Integrating census and residence permit data to estimate annual bilateral migration flows for South America, 1986-2020.

Andrea Aparicio-Castro, University of Manchester
Arkadiusz Wisniowski, University of Manchester
Francisco Rowe, University of Liverpool
Mark Brown, University of Manchester

International migration flows are a difficult component of population change to measure. The difficulty comes from the fact that many countries do not produce consistent flow data, and the available data are usually incomplete and incomparable. Even the two main data sources, censuses and residence permit data(RPD), are subject to data-source-specific systematic biases. The limitations of these sources can be overcome by integrating them to exploit their strengths and compensate for their weaknesses. While census data strengths lie mainly in their comparability due to them being collected following commonly shared guidelines, RPD strengths relate to their production frequency. This paper develops a Bayesian hierarchical model that integrates census and RPD to estimate annual bilateral migration flows. The model corrects the limitations of the data.Specifically, it overcomes the fact that some censuses only provide information on residency 5 years before the census date, measure migration by counting events, and do not supply data for intercensal periods. Likewise, our model handles the dissimilarities of RPD caused by being tied to country-specific legislation and collection systems. We illustrate our model with South American data from 1986 to 2020. The output is a set of synthetic estimates of migration flows with measures of uncertainty.

Keywords: Bayesian methods / estimation, Harmonized data sets, International migration, Population projections, forecasts, and estimations

See extended abstract.

  Presented in Session 169. Creating and Using International or Historical Datasets