数字金融政策与城市全要素生产率:基于BERTopic的深度学习分析  

Digital Financial Policy and Urban Total Factor Productivity:A Deep Learning Analysis Based on BERTopic

作　　者：黄徐亮 Huang Xuliang

机构地区：[1]中国社会科学院大学应用经济学院

出　　处：《中国社会科学院大学学报》2025年第3期126-143,148,共19页Journal of University of Chinese Academy of Social Sciences

摘　　要：以数字金融推动全要素生产率提升是建设金融强国战略的内在要求和重要着力点。本文利用深度学习模型BERTopic挖掘2009—2022年我国数字金融政策的核心主题,并构建了城市数字金融政策指数。实证研究发现,数字金融政策显著促进了城市全要素生产率提升,其中安全监管和服务平台主题的数字金融政策对城市全要素生产率的影响更为显著。在影响机制上,数字金融政策通过增强创新研发应用潜力、降低要素交易成本、提升管理决策能力、提高服务精度广度和优化生产供应协同,促进了城市全要素生产率提升。进一步分析表明,数字金融政策对城市全要素生产率的提升效应存在明显的营商环境和资本错配差异。鉴此,应深化金融供给侧结构性改革,推动金融与数字技术的深度融合,不断提升金融对实体经济的服务质效。Promoting total factor productivity through digital finance is an inherent requirement and an important focal point of the strategy to build a strong financial nation.This paper employs the deep learning model BERTopic to explore the core themes of digital financial policies from 2009 to 2022 and constructs an urban digital financial policy index.Empirical research finds that digital financial policies significantly enhance urban total factor productivity,with the themes of safety regulation and service platforms having a particularly pronounced impact.In terms of the mechanisms of influence,digital financial policies promote urban total factor productivity by enhancing the potential for innovation and research application,reducing transaction costs of factors,improving management decision-making capabilities,broadening the precision and scope of services,and optimizing production and supply collaboration.Further analysis reveals significant differences in the effects of digital financial policies based on the business environment and capital misallocation.In light of this,it is necessary to deepen the structural reform of the financial supply side,promote the deep integration of finance and digital technology,and continuously improve the quality and efficiency of services to the real economy.

关 键 词：数字金融政策 全要素生产率 金融供给侧结构性改革 BERTopic模型 

分 类 号：F832[经济管理—金融学]