Evaluating Heuristics Used When Designing Product Costing Systems

Please do not cite without permission We appreciate the many helpful comments from Design choices regarding the features of a product costing system significantly influence the accuracy of reported costs used in product-and capacity planning decisions. Using simulations, we examine how an amalgam of system design choices influences the error in reported costs. We focus on popular heuristics / rules for grouping resources into cost pools and on methods for determining the driver for cost pools that contain many resources. We also model alternate ways to translate the vague guidance in the literature to implementable rules. We find that correlation-based methods (that follow the prescription to combine " like " resources) are superior to size-based methods (such as the Willie Sutton rule) when selecting the resources combined to form an activity cost pool. Taking information requirements into account, we provide guidance on how such correlation-based rules might be implemented. In addition, we find that combining information from several resources in an index to use as the cost driver achieves economically significant gains in accuracy relative to using a driver based on the largest resource and might beat a driver that considers all resources in an activity pool. By varying properties of the underlying production environment, we also speak to the generalizability of our findings.