The Myth of the (Ir)rational Voter? Theoretical and Methodological Advancements for Studying Voter Rationality

Research question and goal

According to the democratic ideal, voters elect politicians who respond to the will of the people by implementing policies that voters want. However, if voters do not know what they want or constantly change their minds about what they want, governments cannot be responsive. The quality of democratic government, thus, depends on the rationality of the electorate. Are voters up to this task? Or are they, as many prominent scholars claim, irrational? This is the principal research question that we address in this project.

The view of voters as rational, meaning that they have stable preferences resting on wellfounded beliefs, has been heavily contested within political science in the last two decades. For example, in their highly influential study Democracy for Realists, Achen and Bartels (2017) argue that voters are irrational and that, in the unlikely event that voters should have stable preferences, these are characterized by blind following of their favoured parties, cognitive biases and misinformation. This has severe normative implications. If voters do not know what they want or do not want what is best for them, why should they have a voice at all? This has led scholars to argue that the influence of voters, and thereby democracy itself, should be curtailed in order to improve policy making (e.g., Caplan, 2011). For example, in his 2017 book Against Democracy, published on Princeton University Press, Brennan (2017) argues that suffrage should be restricted to knowledgeable citizens for exactly this reason.

We believe that scholars jump to conclusions, when they recommend circumscribing fundamental democratic rights in face of alleged voter irrationality. We argue that scholars arrive at the conclusion that voters are irrational by holding voters to an impossible, and even undesirable, ideal of rationality that few scholars themselves would live up to. As a consequence, much of the empirical evidence of voter irrationality is inconclusive and, at times, may even support the alternative hypothesis that voters are rational.

Replace a naive notion of ”perfect rationality”

We improve on the state of the art by theoretically and methodologically incorporating insights from the theory of bounded rationality (Simon, 1957). The goal is to replace a naive notion of ”perfect rationality” with a form of practical rationality that is compatible with the capacities that are actually possessed by real people.

We approach the overarching question of voter rationality through two subquestions:

1. Do voters have meaningful and stable opinions? We argue that voters do not need to have meaningful preferences over all policies (cf. Converse, 1964; Freeder, Lenz and Turney, 2019), but only on the policies that they personally perceive to be important. On these issues, however, their preferences should be ideologically consistent and they should not easily be persuaded to change their minds.
2. Do voters’ preferences rest on wellfounded beliefs? We argue that the literature on confirmation bias, the apparent pattern that voters seek out information which confirms their beliefs (cf. Nyhan and Reifler, 2010; Taber and Lodge, 2006), is confounded and may well be explained by voters seeking out sources they trust. When information comes from sources they trust, however, they should be willing to learn from this information and change their minds even if this implies psychological discomfort for the voter.

Are voters irrational?

This is one of the major and perennial questions within political science, yet, the state of the art cannot offer a conclusive answer. We address the gaps in this literature with three carefully designed experimental studies and one theoretical synthesis, which all engage directly with this important question. With the theoretical and methodological advancements in this project, we believe that we are providing voters with a fair chance to reclaim their rationality. Should voters also fall short when measured against the realist ideal of rationality that we purport, this will be a significant blow to the democratic ideal. Regardless of the empirical results, we are convinced that this project will fuel both theoretical, empirical and normative investigations of the rationality and desirability of democracy.