Evidence Games: Lying Aversion and Commitment
Decisions in fields such as medicine, finance, and law often rely on hard evidence. Imagine a situation where one party (Person A) has evidence about the quality of their own work, and another party (Person B) sets Person A’s compensation. Person A wants to receive as much compensation as possible regardless of what their evidence shows, while Person B wants the compensation to accurately reflect Person A’s true performance. Person A may choose to disclose all their evidence, some of it, or none of it, but cannot produce evidence they do not actually have. Situations like this are called “evidence games.”
Now suppose Person B can commit in advance to compensation that specifies the reward tied to each possible piece of evidence—including the outcome where no evidence is provided. In “Evidence Games: Lying Aversion and Commitment,” Elif Osun and her co-author examine how commitment affects the outcomes of evidence games. Commitment means that Person B sets the rules for how evidence will be rewarded before they receive any information. Theory suggests that commitment should not change the outcome: Person B ends up giving a reward that cannot distinguish low-value from high-value evidence providers, with or without commitment.
In their paper, however, the authors test these theories and find that commitment does influence outcomes. They show that taking lying aversion—a well-known behavioral tendency—into account can explain why. They illustrate that lying-averse agents may share their negative evidence even if revealing it leads to a lower reward. In turn, Person B can benefit from taking this behavioral tendency into account when designing a reward scheme, if they can commit to it in advance. The authors find that commitment not only increases Person A’s overall reward compared to situations without commitment, but also helps Person B evaluate Person A more accurately.
The full article is available here.
Senior Economist
