Exploring the entropy‑complexity nexus. Evidence from Italy

Combining evolutionary economic geography and the recombinant knowledge approach, we test whether the level of entropy in available capabilities affects the level of a region’s economic complexity. Merging different data sources on 103 Italian NUTS-3 regions and 9 years (between 2008 and 2016), our panel regressions show that a higher level of economic complexity in a region is explained by a higher average level of entropy in the region’s capabilities. We also find this effect stronger in the northern and central regions of Italy, and it is due only to entropy in the available technologies, not in capital or skills. Using structural equation models, we show that variety of skills and relative capital endowment are indirectly linked to the level of economic complexity in Italian regions, since they first stimulate a greater technological diversification.