The Internal Revenue Service tracks movements of tax filers annually from county to county. If a filer listed their address as Orange County, CA in 2007, and Maricopa County, AZ in 2008, then that person is counted as a 2007-08 migrant from Orange County to Maricopa County. In the downloadable file, we have reorganized some of the IRS data into tables containing in-migration, out-migration, and net migration by MSA from 1992-93 to 2007-08. To do this we first netted out intra-MSA migration. The Northport-Bradenton-Sarasota MSA in Florida, for example, contains two counties, Manatee and Sarasota. We netted out the migration between Manatee and Sarasota. For the Tampa-St. Petersburg-Clearwater MSA, which has four counties, we netted out all the intra-MSA flows. Second, we checked for outliers. Roughly speaking, we checked all outliers larger than two standard deviations. When we found them, we compared out-migration from the MSA to in-migration to all other counties from that MSA. That comparison is limited because the IRS censors any county-to-county data for which there are fewer then ten migrants, and sometimes on other criteria. If that comparison made it clear that the number was wrong, we estimated it by looking at the median ratio of the numbers for in-migration over the years and using that to estimate a replacement for the erroneous number. You can tell when that was done by a number’s not being an integer. (We are exploring improved interpolation methods.) Third, we organized the data by MSA and by year, so that when you find an MSA, the numbers for all 16 years are conveniently available. Obviously these data have imperfections, the most important being that they are only tax filers. The proportion of migrants who file tax forms probably changes over time and varies across places. Also, for many analyses you would not want to use our interpolations. We have future plans to include individual origin and destination MSAs, and are considering including data about building permits and house prices. Please feel free to use this data set, prepared by Katherine Thomason and Henrique Romero, in any way you wish, with the caution that it is a beta version. Please direct inquiries and/or data suggestions to Henrique Romero, Assistant Research Economist, Bureau of Economic and Business Research, University of Florida, firstname.lastname@example.org.
Annual IRS In- and Out-Migration Data for MSAs, 1985-86 to 2007-08
Wednesday, July 21, 2010
Analysis Type and Topic: