Ecological Influences on the Formation of the Hiring Network in the Communication Job Market,2015 to 2019

Abstract

This study investigates ecological factors that drive hiring decisions in the academic marketplace. Faculty hires between institutions are conceptualized as interorganizational network ties. Drawing on theoretical insights from network inertia and niche processes in organizational ecology, the current study builds an ecological framework to explain the formation mechanisms of the faculty hiring network among 81 U.S. institutions granting PhDs in communication. Consistent with the predictions of the ecological model of hiring decisions, the empirical results of exponential random graph models (ERGMs) revealed that past behavior (or the presence of previous ties), niche width (or the number of research specializations), and niche overlap (or the degree of shared research specializations) significantly constrained the likelihood of tie creation during the 2015 to 2019 period. These effects held true even when traditional explanations such as network self-organization and status-based hiring patterns were taken into account. Theoretical and practical implications are discussed.

Keywords

tie formation organizational ecology organizational networks network analysis faculty hiring

Human resources are one of the most valuable assets of organizations. The recruitment of high-quality employees is especially important to industries in which organizational outcomes depend heavily on employees’ direct communication with stakeholders (Van den Bulte & Wuyts, 2007). This is also evident in the higher education system where faculty hiring decisions generate implications for an institution’s scholarly quality, perceived status and prestige, and possessed social and economic capital (Burris, 2004).

Employee recruitment has received considerable scholarly attention over the past few decades. The predominant perspectives on hiring decisions have emphasized the importance of communication channels and networks (Mai et al., 2015), organizational characteristics (Barnett et al., 2010), individual achievements (Headworth & Freese, 2016), and trust and support from common connections (Jackson, 2019). Although the existing literature has advanced our understanding of organizational behavior in labor markets, what remains unaddressed is how ecological factors constrain decision-making. In this sense, theories of organizational ecology can add to this stream of research by clarifying how organizational responses to environmental conditions are manifested in recruitment patterns. Despite the prevalence of ecological interdependence in organizational communities (Aldrich et al., 2020), few attempts have been made to explicitly examine ecological influences on hiring decisions.

The purpose of this study is to introduce ecological theories (i.e., structural inertia theory and niche theory) as conceptual frameworks for understanding the hiring decisions of PhD-granting institutions in communication. Faculty hires between institutions are conceptualized as directed network ties at the interorganizational level. The ecological model of hiring decisions proposes that inertial forces and niche processes significantly shape faculty hiring patterns. In doing so, the current study contributes to developing a more nuanced view of the ecological foundations of network emergence.

The article is organized as follows. The first section reviews the literature on networks, communication, and the academic marketplace. Guided by structural inertia theory and niche theory in organizational ecology, this study proposes three hypotheses regarding the influences of past behavior, niche width, and niche overlap on the likelihood of tie creation in the hiring network. The second part introduces data collection, measurement, and analytical strategies. Following a social network analysis of the formation mechanisms of the faculty hiring network among 81 U.S. universities during the 2015 to 2019 period, the final part concludes with a discussion of the theoretical and practical implications of the empirical findings.

Literature Review

Networks, Communication, and Labor Markets

There is abundant literature at the intersection of networks, communication, and labor markets. The first line of research emphasizes the roles of social networks and communication processes in job search, acquisition, and retention. Granovetter’s (1973) classical strength-of-weak-ties theory posits that weak ties provide job seekers with more diverse, non-redundant, and novel information resources than strong ties, so an effective strategy to land good positions is to develop and maintain as many weak ties as possible. While the major findings have been replicated in many contexts, the theory has been criticized for overlooking the importance of referrals, recommendations, and intermediaries in shaping the hiring decision (Bian, 1997). The strategic use of indirect ties in the job market becomes even more salient in the contemporary media landscape. With the emergence and development of new technologies, individuals have been increasingly relying on digital platforms (e.g., Indeed) and social networking sites (e.g., LinkedIn) to directly access job information, reducing the importance of traditional interpersonal channels (Piercy & Lee, 2019). As a result, job seekers may be exposed to the same information online and then apply for the same positions. In this respect, strong referrals and recommendations from current employees or overlapping connections can help a job applicant to stand out from the pack by convincing an employer of the applicant’s reliability, potential performance, and other personal attributes (Jackson, 2019). Employers also benefit considerably from recruitment through social networks or word-of-mouth. Specifically, workers hired through referrals achieve better job performance and have lower turnover rates than do those hired through other channels (Pallais & Sands, 2016).

The second stream of studies treats talent flows between institutions in the labor market as interorganizational network ties. Up to now, much attention has been paid to faculty hires in the academic marketplace. As hiring results are accessible to the general public, hiring decisions in academia are communicative, reflecting collective assessments of educational outcomes of institutions (Clauset et al., 2015). Past research has also shown empirical evidence of status-based hierarchies in faculty hiring networks across a variety of disciplines (Gondal, 2018; Kawa et al., 2019). Burris (2004) observed that the academic marketplace was similar to a caste system in which elite schools or departments followed the principle of exclusivity and hired PhDs from those in the same status hierarchy. Additionally, high-status institutions established and maintained an extended network of partners by placing graduates in middle- or low-status institutions, facilitating the accumulation and reproduction of social capital over time (Burris, 2004). In the communication discipline, Barnett et al. (2010) found that a program’s ability to place graduates was positively correlated with its reputation. Mai et al. (2015) lent empirical support to self-organization mechanisms of reciprocity, transitivity, and cumulative advantage in tie-formation processes. When these structural effects were taken into account, the association between placement centrality and reputational rankings became small.

The current study adds to the two lines of literature by offering an ecological model of hiring decisions. While communication scholars have demonstrated the utility of the ecological perspective in explaining network evolution (e.g., Margolin et al., 2015; Monge & Contractor, 2003; Xu, 2021), little work has been conducted to investigate ecological constraints on networked hiring patterns. This paper also contributes to the dialogue between ecological and network perspectives on organizational behavior by shedding light on the ecological foundations of network emergence.

