While learning by doing has been shown to be the primary driver of productivity, there is ambiguity regarding whether learning in one context can be transferred to another. Individuals working in service settings such as call centers, restaurants and hospitals are subjected to the practice of working in both normal workload and heavy workload shifts. In heavy workload contexts, individuals perform the same tasks as normal workloads but at a faster pace with higher throughput expectations due to the increase in demand. We
use arguments from the organizational learning and task design literature to understand how learning by doing in a normal workload influences employee performance in heavy workloads and vice-versa. Data for our study comes from a medical emergency call center wherein dispatch officers (DO) process call requests and assign appropriate follow-up actions. This includes 15,627 DO-shift observations pertaining to about 360 DOs over a year. We measure individual performance in terms of the proportion of errors made by the DO during that shift. Our analyses suggest three important sets of findings that inform the current literature. First, learning is context-specific, and its transferability across contexts is limited. Second, increasing experience across the different contexts are substitutes, not complements. Third, when switching between the two contexts, we find that experienced employees experience lower switching. Our post-hoc analyses
suggest that these switching costs are especially lower for employees with high normal workload experience compared to those with high heavy workload experience. In other words, it appears that experience working in normal workload environments is more fungible across different work environments. Taken together, our research informs both organizational learning theory as well as the practice of worker staffing on the
benefits of moving workers around different jobs within an organization.