measuring global economic activity (replication data)

A number of economic studies have used a proxy for world real economic activity derived from shipping costs. This measure turns out to depend on a normalization that has substantive consequences, of which users of the index had been unaware prior to this paper. This paper further evaluates this and alternative measures in terms of treatment of trends, coherence with world output, and correlation with commodity prices. I conclude that measures derived from world industrial production offer a better indicator of global real economic activity.