Energy-Mix Entropy and the Scale–Composition Divide in Urban Renewable Transitions: Evidence from Chinese Cities

Cities can consume more renewable energy without achieving an equivalent shift in their energy mix. We examine this scale–composition divide in a balanced panel of 284 Chinese cities from 2007 to 2019 and use normalized Shannon entropy to characterize the state of each city’s energy portfolio. The analysis distinguishes absolute renewable consumption, total renewable share, and wind plus solar share; it also separates digital infrastructure, digital financial inclusion, and environmental digital governance. Total energy use is strongly positively associated with absolute renewable consumption, negatively associated with total renewable share, and positively associated with wind plus solar share. Energy-mix entropy has a nonlinear association with renewable composition that persists across lagged and alternative diversity measures. By contrast, upper–lower quantile differences, formal subgroup contrasts, and capability interactions based on entropy excluding renewable carriers do not survive multiple testing adjustment. The digital measures also diverge in a common sample: digital financial inclusion is negative, environmental digital governance is positive, and digital infrastructure is imprecisely estimated. Urban renewable transition is therefore better understood as a compositional process associated with portfolio state than as the uniform accumulation of city capabilities.