LAUSR

Natural gas, as a cleaner transition energy than other fossil fuels, has been widely used in China and, thus, further investigation of its potential determinants is particularly useful. For this purpose, this study explores the influencing factors for the consumption of natural gas in China and tests whether the natural gas Kuznets curve (GKC) hypothesis exists, using a panel of 30 provinces covering 1997–2015. To fully capture the significant differences in the natural gas markets across regions, the 30 provinces in China are re-divided into four regions in accordance with their natural gas consumption and economic level. Additionally, considering the possible cross-sectional dependence and slope heterogeneity of the data, a series of econometric approaches that does accommodate cross-sectional dependence and heterogeneity is utilized. For the full panel and four regions, cross-sectional dependence and slope heterogeneity exist. The results indicate that, natural gas consumption is significantly and positively affected by economic growth, urbanization level, energy intensity, and economic structure in the long run, while energy consumption structure has a negative effect. In addition, only for the full panel, Regions I and II, is the GKC hypothesis valid. Also, the results of a panel causality test for the four regions are mixed, again confirming the presence of significant regional differences.