How to Distinguish Motivated Reasoning from Bayesian Updating

Abstract

Many experimental and observational studies use the way that subjects respond to information as evidence that partisan bias or directional motives influence (or do not influence) political beliefs. For a natural and tractable formulation belief formation with both accuracy and directional motives, this is not possible. Any subject influenced by directional motives has a "Fully Bayesian Equivalent" with identical beliefs upon observing any signal. As a result, comparing how individuals or groups with different partisanship or priors respond to information has no diagnostic value in detecting motivated reasoning, even in a multivariate or dynamic setting. Conversely, providing a ``Bayesian rationalization'' consistent with a pattern of updating is not meaningful evidence for a lack of directional motives. These results have theoretical implications for the convergence of beliefs among those with directional motives and practical implications for empirical studies that aim to detect directional motives.

Formal Models of Authoritarian Regimes: A Critique

“…19 At stake here is the axiom of the independence of irrelevant alternatives. 20 On the general difficulty of distinguishing beliefs driven by "directional motives" from those based on different priors, see Little (2022). On the importance of emotions, see Greene and Robertson (2020).…”

Selection Neglect and Political Beliefs

“…Motivated reasoning is similarly difficult to distinguish from selection neglect (and is often difficult to distinguish from rational updating as well, see Little, 2021). Section 5 includes several examples of empirical findings that are consistent with both motivated reasoning and selection neglect as the underlying cognitive mechanism, including, for example, the finding that partisans believe that almost twice as many out-party members hold extreme views on a wide array of issues including immigration, gun control, racism, sexism, and police than they do in reality (Yudkin et al, 2019).…”

Biased Learning from Elections

“…is the Kullback-Leibler divergence from the objective belief to subjective belief f . This term captures the penalty for moving the motivated belief away from the objective belief; see Little (2021) for further discussion.…”

“…While there is certainly some truth to this, we argue there is value in considering that parties do sincerely hold beliefs with at least some directional motives. Little (2021) shows that while there are challenges to detecting motivated reasoning of this form (see also Koehler, 1993;Gerber and Green, 1999;Druckman and McGrath, 2019), some empirical designs do provide strong indirect evidence of motivated beliefs. And, as mentioned above, there is good reason to suspect that party elites have stronger directional motives about the popularity of their platforms than rank-and-file voters, in addition to evidence of biased beliefs consistent with what we find (e.g., Enos and Hersh, 2017;Broockman and Skovron, 2018).…”

“…We study how motivated reasoning affects parties' ability to learn from election results using the electoral model of Mack (n.d.) and the belief model developed in Little (2019Little ( , 2021. Electoral competition is a natural domain to study biased belief formation because election results typically reflect both the observed result and other features of the election.…”