Multi-level DEA for the construction of multi-dimensional indices

Data Envelopment Analysis (DEA) is a non-parametric, mathematical programming method that is used to evaluate the performance of Decision Making Units. One variation of the method is focused on expanding the number of stages: inputs are transformed into intermediate measures and in turn those are transformed into outputs. DEA and its variations have been used to construct composite indicators. The purpose of the current paper is to propose a new variation of DEA that relies on a two-stage model for the construction of multi-dimensional indices. The proposed variation: • Uses a two-stage DEA model for the calculation of each sub-indicator that will be integrated into the final index; • All the sub-indicators are integrated into the final index with the use of a Benefit-of-the-Doubt mathematical programming model. As it was mentioned, the proposed method can be used for the construction of multi-dimensional indicators and in the current paper is used to calculate the sustainability of the EU-28 countries.