The visual portrayal of groups in media reinforces stereotypes, potentially leading to discriminatory actions. That is particularly true for stigmatized groups such as migrants. Further, this representation is marked by the relationship between space and society. For instance, contextual events in different countries influence the migrants’ portrayal in media. Therefore, it is critical to study the difference in the migrants’ portrayal among countries. Yet, there is a lack of comparative studies analyzing the portrayals of migrants in different media outlets and countries differing in their political and media systems, net migration figures, and migration policies. This study aims to fill this gap by investigating the portrayal of migrants across ten countries with diverse political and media systems and migration figures. It does so with an interdisciplinary approach to the visual portrayal of migrants using Deep Learning techniques and analyzing results through the lenses of migration and gender studies and visual geography. We have analyzed the demographic and emotional information of the people portrayed in 18.000 pictures across ten countries using an intersectional approach (including gender, age, physical features, and emotions). The portrayals of the general group “migrants” were compared with those of the specific groups “refugees” and “expats”. Results show differences in the portrayal of groups within and between countries. The demographics in the portrayals do not match the official statistics. The general portrayal of migrants predominantly associates them with asylum seekers, and migrants are mostly depicted as crowds. For expats, we found an over-representation of “white” and an under-representation of “Asian” faces, while for migrants and refugees, depictions align with the demographics of low-skilled migrants. Women are portrayed as younger than men. All these effects differ per location.

Camilla Spadavecchia and Juan Sebastian Olier Jauregui
Tilburg University

---

 
ID Abstract: 702