Determinants of maritime transport costs - a panel data analysis for Latin American trade

This paper analyses the determinants of maritime transport costs for intra-Latin American trade over a period of 6 years (1999-2004). The data refer to yearly disaggregated (five-digit level Standard International Classification) containerised maritime trade flows on 277 trade routes. With this data-set, a transport costs equation is estimated using linear regression analysis in a panel data framework. The first contribution to the literature is to exploit the greater variability present in the data and to control for unobservable heterogeneous effects. The second is to investigate the role of distance as a determinant of international maritime transport costs in comparison to the location within the liner services network and the potential impact of being peripheral in the liner shipping network. Finally, the work analyses influence of open registries on the variability of maritime transport costs. To the authors' knowledge, this has not been done previously. Three groups of explanatory variables are considered. Firstly, time variant variables: use of open registries and trade imbalance. Secondly, variables related to liner shipping network structures: number of liner services; shipping opportunities; deployed ships and deployed TEUs. Finally, product-related variables such as volume of shipment, value of product and special characteristics of the cargo (i.e. refrigerated cargo). The results allow the quantification of the effects of explanatory variables on international maritime transport costs and to compare the obtained elasticities with previous cross-section analyses. In particular, estimating the impact of the use of open registries on maritime transport cost is a new contribution in this field that could provide policy-makers with valuable information to be used in the implementation of economic policies.