Abstract
Organizational socialization is the process by which newcomers adapt to their organization and learn how to become productive organizational members. Organizations employ tactics such as classroom training, mentoring, orientation sessions, and on-the-job training to assist new employees in their transition. At the same time, newcomers engage in their own proactive efforts to seek information and establish ties with experienced organizational members who have access to valued resources. While scholarship has noted the importance of both socialization and employee intraorganizational networks within public sector organizations, little research has focused on these areas. This article links the organizational socialization and social network frameworks by examining how the networks of new employees in a state agency change over time. Based on our findings, we offer implications for rewarding core advisors, mentoring programs, and fostering interactions between newcomers and experienced organizational members.
Organizational socialization is the process by which newcomers become productive members of an organization (Louis, 1980; Romzek, 1990; Van Maanen & Schein, 1979). Organizations employ a series of tactics to instill organizational values, culture, behavioral expectations, and necessary job skills (Louis, 1980; Van Maanen & Schein, 1979). The new employee is tasked with absorbing as much as possible while defining his or her role within the organization and generally learning how to fit in. The newcomer must sift through the varied organizational and job-related information and messages to determine what it all means, what is essential and what is not, and who within the organization is important in terms of power, influence, and access to resources.
In a seeming paradox, although the organization sends a myriad of information bits and normative signals to newcomers through the socialization process, in reality newcomers receive only a mere fraction of the knowledge they require to fully become an organizational member (Miller & Jablin, 1991; Morrison, 1993). New employees thus engage in proactive efforts to gain access to more job-related information, organizational knowledge, and feedback to help smooth their adaptation (Miller & Jablin, 1991; Morrison, 1993; Ostroff & Kozlowski, 1992; Saks & Ashforth, 1997a, 1997b). One way in which new employees proactively acquire information is by establishing connections to seasoned organizational members; these ties offer opportunities to gain tacit knowledge not readily available from formal organizational resources (Ostroff & Kozlowski, 1992). The social ties established by newcomers can be conceptualized as personal, or egocentric, social networks that offer access to organizational and job-related information, resources, and support. While organizational socialization studies have begun to examine newcomers’ proactive information seeking through social ties (Miller & Jablin, 1991; Morrison, 1993, 2002), there is still a great deal we do not know.
This article investigates changes in newcomers’ social networks as they are integrated into a large state agency. We seek to add to the nascent literature on employee networks and organizational socialization within public sector organizations. Although public management scholars have noted the importance of organizational socialization to public sector organizations, particularly with respect to public service motivation, to date it has garnered little attention within the public management literature (Moynihan & Pandey, 2007; Perry, 2010; Perry & Wise, 1990; Romzek, 1990; see Oberfield, 2010 for one example). Similarly, although several studies have examined interorganizational networks in public management, few studies have considered or examined public sector intraorganizational networks (Moynihan & Pandey, 2008). Recent work by Hatmaker, Park, and Rethemeyer (2011), Lee and Kim (2011), and Moynihan and Pandey (2008) are a few exceptions. Moynihan and Pandey (2008) in particular note that studying employee networks can lend insight into employee behavior and attitudes, and they call for increased research in this area. More specifically, social network analysis enables the examination of the patterns of relationships between employees, rather than individual actors in isolation, and these structured patterns of interaction represent both opportunities for and constraints on behavior (Brass, Galaskiewicz, Greve, & Tsai, 2004).
This article uniquely contributes to this literature in a few ways. First, we use longitudinal, egocentric employee social network data from a large, regionally dispersed state agency. We are aware of no other longitudinal studies of new employees’ egocentric networks in a public sector organization. In addition, our article follows newcomers and their networks for more than 3 years, longer than most longitudinal socialization studies (Morrison, 1993), enabling us to distinctively examine how their relationships change over time.
Our article also adds to the management literature on organizational socialization. Few studies have examined new employee egocentric networks in the course of organizational socialization within the public management or the management and organizations literature (see Hatmaker et al., 2011; Morrison, 2002). As such, our article works from broad research questions to inductively examine the nature of new employees’ social networks in the context of organizational socialization. This article first reviews relevant literature and outlines research questions that will be explored using our data. We then describe the organizational context, data collection, and analytic methods used for this study. After presenting our findings and analysis, we conclude with a discussion of implications for public managers and administrators.
Organizational Socialization and Proactive Behavior
Organizational socialization is the process by which newcomers learn the knowledge, behaviors, and values required to adopt a role within an organization (Van Maanen & Schein, 1979). It is a boundary spanning process by which newcomers become full-fledged organizational insiders by being “given broad responsibilities and autonomy, entrusted with “privileged” information, included in informal networks, encouraged to represent the organization, and sought out for advice and counsel by others” (Louis, 1980, p. 231). It is also a learning process; newcomers must become proficient in and understand the norms, values, tasks, and roles that define organizational membership (Ashforth, Sluss, & Harrison, 2007). Organizational socialization efforts have two primary goals: (a) inculcating organizational values, norms, and culture in newcomers, and (b) transferring knowledge and information required by newcomers to perform effectively and complete their job tasks (Bauer, Morrison, & Callister, 1998; Louis, 1980; Romzek, 1990; Van Maanen & Schein, 1979).
Both newcomers and organizations can benefit from effective organizational socialization strategies and tactics. Socialization tactics have been shown to positively affect organizational commitment, job satisfaction, person-to-organization fit, role clarity, job embeddedness, task mastery, and adoption of the organizational culture (Allen, 2006; Anakwe & Greenhaus, 1999; Ashforth et al., 2007; Bauer et al., 1998; Jones, 1986; Kim, Cable, & Kim, 2005; Romzek, 1990; Saks & Ashforth, 1997a). These factors can influence the likelihood that a person may leave a job, and researchers have also found that socialization is linked to lower turnover and intentions to quit (Allen, 2006; Ashforth et al., 2007; Saks & Ashforth, 1997a). Employees who are socialized effectively are skilled, committed individuals who fit in well, perform well, and may be likely to serve the organization for a long tenure.
