Equally, even formal game theoretical models rely on assumptions being made about the preferences of the players, while Delphi and prediction markets are self-evident instances of collective judgment, the accuracy of which also depends on the way in which questions are framed, an aspect they share with opinion polls, and it is safe to say that both expert opinion and scenario analysis are based on subjective processing of data. In this vein, it can be argued that the role of expert judgment is cross-cutting – after all, its importance as a source of political intelligence has regularly emerged in the literature, as explained in the previous

chapters. In addition, as already suggested by Bunn and Wright (1991) and confirmed by Tetlock and Gardner (2015), interaction between judgmental and statistical forecasting methods is desirable and fruitful. Thus, in view of its relevance with respect to the whole body of knowledge related to PRA, the subject deserves specific attention. The rest of the chapter unfolds as follows: Section 2 explores the diverse meanings attached to the concept of expertise, focusing on references to ‘experts’ in the political risk and forecasting literatures, drawing a clear-cut distinction between expert judgment in the sense of forecasting accuracy on the one hand and expert judgment based on political knowledge needed for the framing and modeling of political risks on the other. To better illustrate how these two concepts can and indeed should be kept separate, Section 3 discusses the role of political knowledge in a realm which is distinct from forecasting yet very salient to PRA (that, is the operationalization of political variables), while Section 4 discusses the most critical aspects in ‘human’ – that is, not depending on specific knowledge – judgment, which have to do with forecasting skills irrespective of (political) knowledge. Section 5 concludes arguing in favor of a reappraisal of the concept of political expertise based on enhanced accountability of the expert.