A method to align heterogeneous migration data in the US census and American Community Survey
摘要
This paper designs a method to align migration flows that have heterogeneous time horizons in the US census and American Community Survey microdata dating back to 1940. The method enables credible conversion of five-year flows to one-year flows, leveraging the availability of both types of migration data in one year of the sample period. Resulting estimates of prior-year migration can replicate established trends such as the recent decline in interstate migration, thus supporting estimate validity. Given the enabled wide sample period and broad availability of survey covariates, this alignment technique can serve as a useful resource for migration research.