Impact of global terminal operators on port efficiency: a tiered data envelopment analysis approach

Abstract

This study evaluates how different types of global terminal operators (GTOs) – global stevedores (GS), global hybrids (GH), and global carriers (GC) – shape port efficiency. Both crane efficiency and relative technical efficiency of ports, measured through tiered data envelopment analysis, are examined based on levels of GTO participation in container production. The participation of GS in the terminal operation of a port induces higher crane efficiency. Yet this efficiency is not effectively transformed into port level efficiency. This is related to the GS's intensive utilisation of ship-to-shore gantry cranes, while the landside operation is greatly influenced by different regional practices and hinterland distribution networks. In contrast, carrier-controlled global operators and dedicated leased terminals do not produce the same efficiency levels as other GSs do. Their major incentive to participating in terminal operation is to secure accessibility to key regional port facilities. The implication is that port authorities may opt to increase levels of intra-port competition to address efficiency concerns at a port level.

Keywords:

• global terminal operators
• tiered data envelopment analysis
• port efficiency
• container ports

Acknowledgements

I sincerely thank the guest editors and anonymous reviewers for their helpful comments on this article.

Notes

After calculating HS _i , the natural log was taken such as ln(HS _i +1) to make Equation (2) as an index having values larger than zero for each port.

In estimating the index, it is assumed that the economic activities of country J are universally distributed throughout the country. This allows us to measure the distance between a port and a country based on the central point of a country. This may be a source of error, especially for large countries like Russia, India, US, Mexico, and Brazil, if their economic activities are unequally distributed. One possible approach to correcting this would be to divide the world (or these countries) into smaller zones, with economic activity data for each zone, and then calculate the index. However, the data required for the ideal calculation is difficult to acquire for all countries.

If the measurement method is not the same for a certain country between the two sources, we adjusted the CIA data to match the WDI data.

Access to the database from October 2004 to January 2005.

Based on the model with three inputs, I also conducted the same analyses presented in the sections 4.2. The results are consistent with the results based on the model with four inputs, while they are not presented here.

Welch's test of equality of means is more robust than ANOVA when group variances and sizes are unequal. Brown and Forsythe's F-test of equality of means is more robust than ANOVA when groups are unequal in size and deviations from the group means are highly skewed.

The definitions of specific regions are based on Containerisation International Yearbook 2005.

Table 1 is organised as a table based on the author's summary of DSC (2004).