A Longitudinal Analysis of Concentration Developments for Container Terminals in Northern Vietnam

Characterizing the Evolution of the Container Traffic Share in the Mediterranean Sea Using Hierarchical Clustering

“…H* shows that Med is an unconcentrated port system (H* < 0.1). The range of the H* in both Med regions during 2000-2015 is lower in comparison to other port systems investigated (e.g., in References [4,22]). While a direct comparison with other areas may be far fetched, the temporal variability of the traffic share tends to be relatively stable during the period investigated.…”

“…The concentration indices and hierarchical clustering based in CoDa have a different underlying foundation, so their comparison is interesting. Understanding the concentration pattern may help to identify the future development of a port system [22]. Several cases have illustrated that the past concentration data may infer future tendencies on port system throughput (e.g., Reference [8]).…”

“…Concentration indices are well-known tools oriented to assess the level of concentration in a market. They have been used widely in port systems to evaluate port competition, estimate the historical trend of the concentration, establish models of port development or as general descriptors of the market (e.g., [4,7,8,22]). However, some authors give reasons to doubt the suitability to apply indices based only on traffic shares without taking into account their geographical locations and economic characteristics [23].…”

“…The Herfindahl-Hirschman Normalized index (H*) is a widely accepted measure of market concentration or asymmetry in the market shares and conceptually has been used to describe regional port systems in similar formulations [3,4,11,22,23]. The H* is a function of the number of ports included in the analysis and the container throughput for each one.…”

“…This index has traditionally been used to assess the industry concentration and income distribution, and compares equally concentrated levels of port ranges with a different number of terminals [3]; this means that the Gini coefficient measures the statistical dispersion of container throughput distribution. The Lorenz curve presents the variation in the cumulative throughput of all container terminals [22]. Thus, the Gini coefficient is calculated by the ratio of the area between the Lorenz curve and a diagonal line of equal distribution, and the area below the line of equal distribution.…”