The impact of the 2015 refugee crisis on the positional decision-making processes of political parties in multi-party systems across Europe

The impacts of the 2015 refugee crisis on immigration and integration policies of national parties in multi-party systems across Europe have drawn the attention of various scholars. In reaction to increased refugee-inflows the political landscape across different countries has seen mechanisms of transformation and adaption. As this event created unique conditions for national party competition, I propose that mechanisms of cohesion and polarization were set into motion considering party type, governmental position, and previous positional shifts. To test my arguments, I present a novel approach to estimate party positions on issue-specific topics. Using a supervised machine learning approach, utilizing manually coded immigration and integration categories based on the Comparative Manifesto Project, I apply multiple binary text classification models to identify additional and previously unidentified topic-related text passages within party manifestos from 2005-2020. In the next step, I run Wordfish and Wordscores models on this newly generated text corpus to estimate individual national party positions on immigration- and integration-issues. The results provide a novel dataset of topic specific positions of national parties, supporting the effectiveness of supervised text analysis on small amounts of ideologically loaded textual data. Finally, I analyse the cross-national determinants of positional shifts in the wake of the refugee crisis. The results give insight how parties adapt their positions in response to this unique event with respect to the dynamics it created in relation to party-systemic and party-specific preconditions.