Carbon emission efficiency of China's industry sectors: From the perspective of embodied carbon emissions

Carbon emission efficiency directly determines the overall carbon emissions level. Previous studies mainly investigated carbon emission efficiency from the perspective of direct carbon emissions without considering carbon emissions associated with intermediate production and consumption, resulting in deviations in the measurement results of carbon emission efficiency. This study distinguishes the differences between direct carbon emissions and embodied carbon emissions, including intermediate production and consumption, and constructs a noncompetitive input-output model to measure the embodied carbon emissions from 28 industry sectors in China between 2005 and 2017. The superefficiency slack-based measure model was used to analyze the carbon emission efficiency of 28 industry sectors in embodied carbon emissions and direct carbon emissions scenarios. The results are as follows. First, China's embodied carbon emissions and direct carbon emissions are generally increasing, with large differences in the total amount, growth rate, and rate of change. Second, China's overall efficiency of direct carbon emissions and embodied carbon emissions has improved. However, there was significant heterogeneity among various industry sectors. Third, in the 28 industry sectors, the change in embodied carbon emissions efficiency was generally gradual, while its average value was significantly lower than that of direct carbon emissions efficiency, indicating that indirect carbon emission efficiency is lower than direct carbon emissions efficiency. Fourth, the reduction potential of embodied carbon emissions for the 28 industry sectors was less than that of direct carbon emissions, although both have decreased, with a greater magnitude of decrease in the former. Finally, this paper proposes countermeasures to promote low-carbon development of China's industries based on improving intermediate production technology and implementing dynamic supervision. (C) 2020 Elsevier Ltd. All rights reserved.