27 Aug 2009Working Paper Summaries

Measuring and Understanding Hierarchy as an Architectural Element in Industry Sectors

by Jianxi Luo, Daniel E. Whitney, Carliss Y. Baldwin and Christopher L. Magee

In an industry setting, classic supply chains display strict hierarchy, whereas clusters of firms have linkages going in many different directions. Previous theory has often assumed the existence of the hierarchical relationships among firms, and empirical industry studies tend to focus on a single-layer industry, or a two-layer structure comprising buyers and suppliers. And yet, some industries have a multilayer structure with a multistep supply chain. Others comprise a cluster of complementary firms producing different parts of a large system. HBS professor Carliss Y. Baldwin and colleagues use network analysis to study multilayer industries both empirically (in the case of Japan) and theoretically and to explore how industries are organized at the sector level in an attempt to reveal the underlying rules that determine how industry architectures form and change. Key concepts include:

Empirical analysis shows that the automotive sector in Japan exhibits a significantly higher degree of hierarchy and higher transaction breadth (average number of customers per firm) than the electronics sector.
The degree of hierarchy in an industry sector may be traced back to fundamental properties of the underlying technologies.
This research helps points the way to new approaches for understanding industry architectures and the factors that influence the architecture of industry sectors.

Author Abstract

Hierarchy is a generic structure in which levels are asymmetrically ordered. In an industry setting, classic supply chains display strict hierarchy, whereas clusters of firms have linkages going in many different directions. Previous theory has often assumed the existence of the hierarchical relationships among firms and empirical work has focused on a single level of an industry or bilateral relationships. However, quantitative evidence on the deep hierarchy in large industrial sectors is lacking. In this paper, we develop metrics and methods to define and measure the degree of hierarchy in transactional relationships among firms, and apply the methods to two large industrial sectors in Japan: automotive and electronics. We compiled the networks of firms connected by transactional relationships. Our empirical analysis shows that the automotive sector exhibits a higher degree of hierarchy than the electronics sector. We further analyze the differences in hierarchy using a simulation model based on transaction breadth and transaction specificity. The empirical measurement and model analysis together indicate that it is the low transaction specificity that drives down the degree of hierarchy in the electronics sector. Differences in transaction patterns in turn may result from the differences in the power level of underlying technologies, which affect product specificity and asset specificity. Thus, the degree of hierarchy in an industry sector may be traced back to fundamental properties of the underlying technologies. 47 pages.