Orchinik, R., Dubey, R., Gershman, S., Powell, D., Bhui, R. (2023). Learning about scientists from climate consensus messaging. Proceedings of the Annual Meeting of the Cognitive Science Society.

Abstract: Informing people of the overwhelming consensus among climate scientists that human-caused climate change is occurring increases belief in the proposition and the importance of policy action. However, consensus may not be interpreted in the same way; it could emerge from skilled experts converging on the truth, or a biased cabal working for their own gain. We show that the weight that an individual places on the skill and bias of experts affects whether they are persuaded by strong consensus. We demonstrate that beliefs about the skill and bias of pro-consensus scientists (those who express that climate change is occurring) and anti-consensus scientists (those who do not) are central components of a belief system about climate change, determining what individuals learn from climate scientists. However, these characteristics are not fixed as individuals also learn about scientists from consensus. In this way, people learn both from and about climate scientists given consensus.

Hattersley, M., Orchinik, R., Ludvig, E., Bhui, R. (2023). Preferences for descriptiveness and co-explanation in complex explanations. Proceedings of the Annual Meeting of the Cognitive Science Society.

Abstract: Good explanations can be distinguished from bad ones in different ways, for instance by how much of the available information they can explain (i.e., maximise the likelihood of) the available data. Here, we consider two different components of likelihood: descriptiveness (the likelihood of the individual data points) and co-explanation (the likelihood of the specific subset of data under consideration). We consider whether people prefer explanations that are high in descriptiveness vs. coexplanation. Moreover, we consider whether people who endorse conspiracy theories prefer explanations for either quality. In a medical diagnosis task, participants make binary choices between two fictional disease variants: one higher in descriptiveness versus another higher in co-explanation. Overall, participants displayed a weak preference for descriptiveness. This preference, however, did not vary across increasing levels of descriptiveness. Moreover, such preferences were unrelated to conspiracy mentality. Thus, both explanatory virtues may play a role in the appeal of likely explanations.

Misinformation

Uncommon Errors: Adaptive Intuitions in High-Quality Media Environments Increase Susceptibility to Misinformation, with Cameron Martel, David Rand, Rahul Bhui
[Invited resubmission: Management Science]

Abstract: Belief in misinformation has been linked in part to digital media environments promoting reliance on intuition -- which in turn has been shown to increase belief in falsehoods. Here we propose that this apparently irrational behavior may actually result from ecologically rational adaptations to complex environments. In a large survey experiment, we test whether intuitive belief in misinformation may result from these rational adaptations by randomizing participants to be shown either a largely true or largely false news feed. We show that individuals make more frequent and quicker errors on the less common headline type, and less frequent errors on the more common headline type. After seeing many true headlines, a participant is more likely to misidentify a subsequent false headline as true, and vice versa after seeing many false headlines. This pattern is consistent with adaptation to the proportion of true and false content (the veracity base rate).  We use computational modeling to show that these differences are driven by intuitions, which correspond to Bayesian priors, about the veracity of the content -- intuitions which then spill over into new environments. Our results, when paired with the observation that the news consumed by most Americans is overwhelmingly true, suggest that belief in misinformation and the intuitions that underlie it are not necessarily a failing of humans in digital environments but can be a byproduct of rational adaptations to them.

Blatantly false news increases belief in more plausible falsehoods, with David Levari, Cameron Martel, Paul Seli, Rahul Bhui, Gordon Pennycook, and David Rand
[Under review: PNAS]

Abstract: What are the consequences of exposure to blatant falsehoods and “fake news”? Here we show exposure to highly implausible claims can increase belief in more ambiguous false claims, as they seem more believable in comparison. Participants in five preregistered experiments (N=5,476) were exposed to lower or higher rates of news headlines that seemed blatantly false, as well as some more plausible true and false headlines. Being exposed to a higher prevalence of extremely implausible headlines increased belief in more ambiguous headlines, regardless of whether they were actually true or false. The effect persisted for headlines describing hypothetical events and actual news headlines, whether headlines were actively evaluated or simply read passively, among liberals and conservatives, and among those high or low in cognitive reflection. Our findings emphasize the importance of reducing exposure to fake news.

Pre-registered Replication and Conceptual Extension of Effron and Raj (2020), with Rahul Bhui and David Rand
[Invited Resubmission (Proposal): Nature Communications]

Abstract: Effron and Raj (2020) found that repeat exposure to misinformation reduces moral condemnation of those falsehoods. This finding is concerning, given that moral condemnation can play an important role in stopping the spread of online misinformation. We propose a registered report conceptually replicating this finding and an investigation of its generalizability, using an updated and much larger set of false headlines. We will also investigate whether asking for accuracy evaluations of the headlines, a type of accuracy nudge present in the original paper, alters the effect as inattention to the veracity of headlines may decrease outrage and, thus, moral condemnation.

