Will Digital Finance Reduce Agricultural Total Factor Productivity? Evidence from China

Using a city-level panel for China (2011–2021), this paper estimates agricultural total factor productivity (TFP) with a stochastic-frontier approach and identifies the effect of digital finance through two-way fixed effects and instrumental-variable strategies. We document a statistically and economically significant negative association: a 1% increase in the digital finance index is linked to a decline of 1.5 in agricultural TFP. Evidence points to capital misallocation as the dominant channel, with the adverse effect most pronounced where agricultural capital markets are highly distorted. Heterogeneity analyses show stronger negative impacts in labor-intensive areas, non-major grain regions, and small-scale farming systems. Results are robust across alternative specifications and IV estimations. By moving from provincial aggregates to city-level variation, this study sharpens identification and uncovers within-province patterns that are invisible in coarser data. The findings highlight an important unintended consequence of digital financial expansion for agriculture and underscore a policy priority: improving the allocation and targeting of digital credit within rural economies to support productivity and sustainable development.