Behavior seminar

We revisit the Easterlin paradox about the flatness of the happiness trend over the long run, in spite of sustained economic development. With a bounded scale that explicitly refers to “the best possible life for you” and the worst possible life for you, is it even possible to observe a rising trend in self-declared life satisfaction? We consider the possibility of rescaling, i.e. that the interpretation of the scale changes with the context in which respondents are placed. We propose a simple model of rescaling and reconstruct an index of latent happiness on the basis of retrospective reports included in unexploited archival data from the USA. We show that national well-being has substantially increased from the 1950s to the early 2000s, on par with GDP, health, education, and liberal democracy. We validate our new index on several datasets, and find that it captures important changes in personal life circumstances over and above nominal life satisfaction. Our model sheds light on several well-documented happiness puzzles, including why life satisfaction did not drop during the COVID-19 pandemic, why residents of Ukraine report the same life satisfaction today than before the war, why people obstinately think that they will be better off in the future, and why people take life-changing decisions – like having kids – that seem to make them less happy.

Preferences and beliefs about different age groups shape social, political, and economic outcomes. This paper provides strong evidence of “youngism”, which refers to systematic bias in social preferences and unfavorable stereotypes against young adults. Among nationally representative samples from the United States and Czechia, we show that participants in both countries are substantially less pro-social in controlled allocation tasks towards young adults relative to other age groups. This preference bias is widespread, similar in size to discrimination against immigrants, and increases with age. Next, we show that young adults are perceived as more immoral, less helpful, less responsible, less hard-working, and enjoying easier lives than other age groups. Finally, we provide suggestive evidence that these unfavorable stereotypes about young adults feed into the preference bias.

While substantial research has been conducted on how humans infer hidden preferences of others from observing their choices, the effect of their response times (RT) on such inferences has received little attention. We recently investigated this question and showed that humans can infer other people’s social preference, even when observing their RT alone. In parallel, a recent EEG study identified a brain signal which encodes the mismatch between the expected and actual length of elapsed time. Here, we investigate whether this neural marker of time estimation is also relevant when judging the speed of a decision and thus involved in inferring other people’s preferences from RT.

Anti-immigration attitudes remain widespread across Western societies, raising concerns for social cohesion. This study investigates whether correcting misperceptions about immigrants’ perceived deservingness—using cues such as intent to contribute and efforts to overcome socio-economic challenges—can counter anti-immigration prejudice. In Study 1 (N = 474), a factorial survey experiment showed that low-status immigrants were viewed more favorably when they exhibited deservingness traits. Studies 2a (N = 1,506) and 2b (N = 1,255), conducted as one-week longitudinal studies during the 2024 European and French parliamentary elections, revealed that an information treatment emphasizing deservingness cues strongly reduced misperceptions about immigrants, modestly reduced opposition to immigration, and had an even smaller effect for anti-immigration voting. These findings suggest that while immigrants are often perceived negatively, emphasizing perceived deservingness can mitigate prejudice, presenting a promising strategy for reducing anti-immigration bias.

We study attitudes toward offering monetary payments for health behaviors. We develop the Policy Lab, an experimental paradigm to characterize policy preferences in which participants decide whether to implement interventions to influence others’ real-world behavior. In two studies with representative samples of the Swedish population (N=2,010) and one with Swedish policymakers (N=2,008), participants decide whether to provide others (N=1,529) with monetary incentives for vaccination. A majority of participants oppose using monetary incentives. Despite the widespread perception that such incentives are an effective policy instrument, which is supported in our data, opposition to their use is driven by perceptions that they are coercive and unethical. Policymakers exhibit, if anything, greater opposition to the use of monetary incentives. We also document that opposition to incentives extends beyond vaccination to other health domains. Our study provides evidence that the public may oppose polices that they correctly perceive as effective, potentially creating barriers to their adoption. We further introduce a novel method to elicit policy preferences, applicable whenever researchers conduct randomized trials.

This paper investigates the effects of the staggered roll-out of a pollution reduction programme introduced in the UK in the 1950s. The policy allowed local authorities to introduce so-called “Smoke Control Areas” (SCAs) which banned smoke emissions. We start by digitizing historical pollution data to show that the policy led to an immediate reduction in black smoke concentrations. We then merge data on the exact location, boundary and month of introduction of SCAs to individual-level outcomes in older age using individuals’ year-month and location of birth. We show that exposure to the programme increased individuals’ birth weights as well as height in adulthood. We find no impact on their years of education or fluid intelligence.

In this talk I will present the results and reflect on the implications of a number of studies on the construct of risk preference in which I have been involved over the years. In one way or another, they have addressed the nature and measurement of risk preference, one of the building blocks of economic theories of human choice. We will touch on questions of how to (properly) measure risk preferences, how stable they are, how they compare between humans and primates, and what shapes them.

