Session 18: Psychology and Economics

We conduct a comprehensive experiment on decision-making for binary lotteries—specifically, lotteries that yield a positive amount or zero. Such lotteries form the basis of many empirical results revealing violations of expected utility (EU) and motivating behavioral alternatives. However, the space of binary lotteries has not been comprehensively explored, and thus the predictions of various behavioral alternatives to EU have not been fully assessed even within this limited domain. We provide this exploration, and discover empirical patterns that stand in stark contrast to the predictions of existing behavioral models. In particular, the data indicate that risk attitudes are driven largely by relative probability comparisons between two options, with absolute magnitudes of probabilities playing a minor role. We show that the data is largely consistent with the model of “upside potential” proposed by McGranaghan et al. (2025).

Aggregate productivity depends on whether productive inputs are allocated to their highest-value uses. We document a novel form of misallocation within a single individual performing a cognitive task. In an online experiment, participants complete a 30-question mathematics exam with random question order and randomly timed enhanced incentives (“bonus boosts”). Performance falls by 9.5 percentage points from the first to the last question, and 82 percent of participants exhibit a decline. Yet 58.0 percent of participants report no preference over when to receive bonus boosts, and willingness to pay for these boosts is constant across timing options. This indifference is costly: assigning boosts early rather than late raises expected bonus earnings by 9 percent. The misallocation aligns with miscalibrated beliefs about productivity: pre-exam forecasts imply constant performance, and post-exam hindcasts capture only 55 percent of the actual decline. These findings identify miscalibrated beliefs about within-worker productivity trajectories as an under-recognized source of misallocation.

Traditional theory assumes firms set prices based on demand elasticity, which they can learn from data. However, sales data reflect both elasticity and intertemporal substitution (IS): consumers shifting purchases toward anticipated discounts. Surveying 15,000 gas station managers, many have a simple model of consumer response to price changes that neglects IS, which they use to interpret data even when it shows stark IS patterns. Manager explanations of the data are consistent with two learning impediments: The simple model directing attention away from data patterns that would challenge the model; noticing data patterns but augmenting the wrong model to explain these. Both are linked to lower cognitive skills. Managers neglecting IS believe fuel demand is more elastic, and charge lower fuel prices. An online experiment provides additional, causal evidence on the mechanisms impeding learning.

We study how firms choose among incentive contracts and how accurately managers predict their e!ects. Using a survey of managerial beliefs and a large-scale field experiment, we randomly assign agents at the market level to several widely used, expenditure-equivalent incentive schemes. In the field, the best-performing contract increases agent output and firm revenue by over 20% relative to the status quo, despite being ranked lower by managers, whereas the worst-performing contract performs as predicted. Managers correctly identify underperforming contracts but systematically underestimate top- performing ones. We document the sources of performance differences—labor supply responses rather than selection or pricing and the determinants of managerial predictability: contract complexity and managerial hierarchy. Our results highlight the importance of contract design for firm performance and reveal systematic limits to managerial cognition in shaping incentives.

Labor supply depends on wages and amenities, and standard models implicitly assume that firms hold accurate beliefs about workers’ amenity valuations. In a survey with firms and workers in Germany, we measure workers’ valuations of amenities and firms’ beliefs about workers’ valuations. We find that firms systematically underestimate workers’ valuations of all amenities. These misperceptions are driven by interpersonal projection: managers project their own preferences—they value amenities less—onto workers. Through the lens of a simple model of imperfect competition, we show that firm misperceptions result in (i) labor shortages and (ii) excess labor costs for biased firms, and increase the market power of unbiased firms. Empirical tests confirm these predictions: a simple calibration suggests that non-providing firms could reduce their labor costs by 5% by providing amenities.

This paper studies when strategic understanding acquired in one mechanism can be transferred to another. We introduce a framework in which agents’ knowledge is represented as a set of payoff comparisons they can make, and use it to formalize what it means to understand that a strategy profile is an equilibrium. We first apply this framework to mechanisms that are strategically equivalent—that is, share the same game form up to relabeling of actions—and show that agents’ understanding of equilibrium transfers across such mechanisms once the relevant action correspondences are explained to them. We then define strategic analogy, a weaker notion that allows not only actions but also types to be remapped, and show that understanding of equilibrium transfers across strategically analogous mechanisms once agents recognize how actions and types correspond. Applications include single item auctions, scoring auctions, and nonlinear pricing with capacity constraints.

