Florian Zimmermann

This paper experimentally studies the role of associative memory for belief formation. Real-world information signals are often embedded in memorable contexts. Thus, today’s news, and the contexts they are embedded in, may cue the selective retrieval of similar past news and hence contribute to the widely documented pattern of expectation overreaction. Based on a stylized version of models of associative memory in the literature, we develop a simple and tightly controlled experimental setup in which participants observe sequences of news about the stock market value of hypothetical companies. Here, identical types of news are associated with identical stories and images. In this setup, participants’ expectations strongly overreact to recent news. We successfully verify the model’s predictions about how the magnitude of overreaction should depend on the history of news. For example, once today’s news are associated with the stories and images of previous opposite news, expectations systematically underreact. By exogenously manipulating the scope for imperfect and associative recall in our setup, we further provide direct causal evidence for the role of memory in belief formation and overreaction. Finally, we use our experimental data to estimate the model parameters that govern the strength of imperfect and associative recall over different time horizons.

This study presents descriptive and causal evidence on the role of the social environment in shaping the accuracy of self-assessment. We introduce a novel incentivized measurement tool to measure the accuracy of self-assessment among children and use this tool to show that children from high socioeconomic status (SES) families are more accurate in their self-assessment, compared to children from low SES families. To move beyond correlational evidence, we then exploit the exogenous variation of participation in a mentoring program designed to enrich the social environment of children. We document that the mentoring program has a causal positive effect on the accuracy of children’s self-assessment. Finally, we show that the mentoring program is most effective for children whose parents provide few social and interactive activities for their children.

Recent models assume that the anticipation of future consumption can have direct utility-consequences. This gives rise to preferences over the timing and structure of information. Using a novel and purposefully simple set-up, we study the determinants of preferences for sooner versus later information. Our main results are as follows: We find that the majority of subjects prefer receiving information sooner. This preference, however, is not uniform but depends on context. When the environment allows subjects to not focus attention on (negative) consumption events, later information becomes more attractive. Finally, variations in prior beliefs do not seem to have strong effects on information preferences.

Throughout the Western world, people’s policy preferences are correlated across domains in a strikingly similar fashion. Based on a simple model, we propose that what partly explains the particular internal structure of political ideology is heterogeneity in moral universalism: the extent to which an individual’s altruism and trust remain constant as social distance increases. In representative surveys with 15,000 respondents, we measure universalism using structured choice tasks. In the data, heterogeneity in universalism descriptively explains a substantial share of desired government spending levels for welfare, affirmative action, environmental protection, foreign aid, health care, military, border control, and law enforcement. Moreover, the canonical left-right divide on issues such as the military or redistribution reverses depending on whether participants evaluate more or less universalist versions of these policies. These patterns hold in the United States, Australia, Germany, France, and Sweden, but not outside the West. We confirm the idea of higher universalism among the Western political left by estimating the universalism of U.S. regions using large-scale donation data and linking this measure to local vote shares.

Many applied economic settings involve tradeoffs between in-group members and strangers. To better understand decision-making in these contexts, this paper measures and investigates the economic relevance of heterogeneity in moral universalism: the extent to which people exhibit the same level of altruism and trust towards strangers as towards in-group members. We first introduce a new experimentally-validated survey-based measure of moral universalism that is simple and easily scalable. We then deploy this tool in a large, representative sample of the U.S. population to study heterogeneity and economic relevance. We find that universalism is a relatively stable trait at the individual level. In exploratory analyses, heterogeneity in universalism is significantly related to observables: older people, men, the rich, the rural, and the religious exhibit less universalist preferences and beliefs. Linking variation in universalism to self-reports of economic and social behaviors, we document the following correlations. Universalists donate less money locally but more globally, and are less likely to exhibit home bias in equity and educational investments. In terms of social networks, universalists have fewer friends, spend less time with them, and feel more lonely. These results provide a blueprint for measuring moral universalism in applied settings, and suggest that variation in universalism is relevant for understanding a myriad of economic behaviors.

Social interactions pervade daily life and thereby create an abundance of social experiences. Such personal experiences likely shape what we believe and who we are. In this paper, we ask if and how personal experiences from social interactions determine individuals' inclination to trust others? In a laboratory experiment, we manipulate social experiences prior to measuring participants' willingness to trust others. We contrast this situation with a non-social control condition where we keep all aspects of the prior experiences identical, except that we remove the social dimension. Subjects' willingness to trust is substantially higher after a positive social experience relative to a negative social experience. No such effect is obtained in the non-social control condition. Our results cannot be explained by rational learning as well as income and reference point effects. Delving into the underlying mechanisms, we provide evidence that non-standard belief patterns are an important driver of experience effects.

