Many important decisions, for example, applying to college, require an individual to simultaneously submit several applications. These decisions are unique as each application is risky because acceptance is uncertain while also being rival as one can only attend a single college. In an influential theoretical analysis of these problems, Chade and Smith (2006) establish the No Safety Schools Theorem which suggests larger portfolios are riskier than single choices. We offer experimental evidence, using several experiments, that this theorem is routinely violated. In fact, the majority of our subjects violate this theorem. However, performance improves with practice, advice, and feedback.

This paper discusses the importance of incorporating personal assistance into interventions aimed at improving long-term education and labor market success. While existing research demonstrates the cost-effectiveness of low-touch behavioral nudges, this paper argues that the dynamic nature of human capital accumulation requires sustained habits over time. To foster better habits, social connections are critical for encouraging enduring effort and intrinsic motivation. The paper links the role of personal assistance to economic theories of human capital investment and decision-making, and showcases examples from various stages of skill accumulation, including early childhood, adolescence, and adulthood, in which interventions that incorporate personal assistance substantially out-perform less intensive nudges. We underscore the importance of interactive support, guidance, and motivation in facilitating significant progress and explore the challenges associated with implementing cost-effective policies to provide such assistance.

I study the optimal design of monetary incentives in experiments where incentives are a treatment variable. I propose a novel framework called the Budget Minimization problem in which a researcher chooses the level of incentives that allows her to detect a predicted treatment effect while minimizing her expected budget. The Budget Minimization problem builds upon the power analysis and structural modeling. It extends the standard optimal design approach by explicitly incorporating the budget as a part of the objective function. I prove theoretically that the problem has an interior solution under fairly mild conditions. To showcase the practical applications of the Budget Minimization problem, I provide examples of its implementation in several well-known experiments. I also offer a practical guide to assist researchers in utilizing the proposed framework. The Budget Minimization problem contributes to the experimental economists’ toolkit for an optimal design, however, it also challenges some conventional design recommendations.

While individuals are expected to perceive similarly identical quantities, regardless of the used units (e.g., 1 ton or 1000 kg), several scholars suggest that consumers over-infer quantities when they are presented in bigger and phonetically longer numbers. In two experimental studies, we examine this numerosity bias in the context of household food waste. Unlike previous scholars, manipulating numerosity revealed no effect: perceptions of food waste volume and likelihood to reduce it are not influenced by the used numeric value (2500 g vs. 2.5 kg; Study 1) nor the number of syllables (two kilos eight hundred seventy-five grams vs. three kilograms; Study 2).

This paper is a single-project meta-analysis of four experiments that model charitable giving as individual contributions to a multiplicity of competing threshold public goods. We pool 17,136 observations at the individual level to summarize the project and investigate the role of learning, gender, and risk attitude, since the included studies are inconclusive in this regard. We find that equally effective coordination devices are the existence of a single contribution option that stands out on its merits, learning, and delegation as long as the intermediary is formally obliged to pass along a high enough percentage of the transferred resources. Women delegate less than men, and consequently prefer direct contributions. Risk tolerance increases overall donations but decreases individual earnings. We discuss possible implications of our findings.

We design an experiment to study the implications of introducing position uncertainty in a social dilemma where eight players decide to contribute to a public good sequentially. Contributions are significantly higher when players make sequential decisions to contribute or not, are uncertain about their position in the sequence, and observe a sample of their predecessors’ choices compared to the simultaneous-move game. Yet, contribution rates remain invariant to the number of agents sampled. Consequently, contributions don’t unravel even with position certainty, and there is no incremental benefit of introducing position uncertainty, contrary to the theoretical prediction. Furthermore, controlling for the sum of contributions observed, individuals contribute less the later in the sequence they are.

Physical inactivity is a leading cause globally of noncommunicable diseases such as diabetes, heart attacks, and strokes. Here, we present the results from a 4-week-long experimental test of a nudge designed to promote physical activity among 206 seniors in Abu Dhabi, United Arab Emirates—a population with one of the highest rates of physical inactivity in the world. We find that the “Forever Fit” nudge—a booklet containing a simple exercise program and information about the health benefits of physical activity—has a large positive effect on 93 previously inactive seniors. The nudge increases the time previously inactive participants spend being physically active from about 5 to about 15 minutes per day.

