On the determinants of robustness of costing system designs to errors
Both academics and practitioners accept that costing systems are hardly error-free. Little is known, however, about which costing system design parameters affect the robustness or sensitivity of costing systems to these errors. Using asimulation method and fully parameterizing costing system designs, we present newresults on the determinants of robustness of costing system designs. We find that
- increasing homogeneity in the distribution of resources over resource costpools and in the distribution of percentages allocated by cost drivers at costpools always improves robustness to errors
- increasing the number of cost driverlinks does not have a monotonic effect on overall accuracy
- increasing homogeneity in the spread of cost drivers over cost pools often improves robustness toerrors, other than in the case where there is measurement error on the cost drivers.
We also find that, under measurement error on the cost drivers, the impact of stage II design parameters is larger than the impact of stage I design parameters. Based on these and other insights, we offer costing system designers and users of costing information guidelines to assess and/or improve the robustness of theircosting systems.