Taking the green pill: Psychological distance and the climate crisis in the European Parliament

Abstract

Climate change presents an existential risk to global society, yet public engagement has not risen to meet this challenge. Although research suggests that manipulating psychological distance could enhance the effectiveness of climate communication, it remains unclear whether politicians employ this strategy. This paper addresses this gap by developing and validating automated methods to measure psychological distance in political speech. I introduce a dataset of 35,000 speeches from the European Parliament on climate-related topics from 2014 to 2023, translated into English to bypass multilingual modelling challenges. I present a novel tool designed to streamline the annotation of long-form texts, enhancing efficiency and consistency among annotators. Finally, I explore and validate the use of generative language models to synthesise training data for fine-tuning bidirectional encoder models in order to measure psychological distance in text. This approach demonstrates strong performance and allows for scalable analysis of complex phenomena in natural language. It paves the way for future substantive research on the dynamics of European climate politics.