Industrial Network Analysis Using Inter-Firm Transaction Data

Abstract

Background/Objectives: Inter-firm transaction data can be used as basic information to grasp the effectiveness and robustness of country’s economic activity. Methods/Statistical analysis: In this paper, we constructed and analyzed an inter-firm network and inter-industry network using inter-firm transaction data from KED. From the result of analysis, we found that the distribution of transaction frequency between companies tends to follow the power-law distribution. This is because a large number of companies have trading connections few firms such as a conglomerate, implying that most transactions are concentrated in the major companies. And we analyzed the transaction type of each industry using the E-I index. According to the result, the companies which belong to the service-related industries tend to trade with the companies in other industries. Otherwise, the companies which belong to the motor manufacture-related industry and the electronic-related industry do a lot of trade with the companies in same industries. Especially, the network structure of these industries is hierarchized as a tree structure in comparison with other industries. Findings: From the inter-industry transaction network, we also found that the industries in Korea are largely divided into two groups: domestic service industries and export manufacturing. These two sub-networks commonly form a tree structure representing the hierarchical flow of transactions where a transaction flows from leaf nodes to root node in the inter-industry network. Application/Improvements: The General Construction industry is the root node that is located at the top in the network of domestic service industries. And the electronic and computer-related industries are the root nodes in the network of export manufacturing.