How Observed Queue Length and Service Times Drive Queue Behavior in the Lab
45 Pages Posted: 3 Jun 2019 Last revised: 28 Nov 2022
Date Written: October 1, 2022
Abstract
We study decisions to renege from observable queues using controlled laboratory and online experiments. We focus on the role observed service time sequences and queue length play in decisions to renege. Experimental evidence shows that the survival time is longer for queues starting Fast (with shorter than expected service times) and shorter for physically long queues, controlling for total waiting time. Furthermore, providing information on the expected waiting time mitigates this effect of the queue length. Online experiments show that the effect of observing a fast depleting queue at the beginning persists so that people can wait for queues that are physically longer and/or have a longer total wait relative to shorter steady progressing queues. The additional patience is attributed to a relative progress effect, driven by the proportion of the queue length remaining. We use a simulation model to explore the system-level effects of these individual decisions, where different service time sequences arise naturally due to randomness. Simulation analysis shows that individuals' response to service time sequences is reflected on congestion-related metrics at the system level.
Keywords: Behavioral queues: balking and reneging; Laboratory Experiments; Queue length, experienced wait
Suggested Citation: Suggested Citation