Network-building behavioral tendencies, range, and promotion speed

We studied 459 Ukrainian civil servants to determine how career network-building behavioral tendencies relate to network range and promotion speed. We identify two main behavioral tendencies for initiating social relationships: (a) networking within formal structured groups organized around activities created specifically to encourage members to form personal bonds (structured foci) and (b) individually driven networking outside these structured foci. The study shows that individually driven networking is related to broader network range, while structured foci networking has an inverted-U relationship to network range. The optimal networking for range involves a moderate level of structural foci networking and high levels of individually driven networking. Broad network range is related to faster promotion speed to higher organizational levels. Extroverts have a tendency toward individually driven networking, while high Machiavellians have a tendency toward both individually driven and structured foci networking. © 2014 Elsevier B.V. All rights reserved. Social network researchers argue that individuals’ human capital only partly determines their career outcomes and that social capital, as reflected in networks of social relationships, can speed promotions within organizational hierarchies (e.g., Burt, 1992; Brass, 1985). One major research tradition, the networks-asresources perspective, examines how social networks enhance career success, such as through faster promotions (e.g., Ibarra, 1995; Campbell et al., 1986; Granovetter, 1973). Rather than focus on the content that flows through social ties (e.g., friendship, advice), this perspective focuses on the structure of individuals’ ties, specifically their personal network range. Range has been conceptualized in three ways (Campbell et al., 1986): personal network density, or the extent to which one’s ties are themselves connected (e.g., Burt, 1992); size, or the individual’s total number of ties; and tie diversity or heterogeneity (e.g., Seibert et al., 2001). According to the networks-as-resources perspective, individuals who have broader range can learn more non-redundant information from diverse groups than can others lacking that access in the organization. Organizations will value most those few employees who have access to diverse information because they potentially understand ∗ Corresponding author. Tel.: +1 859 257 3741. E-mail addresses: [email protected] (A. Shipilov), [email protected] (G. Labianca). 1 Tel.: +33 1 60 72 44 24. organizational goals more broadly and may create unique solutions to organizational problems (Burt, 2004). We adopt the approach of conceptualizing personal network range in the most direct fashion – individual’s heterogeneity of ties to diverse information clusters outside their focal group (Campbell et al., 1986; Seibert et al., 2001). The main information clusters in an intra-organizational context are the units created by the formal organization structure, and range in this context refers to ties crossing these formal unit boundaries (e.g., Oh et al., 2004; Thompson, 1967). Such boundaries demarcate pools of knowledge necessary for the successful functioning of an organization as a whole, but which are separated due to specialization (Burt, 1997; Ibarra, 1995; Tushman and Scanlan, 1981). This separation creates an organizational need for integration, and individuals with broad network range are positioned to satisfy that need and to reap rewards in the form of faster promotions. Because any tie that crosses formal boundaries helps organizations to integrate knowledge and enhances employees’ value, it is particularly relevant to examine the heterogeneity of individuals’ contacts to understand how range affects career-related rewards. What behavioral tendencies lead individuals to become embedded in diverse networks that are broad in range? What are the psychological determinants of these behaviors? The networksas-resources perspective, although important to organizational research, lacks detail in describing the process individuals use to develop their network structure. Rather, the perspective creates prescriptions, such as suggesting that individuals should develop http://dx.doi.org/10.1016/j.socnet.2014.03.006 0378-8733/© 2014 Elsevier B.V. All rights reserved. Author's personal copy 72 A. Shipilov et al. / Social Networks 39 (2014) 71–83 networks with broad range, without suggesting how to do it or who is more likely to engage in the behaviors that maximize range. Our study contributes to the networks-as-resources perspective by elaborating existing theory, increasing understanding of the social contexts in which individuals meet others, and offering a more comprehensive understanding of both network building behavioral tendencies that maximize range and their psychological determinants. We argue that individuals exhibit two main approaches when building network ties: meeting people through joint involvement in structured formal groups or associations organized specifically to encourage meeting others and developing norms of mutual obligation and reciprocity; or meeting network partners through individually oriented activities that are not