Incentive Effects of Inter-Agent Monitoring: Implications for Responsibility Accounting and Auditor-Consultant Interaction (Centralized, Delegated, Equilibrium).

This research concerns a previously unexplored area in the study of agency theory--the effects of private information on the effort and output of agents. To fully explore this question, three different contracting structures are discussed and contrasted. They are the centralized-monitoring structure, the delegated monitoring structure, and the delegated-contracting structure. The contracting structures differ from one another, primarily, in the way that the principal chooses to reward his agents. In the centralized-monitoring structure, an agent is rewarded solely on the output that he produces. Under the delegated-monitoring structure, rewards are based on a report, made by a pre-selected monitoring agent, that describes the monitored agent's effort level. The monitoring agent is rewarded contingent on the stated effort level in the report. When the reporting or "monitoring" agent reports the monitored agent's effort to be high, the monitoring agent is rewarded based on their combined output. When the monitoring agent reports the monitored agent's effort to be low, the monitoring agent is rewarded based solely on his own output. In either case, the monitored agent is rewarded solely on the reported effort. Finally, under the delegated-contracting structure, the monitoring agent is given full authority over the monitoring agent. He establishes an optimum effort level for the monitored agent and rewards him based on his success at maintaining such a level. The monitoring agent has full responsibility for maintaining his own and the monitoring agent's output level. The monitoring agent's rewards are based on the joint outputs even if the monitored agent supplies a low effort. In each contracting structure, using the Nash solution concept, the principal's expected utility maximization problem is constrained by agents' market reservation and incentive-compatibility conditions. Myerson's revelation principle is employed in the delegated-monitoring structure. On the equilibrium path, the optimal indirect-revelation mechanism suppresses the monitoring agent's coalitional incentive with the monitored agent. Employing the Kuhn-Tucker conditions, each constraint is examined. The expected utility comparison of contracting structures establishes the necessary and sufficient conditions for Pareto-superiority. The findings from this study are that: (1) the delegated-monitoring structure is equivalent to the delegated-contracting structure, (2) under some circumstances the delegated-monitoring structure is infeasible, and (3) by delegating monitoring authority to an agent (i.e. monitoring agent) the principal may increase his welfare when the monitoring agent is relatively less risk-averse, and the monitored agent is relatively more risk-averse.