Adaptive robust bayesian spatial panel estimation of economic resilience under environmental and energy shocks

This paper develops a robust and adaptive Bayesian spatial panel methodology to assess economic resilience among European countries facing environmental and energy shocks. The proposed framework is designed for high-dimensional panels, integrating spatial spillovers, temporal dynamics, and variable selection under uncertainty. The approach combines a fixed spatial weights matrix with time-varying spatial dependence parameters and Bayesian filtering to capture evolving interdependencies across countries. Variable selection is handled through posterior inclusion probabilities, ensuring robustness to outliers and structural breaks. The empirical analysis is conducted on a balanced panel of 28 European countries over the period 2004-2023, using the Solow residual as a measure of total factor productivity, together with indicators related to greenhouse gas emissions, energy prices, electricity generation structure, and energy trade openness. The findings highlight significant cross-country heterogeneity in resilience: nations with carbon-intensive energy systems experience deeper and more persistent productivity losses, whereas those investing in renewable energy sources show relatively more adaptive responses despite transitional inefficiencies. The structure of energy prices and the mix of electricity generation play distinct roles, and openness to electricity trade is found to enhance resilience by fostering cross-border complementarities. The methodology provides a decision-support tool for guiding energy transition policies under complex and uncertain conditions.