Chunyang Tong, Mahesh Nagarajan, Yuan Cheng

In many service settings, constraints associated with resources used in the service generate a natural trade-off between responsiveness and the service quality experienced by customers. That is, the service provider (SP) may compromise the service quality in order to ease the congestion level of the service system in the presence of resource constraints. This is a trade-off that many public-funded health care service systems face on a day to day basis. In such health care service systems, the service offered to patients usually consists of multiple steps and certain key limited resources are shared among these steps. It is not immediately clear what a system manager would do if a service innovation frees up some time for completing the task in one of the steps. In the presence of limited and shared resources among multiple service steps, the system manager may elect to use this freed up time in one step to spend more time in other steps so that customers would experience a higher service quality, or choose to simply admit more arrivals and possibly compromise the service quality. Indeed, empirical as well as anecdotal evidence shows that the effect of such local innovations on the overall service quality is mixed. 

To study the above issue, this study consider a simple two-step service system. A typical customer experiences both steps one after the other. The author look at systems where the first step provides a basic service and the second step involves a value-added service. Service managers often refer to this second step as a discretionary step. In the context of the settings, the author interpret this second step as one where the service capacity (and therefore the service time) is set carefully and adds significant value to the customer. The author hereafter interchangeably refer to the base step as step 1 and the value-added discretionary step as step 2 of the service processes, respectively. Meanwhile, in order to study the effect of local innovations that decrease the service time of step 1 on the overall service quality and the congestion of the service system under different market settings, a stylized queuing model was built that incorporates service time vs. quality curves and we solve for the optimal service delivery of the system. Then, this study analyzes the optimal service delivery in a short-run decision setting in which the capacity as dictated by the number of servers is inflexible, and expands the model by studying the optimal service delivery in a long-run decision setting in which the capacity becomes flexible. Next, using a model of competition, the author finds that this effect continues to hold in settings where SPs compete for arrivals. Lastly, the robustness of the main findings using a model with tandem queues was explored. 

In the service encounters of many service systems, jobs go through a basic step followed by some level of secondary value-added service. Through the analytical research, the author shows that the service time of the two steps follow a relationship that is determined by numerous factors such as demand being exogenous, short term and long term decisions as well as competition. Thus, the effect of innovations in an upstream step is not apparent. a novel model of competition was provided that uses a structured demand model but sufficiently abstracts from details to provide a characterization of the Nash equilibria. The author shows properties of the equilibria and examine the effects of local innovations that decrease service times under a competitive setting. It is well known queueing duopoly games are notoriously difficult, but we show statics of the equilibria that confirm our monopoly results. Finally, the author believes that there are many economic issues that are not addressed in their study of service quality in service systems. For instance, their analysis is from a social planner’s viewpoint and thus their theoretical predictions present a “should-be” scenario when a treatment technology advance emerges in health care systems. However, many other factors influence the quality of service provided by a health care system. Apparently, the service-quality provision may be directly linked with the incentive mechanism in place. There have been debates over “pay for volume” or “pay for outcome” which are obviously relevant for service delivery. Further work based on the approach proposed in this study, will produce interesting and insightful results.