What drives organic certification? Practices, capabilities, and governance in the wine sector

Despite being a central instrument of European agri-food sustainability policy, organic certification uptake remains uneven even among producers that implement environmental practices. This paper examines the factors associated with organic certification in the wine sector, focusing on environmental practices, organizational capabilities, and relational governance. We conceptualize certification as an institutional outcome that formalizes accumulated sustainability engagement rather than as a direct proxy for practice adoption. Using cross-sectional survey data from 232 Hungarian wineries, we estimate logit models in a context characterized by heterogeneous production structures and dense relational networks. The results show that green field practices and green technological practices are strongly associated with certification: a one-standard-deviation increase in these variables is associated with 11.6 and 9.4 percentage point higher probabilities of certification, respectively. Absorptive capacity—the ability to identify, integrate, and apply external knowledge—is also positively associated with certification, corresponding to an 11.0 percentage point higher probability. By contrast, relational quality with business partners is associated with a 12.9 percentage point lower probability of certification. This finding is consistent with the interpretation that trust-based relationships may reduce reliance on formal certification as a credibility mechanism. Interaction analysis indicates that relational quality does not significantly moderate the association between green field practices and certification; rather, it is associated with lower certification probabilities across different levels of green practice engagement. These findings suggest that certification-based sustainability policies may need to address organizational learning constraints and the governance contexts in which producers establish credibility. Because the data are cross-sectional, the results should be interpreted as associations rather than causal effects.