An empirical field study of statistical and management prediction models and the effects of a budget-based incentive program

Abstract

Previous research has shown that the use of budgets, coupled with significantly valued rewards, to motivate personnel and evaluate performance in managerial accounting settings results in a greater likelihood that managers will report performance at or near the target level. The research reported in this dissertation suggests that organizational and situational factors serve to mitigate the effects of incentive compensation plans on management predictive behavior. The predictive behavior of managers forecasting a variable important to corporate performance and bonus compensation was evaluated before and after such a program was implemented in a Fortune 500 manufacturing company. The predictions were compared to those of two control groups involved in similar activities, but not rewarded by incentive compensation, and two statistical models. Results of the analysis indicated the reward system affected the predictive behavior of the experimental group consistent with previous findings. However, the effects were not consistent throughout the year and appeared to be influenced by organizational considerations which prohibited the group from taking full advantage of bonus payment opportunities. In addition, statistical models were found to be generally better forecasters of the key variable than management models.