I am an Associate Professor of Economics at UC Davis. My research fields are Economic Theory, Behavioral Economics, and Experimental Economics. I am also affiliated with the Joint Initiative for Latin American Experimental Economics and the Indian Institute of Technology, Kanpur.
If you are a student at Davis and would like to chat in person, you can find me in my office at 1103 SSH Building. Here is my CV.
[1] External and Internal Consistency of Choices made in Convex Time Budgets [go to paper] Experimental Economics 2017 (with Evan Calford, Yoram Halevy and Guidon Fenig)
We evaluate data on choices made from Convex Time Budgets (CTB) in Andreoni and Sprenger (2012a) and Augenblick et al. (2015), two influential studies that proposed and applied this experimental technique. We use the Weak Axiom of Revealed Preference (WARP) to test for external consistency relative to pairwise choice, and demand, wealth and impatience monotonicity to test for internal consistency. We find that choices made by subjects in the original Andreoni and Sprenger (2012a) paper violate WARP frequently; violations of all three internal measures of monotonicity are concentrated in subjects who take advantage of the novel feature of CTB by making interior choices. Wealth monotonicity violations are more prevalent and pronounced than either demand or impatience monotonicity violations. We substantiate the importance of our desiderata of choice consistency in examining effort allocation choices made in Augenblick et al. (2015), where we find considerably more demand monotonicity violations, as well as many classical monotonicity violations which are associated with time-inconsistent behavior. We believe that the frequency and magnitude of WARP and monotonicity violations found in the two studies pose important confounds for interpreting and structurally estimating choice patterns elicited through CTB. We encourage researchers employing CTB in present and future experiments to include consistency tests in their design and pre-estimation analysis.
[2] The Relation between Behavior under Risk and over Time. [go to paper] [WP version] AER-Insights 2020 (with Yoram Halevy and Kota Saito)
The paper establishes a tight relation between non-standard behaviors in the domains of risk and time, by considering a decision maker with non-expected utility preferences who believes that only present consumption is certain while any future consumption is uncertain. We provide the first complete characterizations of the two-way relations between the certainty effect and present biased temporal behavior, and between the common ratio effect and temporal reversals related to the common difference effect.
Present bias is the inclination to prefer a smaller present reward to a larger later reward, but reversing this preference when both rewards are equally delayed. This paper investigates and characterizes the most general class of present-biased temporal preferences. We show that any present-biased preference has a max-min representation, which can be cognitively interpreted as if, the decision maker considers the most conservative present equivalents in the face of uncertainty about future tastes. We also discuss empirical anomalies which temporal models like beta-delta or hyperbolic discounting cannot account for, but the proposed general representation can accommodate.
[4] Motives Behind Cooperation in Finitely Repeated Prisoner's Dilemma [go to paper] [WP version] Games and Economic Behavior 2023
Selfish preferences cannot explain the significant instances of cooperation observed in Finitely Repeated Prisoner's Dilemma experiments. This paper deploys a novel experiment to compare four theories that explain cooperation. The four theories capture the following four non-pecuniary motives: caring about others (Altruism), being conscientious about cooperation (Duty), enjoying the pleasure of collaborative cooperation (Reciprocal Cooperation), and concerns for reciprocal kindness (Sequential Reciprocity). Our experimental design varies the decline-rate of future rewards, under which these theories make contrasting predictions. We find that our data is best explained by Reciprocal Cooperation-type subjects.
[5] Belief Elicitation in Political Protest Experiments: When the Mode Does Not Teach Us About Incentives to Protest [go to paper] [WP version] Journal of Economic Behavior and Organization 2023 (with Nathan Canen)
Many recent experiments studying political protests elicit subjects’ beliefs and actions, before and after an information intervention, to assess the causal role of beliefs on actions. We show that unless beliefs are symmetric and unimodal, using a belief elicitation scheme that is mismatched with the research question may affect the magnitude and even reverse the sign of identified effects. We provide a simple characterization of when such a sign reversal occurs. As an example, we revisit Cantoni et al. (2019)’s influential study of whether political protests are strategic complements or substitutes. We show how their belief elicitation method allows, in theory, a novel and alternative interpretation of their results, which could have been avoided with a different method.
