Performance evaluation of Turkish disaster relief management system in 1999 earthquakes using data envelopment analysis

This study estimates the relative efficiency of disaster relief organizations that participated in the relief activities following the Marmara and Düzce earthquakes that took place in Turkey in 1999. For this purpose, the activities in the response and relief phases of the disaster management cycle following the 1999 earthquakes are classified into four groups: “search and rescue,” “infrastructure rehabilitation and debris removal,” “health care” and “providing basic needs to the disaster victims.” Next, the unbounded data envelopment analysis (DEA) model is applied to estimate the efficiency scores and optimal input/output weights of the decision-making units (DMUs) in these groups. The efficient and inefficient units are then determined, and the target input and output values are determined for the inefficient units. However, some inputs and outputs are found to be zero or are outlier values in the unbounded model results. To eliminate this deficiency, the bounded model is applied by including additional constraints in the unbounded model. After determining a solution with this model, the optimal input/output weights, the efficiency scores of the DMUs and the efficient and inefficient units are determined. The results of the bounded version of the DEA are more accurate than those of the unbounded model. More accurate and reliable results are obtained from the unbounded model because zero and outlier values are eliminated. Copyright Springer Science+Business Media Dordrecht 2015