A key question in the literature on motivated reasoning and self-deception is how motivated beliefs are sustained in the presence of feedback. In this paper, we explore dynamic motivated belief patterns after feedback. We establish that positive feedback has a persistent effect on beliefs. Negative feedback, instead, influences beliefs in the short run, but this effect fades over time. We investigate the mechanisms of this dynamic pattern, and provide evidence for an asymmetry in the recall of feedback. Finally, we establish that, in line with theoretical accounts, incentives for belief accuracy mitigate the role of motivated reasoning.

A key question in the literature on motivated reasoning and self-deception is how motivated beliefs are sustained in the presence of feedback. In this paper, we explore dynamic motivated belief patterns after feedback. We establish that positive feedback has a persistent effect on beliefs. Negative feedback, instead, influences beliefs in the short run, but this effect fades over time. We investigate the mechanisms of this dynamic pattern, and provide evidence for an asymmetry in the recall of feedback. Finally, we establish that, in line with theoretical accounts, incentives for belief accuracy mitigate the role of motivated reasoning.

We conduct a large-scale field experiment in the German labor market to investigate how information provision affects job seekers’ employment prospects and labor market outcomes. Individuals assigned to the treatment group of our experiment received a brochure that informed them about job search strategies and the consequences of unemployment, and motivated them to actively look for new employment. We study the causal impact of the brochure by comparing labor market outcomes of treated and untreated job seekers in administrative data containing comprehensive information on individuals’ employment status and earnings. While our treatment yields overall positive effects, these tend to be concentrated among job seekers who are at risk of being unemployed for an extended period of time. Specifically, the treatment effects in our overall sample are moderately positive but mostly insignificant. At the same time, we do observe pronounced and statistically significant effects for individuals who exhibit an increased risk of long-term unemployment. For this group, the brochure increases employment and earnings in the year after the intervention by roughly 4%. Given the low cost of the intervention, our findings indicate that targeted information provision can be a highly effective policy tool in the labor market.

Beliefs are often found to be sticky and rather immune to new information. In this paper we highlight a specific mechanism that raises resistance to incorporate new information. We provide results from a lab study in the context of an estimation task where subjects need to provide an estimate about an objective state of the world. In this context we provide causal evidence that commitment to a first opinion leads to a neglect of new and challenging information. Investigating the sources of this effect, we show that our findings are well explained by an internal desire to act consistently. We also present a simple model that formalizes how a desire for consistency can produce our pattern of findings.

Theories of expectations-based reference-dependent preferences have provided a critical modeling innovation, incorporating a structured theory of the formation of reference points. An important prediction of these models is a monotone response in behavior to changes in expectations. To test such models we conduct a real-effort experiment manipulating expectations and examining consequences on effort provision. In contrast to the theory, we document substantial non-monotonicities in the effort response to changing expectations. Our results provide some evidence on the limitations of expectations-based reference dependence.

This paper studies the role of consistency as a signaling device. We propose a two-period model that highlights the informativeness of consistency as a signal of skills and allows for the analysis of consequences for behavior. In a simple principal–agent experiment, we test the basic intuition of the model. We show that consistency is indeed associated with skills. Consequently, consistency is valued by others, inducing people to act consistently.

In this paper we analyze how image utility can lead to misreporting of private information in contexts where truthful reports maximize monetary outcomes. In a controlled experiment, subjects go through a series of quiz questions and subsequently report a performance measure. We vary whether reports are made to an audience or not. In an additional feedback treatment, reports are also stated to an audience and afterwards the experimenter publicly verifies whether reports were correct. We find that in the audience treatment, stated reports are significantly higher relative to the private treatment as well as the feedback treatment. Our findings suggest that overconfident appearance might be a consequence of social approval seeking.

In this paper we examine individuals’ attitudes toward the timing of information. We test a theoretical prediction that people prefer to get information “clumped together” rather than piecewise. We conduct a controlled lab experiment where subjects participate in a lottery and can choose between different resolutions of uncertainty (clumped or piecewise) and analyze which kind of resolution is preferred. Two additional treatments allow us to get a quantitative measure of subjects’ preferences over different information structures. Our data provide little support for a systematic aversion to piecewise information on the aggregate level. In additional treatment conditions, we demonstrate the robustness of our findings and explore potential explanations.