How Data Trading Platforms Empower New Forms of Digital Tourism in China: A Causal Inference Based on Double/Debiased Machine Learning

As the “fifth major factor of production,” data plays a crucial role in fostering China’s tourism industry, advancing high-quality economic development, and gaining competitive market advantages. Serving as institutional infrastructure for data factor rights confirmation, pricing, trading, and value conversion, data trading platforms are central to the market-based allocation of data factors. The efficient flow and value realization of data elements have paved the way for the rapid development of digital tourism; new forms of digital tourism represent a profound transformation of the industry resulting from integration and innovation with other sectors. Based on the platform ecosystem theory, we select the panel data of 297 Chinese cities from 2012 to 2024 and innovatively use the Double/Debiased Machine Learning (DDML) model to empirically test the impact of data trading platforms on the new forms of digital tourism and its mechanisms. It is found that the construction of data trading platforms effectively empowers the development of new forms of digital tourism, and this conclusion still holds after a series of robustness tests, such as changing the sample split ratio, replacing the machine learning algorithm, and the instrumental variables method. Mechanism analysis indicates that data trading platforms significantly promote new forms of digital tourism through dual pathways of talent agglomeration and technological innovation, an effect further strengthened by increased government support. Heterogeneity analysis found that the empowerment effect is more significant in cities with lower resource endowment and common administrative level and historical cities, which can be effectively transformed into an employment support effect. Spatial effect analysis reveals that the establishment of data trading platforms exerts a positive pull effect on new forms of tourism in surrounding cities within a 30 km core zone. However, this effect gradually weakens with increasing distance, turning into a significant negative siphon effect beyond 60 km. The findings provide theoretical basis and empirical support for regionally differentiated digital tourism development policies.