What do fragility indices measure? Assessing measurement procedures and statistical proximity

This article examines nine fragility indices. Their content validity is assessed by reviewing conceptualization, measurement and aggregation methods. Their convergent/divergent validity is assessed via principal component analysis and multidimensional scaling. These techniques are capable of determing the dimensionality of and the statistical proximity within the examined sample of indices. Both the conceptual and the statistical analysis support the hypothesis that there is a group of "holistic" fragility indices which are of little use for investigating the causes and consequences of fragility. The remaining indices address more specific aspects of fragility and produce empirically distinguishable results.