Alternative versions of the global competitive industrial performance ranking constructed by methods from social choice theory

The Competitive Industrial Performance index (developed by experts of the UNIDO) is designed as a measure of national competitiveness. Index is an aggregate of eight observable variables, representing different dimensions of competitive industrial performance. Instead of using a cardinal aggregation function, what CIP’s authors do, it is proposed to apply ordinal ranking methods borrowed from social choice: either direct ranking methods based on the majority relation (e.g. the Copeland rule, the Markovian method) or a multistage procedure of selection and exclusion of the best alternatives, as determined by a majority relation-based social choice solution concept (tournament solution), such as the uncovered set and the minimal externally stable set. The same method of binary comparisons based on the majority rule is used to analyze rank correlations. It is demonstrated that the ranking is robust but some of the new aggregate rankings represent the set of criteria better than the original ranking based on the CIP.