Uniquement disponible en anglais.
Par Coline Metta-Versmessen.
Environmental Policy Stringency (EPS) scores are increasingly used in econometric studies to analyze the effects of climate policy mixes on reducing greenhouse gas (GHG) emissions. These studies often rely on single indicators, such as the OECD measure, as neutral and stable proxies over time and across countries. However, econometric results obtained with such indicators may depend on the aggregation choices of these composite indicators. This study examines this issue by introducing sectoral EPS indicators with country-year specifications, based on the Climate Actions and Policies Measurement (CAPMF) database. These indicators capture national differences and extend sectoral coverage, including diffuse emissions. We test two key hypotheses: 1) the Proxy assumption, which considers EPS a neutral and robust proxy for climate policy stringency over time and across countries and sectors; 2) the Effectiveness assumption, which posits that stringency is associated with the effectiveness of policy mixes, implying a causal link between EPS and emissions reduction. Using two-way fixed effects panel regression with a Shift-Share Instrumental Variable (SSIV), we confirm the Effectiveness assumption: higher EPS is generally associated with reduced GHG emissions. However, we find that all indicators exhibit small regional and temporal instabilities, raising concerns about their reliability, with country-year-specific EPS outperforming indicators with equal weighting. Lastly, we demonstrate that aggregation choices in EPS construction are not neutral and directly influence regression results. These findings underscore the importance of using multiple indicators to enhance the robustness and comparability of EPS measures across sectors and national contexts.