🌍 ClimaParl

Climate change increasingly shapes contemporary party politics. Yet, scholarship often conflates climate change with broader environmentalism, limiting our understanding of its distinct, cross-cutting nature. In this letter, we introduce a novel approach to identify climate-specific content in political texts. Methodologically, we fine-tune XLM-RoBERTa, a multilingual language model, on manually annotated parliamentary speeches from 27 European countries. This classifier distinguishes between narrow and broad references to climate change. Empirically, we demonstrate the face validity and the usefulness of our classifier by mapping climate salience in European parliaments and revisiting two studies on the effect of climate protests on references to climate change in parliamentary speeches. Our contribution provides a resource-efficient tool for scholars and supplies sentence-level predictions for European parliaments allowing researchers to study short- and long-term dynamics. Thus, we advocate for a more precise study of climate change in political discourse, facilitating deeper insights into how political actors engage with this defining issue.