We combine theory and a survey-experimental study in Austria, Germany, and the UK to investigate how sorting fact from fiction and updating from news are influenced by cognitive ability, motivated reasoning, and overconfidence in complex topics, such as climate change and science. We predict and find that cognitive ability (i.e., both IQ and education) improves news discernment. The positive effect of cognitive ability is robust and immune to motivated reasoning. In particular, the ability to give correct answers that counter one’s existing issue opinions and biases increases in IQ and education. These novel results are good news, suggesting the malleability of news discernment. However, when we disaggregate data by news topic, we find that higher cognitive ability may sometimes boost motivated decision making. Our findings suggest that institutions matter. For example, trust in institutions reduces the magnitude of motivated reasoning, which likely helps limit opinion polarization in the longer term.

La protection sociale, historiquement très soutenue en France, est sous pression : les besoins sont plus dynamiques que les recettes. Une augmentation des prélèvements sociaux est-elle politiquement soutenable ? A partir du baromètre de la Drees, qui interroge chaque année 4000 individus sur la protection sociale en France, nous analysons l’impact des réformes du financement de la protection sociale sur l’opinion publique sur la période 2000-2020. Deux types de chocs sont étudiés : l’évolution des taux de cotisations sociales (chômage, retraite) et le choc sur la CSG, cet impôt étant exclusivement affecté à la protection sociale. Nous tirons parti du fait que certains groupes de travailleurs et/ou de retraités ne sont pas affectés par ces hausses de taux pour mener une analyse en double différence. Nous cherchons ainsi à déterminer si, après la réforme, l’opinion des groupes affectés par les changements de prélèvements connaît une évolution significativement différente de celle des groupes non affectés (alors que ceux-ci suivaient une tendance commune avant la réforme). Nous n’observons pas de sentiment de « révolte des cotisants » : une augmentation modérée ou graduelle de prélèvements affectés, qui font partie des structures habituelles de financement de la protection sociale, n’altère pas l’adhésion des populations concernées. En revanche, une réforme atypique du financement de la protection sociale, comme la hausse particulièrement rapide de la CSG en 2018 pour compenser la suppression des cotisations salariales chômage, sans contrepartie directe pour les retraités ou les détenteurs d’actifs financiers et de revenus du patrimoine, a déclenché une réaction de moindre soutien au financement de l’assurance chômage dans l’année qui a suivi la réforme.

Under which conditions do humans cooperate with machines? This study explores human and algorithmic cooperation in the infinitely repeated Prisoner’s Dilemma, focusing on a critical threshold for cooperation. Using online experiments, we assess how replacing human opponents with learning algorithms – either human-mimicking or profit-maximizing – affects cooperation rates. Our results show that human cooperation is influenced by opponent type and strategy, with a higher predictive power of the theoretical threshold when interacting with AI maximizing profits.

We examine how social learning and social identity interact to shape how people respond to the observed behavior of others. Across cultures and decision domains, individuals follow their ingroup more than their outgroup, but this tendency depends on subjective social identification. Strong identifiers not only follow their ingroup more but often choose the opposite of whatever the outgroup is doing. While Bayesian social learning is important, ingroup conformity accounts for a large share of ingroup influence. Outgroup differentiation varies with the strength of ingroup norms, which affect what can be learned from outgroups. Parochial learning (biased perception of ingroup expertise) plays a relatively minor role. Our findings explain why the same observed behavior might be imitated by some individuals but counteracted by others, with implications for polarization and information transmission.

The measurement of intelligence should identify and measure an individual’s subjective confidence that a response to a test question is correct. Existing measures do not do that, nor do they use extrinsic financial incentive for truthful responses. We rectify both issues, and show that each matters for the measurement of intelligence, particularly for women. Our results on gender and confidence in the face of risk have wider applications in terms of the measurement of “competitiveness” and financial literacy. Contrary to received literature, women are more intelligent than men, compete when they should in risky settings, and are more literate.

We consider how a physical disability alters patterns of time use. A disability may raise the time cost of all activities; of some—making them differentially less worth doing; or it may make switching activities more costly. The first yields no predictions about time use, but the latter two possibilities both predict that fewer activities will be undertaken, with more time spent on each. These explanations describe our findings based on non-working ATUS 2008-22 respondents ages 70+, 32% of whom self-assess a disability. Data from the Polish Time Use Survey, where disability is medically certified, show similar results; and they demonstrate the same loss of variety over multiple days. Remarkably similar results are found using Italian and Spanish data, which also demonstrate that having a disability leads to much greater feelings of being rushed for time. Using Polish data we find that a husband responds to his own and his wife’s disability with the same reduction in variety; in contrast a wife enjoys less variety if she has a disability but does more different things if her husband is disabled. Overall, a mobility/physical disability leads an otherwise identical person to engage in over 10 percent fewer activities on a typical day. The lost variety and the additional time pressure represent extra costs equivalent, in the U.S., Polish, and Spanish data to as much as $240,000 per year.