Legitimizing property rights over the resources that participants use in dictator and ultimatum games has been shown to significantly alter behavior. However, a similar impact has not been observed in public good experiments. We employ an interior public good design with thirty periods of peer punishment, which allows groups to choose between plausible contribution norms without conflicting with efficiency. Across our Unearned and Earned treatments, endowments are randomly allocated or earned through a real effort task. In Unearned, both High and Low types adhere to a norm of contributing an equal proportion of one’s endowment. In contrast, in Earned, only Low types adhere to the proportional contribution norm, while High types contribute less than an equal proportion. Notably, deviations from the proportional contribution by High types are punished significantly less in Earned, suggesting a greater tolerance to such deviations when property rights are earned.

Although compulsory insurance mitigates the negative externalities caused by uninsured individuals, it raises the issue of insurance crowding out prevention. However, at the theoretical level, compulsory insurance and self-insurance (preventive investments dedicated to loss reduction) are know to be substitutes for risk averters but complements for risk lovers. This paper aims to empirically test these opposite predictions through a laboratory experiment using a model-based design. Our experimental results confirm the theoretical predictions: compulsory insurance and self-insurance are complements for risk lovers and substitutes for risk averters. This study strongly supports public policies advocating mandatory insurance implementation as they enhance risk lovers’ self-insurance investments. Therefore, a risk management scheme combining voluntary top-up and compulsory partial insurance guarantees an optimal risk allocation for risk-averters and increases the investments in self-insurance for risk-lovers.

The money-burning game (MBG) is widely used to study anti-social or destructive behavior. We extend the design of the MBG to separate three motives that could lead subjects to burn their partner’s money – spite, reciprocity, and inequality aversion. We detect that reciprocity is the dominant reason: Most of our subjects would only burn their partner’s money if they believed that their partner would burn theirs. This finding has important implications for the interpretation of the behavior of the game.

Based on a representative sample of the French population (n = 1,154), we show that there is a positive association between risk tolerance and trust. We rely on, the World Value Survey WVS binary trust measure, and a ‘0 − 10’ scale that we decline in three domains: trust in general, family, and co-workers. We also vary the measure of risk tolerance, by considering an incentivized investment task, and a ‘0 − 10’ stated preference scale that we decline in three domains: risk tolerance in general, in finance, and health. These variations allow us to test 16 different relations, by crossing four dependent trust variables with four different risk tolerance covariates. After adjusting for multiple testing, we found nine combinations with a strong positive link between risk tolerance and trust in the general population, and that stated risk tolerance measures predict stated trust better than elicited risk measures.

Do large language models (LLMs) – such as ChatGPT-3.5 Turbo, ChatGPT-4.0, and Gemini 1.0 Pro, and DeepSeek-R1 – simulate human behavior in the context of the Prisoner’s Dilemma (PD) game with varying stake sizes? Through a replication of Yamagishi et al. (2016) ‘Study 2,’ we investigate this question, examining LLM responses to different payoff stakes and the influence of stake order on cooperation rates. We find that LLMs do not mirror the inverse relationship between stake size and cooperation found in the study. Rather, some models (DeepSeek-R1 and ChatGPT-4.0) almost wholly defect, while others (ChatGPT-3.5 Turbo and Gemini 1.0 Pro) mirror human behavior only under very specific circumstances. LLMs demonstrate sensitivity to framing and order effects, implying the need for cautious application of LLMs in behavioral research.

We compare different forms of communication in the context of cheap talk sender-receiver games. While previous experiments find evidence supporting the comparative statics prediction that more preference divergence leads to less information transmission, there is also a consistent pattern of overcommunication and exaggeration, not predicted by theory, in which subjects convey more information than predicted in equilibrium. The latter of these findings may be due to the restricted nature of the message space in most experimental cheap talk games, encouraging subjects to engage in exaggeration artificially, rather than allowing it to emerge naturally. We tested this hypothesis with an incentivized lab experiment, and found evidence both phenomena persist with natural language (text-based) communication. Moreover, we probe the consequences of this expanded message space for outcomes, showing that senders benefit more than receivers, but that the most notable effect is that text messages improve efficiency.

