Which ports differ in the copycat effectiveness of the hinterland strategy? Based on port-hinterland distance correlation

With increasing cargo and intensifying competition, port differentiation needs to be considered for port expansion, investment and other decisions, as the effectiveness of the same strategies may be different for different ports. Under the concept of hinterland, this paper tries to distinguish large ports by the distance correlation with their hinterland. According to the evolution history of hinterland's spatial structure, only the continent level or more advantaged port are uncorrelated with its continent level hinterland in distance. Following this idea, Granger causality test is applied to differentiate the ports. This paper uses Nighttime Light (NTL) data to represent the human activities of the hinterland for higher spatial accuracy and statistical consistency, rather than common economic indicators such as GDP. In the case of the European continent and surrounding ports, this paper has differentiated the container port by cargo throughput (TP) and liner shipping connectivity (LSC). In result, ports are divided by the presence and absence of port-hinterland distance correlation, and additionally these ports with port-hinterland correlations are divided into 4 groups. Ports can copycat strategies with each other only in same division, because the port-hinterland relationships among different divisions are not same.