Towards a new measure of a country’s competitiveness: applying canonical correlation

This paper aims to develop and implement a new approach of ranking countries according to their level of economic competitiveness.

Competitiveness is conceptualized as the degree of participation in global integration, measured by the levels of exports, capital inflows and immigration. Based on a wide range of development indicators, composite indicators are constructed by means of a principal component analysis. In turn, these composites enter a canonical correlation analysis and estimated canonical weights are used as weights in the calculation of index scores.

Measures of institutional and political quality are shown to be most closely connected to a country’s ability to export and to attract foreign factors of production. Inflows of foreign capital turn out to correlate much less with our weighting composites than export volumes.

Regarding prospects for future research, a broad range of further applications of this approach in the macro area is conceivable, including measures of a country’s quality of schooling, the quality of its health care system or of its governmental institutions.

In comparison to prevailing macroeconomic indices, this method of index construction is superior in at least two respects. First, direct measures of a country’s performance are methodologically separated from mere performance drivers, contributing to higher clarity concerning the concept to be measured. Second, weights of performance drivers are not determined arbitrarily or based on subjective judgment, but emerge from a transparent and well-established statistical procedure.