[6] Ensuring Honest Effort in Peer Grading [go to paper] Journal of Artificial Intelligence Research 2024 (with Swaprava Nath and Jatin Jindal)
We study peer-grading with competitive graders who enjoy a higher utility when their peers get lower scores. We propose a new mechanism, PEQA, that incentivizes such graders through a score-assignment rule which aggregates the final score from multiple peer-evaluations, and a grading performance score which rewards performance in the peer-grading exercise. PEQA makes grader-bias irrelevant. Additionally, under PEQA, a peer-grader's utility increases monotonically with the reliability of her grading, irrespective of her competitiveness and how her co-graders act. In a reasonably general class of score assignment rules, PEQA uniquely satisfies this utility-reliability monotonicity. When grading is costly and costs are private information, a modified version of PEQA implements the socially optimal effort-choices in an equilibrium of the peer-evaluation game. Data from our classroom experiments confirm our theoretical assumptions and show that PEQA outperforms the popular median mechanism.
[7] Higher-order beliefs in a Sequential Social Dilemma [go to paper] Journal of Economic Behavior and Organization 2024 (with Evan Calford)
Do experimental subjects have consistent first and higher-order beliefs about other's preferences? How does any inconsistency affect strategic decisions? We introduce a simple four-player sequential social dilemma where actions reveal first and higher-order beliefs. The unique sub-game perfect Nash equilibrium (SPNE) is observed less than 5% of the time, even though our diagnostic treatments show that a majority of our subjects are self-interested, higher-order rational and have accurate first-order beliefs. In our data, strategic play vastly deviates from Nash predictions because first-order and higher-order beliefs are inconsistent for most subjects.
[8] The value of and demand for diverse news sources [go to paper] [ANU media coverage] Games and Economic Behavior 2025 (with Evan Calford)
We study the value of and the demand for instrumentally-valuable information in an environment where signals are transparently biased. We observe remarkable sophistication in information aggregation and acquisition. A majority of our subjects (63%) made unbiased reports even when faced with biased signals. When allowed to buy pairs of opposite or similarly biased information sources, subjects actively shopped for diverse information at personal costs, and their demand for diverse information reacted rationally to its value and cost. Subjects who were worse at aggregating information, were more likely to purchase diverse signals. Our results advocate for greater transparency in media bias.
[9] From Me to We: Beating Procrastination in Teams [go to paper] Journal of Economic Behavior and Organization 2025 (with Guidon Fenig)
Can team incentives increase worker's productivity and decrease procrastination in intertemporal tasks? We recruited just under 600 online workers to engage in tedious tasks over three days. They were randomly assigned to either individualistic (Solo) incentives or to one of two team-based incentives (Cooperative and Competitive). Contrary to theoretical predictions, workers under Cooperative incentives surpassed the performance of those working under either Solo or Competitive incentives. Productivity on Day 1, which in theory should inversely relate to procrastination, was also significantly higher in both team treatments. Our structural analysis confirms that teams increase productivity by enhancing intrinsic motivation and by reducing the tendency to delay work. Finally, teams increase productivity further under Competitive incentives, when workers can observe and react to the efforts of their team members.
[10] Noisy Foresight (with Chad Kendall) [go to paper] Journal of the European Economic Association 2025
Rational agents must perform backwards induction by thinking contingently about future states and actions, but failures of backwards induction and contingent reasoning are ubiquitous. How do boundedly-rational agents make decisions when they fail to correctly forecast actions in the future? We construct an individual decision-making experiment to collect a rich dataset in which subjects must reason only about their own future actions. We demonstrate substantial mistakes relative to the rational benchmark, and use the rich dataset to estimate several possible models of boundedly-rational foresight. We find that a model in which subjects expect to make more mistakes when the payoff consequences of their future actions are more similar best explains behavior.
[11] Estimating Present Bias and Sophistication over Effort and Money Experimental Economics 2026 (with Claudia Cerrone, Hyok Jung Kim, Leonhard Lades) [go to paper]
We use a real-effort experiment to jointly estimate present bias and sophistication parameters, separately over money and effort. Our novel incentive structure aligns the choice scenario with the canonical assumption of choices being the interior optima of a concave utility-maximization exercise. Participants choose to (and predict to) complete 14% (and 10%) fewer tasks on the same day than on a future day, leading to an estimated βe between 0.70 and .79 (and ̂βe between 0.80 and .88). We find no evidence of present bias or sophistication over money.
[12] Uniquely Justifiable Strategies and Mechanism Design with Boundedly Rational Players (with Chad Kendall) [go to paper]
A strategy is uniquely justifiable (UJ) if, at each decision point, the prescribed action is the unique best response to any contingency the player may consider. We define UJS mechanisms as mechanisms in which the dominant strategy is also uniquely justifiable, so that boundedly rational players do not have to perform contingent reasoning. By contrast, the leading alternative, Obviously Strategy Proof (OSP) mechanisms, require players to contingently reason about their own future actions. The presence or absence of uniquely justifiable strategies neatly organizes existing evidence of adherence — or non-adherence — to (obviously) dominant strategies across mechanisms, from binary allocation problems to matching problems. We characterize the set of UJS mechanisms for binary allocation problems to show that they resemble a generalized Multiple Price List. To compare OSP and UJS mechanisms in a setting that avoids confounds, we conduct an experiment using novel elicitation mechanisms of each type. We identify a tradeoff: OSP mechanisms fail because players fail to forecast their future actions, while UJS mechanisms allow other forms of boundedly rational mistakes.
