Digital strategy, ESG integration, and green supply chain efficiency in the Thai Agri-food industry: evidence from an undesirable-output SBM-DEA and system GMM analysis

Agri-food processing drives Thailand's exports but is also a significant source of greenhouse gas emissions. This study examines whether digital strategy and environmental, social and governance (ESG) integration raise objectively measured green supply chain efficiency, and whether the two levers are complementary. A non-oriented, variable returns to scale modified slacks-based measure with Scope 1 and 2 emissions as an undesirable output is estimated for listed agri-food firms over 2021 to 2025; because emission disclosure is incomplete, efficiency is obtained for 172 firm-years of 42 firms from a frame of 51 firms. The score, interpreted as focal-firm supply chain eco-efficiency, is modeled with two-step system GMM on 86 firm-years of 33 firms, complemented by pooled, bias-corrected fixed effects, fractional probit, and selection-corrected estimates. Mean efficiency is 0.548. In the system GMM, digital intensity is negatively associated with efficiency (β = −0.677, p = 0.005), while ESG integration and the digital-ESG interaction are not significant. The interaction is nevertheless positive in almost every specification and significant in pooled (β = 0.184, p < 0.001), bias-corrected fixed effects ( p < 0.001), and fractional probit ( p = 0.037) estimates, and the marginal effect of digital intensity turns positive only at the highest ESG ratings. Productivity growth stems from catching up rather than frontier advance. The evidence is consistent with complementarity, in which digital investment improves eco-efficiency only when directed by strong sustainability integration, although the short panel limits the precision of the dynamic estimates. The study provides replicable frontier evidence from an emerging-market industry.