ABSTRACT This paper investigates the paradox of how Italy's fragmented, SME‐dominated wine industry achieves global export success. Moving beyond purely firm‐centric explanations, we test whether export intensity is spatially dependent, clustering geographically in regional ecosystems. Using a unique nationwide dataset of 1091 geolocated wineries, we employ a Bayesian Spatial Autoregressive (SAR) model to quantify this relationship. Our approach addresses the classic identification challenge in agricultural settings of disentangling economic spillovers from shared natural endowments (terroir) by using official DOC/DOCG status as a robust institutional proxy for geographic identity. Controlling for a rich set of firm‐level characteristics, we find robust evidence of significant positive spatial autocorrelation: a winery's export intensity is positively and significantly associated with the export intensity of its geographic neighbours, even after accounting for its own resources. This result provides strong, micro‐level evidence for the role of agglomeration economies in driving international competitiveness. The findings imply that place‐based policies aimed at strengthening local industrial clusters may be more effective for enhancing export performance than aspatial, firm‐focused strategies.
Chauvin et al. (Mon,) studied this question.