Research

We study risky choice in a field setting where employees choose among goal-reward contracts resembling financial lotteries and where we observe both choices and beliefs. We find risk aversion and choice heterogeneity far exceeding expected utility predictions and unexplained by prominent behavioral motives like overconfidence, nonlinear decision weights, and loss aversion. We propose and experimentally validate a heuristic explanation for risk taking involving contingency neglect during pairwise evaluation. The heuristic fits the field and lab data better than leading alternative models, uniquely predicts the belief distortions and framing effects we document, and offers a potential explanation for empirical insurance puzzles.

Flood risk disclosure is expanding rapidly in real estate markets, but if disclosures move demand by commanding attention rather than by conveying information, their welfare effects are ambiguous. We examine this tension in a discrete choice experiment where 1,498 respondents make hypothetical choices among real Gulf Coast properties under a control condition and four disclosure formats varying in informational content and presentation. Absent disclosure, insurance costs barely influence choices; by contrast, when respondents are shown precise flood insurance estimates, they are willing to pay $3.34 in annualized housing costs to avoid each annual premium dollar. Using a separate beliefs elicitation, we estimate each format’s technical information content and construct a Bayesian benchmark for how much choices should respond. Observed responses exceed this benchmark by a factor of four to six; informational content accounts for at most 30 percent of the response to any format.

Conspiracy beliefs are the subject of a rich literature in psychology, which explains them through personal dysfunction. The corresponding economics literature is scant. We offer a demand-side model of belief in specific conspiracy theories with no biases, misspecification, or strategic manipulation. A conspiracy, by its nature, generates equivocal evidence about its own existence, because the same strength that produces more to detect also produces better concealment. We model belief in a specific conspiracy theory as learning about the strength q of a cabal, a stand-in for the secretly coordinating actors of any conspiracy theory. In the baseline the cabal influences events with probability q and conceals that influence with probability q, so public evidence arrives at rate q(1−q) and cannot distinguish a weak cabal from a strong, well-concealed one. Bayesian agents converge to two-point beliefs on the pair of observationally equivalent states, with weights given by the ratio of their prior densities there, and their optimal actions diverge accordingly. Divergence requires only that concealment improve elastically with strength over some range, not exact symmetry. It is also invisible to standard belief elicitation: agents with indistinguishable measured priors can be driven to opposite camps by identical public information. Polarization, opposite responses to common information, and stable adherence follow from Bayes’ rule alone. The model provides a complementary account to the one offered by the psychology literature.

Prior research has documented a tendency for insurance demand to rise abruptly in the aftermath of natural disasters. This may be attributable in part to the information shock that a disaster represents. Understanding how beliefs about risk evolve is of first-order importance for projecting climate adaptation in coming decades. Using a three-wave survey of 815 residents of flood-prone coastal counties in the southeastern United States, including 436 for whom we have longitudinal data, we study how individuals’ beliefs evolve after two very different types of informational shocks about future local flooding risk: viewing flood maps, and experiencing a local flooding disaster. We find that, on average, exposure to maps causes respondents to update their beliefs of their homes’ ten-year flood risk downward substantially (38%, 95% CI 13%-56%) and reduces their reported level of worry on a ten-point scale. We then leverage the natural experiment created by differential flood exposure during the hurricane seasons of 2017 and 2018, the latter of which fell between the initial wave of our survey and two follow-up waves. We find that flood disasters systematically increase flood concern in affected counties, as expected. More surprisingly, we find that those who experienced a flood before Wave 1 saw greater increases in concern between waves than those who did not – a pattern inconsistent with standard Bayesian updating and more consistent with the conservatism hypothesis, whereby experiential information shocks are incorporated into beliefs only gradually. This delayed updating process has implications for projections of climate adaptation among flood-prone populations.

From heat waves to hurricanes, tangible weather experiences strengthen personal belief in climate change. We ask whether a high temperature day that breaks local heat records — a mathematical construct not directly accessible to the senses — has additional impacts on perceptions of worsening heat, beyond the absolute temperatures themselves. Matching historical heat records to survey data from the United States, we find that each record heat day in a county in 2022 increases perceptions that excessive heat is getting worse, controlling for average temperatures, the number of extreme heat days, and demographics. Exposure to sixteen record heat days predicts roughly the same difference in heat perceptions as that between the average Democratic respondent and a political independent.