How robust is the Foreign Policy/Kearney index of globalisation?

This short paper argues that the widely-known Kearney/Foreign Policy (KFP) index of globalisation is constructed by making some problematic assumptions about the normalisation and weighting of the variables included in the index. We propose alternative normalisation and weighting rules, and using these rules, recalculate the ranking of the fifty countries, using the original KFP data. Specifically, we use the statistically optimal weights obtained by principal components analysis, and a normalisation rule that treats different years of observations separately. We find that up to 85% of countries change position under the new rules, some of them substantially