The Uses and Abuses of Election Models for Political Participation

Paulina Oppermann

Leibniz Universität Hannover

Often cited as one of the successes of predictive modelling in the social sciences, election models have recently taken center stage in a critical discussion about predictive limitations, false interpretations of certainty, as well as self-reinforcement between models and their explananda. The aim of this talk is to apply established theories in philosophy of modelling to different types of election models, focussing on their ethical and political implications. I will provide an ontological differentiation between predictive and explanatory types of election models, their epistemic merits and shortcomings, and their influence in shaping political discourse. Focussing on the case study of the 2016 US presidential election, I will argue that predictive election models in particular are of interest because of a public misconception of their modelling abilities, namely that these models are ontologically build not to model an election outcome, but a hypothetical outcome of an election held under current circumstances. Combined with misconceptions of certainty and probability, this leads to a problematic influence of models on voting behavior, as well as to a gap between the a priori interpretation and the a posteriori outcome of an election. On the other hand, one might argue that these election models still provide necessary background for an informed and rational decision of voters, thus contradicting the notion that election models (and opinion polls in particular) need to be strictly regulated or perhaps even banned within the critical phase of an election campaign. For example, in multi-party first-past-the-post elections, a rational agent prone to vote against a certain party might wish to know which candidate has the best chance of winning against said party in his electoral district. I will conclude my remarks by connecting the issues discussed to options for further research into related fields in political science, political philosophy, and model theory.

Chair: Robin Waldenburg

Time: 05 September, 14:00 – 14:30

Location: SR 1.004