Peter Schwardmann
Peter Schwardmann
Associate Professor of Behavioral Economics
Department of Social and Decision Sciences
Carnegie Mellon University
Porter Hall 319
4815 Frew St.
Pittsburgh
Embedded Files
WORKING PAPERS
Talking across the Aisle, with Luca Braghieri and Egon Tripodi, 2025
This paper presents the results of a large-scale experiment featuring natural face-to-face video conversations between Democrats and Republicans in the United States. We investigate both the drivers of self-selection into politically homogeneous conversations (echo chambers) and the effects of co- versus cross-partisan conversations on information aggregation and affective polarization. We identify a relative preference for co-partisan conversations that is explained by participants' pessimism about the hedonic and informational value of cross-partisan conversations. Participants' pessimistic expectations about the extent to which they can learn from counter-partisans are qualitatively correct, as they do by and large learn less in cross-partisan conversations. We show that this gap in learning is driven not by a lower potential for learning anchored in the way knowledge is distributed across party lines, but by the greater difficulty of extracting knowledge from counter-partisans. Participants' pessimism about the hedonic value of conversations is less warranted, as co- and cross-partisan conversations are deemed equally enjoyable ex-post. Moreover, cross-partisan interactions lead to a reduction in affective polarization that lasts for more than three months after the end of our experiment. Taken together, our findings suggest that policies that encourage cross-partisan interactions with the aim of reducing affective polarization and fostering information aggregation might be more successful at the former than the latter objective.
Rationalizations and political polarization, Yves Le Yaouanq and and Joël van der Weele, R&R at the Journal of Political Economy, 2025
We present a self- and social-signaling model formalizing findings in political psychology that moral and political judgments stem primarily from intuition and emotion, while reasoning serves to rationalize these intuitions to maintain an image of rationality. In social interactions, agents' rationalizations are strategic complements: others' rationalizations weaken their ability to judge critically and make their actions less revealing of (inconvenient) truths. When agents are naive about their own rationalizations, our model predicts ideological and affective polarization, with each side assigning inappropriate motives to the other. Cross-partisan exchanges of narratives reduce polarization but are avoided by the agents. In within-group exchanges agents favor skilled speakers, whose narratives worsen polarization. Our model explains partisan disagreements over policy consequences, aligns with empirical polarization trends, and offers insights into efforts to disrupt echo chambers.
Depression Stigma, with Chris Roth and Egon Tripodi, 2024
People with mental illness are frequently discriminated against and stigmatized. We introduce a new measure of perceived depression stigma and examine the causal effect of perceived stigma on help-seeking in a sample of 1,844 Americans suffering from depression. We document that a large majority of our participants overestimate the extent of stigma associated with depression. An information intervention aimed at reducing perceived social stigma decreases the demand for psychotherapy. Merely prompting individuals to think about stigma has a similar effect. A mechanism experiment reveals that our interventions increase optimism about future mental health, thereby reducing the perceived need for therapy.
Correcting Consumer Misperceptions about CO2 Emissions, with Taisuke Imai, Davide Pace and Joël van der Weele, R&R at the Journal of the Association of Environmental and Resource Economists, 2024
Policy makers put great emphasis on the role of information about carbon emissions in achieving sustainable decisions by consumers. We conduct two studies to understand the optimal targeting of such information and its effects. First, we conduct an incentivized and representative survey among US consumers (N = 1,022) to investigate awareness of climate impact and willingness to mitigate it. We find a large variation in the perceptions of the carbon emissions of different consumption behaviors, with an overall tendency to underestimate these emissions. We also find a positive but highly concave willingness to mitigate climate impact. We combine elicited misperceptions and willingness to mitigate in a structural model that delivers sharp predictions about where to best target information campaigns. In an experiment with actual consumption decisions (N = 2,081), we then test for the effect of CO2 information on the demand for beef, a product predicted to be a productive target for information. Correcting misperceptions has no effect on the demand for beef, both in absolute terms and compared to a predictably less productive target of information, i.e. the demand for poultry. Our dataset allows us to hone in on the underlying reason for this null effect.
Social Anxiety and Evaluative Interviews, with Samantha Horn and Egon Tripodi, 2025, slides
We study how social anxiety shapes behavior, beliefs, and performance in evaluative interviews. In a controlled online experiment (N = 933), applicants decide whether to participate in a live, five-minute interview with a trained interviewer that determines a monetary hiring bonus. Despite identical objective task ability and indistinguishable interviewer assessments, socially anxious applicants are substantially less willing to interview and hold markedly more pessimistic expectations about task performance, likability, and hiring. These beliefs strongly predict avoidance. Moreover, a non-discriminatory interview experience does not eliminate pessimism. Specifically, socially anxious applicants update more negatively than socially secure individuals, in a way that we show to be inconsistent with Bayesian updating under similar average signals. Applying machine learning techniques to transcripts and video data, we show that belief gaps arise from differential interpretation (updating) rather than differential treatment by interviewers (signal). Consistent with inequities emerging through self-selection rather than discrimination, we find that reducing frictions to showing up is more effective than altering interview environments in reducing mental health inequities.
