随着数字经济发展，数据资产已成为企业核心生产要素。基于中国上市企业数据，构建双重机器学习模型，探究数据资产对提升企业绿色创新能力的因果效应及作用机制。结果表明，数据资产显著提升企业绿色创新能力，经过稳健性检验后该结论依然成立。机制分析表明，数据资产通过优化人力资本结构促进绿色创新能力提升；ESG表现、数字基础设施、行业竞争正向调节数据资产对绿色创新能力的促进作用。异质性分析表明，数据资产对企业绿色创新能力的提升作用在超大/特大/大城市企业、高科技企业、国有企业和规模较大企业中更显著。研究不仅揭示了数据资产提升企业绿色创新能力的机制路径，更为通过优化数据要素配置促进企业绿色转型提供了政策启示。

Under the guidance of China's "dual carbon" goals, green innovation has emerged as a core driver for the low-carbon transformation of the economy and society, while the traditional pollution control model is shifting toward a more value-creating green development paradigm that emphasizes sustainable growth and innovation. The intensification of global trade frictions has hindered the international flow of green technologies, making the enhancement of independent innovation capabilities a critical pathway to achieving carbon neutrality. However, green innovation is characterized by high investment, long cycles, and significant risks, and these characteristics frequently lead to unsustainable corporate investment decisions, particularly under short-term operational pressures, thereby creating persistent bottlenecks in innovation advancement. The rapid adoption of digital technologies in recent years offers a new breakthrough to address this challenge. Against the backdrop of digital transformation, data, as a core resource, is permeating various industries, becoming the key link between digital industrialization and industrial digitization. Data assets, with their attributes of infinite supply, low-cost reuse, and cross-spatiotemporal sharing, can effectively reduce the trial-and-error costs of green innovation, enhance technological synergies, and drive breakthroughs in energy conservation, emission reduction, and circular economy.
Meanwhile, the transformation of data assets into green innovation outcomes is influenced by multiple factors. High-quality digital talent is essential for unlocking data value, while robust digital infrastructure improves data allocation efficiency. Moderate market competition drives corporate innovation, and strong ESG performance steers data resources toward sustainable development. Existing research has examined the impact of digital finance, the digital economy, and digital policies on green innovation. However, studies on data assets remain largely theoretical, focusing on accounting, valuation, and economic effects, and empirical research in this domain is still in its infancy, particularly concerning the intricate relationship between data assets and corporate green innovation. A deeper understanding of how data assets empower green innovation is crucial for enriching the theoretical framework surrounding data marketization but also for providing actionable insights to policymakers. By better understanding this relationship, policymakers can more effectively align digital economy strategies with sustainability objectives, which is of paramount importance in the context of achieving China's “dual carbon” goals.
Against this backdrop, this study employs the double machine learning model to examine the causal effects and mechanisms through which data assets influence corporate green innovation capabilities. Empirical results demonstrate that data assets significantly enhance green innovation, a finding that remains robust across a series of tests. The analysis reveals that data assets improve green innovation by optimizing human capital structure, particularly in firms with higher ESG performance, where the effect is stronger due to their emphasis on sustainable development and technological advancement. Additionally, digital infrastructure strengthens this relationship by improving the allocation efficiency of data resources, while intensified industry competition further amplifies the effect by increasing firms' sensitivity to technological innovation. Nevertheless, the magnitude of this impact varies across different firm types and regions due to differences in resource endowments and technological readiness: enterprises in megacities, high-tech firms, state-owned enterprises, and large corporations benefit more significantly, likely due to their greater access to resources and technological infrastructure; whereas small and medium-sized city enterprises, non-high-tech firms, non-state-owned enterprises, and smaller businesses exhibit weaker effects due to differences in resource endowments and technological readiness. This study not only uncovers the mechanisms driving green innovation through data assets but also offers policy insights for facilitating corporate green transformation through optimized data resource allocation.
The contributions of this study lie in its integration of data assets as a new production factor and its methodological innovation through the application of double machine learning, addressing limitations in traditional econometric approaches. By clarifying how skilled labor mediates the relationship between data assets and green innovation, it demystifies the role of human capital in fostering technological breakthroughs. Moreover, by examining contextual moderators such as ESG performance, industry competition, and digital infrastructure, the study provides actionable guidance for firms across diverse sectors and regions, informing strategies for leveraging data assets to advance sustainable development objectives.