
Presented in Session 14. Measurement Issues in Historical Data