driven by formal group membership, such as going to a friend’s party and developing a relationship with someone met there by chance. Our main contribution will be to illustrate that individuals’ tendencies to build network ties through either or both of these different approaches are related to differences in individuals’ range in the workplace. Specifically, we use the theory of social organization (Feld, 1981) and group homophily theory (McPherson and Smith-Lovin, 1987) to argue that individuals’ tendency to meet people through an individually driven approach increases personal network range in the workplace. We also argue that the tendency to meet people through membership in formal groups and other structured foci has a negative curvilinear effect on individuals’ network range at work, with those engaging in an above average amount of networking in structured contexts suffering dramatically lower range. We suggest that there is an optimal combination of networking behaviors that maximizes range: moderate amounts of structured foci networking and high amounts of individually driven networking. This optimal combination leads to the fastest promotion speed to higher organizational levels. Finally, we show that these networking behaviors are tied to psychological characteristics rooted in personality, such as extroversion and Machiavellianism. 1. Theory development 1.1. Range and promotions Network range is important because it is a major determinant of a critical career outcome: promotion speed (Seibert et al., 2001). Organizations divide labor into production specializations, so that individuals and organizational units focus on their most immediate tasks and exclude related tasks, which are then assumed by other individuals and units (Burt, 1992). This tendency toward unit specialization is particularly acute in large private and public bureaucratic organizations (Tushman, 1977; Tushman and Scanlan, 1981). Individuals in specialized units tend to lose track of what their colleagues in other units are doing, even though such knowledge can benefit the operation as a whole (March and Simon, 1958; Mors et al., 2008). Employees with ties to numerous formal units outside their focal unit have networks with high range. Occupying boundaryspanning positions between formal units allows them to derive more personal benefits relative to their human capital, compared with others outside such positions. For example, boundary spanners can take ideas from other units and apply them to their own units, so that supervisors will see them as being more technically competent than their peers (Tushman and Scanlan, 1981; Tushman, 1977). Moreover, boundary spanners are thought to have the most innovative ideas (Burt, 2004) and can control resource flows between formal units in their organizations (Burt, 1997). Studies that have directly measured personal network range as formal group tie diversity have generally found that it relates positively to individuals’ career outcomes. In a study of a contract research and development firm, Reagans and McEvily (2003) demonstrate that individuals with wide range can ease the transfer of information between units, which ultimately makes them more valuable because they are the integrative glue that keeps the organization learning and growing. Furthermore, Ibarra (1995) suggests that being connected to diverse groups allows individuals to hear about opportunities for advancement more quickly. She also finds that wide-ranging networks give individuals broad-based political support throughout the organization. Finally, Seibert et al. (2001) indicate that employees who have contacts with different functional units are better able to acquire needed resources from disparate parts of the organization; consequently, they are likely to be rewarded with faster promotion and experience greater career satisfaction. Related studies that did not measure network range as formal group tie diversity, but rather examined how people benefit from having ties that span informal groups within organizations, also suggest a positive relationship between inter-group spanning and career outcomes. For example, Mehra et al. (2001) find that employees positioned between informal groups in a high-technology firm receive high supervisory performance evaluations. Brass (1985) finds that non-supervisory employees who span informal groups in their work units achieve not only high supervisory ratings but also fast promotions. Fleming and Waguespack (2007) find that members of an open source development community with ties spanning more working groups emerge as community leaders. Finally, Oh et al. (2004) show that units whose members have ties spanning other units exhibit high task effectiveness, while Tsai and Ghoshal (1998) suggest that those units are more innovative. Thus, the existing literature suggests that individuals with contacts to many organizational units will be able to access heterogeneous knowledge residing in different organizational silos, will have informed perspectives on what is transpiring elsewhere, will control information flows across units, will enjoy broad political support in the organization, and will ultimately perform more effectively in their jobs. These positive factors will be associated with faster promotions for individuals with higher network range. While Campbell et al. (1986) suggest that range could also be conceptualized as the density of ties between contacts or as personal network size, neither approach is as proximal in capturing the potential resources available to individuals as is examining the spanning of formal inter-unit boundaries. 