Recent results indicate that, in spite of the skewness of citation distributions, the ranking of research units that focus on the upper tail of citation distributions is quite similar to the ranking one obtains with average-based indicators. This paper explores the conjecture that this can be explained because country differences in international comparisons have a strong scale factor component. If this is the case, it is argued that the effect on overall citation inequality of differences in citation impact across countries should be drastically reduced when raw citation counts are normalized with the countries’ mean citations. This is what we find for a large Thomson Reuters dataset of articles in Physics and the all-sciences case published in 1998-2003 with a five year citation window, and a partition of the world into 36 countries and two residual geographical areas. We conclude that international comparisons in terms of countries’ mean citations appear to capture most of the differences over the entire support of country citation distributions
