Matching Collaboration to Disruptions: Relational Event Modeling of Inter-Team Collaboration During Organizational Disruptions

The majority of organizations regularly face disruptions, resulting in substantial financial and reputation damages. Because such actions usually require the collaborative efforts of several teams from different organisations, the members of these teams need to interact with other teams inside and outside their organisation to effectively coordinate tasks, exchange information, and represent teams’ decisions and activities. While most of the team research has been optimistic about the effectiveness of collaboration between teams, less is known about in which situations is higher frequency of inter-team collaboration beneficial, and in which it might lead to a collaborative overload. Following insights from team research and the contingency approach, we aim to examine how can teams optimize the way they collaborate to effectively deal with different types of disruptions. We suggest that the success of disruption resolution depends on the effective alignment between teams’ collaboration and type of disruption. We argue that more complex or severe disruptions, which pose greater demands on the teams, require more centralized collaboration structures. In these circumstances, one or a few team members initiate collaboration on behalf of all teams. Conversely, during less severe or complex disruptions, we expect that teams relying on decentralized collaboration would be more effective in resolving disruptions. We test our conceptual model using data from a public water-supply company in the Netherlands. Collected longitudinal data consists of detailed information about the characteristics of over 8000 disruptions, including indicators of disruption severity, and the effectiveness of disruption resolution, operationalized by disruption duration. In addition, we obtained anonymized information on the telephone communication between organizational teams that dealt with the disruptions, and used it to measure centralization of collaboration. We treat these telephone records as relational event data, temporarily ordered in a sequence of telephone calls. To test the expectations related to the effects of different collaboration patterns on disruption duration, for more and less complex disruptions, we apply relational event modelling. The present study aims to demonstrate that the collaboration forms need to be adapted to the disruption characteristics, challenging the widely held belief that increased collaboration is beneficial in all situations. Most of to-date studies on resilience have relied on some form of subjective data, while the existing empirical studies predominantly used cross-sectional design. We rely on the objective measures of team collaboration, as we analyse interaction behaviour of teams dealing with real disruptions, in their natural surroundings.We leverage the richness of our time-stamped data, and apply relational event modelling to contribute to the recent attempts to investigate team dynamics using event-based approach. Practically, our study provides recommendations about how can practitioners design collaboration procedures between teams to effectively handle varying disruptions.