1.2. The theory of social organization Although achieving range is important for promotion in the workplace, what factors determine an individual’s range? Feld’s (1981, 1982) theory of social organization is the main sociological perspective attempting to explain the origins of personal network diversity. Central to this theory is the focus—a legal, social, physical, or psychological entity around which joint activities are organized. Feld’s focus concept is extremely inclusive: among the examples he mentions are formal and informal entities such as work organizations, formal and informal voluntary organizations, and kinship entities such as families or clans, physical locations such as “hangouts,” city neighborhoods, or courts in the middle of a housing project, and gatherings such as football games. “Foci may be many different things, including persons, places, social positions, activities, and groups. They may actively bring people together or passively constrain them to interact (Feld, 1981: 1018).” Feld argues that individuals meet others either by networking in these foci or in more random, chance encounters, and this has a profound effect on the diversity of their personal networks. He proposes that if individuals build ties around a few foci, their personal networks are likely to be very dense, with many of their social ties Author's personal copy A. Shipilov et al. / Social Networks 39 (2014) 71–83 73 also being tied to each other because foci encourage density (Feld, 1981: 1024; 1982: 797). Foci can be more or less constraining, based on the extent to which they obligate individuals to interact regularly and create social ties that demand “time, effort and affect” (e.g., Feld, 1981: 1023). The constraining nature of foci, in combination with individuals lacking an infinite capacity to maintain large and diverse sets of ties, makes their networks denser and more homogeneous. The result of this density and homogeneity should be a network with narrower range, and one that would be less likely to result in fast promotion in one’s career. One difficulty with Feld’s theory of social organization is that the expansive definition of foci as any legal, social, physical, or psychological entities around which individuals organize their activities makes it difficult to conceive of how individuals would not meet each other within a focus, and it could lead to nearly anything being labeled a focus. As an example, an individual who meets another while walking their dogs could attribute their new tie to that dogwalking focus (a psychological entity) or perhaps to the architect that designed the walking path that enticed them to walk their dogs along the same road (a physical entity). We will, therefore, narrow our investigation to a specific type of foci that are most likely to be network-constraining – “structured foci,” which are ongoing formal entities that actively and regularly bring individuals together to engage in organized joint activities created specifically to encourage members to form personal bonds. By restricting our inquiry to these structured foci, we will be able to understand the extent to which individuals’ tendencies to favor meeting others in these formal entities might encourage personal network density and homogeneity, and whether this is related to workplace range and promotion speed. We also elaborate Feld’s theory of social organization in another manner. While Feld attempts to explain the relationship between foci networking and its resulting network diversity, he avoids elaborating an alternative to networking through foci. Thus he does not articulate the potential tradeoffs between meeting others that become part of a person’s network within foci versus outside of foci, which is critical when we consider that individuals have limited resources to invest in their social networks. Third, if building ties in foci reduces the diversity of one’s network, how can individuals build relationships to enhance their network diversity? Research shows that individuals’ networks have considerable range variations; some people have many range-diversifying ties while others do not (e.g., Reagans and McEvily, 2003). Moreover, evidence shows that some individuals actively network to enhance their range (Ibarra, 1995: 679; Oh et al., 2004; Vissa, 2012), although the specific nature of these behaviors is unclear from the theory of social organization. When a theoretical perspective as well-established as the theory of social organization does not have enough clarity to apply to your research agenda, Lee et al. (1999) suggest using inductive fieldwork to elaborate and redirect it,2 and we pursue that approach here. To identify networking behavioral tendencies and consequences, we administered open-ended surveys and interviewed multiple samples of professionals in the United States and in the Ukraine and asked the respondents to