Abstract

Over the past two decades, there has been a growing interest in composite indicators as a means of synthesizing information about multidimensional phenomena, which can aid policy decisions. However, the normative nature of composite indicators has been criticized, as has the use of a single indicator to describe complex phenomena. In this paper, we propose a new model called the Ultrametric Composite Indicator (UCI) that uses an extended ultrametric correlation matrix to reconstruct the hierarchical relationships among manifest variables and their associated latent concepts. The UCI model generates a tree structure where the leaves represent the manifest variables, internal nodes correspond to first-level and higher-level Specific CIs (SCIs) associated with specific and broader dimensions, and the root represents the General CI that describes the multidimensional phenomenon. We also introduce a test for identifying relevant dimensions and retaining statistically significant higher-level SCIs. We apply the UCI model to study circular economy and provide information on EU countries' behavior regarding circular material use, which can help guide policy decisions for EU governments.