Greentech homophily and path dependence in a large patent citation network

We propose a method to identify the main technological trends in a very large (i.e., universal) patent citation network comprising all patented technologies. Our method builds on existing literature that implements a similar procedure, but for much smaller networks, each covering a truncated sub-network comprising only the patents of a selected technology field. The increase of the scale of the network that we analyse allows us to analyse so-called macro fields of technology (distinct technology fields related by a coherent overall goal), such as environmentally friendly technologies (Greentech). Our method extracts a so-called network of main paths (NMP). We analyse the NMP in terms of the distribution of Greentech in this network. For this purpose, we construct a number of theoretical benchmark models of trajectory formation. In these models, the ideas of homophily (Green patents citing Green patents) and path dependency (the impact of upstream Green patents in the network) play a large role. We show that a model taking into account both homophily and path dependence predicts well the number of Green patents on technological trajectories, and the number of clusters of Green patents on technological trajectories.

JEL Classification: Q55, Q54, O31, O33, O34

Keywords: patent citation networks; green technology; climate change mitigation