We collect individual participant data from 70 papers that use laboratory experiments to examine individual tax evasion behavior (or “Tax Evasion Games”), in order to use meta-analysis to estimate the impacts of different public policy, experimental design and individual level variables on tax evasion choices. Our results show that standard enforcement variables like audits (including audit rules) and fines perform differently on the extensive and intensive margins. We find that other fiscal variables like a flat tax system, tax rates, and tax amnesties have unambiguous negative impacts on tax compliance, and that specific features of the experimental setting, such as how subjects are directed to report income, or whether taxes are redistributed to the participants or to a real life public good, have significant impacts on tax compliance. Our results also indicate that the demographic characteristics of the subjects (e.g., gender, experimental income, occupation, risk attitude) affect compliance.

Decisions under risk are often embedded in a social context that we usually abstract from when studying decision-making in the laboratory. In contrast to that practice, our experiment investigates whether risk-taking is affected by social comparisons. In particular, we focus on situations where some amount of money has to be allocated to two parties: either the amount can be shared, or a random device allocates the entire amount to one of the parties. We find that the social context of the decision matters strongly: When participants are in a disadvantageous initial position compared to the other party, they select the risky option much more often than in a purely individual decision, identical in all other respects. Overall, we find that individuals are relatively more risk-seeking in the socially unfavorable domain than in isolation, in contrast to the favorable one, where we find no or little change in elicited risk attitudes in comparison to an isolated decision.

We examine how taxes impact charitable giving and how this relationship is affected by the degree of wasteful government spending. In our model, individuals make donations to charities knowing that the government collects a flat-rate tax on income (net of charitable donations) and redistributes part of the tax revenue. The rest of the tax revenue is wasted. The model predicts that a higher tax rate increases charitable donations. Surprisingly, the model shows that a higher degree of waste decreases donations (when the elasticity of marginal utility with respect to consumption is high enough). We test the model’s predictions using a laboratory experiment with actual donations to charities and find that the tax rate has an insignificant effect on giving. The degree of waste, however, has a large, negative and highly significant effect on giving.

We study a setting where individuals prefer to coordinate with others but they differ on their preferred action. Our interest is in understanding the role of link formation with others in shaping behavior. So we consider the situation in which interactions are exogenous and a situation where individuals choose links that determine the interactions. Theory is permissive in both settings: conformity (on either of the actions) and diversity (with different groups choosing their preferred actions) are both sustainable in equilibrium. We conduct an experiment to understand how link formation affects equilibrium selection. Our experiment reveals the powerful effect of linking on equilibrium selection: with an exogenous complete network, subjects choose to conform on the majority’s preferred action. By contrast, with endogenous linking—irrespective of the costs of linking—subjects always opt for diversity of actions.

We provide a novel but intuitive explanation for expected utility violations found in the Allais paradox: individuals are commonly averse to receiving nothing. We call this phenomenon the zero effect. Our laboratory experiments show support for the zero effect. By contrast, the evidence for the certainty effect is weak to nonexistent.

We study self-selection into earning money in an honest or dishonest fashion based on individuals’ attitudes toward truthful reporting. We propose a decision-theoretic framework where individuals’ willingness to pay for honest earnings is determined by their (behavioral) lying costs. Our laboratory experiment identifies lying costs as the decisive factor causing self-selection into honest earning opportunities for individuals with high costs and into cheating opportunities for those prepared to misreport. Our experimental setup allows us to recover individual lying costs and their distribution in the population.

Recently, social science research replicability has received close examination, with discussions revolving around the degree of success in replicating experimental results. We lend insight to the replication discussion by examining the quality of replication studies. We examine how even a seemingly minor protocol deviation in the experimental process (Camerer et al. in Science 351(6280):143–1436, 2016. https://doi.org/10.1126/science.aaf0918), the removal of common information, can lead to a finding of “non-replication” of the results from the original study (Chen and Chen in Am Econ Rev 101(6):2562–2589, 2011). Our analysis of the data from the original study, its replication, and a series of new experiments shows that, with common information, we obtain the original result in Chen and Chen (2011), whereas without common information, we obtain the null result in Camerer et al. (2016). Together, we use our findings to propose a set of procedure recommendations to increase the quality of replications of laboratory experiments in the social sciences.

