An Agency Model Free from the Utility Theory
47 Pages Posted: 7 Mar 2018 Last revised: 19 Sep 2020
Date Written: June 27, 2014
In most of the agency literature, risk aversion or dis-utility of effort are necessary to form a nontrivial problem. However, the utility theory has been challenged in many studies. This paper examines two alternative sources of moral hazard other than risk or effort aversion. Here, the agent is risk-neutral and does not incur (different levels of) dis-utility from making decisions. He is responsible to allocate a finite resource to multiple tasks or divisions. The resource has a concave effect on the expected output and the performance of each task, reflecting its diminishing return. Consequently, the resource scarcity imposes an ellipsoidal constraint on the expected output levels. The value created and the agent’s payment (contract) both linearly depend on the measurable outputs of the tasks.
The optimal linear contracts for different situations are found using KKT conditions. We see information asymmetry does not necessarily induce moral hazard when there is no other source of tension. After imposing a limit on salary, we see that it is not optimal to pay the agent based on his contributed value even though this value is measurable from the outputs. The solutions to the complete model debunk the common “incentive alignment” paradigm that the performance-based compensation should be proportional to the value the agent creates. In fact, gain-sharing plans are not optimal, however measurable the gain is. Furthermore, maximizing the principal’s net profit is not equivalent to maximizing the value created.
Keywords: Incentive Mechanism, Multi-Tasking, Compensation, Quadratic Constraint
JEL Classification: C72, D82, D86, J33, M12
Suggested Citation: Suggested Citation