Organizational socialization involves efforts by both the organization and the new employee. Training, orientation sessions, apprenticeships, on-the-job learning, and mentoring are all forms of discrete socialization tactics employed by organizations. In their seminal work, Van Maanen and Schein (1979) identified six types of socialization tactics implemented by organizations that influence newcomer integration. Jones (1986) then classified these tactics as either institutionalized or individualized. Institutionalized tactics are those in which the organization drives and controls the newcomers’ integration, leading newcomers to accept predefined roles, while individualized tactics are those tactics in which newcomers “develop their own approach to their roles” (Ashforth et al., 2007; Jones, 1986; Saks & Ashforth, 1997b, p. 49) The efficacy of institutionalized socialization efforts is at least in part a function of the newcomer’s individual agency, personal characteristics, and self-efficacy (Filstad, 2004; Jones, 1983, 1986). These same characteristics and traits may be even more important for those employees whose organizations do not employ institutionalized socialization tactics since they must proactively initiate their own efforts for adaptation. Speed of transition is also influenced by these factors; how quickly a newcomer transitions into an organizational insider is dependent in part on the success of newcomers’ proactive efforts (Reichers, 1987).
Early research on organizational socialization largely focused on the organizational efforts to integrate newcomers, but more recent research considers the individual agency of newcomers and examines their proactive behaviors (Ashforth et al., 2007; Ashforth, Sluss, & Saks, 2007; Bauer et al., 1998; Kim et al., 2005; Jones, 1983; Morrison, 1993; Ostroff & Kozlowski, 1992; Saks & Ashforth, 1997a, 1997b). From the newcomers’ perspective, socialization is a sensemaking process in which newcomers interpret and attach meaning to events in their new setting (Louis, 1980). A newcomer’s entry into the organization is characterized by uncertainty, and proactive information seeking can help them to reduce unknowns by enabling them to better understand the environment (Miller & Jablin, 1991; Morrison, 1993). These proactive efforts include obtaining feedback, information, and emotional support from supervisors, peers, members of professional organizations, and even upper level management (Miller & Jablin, 1991; Morrison, 1993; Ostroff & Kozlowski, 1992; Saks & Ashforth, 1997a, 1997b).
Newcomers require and seek several different types of information to master their jobs and become organizational members. Researchers have categorized this information into three categories that are linked to newcomer learning and that are needed to become an organizational member: (a) organizational information about norms, policies, politics, history, and structure; (b) job information that helps one to understand how to perform one’s job, and (c) role information that helps one to understand the expectations and constraints associated with one’s formal organizational role (Morrison, 1993, 2002; Ostroff & Kozlowski, 1992). Over time, the types of information required by newcomers may change. Morrison (1993, p. 583) notes that “as time progresses newcomers become less concerned about fitting in and more concerned about how well they are performing.” This transition could be an indicator of job mastery, or they may believe that over time it is less acceptable to seek task-related information because it is knowledge they should have gained early on in their tenure.
The relationship between newcomers and insiders (i.e., more experienced members of the organization) plays a critical role in new employees’ socialization (Higgins & Kram, 2001; Morrison, 2002; Podolny & Baron, 1997). Experienced colleagues play an important role in socialization efforts, positively effecting task mastery, role clarity, and the adoption of the organizational culture (Anakwe & Greenhaus, 1999). Even while newcomers are experiencing institutionalized socialization tactics, they actively engage in individualized tactics in which they seek out information from their own select sources (Ashforth et al., 2007; Miller & Jablin, 1991; Morrison, 1993, 2002; Ostroff & Kozlowski, 1992). Developing ties to peers, more senior coworkers, supervisors, and upper level managers enables newcomers to access resources that may not be available formally through socialization channels (Louis, 1980; Morrison, 1993, 2002; Ostroff & Kozlowski, 1992). They are able to “fill in the blanks” that remain even after attending orientation sessions, participating in training, perusing organizational manuals, or experiencing other organization-sponsored socialization tactics.
Often individuals with extensive experience and tenure within an organization hold a great deal of tacit knowledge about the organization and job tasks that is not readily available elsewhere. To access this information, newcomers must seek out these individuals and develop ties with them. Newcomers may survey the field of coworkers, peers, and superiors to understand who has what experience, knowledge, and other resources to be accessed, and then determine how to establish connections with these individuals. These proactive behaviors enable newcomers to develop a repertoire of task information, organizational knowledge, and social support which, ultimately, may lead to opportunities for advancement (Higgins & Kram, 2001; Morrison, 1993, 2002; Podolny & Baron, 1997).
About Egocentric Social Networks
Since little public management literature discusses egocentric networks, this section provides a brief discussion of egocentric networks and their analysis. Egocentric, or personal, networks are the set of social ties that a particular individual establishes with others (Marsden, 1990; Wellman, 1983). Egocentric network analysis strives to understand and explain behavior by examining an individual’s (ego’s) connections to others (alters; Marsden, 1990). By analyzing personal networks, we attempt to understand and explain differences across individuals based on their relationships and the access to opportunities, information, and other important resources that result from these relationships (Marsden, 1990). Interpersonal networks have been found to influence job satisfaction, organizational commitment, turnover, one’s ability to get a job, performance, and career advancement (Brass, 1984, 1985; Brass, Galaskiewicz, Greve, & Tsai, 2004; Burt, 1992; Granovetter, 1973; Krackhardt & Porter, 1985; Lee & Kim, 2011; Morrison, 2002; Moynihan & Pandey, 2008; Podolny & Baron, 1997).