PUBLICATIONS
Uncertainty about Carbon Impact and the Willingness to Avoid CO2 Emissions, with Taisuke Imai, Davide Pace, and Joël van der Weele
Ecological Economics, 2025, 227, 108401
Using data from a large representative survey, we document that consumers are very uncertain about the emissions associated with various actions, which may affect their willingness to reduce their carbon footprint. We then experimentally test two channels for the behavioral impact of such uncertainty, namely risk aversion about the impact of mitigating actions and the formation of motivated beliefs about this impact. In two novel large online experiments (N=2,219), participants make incentivized trade-offs between personal gain and (uncertain) carbon impact. We find no evidence that uncertainty affects individual climate change mitigation efforts through risk aversion or motivated belief channels. The results suggest that reducing consumer uncertainty through information campaigns is not a policy panacea and that communicating scientific uncertainty around climate impact need not backfire.
Anticipatory Anxiety and Wishful Thinking, with Jan Engelmann, Maël Lebreton, Nahuel Salem, Joël van der Weele
American Economic Review, 2024, 114 (4), p. 926-60.
Across five experiments (N=1,720), we test whether people engage in wishful thinking to alleviate anxiety about adverse future outcomes. Participants perform pattern recognition tasks in which some patterns may result in an electric shock or a monetary loss. Diagnostic of wishful thinking, participants are less likely to correctly identify patterns that are associated with a shock or loss. Wishful thinking is more pronounced under more ambiguous signals and only reduced by higher accuracy incentives when participants' cognitive effort reduces ambiguity. Wishful thinking disappears in the gain domain, indicating that negative emotions are important drivers of the phenomenon.
Misperceived Effectiveness and the Demand for Psychotherapy , with Chris Roth and Egon Tripodi
Journal of Public Economics, 2024 , 240, p. 105254
While psychotherapy has been shown to be effective in treating depression, take-up remains low. In a sample of 1,843 depressed individuals, we document that concerns about effectiveness are top of mind when respondents consider the value of therapy. We then show that the average respondent underestimates the effectiveness of therapy and that an information treatment that corrects this misperception increases participants’ incentivized willingness to pay for therapy. Information affects therapy demand by changing beliefs rather than by shifting attention. Our results suggest that information interventions that target the perceived effectiveness of therapy are a potent tool in combating the ongoing mental health crisis.
Selling dreams: Endogenous optimism in lending markets, with Luc Bridet
American Economic Journal: Micro, 2024, 16(3), p. 492-524
We propose a simple model of borrower optimism in competitive lending markets with asymmetric information. Borrowers in our model engage in self-deception to arrive at a belief that optimally trades off the anticipatory utility benefits and material costs of optimism. Lenders' contract design shapes these benefits and costs. The model yields three key results. First, the borrower's motivated cognition increases her material welfare, regardless of whether or not she ends up being optimisitic in equilibrium. Our model thus helps explain why wishful thinking is not driven out of markets. Second, in line with empirical evidence, a low cost of lending and a booming economy lead to optimism and the widespread collateralization of loans. Third, equilibrium collateral requirements may be inefficiently high.
Cursed Consumers and the Effectiveness of Consumer Protection Policies, with Alessandro Ispano
Journal of Industrial Economics, 2023, 71(3), p. 407-440
We model firms' quality disclosure and pricing in the presence of cursed consumers, who fail to be sufficiently skeptical about undisclosed quality. We show that neither competition nor the presence of sophisticated consumers necessarily protect cursed consumers from being exploited. Exploitation arises if markets are vertically differentiated, if there are few cursed consumers, and if average product quality is high. Three common policy measures aimed at consumer protection, i.e. mandatory disclosure, third party disclosure and consumer education may all increase exploitation and decrease welfare. Even where these policies improve overall welfare, they often lead to a reduction in consumer surplus.
Learning about One's Self, with Yves Le Yaouanq
Journal of the European Economic Association, 2022, 20.5, p. 1791–1828
Pre-analysis plan, Mapping of PAP into WP
To better understand why naiveté about present bias is so prevalent and persistent, we investigate people's (in)ability to learn from their past behavior. Participants in our experiment repeatedly decide how much to work on an unpleasant task and are asked to predict their future effort. We find that participants are naively present biased at first, but update their beliefs once they gain experience with the task. Moreover, our methodology allows us to establish that the amount of updating we observe would eliminate naiveté in the long run. A treatment in which we vary the nature of the task after an initial experience shows that learning is unencumbered by a change in environment. Taken together, our results suggest that persistent naiveté results neither from a fundamental inferential bias nor from an inability to transport newly acquired self-knowledge to new settings. However, participants exhibit another bias: they underestimate their future learning, which may lead to underinvestment in experimentation and a failure to activate self-regulation mechanisms.
Self-persuasion: Evidence from International Debating Competitions, with Egon Tripodi and Joël van der Weele
American Economic Review, 2022, 112.4, p. 1118-46.
