An Extended Model of Social Embeddedness: Applying Social Network Theory to Enrich Job Embeddedness Theory

An Extended Model of Social Embeddedness: Applying Social Network Theory to Enrich Job Embeddedness Theory An Extended Model of Social Embeddedness: Applying Social Network Theory to Enrich Job Embeddedness Theory Employee turnover has long captivated the attention of practitioners and scholars alike (Griffeth & Hom, 2001). Employers worry about exorbitant attrition, which incurs personnel costs (e.g., recruiting and training replacements; Hom & Griffeth, 1995), hampers progress toward a diversified workforce (Hom, Roberson, & Ellis, 2008), and decreases corporate effectiveness (Kacmar, Andrews, Rooy, Steilberg, & Cerrone, 2006). Since the dawn of the 20th century, researchers have investigated turnover as reflective of basic motivational processes or as a phenomenon meriting separate scrutiny given March and Simons (1958) original pronouncement that motivating employees to participate is as vital as motivating them to produce. Over the years, scholars have proliferated and tested increasingly complex theories of turnover determinants (Hom & Griffeth, 1995). Despite greater breadth, existing multivariate models continue to modestly predict withdrawal, accounting for less 25% of the variance (Maertz & Campion, 1998). To break through this ceiling on variance explained, Mitchell, Holtom, Lee, Sablynski, and Erez (2001) explain why job incumbents stay rather than leave. Conventional perspectives on why people quit miss the different emotional and psychological processes implicated in staying. Lending credence to Mitchell et al.s (2001) claim that staying is not the obverse of leaving, behavioral intention models posit that both reasons for and reasons against performing an act are needed to fully comprehend behavioral decisions (Prestholdt, Mathews, & Lane, 1987; Westaby, 2005), while regulatory focus theorists hold that approaching positive goals (e.g., promoting staying) elicit different emotions and strategies than avoiding negative goals (preventing leaving; Brockner & Higgins, 2001; Higgins, 1997). Mitchell et al. (2001) thus pioneer job embeddedness, which comprises three classes of forcesboth on and off the jobthat bind employeesnamely, fit (how well they fit the job and community), links (work and extrawork connections), and sacrifice (job and community amenities relinquished upon leaving). This novel conception inspired mounting evidence for its validity for predicting quits in various jobs (Crossley, Bennett, Jex, & Burnfield, 2007; Mitchell et al., 2001) and China (Hom et al., in press). Other researchers find that embeddedness promotes corporate-wide retention, performance, and citizenship (Felps et al., in press; Lee, Mitchell, Sablynski, Burton, & Holtom, 2004). Our project extends this provocative conceptualization of embedding forces by applying social network theory and methodology to identify additional relational constraints on leaving. Indeed, Holtom, Mitchell, Lee, and Eberly (2008) suggest that this approach mightclarify more precisely how the links described in social networks influence ones propensity to stay on their job (p. 30). Drawing from the network literature, we enrich Mitchell and Lees (2001) conception of links as volume of links by considering how tie strength (Mossholder, Setton & Henagan, 2005), dense networks (network closure, Burt, 2005), and social capital (resources embedded in relationships; Lin, 2001; Siebert, Kramer, & Liden, 2001) can enhance loyalty. Moreover, we address an oversight in Mitchell and Lees theory by exploring how links might de-embed incumbents. Specifically, theory and work on normative prescriptions (Maertz & Griffeth, 2004; Westaby, 2005) and turnover contagion (Felps et al., in press) find that social contacts can stimulate departures through their words (verbal persuasion) or deeds (their quits can snowball into more quits; Krackhardt & Porter, 1986). Further, the mentoring and occupational attainment literatures imply that external constituents can encourage exits by advising incumbents to change careers, identifying job leads, or securing jobs for them (Higgins & Kram, 2001; Lin, 2001). In sum, we propose and test a model integrating v