Structural Inertia and Network Inertia

Structural inertia theory in organizational ecology provides a useful baseline for understanding path-dependent features of organizational network ties. The theory states that organizations face strong inertia pressures and are thus resistant to change (Hannan & Freeman, 1984). Inertial forces typically increase with organizational age, size, and complexity. Members in older, larger, and more complex organizations have stronger tendencies to oppose changes to existing architectural features to which they have been exposed (Hannan & Freeman, 1984). Organizational ecology proposes that inertia is a by-product of rather than a precondition for natural selection (Hannan & Freeman, 1984). As an indicator of good management, structural inertia reflects the retention, duplication, and institutionalization of well-tuned organizational architectures that exploit existing competence and competitive advantage. Empirical evidence has also indicated that successful organizations tend to be inert in dynamic environments (Stieglitz et al., 2016). Psychological research complements this ecological view of organizational adaptation. For instance, Brown and Starkey (2000) identified organizational defense against anxiety as a driving force behind inertia (Brown & Starkey, 2000). According to their psychodynamic perspective, organizations act conservatively to preserve their existing self-concepts, which may impede the ability to engage in exploratory activities and critical self-reflexivity.

Drawing upon structural inertia theory, Kim et al. (2006) conceptualized “network inertia” as “a persistent organizational resistance to changing interorganizational network ties, or the difficulties an organization faces during network transformation” (p. 715). The network inertia perspective acknowledges internal and external constraints on network change, challenging the rationality assumption that organizations develop their network ties through simple cost-benefit calculations. Even if there is a need for change due to efficiency considerations, a focal organization may already be locked into network ties with current partners, making the dissolution of existing ties extremely difficult (Maurer & Ebers, 2006). Network inertia can also be considered as a result of successful experiences in the past (Kim et al., 2006). Transitions from old partners to new ones are full of uncertainty. Retaining existing interorganizational ties can reduce the likelihood that an organization will suffer from unpredictability and ambiguity. Interorganizational hiring networks are likely subject to inertial forces. For instance, one common hiring practice in academia is to interpret the prestige of an applicant’s PhD institution as a cue for individual performance. However, the accuracy of this quality heuristic is not necessarily ensured (Headworth & Freese, 2016). Empirical evidence showed that individual achievements (including prestigious publications and awards) did not well explain job placement and accounted for only a small portion of placement differences among institutions with varying levels of status (Headworth & Freese, 2016). The taken-for-granted practice of using graduate education prestige as a quality signal may be further compounded by unconscious biases of decision makers, thus reinforcing resistance to tie change in the faculty hiring network.

Communication scholars have employed structural inertia theory to understand organizational and community change. For instance, Weber (2017) developed a theory of organizational speciation to explain the emergence of new organizational forms. A high degree of inertia within organizational communities is considered a precondition for speciation processes. As established organizations are subject to strong inertial forces that resist fundamental change, they are less likely early adopters of new innovations. This slow reaction enables entrepreneurs to take advantage of unexplored opportunities to create new industries. Weber and Monge (2017) empirically examined the consequence of structural inertia for traditional newspapers facing environmental disruption and transformation. It was confirmed that adaptation to new digital technologies increased survival rates of inert organizations in the media industry.

This study complements the existing organizational, network, and communication literature on structural inertia by demonstrating the persistent nature of faculty hiring ties in the academic marketplace. In the present case, network inertia is manifested in replications of network ties over time. Suppose that university i placed a graduate in university j at time t. Driven by strong inertia, j would tend to hire another graduate from i in the subsequent period because it consolidates existing relation-specific assets (Kim et al., 2006). As the exploration-exploitation trade-off (March, 1991) suggests that consequences of experimenting with new alternatives are less well known than those of replicating past actions, hiring somebody from a university where there is no pre-existing recruitment tie can be risky. Previous hires create informal communication channels between universities (Mai et al., 2015) and facilitate the replications of faculty hiring ties through cumulative advantage (Burris, 2004), trust development (Jackson, 2019), and relational and cognitive lock-in (Maurer & Ebers, 2006). Hiring through established interorganizational networks also reduces uncertainty because search committee members can easily access inside information about personality and performance that is not contained in observable characteristics from an applicant’s curriculum vitae and cover letter. From an evolutionary perspective, maintenance of positively selected activities allows “organizations to capture value from existing routines that have proved – or have been perceived as – beneficial” (Aldrich et al., 2020, p. 25). This suggests that inertial forces are especially strong if faculty members hired from a peer institution have achieved career success at the current institution. Hiring schools or departments thus infer that graduates from the same institution can also thrive in the working environment. It is predicted that organizations that have formed a faculty hiring tie in the past will replicate the tie in the future.

Hypothesis 1: The presence of previous ties will increase the likelihood of subsequent tie formation in the hiring network.

Niche and Niche Width

Initially borrowed from biological evolution, the concept of “niche” offers a resource-based view of organizational interdependence and outcomes. An ecological niche is a multidimensional resource space that supports the survival of organizations (McPherson, 1983). Niches have been defined in terms of gratification opportunities (Dimmick, 2003), mission and goals (Shumate & Lipp, 2008), and functional or spatial dimensions (Lee & Monge, 2011). As an organizational attribute, niche width is the number of niches an organization inhabits out of the number of niches that exist (Hannan et al., 2003). Organizations having high (vs. low) niche width are classified as generalists (vs. specialists). Niche theory in organizational ecology holds that generalist organizations possess and maintain excess capacity that allows them to perform better than specialist organizations in uncertain environments (Hannan & Freeman, 1989).

Communication scholars have developed and refined niche theory by highlighting the importance of niche width in driving the evolution of media species and organizational communication networks. For instance, Scolari (2013) argued that a medium in danger of extinction can still survive fierce competition with other media forms through niche-width strategy. Shumate and Lipp (2008) confirmed that niche width, operationalized as the heterogeneity of organizational goals, was associated with an increased likelihood that an NGO would become an authority, broker, or initiator in interorganizational hyperlink networks. Lee and Monge (2011) found that organizations with a broad geographic scope had a greater chance of building multiple linkages with the same partner than did those targeting narrow niches.

Despite repeated attempts to theorize communication dynamics through a niche perspective, the identity-based view of niche processes in organizational ecology has rarely informed communication research (see Yang, 2020, for a notable exception). This theoretical approach portrays organizational identity as grades of memberships in socially constructed categories (Hsu & Hannan, 2005). Organization-category affiliations capture the duality of social systems—that is, an organization (or category) is constituted by the set of categories (or organizations) to which they belong. The notion of duality suggests that an organization’s affiliations signal its reference points in social space (Breiger, 1974), so niches can be reflected by organizational affiliations with institutionalized categories (Hsu, 2006). Accordingly, niche width is the count of categories assigned to a focal organization. Consistent with a sociological perspective on organizational identity (Gondal, 2018; Vergne & Wry, 2014), this study relies on self-reported category memberships to determine organizational niche width.

