Tasting place: Cultural identity and the uneven geography of European geographical indications

The European food landscape features a rich diversity of Geographical Indications (GIs), yet their distribution remains highly uneven, with Southern Europe accounting for a disproportionate share of registered products. This study examines cross-national associations between structural, institutional, economic, agricultural, and cultural indicators and GI density across 26 EU Member States. It introduces a demand-side cultural perspective to complement existing supply-side explanations of GI development. Using harmonised data from Eurostat, FAOSTAT, the European Commission, the IMF, and commercial restaurant databases, we apply exploratory regression-based modelling and variable-selection techniques to identify indicators associated with national variation in GI density. Given the limited sample size and cross-sectional design, the analysis is explicitly exploratory and should be interpreted as identifying patterns of association rather than causal relationships. The modelling combines multicollinearity checks, theory-guided variable selection, backward stepwise selection, and LASSO regularisation. The results indicate a pronounced North–South divide in GI density. Among the indicators considered, the share of restaurants serving national cuisine in capital cities is the only variable consistently associated with higher GI density after accounting for economic, agricultural, and institutional indicators. We interpret this restaurant-based measure not as a causal factor, but as a proxy for the public visibility of national food culture and broader societal engagement with culinary heritage. The findings suggest that cultural context may be an important but underexplored dimension in understanding the uneven geography of European GIs. Rather than demonstrating that food culture drives GI development, the study highlights how visible expressions of culinary identity are associated with origin-based quality schemes and points to the need for more spatially disaggregated and longitudinal research.