The Determinants of total electricity-generation energy diversification in G20 economies over 2000–2021

This study investigates the determinants of total electricity generation diversification in G20 economies over 2000–2021. Total and renewable diversification are initially measured using the inverse Herfindahl–Hirschman Index constructed from technology-level electricity generation shares, while normalized Shannon–Wiener indices are employed as alternative measures for robustness. The analysis evaluates how renewable energy diversification, energy transition, geopolitical risk, and energy intensity are associated with diversification in both the short and long runs. To address cross-sectional dependence, slope heterogeneity, mixed integration orders, and dynamic adjustment, this study applies cross-sectionally augmented ARDL, augmented mean group, and cross-sectionally augmented panel quantile ARDL estimators. The results indicate a stable long-run relationship among the variables. Renewable energy diversification shows a positive and relatively stable association with total electricity generation diversification, suggesting that variety across renewable technologies tends to coexist with a broader electricity generation portfolio. Geopolitical risk generally promotes diversification, although its effect varies across specifications and quantiles. Energy transition and energy intensity exhibit specification-, measure-, and quantile-dependent effects. The quantile estimates reveal substantial heterogeneity, while the Shannon-based results show that the precise distributional location of some effects is sensitive to index construction. The error-correction coefficients are negative and significant in the mean-based models, confirming adjustment toward long-run equilibrium. Overall, the alternative-measure results broadly support the central findings while also demonstrating that some control variables and quantile effects remain measure-sensitive. Because this study relied on observational data, the findings should be interpreted as conditional associations rather than causal effects. The results suggest that diversification policies should be stage-specific and aligned with country-level electricity system conditions.