Press: Süddeutsche Zeitung
Laboratory evidence shows that when people have to argue for a given position, they persuade themselves about the position’s factual and moral superiority. Such self-persuasion limits the potential of communication to resolve conflict and reduce polarization. We test for this phenomenon in a field setting, at international debating competitions that randomly assign experienced and motivated debaters to argue one side of a topical motion. We find self-persuasion in factual beliefs and confidence in one’s position. Effect sizes are smaller than in the laboratory, but robust to a one-hour exchange of arguments and a ten-fold increase in incentives for accuracy.
Spin Doctors: An Experiment on Vague Disclosure, with Marvin Deversi and Alessandro Ispano
European Economic Review, 2021, 193
Unfavorable news are often delivered under the disguise of vagueness. But are people sufficiently naive to be fooled by such positive spin? We use a theoretical model and a laboratory experiment to study the strategic use of vagueness in a voluntary disclosure game. Theory predicts that, when facing a possibly naive receiver, the sender uses messages that clearly separate her from worse but not from better types. Senders in the experiment adopt this strategy and some (naive) receivers are systematically misled by it. Imposing precise disclosure leads to less, but more easily interpretable, disclosure. Both theory and experimental data further suggest that imposing precision improves overall information transmission and is especially beneficial to naive receivers. Our results have implications for the rules that govern the disclosure of quality-relevant information by firms, the disclosure of research findings by scientists, and testimonies in a court of law.
Gender differences in social interactions, with Guido Friebel, Marie Lalanne, Bernard Richter and Paul Seabright
Journal of Economic Behaviour & Organization, 2021, 183, p. 33-45
We study how the random assignment of new students to introductory-week groups shapes subsequent friendship networks. Both women and men report being much more likely to be friends with same-gender students with whom they were (randomly) assigned in a group during their first week on campus, and the effect is much stronger for women. When students from the same cohort play a repeated trust game in the experimental laboratory, their behavior helps explain what we observed in the field. Women display more stability and less flexibility than men in their interactions with individuals with whom they had previously played. This difference is enough to generate homophily in the observational data even though subjects show no intrinsic preference for same-gender interaction.
Deception and Self-Deception, with Joël van der Weele
Nature Human Behavior, 2019, 3, p. 1055-1061
Supplementary Information, Data, Codes and Materials.
Media: ABC News, Phys.org, MSN, FAZ, Folia, WirtschaftsWoche; Blog posts: Nature Research, FehrAdvice, Psychology Today, Darwinian Business; Podcasts: BAdW.
There is ample evidence that the average person thinks he or she is more skillful, more beautiful, and kinder than others and that such overconfidence may result in substantial personal and social costs. To explain the prevalence of overconfidence, social scientists usually point to its affective benefits, like those stemming from a good self-image or reduced anxiety about an uncertain future. An alternative theory, first advanced by evolutionary biologist Robert Trivers, posits that people self-deceive into higher confidence in order to more effectively persuade or deceive others. We conduct two experiments (combined N=688) to test this strategic self-deception hypothesis. After performing a cognitively challenging task, half of our subjects are informed that they can earn money if, during a short face-to-face interaction, they convince others of their superior performance. We find that the privately elicited beliefs of the group that was informed of the profitable deception opportunity exhibit significantly more overconfidence than the beliefs of the control group. To test whether higher confidence ultimately pays off, we experimentally manipulate subjects' confidence by means of a noisy feedback signal. We find that an exogenous shift in confidence does make subjects more persuasive in subsequent face-to-face interactions. Overconfidence emerges from these results as the product of an adaptive cognitive technology with important social benefits, rather than some deficiency or bias.
Motivated health risk denial and preventative health care investments
Journal of Health Economics, 2019, 65, p. 78-92
People deny health risks, invest too little in disease prevention, and are highly sensitive to the price of preventative health care, especially in developing countries. Moreover, private sector R&D spending on developing-country diseases is almost non-existent. To explain these empirical observations, I propose a model of motivated belief formation, in which an agent's decision to engage in health risk denial balances the psychological benefits of reduced anxiety with the physical cost of underprevention. I use the model to study firms' price-setting behavior and incentive to innovate. I also show that tax-funded prevention subsidies are welfare enhancing.
Cooperating over losses and competing over gains: a social dilemma experiment, with Alessandro Ispano
Games and Economic Behavior, 2017, 105, p. 329 -348
Evidence from studies in international relations, the politics of reform, collective action and price competition suggests that economic agents in social dilemma situations cooperate more to avoid losses than in the pursuit of gains. To test whether the prospect of losses can induce cooperation, we let experimental subjects play the traveler's dilemma in the gain and loss domain. Subjects cooperate substantially more over losses. Furthermore, our results suggest that this treatment effect is best explained by reference-dependent risk preferences and reference-dependent strategic sophistication. We discuss the implications of our results and relate our findings to other experimental games played in the loss domain.
Google Sites
Report abuse