provide examples of behaviors that helped them meet people who assisted their careers (see Section 2 for more details on selection, interviews, and processes for analyzing responses). Our fieldwork allowed us to narrow the definition of foci. When we asked participants how they met people who assisted with their 2 Theory elaboration differs from grounded theory development in that it takes an existing theory as its foundation and then uses the insights from the qualitative fieldwork for elaboration and extension (see Zott and Huy, 2007; Maguire et al., 2004; Uzzi, 1996, 1999, for examples of this approach). career advancement, many named specific foci, such as “clubs,” “associations,” or “forums”—ongoing formal entities that actively and regularly bring individuals together to engage in organized joint activities and created specifically to encourage members to form personal bonds. To distinguish these examples from the theory of social organization’s general notion of foci, we termed these types of networking behaviors structured foci networking (SFN). We then studied the extent to which professionals had a tendency to engage in these behaviors in additional independent samples, and related this to their workplace network range and career promotion speed. We found that networking outside of structured foci was critical for individuals to meet people who ultimately helped advance their careers. Many respondents reported they met people by “striking up conversations while traveling,” “inviting people to dinner,” or “going out for drinks.” These alternative non-structured foci networking behaviors involve people using their own initiative to meet others directly, without the intermediation of ongoing structured entities. For example, airline passengers are brought together on a one-time basis; meeting and forming a relationship is incidental to the flight’s purpose. But if two passengers, on their own initiative, engage in direct interpersonal networking during the flight, this type of networking behavioral tendency merits its own label. We call it individually driven networking (IDN). 1.2.1. Structured foci networking (SFN) and range By narrowing the notion of foci to structured foci, we can extend Feld’s theory on how networking in foci might affect range. Our introduction of SFN and IDN and our recognition that these two types of networking behavioral tendencies might be interdependent and involve tradeoffs means that we must consider the costs involved in meeting people in either type of setting and potential effects on range. Networking incurs three general types of costs: initiation, maintenance and opportunity costs (Bala and Goyal, 2000; Burt, 1992, 2002). Initiation costs are incurred when individuals search for new partners (Bala and Goyal, 2000; Gargiulo and Benassi, 1999; Portes and Sensenbrenner, 1993). Maintenance costs are incurred when individuals invest time, affect, attention, and other resources to maintain ties (Bala and Goyal, 2000; Burt, 2002: 344; Feld, 1981). Opportunity costs are levied by the extent to which social ties constrain an individual’s ability to form new ties (Burt, 1992; Gargiulo and Benassi, 1999, 2000; Portes and Sensenbrenner, 1993). An individual’s tendency to SFN will entail lower initiation costs for his or her relationships, whether or not these relationships enhance their network diversity. By joining the focus and attending an organized function, parties signal that they have common interests, are willing to socialize, and perhaps want to initiate a relationship (Feld, 1982). As an example, finding a romantic partner is easier at a dating club than through a random nightclub encounter. The same principle applies in an organizational setting—meeting someone on a cross-unit task force is more likely to lead to a relationship than meeting in the company cafeteria. Common-focus membership also encourages partners to introduce individuals to other focus members (Gargiulo and Benassi, 2000), and this process of transitivity also expands an individual’s network (Feld, 1981) by lowering initiation costs. However, although SFN lowers the initiation costs of relationships, it increases the maintenance and opportunity costs for range-diversifying ties. Individuals building relationships through SFN will be subject to constraints—behavior norms, expectations, and obligations—the price they must pay for continued group membership (Feld, 1981, 1982; also see Hackman, 1992; Schein, 1965; Portes and Sensenbrenner, 1993 for related discussions). Membership means that the focus will constrain their actions; it will organize and dictate more of their social life and reduce their Author's personal copy 74 A. Shipilov et al. / Social Networks 39 (2014) 71–83 scope of activities (Krackhardt, 1999). Foci will also tend to increase cohesion by pressuring members to initiate relationships with other group members, taking their time and attention away from forming ties with outsiders. Finally, foci encourage the maintenance and renewal of intra-group relationships, as well as continued reciprocity among group members, making the ties themselves stronger (Granovetter, 1982; Levin and Cross, 2004), but also creating relational inertia (Gargiulo and Benassi, 2000). Adherence to such