Information Technology Reference
In-Depth Information
8.5
Conclusion
This chapter has for the first time applied a multilevel analysis to maritime transport
and, in particular, to liner-shipping networks formed by the daily circulation of
container vessels on a global level. What can be learned from such an attempt?
First, we confirm that liner shipping is characterized by high complexity, while
the network properties of seaports interestingly are in line with the usual port
hierarchies provided by economic intelligence and earlier academic studies on port
performance. Second, the clustering methodology applied to the world graph of
direct and indirect inter-port links, with or without traffic, provides a rather unusual
combinations of ports that do not correspond to our vision of the world delineated
by political and cultural criteria.
Thus, retrieving the inherent functional, spatial, economic - and perhaps acci-
dental - logic of each cluster would necessitate enormous efforts of data checking
from the raw dataset itself. What should port specialists understand from the results?
Some clusters are not easy to “define” because they do not exactly overlap with the
usually defined geographical areas or systems in port geography, such as port ranges
and hub-port regions. For instance, Yokohama appears as the most central port in
a Northeast Asian cluster, whereas all indications in the literature are that Busan
(South Korea) has the strongest position in this region as a hub. This is perhaps
where the methodology follows a somewhat different direction than the mainstream
literature: hub ports (i.e., stars) where many links converge have a relatively lesser
importance in the results because instead of forming - or belonging to - small
worlds, they tend to create hierarchy and polarity in the system.
However, there are very positive lessons to learn from the application of such
methods to liner shipping for port and maritime geography. Despite the “automatic”
dimension of clustering, strong coherence is found in a number of clusters where
either geographic or commercial/cultural logic dominates, as seen with transatlantic
groupings or South-South groupings. Surprisingly, we did not find many clusters
including the biggest ports of the Triade as altogether dominant poles of a global
system (e.g., North America, Western Europe, and East Asia). This indicates that
the world maritime system is regionally polarized but not globally polarized: some
corridors link different regions through a complex mix of pendulum, round-the-
world, and other line-bundling services, while intraregional services tend to be most
concentrated on a few ports. This confirms that liner shipping does not exactly
overlap with trade routes: the hub strategies of global ocean carriers have rerouted
the dominant flows toward many hub ports in the Caribbean, the Mediterranean,
and Asia as a whole, for the aforementioned reasons. Perhaps further research
focused on finding a better match between the world economic system and the world
maritime system would “delete” such hub ports from the map. Thus, the question is
as follows: which ports are more hubs than others? Unfortunately, the answer or the
methodology to answer is not yet available in the literature.
Further research should follow several key directions to improve the analysis.
First, selecting only a category of vessels of a certain size would allow for the results
Search WWH ::
Custom Search