Invention and Collaboration Networks in Latin America: Evidence from Patent Data
Carlos Bianchi, Pablo Galaso, Sergio PalomequeThe relevance of collaboration networks in innovation has been widely studied from different streams of research. In recent decades, various studies on this subject have combined social network analysis with patent data to analyze collaboration networks. These works have made important empirical and theoretical advances, however, they have focused almost exclusively on analyzing cases in developed countries (mainly in Europe, North America and Asia). Meanwhile, underdeveloped regions have been neglected by the literature. This paper aims to fill this gap analyzing collaboration networks associated with the processes of invention and patenting in Latin American countries between 1970 and 2018.
We use data from the US Patent & Trademark Office (USPTO) retrieved from the PatentsView database. This database incorporates disambiguated inventor and owner’s identifiers and geolocation. We select patents with, at least, one inventor located in Latin America, and elaborate two types of collaboration networks: networks of inventors and networks of innovators (i.e. patent owners). We study both networks at two different geographical levels. First, at the regional level, we analyze interaction patterns among all actors in Latin America, identifying intra and/or extra regional collaboration dynamics. Second, at the national level, we study networks for each country, searching for territorial differences and particularities in networks that may be associated with national innovation systems. We analyze different network topological properties, focusing particularly on two aspects: internal connectivity (i.e. the extent to which nodes are connected to each other within each country) and external openness (i.e. the extent to which local actors are connected to external nodes). We are also interested in comparing the two types of networks (i.e. inventor vs. innovator networks).
Our findings reveal that almost all Latin American networks are highly fragmented and disconnected. However, we also find that internal connectivity in national networks evolved differently during the period under study. We manage to identify and explain interesting cases (some of them unexpected) of network evolution. For example, while Brazil and Mexico present the largest networks, Cuba is the only country with a giant component, which emerged abruptly in the 2000s and remained connected until today. On the other hand, Argentina and Chile, stand out mainly for the connectivity of their innovator networks. We have also documented cases such as Venezuela, where a giant component emerged in the 1990s and subsequently disintegrated, possibly because of both internal and external institutional shocks experienced by the country.
We also find that Latin American networks are externally oriented. National collaboration networks are strongly connected to external actors. Yet, in certain countries such as Peru or Venezuela, this orientation seems to indicate a strong dependence on international collaborations. Furthermore, we find a clear extra-regional orientation of links, which is consistent with the great fragmentation of the collaboration network at the Latin American level, evidencing the absence of a regionally-integrated innovation system. Our results contribute to identify weaknesses and strengths of collaboration patterns related to innovation in Latin American countries and, therefore, they may be considered by policymakers interested in industrial and innovation policies.