High-resolution estimates of the foreign-born population and international migration for the United States

TL;DR

This paper develops a detailed statistical method to produce high-resolution estimates of the foreign-born population in the U.S. from 2000 to 2018, combining survey data with auxiliary information to improve accuracy.

Contribution

It introduces a novel evidence synthesis approach that integrates multiple data sources and constraints to produce precise, disaggregated migration estimates for the U.S. foreign-born population.

Findings

High-resolution estimates of foreign-born population from 2000-2018

Improved accuracy through bias correction and data pooling

Validated predictive performance of the statistical model

Abstract

Detailed estimates of migration stocks and flows provides evidence for understanding population dynamics, and the impact of economic and political changes that influence migration. Using data from the 2000 decennial census and 2001-2016 American Community Survey (ACS), this study derives highly-disaggregated estimates of the foreign-born population residing in the United States for the period 2000-2018, and annual foreign-born entries to the ACS population as a measure of immigration volume. These estimates are derived from an evidence synthesis combining pooled survey data with auxiliary data on potential biases in raw survey estimates and other trends affecting the foreign-born population. For an individual population stratum (defined by current age, entry year, country of origin, and calendar year) direct estimates using survey data can have substantial sampling uncertainty. By imposing logical and probabilistic constraints, data are pooled across survey years to produce more precise estimates. Corrections are implemented for respondent misreporting of demographic information, and undercount of the foreign-born population in the ACS. This paper describes the statistical approach used to model population change, demonstrates the validity of the approach via in- and out-of-sample predictive performance, provides the population estimates, and highlights potential applications.