统计与数据科学学院知识库

随着全球气候变暖问题日益突出，世界各国正大力推广低碳生活模式，并积极采取措施应对气候变化，致力于减少温室气体排放。中国同样对此作出郑重承诺，提出了“碳达峰”和“碳中和”的减排目标。中国正处于新发展阶段，数字经济成为社会发展提质增效的关键驱动因素，已深刻融入各类社会活动中，大数据、云计算等新兴数字技术的应用给碳减排提供了新的机遇。中国正向数字经济时代迈进，同时双碳目标是现阶段所面临的严峻挑战，数字经济降低碳排放的作用机理、数字经济碳排放量的测算及影响效应分析成为当前亟待探讨的问题和研究热点。

本文以数字经济的数字产业化和产业数字化分类为基础，结合2007、2012和2017年中国30个省份投入产出表准面板数据，编制中国数字经济投入产出表，测算各省份数字产业化部门和产业数字化部门的能源消耗量，由能源消耗量和该类能源对应的碳排放系数计算得出各省份数字产业化和产业数字化部门的碳排放量。构建出数字经济碳排放效应的LMDI分解模型，从能源结构强度效应、能源消费效率效应、数字经济增长效应以及人口规模效应四个方面对全国层面和省域层面的数字经济碳排放效应进行因素分解。有助于推动各省联合减排，为制定有差别的、有针对性的碳减排政策提供依据，对中国提高碳减排效率、改善能源消费结构、推动低碳经济发展、实现碳减排目标具有重要意义。

研究结论显示：（1）产业数字化部门能源消耗量在数字经济能源消耗总量中占较大比重。（2）化石能源消耗产生的碳排放是数字经济能源消耗碳排放的主要构成部分。（3）数字产业化部门能源消耗多为清洁能源，产业数字化部门能源消耗多为化石能源。（4）能源消费效率效应对数字经济碳排放总量的增长具有抑制作用，能源结构强度效应、数字经济增长效应、人口规模效应对数字经济碳排放总量的增长起促进作用。结合上述结论，提出有针对性的对策建议。

With the growing prominence of global climate change, countries worldwide are vigorously promoting low-carbon lifestyles and actively implementing measures to address climate change, focusing on reducing greenhouse gas emissions. China has also made a solemn commitment in this regard, setting forth its "carbon peak" and "carbon neutrality" emission reduction targets. In its new stage of development, digital economy has become a key driver for improving societal efficiency, deeply integrating into various social activities. The application of emerging digital technologies such as big data and cloud computing presents new opportunities for carbon reduction. As China strides towards the digital economy era, the dual carbon goals pose significant challenges. Understanding the mechanisms by which the digital economy reduces carbon emissions, calculating the volume of carbon emissions from the digital economy, and analyzing its impact effects have become pressing issues and research hotspots.

In this study, based on the categorization of digital industrialization and industrial digitization, we compiled a China Digital Industry Input-Output Table using standardized panel data from China's input-output tables in 2007, 2012 and 2017 across 30 provinces. We calculated the energy consumption of digital industrialization sectors and industrial digitization sectors towards coal, oil, natural gas, and power sectors. By multiplying each province's energy consumption with the corresponding carbon emission coefficients for that type of energy, we obtained the carbon emissions from both digital industrialization and industrial digitization sectors in each province. We constructed an LMDI decomposition model to analyze the factors influencing the carbon emission effects of the digital economy at both national and provincial levels, breaking it down into four aspects: energy structure intensity effect, energy consumption efficiency effect, digital economic growth effect, and population scale effect. This approach helps promote joint emission reductions among provinces, providing a basis for formulating differentiated and targeted carbon reduction policies, thereby significantly contributing to enhancing China's emission reduction efficiency, achieving its carbon reduction goals, improving energy consumption structures, and fostering a low-carbon economy.

The research findings indicate: (1) The energy consumption of the industrial digitization sector accounts for a larger proportion in the total energy consumption of the digital economy. (2) Carbon emissions resulting from fossil fuel consumption constitute the primary component of carbon emissions from the energy consumption of the digital economy. (3) Energy consumption in the digital industrialization sector is mostly clean energy, while energy consumption in the industrial digitization sector tends to be fossil fuels. (4) The energy consumption efficiency effect exerts an inhibitory role in the overall growth of carbon emissions from the digital economy; meanwhile, energy structure intensity effect, digital economic growth effect, and population scale effect all contribute to the increase in the total carbon emissions from the digital economy. Armed with these conclusions, targeted policy recommendations are proposed.