ABSTRACT
Accounting conservatism has been described as deriving from preferences for reporting errors to be in the direction of understatement rather than overstatement. We pair reporters with users (who rely on reporters’ information) in a multiperiod experiment. We posit that, under misaligned incentives that motivate aggressive reporting, users view aggressive reports as reflecting exploitative intent and expect that a norm prohibiting aggressive reporting applies. We predict that users use noisy reporting errors to gauge reporters’ norm compliance. We find that, ceteris paribus, users prefer not to be paired with reporters producing overstatement errors likely to reflect aggressive reporting relative to reporters producing understatement errors likely to reflect conservative reporting; alternatively, we find no such asymmetric preferences when the agents’ motives are aligned. The asymmetric preferences cannot be explained by agency theory predictions of payoff maximization or loss aversion. These moral preferences provide an initial condition from which conservatism can endogenously emerge.
Data Availability: Data are available from the authors upon request.
JEL Classifications: B52; D81; D82; M41.