Experienced and Prospective Wait in Queues: A Behavioral Investigation

Problem definition: Rational analysis predicts that the cost of queuing depends on the remaining wait time, but research suggests that other parameters such as queue length, service speed, experienced length, and elapsed wait also matter. However, it is challenging to isolate the distinct contribution of these parameters on a monetarily translatable metric because of uncertainty, missing controls in analyses that may lead to omitted variable bias, and differences in outcome variables across existing studies. Methodology/results: We introduce the use of the Becker-DeGroot-Marschak (BDM) mechanism to elicit queue-completion cost. This incentive-compatible, continuous metric is used in a deterministic, full-information setting that eliminates possible indirect effects to isolate the direct contribution of the above parameters. Contrary to previous literature, we find that completion cost is an additive function of prospective queue length and service time, and not of their interaction—remaining wait time—or experienced queue features. We also show how our data, given alternative specifications that do not control for certain variables, would lead to previous results. Managerial implications: Organizations regularly offer customers paid options to shorten their wait time, present them with alternatives to joining a line, or use incentives to manage congestion. Our study—with 1,163 unique subjects across 31 queue variations—provides them with a granular and monetary metric of completion costs, as well as an understanding of their dependence on queuing factors. In the case of fully informed consumers in visible, deterministic queues, we find that compensation can be determined based on what is ahead instead of what has already occurred. The fact that completion costs are not proportional to waiting time also has implications for the design of queuing systems; for example, people may react more negatively to a fast-long queue than predicted by rational theory. Funding: This work was supported in part by the U.S. National Science Foundation [Grant CNS-1739413], the Office of the Dean at the University of Pittsburgh School of Business, and the Fundamental Research Funds for the Central Universities [Grant FRF-TP-24-029A]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2022.0352 .