Abstract:
|
The American Community Survey (ACS) is the largest collection of annual mobility data in the U.S. Data on residential mobility are used for community development, environmental planning and by the private sector. The data are inputs for federal funding programs, affecting millions of dollars in program allocations. These data must be complete and accurate, supporting analyses at low levels of geography. Tax data have long been used in domestic migration population estimates. Individual income tax returns include addresses for filers, but not all U.S. residents are filers. Published county-to-county migration estimates produced from IRS and ACS data do differ. This paper compares respondent-provided residence data in the ACS to a short panel of tax returns for the same individuals. Matching at the person-level, ACS current and previous year addresses are compared to tax return data. This paper explores coverage and matching issues involving tax data and data quality issues involving the ACS, looking to explain differences in estimates. An additional comparison to tax data for the following year provides a look at consistency and stability of tax data for migration studies.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.