We conduct a series of experiments and document a robust behavioral pattern whereby people behave more morally in uncertain environments than degenerate deterministic ones. We show that this pattern is weakened when the moral implication of behavior is diminished or when uncertainty pertains to others rather than oneself. These findings are incompatible with standard models that respect dominance. We propose a mechanism based on the anxiety aspect of uncertain environments whereby people act morally as if their moral behavior can help deliver a better outcome. We further delve into the complexity aspect of uncertainty to arrive at a more comprehensive understanding of these findings.

Andreoni and Sprenger (2012a,b) observe that utility functions are distinct for risk and time preferences, and show that their findings are consistent with a preference for certainty. We revisit this question in an enriched experimental setting in which subjects make intertemporal decisions under different risk conditions. The observed choice behavior supports a separation between risk attitude and intertemporal substitution rather than a preference for certainty. We further show that several models, including Epstein and Zin (1989); Chew and Epstein (1990); and Halevy (2008) exhibit such a separation and can account for the overall experimental findings.

We extend Ellsberg's two-urn paradox and propose three symmetric forms of partial ambiguity by limiting the possible compositions in a deck of 100 red and black cards in three ways. Interval ambiguity involves a symmetric range of 50 − n to 50 + n red cards. Complementarily, disjoint ambiguity arises from two nonintersecting intervals of 0 to n and 100 − n to 100 red cards. Two-point ambiguity involves n or 100 − n red cards. We investigate experimentally attitudes towards partial ambiguity and the corresponding compound lotteries in which the possible compositions are drawn with equal objective probabilities. This yields three key findings: distinct attitudes towards the three forms of partial ambiguity, significant association across attitudes towards partial ambiguity and compound risk, and source preference between two-point ambiguity and two-point compound risk. Our findings help discriminate among models of ambiguity in the literature.

We run an experiment to elicit preferences over state-contingent timed payouts. We analyze the data using a new revealed preference method (building on Nishimura et al., 2017) that can test for consistency with utility functions that increase with a given preorder. We find that correlation neutrality, a property implied by discounted expected utility, is widely violated and there is, instead, strong evidence of intertemporal correlation averse behavior. Our results suggest that utility is not additive across both states and time and that credible models of choice need to allow people to prefer negative correlation in timed payouts.

We examine gender differences in the response to competitive pressure using data from the most competitive entrance exam—China's Gaokao. Compared to male students, females underperform on the competitive and high-stakes Gaokao, relative to their performance on the low-stakes mock examination. Moreover, women's performance suffers more than men's in response to negative performance shocks in an earlier exam on the same day. These effects are more pronounced for subgroups of students where the stakes matter more. Overall, these findings appear to be best explained by women's lower tolerance for pressure and weaker incentives to do well in high-stakes settings.

We examine path dependency in physician decisions in an emergency department setting, and find that physicians’ treatment decisions for the current and previous patients are positively correlated. We show that the positive autocorrelation is higher when the current patient is of greater medical uncertainty or more similar to the previous patient in terms of observed characteristics and when the physician is less experienced or more fatigued. We then show that these patterns are highly consistent with the memory and attention model, whereby the physician’s current decision is anchored to her previous decision. The results from both reduced-form analyses and structural estimations provide further support for the importance of memory and attention in physician decision-making.