Environmental efficiency of seaports correcting estimation biases: Weak disposability and bootstrapping method

Previous studies on port environmental efficiency using DEA tended to be biased due to two main sources: (1) erroneous disposability assumption between desirable and undesirable outputs and (2) ignorance of data sampling bias. This study developed a robust novel methodology for measuring environmental efficiencies to correct for both estimation biases. First, to address the disposability bias between desirable and undesirable outputs, the weak disposability condition is imposed on the model-building process. Also, to solve inherent bias caused by the data sampling process, a bootstrapping method was adopted. Building on slacks-based measurement (SBM)-DEA models, a subsampling technique is adopted as a bootstrapping method. This paper suggests a novel approach to evaluating the volatility of the confidence intervals of the measured efficiency score. As a result of empirical study on the environmental efficiency of 9 Korean ports in 2019 ∼ 2021, we can examine the impact of environmental policies on port performance.