Time Is Relative: How Framing of Time Estimation Affects the Accuracy of Cost Information

Abstract

Accurate cost information is critical to effective decision making within organizations. Cost computations often rely on subjective judgments by employees regarding the amount of time that different tasks consume. In an experimental setting, we examine the accuracy of two common approaches to eliciting subjective time estimates vital for accurate cost information. Specifically, we compare estimation error when employees estimate (i) the total time for all iterations of a task (the pool approach) versus (ii) the average time for one iteration of a task (the unit approach). These two approaches have received interest by both practitioners and researchers and are at the heart of the difference between conventional activity-based costing (ABC) and time-driven ABC. While mathematically equivalent, we hypothesize and find that the two approaches evoke different cognitive processes that lead to differences in estimation error. Relative to the unit approach, the pool approach produces larger error in the allocation of time among different tasks, but only when the number of iterations per task varies across tasks. Further, the pool approach results in overestimation of productive time, whereas the unit approach leads to underestimation of productive time. Our findings are robust to different response modes of the pool approach (estimates in absolute time units and in percentages). This study is relevant for designers and users of cost and performance-measurement systems in that allocation errors lead to cost cross-subsidization and poor resource-allocation decisions, while overall errors undermine capacity utilization decisions. This paper was accepted by Brian Bushee, accounting.