The Causal Effect of Intangible Cultural Heritage on Corporate Performance—Evidence from a Double Machine Learning Model

While culture is recognized as a foundational informal institution, quantifying its causal impact on corporate outcomes remains a challenge. This study focuses on place-based social culture, specifically using a firm’s local Intangible Cultural Heritage (ICH) as the core cultural variable. We aim to identify its causal effect on firm performance (measured by ROA) and contrast it with another pivotal cultural force—the Spirit of the Long March. Leveraging the Double Machine Learning (DML) method, we find that ICH exerts a significant positive effect on ROA, with the estimated coefficient ranging from 0.0018 to 0.0020 ( p < 0.01). In contrast, the effect of the Spirit of the Long March is statistically insignificant. Cross-validation using multiple machine learning algorithms confirms the robustness of our findings. Heterogeneity analysis reveals that the effect of ICH is more pronounced in state-owned enterprises and firms with older CEOs. Mechanism analysis further uncovers that ICH enhances performance primarily by fostering financial conservatism, specifically through a reduction in corporate leverage. Our study provides robust causal evidence on how place-based traditional culture shapes corporate financial policies and underscores the economic value of preserving intangible cultural assets.