On the Appropriate Interpretation of Revealed Comparative Advantage Indexes

Abstract

In this study, Balassa’s revealed comparative advantage index and the indexes of logarithmic revealed comparative advantage, symmetric revealed comparative advantage, weighted revealed comparative advantage, which are defined as an increasing function of the initial one, have been investigated theoretically and empirically from the point of consistency. In theoretical investigating the apparatus for evaluating consistency is the notion of Pearson’s sample correlation. Empirical investigating has been based on the foreign trade data for the year of 2020 of four digits subsectors of Turkish manufacturing industry classified according to ISIC (Rev. 4). The findings obtained from both theoretical and empirical investigating can be listed as follows: If any pair of the indexes mentioned above depend on each other linearly, then they are perfectly consistent as cardinal measure; but if not, they are not perfectly consistent. Moreover, even if they are depend on each other linearly or not, the following result is valid all the time: They are perfectly consistent without no condition when they are interpreted as ordinal measure; and are also perfectly consistent when they are interpreted as dichotomous measure provided that their neutral points are compatible.