In academia, sub-disciplinary specializations, defined as “the self-identified areas of research expertise in which departments hire faculty, train graduate students, and conduct research” (Gondal, 2018, p. 202), are a major form of institutionalized categories that serve as collective identity codes. Institutions rely on such codes to make strategic identity claims in the disciplinary space. As self-claimed niche broadness signals a complex or diffuse organizational identity (Vergne & Wry, 2014), generalist organizations face weaker constraints on the range of legitimate activities they can pursue than do specialist organizations (Hannan et al., 2003).

Atypical hiring is a source of uncertainty for academic institutions. While the communication field has a strong multidisciplinary tradition, only a small proportion of communication faculty members earned their PhDs in other disciplines (Barnett et al., 2010; Mai et al., 2015). This suggests that not every institution has developed a preference for hiring practices that defy institutionalized disciplinary boundaries. Hiring outside of the communication field can be risky because such candidates are expected to share few overlapping contacts with a focal hiring institution and may be unwilling to engage with the community of communication scholars. According to niche theory (Hannan & Freeman, 1989), high niche width provides generalists with extra capacity that helps them buffer against uncertainty. As a result, generalist organizations tend to deem atypical practices as less risky. This ecological interpretation is also in line with structural hole theory (Burt, 2004), which emphasizes that social brokers are aware of diverse options and adopt alternative ways of thinking. In the present case, a niche width of N is translated into N (N – 1)/2 possible pairs of research specializations offered by a PhD-granting university. For an increase in niche width from N to N + 1, the net increase in the number of possible pairs is N. Therefore, generalist institutions with memberships in multiple research specializations are more likely to access diverse interpretations of communication scholarship, resulting in a high tolerance for atypical hiring practices. Thus, this study predicts that generalist institutions will have fewer conventional hires than specialist institutions.

Hypothesis 2: An increased niche width will lead to a decreased likelihood that an organization will receive ties in the hiring network.

Niche Overlap

As another important component in niche theory in organizational ecology, niche overlap is “the fraction of the focal organization’s niche covered by the niche of the other” (Hannan et al., 2003, p. 325). The identity-based view of niche processes defines organizational niches in terms of category memberships (Hsu, 2006). As the degree of shared identity between organizations is a function of their category overlaps (Breiger, 1974), pairwise niche overlap can be reflected by co-memberships in socially constructed categories. Niche overlap is not necessarily symmetric. It is possible that the niche of organization i is completely occupied by that of organization j, but not vice versa (McPherson, 1983).

Previous communication research has generally conceived of niche overlap as a proxy for the degree of competition between entities within a common resource space. For instance, a new medium is expected to produce a competitive displacement effect on an old medium if the former is functionally superior to the latter in the overlapping niche area (Dimmick, 2003). The organizational speciation model (Weber, 2017) states that the growth of new forms of organizations requires additional environmental space. As a result, new industries need to increase their niche overlap with established ones to battle for the same set of resources. While the communication literature largely treats niche overlap as an indicator of competitive intensity, ecological research has acknowledged that two entities in overlapping niches can partially or fully benefit from each other even in the presence of competition (Aldrich et al., 2020). Another issue is that competitive dynamics are not necessarily driven by niche overlap at a single level. For instance, organizations exploiting the same type of human resources but not the same pool of time are not direct competitors (Shi et al., 2017).

Going beyond the conventional view relating niche overlap to competition, this study argues that the degree of mutualistic interdependence between institutions also increases with niche overlap in socially constructed categories. In the present case, faculty hiring ties in the academic marketplace constitute a mechanism for transferring and exchanging economic capital (Burris, 2004). PhD students can be trained for employment beyond academia (Kawa et al., 2019), but a discipline’s sustainable development still depends on a stable and sufficient demand for faculty positions. In this sense, those occupying the same niche areas can support each other through hiring practices. Job seekers are likely to apply for and acquire positions in institutions that offer and value the same set of research expertise. Recently, Gondal (2018) empirically investigated the impact of overlapping memberships in research specializations on the formation of the faculty hiring network among sociology departments. The findings demonstrated that institutions without shared interests were unlikely to form a network tie. The above discussion leads to the following hypothesis:

Hypothesis 3: An increased niche overlap will lead to an increased likelihood of tie formation between organizations in the hiring network.

Method

Data Collection

Data were collected through a whole-network approach. The faculty hiring network consists of U.S. universities that offer PhD programs in communication. Nodes are universities, and ties indicate the flow of faculty hires between pairs of universities during the 2015 to 2019 period. Following a similar procedure adopted by Barnett et al. (2010) and Mai et al. (2015), this study utilized the NCA Doctoral Program Guide as a baseline for constructing the network boundary. As of the end of May 2020, NCA had listed 87 programs from 81 universities. However, many PhD programs in media, journalism, mass communication, or related fields offered by the same set of universities in the directory were still missing. Interestingly, even NCA (2019) itself acknowledges that the broad classification of communication covers these subareas. Thus, a methodological concern is that failure to include these programs may result in biased estimates. Suppose that a university had two separate doctoral programs in communication. One was on the NCA list; the other was not. Hiring practices from the university would be underrepresented if faculty members from the latter program were excluded from analysis. In view of this, the current study supplemented the existing NCA list, leading to a total of 108 PhD programs from the same 81 universities.

The next step was to collect faculty information from each university. The qualified tenured or tenure-track faculty members were affiliated with at least one of the 108 PhD programs in communication and did not enter retirement at the time of data collection (March 2020). Those without doctoral degrees were excluded. This procedure led to a total of 2,192 cases. The coded information included (a) name, (b) current institution of employment, (c) years of employment, (d) rank (full, associate, or assistant professor), (e) institution where the doctoral degree was earned, (f) field of doctoral study, and (g) year of graduation. For those who changed positions in their academic careers, (h) previous institution(s) of employment, (i) the start year of employment, and (j) the end year of employment were recorded. The above information was retrieved from faculty profiles and CVs on school, department, or personal websites, LinkedIn profiles, institutional affiliations reported in journal articles or books, and the ProQuest Dissertations & Theses Global database. In the dataset, 77.83% (1,706 out of 2,192) of faculty members received their PhDs in the broadly defined communication field, the vast majority of whom (1,664 out of 1,706, 97.54%) completed communication doctoral programs offered by the universities on the NCA list.

Network Construction

The observed network contains faculty hiring ties among 81 U.S. universities from 2015 to 2019. The start year of a position was the criterion to determine when a specific hiring event occurred. Year 2015 was chosen as the start of the observation period because Barnett et al. (2010) and Mai et al. (2015) had already examined hiring patterns in communication before 2015. The 81 universities hired 630 people and filled 647 tenure-track or tenured faculty positions. Each university in the sample hired at least one faculty member in the observation window.

