Journal of Behavioral and Experimental Economics最新文献

We explore whether people trust the accuracy of statements produced by large language models (LLMs) versus those written by humans. While LLMs have showcased impressive capabilities in generating text, concerns have been raised regarding the potential for misinformation, bias, or false responses. In this experiment, participants rate the accuracy of statements under different information conditions. Participants who are not explicitly informed of authorship tend to trust statements they believe are human-written more than those attributed to ChatGPT. However, when informed about authorship, participants show equal skepticism towards both human and AI writers. Informed participants are, overall, more likely to choose costly fact-checking. These outcomes suggest that trust in AI-generated content is context-dependent.

A growing body of evidence suggests poor mental health is associated with sub-optimal economic preferences and behaviours. However, much of this evidence is correlational, and this paper aims to address a call for more research on dynamics. We use rich panel data to evaluate within-individual dynamics in financial management and mental health. We observe increased (decreased) difficulty in managing financial affairs following periods of increased (reduced) psychological distress. Importantly, these dynamics occur in the absence of observable changes to objective financial circumstances (i.e. income and wealth shocks), and are robust with respect to a range of alternative approaches to measurement. Further, we show that most of the association between change in psychological distress and change in financial precariousness (concurrently and subsequently) can be attributed to change in self-efficacy. This supports the idea that investment in mental health will yield important additional benefits via the capacity to empower individuals to make better decisions and escape mental health-based poverty-traps.

Nudges have been shown to influence behavior across various domains, including tax payments. Although nudges that exploit the deterrence motivation are generally considered most effective, evidence of their effectiveness in terms of the degree of deterrent information included is limited. To address these gaps, we conducted a natural mailing experiment to measure compliance with TV fee payments among a sample of households that failed to respond to an initial communication. In addition to studying the strength of deterrence, our study investigates the impact of treatment timing and content simplification on the compliance rate. The results reveal substantial effects of the strongest deterrent information, resulting in a 25 % increase in the payment registration rate. Treatments with a lower deterrence level did not show improvement from the baseline. We also demonstrate that timing matters, as reminders sent closer to the original response deadline (i.e., earlier) motivate 18 % more households to register for payment. The combined effect of the strongest deterrence and simplification leads to a 59 % increase in registrations. This implies that people are more likely to exhibit compliant behavior when the potential consequences of noncompliance are made more salient, and information is conveyed in a timely and simplified manner.

We experimentally examine the effects of advice on decision making in a dictator game in which participants receive no advice, selfish advice, or fair advice before making decisions and the advisors do not benefit from the decision makers’ actions. We find that participants receiving fair advice are more inclined to exhibit fair behavior, whereas those receiving selfish advice are more likely to act selfishly; however, no significant difference in magnitude between the effects of fair and selfish advice is found. We also discover that both types of advice have a more pronounced impact on participants with relatively lower social tendencies, as indicated by the social value orientation test. This study is the first to directly demonstrate the impact of fair and selfish advice in moral dilemmas.

We experimentally investigate whether and how the potential presence of algorithmic trading (AT) in human-only asset markets can influence humans’ price forecasts, trading activities and price dynamics. Two trading strategies commonly employed by high-frequency traders, spoofing (SP) - associated with market manipulation - and market making (MM) - seen as liquidity provision - are considered. These experiments reveal that, first, the mere expectation of SP traders can, at first, impair price convergence towards fundamentals. Second, the expected presence of AT, especially MM traders, induce larger initial price forecasts deviations from fundamentals. Third, despite the absence of AT in our experiments, the information about the presence of AT, employing MM strategy, is sufficient to alter subjects trading behavior over time and the impact of past realized prices on subjects’ order prices.

Ambiguity aversion is the interpretation of the experimental finding (the Ellsberg paradox) that most subjects prefer betting on events whose probabilities are known (objective) to betting on events whose probabilities are unknown (subjective). However in typical experiments these unknown probabilities are known by others. Thus the typical Ellsberg experiment is a situation of asymmetric information. People may try to avoid situations where they are the less informed party, which is normatively appropriate. We find that eliminating asymmetric information in the Ellsberg experiment while leaving ambiguity in place, makes subjects prefer the ambiguous bet over the objective one, reversing the prior results.

Decisions under risk and ambiguity are frequently encountered in our daily lives, and the associated preferences are often quantified. This article deals with the relationship between individuals’ preferences towards risk and ambiguity. In particular, we question the correlation of the preferences and their alignment. For that purpose, we analyze experimental data that measures individuals’ risk and ambiguity preferences through a lottery choice experiment. The results indicate that most of the subjects are risk-loving and ambiguity-averse. We also show that risk preferences and ambiguity preferences are negatively correlated. Finally, we prove that the majority of subjects have preferences towards ambiguity that are aligned with those towards risk, either perfectly or weakly. Thirty-four percent of the sample has preferences towards ambiguity that are the opposite of those towards risk.

Social and moral motivations can influence households’ decisions regarding pro-environmental behavior, such as recycling. In a theoretical framework that allows for these motivations, we analyze how policies such as unit pricing and mandatory recycling affect whether and, in the presence of heterogeneous households, the extent to which a society recycles. We show that unit pricing enhances the effect of intrinsic motivation while mandatory recycling can erode it (depending on the marginal utility of self-image and the recycling cost-to-benefit ratio). We empirically investigate the relationship between policy and intrinsic motivation, using different cutoffs for recyclers and non-recyclers, and find support for our theoretical predictions.

The aim of this study is to test the intergenerational transmission of frequently observed financial biases. We elicited financial biases of first-year university students and their parents. Results indicate the intergenerational transmission of illusion of control, loss-aversion, and overconfidence from parents to children as well as transmission of affinity bias only from parents to daughters. Financial knowledge and saving habits of students have a strong influence on the transmission of financial biases. In addition, we found a moderate effect of economic preferences in the transmission process.

A field quasi-experiment was conducted in cooperation with the local public administration on the waste collection agenda, testing two versions of a standard notice were tested against a control notice targeted at debtors. Both simplified versions, which included a deterrence message, were more effective than the control notice in promoting compliance. However, the version that emphasized the negative consequences of nonpayment, specifically the risk of high interest rates, was more effective in increasing tax compliance than the version that included a descriptive social norm to soften the communication toward fee debtors. The study suggests that adding a social norm to a simplified notice with deterrence does not further increase tax compliance. In addition, the notice had a greater positive effect on fee debtors who had missed only one payment than on those who had missed multiple payments of waste collection fee.