Using population-wide Danish administrative registers on housing transactions, I document an asymmetric, hockey-stick relationship between origin market prices and overpay- ment for comparable homes. Quantitatively, the elasticity of overpayment with respect to the origin-destination price difference is 3.9 percent (p < 0.01) when movers relocate from more expensive to cheaper housing markets. In contrast, buyers moving to more expensive locations exhibit little systematic overpayment, and their purchase prices are unrelated to prices at origin. I interpret these patterns through a housing search model in which buyers enter with price beliefs anchored in their origin market and update those beliefs gradually during search. Despite homogeneous learning, endogenous stopping generates the observed asymmetry at purchase: buyers predisposed to overpay transact quickly before fully learning the local price level, while those predisposed to underpay search longer and converge toward local prices. The model yields additional predictions that I test using administrative and survey data. The evidence supports origin-based price extrapolation with subsequent learning rather than preference-based explanations such as reference dependence.

This paper studies persuasive behavior when the sender can control both the information the receiver observes and the model through which it is interpreted (the narrative). Even when the receiver begins with a correctly specified model and understands the sender’s strategic incentives, the sender can manipulate him and often secure her preferred action with probability one. The key mechanism highlights a strong complementarity between strategic communication of information and narratives, allowing the sender to strictly outperform a Bayesian persuader with commitment power. We fully characterize the sender-optimal equilibrium for a broad class of information technologies. Softer information lowers the bar for full manipulation, while harder information expands the set of environments where any manipulation is possible. The results provide a formal foundation for understanding the widespread success of disinformation.

Conveying private information to interested parties is central to many economic and social activities. In such interactions, the sender may lie by misreporting the truth, but may also deceive by inducing inaccurate beliefs about the payoff- relevant state. While a huge experimental literature documents aversion to lying, there is little evidence regarding aversion to deceiving others. Deception aversion is conceptually difficult to document because it depends on unobserved second-order beliefs: the sender’s belief over the receiver’s belief (over the payoff-relevant state). In this paper, we introduce a novel game and show theoretically how to identify deception aversion from choice data alone, with minimal assumptions on second-order beliefs. We run a laboratory experiment and find strong support for deception aversion that is robust to several natural variations of the game. Many subjects lie in order to avoid deception, and structural estimates imply that 30% of subjects are deception-averse.

Abstract. We model an agent who updates her beliefs over a set of variables after observing some of them without fully propagating their implications. We provide a representation of updated beliefs that exhibit limited propagation along a directed acyclic graph, and show that it is implemented by a variant on a standard propagation algorithm. Failures of contingent thinking occur when the agent’s inferences travel through fewer graph paths from hypothetical variables relative to given ones. We characterize the model’s relationship to Bayesian updating and familiar non-Bayesian benchmarks. Contingent thinking is necessary for Bayesian updating, and failures cause correlation neglect and violations of iterated expectations. Our frame- work offers a new perspective into experimental evidence on contingent thinking, reinterpreting effects such as the winner’s curse or the Monty Hall fallacy. We illustrate the framework with applications, ranging from public good contribution games to the recreational puzzle Kakuro.

Theory and empirical research attribute low household stock market participation partly to market frictions and behavioral biases, but substantial cross-sectional variation remains unexplained. This paper examines whether historical race-based trauma shapes current household financial market participation. Our analysis exploits generational exposure to the Freedman’s Savings Bank (FSB), established in 1865 to encourage Black Americans to save. The bank collapsed in 1874 due to fraud and mismanagement. Using restricted-use Panel Study of Income Dynamics (PSID) data, we link present-day stock ownership to historical FSB branch locations. We show that personal, parental, and grandparental exposure to the FSB reduces stock market participation among Black individuals. These effects persist after controlling for socioeconomic and geographic differences, migration, and broader patterns of racial exclusion. Our findings reveal intergenerational transmission of race-based financial trauma and a robust mechanism perpetuating the racial wealth gap.

