Bayesian Hierarchical Modelling for Migration Flow Size Estimation: Application to Italian Data

Charlotte Taglioni , University of Padova
Brunero Liseo, University of Rome La Sapienza

Migration flow size estimation represents a challenge both for statisticians and National Statistical Institutes. An estimation of migration flow for Italy is proposed applying the Bayesian model initially proposed by Bryant and Graham (2013). The model allows for integration of different data sources both for data per flow and net migration data, and it is very flexible and complex at the same time. Applications of this model to the Italian migration flows are performed, highlighting advantages and limits. Data and metadata used come essentially from Istat (Italian National Statistical Institute) but also from other sources. The use of different data sources, along with assumptions on their reliability, results in a complex estimation embedding official sources with rather complete coverage as well as precise but limited coverage sources. The estimation of the flow size divide the migration flow by year, sex, age and region of migration. International migration is generally easier to estimate than internal migration even if, the technological development in Istat processes is making easier to notify and document changes of residence. The Bayesian estimation includes credible intervals giving a measure of uncertainty of the estimates. The flows can be estimated both alone or within the whole demographic account.

See extended abstract

 Presented in Session P3. Poster Session Migration, Economics, Environment, Methods, History and Policy