This study examines how demand-side drivers are translated into green innovation outcomes under varying supply-side conditions, thereby addressing limitations in prior research that relies on broad supply-demand models and overlooks configurational mechanisms.
Grounded in a mechanism-based perspective, this study employs dynamic qualitative comparative analysis (QCA) on panel data from 30 Chinese provinces over the period 2014–2024. In addition, a long short-term memory (LSTM) model is used to analyze the temporal evolution and predict future trends (2025–2026).
The results show that green innovation is not driven by individual factors, but by specific combinations of demand drivers and enabling conditions. Three underlying mechanisms are identified: (1) an enabling mechanism, in which strong supply conditions support demand effectiveness; (2) a synergistic mechanism, in which supply and demand jointly reinforce innovation outcomes; and (3) a compensatory mechanism, in which strong demand partially offsets weaker supply conditions. The findings also reveal significant temporal dynamics and regional heterogeneity, with demand-driven pathways playing a more prominent role in less-developed regions.
This study reconceptualizes green innovation as a conditional process in which demand-side drivers require enabling supply-side conditions to be effective. By identifying a limited number of mechanism-based configurational pathways, it provides a more parsimonious and theoretically grounded explanation compared with broad supply–demand models. The integration of dynamic QCA with time-series prediction further offers new insights into the temporal evolution of green innovation.
