摘要 : This paper uses the logarithmic mean Divisia index (LMDI) approach to decomposition analysis to identify the factors that influence changes in carbon dioxide (CO_2 ) emissions in the United States. The LMDI approach decomposes CO_... 展开 This paper uses the logarithmic mean Divisia index (LMDI) approach to decomposition analysis to identify the factors that influence changes in carbon dioxide (CO_2 ) emissions in the United States. The LMDI approach decomposes CO_2 emissions into specific determinants. The data set includes the 50 states plus the District of Columbia from 1998 - 2018. The five factors that influence the change in CO_2 emissions include the emissions per unit of fossil fuel consumption, share of fossil fuels in total energy consumption, energy intensity, GDP per capita, and population. The results indicate that, during the 20-year period, CO_2 emissions declined in 36 states plus the District of Columbia. The reduction in energy intensity served as the most important factor in the change of CO_2 emissions, offsetting 63 percent of the effects of per capita GDP and population. From the perspective of climate change, the importance of a change in energy intensity demonstrates the effectiveness of a decrease in primary energy consumption per unit of GDP. 收起