Short- and long-term stability in organizational networks: Temporal structures of project teams

•We propose that social processes operate over different time frames.•Reciprocity and closure have distinct meanings over different time frames.•We investigate the temporality of these processes with a relational event model.•We examine cohesion in two project teams marked by large compositional changes.•Long-term closure reflects cohesion while short-term reflects adaptation to change. Network research focuses on patterns of stable relationships, where stability represents the unfolding of social processes over long time frames. We argue that social interactions exhibit important regularities in different time frames (short and long term), reflecting distinct social processes. We illustrate the value of this distinction through a comparative case study of technology-mediated communication, within two project teams in a digital marketing agency. We examine how the embedding of interpersonal interactions in processes of reciprocity and closure over different time horizons enables the emergence of cohesion in the face of constant compositional changes. We propose that the time frames in which stable patterns of interactions develop are the key to understanding the nature of the underlying social processes with short-term patterns of closure and reciprocity representing adaptation to change while longer term patterns indicate cohesion. Our results are supportive of this argument and show that the two teams exhibit the same regularities in interactions but across different time horizons. We discuss the implication of our findings and argue that distinguishing between short- and long-term stability of social networks offers a novel and promising avenue for network research.