Measuring composite indicators using a multiplicative and multilayer DEA model: a case study on the Digital Economy and Society Index (DESI)
Lamriq Rabii, Doukkali Abdelaziz, Belkhayat Najib

Abstract
A Composite indicator based on data envelopment analysis (DEA) has gained significant consideration in recent years. However, the large number of indicators as well as the ratio measures makes the use of DEA models less effective in the composite indicators. In the literature, many articles have proposed approaches based on DEA models to address the problem of ratio measures. Others researches to address the problem of discrimination related to the data dimension. In this paper, we have developed an indicator composite based on a multiplicative and multilayer DEA model (DEA-MM). The multiplicative DEA model is based on the concept of the geometric mean of invariant units which is highly desirable when all measures are in ratios. In addition, the use of the hierarchical multilayer structure of the indicators allows the DEA-MM model to produce a high degree of discrimination between the scores of the decision units. A numerical application on the Digital Economy and Society Index (DESI) is also illustrated in this paper.

Full Text: PDF