Modelling the Relationship Between Internal Migration and Regional Development

Becky Arnold, Netherlands Interdisciplinary Demographic Institute (NIDI)
Ewoud Jansma, NIDI
Leo van Wissen, Netherlands Interdisciplinary Demographic Institute (NIDI) and University of Groningen

Regional development is a key consideration for internal migrants as they pursue favourable economic conditions and high standards of living. In this project regional development is quantified in three dimensions (economic, social, and living environment) to produce regional development profiles at the NUTS 3 level in the EU 27 countries. The degree to which different types of migrants (by age, sex, and citizenship) prioritise these different dimensions is modelled. To do this the expected number of migrants is expressed as a function of migrant priorities and regional development scores. MCMC fitting is is then used to estimate migrant priorities where complete historic internal migration data available, e.g. Norway where access to person-level register data has already been granted. The MCMC (Markov-­Chain Monte-­Carlo) methodology uses random walkers to explore parameter space and identify minima, making it well suited for studying multidimensional non-linear functions. In countries where internal migration data is less abundant a Bayesian approach will be used to incorporate the data that is available, using the priorities fitted in data-rich countries as priors. Migrant priorities, estimated demographic changes, and projected regional development evolution will then be used to model internal migration in the EU 27 for the next 15 years.

Keywords: International Migration, Population and Development, Data and Methods, Inequality, Disadvantage and Discrimination

See extended abstract.