Abstract. This paper is concerned with Caplan’s (2000, 2001, 2006) rational irrationality model and its application on electoral preference formation. According to this model, individuals consume irrationality in a rational manner i.e. they maximize their total welfare through an exchange between their material and psychological welfare. Using the contributions of an important debate from the philosophy of economics regarding the imposition of ontological constraints on models, unrealistic assumptions and explanatory unification, two arguments are formulated: First, Caplan’s model is ontologically bewildered – it contains opposed statements about the psychological descriptiveness of rational irrationality. Second, even though Caplan is apparently aware of the tension between ontological and tractability considerations he seems to privilege the latter.

Keywords: rational irrationality, unrealistic assumptions, realism, ontological constraints, electoral preferences.