Evaluating Heuristics Used When Designing Product Costing Systems

Abstract

The academic and practitioner literature justifies firms' use of product costs in product pricing and capacity planning decisions as heuristics to address an otherwise intractable problem. However, product costs are the output of a cost reporting system, which itself is the outcome of heuristic design choices. In particular, because of informational limitations, when designing cost systems firms use simple rules of thumb to group resources into cost pools and to select drivers used to allocate the pooled costs to products. Using simulations, we examine how popular choices in costing system design influence the error in reported costs. Taking information needs into account, we offer alternative ways to translate the vague guidance in the literature to implementable methods. Specifically, we compare size-based rules for forming cost pools with more informationally demanding correlation-based rules and develop a blended method that performs well in terms of accuracy. In addition, our analysis suggests that significant gains can be made from using a composite driver rather than selecting a driver based on the consumption pattern for the largest resource only, especially when combined with correlation-based rules to group resources. We vary properties of the underlying cost structure (such as the skewness in resource costs, the traceability of resources to products, the sharing of resources across products, and the variance in resource consumption patterns) to address the generalizability of our findings and to show when different heuristics might be preferred. This paper was accepted by Stefan Reichelstein, accounting.costing, estimation, activity-based costing, cost drivers, cost pools

A Comparison of Activity-Based Costing and Time-Driven Activity-Based Costing

“…The first effect has been found in many prior numerical experiments (Labro and Vanhoucke 2007;Balakrishnan et al 2011). The parameters at the beginning of the costing system, the resource-to-activity mapping parameters, have much less influence on product costing errors than the parameters from near the end of the costing system, the activity-to-product mapping.…”

“…They find that the hybrid TDABC system results in more accurate cost estimates than the aggregated ABC system. Balakrishnan et al (2011) (hereafter BHL) is the most important prior paper for our work. They incorporate simulations to compare different heuristics used to generate cost pools and select allocation bases.…”

“…We expand the simulation in Balakrishnan et al (2011) Upper bound for correlation between resource consumption patterns We use two levels {0.2, 0.8}. The correlation parameter in our simulation is drawn from the interval {mincor1, maxcor1}.…”

“…As a first step, researchers have focused on how the product costing errors are generated. We follow prior work Vanhoucke 2007, 2008;Balakrishnan, Hansen, and Labro 2011) and examine how the use of heuristically designed product costing systems affects the accuracy of reported product costs, but not how these product costs affect particular decisions such as setting prices.…”

Simulation-based research in management accounting and control: an illustrative overview

“…More fatal are the results presented with respect to pricing errors: the simulation experiments show that prices are 50-70 percent higher compared to "correct" prices. Balakrishnan et al (2011) also focus on the design of costing systems. In particular, they argue that practitioners utilize rules of thumb when grouping resources into cost pools, and when selecting drivers for cost allocation.…”