Employee Turnover and Tacit Knowledge Diffusion: A Network Perspective

Abstract

An enormous amount of information and knowledge resides in the minds . . . of key people, but this material is rarely organized in a fashion that allows for its transmission to others (Powell, 1998: 237). Just as the apprentice learns the tools of the trade from a master, businesses gain from the knowledge shared by mentors, supervisors, coworkers, project team members, and long-tenured employees. Yet the business world is in the midst of an era characterized by the boundaryless career (Arthur and Rousseau, 1996)-- one where median employment tenure is just four and a half years, new job creation accounts for only one tenth of all career moves, and large firm decentralization is a regular occurrence. know I can't stop people from walking out the door--but how do I stop them from taking their knowledge with them? (Labarre, 1998:48). That is, when leave, companies lose not only human capital, but also accumulated knowledge. This is a common problem firms face in the knowledge economy and the central issue addressed in this article. As consulting, research, and information technology firms are realizing, their whole business is pretty much locked away in the minds of employees (Koudsi, 2000: 233), yet this knowledge is rarely shared, swapped, traced, and fertilized to ensure that it remains, at least in part, with the firm when leave. The problem's significance is shown by the fact that many businesses are spending millions of dollars each to develop and purchase solutions to combat knowledge exodus (Koudsi, 2000; McCune, 1999). Companies, recognizing knowledge as a valuable asset, are busily devising ways to capture it, from narrative re-creations of past triumphs to rewards for in formation gleaned in exit interviews (Branch, 1998). Organizational knowledge and employee turnover have been studied extensively. Our contribution is a link between the two, whereby social networks explicate the connection between employee turnover and tacit knowledge loss. Closely related to social networks is the concept of social capital. We adapt the meaning suggested by Tsai and Ghoshal (1998) in defining social capital as resources embedded in social relationships as well as the norms and values inherent in such relationships. Others (e.g., Dess and Shaw, 2001) have suggested that employee turnover can negatively affect firm performance through loss of social capital. We expand this by taking into account the tacit knowledge that firms lose when leave. In light of employee turnover, we focus on social network structures likely to lead to retention of the tacit knowledge embedded in employees' minds. We offer propositions concerning the problem of tacit knowledge loss and encourage the development of solutions that take into account the social n etwork structure of organizations. Specifically, we posit that 1) tacit knowledge can be preserved, in part, when firms promote employee interaction, collaboration, and diffusion of non-redundant tacit knowledge, and 2) characteristics of a firm's social network, including density and an optimal mix of weak and strong ties, promote interaction, collaboration, and non-redundant tacit knowledge diffusion. This paper is divided into three major sections. First, we introduce the general theoretical background, followed by more specific theoretical discussions of the tacit knowledge, knowledge-based view of the firm, and employee turnover literature. Second, we frame our propositions in the context of a firm's social structure, highlighting the interplay among diffusion, interaction and collaboration, and non-redundant information. Third, we provide a summary, implications, and future research directions. BACKGROUND A major challenge facing organizations is uncovering the most effective methods of gathering and applying knowledge en route to economic value creation (Miles et al., 1998). In our technological, global society, this need for knowledge is more salient than ever before. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 214 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.