We design an experiment to study the effect of asymmetry in the context of group lending with joint liability. The performance of group loan contracts crucially hinges on borrowers engaging in peer monitoring and the common practice is to offer participants of a group loan symmetric contract terms. Our experiment shows that asymmetric contracts, in which monitoring is a dominant strategy for one borrower, increase the monitoring rate, and thus the repayment rate, without leaving borrowers substantially worse off. In addition, asymmetric contracting also raises expected profits of the lending institution. Overall, our experiment reveals that asymmetric group loan contracts are worth considering as part of a policy to maintain both financial stability and higher lender profits.

We study experimentally contests in which players make investment decisions sequentially, and information on prior investments is revealed between stages. Using a between-subject design, we consider all possible sequences in contests of three players and test two major comparative statics of the subgame-perfect Nash equilibrium: The positive effect of the number of stages on aggregate investment and earlier-mover advantage. The former prediction is decidedly rejected, as we observe a reduction in aggregate investment when more sequential information disclosure stages are added to the contest. The evidence on earlier mover advantage is mixed but mostly does not support theory as well. Both predictions rely critically on large preemptive investment by first movers and accommodation by later movers, which does not materialize. Instead, later movers respond aggressively, and reciprocally, to first movers’ investments, while first movers learn to invest less to accommodate those responses.

There is existing evidence that many individuals have preferences regarding selection of numbers in lottery games. Lottery data indicate that the percentage of players who choose their numbers, instead of having numbers randomly assigned, varies widely by lottery game. Differences in number selection mechanisms between games and an expected return maximization motive only present for parimutuel games are both reasons that can explain the variation. Differences in the payoff distributions between lottery games could also be contributing to the observed variation, a novel proposition. An experiment is designed to control for differences in number selection mechanisms and remove the expected return maximization motive, to test for the presence of distribution-dependent number preferences. Results indicate that 40% to 50% of subjects may display such preferences. It is therefore possible that distribution-dependent number preferences contribute to the empirical variation in number selection percentages in lottery games.

College students gain a considerable amount of weight by consuming unhealthy food. Many universities adopt costly programs to alleviate this problem. We study the effect of a simple, inexpensive option: moving unhealthy items out of sight. The opportunity to investigate this intervention comes from the decision of a dining hall in the University of New Hampshire that relocated cookies from a main section in plain sight to an out-of-the way corner. The cost of cookies did not change, since the dining hall operates as an “all that you can eat” restaurant. Relative to pizza, a product that did not change location, the consumption of cookies dropped by up to 22% relative to their predicted level had the relocation not taken place. We see this as evidence that simple changes in design can nudge students towards healthy eating.

Bayesian estimates from experimental data can be influenced by highly diffuse or “uninformative” priors. This paper gives examples of how diffuse priors can affect estimates, and discusses how practitioners can use their expertise to critique and select a prior.1

Economic policy and research rely on the accurate measurement of welfare. In nearly all instances, measuring welfare requires collecting data via long household surveys. If survey response patterns change over the course of a survey to introduce measurement error, this measurement error can be either classical (i.e., changing distributions, leading to noise) or non-classical (i.e., changing expectations, leading to bias). We embed an experiment in a survey by randomly assigning a questionnaire with either the assets module near the beginning of the survey or the assets module at the end of the survey, delaying enumeration of assets by about 60 minutes. We find no evidence in the full sample that survey ordering introduces differential response patterns, either in the number of reported assets or the reported value of those assets. In exploratory analysis of heterogeneity, we find evidence of non-classical measurement error due to survey ordering within sub-samples of respondents who (i) are from larger households or (ii) have low levels of education. Our experimental design can be generalized to serve as an ex-post test of data quality with respect to questionnaire length.

We report the results of an experiment on selective exposure to information. A decision maker interested in learning about an uncertain state of the world can acquire information from one of two sources that have opposite biases: when informed on the state, they report it truthfully; when uninformed, they report their favorite state. A Bayesian decision-maker is better off seeking confirmatory information unless the source biased against the prior is sufficiently more reliable. In line with the theory, subjects are more likely to seek confirmatory information when sources are symmetrically reliable. On the other hand, when sources are asymmetrically reliable, subjects are more likely to consult the more reliable source even when prior beliefs are strongly unbalanced and this source is less informative. Our experiment suggests that base rate neglect and simple heuristics (e.g., listen to the most reliable source) are important drivers of the endogenous acquisition of information.