Egocentric networks may be conceptualized as instrumental or expressive networks (Ibarra, 1992). Instrumental networks may enable one to gain access to information, support, and resources that are important to successfully doing one’s job, advancing one’s career, or gaining a sense of belonging within a profession or organization (Brass, 1984; Burt, 1992; Ibarra, 1992, 1997; Podolny & Baron, 1997). In our study, instrumental networks focus on task or technical advice, organizational information, and role models. Expressive networks have an affective component; they offer socioemotional resources that can help an individual feel they fit in well within the organization or work group, such as friendship and social support (Ibarra, 1992, 1997). Friendship and social support networks are expressive networks comprised of individuals with whom one socializes, or to whom one goes to for friendship or social support.
To analyze egocentric networks, researchers examine both the structure of the networks as well as measures regarding the nature of the connections. One network measure that offers insight into the nature of an egocentric network is density, or the connectedness of the network actors (Marsden, 1990). When network members are all connected to each other, the network consists of many redundant ties—members are able to access the same information and resources through several different alters. In sparse networks with fewer connections, individuals are less likely to get the same information from multiple sources and thus may be more efficient in gathering resources (Burt, 1992).
To understand the network composition, we can examine who the network members are, for example in terms of organizational position or status within the organizational hierarchy, and the nature of the ties themselves. The greater the diversity of a network’s members, the more likely an individual has access to different types of resources and information, giving that person an advantage over those who have less diverse connections that do not provide varied information and resources. Network size, or, the number of network members, can indicate how effective an individual is at establishing connections, how well integrated an individual is or even simply an individual’s popularity (Marsden, 1990). When combined with the diversity measure, network size can also indicate how effective one is at creating ties that can provide different types of resources.
Examining the ties between individuals can also reveal information about the nature of the relationship between network members, such as how strong the connection is. Tie strength can refer to the amount of time, frequency of communication, emotional intensity, and reciprocity in a connection between two individuals in a network (Granovetter, 1973, p. 1361; Marsden, 1990). Weaker ties can provide bridges to more distant contacts which can offer access to information that an individual would not receive from a direct tie (Granovetter, 1973).
Socialization via Egocentric Networks
Newcomers’ network relationships offer access to a variety of resources instrumental for the socialization process and can facilitate role transition. Connections to more experienced organizational members offer job-related task and organizational information. Role models may be particularly helpful with assisting newcomers in their adaptation (Filstad, 2004). Role model clusters can be conceptualized as social networks that shape newcomers’ professional identity as they try out many “possible selves” at work (Ibarra, 1999). Network characteristics such as the number and diversity of role models, the closeness of relationships, and the extent to which role models share salient characteristics with the newcomer can influence the adaptation process by affecting which possible selves one is likely to try (Ibarra, 1999).
Network structures may not be static. Interpersonal networks may be either redefined or shifted as newcomers adapt to their roles (Ibarra, Kilduff, & Tsai, 2005). In fact, there may be distinct phases in a new employee’s socialization and the organization’s attempts to mould socialization that lead to changes in network structure, composition, and nature over the early months of a person’s socialization. Morrison’s (2002) study provides a useful framework to examining newcomers’ information-seeking behavior through networks. In her study on newcomers’ information-seeking behavior, Morrison found that structural characteristics such as informational network size, density, range, and status have strong and continuous impacts on socialization as measured by gains in organizational knowledge, task mastery, and role clarity. An individual with large, sparse networks may be more successful in securing organizational knowledge (Morrison, 2002). By contrast, a person with a small, dense network of “strong” ties may be better able to acquire high levels of task mastery and role clarity (Morrison, 2002). Friendship networks may facilitate new employees’ social integration and organizational commitment and individuals with stronger personal networks may feel more integrated into the organization (Baldwin, Bedell, & Johnson, 1997; Wayne, Shore & Liden, 1997). As newcomers learn more about their job and organization, define their roles, and identify key organizational insiders, it is likely that the structure of newcomers’ networks do not remain static throughout socialization. Our first research question focuses on the nature of these structural changes over time.
Next we expand upon Morrison’s (2002) framework to explore how the composition of each type of social network changes as employees gain early-career experience. As an individual’s socialization progresses, they might have different relational needs. For instance, a person might seek more social supports from colleagues at an early stage of role entry, while she or he might want to develop her or his task competence later. Relations useful for one form of support, growth, and socialization may not be relevant or productive to another form. Emerging needs might also affect the structure and composition of new employees’ social networks within an organization. Our next two research questions address the dynamics of social relationships over time in the socialization process.
Multiplexity refers to the degree to which actors have overlapping relationships with particular alters in more than two domains of relationships, or “the degree to which network relationships are multidimensional” (Burt 1983; Ibarra, 1995, p. 675). For instance, a member of one’s role model network may be also a member of one’s task advice network. Multiplex ties are those with high levels of trust, but relationships take time to develop and cultivate and trust develops over time through repeated interactions (Ibarra, 1992). Previous studies suggest that multiplex ties, in contrast to “weak ties” (Granovetter, 1973) are significant in one’s organizational life because they are stronger, stable, reciprocal, and associated with high levels of trust and reliability (Brass, Butterfield, & Skaggs, 1998; Ibarra, 1992, 1995). Multiplexity and its role in newcomers’ socialization has received little attention in previous literature. As a first step, understanding multiplexity patterns from a purely structural perspective can generate a basis toward understanding the content of these exchanges. In our last research question, we explore the proportion of multiplex ties in each network and how they change over time.
