Communicating the uncertainty of synthetic indicators: a reassessment of the HDI ranking

Composite indicators convert information about different facets of a given phenomenon into a single figure. Unavoidably, the “conversion process” involves a high level of arbitrariness, which, in general, makes the results not robust. The approach to composite indicators used in this paper aims at mitigating this problem and makes final users more aware of the unavoidable uncertainty of the results (e.g. rankings) based on a given composite. We illustrate our approach by applying it to the Human Development Index.