Complex Network Analysis of Imported and Exported Goods in the United States of America

In this paper the network structure of imported and exported freight flows in United States in 2022 is explored. The data used are from the Freight Analysis Framework (FAF) that represents freight movement among states and major metropolitan areas by all modes of transportation. Using tools from network science hubs and central nodes making the network of freight flow vulnerable are identified and communities of nodes based on shipments of goods in value of tonnage flows are extracted to distinguish different functional regions. Different network topology in 2022 is found analyzing imported goods and exported goods. The networks are graphs with 132 nodes that are spatial subdivisions based on major Metropolitan Statistical Areas (MSAs) and the non-MSA parts of their States. The links are the directional movement of freight flows among the nodes. The imports graph explored here has 10739 directional links among the 132 nodes. The graph of international exports is smaller with 9722 links among the 132 nodes and displays different freight flow behavior and community clustering than the much larger 2022 domestic freight flow network that has 17356 directional links. The analysis here finds some similarities in the location of vulnerable nodes. Similarities are found in functional regions formed by data-driven clusters of nodes dividing the US into Northeast, Southeast, Southcentral, West, and Midwest with differences among the domestic, import and export networks. The number of distinct clusters in the imports and exports networks is higher than the domestic network clusters and in a few clusters node membership violates regional contiguity and proximity. This is due to the infrastructure connectedness and economic complementarity (e.g., nodes that exchange goods for auto manufacturing with Detroit, Michigan, and Laredo in Texas that imports goods from Mexico).