Bayes’ statistical rule remains the status quo for modeling belief updating in both normative and descriptive models of behavior under uncertainty. Some recent research has questioned the use of Bayes’ rule in descriptive models of behavior, presenting evidence that people overweight ‘good news’ relative to ‘bad news’ when updating ego-relevant beliefs. In this paper, we present experimental evidence testing whether this ‘good-news, bad-news’ effect is present in a financial decision making context (i.e. a domain that is important for understanding much economic decision making). We find no evidence of asymmetric updating in this domain. In contrast, in our experiment, belief updating is close to the Bayesian benchmark on average. However, we show that this average behavior masks substantial heterogeneity in individual updating behavior. We find no evidence in support of a sizeable subgroup of asymmetric updators.

Bayesian optimal experiments that maximize the information gained from collected data are critical to efficiently identify behavioral models. We extend a seminal method for designing Bayesian optimal experiments by introducing two computational improvements that make the procedure tractable: (1) a search algorithm from artificial intelligence that efficiently explores the space of possible design parameters, and (2) a sampling procedure which evaluates each design parameter combination more efficiently. We apply our procedure to a game of imperfect information to evaluate and quantify the computational improvements. We then collect data across five different experimental designs to compare the ability of the optimal experimental design to discriminate among competing behavioral models against the experimental designs chosen by a “wisdom of experts” prediction experiment. We find that data from the experiment suggested by the optimal design approach requires significantly less data to distinguish behavioral models (i.e., test hypotheses) than data from the experiment suggested by experts. Substantively, we find that reinforcement learning best explains human decision-making in the imperfect information game and that behavior is not adequately described by the Bayesian Nash equilibrium. Our procedure is general and computationally efficient and can be applied to dynamically optimize online experiments.

We study the emergence of bubbles in a laboratory experiment with large groups of individuals. The realized price is the aggregation of the forecasts of a group of individuals, with positive expectations feedback through speculative demand. When prices deviate from fundamental value, a random selection of participants receives news about overvaluation. Our findings are: (i) large asset bubbles are robust in large groups, (ii) information contagion through news affects behaviour and may break the coordination on a bubble, (iii) time varying heterogeneity provides an explanation of bubble formation and crashes, and (iv) bubbles are strongly amplified by coordination on trend-extrapolation.

We develop a method to assess population knowledge about any given topic. We define, and rationalize, types of beliefs that form the ‘knowledge spectrum’. Using a sample of over 7000 UK residents, we estimate these beliefs with respect to three topics: an animal-based diet, alcohol consumption and immigration. We construct an information-campaign effectiveness index (ICEI) that predicts the success of an information campaign. Information resistance is greatest for animal-based diets, and the ICEI is highest for immigration. We test the predictive power of our ICEI by simulating information campaigns, which produces supportive evidence. Our method can be used by any government or company that wants to explore the success of an information campaign.

Welfare-enhancing policies such as congestion pricing are argued to improve efficiency in situations with externalities. Unfamiliarity and lack of any personal experience with such policies, however, can hinder their implementation; particularly the ex-ante uncertainties of incidences of gains and losses as well as debates regarding equity concerns and how to recycle revenues often stymies implementation. This paper employs a laboratory experiment with heterogeneous users to investigate the effectiveness and acceptability of a toll in a six-player-two-route congestion game. To measure acceptability and how it is affected by experience with the toll, we conduct referenda before, during, and after subjects experience a congestion problem and a toll. The experiment employs a $2\times 2$ design that varies two treatments: the rate of revenue reallocation and the level of information before the final vote. After an experiential learning phase, congestion pricing is found to curb congestion effectively, and although some subjects do not vote in their monetary self-interest initially, the majority does so after experiencing the congestion pricing policy. Data on worldviews and beliefs are collected and matched to voting behavior to examine the evolution of how experience determines acceptability. Some worldviews and beliefs can predict voting behavior and the timing of when an individual finds a toll (un)acceptable.

The paper surveys the experimental literature on centralized matching markets, covering school choice and college admissions models. In the school choice model, one side of the market (schools) is not strategic, and rules (priorities) guide the acceptance decisions. The model covers applications such as school choice programs, centralized university admissions in many countries, and the centralized assignment of teachers to schools. In the college admissions model, both sides of the market are strategic. It applies to college and university admissions in countries where universities can select students, and centralized labor markets such as the assignment of doctors to hospitals. The survey discusses, among other things, the comparison of various centralized mechanisms, the optimality of participants’ strategies, learning by applicants and their behavioral biases, as well as the role of communication, information, and advice. The main experimental findings considered in the survey concern truth-telling and strategic manipulations by the agents, as well as the stability and efficiency of the matching outcome.

