Optional Budget Mechanisms with Verifiable Cost Signals: An Experiment
Loading...
Date
2009-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The Ohio State University
Abstract
Past accounting research has indicated that budgeting systems are inefficient and fail to account for participant's non-pecuniary motivations in their decision making process. This study uses an experiment to investigate the effects of a verifiable cost signal submitted by the agent on a budgeting agency model in order to mitigate a management control problem induced by asymmetric information. The agent's cost signal is constrained to be truthful and can vary in accuracy in relation to a projects actual cost. The agent has private knowledge of cost, incentive to overstate it, and the ability to do undetected by the owner. This experiment consists of three treatments: (1) the agent is given the option of both choosing the signal and choosing the accuracy of the signal, (2) the agent is forced to submit a signal that is a truthful but coarse indication of actual cost, (3) the agent is forced to submit a signal that is a truthful but fine indication of actual cost. The three treatments are compared with a benchmark treatment from Rankin-Schwartz-Young 2003 where the agent has no cost signal available and can only submit a budget request. In addition, the three treatments will be compared to each other in order to determine the incremental effects of an optional signal and the accuracy of the signals available. The experiment took place in Winter 2009. This study has four main findings: (1) That the existence of a cost signal leads to a higher number of approvals when actual costs are high. (2) That, when given many signaling options, the agent strategically chooses the signal that is appropriate for observed cost in order to get the investment approved. This includes both the use of the fine signal in the high cost case and the use of no signal when costs are low. (3) That forcing an agent to signal provides the principal with enough information to set a maximum amount over the cost signal that the principal sees as unfair. I found that this signal exists at approximately 50 greater than the upper bound of the cost signal. (4) That, when an agent has the option of signaling and chooses not to, the agent also sends a lower budget request then when no signal was possible.
Description
Keywords
Capital Budgeting, Signaling, Experiment, Agency