Study Context 1
This article is based on data from a division of a large state audit agency which we have pseudonymed the “Department of Audit and Control” (DAC). 2 The division is divided into eight regional offices located throughout the state plus one central office. As a result of a legislative mandate that set a stringent deadline for completion of audits within one of its functions, the agency hired 90 new auditors. It also promoted several experienced auditors to supervisory positions. As of the beginning of our study period in the fall of 2006, the agency had hired 75 new auditors and promoted 22 employees to management positions. The new auditors were brought into the agency in cohorts, and each cohort included new auditors from multiple regional offices.
Newcomers learned about the organization and how to perform audits through multiple mechanisms instituted by the agency. First, each incoming cohort of new auditors attended a 2-week training session, the “Auditors Academy,” to learn specific auditing protocols, techniques, and tools. After completion of the 2-week training session, each cohort formed a community of practice. Although the communities of practice were formally instituted by the organization to continue the newcomers’ adaptation, each individual community had the autonomy to decide how often, where, and how (i.e., in person or via conference calls) it met. Each community also controlled the substantive content of its meetings; many meetings were devoted to additional training sessions in which senior level managers were invited to present as well as a round table sharing of audit and organizational experiences amongst the newcomers. Elsewhere we discuss the communities of practice and how the communities interacted with the newcomers’ egocentric networks (Hatmaker et al., 2011).
A third mechanism used to integrate the newcomers and help them to learn their job was on-the-job training. These efforts were a central component of the new auditors’ prolonged learning. Since the division was tasked with auditing entities throughout the state, employees spent the majority of their time in the field on projects. New auditors were paired with experienced auditors and supervisors on project teams. In this hands-on experience, they applied knowledge gained in training and learned how to manage relationships with the auditees. Through these interactions, tacit knowledge—the knowledge accumulated by auditors through their experiences that was not written in manuals—was also transferred to the new auditors from the more experienced organizational members.
To understand how knowledge was actually being transferred in light of these socialization activities, the agency contracted with the authors’ institution 3 to conduct a study of the new auditors’ learning and knowledge acquisition and the transition of the newly established managers. The agency was interested in understanding how the newcomers were actually learning beyond the formal training mechanisms to inform the socialization process for future cohorts.
Data and Method
The initial study population consisted of a sample 25 new auditors stratified across eight regional offices and one statewide office in the agency. Of the new auditors in the study, 13 are women and 12 are men. This article is based on a final sample size of 22 auditors for whom we have six waves of network data.
Initially the project team designed the study to collect four waves of semistructured interview data between October 2006 and December 2007. The authors added a social network questionnaire to the protocol to map the new auditors’ building of support relationships. We collected network data over six waves, and the authors also conducted an additional final (fifth) round of semistructured interviews in April and May 2010 to close out the data collection. We collected the first wave of network data via in-person interviews; we collected the remaining waves by phone. For the first wave, we conducted the network interviews after the first in-person semistructured interviews during October to December 2006. The final interviews asked participants to describe times when they sought advice or support from their contacts and whether they thought their network contacts would be important for their professional growth over the long term. This article reports findings based on the six waves of network data, collected in October to December 2006, March to April 2007, May to August 2007, January 2008, August 2008, and January to February 2009 supplemented by interview data.
The network questionnaire was designed to capture the participants’ egocentric networks across three instrumental networks and one expressive network. The choice of networks was based on previous network research (e.g., Ibarra, 1992, 1997; Morrison, 2002; Podolny & Baron, 1997). The three instrumental networks were chosen because they represent types of information that have been shown to be important for newcomers: (a) information about how to perform job-related tasks (task information and advice), (b) information about organizational goals, norms, structure, politics, and so forth. (organizational information), and (c) role expectations (role models; Morrison, 2002, 1993; Ostroff & Kozlowski, 1992). Expressive networks (friendship and social support) offer newcomers social support and a sense of belonging (Brass, 1984; Ibarra, 1992; Podolny & Baron, 1997), and research has indicated that this social integration is needed for effective socialization (Bauer et al., 1998; Morrison, 1993, 2002; Reichers, 1987). The appendix includes the name generator (Burt, 1984) questions asked for each of these networks. Respondents were asked questions to measure the intensity of ties and name interpreter questions (Burt, 1984) were asked to characterize the alters demographically (e.g., organizational position, how the participant met the alter, the alter’s region, and gender). To generate the sociomatrix for the network alters, participants were asked which alters knew each other. Waves 2 through 6 of data collection reviewed the list of alters and asked who should be removed or added to each network (and why).
To analyze the network data, we used Microsoft Excel and E-Net, a software package developed specifically to manage and analyze egocentric network data (Borgatti, 2006). E-Net produces visualizations of ego-nets and reshapes data for easier analysis in other packages. We generated a series of measures and summaries, including cross-tabulations of alter characteristics across each wave of data collected. We calculated the size of the networks as the average number of alters per ego across each wave of data collection for each of the four network types. As a measure of tie strength, we calculated the average frequency of communication across each wave for each of the four network types. We also calculated the proportion of alters in each network who were higher in the organizational hierarchy than the new employees. We examined the multiplexity of ties across the four different types of networks by calculating the proportion of alters in each network that are shared across at least one other network. We calculated these measures for each network at each wave of data collection to compare and track changes over time. We analyzed the interview data using open and axial coding (Corbin & Strauss, 2008); this article is based predominantly on the network data, with the interview data used to supplement and offer possible explanations for the network findings.