How harmful is stigma in the “real world”? Answers are elusive because stigma is difficult to measure in observational data, and isolating its effects requires exogenous variation in stigma without variation in the stigmatized trait. This study addresses these challenges by focusing on a widespread form of stigma — weight stigma — in the high-stakes setting of inpatient healthcare. BMI categories are displayed prominently to providers in electronic medical records, and obesity is heavily stigmatized socially. The “obese” cutoff may thus discretely shift stigma while keeping constant the underlying trait. Using a regression discontinuity design that exploits this institutional feature, I find a discontinuous increase in in-hospital mortality at this cutoff, though patient health does not change. Two patterns suggest stigma-based discrimination as the mechanism. First, just-obese patients receive lower diagnostic effort than almost-obese patients. Second, a physician-validated LLM identifies a rise in stigmatizing language in clinical notes at the cutoff — specifically, statements that impose moral judgment, undermine patient credibility, and stereotype patients — that closely tracks mortality effects. Overall, this paper establishes stigma as a powerful social force that can have life-or-death consequences.

Cadaveric organ shortages leave thousands without life-saving transplants each year. Countries differ in using opt-in (informed consent) or opt-out (presumed consent) systems for donor registration. Using newly assembled cross-country panel data and an event-study design, this paper provides evidence that presumed-consent laws increase organ donation only when strictly enforced and family veto power is limited; weak opt-out regimes show negligible or even negative effects. A theoretical signaling model provides a plausible mechanism when opt-in or opt-out yields more donations, emphasizing the roles of donation propensity, signaling costs, and the family’s ability to overturn defaults. A large laboratory experiment further tests these mechanisms, showing that opt-in generally produces equal or higher donation rates unless signaling is costly and family veto power is minimal. The results underscore that defaults alone rarely increase donations unless paired with strong institutional enforcement.

According to standard economic arguments, state dependence generates the value of flexibility. This paper proposes that it can instead generate demand for commitment when individuals anticipate that future states will distort their decisions. A conceptual framework models two broad channels—state-dependent valuations (e.g., projection bias) and state-dependent decision mistakes (e.g., scarcity effects)— and shows that sophistication about such future distortions can generate demand for commitment. We test this prediction in a field experiment in Uganda, where we exclude present bias as a source of commitment demand by design. Farmers are offered pay-at-harvest crop insurance for two seasons and can choose upfront whether to commit to second-season insurance or maintain flexibility. Forty percent of farmers choose commitment. An intervention increasing sophistication raises commitment by 11 percentage points. Additional evidence suggests that both channels matter with substantial heterogeneity across farmers. Our results highlight the importance of individuals’ sophistication about future state dependence for welfare analysis and policy design, particularly in environments with high state variability.

Why do partisans prefer like-minded information sources? They may want to learn the truth and believe these sources are the most accurate. Or, they may prefer them for non-accuracy psychological forces such as confirmation bias. We evaluate these motives in two large-scale experiments in which 3,785 participants choose sources to help them predict swing-state winners in the 2024 US presidential election. Partisans exhibit substantial selective exposure, choosing like-minded sources both among real news outlets and among synthetic sources we construct. This behavior remains essentially unchanged under two treatments: (i) increasing incentives for accuracy and (ii) shutting down confirmation motives by having participants delegate their predictions to sources without seeing sources' content. In contrast, participants respond strongly to experimentally-varied source accuracy, even absent incentives. Our results, interpreted in reduced form and through a discrete-choice model, suggest the selective exposure in our experiment can be almost entirely explained by demand for accuracy.

We provide the first systematic quantification of how decision-making frictions—such as failing to claim government benefits, choosing dominated insurance plans, not saving for retirement, and not quitting smoking—aggregate to affect inequality in income, consumption, and wealth. We review the existing literature and combine it with original analysis of survey data to estimate the prevalence and financial impact of 18 frictions across the income distribution. To make these frictions comparable, we develop a framework in which each friction is characterized by three parameters: the share of the population at risk, the share affected by the friction, and the average loss conditional on being affected. Aggregating across the frictions with dollar-loss estimates, the estimated impact on annual income is 7.8% for the bottom quartile of the income distribution relative to 4.2% for the top quartile; the total loss for low-income households is approximately 7.5 times larger than the impact of a major EITC expansion. We then incorporate these frictions into a life cycle model with realistic institutional features, including tax-advantaged retirement accounts, progressive taxation, portfolio choice, and a social insurance system. The model reveals that removing frictions tends to reduce inequality in lifetime consumption, with the largest effects coming from smoking and attending for-profit colleges. Our results suggest that decision-making frictions are a quantitatively important contributor to inequality in income, consumption, and wealth.