Replication data for: What drives high agricultural mechanization for food security under systemic shocks? A configurational analysis of 30 Chinese provinces during COVID-19

This study examines how different configurations of socioeconomic, technological, and institutional conditions contribute to high agricultural mechanization levels (AMLs) in China. The dataset covers 30 Chinese provinces from 2011 to 2021. Data were collected from official statistical yearbooks and publications, including the China Statistical Yearbook, China Agricultural Machinery Industry Yearbook, and statistical publications released by the National Bureau of Statistics of China. Missing values were completed using linear interpolation, and all indicators were standardized using the Z-score method. The agricultural mechanization level index was constructed through factor analysis based on 12 indicators. Three factors were retained, and the composite AML index was calculated using weighted factor scores based on the variance contribution rates. The resulting composite score was used as the outcome variable in the fsQCA analysis. Six causal conditions were included in the fsQCA analysis: land operation scale, agricultural equipment level, rural labor structure, farmers' income, human capital, and fiscal support. To capture potential time-lag effects, the fsQCA analysis used a one-year lagged design: the outcome variable was the provincial AML composite score in 2021, while causal conditions were measured using corresponding provincial data from 2020. All variables were calibrated using the direct method with three qualitative anchors: full membership (0.95), crossover point (0.50), and full non-membership (0.05). The fsQCA results indicate that no single condition is necessary for achieving high AMLs. Instead, multiple configurational pathways lead to high mechanization, revealing three development mechanisms: equipment-fiscal dual driving, resource-driven endogenous development, and policy-capacity compensation. The replication package includes the dataset and scripts required to reproduce the factor analysis and fsQCA results. Data availability: The raw data were derived from official statistical yearbooks of China and are freely accessible from the National Bureau of Statistics of China (https://www.stats.gov.cn/). Processed data and replication code are provided in this package for research and verification purposes. Corresponding author： For any questions regarding the data, interpretation, or reuse, please contact Fen Zhu (School of Marxism, South China University of Technology, Guangzhou, China). Email: [e-mail].