PUB - Publikationen an der Universität Bielefeld

Incoming refugees from Ukraine are currently encountering a wave of solidarity that is seen, according to some, in stark contrast to the solidarity experienced by earlier groups of refugees i.e. from Syria during the so-called 'immigration crisis' in 2015. We aim to inform this debate on solidarity bias by collecting and analyzing quantitative data on (anti-)solidarity statements posted on Twitter during both waves of refugee immigration. We assess how social solidarity towards refugees differed between 2015 and the current wave of refugees fleeing Ukraine. To this end, we collect and analyze a longitudinal dataset of refugee-related tweets selected via hashtags and covering the period between January 2015 and August 2022. We first annotate the tweets for (anti-)solidarity expressions towards refugees. On these annotations, we train a supervised machine learning model and use it to automatically label over 2.3 million tweets. We assess the automatically labeled data for how statements related to refugee (anti-)solidarity developed and differed for distinct groups of refugees. Our findings show that in relative terms, refugee solidarity was expressed more often in tweets during September 2015 compared to March 2022. However, we find some evidence of solidarity bias in March 2022.