In prosocial decisions, decision-makers are inherently uncertain about how their decisions impact others’ utility – we call this interpersonal uncertainty. We show that people's response to interpersonal uncertainty shapes well-known patterns of prosocial behavior. First, using standard social allocation decisions, we replicate the classic patterns of ingroup favoritism, merit-based fairness ideals, and self-favoring behavior in dictator games. We then show that these patterns also arise in non-social decisions which have no consequences for others and instead solely reflect responses to interpersonal uncertainty. Behavior across social and non-social decisions is highly correlated, and self-reported interpersonal uncertainty predicts behavior in both situations. Moreover, exogenously varying interpersonal uncertainty shifts prosocial behavior in the direction that avoids such uncertainty. Our results quantify how beliefs in the form of interpersonal uncertainty influence prosocial behavior, which we estimate to be of similar importance to social preferences.
Leading measures of democracy rely on evaluations from small panels of experts. This measurement strategy introduces two features that can limit the statistical value of aggregation: expert judgments may be highly correlated, and expert panels are often small. We model democracy as a latent, unobservable state and propose a new ap- proach rooted in the logic of the “wisdom of crowds” to address these limitations. We show that including a broader and more diverse set of observers – including non-experts – can decrease errors in measuring democracy. Our theoretical results quantify when such larger, decorrelated samples can outweigh the costs of noisier individual judgments, and thus provide more accurate estimates of the true state. We complement the theory with extensive simulations, discuss the empirical implications of this approach and develop an algorithm for practical implementation.
[15] Validated RIS: A within subject test of incentive compatibility in the random (lottery) incentive scheme (with Evan Calford) [go to paper]
We propose the Validated Random Incentive Scheme (VRIS), an enhancement of the Random Incentive Scheme (RIS) that enables internal validation of incentive compatibility. By observing each subject under both RIS and single-choice incentives, VRIS expands the information recoverable from a single design. In an online experiment with three risky decisions and a perception task, we find that RIS distorts marginal choice distributions in two decisions and joint distributions in five of six pairs. VRIS instead accurately measures marginal behavior, enables tests of joint incentive compatibility, and recovers undistorted joint distributions even when RIS distorts individual decisions.
[16] Learning About Outgroups: The Impact of Broad Versus Deep Interactions (with Arkadev Ghosh, Matt Lowe, and Gareth Nellis) [go to paper]
We hypothesize that broad contact, involving brief interactions with multiple outgroup members, and deep contact, meaning longer interactions with a single outgroup member, play distinct roles in shaping intergroup relations. We set up a factory in India and recruited Hindu and Muslim men to work in pairs on joint production tasks. We randomly assigned participants to work either with the same ingroup or outgroup partner daily (deep contact), a different outgroup partner each day (broad contact), or to a control group. While deep contact strengthens social and economic ties with the outgroup partner interacted with, only broad contact reduces misperceptions about outgroup strangers. These findings align with a model in which independent sampling (observing multiple outgroup members) promotes learning about outgroups more than prolonged interaction with a single individual does. Nevertheless, neither type of contact changes behavior toward the wider outgroup.
Many group identities that influence economic behavior are imperfectly observed. Individuals and institutions often conceal identity markers to limit discrimination. Yet concealment also creates uncertainty about group membership, hampering coordination in social interaction. To study this tradeoff, we paired high- and low-caste men for collaborative data entry work in North India. We randomly assigned each mixed-caste pair to either be: (i) introduced by full names, making caste common knowledge; (ii) introduced by first names only, making caste disclosure a choice; or (iii) instructed not to disclose caste. The two concealment conditions substantially reduce the accuracy of beliefs about a partner’s caste and confidence in those beliefs. They also weaken workplace rela- tions, lowering trust, willingness to interact, and perceived productivity—consistent with identity helping struc- ture social coordination. Evidence on mechanisms shows that identity concealment inhibits authentic interaction, making workers less able to express their “true selves,” while certainty about a partner’s identity is associated with stronger workplace ties. Concealment leaves a sizable caste disparity in higher-status role assignment intact, implying minimal impacts on discrimination. We conclude that where group identities are socially entrenched, reducing their legibility may undermine intergroup relations.
[i] Arguing about Tastes: Modeling How Context and Experience Change Economic Preferences by David M. Kreps, Journal of Economic Literature [go to article]