人工智能创新发展缓解了供应链长鞭效应吗？

陈蕾　周锴　董惠敏

（首都经济贸易大学）

摘要：供应链长鞭效应导致的需求信息扭曲与效率损失已成为产业链高质量发展的重要瓶颈，而人工智能发展可能有助于这一问题的缓解。本文以2010—2023年沪深A股上市公司及其所在地级及以上城市为样本，采用双重差分模型，探讨国家新一代人工智能创新发展试验区对企业供应链长鞭效应的影响。研究发现：国家新一代人工智能创新发展试验区的建设有助于企业提升库存管理能力和供应链协同水平，进而缓解供应链长鞭效应；企业信息披露质量和成本粘性对人工智能创新发展政策的影响具有正向的交互作用；国家新一代人工智能创新发展试验区对实体产业技术偏向、多元化经营、平台生态嵌入程度高的企业面临的供应链长鞭效应缓解作用更加明显。本文的研究结论为国家新一代人工智能创新发展试验区建设优化路径与企业供应链管理提供了政策启示。

关键词：人工智能；供应链管理；供应链协同；长鞭效应；平台生态嵌入

作者简介：陈蕾，首都经济贸易大学财政税务学院教授、博士生导师，北京，100070；周锴，首都经济贸易大学财政税务学院博士研究生；董惠敏，首都经济贸易大学财政税务学院博士研究生，通信作者。

基金项目：国家社会科学基金一般项目“数据要素市场化配置的流通生态、效率测度与治理优化研究”（24BGL236）

引用格式：陈蕾,周锴,董惠敏.人工智能创新发展缓解了供应链长鞭效应吗?[J].经济与管理研究,2026,47(5):98-114.

Does AI Innovative Development Alleviates Bullwhip Effect in Supply Chains?

CHEN Lei, ZHOU Kai, DONG Huimin

(Capital University of Economics and Business, Beijing 100070)

Abstract: The bullwhip effect disrupts supply chain stability by amplifying minor demand fluctuations into significant upstream distortions, leading to excess inventory, stockouts, and rising costs. Traditional mitigation strategies struggle to effectively address supply chain bullwhip effects due to coordination challenges and response delays. Artificial intelligence (AI) offers a potential solution to this dilemma. The new generation of national AI innovative development pilot zones, launched in 2019, provides comprehensive support for firms’ digital and intelligent transformation. This paper aims to evaluate the economic impact of AI pilot zones in reducing the bullwhip effect, offering a theoretical understanding and practical strategies for AI-enabled supply chain resilience.

Using data from China’s A-share listed companies and their prefecture-level cities from 2010 to 2023, this paper employs a difference-in-differences model to examine the impact of AI pilot zones on the bullwhip effect in corporate supply chains. The findings confirm that these zones have a positive impact on alleviating the bullwhip effect in supply chains. The conclusion holds after addressing endogeneity issues and conducting robustness checks. Mechanism tests indicate that these zones strengthen inventory management by promoting AI adoption and operational efficiency, and improve supply chain collaboration by enhancing relationship stability. Information disclosure quality and cost stickiness reinforce the policy effectiveness. Heterogeneity analysis shows that the effects are more pronounced for firms that focus on real industries, engage in diversified operations, and embed into the platform ecosystem.

The potential marginal contributions of this paper are as follows. First, it breaks through the digital technology empowerment perspective of traditional supply chain management research, focusing on AI innovation development policy tools and analyzing their institutional impacts. Unlike market-driven digital technology R&D and application, the establishment of AI pilot zones has clear policy objectives, admission criteria, and resource allocation orientation, providing targeted support for computing resources, data factors, application scenarios, and collaborative innovation mechanisms. By promoting the introduction of intelligent forecasting, intelligent decision-making, and data collaboration systems, it more specifically alleviates information asymmetry in supply chains and reduces transaction costs arising from demand misjudgment, information lag, and coordination friction. Second, it enriches the theoretical system of bullwhip effect governance in supply chains, supplementing the understanding of supply chain fluctuations by focusing on innovation development policy rather than merely on technology application. It provides a new theoretical perspective for analyzing how institutional innovation affects corporate supply chain operations. These contributions enrich supply chain management theory and provide references for the strategic deployment of AI-driven initiatives.

Keywords: artificial intelligence; supply chain management; supply chain collaboration; bullwhip effect; platform ecological embedding

下载全文