Public attitudes and actions towards renewable energy in the EU countries: A micro-level analysis during the energy crisis

Background: Despite ambitious EU decarbonization targets, household adoption of renewable energy (RE) remains uneven, particularly during systemic energy price shocks.Methods: Using Flash Eurobarometer microdata ($N \approx 23,000$) spanning 27 EU countries linked with regional macro indicators, we combine binary logistic regression with machine learning classifiers (random forests and decision trees) to evaluate adoption drivers and behavioural discrepancies.Results: We identify both a significant attitude–behaviour gap and an inverse attitude–behaviour gap in specific demographic segments. Machine learning feature importance confirms that structural factors (homeownership, regional climate requirements) and capital constraints strongly dominate psychological readiness in determining realized RE investments.Conclusions & Policy: Rising energy prices heighten adoption interest but exacerbate disparities. Effective decarbonization policy requires tenure-specific incentives (targeting renters), targeted upfront subsidies for low-income and elderly households, and place-based financial support for regions facing disproportionate energy burdens.