Abstract:
|
This paper examines estimates of the U.S. foreign-born population -- Census survey estimates of citizens and noncitizens and DHS estimates of three noncitizen subsets (legal permanent residents, legal temporary residents, and unauthorized) -- finding a discrepancy of 3.4-4.4 million in each of the years 2007-2012. The DHS implied estimate of noncitizens, based on both Census survey information and DHS administrative records, is always larger than the Census estimate. Insights from recent immigration literature about behavioral and coverage differences between pre-1980 and since-1980 entrants, coupled with availability of separate Census and DHS estimates for the two periods, make it possible to designate some of the estimates as preferred and derive their logical implications, which may be regarded as reconciled estimates. The reconciled estimate of citizens is smaller than the Census estimate by 2.2-3.2 million (due to unauthorized misclassified as citizens) and the reconciled estimate of unauthorized is smaller than the DHS estimate by 1.2-1.7 million. The paper further considers the effects of (1) possible undercount and (2) omission of temp RAPs and TPS in the estimates.
|