18 Apr 2013  Working Papers

The Impact of Pooling on Throughput Time in Discretionary Work Settings: An Empirical Investigation of Emergency Department Length of Stay

Executive Summary — Improving the productivity of their organizations' operating systems is an important objective for managers. Pooling—an operations management technique—has been proposed as a way to improve performance by reducing the negative impact of variability in demand for services. The idea is that pooling enables incoming work to be processed by any one of a bank of servers, which deceases the odds that an incoming unit of work will have to wait. Does pooling have a downside? The authors analyze data from a hospital's emergency department over four years. Findings show that, counter to what queuing theory would predict, pooling may actually increase procesdsing times in discretionary work settings. More specifically, patients have longer lengths of stay when emergency department physicians work in systems with pooled tasks and resources versus dedicated ones. Overall, the study suggests that managers of discretionary work systems should design control mechanisms to mitigate behaviors that benefit the employee to the detriment of customers or the organization. One mechanism is to make the workload constant regardless of work pace, which removes the benefit of slowing down. Key concepts include:

  • This research offers practical insights for workplace managers and health care policymakers.
  • In workplaces where workers have discretionary control, the potential negative effects of designing pooled systems must be carefully considered. This has implications for designing and managing staffing structures and workflows, particularly in the context of service delivery organizations.
  • Managers should consider implementing group incentives rather than individual incentives to motivate workers. This may encourage fast workers to reduce their speed just enough so that they will not negatively affect the productivity of others by over-utilizing shared resources.
  • While workplaces often seek to incentivize workers through pay-for-performance programs that focus on individual productivity, a group-level approach may help counteract the negative effects that fast workers exhibit on overall productivity levels.
  • In health care, emergency departments may benefit from implementing non-pooled work systems in which patients are assigned to a doctor-nurse team immediately upon arrival.


Author Abstract

We conduct an empirical investigation on the impact of pooling tasks and resources on throughput times in a discretionary work setting. We use an Emergency Department's (ED) patient-level data (N = 234,334) from 2007 to 2010 to test our hypotheses. We find that when the ED's work system had pooled tasks and resources, patients' lengths of stay were longer than when the ED converted to having dedicated tasks and resources. More specifically, we find that dedicated systems resulted in a 9% overall decrease in length of stay, which corresponds to a 25-minute reduction in length of stay for an average patient of medium severity in this ED. We propose that the improved performance comes from a reduction in social loafing and a more distributed utilization of shared resources. These benefits outweigh the expected efficiency gains from pooling, which are commonly predicted by queuing theory.

Paper Information