I show that the effect of heterogeneity on contest investments depends on the structure of the competition, which implies that heterogeneity matters for optimal contest design. This insight helps to explain empirical evidence on the comparison of different contest structures.

The article discusses Stein Rokkan's research strategy, i.e. that of retrospectively explaining structural variations, in the light of the crucial problem of modern social science, namely: how to reconcile the sociology of structures and the sociology of action. The authors' hypothesis is that...