Abstract: In this article, we discuss some aspects of Eliaz and Ok ’s (Games Econ Behav 56:61–86, 2006) choice theoretical foundations of incomplete preferences. Our aim is to clarify some aspects of their regularity condition for preorders, showing that, as far as rationalization of a choice correspondence alone is concerned, no further restriction is imposed by requiring the preorder to be regular. However, if one is also interested in capturing other observable aspects of the individual’s choice procedure, such as Eliaz and Ok’s notion of c-incomparability or the notion of behavioral indifference introduced in this paper, then the only option is to use a regular preorder. We also argue that their weak axiom of revealed non-inferiority (WARNI) is too strong a property if our aim is to characterize the rationalizability of a choice correspondence by a (possibly incomplete) preorder. Finally, as we have mentioned above, we introduce the notion of behavioral indifference and argue, not only that Eliaz and Ok’s notion of c-incomparability (observable incomparability) can be derived from it, but also that it has a wider range of applicability than their notion of observable incomparability. 

Click here for the paper.

Abstract: People often need to plan how to allocate their attention across different tasks. In this paper, we run two experiments to study a stylized version of this attention allocation problem between strategic tasks. More specifically, we present subjects with pairs of 2x2 games, and, for each pair, we give them 10 seconds to decide how they would split a fixed time budget between the two games. Then subjects play both games without time constraints, and we use eye-tracking to estimate the fraction of time they spend on each game. We find that subjects’ planned and actual attention allocation differ and identify the determinants of this mismatch. Further, we argue misallocations can be relevant in games where a player’s strategy choice is sensitive to the time taken to reach a decision.

Last update: 11/15/2022

Click here for the paper.

Abstract: Economists often rely on people’s choices to infer their preferences. However, inferring preferences from choices is problematic when people face unobserved constraints. In this paper, I study how to (cautiously) infer preferences from choices when the choices of the members of a social group are subject to a constraint that we only imperfectly know. When this happens, heterogeneity in the group’s choices and more information about the constraint help recover group members’ preferences, whereas homogeneity in choices hinders it. I further show that a well-known approach for choice-based welfare analysis makes misleading inferences about preferences and efficiency in this context.

Last update: 03/12/2025

Click here for the paper.

Abstract: This paper proposes a novel incomplete ordering to compare the rationality of choices, where rationality is defined as compatibility with preference maximization. I first characterize this rationality ordering, showing that it is the most complete ordering that satisfies two intuitive properties. I then study how it relates to indices of incompatibility, proving that an index of incompatibility agrees with it if and only if it is a monotonic aggregator of a suitably defined notion of predictive mistakes. Using this notion of predictive mistake, I show that the well-known Houtman-Maks index only satisfies a weaker notion of agreement with the ordering. I then propose a method to adjust indices that satisfy this weaker notion of agreement while incorporating our judgments about the relative importance of different violations of rationality. Applying this method to the Houtman-Maks index yields a new family of indices, which I compare to the Houtman-Maks index using an existing choice experiment. Finally, I show that the ordering is meaningfully connected to several models of boundedly rational choice.

Last update: 16/02/2026

Click here for the paper.

Abstract: Will competition between firms in the job market increase efficiency and reduce discrimination? We propose a simple hiring model in a market where the worker's productivity is unknown and the minimum wage is not too low. We then show that, under these conditions, competition leads to a constrained-inefficient and potentially discriminatory outcome because it creates a hold-up problem that reduces the firm's incentives to experiment. Since more competition amplifies this hold-up problem, workers of social groups that face greater informational disadvantages---such as being perceived as less productive or having worse signaling technologies---are more likely to remain unemployed as competition increases, increasing discrimination.

Last update: 29/01/2025

Click here for the paper.

Abstract: There are two ways in which people usually engage with contracts—compensation schemes to execute tasks. They can choose between them (contract choice) or allocate time across them (contract time allocation). In this paper, we study how people behave in each of these problems. A standard model suggests that drafting a cost-effective contract that both induces an agent to choose it and allocate time to it presents a significant challenge. However, our experimental results indicate that this tradeoff might be less pronounced than the model predicts due to what we call the attractiveness bias–a tendency for subjects to allocate more time to contracts they find appealing, even when the model suggests they should get relatively little time.

Last update: 02/04/2025

Click here for the paper.

Abstract: We study how to test whether choices are compatible with maximizing an incomplete preference when we cannot choose the menus from which to observe choices, both theoretically and empirically. Theoretically, we contrast the testable restrictions of the complete and incomplete preference maximization models, showing that once we drop completeness, testing for compatibility becomes computationally hard and may even require an infinite dataset. Empirically, we propose a toolkit for testing compatibility with maximizing an incomplete preference, addressing cases in which the analyst might observe only some of the choices a person would make from a menu. We apply this toolkit to two laboratory experiments to document the fraction of subjects that maximize a transitive, but possibly incomplete, preference.

Last update: 16/02/2026

Click here for the paper.