rules, often necessary for maintaining an individual’s focus membership (Baker, 1994), directs the individual’s network formation more internally to the focus (McPherson and Smith-Lovin, 1987). As an individual engages in SFN, the opportunity costs and maintenance costs for range-diversifying ties increase because the pressures toward homophily among group members tend to be stronger than among people who meet outside of foci (Feld, 1981; McPherson and Smith-Lovin, 1987). McPherson and Smith-Lovin (1987) find that groups introduced stronger homophily across more categories than dyadic homophily not attributable to the group. Although some perceived similarity is likely behind any connection, including one initiated via individually driven networking, the key difference is that IDN efforts are likely to occur in a broad array of contexts. Consequently, when a structured focus is the site of the networking, the actors are likely to be more similar. Group homophily theory identifies four other mechanisms that encourage homogeneity in groups, including homophilous recruitment based on existing members recruiting new members (e.g., Babchuk and Booth, 1969); a desire for low intra-group conflict that biases toward greater homogeneous membership (Davis, 1963); inter-group competition for members that forces groups to develop distinctive social niches (McPherson, 1983); and taskbased pressures that make some individuals more suitable for membership (McPherson and Smith-Lovin, 1987). Beyond these mechanisms, these groups create rules and obligations that attract certain people and drive other non-conforming members away (cf. Schneider et al., 1995). Groups can also limit the member’s ability to form range-diversifying ties, either through norms that define who is socially unacceptable or because when individuals adhere to intra-group obligations, they lose time and attention that they could otherwise invest in seeking range-diversifying partners. Consequently, as Feld (1982: 797) points out, “social structuring of an activity leads people to develop relationships with others like themselves.” Thus as individuals engage in extensive SFN, it becomes increasingly difficult for them to build other relationships that will diversify their range. At low levels of SFN, the higher opportunity and maintenance costs of building range-diversifying ties might be hidden. Individuals engaging in minimal SFN behaviors and having networks mainly within their organizational units have relatively low costs for potentially missing out on opportunities to meet partners outside of their units. If the employee is new to an organization, for example, meeting others through a structured focus, such as a company-sponsored softball team, will help develop some rangediversifying relationships. As a result, low levels of SFN will initially be associated with an increase in range. However, as the employee begins to engage in more SFN, each focus will generate pressures of greater constraint and higher overall costs for the individual’s ability to diversify networks—simply because the employee will need more time and attention to maintain membership in multiple groups and satisfy their expectations. The likelihood also increases non-monotonically that the groups’ norms and obligations will conflict. The costs of reducing these conflicts, while still fulfilling the obligations necessary to maintain good standing in each of them, will drain energy and resources (Hackman, 1992; Simmel, 1955; see also Krackhardt, 1999). In response, the individual will attempt to maintain multiple group memberships by relating with partners who are acceptable across the groups—reducing his or her personal network range. In summary, small amounts of networking in the SFN context are beneficial for increasing range because of minimized initiation costs for establishing new contacts. But as SFN increases beyond low to moderate levels, the individual suffers greater time and attention pressures, more constraints driven by the increasing likelihood of inter-group conflicts, and added homogenizing pressures inside the groups. These constraints increase the opportunity and maintenance costs of range-diversifying ties, resulting in the following relationship: Hypothesis 1. Individuals’ tendency to engage in structured foci networking behaviors will have a negative curvilinear (inverted-U) relationship with their network range. 1.2.2. Individually driven networking (IDN) and range IDN occurs outside a group’s active intermediation, so IDN and range should show a different relationship. Specifically, IDN should incur higher costs of initiating relationships precisely because such ties are built outside of foci. Without group-based referrals, IDN lacks bounded solidarity and enforced trust (Portes and Sensenbrenner, 1993); people tend to be more wary of initiating new relationships on their own. Moreover, without third parties to encourage tie formation as happens in a group, IDN offers fewer chances of initiating new ties. However, in contrast to SFN, IDN does not increase maintenance or opportunity costs for range diversifying ties. This happens because individuals are not consistently bound together by recurring organized activity and are less pressured to combine their