Problematic Content in Spanish Language Comments in YouTube Videos about Venezuelan Refugees and Migrants

Abstract

On YouTube, we found extensive content relating to the recent Venezuelan refugee movement that mostly affects neighboring countries like Peru and Ecuador. While there are several studies on general hate speech on social media, only a few have focused on the online discussion of the Venezuelan migration crisis representing the Latin American perspective. Here, we analyzed via manual coding and computational text analysis 235,251 comments from 200 YouTube videos (selected according to theoretical criteria) in the Spanish language on the Venezuelan refugee crisis. In our sample, we found a high number of problematic comments in videos on Venezuelan refugees and migrants, of which 32% were offensive comments and 20% were hateful comments. The most common linguistic patterns revealed references to xenophobic, racist, and sexist content, and showed that offensive content and hate speech are not easy to separate. Only a small amount of around 8% of highly active users is responsible for about 40% of the problematic content and these users actively comment on multiple videos, indicating a network structure in our sample. Our results enlighten a much-neglected topic in the discussion about Venezuelan refugees and migrants on YouTube and contribute to an enhanced understanding of online hate speech from a Latin American perspective for better and early detection.