Synergistic Small Worlds that Drive Technological Sophistication

TL;DR

This paper uses partial information decomposition to analyze the internal interactions within production processes, revealing that highly sophisticated industries have modular small-world network structures centered around a connective core, which correlates with higher economic complexity.

Contribution

It introduces a novel application of partial information decomposition to uncover the network structure of industrial interactions, linking network topology to technological sophistication.

Findings

Highly sophisticated industries exhibit modular small-world network topologies.

The tertiary sector acts as a central connective core in these networks.

Well-connected cores and specialized modules correlate with higher economic complexity.

Abstract

Advanced economies exhibit a high degree of sophistication in the creation of various products. While critical to such sophistication, the nature and underlying structure of the interactions taking place inside production processes remain opaque when studying large systems such as industries or entire economies. Using partial information decomposition, we quantify the nature of these interactions, allowing us to infer how much innovation stems form specific input interactions and how they are structured. These estimates yield a novel picture of the nuanced interactions underpinning technological sophistication. By analyzing networks of synergistic interactions, we find that more sophisticated industries tend to exhibit highly modular small-world topologies; with the tertiary sector as its central connective core. Countries and industries that have a well-established connective core and specialized modules exhibit higher economic complexity and output efficiency. Similar modular networks have been found to be responsible for maintaining a balance between integration and segregation of information in the human brain, suggesting a universal principle underlying the organization of sophisticated production processes.