The empirical application of the spatial theory of voting in multiparty systems with random utility models

Spatial models of voting predominate in the formalization of political decisions and continue to be a growth industry in political science. But strict empirical applications of this theory have been rare. Only recently, conditional discrete choice models have been proposed to fill the gap between formal and empirical models and to predict the individual voting decision in multi-candidate/multiparty contests on the basis of the spatial model. This article highlights several flexible features of these models that are well known in transportation economics and applied marketing science but not yet discussed in the electoral studies community. Empirical illustrations are provided on the basis of (nested) multinomial logit.