Mapping the Energetic Structure of Climate Transitions for Policy Relevant Regime Detection

TL;DR

This paper develops a regime-diagnostic framework to identify different structural states in climate-economy systems, enabling more nuanced analysis of policy impacts and transition dynamics across countries.

Contribution

It introduces a novel empirical method to detect latent regimes in climate-economic data, improving understanding of heterogeneous transition processes.

Findings

Identified distinct regimes with varying stability and responsiveness.

Revealed heterogeneity in emission-growth relationships across countries.

Provided a tool for conditionally applying econometric and machine learning models.

Abstract

Understanding how climate and innovation policies perform during socio-technical transitions remains a central challenge in innovation studies. Empirical analyses of the relationship between economic growth and carbon emissions continue to yield conflicting results, partly because they rely on pooled models that implicitly assume stable and homogeneous dynamics. Transition theory, by contrast, emphasizes that decarbonization unfolds through heterogeneous regimes characterized by varying degrees of stability, inertia, and reconfiguration. Yet, empirical tools capable of identifying these regimes prior to policy evaluation or forecasting remain limited. This paper introduces a regime-diagnostic framework designed to condition empirical analysis on the structural state of the climate-economy system. Rather than estimating causal effects or generating forecasts directly, the framework reconstructs latent transition regimes from the time varying responsiveness of emissions to economic activity. These diagnostics are used as a pre-modeling step, allowing econometric and machine learning tools to be applied conditionally on empirically identified regimes. Using a global panel of approximately 150 countries over the period 1991-2022, we apply the framework to the emission-growth relationship.