Under increasing land constraints and food security pressures, understanding the direction of agricultural technological change is essential for improving land use efficiency. This study investigates the regional disparities, distributional dynamics, and spatial convergence of biased technological change in Chinese agriculture. Dagum Gini decomposition is used to identify regional differences and their sources, kernel density estimation examines distributional dynamics, and spatial econometric models test convergence patterns. The results show that agricultural technological progress in China is predominantly biased toward labor and capital, with labor–land bias being the strongest and continuously increasing. This indicates a gradual shift from land-dependent growth toward more intensive use of non-land inputs under farmland constraints. Regional disparities in technological bias have widened over time, mainly driven by interregional differences and distributional overlap. Kernel density analysis reveals dynamic but largely non-polarized evolution, suggesting gradual adjustment rather than structural divergence. Although no σ-convergence is observed, both absolute and conditional β-convergence exist, with faster convergence in the labor–land dimension and under conditional settings. These findings imply that regions tend to adapt to land constraints through differentiated technological pathways, resulting in uneven improvements in land-related productivity. Overall, biased technological change plays an important role in shaping land use efficiency under resource constraints. The study provides evidence for understanding agricultural adaptation to land scarcity and offers implications for sustainable agricultural development, farmland protection, and long-term food security.
Zhang et al. (2026) studied this question.