Using administrative data to improve the estimation of immigration to local areas in England

• Published in: Journal of the Royal Statistical Society. Series A, Statistics in society Statistics in society, 2010-10, Vol.173 (4), p.707-731
• Main Authors: Boden, Peter, Rees, Phil
• Format: Article
• Language: English
• Subjects: Administrative data sets; Applications; Census; Censuses; Combinatorics; Combinatorics. Ordered structures; Countries; Data analysis; Data collection; Datasets; Designs and configurations; Drivers; Emigration; England; Estimates; Estimating techniques; Estimation; Estimation methods; Exact sciences and technology; Foreign students; General topics; Higher education; Human migration; Immigration; Immigration estimation; Insurance, economics, finance; International migration; Mathematics; Measurement; Methodology; Migration; Modelling Migration Flows; National insurance; Noncitizens; Offices; Population censuses; Population estimates; Probability and statistics; Sampling theory, sample surveys; Sciences and techniques of general use; Statistical estimation; Statistics; Surveys; United Kingdom
• ISSN: 0964-1998; 1467-985X
• DOI: 10.1111/j.1467-985X.2009.00637.x

International migration is now a significant driver of population change across Europe but the methods that are available to estimate its true effect on subnational areas remain inconsistent, constrained by inadequate systems of measurement and data capture. In the absence of a population register for England, official statistics on immigration and emigration are derived from a combination of survey and census sources. The paper demonstrates how administrative data systems such as those which capture registrations of recent migrants with a local doctor, national insurance number registrations by workers from abroad and the registration of foreign students for higher education can provide data to understand patterns and trends in international migration better. The paper proposes a model for the estimation of immigration at a local level, integrating existing national estimates from the UK Office for National Statistics with data from these administrative sources. The model attempts to circumvent conceptual differences between data sets through the use of proportional distributions rather than absolute migrant counts in the estimation process. The model methodology and the results that it produces provide alternative estimates of immigration for consideration by the Office for National Statistics as it develops its own programme of improvement to subnational migration statistics.