Periodical measurement of multidimensional phenomena

Abstract

Synthesizing multidimensional phenomena such as well-being in the form of composite indicators has been gaining popularity in recent years. The Mazziotta-Pareto Index is one of the methods of constructing non-compensatory composite indices. The paper proposes a modification of this method, which (unlike the original) enables periodical measurements and can be used to compare countries in research on East European transition. The essence of this modification is the use of anti-pattern normalization, during which only current data is used, and yet after normalization, the indicators are in a certain way comparable over time. The Anti-Pattern Normalized Mazziotta-Pareto Index (APMPI), unlike the original, does not change the previously determined values after the inclusion of new data. Both the imperfection of the original approach and the new proposal are illustrated by an empirical example. Indices of well-being for the OECD countries are constructed. The example shows that although AMPI and APMPI values are not comparable, the rankings based on them are not very different.