Findings and Discussion
Network Structure
To address the first research question, we examined the network visualizations across all six waves for each of the four network types—task information and advice (advice), friendship and social support (friendship), organizational information (information), and role models (role). Each type of network has characteristic ego-net structures and the prevalence of specific structures changes over time. After reviewing each of the graphs, we recognized that there were seven basic types of egocentric networks in the data. Figure 1 provides representative examples of all seven. We categorized networks that had a clear core-periphery structure as “star” networks. In these networks, the core consists of the ego and alters who are all connected to each other, with alters on the periphery that connect to the core but not to each other. “Hub” networks are those in which the study participant (the ego) is connected to each member in the network but the members are not connected to anyone except the ego. 4 Large complete and small complete networks are those in which all possible lines are present; that is, each actor is connected to all the other network actors, or, in network terms, all nodes are adjacent to each other (Wasserman & Faust, 1994). Small cliques are complete networks that contain only three or four nodes. 5 Tables 1 and 2 provide a cross-tabulation of the seven forms by the four network types from Waves 1 and 6.
Seven modal ego-net configurations.
Cross-Tabulation of Network Structures by Network Type in Wave 1.
Note: The first and second most prevalent forms are shaded in gray. The Hybrid column represents a count of both “Hybrid 1” and “Hybrid 2” forms shown in Figure 1.
Cross-Tabulation of Network Structures by Network Type in Wave 6.
Note: The first and second most prevalent forms are shaded in gray. The Hybrid column represents a count of both “Hybrid 1” and “Hybrid 2” forms shown in Figure 1.
The data tell an interesting story. First, all four networks have a primary ego-net structure and a secondary structure—although in both Waves 1 and 6 friendship networks are more evenly split between large complete, small complete, and other (the first and second most prevalent forms are shaded in gray in Tables 1 and 2). Conversely, there are some form-network pairs that do not occur. For instance, advice networks do not take either the small complete or “other” form, and the star network (with one exception) is found only in advice networks.
Second, two of the forms are relatively rare: hubs and hybrids. Hubs may be rare because it is far more common for alters to know one another, especially given that these networks are concentrated in a workplace where collaboration on projects is practically required. The prevalence—or actually, the lack thereof—of hybrid networks is not readily explained. We labeled graphs as hybrids when the network contained a well-connected alter that knows all of the others in the ego’s network, even though most of the alters are not known to each other. One might expect this form if the ego knows one experienced manager well and then has regionally dispersed alters—for instance, in their information or advice network. A priori, this form seemed likely to result from efforts by DAC to build regional collaboration and knowledge sharing structures, but our data did not find many examples of this form.
Third, two network types are completely omitted in one wave but exist in the other wave represented in Tables 1 and 2. In Wave 1, small cliques did not exist, but they do appear for each network type except for advice networks in Wave 6. And in Wave 6, star networks were essentially eradicated, but were the most prevalent advice network in Wave 1. In lieu of star networks, small complete networks were the most prevalent type of advice network.
Fourth, in Wave 6 the small complete network is the dominant structure across all four networks. In Wave 1, the primary ego-net structure is more diverse across the four networks; star, large complete, small complete, and other are all primary structures in the first wave. The small complete network seems to be the stable form by the end of six periods about two and one half to 3 years into the new auditors’ tenure. The prevalence of complete, closed networks is partly driven by network size: the smaller the network, the more likely all members will know one another.
Why are there characteristic network structures? Our preliminary hypothesis is that structure, network function, and time in role are related. The star structure, which is very prevalent in advice networks in Wave 1, has a core-periphery structure. Workers have a few key advisors that all know one another and a secondary set of advisors that know some but not all of the ego’s core advisors. The periphery may contain people that have specialized knowledge that can be called upon with the help of one or more core advisors. However, by Wave 6, small complete networks are the prevalent structure. The division has recently experienced a great deal of retirements of principals and senior managers, resulting in a considerable loss of expertise. In the later waves, there simply may be fewer individuals with specialized knowledge available to the newcomers. As a result, some network alters higher in the organizational hierarchy are dropping out of subsequent waves.
The development of “complete components,” both large and small, in the information network may be the results of early efforts by auditors to get to know everyone in their region or area of work. Instead of focusing on a few key informants in the first few months, new auditors start by trying to interact with as many people as possible. This early strategy may make sense, insofar as it takes time for new members of the workforce to determine who the key information brokers are. Over time, as alters are dropped from the network large complete networks become less prevalent and are replaced by small complete and small cliques.
The prevalence of small complete and “other” forms in role model networks in Wave 1 and of small complete and small clique forms in Wave 6 suggests primarily that role model networks are substantially smaller. As shown in Figure 2, in Wave 1, the average role model network has about 3 members, whereas information and friendship networks have about 4 members and advice networks average over 5.0 members per ego. As of Wave 6, the role model network has on average about 2.5 members, while friendship and information networks are at 3.75 and 4 members respectively, and information networks average 4.5 members.
Alter count by network type waves 1 through 6.
In general, the overall pattern of “smoothing out” of network forms toward the small complete form may be influenced by both organizational and individual factors. As newcomers rotate through different audit projects and work with different teams, they may be able to be more discriminating about who to add into and remove from their networks, resulting in smaller, more tightly connected networks. As their relationships with coworkers become broader by knowing more people and deeper as they become better acquainted with individuals. Friendship networks seem to be a bit of an exception, as the number of small complete networks in both waves is almost tied with the number of large complete and other networks.
Network Composition
To respond to the second and third research questions, we analyzed the composition of the auditors’ networks across different dimensions. First, we examined the total number of alters in each network. The number of alters changes from Wave 1 to Wave 6 dependent on the type of network. Figure 2 depicts the overall average alter count by network type across each wave.
