International trade and the global economy heavily rely on maritime transport, with ports serving as crucial links between production centers and consumer markets across borders. The size and structure of the economy significantly influence freight flows through ports, and conversely, changes or disruptions in the maritime transport network impact supply chains globally. The intricate interplay between maritime transport and global supply chains is evident at various spatial scales. At the global level, maritime trade demand is driven by factors such as countries' trade demands, supplying nations, and the modal split (share of maritime transport). Events like the COVID-19 pandemic and the 2021 Suez Canal blockage highlighted the tight coupling between these networks, showcasing how disruptions can cause significant trade bottlenecks and supply chain disruptions. Furthermore, events such as Hurricane Katrina (2005) demonstrated how port closures can cause ripple effects across global supply chains, impacting commodity prices and export losses. Prior research on port criticality often focuses on the absolute volume of trade or network characteristics, neglecting the fundamental role of ports as physical infrastructure connecting international supply chains. This study aims to address this gap by developing a framework to comprehensively understand port criticality across trade, transport, and supply chain levels.

Existing macroeconomic studies have examined the evolution of international trade and supply chain interconnectivity, often using Multi-Regional Input-Output (MRIO) tables to describe inter- and intra-industry dependencies. However, these analyses lack insights into the transportation systems facilitating these trade flows. Another body of research has analyzed maritime transport networks using complexity science, focusing solely on port connections without incorporating information on the goods transported or their economic use. This study bridges this gap by integrating global trade and supply chain data with a detailed representation of the maritime transport network to provide a holistic view of port criticality.

This research employs a novel modeling framework to provide a comprehensive understanding of port criticality. The framework consists of three interconnected layers: the trade network layer, the transport network layer, and the port supply chain layer. The trade network layer uses a global modal split model to estimate the share of maritime transport in bilateral trade flows at the commodity level. This model incorporates various factors including cost, time, income, and commodity characteristics. The transport network layer utilizes the Oxford Maritime Transport (OxMarTrans) model, which simulates the allocation of maritime trade flows across a network of 1378 ports and their hinterlands, considering maritime and land routes, transhipments, and observed vessel movements (AIS data). The port supply chain layer links port-level trade flows from OxMarTrans to the EORA MRIO tables to quantify economic dependencies. Two key metrics are introduced: (1) the Port-Level Output Coefficient (PLOC), which captures the total industry output dependent on port trade flows, both domestically and globally, and (2) the Port-Level Import Coefficient (PLIC), which quantifies the marginal change in port-level imports per change in final demand. The base year for the analysis is 2015.

The study reveals that approximately 50% of global trade by value is maritime, varying across sectors (e.g., 76% for mining and quarrying). Low-income countries and Small Island Developing States (SIDS) show significantly higher reliance on maritime trade (1.5 and 2 times the global average, respectively). Every USD flowing through a port contributes an average of 4.3 USD to the global economy. Five macro-critical ports handle goods contributing >1.4% to global economic output, while 40 ports are domestically critical, representing >10% of their serving economies' output. A 1000 USD increase in final demand results in a median 84.6 USD increase in maritime imports across served ports, with some individual ports experiencing >100 USD increases. Significant cross-border dependencies exist due to land connections and transhipment: approximately 35% of global port throughput involves foreign ports. The model identified crucial ports serving landlocked and island nations, highlighting regional infrastructure dependencies. Port throughput is highly concentrated, with a small number of core ports handling a significant share of global trade. Sectoral variations in trade unevenness are also observed. The PLOC metric reveals that ports play diverse roles in supply chains, ranging from raw material exports to integration within domestic manufacturing. Finally, the PLIC metric indicates that ports serving SIDS and countries with limited importing ports exhibit higher import sensitivity to demand changes.

The findings underscore the interconnectedness of ports, maritime infrastructure, and global supply chains. The disproportionate reliance of low-income countries and SIDS on maritime transport highlights the need for robust port infrastructure investments to support economic growth. The identification of cross-border dependencies emphasizes the potential for spillover effects from shocks affecting either economies or the maritime network. The PLOC metric sheds light on the diverse roles ports play in integrating domestic and global supply chains, and the PLIC metric reveals how import requirements change with shifts in final demand. This comprehensive analysis moves beyond aggregate port statistics to reveal criticality dimensions previously overlooked, providing valuable insights for policymakers and researchers.

This study provides a novel, quantitative framework for evaluating port criticality across various dimensions, improving upon existing analyses. The findings emphasize the importance of considering port roles within the broader context of domestic and global economies and supply chains. Future research could integrate this framework with carbon emission models, policy simulations (infrastructure investment, trade facilitation), future trade flow analysis, and disaster impact models to enhance the understanding and management of maritime transport systems.

The study's use of national economies as the spatial aggregation level might bias results due to economic size. The analysis relies on 2015 data, potentially limiting its applicability to current conditions. The model's accuracy depends on the underlying data quality and assumptions.