How Observed Queue Length and Service Times Drive Queue Behavior in the Lab

Join and renege decisions from a single server, observable, first come first served queue are studied via controlled laboratory experiments. The role observed queue length and encountered service time experience plays on these decisions is explored. Reneging behavior by customers influences the performance of many service systems where waiting occurs. An understanding of the drivers of renege behavior is essential to queue design endeavors. Analysis of experiment data shows that, controlling for the total waiting time, the survival time in a queue is affected by the physical queue length as well as experienced service times. Not all subjects are affected by the physical queue length in the same way. Furthermore, this effect is partly mitigated when additional information on waiting time is provided. The experiment design that enables a comparison to a deterministic service time benchmark allows to separate the average pace (speed) set by the expected service time from a relative pace capturing the difference between realized service times and expectations. A fast relative pace is found to have a significant effect on patience which can be reduced by creating a setting that diminishes the motivation for in-queue observation