Emergent Description of the Italian Economic Structure through the Economic Complexity Lens

In this paper, we develop a novel framework that integrates Economic Complexity and spatial statistics to reconstruct and rethink Italy’s industrial geography from the bottom up, leveraging firm-level data from nearly five million geolocated local units across 7,841 municipalities. By statistically validating the co-occurrence of economic activities across space, we construct a bipartite network linking municipali ties and industrial sectors, compute Fitness and Complexity indicators, and identify structural similarities among local production systems. Our results reveal a nested industrial structure, with municipalities and industrial sectors forming spatially and sectorally coherent clusters. Building on this structure, we demonstrate that both po tential extensions of existing Industrial Districts (IDs) and entirely new, self-contained clusters of economic specialization can be identified by combining machine learning with spatial autocorrelation techniques. This approach lays the groundwork for a re newed understanding of regional industrial systems, bridging historical classifications with present-day economic complexity.