Digital economy and green TFP in manufacturing: heterogeneous, agglomeration and spillover effects revisited

Purpose This paper empirically examines how the digital economy affects manufacturing green total factor productivity (GTFP), with particular attention to heterogeneity, non-linearity and spatial spillovers. Design/methodology/approach Using 2012–2021 panel data of 41 cities in the Yangtze River Delta, we estimate a threshold–Spatial Durbin Model where industrial agglomeration acts as the threshold variable and economic distance defines spatial weights. Findings (1) The digital economy improves manufacturing GTFP, but only significantly in cities with strong network effects, non-resource-dependent economies, and simultaneous high-intensity digital and environmental policies. (2) Below an agglomeration threshold of 0.25 the effect is weak (0.114); between 0.25 and 0.67 it rises to 0.279; above 0.67 it reaches 0.443, indicating that high density plus digitization neutralizes congestion. (3) Spatial decomposition shows the indirect (spill-over) coefficient of the digital economy (0.327) is larger than its direct coefficient (0.300), confirming stronger cross-city green gains. Originality/value We pioneer a threshold-SDM framework that embeds agglomeration regimes into spatial spillover analysis, offering a new rationale for coordinated, cluster-based digital manufacturing policies.