| 1,074 | 0 | 565 |
| 下载次数 | 被引频次 | 阅读次数 |
推动数字经济提质增效与能源产出率持续增长协同并举，既是贯彻落实党的二十大精神的重点任务，也是实现碳中和目标和培育新质生产力的应有之义。为系统分析数字经济如何影响能源产出率，本文测度2011—2022年我国省级层面数字经济发展水平，综合利用人工神经网络（ANN）模型和可解释性工具SHapley Additive exPlanations(SHAP)进行数据拟合，创新性地将SHAP贡献度分析纳入中介检验范式框架，重点探究提升能源产出率的核心数字要素与关键传导路径，剖析行业、地区层面数字经济对能源产出率的影响差异。研究发现，其一，数字经济对能源产出率影响呈现先抑制后促进的U型演变轨迹。其二，替代效应、结构效应、监管效应是数字经济作用能源产出率的重要渠道，数字普惠金融有效促进了电能替代与产业升级，而数字技术创新则在强化政府、公众能耗监管层面形成有力支撑。其三，分行业、分地区来看，数字经济在化学纤维制造业、非资源型省级行政区以及数字经济发展引领型地区对能源产出率的拉动效应更为突出。本文不仅在方法学视角对数字经济降耗增效作用研究补充学理支持，而且为数字经济发展与能源产出率提升二者偕行提供政策启示。
Abstract：Enhancing the interplay between advancing the quality and efficiency of digital economy and sustaining growth in energy productivity is not only a key task in aligning with the ideals set forth in the 20 th National Congress of the Communist Party of China, but also essential to achieving carbon neutrality and cultivating new quality productive forces. In order to systematically analyze how digital economy affects energy productivity, this study measures the development level of digital economy at the provincial level in China from 2011 to 2022, using Artificial Neural Networks(ANN) model and SHapley Additive exPlanations(SHAP) for data fitting. Innovatively, SHAP contribution analysis is incorporated into the mediation testing framework. The paper focuses on the core digital factors and key transmission paths that drive the improvement of energy productivity, and analyzes the differences in the impact of digital economy on energy productivity at the industry and regional levels. The research findings are as follows: First, the influence of the digital economy on energy productivity exhibits a U-shaped trajectory. Second, substitution effects, structural effects, and regulatory effects are important channels through which the digital economy affects energy productivity. Digital inclusive finance effectively facilitates electricity substitution and industrial upgrading, while digital technological innovation provides support in enhancing energy consumption supervision at the governmental and public levels. Third, digital economy exerts a more pronounced driving effect on energy productivity in the chemical fiber manufacturing industry, non-resource-based provinces, and regions leading the development of the digital economy. This paper not only supplements the academic support for the research of the energy consumption reduction and efficiency enhancement of the digital economy from the perspective of methodology, but also provides policy insights for the combination of the development of digital economy and the improvement of energy productivity.
[1]戴魁早,黄姿,王思曼.数字经济促进了中国服务业结构升级吗?[J].数量经济技术经济研究, 2023, 40(2):90–112.
[2]戴翔,杨双至.数字赋能、数字投入来源与制造业绿色化转型[J].中国工业经济, 2022(9):83–101.
[3]郭峰,陶旭辉.机器学习与社会科学中的因果关系:一个文献综述[J].经济学（季刊）, 2023, 23(1):1–17.
[4]郭凯明,潘珊,颜色.新型基础设施投资与产业结构转型升级[J].中国工业经济, 2020(3):63–80.
[5]韩晶,姜如玥.数字经济赋能低碳发展:理论逻辑与实践路径[J].统计研究, 2024, 41(4):54–67.
[6]江艇.因果推断经验研究中的中介效应与调节效应[J].中国工业经济, 2022(5):100–120.
[7]李拓晨,石孖祎,韩冬日,等.数字经济发展与省域创新质量——来自专利质量的证据[J].统计研究, 2023, 40(9):92–106.
[8]宋易珈,陈星达,李锡涛,等.数字企业赋能、超大规模市场优势与企业能源利用效率[J].中国工业经济, 2025(6):121–139.
[9]孙传旺,占妍泓,徐梦洁.电力需求响应信号与新能源制造企业绩效[J].管理世界, 2024, 40(12):185–203.
[10]屠西伟,史丹.数字产业集聚与企业能源效率改进[J].数量经济技术经济研究, 2025, 42(10):70–89.
[11]吴力波,杨眉敏,孙可哿.公众环境关注度对企业和政府环境治理的影响[J].中国人口·资源与环境, 2022, 32(2):1–14.
[12]肖争艳,陈衎,陈小亮,等.通货膨胀影响因素识别——基于机器学习方法的再检验[J].统计研究, 2022, 39(6):132–147.
[13]赵涛,张智,梁上坤.数字经济、创业活跃度与高质量发展——来自中国城市的经验证据[J].管理世界, 2020, 36(10):65–76.
[14]周肖肖,贾梦雨,赵鑫.绿色金融助推企业绿色技术创新的演化博弈动态分析和实证研究[J].中国工业经济, 2023(6):43–61.
[15]Antwarg L, Miller R M, Shapira B, et al. Explaining Anomalies Detected by Autoencoders Using Shapley Additive Explanations[J]. Expert Systems with Applications, 2021, 186:115736.
[16]Apley D W, Zhu J. Visualizing the Effects of Predictor Variables in Black Box Supervised Learning Models[J]. Journal of the Royal Statistical Society Series B:Statistical Methodology, 2020, 82(4):1059–1086.
