Knowledge complexity and brokerage in inter-city networks

Localized interactions allow access to external knowledge flows, contributing to the dynamic increase of knowledge complexity. We analyse the knowledge complexity of cities, as a proxy of their knowledge capabilities, and the extent to which capabilities are determined by the interactions between cities. We address this issue studying the evolution of knowledge complexity and the inter-city network in Latin America. We find uneven evolutions of knowledge complexity of cities, which are integrated into sparse and highly outward-oriented networks. Using panel data regression models, we show that cities' knowledge complexity can be explained by the city's position in the network. In particular, we find two determining factors: on the one hand, playing a broker role, i.e. intermediating between disconnected cities, is associated with the level of complexity. On the other hand, network brokerage shows a sort of externality, due to which cities that maintain connections with broker cities also achieve higher levels of knowledge complexity. This evidence contributes to understanding the current changes and challenges in the main Latin American urban agglomerations.