Modelling Asylum Migration Pull-Force Factors in the EU-15

Abstract

We model the relationship between asylum applications across the EU-15 and three key pull-force factors, GDP, the recognition rate (grants of refugee status as a percentage of all refugee decisions) and refugee stocks. A Dynamic Coefficient Random Effects Panel model is employed which captures both the country specific variance and temporal features. This model provides deeper insight into the relationship than has hitherto been possible. We show for Austria, Spain and possibly Belgium that the three predictive factors dominate while for other countries in the EU-15 country specific factors are more important. In relation to temporal effects we show both GDP and the recognition rate at the overall EU-15 level remain significant over time but refugee stock does not. We also show a downward level shift occurred in the elasticity of recognition rate in 2002 resulting in fewer refugees being recognised (i.e., granted refugee status) and, therefore, providing evidence that asylum policies have become more restrictive.