Abstract
This paper presents a novel Directional Distance Function (DDF) Benefit-of-the-Doubt (BoD) model to construct a composite indicator (CI) for measuring Sustainable Development Goal 3 (SDG 3), which aims to ensure healthy lives and promote well-being for all at all ages, in the Organisation for Economic Cooperation and Development (OECD) countries. To simultaneously account for both desirable (e.g., health service coverage) and undesirable (e.g., mortality rates) public health indicators, a DDF version of the BoD model is used. However, the DDF model suffers from limited discriminatory power due to the large number of indicators. To improve the discriminatory power of the DDF, a Joint Variable Selection DDF (JVDDF) model is introduced by linking indicators' weights to the objective function using a penalty function. To validate the robustness of the proposed model, various cross-efficiency BoD and Principal Component Analysis (PCA) models are applied. Finally, a fuzzy programming model is used to aggregate the different CI scores obtained from these models. The results of the fuzzy combination model indicate that Australia, Norway, Sweden, Iceland, and Israel are the top performers in achieving SDG 3's health and well-being targets.