Renewable energy green innovation and digital financial infrastructure shape carbon emissions in the United States
摘要
This paper examines the long-run and short-run determinants of per capita carbon dioxide (CO₂) emissions in the United States over 2000–2023. The analysis focuses on renewable energy consumption, environmental technology innovation, environmental taxation, economic growth, and a digital financial infrastructure proxy for FinTech development. The original title has been refined because the dependent variable is CO₂ emissions rather than a multidimensional sustainable-development index. To account for persistence and smooth structural changes in annual environmental time series, the paper employs a parsimonious Fractional Frequency Fourier ARDL (FFF-ARDL) framework and verifies the long-run results with Fourier-augmented FMOLS, DOLS, and CCR estimators. The results support the existence of a stable long-run equilibrium. Renewable energy consumption significantly reduces emissions, whereas economic growth increases emissions, confirming the persistence of scale effects. Environmental technology innovation and environmental taxation are positively associated with emissions in the long run, while environmental technology innovation displays a delayed negative short-run effect, suggesting that innovation and fiscal instruments may not yet be sufficiently stringent, targeted or diffused to generate dominant long-run mitigation effects. The digital financial infrastructure proxy is insignificant in the long run but positive in the short run, consistent with early-stage digitalisation effects on electricity and transaction activity. Additional diagnostics based on the Excel data file, including Durbin-Watson, Breusch-Pagan-Godfrey, ARCH, VIF and forecast evaluation statistics, are reported to strengthen model validation and to acknowledge the limitations imposed by the small annual sample.