Interstate elderly migration is a central argument in state policy debates regarding tax
incentives, and the ‘return migration’ of the elderly who need assistance has public health
implications. Yet little is known about how patterns of interstate elderly migration have changed
in the 21st century; our study attempts to fill that gap. The replacement of the Census Long Form
(CLF) with the American Community Survey (ACS) requires us to devise a methodology for
reconciling the differences between the two data sources and creating comparable migration
measures. Two commonly used data sources for U.S. migration research -- the Current
Population Survey (CPS) and Internal Revenue Service (IRS) -- prove inadequate for studying
the migration of subpopulations such as the elderly. However, because they span 1980-2010,
they aid in our methodology and help illuminate if detected changes in migration are genuine or
instead an artifact of using the ACS. We find that the ACS can generate comparable migration
data that reveals a continuation of previously identified geographic patterns plus changes unique
to the 2000s. The small number of migrants the ACS yields, however, weakens its usefulness for
analyzing annual migration patterns or for small population states. Most troubling, its changed
definition of residence and survey timing leaves us unable to answer definitively the basic
question of whether elderly migration has increased, decreased or stayed the same in the 21st
century