Decoding Sentiments about Migration in Portuguese Political Manifestos (2011, 2015, 2019)

Abstract

This research conducts a mixed-method analysis of Portuguese political manifestos from 2011, 2015, and 2019, focusing on immigration discourse. We employed Natural Language Processing (NLP) sentiment analysis, with a Multilingual BERT model, alongside qualitative examination of key statements. Findings indicate a generally positive sentiment towards migration among mainstream parties, consistently with Portugal’s socio-economic context. However, the 2019 election highlighted a shift to polarized views, especially from the emerging extreme-right party Chega, mirroring wider European trends. This study underscores the inter- play between political rhetoric, socio-economic realities, and immigration policy, showcasing the applicability and limitations of NLP in political sentiment analysis.