This data suggests that most networks are contracting or staying roughly the same size. In general, these results suggest that as the new auditors settle into their jobs, they begin to sort out the value of various connections and “prune” their network accordingly, much as Burt’s (1992) work on structural holes would suggest should happen. These findings are supportive of the supposition that new workers take a scatter-shot approach to relationship formation in the early days of their tenure and then prune their networks to a size that is more manageable. Early on, newcomers wish to reduce uncertainty and are actively engaged in sensemaking and as such they seek feedback from varied organizational insiders and are more likely to develop larger networks (Fang, Duffy, & Shaw, 2011; Miller & Jablin, 1991; Morrison, 1993; Ostroff & Kozlowski, 1992). Over time, their information needs change. Some participants indicated that they remove alters because they do not need the alters’ advice or information any longer, which may be an indicator of achieving a certain level of job mastery or comfort. The decrease in network size is also linked to attrition within the organization, particularly with respect to retirements. During the course of this project, several higher level experienced employees retired, taking with them a great deal of tacit knowledge. Of the higher level alters that were named by participants in the early waves of data collection, several participants indicated that these individuals were removed due to retirements.
The interview data indicates that alters were also removed due to the realities of the auditors’ project-based work. Participants indicated that some alters, particularly supervisors and principals, were removed over time because they were no longer working together on a project. The interview data indicated that the newcomers hold some of these removed ties “in reserve” and would reactivate them later if needed for a certain task or project. These comments suggest that these ties are in reality not completely broken, but are latent—that is, the ties are simply dormant until the auditor finds them to be of use again (Hatmaker et al., 2011). Thus these relationships may only be temporarily relinquished.
Next, we analyzed the salary grade of the alters reported by each auditor to see whether the salary grade was systematically related to the grade of the ego or the type of network. For this analysis, we calculated the proportion of alters by salary grade for each network across each wave of data collection. There are six different salary grades within this division, resulting in five different job titles (listed in order of increasing authority and experience): auditor trainee (2 grades exist at this level), auditor, supervisor, principal, and division manager. Most new auditors were hired at the auditor trainee level, and were promoted to the auditor level at the end of their probationary period. We also included an “other” category that represented alters named by participants who were from outside the division or organization and as such did not have a corresponding grade. These individuals include, for example, family members, past employees of the agency, and individuals from other divisions within the agency that the new auditor knew from another context. Figure 3 shows the proportion of network alters at the supervisory level or higher level for each wave.
Proportion of alters at supervisory or higher level.
There are several interesting patterns in this data. Role model alters are consistently the most senior—on average, three fourths of role model alters are at the supervisory or higher level. This finding aligns with our a priori expectation that role models would be the most senior, under the assumption that egos seek role models that have experienced success as evidenced by promotion to higher titles/grades. Friendship networks also tend, on average, to include alters that are more senior than the egos; on average nearly a quarter of the alters in the friendship networks are at a supervisory or higher level. Our findings suggest that all networks are built with connections to relatively senior people. In the interviews, many participants indicated that their connections to senior personnel would remain helpful to them as their career progresses within the division. Given that the first wave of this data is for egos with less than a year of experience, it seems that the new auditors have had excellent social access to more senior and experienced people in DAC during the first part of their socializations. How did that occur? One possible explanation is that periodic meetings held by some regional chiefs helped to integrate new auditors, to connect them with senior level individuals, and to make them more comfortable interacting with more experienced and senior personnel. Another explanation is the nature of the work environment. Project teams often consist of a supervisor who leads the team along with experienced and new auditors, and members of project teams rotate across different projects so that new employees have the opportunity to work with a variety of supervisors and principals. Throughout the interviews, participants indicated that their project work enabled them to interact frequently with senior personnel and to learn directly from them.
Multiplexity
To address research question four, we analyzed the multiplexity of ties, the degree to which egos have more than one relationship with alters. Our findings indicate that newcomers develop multiplex relationships early on in their tenure and maintain multiplex ties over time. Figure 4 indicates the proportion of multiplex ties within each network type. Across all six waves of data collection, the majority of network members are also members of at least one other network. The multiplexity within the friendship networks likely reflects the senior personnel included in their network, as well as some alters from the newcomers’ communities of practice, who offered social support as well as some task-related information (see Hatmaker et al., 2011).
Multiplex ties by network for waves 1 through 6.
It is not entirely surprising to find multiplexity across the instrumental networks. Having useful multiplex ties may have been more attractive to the new auditors who were eager to acquire required job skills as quickly as possible. Individuals who can offer one type of instrumental resource are likely also sources of other valued and useful information. For example, one’s supervisor on a project is a resource for job task-related advice, and is also likely to be a role model as newcomers observe their supervisor to learn appropriate ways of problem-solving, decision-making, and communicating with outside organizations being audited. The new employees spent a great deal of time with their supervisors and experienced employees at job sites, so they had ample opportunity to learn what information these individuals can provide and capitalize on those relationships.
Implications for Public Managers and Newcomers
By detailing the structure and composition of social networks built by newcomers, and how this changes over time, our findings lend insight into the types of connections that newcomers develop through their socialization period. These patterns offer several implications for newcomers’ socialization and considerations for both public managers and newcomers. We begin by generally summarizing our findings in this paragraph and then move on to discuss implications for newcomers and public managers related to specific, more detailed aspects of the findings. First, our findings demonstrate that newcomers’ networks assume different structural forms dependent on the type of information or resource being sought, and these forms change as newcomers spend more time in their role. Over time the most prevalent network structure across all network types is the small complete form, a tightly connected network. Second, as their time in the organization increases, newcomers connect with fewer individuals for instrumental support. Across all time periods, role model networks have the fewest network alters. Third, newcomers are connecting to relatively senior people early on and maintaining connections with higher level individuals throughout their tenure. Fourth, over time the majority of network members in the instrumental networks are also members of at least one other network, indicating that newcomers connect with individuals who can provide multiple types of support.
