Using a novel transaction plant-protection UAV data in China, this paper examines how agricultural subsidy payment delays affect agricultural technology adoption. The analysis estimates zero-truncated Poisson correlated random effects models with spatial Durbin extensions. Global and Local Moran's I tests show significant spatial clustering in UAV adoption. In the preferred spatial Durbin zero-truncated Poisson model, subsidy payment delay significantly reduces local plant-protection UAV adoption. This negative effect holds after controlling for spatial dependence and county characteristics. The control function approach further addresses potential endogeneity, and the result remains stable. Adoption is constrained on both the extensive margin and the intensive margin. Heterogeneity analysis shows that the delay effect is not uniform across counties. The negative impact is stronger in counties with lower lagged fiscal expenditure, smaller grain sown area, and higher GDP per capita. The study emphasizes that timely delivery of payments is just as important as the subsidy's nominal value or eligibility rules.
