You are here

Lie for me: An experiment about delegation, efficiency, and morality

29 March 2023
2:00 pm
San Ponziano Complex - Conference Room

In our experimental setting, a Principal can decide to self-report a value in a die-under-the-cup task or delegate the report to an Agent, who, however, has no material interest in the report: a large gain for the Principal is associated with a self-reported even outcome, while a small gain is associated with a self-reported odd outcome. We experimentally manipulate the relative efficiency of the report (technology): the Agent’s lottery associated with the toss of the die either stochastically dominates that of the Principal (efficient technology) or is dominated (inefficient technology). Furthermore, we control for individual social attitudes of the participants via an SVO task.

Within this setting, we address the following main research questions:

  1. Does the Principal delegate the report?
  2. Are delegation choices affected by the relative efficiency of Agent’s technology?
  3. Does an Agent’s social attitude impact the likelihood of misreporting?

We find that Principals do not shift the moral cost of lying to the Agent, in general.

Principals show a high propensity to lie and delegate only when the Agent’s technology is more efficient.

Prosocial Agents and individualist Agents differ in their report’s honesty.

The prosocials are likely to lie when they have access to inefficient technology, and the individualists are likely to lie when they have access to efficient technology. This pattern is compatible with different weights given to relative income and absolute income by the two social types.

Our results shed new light on delegation in tasks involving a moral cost and on the importance of individual social attitudes in such settings. Furthermore, we show that when the Agent’s technology is superior to that of the Principal, delegation leads to honest reports. In contrast, when the Agent’s technology is inferior, principals refrain from delegation and are likely to behave dishonestly.

 

Join at: imt.lu/seminar

relatore: 
Matteo Ploner, University of Trento
Units: 
AXES