In the context of intensifying global competition, productive entrepreneurship plays an important role in industrial upgrading and sustainable economic development. This study explores how the institutional environment affects productive entrepreneurship through different allocation paths. By integrating multiple institutional logics and using PD-QCA (Panel Data fsQCA) methods, we examined data from 76 countries for the period 2012–2022. The results show that institutional complexity provides entrepreneurs with diversified opportunities and resources, but also brings decision-making challenges due to the conflict of institutional factors. We identified three main models: (1) Government-Market Collaborative Model: driven by government support and property rights protection. (2) Market-Led Model: financial support is crucial. (3) Three-Wheel Drive Model: combines property rights protection, market opening, and social norms that promote entrepreneurship. This study contributes to the literature by revealing the multidimensional effects of institutional complexity and provides practical implications for policymakers seeking to promote high-quality entrepreneurial activity. Limitations include the static nature of the data and the need to further explore dynamic institutional interactions over time.

Introduction: Digital Village Construction (DVC) represents a national strategy to integrate digital technologies with agriculture and rural development in China. Whether such a comprehensive, government-led digital transformation can raise agricultural total factor productivity (TFP) remains empirically underexplored, particularly at the county level. This study investigates the causal effect of DVC on agricultural TFP and clarifies the underlying mechanisms and heterogeneous impacts. Methods: We treat China’s National Digital Village Pilot policy as a quasi-natural experiment and construct a time-varying DID design using county-level panel data from 2009–2023. Agricultural TFP is primarily measured via DEA–Malmquist indices based on a land–labor–capital input system and real agricultural output. We include county and year fixed effects and standard controls, and conduct extensive robustness checks. Mechanisms are examined through mediation analyses focusing on non-farm employment opportunities and production technical intensification. Results: Baseline DID estimates indicate that DVC significantly increases agricultural TFP, and the result remains robust across multiple identification and measurement checks. Mechanism analyses suggest that DVC promotes TFP mainly through expanding non-farm employment opportunities and stimulating production technical intensification. Heterogeneity analyses show stronger effects in counties with higher levels of digital financial inclusion, in administratively classified counties, and in areas closer to provincial capitals. Discussion: The findings provide causal evidence that comprehensive digital rural policies can enhance agricultural productivity by improving factor allocation, facilitating labor structural transformation and accelerating technology- and energy-intensive modernization of production. These research conclusions provide empirical evidence for further advancing digital village construction.