STRUCTURAL HOLES IN GUANXI NETWORKS: DO THEY INCREASE EMPLOYEE TURNOVER IN THE PEOPLE’S REPUBLIC OF CHINA? PETER W. HOM Department of Management, Arizona State University Tempe, Arizona 85287-4006 ZHIXING XIAO China Europe International Business School INTRODUCTION Since March and Simon’s (1958) classic book, Organizations, turnover researchers have predominantly investigated job attitudes and labor market conditions as central determinants of voluntary quits (Hom & Griffeth, 1995; Lee & Mitchell, 1994). Despite decades of inquiry, these constructs—or variables of the same type--have only modestly predicted turnover (Griffeth, Hom, & Gaertner, 2000; Maertz & Campion, 1998). In the wake of such findings, scholars are exploring other determinants to enhance turnover prediction and understanding (Maertz & Campion, 2004; Mitchell & Lee, 2001). Foremost among recent theoretical developments is the introduction of relational constructs to capture the quality and nature of workplace relationships (Maertz, 2004; Mitchell, Holtom, Lee, Sablynski, & Erez, 2001b; Mossholder, Settoon, & Henagan, 2005). Given the ascendancy of the knowledge economy, workplace ties are attracting greater scrutiny as professionals are known to bond more with their peers than employing institution (Dess & Shaw, 2001). Though affect for superiors and co-workers have long been recognized as bases for staying (Hom & Griffeth, 1995), present-day scholars are noticing how employees’ direct relational ties to a broad set of organizational (including extrateam) and extraorganizational constituents can induce loyalty (Maertz & Griffeth, 2004; Mitchell & Lee, 2001; Mossholder et al., 2005). Extending research on direct relational ties, our investigation considers how broader social networks encompassing employee-to-colleague relationships as well as colleague-to-colleague relationships (or “indirect ties,” Podolny & Baron, 1997) promote retention in the People’s Republic of China (PRC). Specifically, we examine how constrained or closed networks maintain job incumbency among Chinese nationals (Burt, 1992, 1997, 2001). A constrained network emerges when an individual’s set of contacts communicate with each other or the same third parties, generating redundant and dense interconnections (Burt, 1992). By contrast, an unconstrained network has gaps in a person’s network of relationships (or structural holes), whereby his or her contacts are neither directly nor indirectly connected via common third parties (Burt, 1992). Early network research established that leavers affect incumbents’ quit decisions by decreasing job attitudes (via lost friendships or affect contagion through communications with them or mutual contacts; Shaw, Duffy, Johnson, & Lockhart, 2005; Totterdell, Wall, Holman, Diamond, & Epitropaki, 2004) or by stimulating “snowball” effects where leavers’ actions prompt occupants in similar positions in the social structure (“structural equivalents”) to exit (Feeley & Barnett, 1997; Krackhardt & Porter, 1986). Other studies have shown that network centrality—or average number of network links—can stem departures (Feeley & Barnett, 1997; Mossholder et al., 2005). Though scholarly explorations of voluntary groups reveal that network closure prolongs membership (though these tests assessed contacts’ co-memberships rather than their Academy of Management Best Conference Paper 2006 OB: F1