There were 508 hiring events (508 out of 647, 78.52%) within the network boundary. A total of 121 faculty positions (121 out of 647, 18.70%) were filled by PhDs trained in other disciplines such as political science (n = 14), sociology (n = 13), psychology (n = 12), library and information science (n = 11), English and literature (n = 9), education (n = 6), and computer science (n = 5). In the remaining cases (18 out of 647, 2.78%), employees obtained PhDs in communication from institutions outside of the network boundary (e.g., Columbia University, the University of South Carolina, and the University of Amsterdam).

The 2015 to 2019 hiring network is directed. If university j hired a graduate from university i, a directed tie from i to j was created (i → j). An 81 × 81 valued network matrix was then constructed. Values of off-diagonal cells, S_ij (i ≠ j), indicated the number of i’s graduates placed in j during the five-year period. There were 475 hiring events between i and j (i ≠ j) in the network. The number of unique faculty hiring ties (or the count of non-zero off-diagonal cells) was 399. A non-zero diagonal cell, S_ij (i = j), in the matrix (or a self-loop in the network) indicated that a focal university hired at least one of its own graduates. Twenty-one universities, including Michigan State University (n = 5) and the University of Texas at Austin (n = 4), engaged in self-hires and filled 33 faculty positions. Figure 1 visualizes the faculty hiring network (excluding self-loops) among 81 U.S. universities granting PhDs in communication. Table 1 presents a top 20 list of in-degree and out-degree centrality in the hiring network. Descriptive statistics revealed that the top 20 universities (see the column of weighted out-degree score) were responsible for 69.05% (328 out of 475) of all placements (excluding self-hires) within the network boundary. A university’s hiring capacity (represented by weighted in-degree score) was moderately correlated with its placement capacity (represented by weighted out-degree score; rs(79) = .41, p < .001). The Gini coefficient, ranging from 0 (complete equality) to 1 (complete inequality), was also calculated to capture skewness in degree-based distributions. A low coefficient indicates that universities have an equal chance of placing (or hiring) graduates in (or from) peer institutions. The observed network had a Gini coefficient of .38, .57, .41, and .60 for the distributions of in-degree, out-degree, weighted in-degree, and weighted out-degree, respectively.

Figure 1.

A visualization of the faculty hiring network among 81 U.S. universities, 2015 to 2019.

Table 1.

Top 20 Universities in the Hiring Network Among 81 U.S. Universities, 2015 to 2019.

University	In-degree score	In-degree rank	Faculty hired (weighted in-degree score)	University	Out-degree score	Out-degree rank	Faculty placed (weighted out-degree score)
Alabama	18	1	27	Texas-Austin	25	1	33
Penn State	12	2	14	Missouri	19	2	22
Texas Tech	12	2	12	Wisconsin-Madison	18	3	26
Florida	11	4	15	Indiana	17	4	18
Kentucky	10	5	13	Southern California	15	5	22
Temple	10	5	11	North Carolina-Chapel Hill	15	5	19
Georgia	10	5	10	Pennsylvania	15	5	17
Colorado	9	8	16	Purdue	14	8	18
Minnesota	9	8	12	Northwestern	14	8	15
Indiana	9	8	11	Penn State	13	10	19
Louisiana State	9	8	11	Cornell	13	10	13
Texas-Austin	9	8	11	Michigan State	12	12	16
Kent State	9	8	9	Ohio State	12	12	16
Michigan State	9	8	9	Illinois Urbana-Champaign	10	14	12
Arizona State	8	15	10	Minnesota	10	14	12
Ohio State	8	15	10	Maryland	10	14	11
Oregon	8	15	10	Georgia	9	17	13
Purdue	8	15	9	California-Santa Barbara	8	18	9
Utah	8	15	9	Oregon	8	18	9
Illinois Urbana-Champaign	7	20	12	Florida	8	18	8
Colorado State	7	20	9
Missouri	7	20	8
Tennessee	7	20	8

Measures

Dependent variable

Faculty hiring ties during the 2015 to 2019 period. This study treated the 2015 to 2019 hiring network (excluding self-loops) as the dependent variable (average degree = 4.93, network density = .06, average clustering coefficient = .14, average path length = 2.98). Social network analysis was conducted to predict the likelihood of tie formation.

Independent variables

Previous ties. The current study followed the same procedure outlined in the network construction section to record faculty hiring ties (including a PhD holder’s initial and subsequent job attainment at different career stages) among the same set of institutions in the previous two five-year periods. However, the network data did not include cases in which a faculty member was hired by any of the 81 universities between 2005 and 2014 and then quit academia or moved to an institution outside of the network boundary due to the unobservable nature of such career transitions.

The 2010 to 2014 hiring network (average degree = 3.79, network density = .05, average clustering coefficient = .13, average path length = 2.87) reflected interorganizational hiring patterns in the recent past. The 2005 to 2009 hiring network (average degree = 3.57, network density = .05, average clustering coefficient = .08, average path length = 3.11) was a proxy measure for network inertia in response to positive performance feedback. As the majority of job postings seek junior faculty (NCA, 2019), it can be inferred that the 81 universities hired more assistant professors than the other position ranks during the 2005 to 2009 period. Accordingly, tenure decisions for these newly hired assistant professors should be made no later than 2015. Given that all PhDs in the sample (n = 2,192) still survived in one of the 81 universities at the time of data collection (March 2020), most of the junior faculty recruited between 2005 and 2009 within the network boundary were expected to achieve tenure prior to the 2015 hiring season. The successful tenure cases signaled to an institution’s search committee that alumni or alumnae of the newly tenured faculty members were likely to survive in the institution.

Previous ties were weighted, so cell values in the two network matrices denoted the number of faculty sent from one institution to another. Values of off-diagonal cells, S_ij (i ≠ j), ranged from 0 to 3. All diagonal cells, S_ij (i = j), were removed from model estimation.

Niche width (M = 6.59, SD = 2.97, Min = 2, Max = 14) was measured as a university’s total number of affiliated research specializations. This operationalization is consistent with the conventional approach in organizational ecology that treats niche width as the count of an organization’s target social or market categories (Hannan & Freeman, 1989; Hsu, 2006). The NCA Doctoral Program Guide used 16 institutionalized labels to categorize 87 doctoral programs. The categorization process was based entirely on each program’s self-claimed expertise. A program can be assigned to more than one category. The classification results were directly retrieved from the NCA site in April 2020.

