[Paper Review] Pattern-detection in the global automotive industry: a manufacturer-supplier-product network analysis

Massimiliano Fessina, Andrea Zaccaria+2 more|arXiv (Cornell University)|Oct 16, 2023

Economic and Technological InnovationEconomics, Econometrics and Finance104 references3 citations

TL;DR

This study analyzes the global automotive supply chain using a manufacturer-supplier-product network derived from the MarkLines dataset, applying maximum entropy-based motif detection to uncover structural patterns. It reveals widespread V-shaped and square motifs indicating competitive, supplier-commonality-driven firm relationships, with distinct geographical clusters—especially a fragmented Chinese ecosystem—highlighting structural resilience differences and functional module organization.

ABSTRACT

Production networks arise from supply and customer relations among firms. These systems are gaining growing attention as a consequence of disruptions due to natural or man-made disasters that happened in the last years, such as the Covid-19 pandemic or the Russia-Ukraine war. However, data constraints force the few, available studies to consider only country-specific production networks. In order to fully capture the cross-country structure of modern supply chains, here we focus on the global automotive industry as represented by the MarkLines Automotive dataset. After representing this data as a network of manufacturers, suppliers, and products, we perform a pattern-detection exercise using a statistically grounded validation technique based on the maximum entropy principle. We reveal the presence of a significantly large number of V-shaped and square-shaped motifs, indicating that manufacturing firms compete and are seldom engaged in a buyer-supplier relationship, while they typically have many suppliers in common. Interestingly, generalist and specialist suppliers coexist in the network. Additionally, we unveil the presence of geographical patterns, with manufacturers clustering around groups of suppliers; for instance, Chinese firms constitute a disconnected community, likely an effect of the protectionist policies promoted by the Chinese government. We also show the tendency of suppliers to organize their production by targeting specific functional modules of a vehicle. Besides shedding light on the self-organising principles shaping production networks, our findings open up the possibility of designing realistic generative models of supply chains, to be used for testing the resilience of the interconnected global economy.

Motivation & Objective

• To uncover hidden structural patterns in the global automotive industry’s interfirm production networks, which remain underexplored due to data limitations.
• To address the lack of cross-country, firm-level supply chain analysis by leveraging the comprehensive MarkLines Automotive dataset.
• To identify functional modules and resilience indicators through statistically validated network motifs, moving beyond aggregated or country-specific models.
• To reveal how geographical, industrial, and organizational factors shape supplier relationships and network resilience in global automotive supply chains.
• To inform the development of realistic generative models for stress-testing global economic resilience under supply chain shocks.

Proposed method

• Constructs a tripartite network of manufacturers, suppliers, and products using the MarkLines Automotive Information Platform dataset.
• Applies a maximum entropy principle-based statistical validation technique to detect significant recurring motifs (V-shaped and square-shaped) in the network.
• Projects the network onto manufacturer and product layers by linking nodes that share a significantly large number of common suppliers.
• Uses redundancy and shared neighbor analysis to quantify supplier commonality and infer resilience patterns across firms and regions.
• Performs community detection on network projections to identify functional modules (e.g., powertrain, chassis, electrical) and geographical clusters.
• Validates findings using comparative analysis with existing literature on Chinese automotive industrial policy and fragmentation.

Experimental results

Research questions

• RQ1 What recurring structural motifs—particularly V-shaped and square-shaped patterns—emerge in the global automotive supply chain network?
• RQ2 How do geographical and industrial factors shape the distribution of supplier commonality and network resilience?
• RQ3 To what extent do Chinese automotive manufacturers form a self-contained, less integrated network compared to Western counterparts?
• RQ4 How do functional modules in vehicle production emerge from supplier-product relationships in the network?
• RQ5 Can motif-based analysis reveal organizational and resilience patterns that inform realistic generative models of supply chains?

Key findings

• A significantly large number of V-shaped and square-shaped motifs are detected, indicating that manufacturers typically compete and share many suppliers, rather than being in direct buyer-supplier relationships.
• The Chinese automotive ecosystem forms a disconnected community with fewer shared suppliers, suggesting lower integration and higher internal fragmentation, likely due to protectionist policies.
• Generalist and specialist suppliers coexist in the network, with suppliers organizing production around specific functional modules such as powertrain, chassis, and electrical systems.
• The number of shared suppliers (squares) decreases with increasing node degree, and this pattern strongly separates Chinese from Western firms, with Chinese firms showing fewer shared connections.
• Network projections reveal that Chinese manufacturers are primarily linked only to Chinese joint ventures, confirming limited cross-network supplier sharing.
• Redundancy in supplier networks suggests that manufacturers with many shared suppliers are more resilient to supply shocks, as their production is less vulnerable to localized disruptions.