The Human Development Index with Multiple Data Envelopment Analysis Approaches: A Comparative Evaluation Using Social Network Analysis

Abstract

AbstractThe objective of this work is to use multiple Data Envelopment Analysis (DEA)/Benefit of the Doubt (BoD) approaches for the readjustment and exploitation of the Human Development Index (HDI). The HDI is the leading indicator for the vision of “development as freedom”; it is a Composite Index, wherein three dimensions (income, health, and education), represented by four indicators, are aggregated. The DEA-BoD approaches used in this work were: the traditional BoD; the Multiplicative BoD; the Slacks Based Measure (SBM) BoD; the Range Adjusted Model (RAM) BoD; weight restrictions; common weights; and tiebreaker methods. These approaches were applied to raw and normalized HDI data from 2018, to generate 40 different rankings for 189 countries. The resulting indexes were analyzed and compared using Social Network Analysis (SNA) and information derived from DEA itself (slacks, relative contributions, targets, relative targets and benchmarks). This paper presents useful DEA derived indexes that could be replicated in other contexts. In addition, it contributes by presenting a clearer picture of the differences between BoD models and offering a new way to appreciate the world's human development panorama.