Research specializations of the 21 doctoral programs absent from the NCA list were manually coded through keyword matching. For instance, the description of the Grady College’s PhD program in mass communication at the University of Georgia stated that the program’s areas of expertise included advertising/public relations, journalism, and entertainment and media studies. Consequently, this doctoral program was a category member of “Public Relations and Strategic Communication,” “Journalism,” and “Mass Communication/Media Studies.”

The next step was to aggregate category memberships at the university level. If a university had only one communication PhD program, niche width was computed as the program’s number of affiliated research specializations. If not, the variable was the number of unduplicated category elements across programs. For instance, NCA listed the PhD program offered by the Department of Communication at the University of Georgia. According to NCA’s categorization, this program was affiliated with “Health Communication,” “Interpersonal Communication,” “Rhetoric,” and “Science Communication.” As the Grady College had three unduplicated areas of study (i.e., “Public Relations and Strategic Communication,” “Journalism,” and “Mass Communication/Media Studies”), the university’s niche width was 7.

Niche overlap was operationalized as the fraction of university i’s niches simultaneously occupied by university j (Hannan et al., 2003). This operational definition takes into account size differences in niche width (Sohn, 2001). Mathematically, the numerator was the count of overlapping research specializations between i and j. The denominator was the niche width of i. Suppose that university i was affiliated with “Health Communication,” “Organizational Communication,” and “Political Communication.” University j was assigned to “Computer Mediated Communication” and “Political Communication.” In this example, the only shared category was “Political Communication.” As a result, i’s niche overlap with j was 0.33 (1/3), while j’s niche overlap with i was 0.50 (1/2).

Niche overlap was measured at the dyadic level. An 81 × 81 square matrix was established to capture the degree of niche overlap between pairs of institutions. Values of off-diagonal cells, S_ij (i ≠ j), ranged from 0 (no overlap) to 1 (complete overlap) and were expressed as ratios (M = .52, SD = .26), allowing for meaningful comparisons across dyads (Sohn, 2001). Niche overlap can be asymmetric because S_ij does not necessarily equal S_ji. All diagonal cells, S_ij (i = j), were coded as “0.”

Control variables

Structural effects. Inferential network analysis typically controls multiple mechanisms of self-organization in tie-formation processes. Following the procedures outlined by Lusher et al. (2013), this study controlled the structural effects of arc, reciprocity, popularity, activity, triadic closure, and connectivity. Arc is a baseline propensity for tie formation. Reciprocity reveals the tendency for universities to exchange graduates in the job market. Popularity (or activity) reflects the propensity to hire (or place) graduates from (or in) a large number of peer institutions. Triadic closure indicates that two connected universities have a shared third partner in the hiring network. Connectivity reflects the degree to which tie senders are also tie receivers.

State is a categorical variable. It was included to control the tendency for universities in the same state to form faculty hiring ties. The 81 universities are in 39 states.

Presence of multiple programs is a binary variable indicating whether a focal university offers separate PhD programs in communication. In the dataset, 20 institutions were coded as “1.” Considering that universities occupying broad niches may have more than one PhD program, this control variable can reduce the risk of over- or under-estimating the effect of niche width.

Organizational type distinguishes between private (coded as “1”) and public (coded as “0”) universities. Most institutions (65 out of 81, 80.25%) in the sample are public universities. Including this variable is necessary because Mai et al. (2015) revealed that organizational type played an important role in driving the creation of faculty hiring ties in the field of communication.

Organizational size. Faculty size (M = 27.06, SD = 16.26, Min = 5, Max = 80) was the number of tenured or tenure-track faculty members affiliated with a university’s PhD program(s) in communication. Retired professors were excluded. Student size (M = 36.86, SD = 26.07, Min = 0, Max = 168) was the number of PhDs in communication that a university conferred from 2015 to 2019. The data were retrieved from the Survey of Earned Doctorates (SED) conducted by the National Science Foundation (NSF). Both variables were normalized to have a value between 0 and 1. As they were highly correlated (rs(79) = .68, p < .001), organizational size was calculated as the sum of normalized faculty and student size. Including this variable in model estimation helps to rule out the alternative explanation that tie creation is determined by organizational size (Barnett et al., 2010; Mai et al., 2015).

Organizational prestige (M = 2.81, SD = 2.25, Min = 0, Max = 5) was the number of times a university appeared in the QS World University Rankings (QSWUR) for “Communication and Media Studies” during the 2015 to 2019 period. Previous research typically used the authoritative rankings from the National Research Council (NRC; Barnett & Feeley, 2011; Mai et al., 2015) or U.S. News & World Report (USNWR; Gondal, 2018) to quantify organizational prestige in doctoral education. However, the latest version of the NRC rankings was released in 2010, so the changing landscape in communication education is not captured. In addition, missing data can be a serious problem. More than 20 universities on the NCA list are not on the NRC list. While USNWR updates graduate school rankings annually, the communication discipline is not included. This study relied on QSWUR to construct the prestige measure for two reasons. First, the legitimacy of QSWUR has been acknowledged by NCA. The NCA Doctoral Program Guide reports each university’s standing in QSWUR, while the NRC and USNWR statistics are not shown. Second, QSWUR is updated every year. QSWUR is not without limitations. Like the rankings published by NRC and USNWR, not every university is covered by QSWUR.

Faculty seniority (M = 18.36, SD = 4.02, Min = 9.57, Max = 32.14) was the average number of years since affiliated faculty members earned their doctoral degrees (Mai et al., 2015). A larger number indicates a higher level of seniority. For those who held multiple PhDs, the most recent degree was considered.

Hiring PhDs in other disciplines (M = 1.49, SD = 2.05, Min = 0, Max = 11) was the number of times a university hired faculty trained in disciplines other than communication during the 2015 to 2019 period. Controlling this variable is necessary due to the multidisciplinary tradition of the communication field. An institution’s tendency to hire communication PhDs can be shaped by its simultaneous experience in atypical hiring practices.

Hiring PhDs in communication outside of the network boundary (M = .16, Min = 0, Max = 1) was dichotomized to indicate whether a university hired graduates from communication doctoral programs that were not included in the sample. This variable was dichotomized because only four universities hired at least two PhDs from alternative communication programs during the observation period.