activities with others. For example, two people from different units meet in the company cafeteria and strike up a long-term relationship. Their relationship will impose fewer pressures to avoid other potential network contacts or expectations that they will help each other’s friends (often called pressures for transitivity), as structured foci impose on members (Gargiulo and Benassi, 2000). With lower pressures for transitivity, individuals engaging in IDN enjoy greater personal freedom and more time to pursue relationships with heterogeneous others. They are not prohibited from turning their attention to initiating range-diversifying relationships with people in different units or departments. Homophily is also likely to drive IDN tie formation, but partners who meet via IDN are likely to differ on more dimensions than are partners who meet via SFN (McPherson and Smith-Lovin, 1987), which also increases the probability that IDN will enhance an individual’s network range. Hence, we propose: Hypothesis 2. Individuals’ tendency to engage in individually driven networking will be positively related to their network range. 1.2.3. Personality traits and tendency to network via SFN and IDN Individuals are likely to have certain personality-driven propensities and preferences to engage in structured foci and individually driven networking. These personality characteristics might alter their subjective evaluations of the initiation, maintenance, and opportunity costs for SFN and IDN, making them more or less likely to engage in these networking tendencies. The main personality construct dealing with sociability is extroversion. Individuals higher in extroversion are more gregarious, bold, talkative and unrestrained (e.g., Saucier, 1994; Goldberg, 1999). Extroverts are more likely to proactively socialize when searching for jobs (Wanberg and Kammeyer-Mueller, 2000) and to socialize and attempt to improve their visibility in the workplace (Forret and Dougherty, 2001). Extroverts tend to seek out and enjoy social interaction and feel more comfortable in social situations than introverts (Lucas and Diener, 2003). Extroverts are not as likely to view tie initiation costs as being as high as more Author's personal copy A. Shipilov et al. / Social Networks 39 (2014) 71–83 75 introverted individuals, who tend toward being shy, bashful, timid, retiring, quiet, and more socially aloof (John and Srivastava, 1999). Extroverts would logically orient themselves toward individually driven networking more than introverts, who would find tie initiation more costly and difficult, irrespective of the relative size of their networks (cf., Dougherty et al., 2008). Hypothesis 3. Individuals’ extroversion will be positively related to their tendency to engage in individually driven networking. Individuals high in agreeableness tend to show more prosocial and altruistic behaviors, such as higher empathy (Nettle, 2006), greater willingness to cooperate (Denissen and Penke, 2008), kindness and interpersonal warmth (John and Srivastava, 1999) and greater use of integrative conflict-handling strategies (JensenCampbell et al., 2003). Agreeable individuals have larger social networks (Klein et al., 2004; Selfhout et al., 2010). Theoretical arguments have also been made that agreeable individuals self-select into tight-knit, cohesive groups (Janicik and Larrick, 2005), which might account in part for why groups with more agreeable individuals or with an agreeable leader tend to be more socially cohesive (Barrick et al., 1998; Peterson et al., 2003). We argue that individuals higher in agreeableness are less likely to subjectively perceive the higher maintenance and opportunity costs of networking within structured foci as much of a deterrent, in comparison to those who are lower in agreeableness, owing to a greater affinity for social cohesiveness. Their networking tendency will, therefore, be oriented more toward structured foci networking, which encourages cohesiveness, than those lower in agreeableness. Hypothesis 4. Individuals’ agreeableness will be positively related to their tendency to engage in structured foci networking. Individuals higher in Machiavellianism are more oriented toward gaining power and to taking a more calculating, manipulative approach toward others (Christie and Geis, 1970; Wilson et al., 1996). High Machiavellians would, therefore, be interested to maximize the instrumentality of their social network, including having more social ties (cf., Clifton et al., 2009, which studied a related dark triad personality construct, narcissism). Indeed, others have argued that those higher in Machiavellianism actively seek out exploitable network positions (Janicik and Larrick, 2005; Brass et al., 1998). If this is true, individuals higher in Machiavellianism should be motivated to bear higher initiation, maintenance, or opportunity networking costs than those lower in Machiavellianism in order to maximize their network’s instrumentality. Thus, we expect to see a positive relationship between Machiavellianism and both individually driven as well as structural foci networking.3 Hypothesis 5. Individuals’ Machiavellianism will be positively related to their tendency to engage in individually driven and structured foci networking.