Fragmentation and Visibility in Supply Networks

Abstract

Supply network visibility is critical for resilient supply chains, yet, in practice, visibility beyond tier-1 suppliers is still rare. As a result, many firms are caught off-guard by disruptions to their upper-tier suppliers. Using a state-of-the-art firm-level buyer-supplier relationship dataset, we construct the time-evolving supply network and explore its structure with the goal of improving visibility into the network. We find that the network is highly dynamic, rewiring about 25% of its links each year. The network is also fragmented, i.e., it contains communities of firms that are densely connected to each other yet sparsely connected to other communities via a distinct set of firms. Moreover, this fragmentation has increased over time, meaning that there are more communities in the network and the communities progressively equalize in size. We show that these characteristics have made it harder to attain visibility, hence we propose a network exploration algorithm that utilizes the fragmented structure of the supply network to enable firms to more efficiently find a disrupted supplier in their networks. The performance of the algorithm improves as the network becomes more fragmented.

Keywords: Production networks, supply chains, network analysis, empirical study