Analytical Procedures

ERGMs using Markov Chain Monte Carlo maximum likelihood estimates were employed to test all hypotheses. This inferential technique relies on computer simulations to fit stochastic models that describe structural features of an observed network and capture both regularities and randomness in network formation. Explanatory variables in ERGMs include both endogenous and exogenous parameters. Endogenous parameters estimate pure structural effects such as arc, reciprocity, popularity, activity, triadic closure, and connectivity. Exogenous parameters refer to external relations and nodal attributes. Specifically, ERGMs allow for the inclusion of one or multiple external valued relations as exogenous predictors of the observed network. In addition, researchers can examine whether nodes with certain attributes are more likely to receive or send relations in a focal network. In the present case, tie senders placed graduates in tie receivers. Besides receiver and sender effects, both the match (for categorical attributes) and difference (for binary or valued attributes) parameters can be used to test if nodes with similar attributes are more likely to form a tie (i.e., homophily effects). Following the guidelines for model specification (Lusher et al., 2013), this study estimated the receiver, sender, and homophily effects of all nodal attributes.

Hypotheses 1 through 3 posited that tie-formation processes were driven by exogenous parameters. Specifically, tie presence in previous periods (H₁) and niche overlap (H₃) are external relations, and niche width (H₂) is a nodal attribute. All control variables except structural effects are also nodal attributes. Supplemental Table S1 reports correlation coefficients among non-categorical nodal attributes.

While ERGMs do not assume independent observations, the interpretation of an ERGM is similar to that of a logistic regression. Coefficients are usually expressed in the form of log odds and can be further converted to odds ratios and probabilities. A positive (or negative) and significant coefficient indicates an increased (or decreased) tendency for a specific structural parameter to be observed in a focal network. A parameter is statistically significant if the ratio of its coefficient to standard error is greater than 1.96 in absolute value. Recent advances in ERGMs have enabled researchers to empirically test the formation mechanisms of valued networks (Pilny & Atouba, 2018), but the hiring network was still treated as binary in statistical analysis. One important reason was that there was little variation in non-zero values of network ties (M = 1.19, SD = 0.50, Min = 1, Max = 4). There were 399 unique ties (excluding self-loops) in the observed network, but only 60 of them (15.03%) had a value greater than 1. As a result, cells in the original network were dichotomized before model estimation. The final data analysis was performed using the statnet package (version 2019.6) in R.

Results

Table 2 presents the results of ERGMs predicting tie formation in the faculty hiring network among 81 U.S. PhD-granting universities in the communication discipline from 2015 to 2019. Model 1 provides a baseline estimate for the presence of a tie. Multiple endogenous (i.e., edges, mutual, gwidegree, gwodegree, gwesp, and gwdsp) and exogenous (i.e., receiver, sender, and homophily effects of nodal attributes) control parameters were included in the baseline model. Model 2 adds several exogenous parameters of theoretical interest. Both the sender and difference parameters of niche width were also included as controls. The total amount of explained deviance was 73.41% ([8,983−2,389]/8,983). Only one significant parameter in Model 1 became non-significant in Model 2 (i.e., the difference parameter of organizational size). No new control parameter gained significance in Model 2.

Table 2.

ERGMs Predicting Tie Formation in the Hiring Network, 2015 to 2019.

Network parameters	Model 1		Model 2
Network parameters	Log odds	SE	Log odds	SE
Endogenous parameters
Arc (edges)	−3.32***	0.52	−3.44***	0.56
Reciprocity (mutual)	0.47^†	0.25	0.47^†	0.25
Popularity (gwidegree, α = 0.20)	−0.08	0.57	−0.08	0.58
Activity (gwodegree, α = 0.20)	−0.61	0.46	−0.67	0.45
Triadic closure (gwesp, α = 0.20)	0.05	0.11	0.03	0.11
Connectivity (gwdsp, α = 0.20)	−0.06*	0.02	−0.05*	0.02
Exogenous parameters
State (match)	0.10	0.36	−0.04	0.36
Multiple programs (receiver)	−0.30*	0.14	−0.34*	0.16
Multiple programs (sender)	0.06	0.14	0.08	0.16
Multiple programs (difference)	−0.02	0.12	0.04	0.12
Private university (receiver)	−0.15	0.24	−0.15	0.23
Private university (sender)	0.84***	0.23	0.79**	0.24
Private university (difference)	−0.46*	0.20	−0.41*	0.20
Organizational size (receiver)	3.51***	0.58	3.15***	0.55
Organizational size (sender)	3.79***	0.51	3.16***	0.51
Organizational size (difference)	−0.85*	0.42	−0.64	0.42
Organizational prestige (receiver)	−0.04	0.04	−0.04	0.04
Organizational prestige (sender)	0.21***	0.04	0.20***	0.05
Organizational prestige (difference)	−0.10*	0.04	−0.10*	0.04
Faculty seniority (receiver)	−0.07***	0.02	−0.06***	0.02
Faculty seniority (sender)	−0.01	0.02	−0.01	0.02
Faculty seniority (difference)	0.00	0.02	0.00	0.02
Hiring PhDs in other disciplines (receiver)	−0.01	0.03	0.00	0.04
Hiring PhDs in other disciplines (sender)	−0.02	0.03	0.00	0.03
Hiring PhDs in other disciplines (difference)	−0.07*	0.03	−0.09*	0.04
Hiring PhDs in communication outside of the network boundary (receiver)	0.19	0.19	0.23	0.19
Hiring PhDs in communication outside of the network boundary (sender)	0.09	0.18	0.14	0.19
Hiring PhDs in communication outside of the network boundary (difference)	0.09	0.19	0.08	0.19
H₁: Previous ties, 2010–2014			0.89***	0.16
H₁: Previous ties, 2005–2009			0.68***	0.17
H₂: Niche width (receiver)			−0.10***	0.03
Niche width (sender)			0.00	0.02
Niche width (difference)			−0.01	0.02
H₃: Niche overlap			1.48***	0.38
Null deviance	8,983		8,983
Residual deviance	2,460		2,389
AIC	2,516		2,457
BIC	2,705		2,687

Note. †p < .10, *p < .05, **p < .01, ***p < .001.

The lower part of Table 2 reports the results of hypothesis testing. Hypothesis 1 predicted that the presence of previous ties would increase the likelihood of subsequent tie formation. Model 2 revealed that the number of faculty sent from i to j (i ≠ j) in the past was a positive and significant predictor of tie presence in the 2015 to 2019 hiring network (previous ties, 2010–2014: log odds = 0.89, SE = 0.16, p < .001; previous ties, 2005–2009: log odds = 0.68, SE = 0.17, p < .001). This result provided substantial support for H₁.

