Using the dual-pilot policy of AI experimental zones and low-carbon cities as a quasi-natural experiment, this paper employs a double machine learning method on panel data from 284 prefecture-level cities (2000-2024) to examine the effects and mechanisms of"smart-green integration" on urban green innovation efficiency. Results show: The dual-pilot policy significantly improves green innovation efficiency, remaining robust after alternative dependent variables, outlier exclusion, model re-specification, and controlling for parallel policies. The policy indirectly boosts efficiency through three channels: talent structure optimization, digital-real technology integration, and industry-university-research collaborative innovation. Heterogeneity analysis reveals stronger effects in the commercialization stage (vs. R&D stage), central cities (vs. non-central), non-resource-based cities (vs. resource-based), and high-digital-base cities (vs. low-digital-base). The dual-policy effect significantly outperforms single policies, exhibiting a"1+1>2" synergy.
LU Yang
,
WANG Yu-bao
. The impact of“ smart-green integration” on urban green innovation efficiency: A quasi-natural experiment based on dual pilot programs for AI Experimental Zones and Low-Carbon City[J]. Urban Problems, 2026
, 372(7)
: 51
-64
.
DOI: 10.13239/j.bjsshkxy.cswt.260705
[1] 刘金科,肖翊阳. 中国环境保护税与绿色创新:杠杆效应还是挤出效应?[J]. 经济研究,2022(1):72-88
[2] 王怡颖,王娜. 生成式人工智能赋能产业智能化、绿色化、融合化发展的逻辑进路与实践路径[J]. 经济学家,2026(2):108-117
[3] 钞小静,沈路. 创新价值链视角下人工智能技术对制造业企业绿色创新效率的影响[J]. 经济学动态,2025(4):50-67
[4] Chen X,Chen Y,Huang WL,et al.A new Malmquist-type green total factor productivity measure: An application to China[J]. Energy Economics, 2023, 117:106408
[5] 肖仁桥,高倩,王宗军,等. 营商环境驱动城市绿色创新效率提升的多元路径分析[J]. 科技进步与对策,2026(3):44-55
[6] 袁健,张所地. 知识流动与城市绿色创新效率: “新基建”的调节效应[J]. 城市问题,2024(8):64-73
[7] 付国梅,唐加福,刘娟. 人工智能是否赋能企业绿色转型——基于社会需求理论视角[J/OL]. 南开管理评论,1-17[2026-07-10]. https://link.cnki.net/urlid/12.1288.f.20260518.1446.004
[8] Razia M, Hu Q, Xu J, et al.Greening through AI? The impact of artificial intelligence innovation and development pilot zones on green innovation in China[J]. Energy Economics, 2025,146:108507
[9] 韩先锋,李佳佳,朱承亮. 中国人工智能发展的包容性绿色增长效应[J]. 资源科学,2025(6):1278-1294
[10] 余正颖,李宗洋,李胜会. 数智赋能视角下人工智能发展对城市绿色技术创新的影响研究[J]. 经济问题探索,2026(1):54-70
[11] 付业辉,僧建芬,乐凯迪,等. 人工智能应用与企业价值链升级:效应、机制与情境异质性—大语言模型的文本分析证据[J]. 科技进步与对策,2026(8):26-36
[12] 肖素萍,潘家栋. 数智政策、协同赋能与区域创新水平——来自国家大数据与人工智能“双试点”的经验证据[J]. 浙江学刊,2026(1):214-226,240
[13] Scoble H.From policy to prototype: The role of carbon tax, industry & cities in Australia’s green innovation landscape[J]. Sustainable Futures, 2026,11 :101948
[14] 李毅,胡宗义,周积琨,等. 低碳城市建设的绿色技术创新效应评估——兼论数字经济与制度环境的调节作用[J]. 统计研究,2025(8):59-71
[15] 肖仁桥,马伯凡,钱丽,等. 低碳城市试点政策对企业绿色创新的影响及其作用机制[J]. 中国人口·资源与环境,2023(5):125-137
[16] 郭丰,杨上广,柴泽阳,等. 低碳城市建设能够提升城市绿色技术创新吗?——来自准自然实验的证据[J]. 软科学,2023(1):40-49
[17] 刘文君,刘星诚,惠怀玉. 低碳城市与创新型城市双试点的绿色创新效应[J/OL]. 科研管理,1-14[2026-06-24]. https://link.cnki.net/urlid/11.1567.G3.20260508.1419.013
[18] 鲁锦涛,李振翠,宋马林.“宽带中国”和低碳城市双试点对企业绿色创新的影响研究[J]. 经济社会体制比较,2026(1):102-117
[19] 余东升,余娟娟,陈梦园. 人工智能赋能何以助推新质生产力提升:基于企业绿色全要素生产率视角[J]. 中国软科学,2026(2):137-148
[20] 杨洁,杨阳. 人工智能创新应用先导区的城市减污降碳效应——基于双重机器学习的因果推断[J/OL]. 软科学,1-17[2026-07-04]. https://link.cnki.net/urlid/51.1268.g3.20260630.1617.006
[21] 孟望生,李丁. 创新型城市和低碳城市双试点的绿色发展效应研究[J]. 城市问题,2023(11):55-65
[22] 臧传琴,孙鹏. 低碳城市建设促进了地方绿色发展吗?——来自准自然实验的经验证据[J]. 财贸研究,2021(10):27-40
[23] Hansen M T, Birkinshaw J.The innovation value chain[J]. Harvard Business Review, 2007(6): 121-130, 142
[24] Chernozhukov V, Chetverikov D, Demirer M, et al.Double/debiased Machine Learning for treatment and structural parameters[J]. The Econometrics Journal, 2018(1): 1-68
[25] 洪永淼,汪寿阳. 大数据如何改变经济学研究范式?[J]. 管理世界,2021(10):40-55,72,56
[26] Farbmacher H, Huber M, Lafférs L, et al.Causal mediation analysis with Double Machine Learning[J]. The Econometrics Journal, 2022(2): 277-300
[27] 肖仁桥,陈小婷,钱丽. 异质环境规制、政府支持与企业绿色创新效率——基于两阶段价值链视角[J]. 财贸研究,2022(9):79-93
[28] 吴延兵. R&D存量、知识函数与生产效率[J]. 经济学(季刊),2006(3):1129-1156
[29] 黄先海,高亚兴. 数实产业技术融合与企业全要素生产率——基于中国企业专利信息的研究[J]. 中国工业经济,2023(11):118-136
[30] 权小锋,李静蕾. 人工智能与企业协同创新——基于国家新一代人工智能创新发展试验区的准自然实验[J]. 经济研究,2025(8):72-93
[31] Goldsmith-Pinkham P, Sorkin I, Swift H.Bartik instruments: What, when, why, and how[J]. American Economic Review, 2020(8): 2586-2624
