DEA models for identifying sensitive performance measures in container port evaluation

Abstract

In the container port performance evaluation, the heterogeneity of input or output measures often exist owing to non-controllable external factors such as geographical location, regional economy, political systems and so on. Therefore, it becomes imperative to investigate the individual input or output measures so as to identify their specific impact on the efficiency of a port. In addition, the purpose of performance evaluation cannot be a mere exercise of candidate ranking. The purpose of this article is to study ways of retaining efficiency and improving on inefficiency. The article utilizes the conventional Data Envelopment Analysis model to test the sensitivity of the individual input and output of a decision-making unit (DMU). For an efficient DMU, we measure how much an input can be increased, or an output decreased, without changing its efficiency status. For an inefficient DMU, we measure how much an input should be decreased, or an output increased, to make it achieve the best practice frontier. The new approaches are applied to the efficiency analysis of 77 global container ports. The results indicate that the number of berths and the capital deployed are the most sensitive measures impacting performance of most container ports. The analysis also reveals that container ports located in different continents behave differently.