Maybe It Is Who You Know: Social Networks and Leader–Member Exchange Differentiation

Existing literature on leader–member exchange differentiation (LMXD) offers a meaningful view into the multilevel outcomes associated with leader follower relationships. However, despite the rapid growth of literature on LMXD, scholars lack a complete understanding of its antecedents or the processes that cause leaders to differentiate among team members. We address this issue by using social capital theory to propose that leaders perceive their followers’ social networks as potential resources to grow their own social capital. Because each follower has unique social networks, we propose that leaders differentiate LMX among followers depending on which followers provide them with access to the most social resources. In this conceptual paper, we posit that as leaders gain information about their followers’ social networks, they attribute status to each follower depending on their perception of that follower’s social capital. We then propose that key contingencies, such as a team’s psychological safety climate or a leader’s ambition, influence the relationship between social network characteristics and LMXD. Overall, our model provides a unique multilevel perspective of LMXD and provides important insights for both researchers and practitioners alike.

Building Status in an Online Community

We argue that the actions for which actors receive recognition vary as they move up the hierarchy. When actors first enter a community, the community rewards them for their easier-to-evaluate contributions to the community. Eventually, however, as these actors rise in status, further increases in stature come increasingly from engaging in actions that are more difficult to evaluate or even impossible to judge. These dynamics produce a positive feedback loop, in which those who have already been accorded some stature garner even greater status through quality-ambiguous actions. We present evidence from Stack Overflow, an online community, and from two online experiments consistent with these expected patterns.