Before we begin our discussion, we wish to first note that DAC and the state in which it is located are somewhat unique. The state government is larger than the majority of state governments and DAC is a particularly large agency. Size permits DAC to hire more people and more thoroughly trained people than may be the norm in other situations, making this infusion of personnel unusual. DAC also instituted formal collective socialization tactics and brought newcomers on in cohorts, which may have influenced their network building. The auditors’ work is technical and knowledge-based, and they work primarily offsite from their regional office in project teams. Our repeated data collections may have influenced the subjects’ perception of their own social networks, affecting all data collected after Wave 1.
With these caveats firmly in mind, the new auditors appear to be developing networks that can be advantageous for becoming productive organizational members who fit in well. Since their work is largely executed in project teams in the field, small cohesive instrumental networks are the most likely form to offer ready access to the key task information and advice resources needed to become competent and effective (Morrison, 2002). Mastering their job tasks and clearly understanding their organizational roles may be enhanced by their dense networks (Morrison, 2002).
Their complete, cohesive networks may also be advantageous because they may offer an opportunity for trust to grow, the communication of norms that guide behavior, and a sense of belonging (Coleman, 1990; Podolny & Baron, 1997). For newcomers, these networks can also be instrumental in communicating clear and consistent role expectations, and a clear understanding of what is expected, can positively influence newcomers’ effectiveness (Morrison, 2002; Podolny and Baron, 1997, pp. 674-675). Through such networks, support can be garnered from organizational insiders who provide access to important resources such as sensitive information and whose “buy-in” can facilitate career mobility within the organization (Podolny & Baron, 1997). And by “conveying support, identity, and acceptance, friendship networks made up of strong and interconnected ties will facilitate the development of social integration and organizational commitment” (Morrison, 2002, p. 1151).
Yet complete networks may have some drawbacks in comparison to sparser networks, particularly with respect to long-term career considerations. Large, sparse networks filled with structural holes (positions that connect individuals who are otherwise unconnected) offer a more diverse array of information and resources than complete networks (Burt, 1992). Complete networks often consist of redundant ties; that is, the ego is connected to multiple alters who provide the same, rather than diverse, information, and resources. This redundancy can limit the array of information and support available to an individual, and result in inefficiencies as the ego must expend time and energy to maintain ties which do not necessarily yield different resources. Sparser networks may also contribute more toward career mobility within the organization. Burt (1992) found that networks containing structural holes offer greater diversity of resources and positively impact career advancement, although Podolny and Baron (1997) offer a more granular view in which both complete cohesive networks and sparse networks offer advantages for mobility and performance. In their findings, structural holes in strategic information networks can facilitate mobility, while dense networks of individuals who support an actor in his or her initiatives better facilitates performance (which in turn can support mobility; Podolny & Baron, 1997).
Our network composition and multiplexity results suggest that the newcomers’ instrumental networks are focused around key personnel for each ego—a core set of advisors. Senior organizational members can provide access to a multitude of resources such as tacit, experience-based job knowledge, “insider” information about projects and the organization, long-term career advice, and access to other organizational members who may be instrumental for promotion or advancement. Having supervisors in one’s network contributes to job and role learning and the development of social capital (Fang et al., 2011; Morrison, 2002; Podolny & Baron, 1997, p. 1156). The newcomers’ networks also offer them access to a key resource possessed by influential organizational insiders: power. Newcomers include in their networks individuals who are in authority-positions within the organizational hierarchy, who have technical expertise, and who likely understand the inner workings of the division—individuals who are likely to have power and influence within the organization (Krackhardt, 1990).
Although here we do not examine social capital explicitly, the new auditors’ networks also have implications for the development of social capital that can facilitate job performance, career mobility, and newcomer learning (Burt, 1992; Podolny & Baron, 1997; Fang et al., 2011). As defined by Lin (1999, p. 35, orginal author’s emphasis) social capital consists of “resources embedded in a social structure which are accessed or mobilized in purposive actions.” The newcomers’ network structures and the resources embedded in those structures represent the two key aspects of social capital (Fang et al., 2011). By developing ties early on with higher level organizational members, the new auditors may be accruing social capital from the onset of their tenure. Yet the cohesive networks they appear to be developing may offer limited longer term social capital. The social capital literature consists of competing views as to the optimal network structure: tight, cohesive networks versus networks comprised of structural holes, networks in which the ego has many direct ties, but ties between the alters are sparse (few indirect ties) (Borgatti & Foster, 2003; Podolny & Baron, 1997; Provan & LeMaire, 2012). While the new auditors have done well in developing connected networks, as of our final data collection they had not developed networks that include structural holes.
The new auditors appear to be developing networks that can lead to a sense of belonging, task mastery, and understanding their roles, all important indicators of socialization. Yet for longer term career benefits, they likely need to connect with other higher level individuals who are not as tied to their existing network members. While maintaining a cohesive core offers benefits, a more optimal structure may be a hybrid network that also contains more indirect ties (Podolny & Baron, 1997). However, the newcomers in our study may be somewhat constrained by their organizational configuration and the nature of their work in forming new ties (Podolny & Baron, 1997). The division under study is spread across nine regional offices throughout the state, and much of their work is field-based project work, so that they are interacting primarily with individuals on their project teams within their region. Opportunities to connect with higher level organizational members who are not connected and who offer access to diverse resources do not occur frequently within this organization. 6
Our findings indicate that newcomers prune ties over time based on their project and task needs and they may reactivate those ties as needed. However, our advice to newcomers is that they should give careful consideration to the possible repercussions of pruning. Newcomers may experience negative reputational effects by dropping ties they no longer perceive as valuable (Podolny & Baron, 1997, p. 691). Dropping ties may be viewed by others, especially by those who have been dropped, as a cold and calculated instrumental move on the part of the individual. People may be less willing to establish ties with newcomers who are known to break relationships (Podolny & Baron, 1997) or reactivation of a tie that has been dropped may not be as easy as a newcomer may believe. The prevalence of multiplex ties can also make dropping alters problematic. Multiplex ties deliver more resources efficiently since the costs of maintaining one tie are lower than maintaining multiple ties (Lazega & Pattison, 1999). Yet at the same time they are a source of constraint because if the newcomer wishes to drop one aspect of the relationship, she or he may also risk losing the relationship altogether, along with other types of information and support (Podolny & Baron, 1997). Newcomers should consider ways to maintain or preserve connections no longer needed in the immediate sense to protect their reputation and to easily call upon those relationships in the future (Podolny & Baron, 1997).
