Understanding Voters' Preferences: How the Electorate's Complexity Affects Prediction Accuracy and Wishful Thinking among Politicians with Respect to Election Outcomes



Models of political decision making often assume that politicians are fully informed on voters' preferences. Still, while they may be privileged witnesses of the democratic process, real-world politicians typically act upon imperfect information. Using a large scale survey among (498) Belgian local politicians we analyse whether and to what extent politicians are informed on public opinion. More precisely, we analyse their predictions of the electoral result of their own party in an upcoming election. The focus is on the impact of the electorate's complexity on this prediction accuracy. Complexity is defined in terms of the size and heterogeneity of the electorate. Complexity is found to affect both cognitive (prediction accuracy becomes smaller) and affective (predictions are biased through wishful thinking) processes underlying the politician's prediction. The empirical results show that smaller electorates have better informed politicians (with smaller prediction errors and less subject to wishful thinking), offering a case for decentralized government. Heterogeneity has a bearing on prediction accuracy and wishful thinking. Interestingly, the effects differ between politicians in power and those in opposition.