A hybrid approach for examining the drivers of energy consumption in Shanghai

Increasing attentions have been paid to energy consumption problem due to ecology pressure. This has prompted a growing focus of policy design to explore the driving forces of energy consumption. As the biggest city in China, Shanghai ranks the 1st in economic aggregate in mainland China but faces greater energy shortage challenges than the other cities. In this study, we aim to investigate the energy consumption between 1990 and 2019 in Shanghai to facilitate the sustainable development of population, resources, and environment. Specifically, a hybrid approach coupling LMDI (logarithmic mean Divisia index) and Granger causality is adopted to decompose the influential factors of energy consumption and explores their potential relationships. The hybrid approach not only allows us to verify the contributions of energy consumption changes in the past (1990-2019), but also to identify the relationships between energy consumption and the potential factors. It is found that per urban population GDP (Gross Domestic Product) (accounting for 228.16 % contribution degree), infrastructure investment (accounting for 227.18 % contribution degree) and urbanization rate (accounting for 21.61 % contribution degree) are the top three driving forces of energy consumption. The Granger causality test shows that energy consumption has bilateral causality relationship with per urban population GDP, infrastructure investment and urbanization rate, but it has single Granger causality relationships with infrastructure investment population support coefficient and energy intensity in different lag lengths. Interestingly, our results indicate that the energy consumption, per urban population GDP, infrastructure investment and urbanization rate changes are closely related to urban policies or main events. Therefore, policy makers should not only make appropriate policies in infrastructure investment but also pay attention to the balance between energy consumption socialeconomic effects.