Learning from Many: Partner Exposure and Team Familiarity in Fluid Teams

In services where teams come together for short collaborations, managers are often advised to strive for high team familiarity so as to improve coordination and, consequently, performance. However, inducing high team familiarity, by keeping team membership intact, can limit workers' opportunities to acquire useful knowledge and alternative practices from exposure to a broader set of partners. We introduce an empirical measure for prior partner exposure and estimate its impact (along with that of team familiarity) on operational performance using data from the London Ambulance Service. Our analysis focuses on ambulance transports involving new paramedic recruits, where exogenous changes in team membership enable clean identification of the performance effect. Specifically, we investigate the impact of prior partner exposure on time spent during patient pick-up at the scene and patient handover at the hospital. We find that the effect varies with the process characteristics. For the patient pick-up process, which is less standardized, greater partner exposure directly improves performance. For the more standardized patient handover process, this beneficial effect is triggered beyond a threshold of sufficient individual experience. In addition, we find that the beneficial performance impact of prior partner exposure is amplified during periods of high workload, for both processes. Finally, a counterfactual analysis based on our estimates shows that a team formation strategy emphasizing partner exposure outperforms one that emphasizes team familiarity by about 9.2% in our empirical context.