By Emily R. Barker and Jakub Bijak
Forecasting large changes in the number of asylum applications, which are seen as one constituting element of so-called asylum ‘crises’, is incredibly challenging. Demographers, migration scholars, analysts, and policy makers have tried to build and apply early warning systems for that purpose at least since the large inflow of asylum seekers into Europe in 2015–16, which at the time was relatively unforeseen. Recently, another humanitarian crisis occurred after the Russian invasion of Ukraine in February 2022. In this case, temporary protection offered to Ukrainian nationals in all EU member states superseded asylum applications, and the number of applications would not have necessarily raised alarms for early warning systems based on approaches using a single dependent variable. In this report, they present a model that shows that the warning signs of a crisis could appear in some publicly- available data sources, including some ‘big data’ collections. Their aim is to propose and test an early warning system for asylum applications in the EU that would be easy to use, effective and interoperable for policy makers, and that would give sufficient advance warning that authorities can be prepared for an increase in the number of asylum applications or asylum case load. They look to see if a model can give a warning signal up to six months in advance for two of the most prominent asylum flows from the recent decade, involving people fleeing the wars in Syria and in Ukraine. In addition, they provide an empirical reflection on the definition of a ‘crisis’, and offer some practical recommendations for further work in the area of early warning modelling for migration-related applications.