Bayesian Modelling of International Migration with Labour Force Survey Data

Arkadiusz Wisniowski, University of Southampton

In this paper, a statistical method for estimating international migration flows based on information obtained from Labour Force Surveys (LFS) is developed. The motivation for using LFS data is the general poor quality or absence of data provided by the national statistical institutes. A Bayesian model of flows between two countries is considered, assuming that migration can be measured in the LFS by both sending and receiving countries. The undercount in a particular destination-specific emigration flow is estimated by using the LFS sample of the corresponding receiving country. The model is applied to estimate migration flows from Poland to the United Kingdom in 2002-2008.