Climate Communications

Learning from and about climate scientists, with Rachit Dubey, Samuel Gershman, Derek Powell, and Rahul Bhui

Abstract: Despite the overwhelming scientific consensus that human activities contribute significantly to climate change, public opinion remains divided. To bridge this gap, informative messaging about the consensus has been widely proposed as a persuasive tool. However, it remains challenging to understand how people interpret this information and how it interacts with their wider belief system. Using survey experiments that vary the level of scientific consensus, we find that consensus information not only influences climate change beliefs but also shapes perceptions of climate scientists themselves, consistent with normative principles of Bayesian belief updating. Notably, climate beliefs are strongly linked to perceptions of scientist skill which highlights a potential avenue for communication. By unpacking the belief system underlying one of the most prominent climate communication strategies, our research provides a deeper understanding of public response to consensus messaging, offering guidance for developing more targeted and effective science communication.

Pro-Climate Statements from Elon Musk can Persuade Republicans on Climate Change, with David Rand
[Under Review: Nature Sustainability]

Abstract: Addressing human-caused climate change is of paramount importance but a large minority of Americans either do not believe that human-caused climate change is occurring or do not think it is a policy priority. Resistance to addressing climate change is particularly concentrated on the American political right, a group that is difficult to persuade. Leveraging technologist Elon Musk's embrace of the right, we design and test an intervention that shows Republicans the pro-environmental stances that Musk has taken. We find that the treatment significantly increases Republicans' climate beliefs and intended actions. The predicted treatment effect is positive for almost all participants and there is no impact on participants attitudes towards Musk, suggesting that this is a low-cost and scalable intervention with minimal risk.

Getting Smart on Green Branding, with Santiago Pardo Sanchez and David Rand

Abstract: Implementing existing "green" digital technology at scale could lead to a 20% reduction in global emissions annually. However, climate change has become a central focus of the "culture wars" providing a significant hindrance to the adoption of this technology. Using two survey experiments (N = 3,053), we show that there are partisan differences in purchasing intentions when the same pro-environmental product is called "smart" or "green," with Democrats preferring "green" branding and Republicans preferring "smart." We show that a one-size-fits-all approach to branding pro-environmental products leaves a large environmental impact on the table. Using simulations of targeting regimes, we estimate that showing consumers the branding that aligns best with their political affiliation can increase purchasing intentions by 3%, performing just as well as a machine-learning-based targeting approach. We estimate that this 3% increase in purchasing intentions for just the five products tested in the study would reduce carbon emissions by 900 million tons by 2050.

Economics and Finance

What's the Difference? Measuring the Effect of Mergers in the Airline Industry, with Marc Remer
[Invited resubmission: Journal of Law and Economics]

Abstract: We analyze the effect of four US airline mergers using three retrospective techniques: standard difference-in-differences regression, synthetic control, and nearest neighbor matching. We study if these techniques can be reliably applied to study airline mergers, and find a number issues. For example, routes typically specified as controls are potentially impacted by the merger, and the more advanced techniques may be subject to unobservable variable bias. As such, the estimated effect of a merger does not always align in direction or statistical significance across the three methods. We also expand the analysis beyond the average treatment effect on overlap routes. We find that the increase in multi-market contact from each merger caused prices to increase on a large set of non-overlap routes. Lastly, with the aim of providing guidance to antitrust agencies, we analyze the determinants of route-level price effects on overlap routes. We find that even in mergers where the average effect is different from zero (e.g. American/USAir and United/Continental), there is substantial heterogeneity across overlap routes. The route-level effect is predictable based on pre-merger observables, such as the level of HHI.

Market Structure and the Availability of Credit: Evidence from Auto Credit, with Donghoon Lee and Michael Lee

Abstract: How do changes in market structure impact accessibility to credit? Following the introduction of publicly disclosed Comprehensive Capital Analysis and Review stress tests, market shares of affected banks shrunk by about 2.1 pp. Impact significantly differs across regions, with shares dropping an additional 1.1 pp in non-urban areas. Credit substitution by other lenders is imperfect. Originations drop in areas more reliant on CCAR credit. Borrowing conditions diverge as well: non-urban counties see fewer subprime originations and average borrowing costs rise. Our results suggest that policies geared toward financial stability asymmetrically impacted local credit conditions and inadvertently amplified the urban-rural divide.