Hypothesis 2 stated that an increased niche width would lead to a decreased likelihood that an organization would receive ties in the hiring network. Model 2 showed that the receiver parameter of niche width produced a negative and significant effect (log odds = −0.10, SE = 0.03, p < .001). Specifically, a one-unit increase in organizational niche width would lead to a decrease in the odds of tie formation by 9.52% (1 – exp(−0.10)). Thus, H₂ was supported.

Hypothesis 3 proposed that an increased niche overlap would lead to an increased likelihood of tie formation between organizations in the hiring network. The final results confirmed that niche overlap was a positive and significant predictor of the outcome (log odds = 1.48, SE = 0.38, p < .001). An increase from zero to one in a university’s niche overlap with another would result in an increase in the odds of tie creation by 339.29% (exp(1.48) – 1). H₃ was therefore supported.

A goodness-of-fit (GoF) test was run to check whether the ERGM estimation provided a good representation of the observed network. The detailed GoF results are presented in Supplemental Figure S1. It was confirmed that the final model reproduced the distributions of out-degree, in-degree, edge-wise shared partners, triadic census, and minimum geodesic distance. The GoF diagnostics also showed that the predictors well captured relational features.

Discussion

The objective of this study was to build an ecological framework for understanding the formation mechanisms of hiring networks in the academic marketplace. Faculty hiring patterns among 81 PhD-granting universities in the communication discipline were examined from ecological and network perspectives. Three hypotheses were developed to capture inertial forces and niche processes in network emergence from 2015 to 2019. The empirical results of ERGMs confirmed that the likelihood of tie formation was determined by network inertia, categorical niche width, and niche overlap.

Drawing on structural inertia theory (Hannan & Freeman, 1984) and its extension to social network research (Kim et al., 2006), the current study identifies network inertia as a central mechanism that explains the creation of faculty hiring ties. A theoretical explanation for the strong tendency for previous ties to replicate over time is that experience with old partners produces trust and reduces search costs (Kim et al., 2006). This path-dependent searching strategy also avoids risks associated with faculty hires from new partners due to the availability of inside information that would not otherwise be observable through an applicant’s submitted materials. Current employees may also help job candidates from their alma mater to stand out from the pack through strong recommendations, mobilization efforts, and communicative acts (Jackson, 2019). The inertial-based hiring pattern is also consistent with the evolutionary mechanism of selective retention (Aldrich et al., 2020)—that is, institutions are subject to strong inertial forces, especially when previously hired employees have achieved career success. The positive and significant effect of previous ties from 2005 to 2009 on tie presence in the 2015 to 2019 hiring network provided indirect evidence for this evolutionary account of network inertia. The majority of those employed as assistant professors between 2005 and 2009 were expected to receive tenure prior to the 2015 hiring season. Such positive performance feedback could motivate search committees to hire alumni or alumnae of the newly tenured faculty members.

It is worth mentioning that structural inertia can be detrimental to organizational self-reflexivity and long-term success, even though surviving organizations tend to be inert (Hannan & Freeman, 1984; Stieglitz et al., 2016). This reminds us that inertial-based hiring practices may further reinforce biases in decision-making. Considering that the prestige of an applicant’s PhD institution is not necessarily an accurate quality signal of the applicant’s actual achievements (Headworth & Freese, 2016), the presence of network inertia in the communication job market also cautions against the assumption that placement centrality in the faculty hiring network reflects the quality of doctoral education (Barnett et al., 2010; Barnett & Feeley, 2011).

This study also demonstrates that niche processes play key roles in hiring decisions. An identity-based view (Hsu, 2006; Hsu & Hannan, 2005) was adopted to map organizational niches in the resource space. In the present case, niches were reflected by self-claimed organizational affiliations with 16 institutionalized areas of study specified by NCA. Professional associations (e.g., NCA) typically assist in the institutionalization of social and market categories, facilitating the formation of consensus among stakeholders (Scott, 2013). As one of the leading professional associations in the communication discipline, NCA conferred legitimacy on the 16 research specializations by incorporating them into the Doctoral Program Guide.

Guided by niche theory (Hannan & Freeman, 1989), network brokerage (Burt, 2004), and a sociological perspective on organizational identity (Gondal, 2018; Vergne & Wry, 2014), the current study argues that niche broadness indicates a complex or diffuse organizational identity that allows a generalist institution to develop a tolerance for atypical hiring practices. The empirical results supported the hypothesis that an increased niche width would result in a decreased likelihood that a focal university would receive faculty hiring ties. This effect held true even when the tendency for generalist institutions to offer multiple PhD programs was controlled. The results of Spearman’s correlation analysis (see Supplemental Table S1) further revealed that niche width was positively related to the propensity to hire faculty trained in disciplines other than communication (rs(79) = .23, p = .04). In addition, niche width was not associated with the event of hiring PhDs in communication outside of the network boundary (rs(79) = .03, p = .81). To rule out the alternative explanation that niche width produced a curvilinear effect on hiring decisions, the current study ran a post hoc analysis by including the squared “niche width” variable in model estimation (see Supplemental Table S2). The findings revealed that the receiver (log odds = −0.01, SE = 0.01, p = .28), sender (log odds = 0.00, SE = 0.01, p = .57), and difference (log odds = −0.00, SE = 0.01, p = .57) parameters of the squared term were not statistically significant, offering little evidence of a non-linear phenomenon.

Another important finding related to niche processes is that niche overlap is positively related to the tendency for two universities to form a faculty hiring tie. This result is in line with the increased probability of tie formation between sociology departments with multiple shared research specializations (Gondal, 2018). Compared to the traditional view of communicative niches (e.g., Dimmick, 2003; Weber, 2017), this study emphasizes that niche overlap signals the degree of mutualistic interdependence between organizations (Aldrich et al., 2020). In the present case, this organizational interdependence is manifested in the exchange and transfer of economic capital in the academic marketplace (Burris, 2004). While competing for the same set of scarce resources such as top students and external funding, those occupying overlapping niches can support each other through faculty hires.