In a randomized field experiment, we show that a low-cost intervention, weekly email reminders, increases the weekly exercising frequency of gym members by 13%, with the effect being larger for class training (19%) than for free training (11%). We find that nearly all types of individuals benefit from the reminders (e.g., students, nonstudents, women, men, new members, recurring members), and this effect persists in a three-month posttreatment period. Furthermore, the increase in class visits is partly driven by an increase in the number of bookings of gym classes and a slight decrease in the share of bookings that are canceled. Limited attention and habit formation can best explain these results. In contrast to the increase in gym attendance, we do not find any effect on the duration and renewal of membership contracts.

Preferences over social ranks have emerged as potential drivers of weaker than expected support for redistributive interventions among those closest to the bottom of the income distribution. We compare preferences for alterations of the income distribution affecting the decision maker’s social rank, but not their income, and compare them with similar alterations leaving both rank and income unchanged. Our study fails to find evidence of last-place aversion in a replication of Kuziemko et al. (Q J Econ 129(1):105–149, 2014). However, using a modified design that holds ranks fixed across rounds we find support for both a discontinuously greater disutility from occupying the last as opposed to higher ranks, thus affecting only those closest to the bottom of the distribution, and for a general dislike of rank reversals affecting most ranks. We discuss implications for policy design in both public finance and management science.

An experiment is designed to shed light on how deception works. The experiment involves a twenty period sender/receiver game in which period 5 has more weight than other periods. In each period, the informed sender communicates about the realized state, the receiver then reports a belief about the state before being informed whether the sender lied. Throughout the interaction, a receiver is matched with the same sender who is either malevolent with an objective opposed to the receiver or benevolent always telling the truth. The main findings are: (1) in several variants (differing in the weight of the key period and the share of benevolent senders), the deceptive tactic in which malevolent senders tell the truth up to the key period and then lie at the key period is used roughly 25% of the time, (2) the deceptive tactic brings higher expected payoff than other observed strategies, and (3) a majority of receivers do not show cautiousness at the key period when no lie was made before. These observations do not match the predictions of the Sequential Equilibrium and can be organized using the analogy-based sequential equilibrium (ABSE) in which three quarters of subjects reason coarsely.

We investigate whether informal support is sensitive to the extent to which individuals can influence their income risk exposure by opting into risk. In a laboratory experiment with slum dwellers in Nairobi, we measure subjects’ transfers to a worse-off partner under both random assignment, and self-selection into a safe or risky project. Our experimental design allows us to discriminate between different possible explanations for why giving behaviour might change when risk exposure is self-selected. We find that solidary support is independent of the partners’ choice of risk exposure, which contradicts attributions of responsibility for neediness and ex-post choice egalitarianism. Instead, we find that support depends on donors’ risk preferences. Risk-takers seem to feel less obliged to share the profits they earn from their choices compared to subjects who earn equally high profits by pure luck. Our results have important implications for anti-poverty policies that aim at encouraging risky investments.

We provide the first examination of all-pay auctions using continuous-time protocols, allowing subjects to adjust their bid at will, observe payoffs almost instantaneously, and gain more experience through repeated-play than in previous, discrete-time, implementations. Unlike our predecessors—who generally find overbidding—we observe underbidding relative to Nash equilibrium. To test the predictions of evolutionary models, we vary the number of bidders and prizes across treatments. If two bidders compete for a single prize, evolutionary models predict convergence to equilibrium. If three bidders compete for two prizes, evolutionary models predict non-convergent cyclical behavior. Consistent with evolutionary predictions, we observe cyclical behavior in both auctions and greater instability in two-prize auctions. These results suggest that evolutionary models can provide practitioners in the field with additional information about long-run aggregate behavior that is absent from conventional models.

This paper investigates the usefulness of non-choice data, namely response times, as a predictor of threshold behavior in a simple global game experiment. Our results indicate that the signals associated to the highest or second highest response time at the beginning of the experiment are both unbiased estimates of the threshold employed by subjects at the end of the experiment. This predictive ability is lost when we move to the third or higher response times. Moreover, the response time predictions are better than the equilibrium predictions of the game. They are also robust, in the sense that they characterize behavior in an “out-of-treatment” exercise where we use the strategy method to elicit thresholds.