Uncovering gender dimensions in energy policy using Natural Language Processing

Abstract

There is a lack of empirical analysis of macro-level energy policy through a gender lens, and there exists a research gap in tools for such analysis. This paper proposes a novel empirical approach to explore the gender dimension of the language used in energy policy documents such as the National Energy and Climate Action Plans (NECPs) of European Union (EU) member states. We first review the literature to identify existing approaches that analyse the energy/gender nexus. We then describe how we adapt Natural Language Processing (NLP) techniques to quantify the degree to which the language used in the NECPs is male or female associated. Our work shows the gap in suitable tools. We demonstrate using NLP that all EU NECPs tend to use more male associated language than female. In a ranking of EU member states, the Portuguese NECP demonstrated the most bias, while the Slovenian demonstrated the least. Our work provides novel insights using NLP to understand the genderised use of language, and contributes a methodology to conduct empirical analysis of energy policy documents.