Incentive Mechanism Design Problem Based on Gradient Dynamics

Shi Kuiran,Xiao Tiaojun,Sheng Zhaohan
DOI: https://doi.org/10.1108/03684920910944191
IF: 2.352
2009-01-01
Kybernetes
Abstract:PurposeThe purpose of this paper is to study the incentive mechanism design problem of manager under the assumption of bounded rationality in the theoretical framework of principal‐agent relationship and provide some insights into the relationship between the economic agent's rationality and incentive mechanism.Design/methodology/approachThe framework of principal‐agent relationship is based on the full rationality of the agents. However, the agents cannot make optimal decisions in the uncertain and imprecise environment. To capture the essence of incentive scheme, bounded rationality is introduced in the model and the gradient dynamics is applied in modelling bounded rational learning behavior. The corresponding single‐task principal‐agent problem is solved to obtain the second‐best contracts and the effort levels. Furthermore, the results are generalized for the multi‐task principal‐agent problem.FindingsThe second‐best contract under the assumption of bounded rationality converges to the optimal contract under the assumption of full rationality over time.Research limitations/implicationsAccessibility of data for empirical research is the main limitation which model will be applied.Practical implicationsA useful reference for managerial decisions.Originality/valueThe new approach of bounded rationality modeling and simulation. This paper is aimed at the owners, enterprisers, managers, and workers.
What problem does this paper attempt to address?