The current study did not hypothesize how non-ecological characteristics (e.g., prestige) and network self-organization (e.g., reciprocity, transitivity, and cumulative advantage) would influence the likelihood of tie formation, but the empirical findings presented in Table 2 provide additional insights into hiring patterns in the communication field. Specifically, the positive and significant sender effect of organizational prestige is consistent with the previous finding that a program’s ability to place graduates was positively associated with its reputation (Barnett et al., 2010). Also, the positive and marginally significant parameter of reciprocity replicated the earlier observation of PhD exchange in the communication job market (Mai et al., 2015). The widespread presence of reciprocal ties suggests that hiring institutions may limit their search scope to PhDs from the institutions in which their own graduates have recently been placed. Additionally, the final results did not support the claim that the endogenous factors of transitivity and cumulative advantage drove the creation of faculty hiring ties (Mai et al., 2015), as evidenced by the non-significant effects of triadic closure and activity, respectively. In other words, the 2015 to 2019 hiring network did not exhibit the feature of local clustering. There was also no strong tendency for some nodes to place graduates in a large number of peer institutions. Failure to replicate the finding of cumulative advantage should be discussed in combination with a Gini coefficient of .57 for the distribution of out-degree score in the observed network. This is already a conservative estimate of equality because the first cohort of communication PhD students enrolled in several universities (e.g., Boston University and the University of Delaware) had yet to enter the academic job market during the observation period. Despite the inflated estimate of placement inequality, the Gini coefficient of the observed network (.57) is still lower than that of faculty hiring networks in other disciplines (ranging from .62 to .76, see Clauset et al., 2015; Headworth & Freese, 2016; Kawa et al., 2019). As the Gini coefficient for the network data collected in 2007 is .72 (Barnett et al., 2010), systematic inequality in communication job placements has reduced over time.

Theoretical Contributions

First, this paper adds to the organizational ecology approach in communication (e.g., Xu, 2021; Xu et al., 2021). Communication scholars have employed ecological theories to explain changes in media forms (e.g., Dimmick, 2003; Scolari, 2013), communication networks (e.g., Lee & Monge, 2011; Margolin et al., 2015), and organizational practices (e.g., Weber, 2017; Yang, 2020), but few attempts have been made to draw insights from network inertia (Kim et al., 2006) and the identity-based view of niche processes (Hsu, 2006; Hsu & Hannan, 2005) to conceptualize interorganizational dynamics. By highlighting the importance of institutionalized categories in identity construction, this study sheds light on the niche foundations of network emergence, thus contributing to refining and developing the niche perspective in communication (e.g., Scolari, 2013) and the communicative constitution of organizational identities (e.g., Koschmann, 2013). In the present case, self-claimed areas of study are communicative, signaling an institution’s identity in the evolving disciplinary space.

Second, this study advances the research on networks, communication, and labor markets (e.g., Clauset et al., 2015; Jackson, 2019; Piercy & Lee, 2019) by demonstrating that network inertia and niche processes significantly constrain the formation of faculty hiring ties in the academic marketplace. The mechanism of network inertia suggests that a current employee may act as an intermediary who facilitates communication and trust building between search committee members and job candidates from the employee’s alma mater. In addition, hiring decisions are shaped by a target institution’s ecological position communicated by its affiliations with commonly understood research specializations. Previous research has explored faculty hiring patterns among PhD-granting institutions in communication (e.g., Barnett et al., 2010; Barnett & Feeley, 2011; Mai et al., 2015). The current study contributes to validating an ecological model of hiring decisions while simultaneously considering alternative explanations such as network self-organization (Mai et al., 2015) and status-based hiring patterns (Barnett et al., 2010).

Practical Implications

The empirical investigation of ecological influences on hiring decisions has practical implications for faculty and students in doctoral programs in communication. In the dataset, 81 universities on the NCA list filled 647 tenured or tenure-track positions during the 2015 to 2019 period, and 78.52% of the faculty positions (508 out of 647) were occupied by communication PhDs trained in the same set of institutions. These statistics should be interpreted with caution because senior positions for which early career scholars could not apply were not excluded from analysis. Also, NSF (n.d.) reported that the 81 universities conferred a total of 2,986 PhDs in communication over the same period, suggesting that only a small proportion of graduates got tenure-track positions within the same network boundary. The above results can help inform discussions of training and supervision plans for doctoral students.

The empirical findings also provide insights into the organizing logic of the academic marketplace. The path-dependent nature of faculty hiring ties means that job applicants can mobilize social resources embedded in alumni networks to maximize their chances of receiving offers. Meanwhile, search committee members must be aware of inertial forces underlying hiring practices and promote self-reflexivity to ensure fairness when evaluating candidate quality. In addition, the significant effects of niche width and niche overlap on hiring decisions suggest potential benefits of strategic positioning in the disciplinary landscape. Specifically, PhDs in communication may consider highlighting their inter-disciplinary training backgrounds when applying for job positions offered by generalist institutions. Another potentially useful strategy is to signal multiple shared research specializations to match the expectation of hiring institutions.

Limitations and Directions for Future Research

This study offers a careful analysis of ecological influences on the formation of faculty hiring ties, but it is still limited in several aspects. First, PhD holders may not seek academic employment at PhD-granting institutions. In fact, the majority of the 3,902 job listings over the same five-year period were either non-tenure track positions or posted by institutions absent from the NCA list (NCA, 2019). Another related issue is that not all U.S. PhD-granting communication programs are included in the sample (e.g., Columbia Journalism School’s PhD program in communications). Nonetheless, this may not be a serious concern because 91.58% (2,986 out of 3,262) of all conferred doctoral degrees in communication were from the 81 universities on the NCA list (NSF, n.d.). In addition, there were only 10 cases in which faculty positions were filled by graduates from U.S. communication PhD programs outside of the network boundary.

Second, the analysis does not reflect global faculty hiring patterns. Scholars trained in the U.S. may seek and accept faculty positions in non-U.S. institutions at different career stages. For instance, international students may decide to return to their countries of origin after graduation. Interestingly, the data revealed that the 81 universities hired only eight people who obtained their PhDs in communication outside of the U.S., suggesting an asymmetric flow of talents. Future studies can draw on world system theory (Wallerstein, 1974) to test whether a core-periphery structure applies to the international job market.

Third, the current study does not consider specialties required for each job position. In fact, resources and opportunities are not evenly distributed across areas of study (niches) in the communication job market. For instance, both “Political Communication” and “Public Relations and Strategic Communication” had 40 category members. However, the former had just 10 openings in the 2018 to 2019 academic year, while there were as many as 126 job listings for the latter (NCA, 2019). Future studies should take into account unequal distributions of resource richness in niche space to better capture ecological dynamics in the academic job market.

Another limitation is the sole focus on faculty positions. While the findings are not readily generalizable to other non-academic settings, recruitment processes in academia share many features with those in industry in the aspect that referrals, common connections, mutual fit, and shared experiences play vital roles (Aldrich et al., 2020; