Social Norms and Cheating: An Application to Credence Goods Markets (Job Market Paper)
In credence goods markets, consumers possess less information about whether the products or services fit their needs than the experts. This information asymmetry can lead to fraudulent practices, such as mistreatment and overcharging. This paper explores how behavioral factors shape market outcomes in credence goods markets by focusing on two psychological preferences of the expert: an intrinsic cost to cheating (Cheating Aversion) and a social-image concern (Perceived Cheating Aversion). I develop a formal social-norm-based framework to define cheating and analyze credence goods games within two institutional settings: (i) Verifiability, and (ii) Liability. The findings indicate that when prices are fixed, integrating both concerns about cheating can reduce unethical behavior and improve efficiency in both institutional contexts. However, when the expert has power to set prices, consumer achieves the highest expected utility at moderate levels of sensitivities; overly high sensitivities can be detrimental under Verifiability. In contrast, under Liability, high sensitivities benefit only the expert and have no impact on consumer utility.
Cheating With Externalities and A Regular Audience (With Martin Dufwenberg)
Reporting private information is common, and dishonest reporting often imposes costs or benefits on others. We study how such externalities influence cheating behavior. In a laboratory experiment based on the die-roll paradigm of Fischbacher and Föllmi-Heusi (2013), we introduce a “regular audience” whose payoff depends on the reporter’s self-report. We systematically vary the externality by assigning the audience a payoff that is negatively, neutrally, or positively related to the report. We also manipulate whether the experimenter observes the private draw. This design addresses three open questions: (1) whether externalities influence dishonesty, (2) whether the experimenter could be regarded as a relevant audience, and (3) enabling a direct comparison of competing theories of social image concerns (in particular,comparing the sailing-to-ceiling equilibrium of Dufwenberg & Dufwenberg (2018) to the cutoff-value reporting patterns predicted by Gneezy, Kajackaite & Sobel (2018) and Khalmetski & Sliwka (2019)).
Guilt Aversion and Social-Image Concern in Cheating Games with Negative Externality
This paper develops a model of dishonesty in self-report settings where misreporting imposes harm on others. I focus on two belief-dependent motives: (i) a social-image cost of being suspected dishonest (perceived cheating aversion), and (ii) guilt from disappointing others relative to their expectations (guilt aversion). The model accounts for the common “partial-lying” pattern when image concerns dominate, and predicts downward misreports in pooling equilibria under sufficiently large externalities. A further insight is that strong guilt aversion may paradoxically foster more dishonesty: by strategically lowering others’ expectations, individuals may cheat more to mitigate anticipated guilt.
Deposit, Hold-up, and Vengeance
In scenarios where contracts are incomplete, one party may exploit the other who has already made a relationship-specific investment, thus leading to inefficiency. I establish a hold-up model with bargaining in which the investor (held-up party) holds the “residual rights of control” of the investment. The inefficiency is severe when the investment cost is quite large. Introducing a non-refundable “deposit” can help curb opportunistic behaviors, and reciprocal motivations further mitigate the holdup problem. If the investor is sensitive enough to negative reciprocity but the holding-up party is not, all equilibrium payoffs are efficient, and the holdup behavior disappears from the equilibria.