Econ Papers Daily / 期刊论文
数字智能如何驱动企业绿色创新：来自中国的证据
How Digital Intelligence Drives Corporate Green Innovation: Evidence from China
摘要
Corporate green innovation is essential for sustainable development, but firms often face high technological uncertainty, long innovation cycles, and internal capability constraints. This study examines the impact of digital intelligence (DI) on corporate green innovation using Chinese A-share listed firms from 2007 to 2023. A firm-level DI index is constructed from the full text of annual reports using GPT-4, and a gradient-boosting-based double machine learning framework is employed to control for high-dimensional and nonlinear confounding factors. To mitigate endogeneity concerns, an instrumental-variable DML analysis is conducted using the interaction between historical communication infrastructure and lagged national Internet development shocks as the instrument. The results show that DI significantly promotes corporate green innovation, including both green invention patents and green utility models. Mechanism analysis indicates that DI promotes green innovation through uncertainty perception mitigation, human capital upgrading, and environmental management capability improvement. Moderating effect analysis shows that regional digital infrastructure and regional innovation capacity strengthen the effect of DI. Heterogeneity analysis further indicates that the effect is stronger among larger firms, manufacturing firms, downstream firms, and high-tech firms. These findings provide new evidence on the role of DI in promoting corporate green innovation and digital–green integration.
中文摘要
企业绿色创新对可持续发展至关重要，但企业往往面临较高的技术不确定性、较长的创新周期以及内部能力约束。本研究基于2007—2023年中国A股上市公司，考察数字智能（DI）对企业绿色创新的影响。本文利用GPT-4从年度报告全文构建企业层面的DI指数，并采用基于梯度提升的双重机器学习框架以控制高维和非线性混杂因素。为缓解内生性问题，本文以历史通信基础设施与滞后全国互联网发展冲击的交互项作为工具变量，进行工具变量DML分析。结果表明，DI显著促进企业绿色创新，包括绿色发明专利和绿色实用新型专利。机制分析表明，DI通过缓解不确定性感知、提升人力资本和改善环境管理能力来促进绿色创新。调节效应分析表明，区域数字基础设施和区域创新能力会增强DI的作用。异质性分析进一步表明，该效应在较大企业、制造业企业、下游企业和高科技企业中更强。这些发现为DI在促进企业绿色创新和数字化—绿色化融合中的作用提供了新证据。