Abstract: We examine source dependence in the setting of effort provision. Our first experiment elicits preference over uncertain piece rate schemes to perform a real-effort task. Our second experiment elicits effort after receiving an uncertain gift. We vary the probability of winning and the familiarity of natural sources of uncertainty. We show that subjects are averse to unfamiliar sources for moderate or high probability, but less so for low probability. Moreover, effort exhibits more insensitivity to the probability under the unfamiliar source compared with the familiar source. Our findings support the validity and generalizability of source dependence in applied settings.
Abstract: As large language models (LLMs) like GPT become increasingly prevalent, it is essential that we assess their capabilities beyond language processing. This paper examines the economic rationality of GPT by instructing it to make budgetary decisions in four domains: risk, time, social, and food preferences. We measure economic rationality by assessing the consistency of GPT's decisions with utility maximization in classic revealed preference theory. We find that GPT's decisions are largely rational in each domain and demonstrate higher rationality score than those of human subjects in a parallel experiment and in the literature. Moreover, the estimated preference parameters of GPT are slightly different from human subjects and exhibit a lower degree of heterogeneity. We also find that the rationality scores are robust to the degree of randomness and demographic settings such as age and gender, but are sensitive to contexts based on the language frames of the choice situations. These results suggest the potential of LLMs to make good decisions and the need to further understand their capabilities, limitations, and underlying mechanisms.
Abstract: Uncertainty motivates a wide range of human behavior, as individuals are disposed to pursue a feeling of certainty. We propose that uncertainty motivates morality, and test the hypothesis by examining moral decisions when individuals face an uncertain situation compared with all degenerate certain situations. We find that uncertainty drives individuals to lie less in a dice game experiment and share more in a dictator game experiment. As standard models of choice under uncertainty cannot accommodate these observations, we propose a theoretical framework whereby individuals perceive a connection between the morality of their behavior and the outcome resulting from uncertainty. [Working Paper]
Abstract: Social information, commonly proxied as what people do, has been widely used to direct individual behavior. However, evidence on the effect has been mixed. To understand this puzzle, this paper systematically measures individual response to different social information. In Experiment 1, we investigate whether individual donation is contingent on the willingness to donate in the population. Subjects exhibit substantial heterogeneity: 29% (32%) have a positive (negative) relationship, 11% donate regardless of contingencies, and 19% never donate. When third-party punishment is involved, heterogeneous patterns are similarly observed for donation and punishment behavior. Moreover, there are high consistencies among donation without and with punishment as well as belief in punishment. In Experiment 2, for contingent punishment behavior of third parties, we include more contexts with different levels of controversy of the charities. Our findings highlight the richness of responses to social information, and shed light on the usage of social information in promoting social goods. [Working Paper]
Abstract: This paper investigates the impact of socially responsible investment on individuals’ risk-taking behavior and portfolio rebalancing decisions. We find that concerns for social responsibility do not impact stock market participation and willingness to take risk but alters individuals’ preferences among risky assets. Through our experiment, subjects allocate endowments among one risk-free asset and two risky assets. For one risky asset, we vary the characteristics in four conditions. Relative to the control condition, this risky asset yields additional payments for subjects themselves in one treatment, and for charities in the remaining two treatments. Our results show that additional payments for oneself encourage risk taking behavior and trigger rebalancing across different risky assets. However, payments for charities solely induce rebalancing. This finding is consistent across different interest rates and risk levels. As traditional mean-variance analysis has difficulties in explaining these observations, we propose a framework of narrow framing, whereby individuals only consider social responsibility in the account for risky investments. [Working Paper]
Abstract: While recommendation algorithms have been increasingly used in daily life, little has been done to investigate their effect on decision making in terms of decision quality and preferences. Here we examine this question in an experimental setting whereby subjects from a representative US sample are randomly assigned to five conditions and make sets of binary choices between two lotteries. The two control conditions provide either no recommendation or recommendation based on a randomization device. The three treatment conditions provide recommendations developed by algorithms: one is based on the choice of the majority, and the other two use AI-based recommenders including content-based filtering and user-based collaborative filtering. We find that subjects tend to follow recommended choices and are willing to pay a small fee to receive recommendations for their subsequent decisions. Compared to control conditions, recommendation helps subjects make better and faster decisions and behave in accordance with the independence axiom and the betweenness axiom. These results can be explained by some classes of stochastic choice models. Our work adds to the growing literature on the behavioral underpinnings of algorithms including AI and sheds light on the design of choice architecture for decision making under risk. (AEARCT-0009110)
Abstract: Existing studies on risk attitude point out people’s preference for longshots, lotteries with a small probability of winning sizable payoffs. Collaborating with the IT team at a university, we conduct a field experiment to investigate longshot preference in the domain of motivating behavior. We randomly divide above 60,000 subjects into six groups. Among groups, we vary the lottery incentives for motivating participation in a 3-minute online game, which differ in: (1) sources of uncertainty (familiar index, unfamiliar index, randomly draw one out of 100 participants); (2) skewness of probability; (3) number of prizes; (4) vagueness of information (lucky draw vs. win with probability 1%). The comparison of response rates among groups shows that longshots with simpler sources of uncertainty and vague information provide stronger motivation to play the game. (AEARCT-0004790)