Traditional deficit models in science communication research suggest that our understanding of science is predominantly shaped by scientific institutions. However, it is increasingly evident that political and public discourses also critically influence this understanding. Science becomes essential to political communication when political actors invoke scientific evidence to support their positions. For instance, political debates on the climate crisis frequently employ evidence-based knowledge, yet diverse, partisan interpretations often characterize these debates. This paper investigates the complex interactions between science and politics by categorizing and measuring science-related political communication. We analyzed all parliamentary speeches from the Austrian Nationalrat between 1996 and 2018 (N = 89,926) using automated text analysis methods. Initially, we identified sequences referencing science through a dictionary approach. Subsequently, we used a Large Language Model (LLM) to code the specific nature of these science-related sequences. This coded data enabled a comprehensive statistical analysis of how different political parties discuss and reference science. Our findings indicate that while references to scientific actors and institutions are common, there is variation in how parties frame and engage with scientific evidence. This study provides insights regarding the salience and context of science in political discourse.