Our findings also lead to several items of note for public managers. The first relates to how organizations can foster social interaction opportunities (Moynihan & Pandey, 2007, p. 221), especially for newcomers. Managers should consider using institutionalized socialization tactics in which newcomers are exposed to senior personnel across different departments or offices. Such tactics may assist in helping newcomers develop ties with more insiders who do not know each other directly, thus increasing the structural holes in their networks and the diversity of information available to them. More generally, as recommended by Moynihan and Pandey (2007, p. 221), organizational leaders can consider mechanisms to foster informal social networks within organizations, including mentoring programs, special events such as “lunch and learn” sessions or celebrations of group or individual successes, informal gatherings, and office setups that encourage employee interaction. For organizations that consist of regionally dispersed offices or divisions, holding periodic “all hands” meetings that rotate through different geographies can aid in expanding informal networks. Such exposure offers newcomers more opportunities to diversify their networks.
Public managers should consider establishing a formal mentoring program in addition to venues that facilitate informal interactions between new and experienced employees. Mentors have been shown to be a key part of the socialization process and can lead to increased organizational knowledge (Bauer & Erdogan, 2011; Chao, 2007; Ostroff & Kozlowski, 1992). However, the traditional model of a one-to-one mentor–protégé relationship may not be sufficient for newcomer integration, especially in complex work environments. More recent conceptualizations of mentoring focus on “developmental networks” in which multiple mentors, often through informal mentoring relationships, provide different types of resources, and support (Higgins & Kram, 2001; Kram, 1985). This conceptualization more aptly applies to the networks of organizational newcomers. While formal mentoring programs may facilitate the traditional one-to-one mentoring relationship, developmental networks represent mentoring that occurs more “organically”—that is, without a formal mentoring program’s structure (Bozeman & Feeney, 2007, 2008) and thus may be viewed an informal mentoring. By offering both formal mentoring as well as opportunities for newcomers to establish informal mentoring relationships with experienced organizational insiders, public managers can assist new employees in establishing developmental networks that offer a variety of support and resources.
Another item for public managers to consider is whether core advisors are recognized or rewarded for their activity. The qualitative data leads us to believe these core advisors play a central role in helping the new auditors learn the cultural and social aspects of their new job. Formally assigned mentors may get recognition, but it is not clear that those who emerge organically in networks are equally appreciated. If our conclusion is correct and if, as we further suspect, a relatively limited number of employees fall into these core advisor roles, then DAC and other agencies may wish to identify and reward these activities proactively so that they are continued. It also raises questions regarding the promotion or geographic reassignment of such personnel. Without efforts to identify these core advisors, the organization may inadvertently reassign those who play a critical, if informal, role in integrating new employees.
Conclusion
Understanding the dynamics of socialization is useful to public administrators at least in part due to the impending retirements faced by many government agencies (Perry, 2010). Effective socialization of new employees helps ensure an organization’s survival and continued performance. As Ashforth, Sluss, and Harrison (2006, p. 2) note, “Effective socialization helps transform the newcomer into a contributing member, thereby replenishing if not rejuvenating the organization as a system.” Since social networks can play a central part in shaping newcomers and transferring knowledge, an organization needs “strategies” more than just sets of socialization tactics. Management needs to understand the possible dynamics of this process to leverage, direct, or even control it. Management becomes even more critical in the context of large-scale infusions of personnel, when faulty socialization may impair the assimilation of entire “generations” of new workers.
For the field of public administration, understanding how to successfully integrate employees can assist in developing high performing teams and effective and efficient organizations. Our analysis offers insight into newcomers’ social networks in a public sector organization and how they change over time as the newcomers become acclimated. Our findings contribute to the fledgling literature on socialization and intraorganizational networks in public sector organizations. To offer public administration scholars and practitioners more insight into new employee integration and adaptation, we offer several suggestions for future research. First, understanding more about how the development of social capital via networks influences newcomers’ advancement, knowledge acquisition, and fit within the organization would give insight into the relationship between newcomers’ networks and socialization outcomes (Fang et al., 2011). Second, this infusion of newcomers was somewhat unique in both size and the formal efforts that DAC undertook for socialization; to understand how newcomers develop networks in different hiring and training circumstances, future research should examine cases in which newcomers are not brought onboard in cohorts or where they do not undergo institutionalized or collective socialization activities. Finally, future studies should also focus on different organizational structures and different types of occupations and work. The work of the auditors’ in this agency is heavily technical and knowledge-based, and much of their learning happens on the job while working on projects. We cannot answer the question of whether the socialization and networking of newcomers in less knowledge-based occupations follows similar patterns. Different organizational contexts, such as those organizations in which the work is not primarily executed in field-project based teams, may also reveal differences in network development and socialization. A broad array of future research will further illuminate public sector new employee network development and the socialization process.
Footnotes
Appendix: Egocentric Name Generator Questions
Acknowledgements
We would like to thank Sydney Cresswell and the Program on Local and Intergovernmental Studies at the University at Albany, State University of New York for support in the data collection for this project and Dr. Karl Rethemeyer for input on an earlier draft of this article and for his contribution to the initial data collation. We would also like to thank the anonymous reviewers for their thoughtful comments.
Author’s Note
Not for citation without authors’ prior permission.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.