[17]Athey S, Imbens G W. Machine Learning Methods that Economists should Know about[J]. Annual Review of Economics, 2019, 11(1):685–725.
[18]Białek J, Bujalski W, Wojdan K, et al. Dataset Level Explanation of Heat Demand Forecasting ANN with SHAP[J]. Energy, 2022, 261:125075.
[19]Hao Y, Li Y, Guo Y, et al. Digitalization and Electricity Consumption:Does Internet Development Contribute to the Reduction in Electricity Intensity in China?[J]. Energy Policy, 2022, 164:112912.
[20]Janiesch C, Zschech P, Heinrich K. Machine Learning and Deep Learning[J]. Electronic Markets, 2021, 31(3):685–695.
[21]Lin B, Huang C. Nonlinear Relationship between Digitization and Energy Efficiency:Evidence from Transnational Panel Data[J]. Energy, 2023,276:127601.
[22]Lu J, Li H. Can Digital Technology Innovation Promote Total Factor Energy Efficiency? Firm-level Evidence from China[J]. Energy, 2024, 293:130682.
[23]Lu L, Liu P, Yu J, et al. Digital Inclusive Finance and Energy Transition towards Carbon Neutrality:Evidence from Chinese Firms[J]. Energy Economics, 2023, 127:107059.
[24]Otchere D A, Ganat T O A, Gholami R, et al. Application of Supervised Machine Learning Paradigms in the Prediction of Petroleum Reservoir Properties:Comparative Analysis of ANN and SVM Models[J]. Journal of Petroleum Science and Engineering, 2021, 200:108182.
[25]Pan X, Wei Z, Han B, et al. The Heterogeneous Impacts of Interregional Green Technology Spillover on Energy Intensity in China[J]. Energy Economics, 2021, 96:105133.
[26]Tang X, Ding S, Gao X, et al. Can Digital Finance Help Increase the Value of Strategic Emerging Enterprises?[J]. Sustainable Cities and Society,2022, 81:103829.
[27]Wu J, Lin K, Sun J. Improving Urban Energy Efficiency:What Role does the Digital Economy Play?[J]. Journal of Cleaner Production, 2023,418:138104.
[28]Xu Q, Zhong M, Li X. How does Digitalization Affect Energy? International Evidence[J]. Energy Economics, 2022, 107:105879.
[29]Zhao H, Guo S. Analysis of the Non-linear Impact of Digital Economy Development on Energy Intensity:Empirical Research based on the PSTR Model[J]. Energy, 2023, 282:128867.
[30]Zheng R, Wu G, Cheng Y, et al. How does Digitalization Drive Carbon Emissions? The Inverted U-shaped Effect in China[J]. Environmental Impact Assessment Review, 2023, 102:107203.
(1)资料来源网址为https://www.ndrc.gov.cn/fggz/hjyzy/jnhnx/202507/t20250717_1399273.html;https://www.ndrc.gov.cn/fzggw/wld/lichunlin/zyhd/202507/t20250709_1399092.html。
(2)资料来源网址为https://www.gov.cn/zhengce/zhengceku/202408/content_6970435.htm。
(3)资料来源于国家数据局发布的《数字中国发展报告（2024年）》，网址为https://www.nda.gov.cn/sjj/ywpd/sjzg/0530/ff808081-96b465bf-0197-200a135e-0536.pdf；国务院新闻办公室发布的《中国的能源转型》，网址为https://www.gov.cn/zhengce/202408/content_6971115.htm。
(4)资料来源于中国信息通信研究院发布的《中国数字化绿色化协同转型发展进程报告（2023）》，网址为http://www.caict.ac.cn/kxyj/qwfb/ztbg/202311/P020231110566201824016.pdf。
(1)资料来源于经济日报刊登的《数字化推动电力行业变革》，网址为http://paper.ce.cn/pc/content/202305/15/content_274100.html。
(1)因篇幅所限，各类机器学习模型构建方法、参数设定及拟合优度比较结果以附录1展示，见《统计研究》网站所列附件。下同。
(1)因篇幅所限，能源产出率测算方法以附录2展示。
(2)因篇幅所限，数字经济发展水平测算方法以附录3展示。
(1)因篇幅所限，政府监管效应、公众监管效应测算方法以附录4展示。
(2)因篇幅所限，各类控制变量指标的构成以附录5展示。
(3)本文研究样本不包含我国西藏自治区和港澳台地区。
(1)因篇幅所限，稳健性检验具体方法与结果以附录6展示。
(1)因篇幅所限，全国层面能源产出率、数字经济发展水平、控制变量等数据说明以附录7展示。
(2)因篇幅所限，行业层面数字经济对能源产出率的SHAP结果，代表性行业中数字经济各要素贡献度以附录8展示。
(1)网址为https://www.gov.cn/gongbao/content/2013/content_2547140.htm。
(2)因篇幅所限，资源型与非资源型省级行政区划分，资源禀赋异质性分析结果以附录9展示。
(1)因篇幅所限，数字经济发展水平等级划分，数字经济发展水平异质性分析结果以附录10展示。
基本信息:
DOI：10.19343/j.cnki.11-1302/c.2026.03.004
中图分类号:F426.2;F49
引用信息:
[1]孙传旺,徐梦洁,王博.数字经济影响能源产出率的动因分解与路径识别——基于可解释性机器学习的研究[J].统计研究,2026,43(03):52-66.DOI:10.19343/j.cnki.11-1302/c.2026.03.004.
基金信息:
国家社会科学基金重大项目“能源供给侧与需求侧协同绿色低碳发展机制与实现路径研究”（21&ZD109）
2026-03-25
2026-03-25