Behavior Aware Service Staffing

Empirical studies of service systems have shown that workers exhibit different service rates depending on their assigned workload. In contrast, staffing models typically assume a constant service rate. To address this issue, we model two commonly observed behavioral effects, speedup and slowdown, in a general way that allows us to study their joint impact on service staffing. We fit our behavioral model to a hospital dataset and show that both effects are present across a variety of departments. Given these findings, we incorporate speedup and slowdown behavior into a multiperiod workforce staffing model and show that a workload (defined as the number of jobs assigned to each worker) that maximizes the service rate is typically not optimal. As expected, the effectiveness of the widely practiced single‐ratio workload staffing policy depends on the strength of the speedup and slowdown effects. Interestingly, we find that in the presence of slowdown, weak behavioral effects (where workers work at a relatively constant rate) are the cases where the single‐ratio policy performs the worst; the optimal workload instead varies the most as system demand changes. This result differs from practice in many services such as healthcare, where fixed patient‐to‐nurse ratio workloads are commonly used. We show that the strength of behavioral effects modulates the trade‐off between a steady workload and the number of schedule adjustments.