Efficiency assessment of container operations of shipping agents in Spanish ports

In this paper a two-stage Data Envelopment Approach is used to assess the relative efficiency of container shipping agents operating at Spanish ports, and studying the factors influencing it. In the first stage, an input-oriented, Variable Returns to Scale (VRS) model is used to compute efficiency scores of the different shipping agents. The model considers labor as input, and numbers of loaded and unloaded containers handled as outputs. Scale efficiency, returns to scale and average efficiency of shipping agents at each port are reported. In the second stage, different regression approaches are applied to relate the efficiency scores obtained to a number of exogenous variables. The results identify as significant some of these variables such as the number of container lines with which it operates. Belonging to each of four clusters identified from the dataset seems also to have a significant influence on the efficiency of the studied agents.