The Impact of Forecast Errors in Multistage Production Systems

In their pioneering work on aggregate production planning, Holt, Modigliani, and Simon (1955) and Bowman (1963) demonstrated that forecast errors can significantly increase manufacturing costs. Without having a perfect forecast, total costs increased by 10 percent. More recent work on the cost of forecast errors goes beyond the simplifying assumptions of a single product and single homogeneous work force. Forecast errors are being studied in the more realistic multistage manufacturing setting where parent-component relationships exist and there are capacitated work stations. However, the results to date imply that the impact of forecast errors depends on the manufacturing environment. The purpose of this paper is to find out more about these contingencies in realistic manufacturing settings, thereby adding to the growing literature base on how forecast errors affect manufacturing performance.