Introduction

This research investigates the evolving structure of production within the electronics industry, specifically examining Japan’s strategic position from a network science perspective. Conventional aggregate trade statistics often fail to capture the architecture of firm-level relationships where a limited number of upstream suppliers exert significant system-wide influence. This is particularly relevant in East Asia, where cross-border production has evolved into a densely interconnected network linking Japan, China, South Korea, and Taiwan. While Japan was a dominant force in semiconductors during the 1980s, its position has been challenged by the rise of South Korea and Taiwan, transitioning Japan into a critical actor in strategically important upstream segments rather than an overwhelmingly dominant producer.

Data collection

The study utilizes firm-level supply chain data from the S&P Capital IQ Pro database, focusing on inter-firm business relationships. The analysis restricts itself to relationships identified from publicly available sources, such as company reports and official filings, to ensure transparency and consistency. The initial sample of electronics firms was expanded to include those in closely related sectors, specifically the automotive and aerospace-defense industries. The resulting curated dataset comprises 19,135 firms, of which 7,002 are publicly listed companies. Cleaning filters were applied to standardise names, harmonise country data, and remove terminated or duplicate links.

Construction of supply chain network

A directed production network was constructed where each firm represents a node and links represent the directional flow of goods or services. The network is fragmented, but its largest weakly-connected component (WCC1) contains 15,292 nodes and 27,751 links. Within this component, the top five countries—the United States, China, India, Taiwan, and Japan—collectively account for over 60% of all firms. Japan maintains a significant position in upstream segments, including electronic components, manufacturing equipment, and precision instruments.

Network analyses

Initial analysis shows the network is extremely sparse, with a density of 0.00012. Path analysis indicates high efficiency in connecting distant firms, with an average path length of 5.89, which may also facilitate the rapid propagation of shocks. Bowtie and Helmholtz–Hodge decompositions reveal that Japan and China have a higher concentration of upstream firms compared to the United States. Mean values of the Hodge potential are 0.239 for Japan, 0.209 for China, and -0.119 for the US, where higher values indicate more upstream roles.

Community detection partitioned the network into 300 distinct communities with a high modularity of 0.747, indicating strong local reliance. The largest community (Com1) serves as the global backbone for semiconductors and ICT. In this community, the relative shares of Taiwan and South Korea increased significantly, while Japan’s share declined from 7.96% in the overall network to 7.53% in Com1. This signifies Japan’s weakening status in the mainstream semiconductor industry. Conversely, the second-largest community, centered on the automotive-mobility-energy supercluster, shows Japan’s share rising to 11.21%, indicating sustained leadership.

The analysis further identified a Japan-based electronics–automation–utilities cluster (Com5) where Japanese firms hold a 58% share. In contrast, the third-largest community (aerospace and defense) is dominated by US and European firms, with negligible Japanese involvement and the apparent strategic exclusion of China. Connector hubs, such as Amazon, Samsung, and Bosch, were identified as critical bridges essential for global network cohesion.

Summary

The findings confirm that while Japan remains strong in upstream segments like electronic components and manufacturing equipment, its influence in mainstream semiconductor manufacturing has weakened relative to competitors in the US, Taiwan, and South Korea. However, Japan maintains a stable, influential position in the automotive sector and dominates a unique industrial ecosystem linking electronics with factory automation. This network-based approach provides a quantitative basis for stress testing and policy analysis amid increasing geopolitical risks, such as export controls and friend-shoring.