On the use of composite indices in economic history. Lessons from Italy, 1861-2017

We argue against the use of composite indices, such as the Human Development Index, in economic history. We show that composite indices can be interpreted as paternalistic social welfare functions (PSWF), and therefore are nothing more than a formal representation of the analyst’s ethical system. This contrasts with the use economic historians typically make of composite indices, as tools to lend objectivity to the measurement of multidimensional phenomena. We support our claim by introducing a new constant-elasticity-of-substitution SWF family, which a) encompasses all composite indices put forth by the literature, and b) identifies the analyst’s implicit preferences by means of standard tools, e.g. marginal rates of substitution and elasticity of substitution parameters. The theoretical framework is illustrated by an empirical investigation of the long-run evolution of Italians’ living standards (1861-2017). We show how any history based on composite indices is one where both data and history play a minor role, if any.