This research determines the intertemporal relationships caused by the coal, oil, and natural gas consumption in the carbon dioxide emission by the G7 countries from 1965 to 2018. Auto-regressive and Distributed Lags models and Bound test were used to detect cointegration and understand the dynamic effect. Due to structural breaks occurred in the variables, two dummy variables for the periods of breaks, 1978 and 1990 were incorporated respectively. Positive causality was identified, in the sense that the consumption of fossil fuels provides an increase in carbon dioxide emissions. Short-term elasticities indicate that an increase of 1 percentage point in the consumption of oil, coal, and natural gas will cause, respectively, an increase of 0.4823%, 0.3140%, and 0.1717% in carbon dioxide emissions. In the long run, the increase of 1 percentage point in the consumption of oil, coal, and natural gas will cause, respectively, an increase of 0.4924%, 0.2692%, and 0.1829% in carbon dioxide emissions. The error correction model (ECM = −0.4739) indicates that 47.39% of a shock in the carbon dioxide emissions variable is resolved in one year and after 2 years, carbon dioxide emissions return to long term equilibrium.

本研究确定了 1965 年至 2018 年 G7 国家二氧化碳排放中煤炭、石油和天然气消耗引起的跨期关系。自回归和分布式滞后模型和边界检验用于检测协整并了解动态影响。由于变量出现结构性断裂，分别纳入了1978年和1990年两个断裂期的虚拟变量。确定了正因果关系，因为化石燃料的消耗导致二氧化碳排放量增加。短期弹性表明，石油、煤炭和天然气消费量每增加1个百分点，二氧化碳排放量将分别增加0.4823%、0.3140%和0.1717%。从长远来看，石油、煤炭、天然气消费量每增加1个百分点，二氧化碳排放量将分别增加0.4924%、0.2692%和0.1829%。误差修正模型 (ECM = -0.4739) 表明，二氧化碳排放变量中 47.39% 的冲击在一年内得到解决，两年后，二氧化碳排放恢复到长期平衡。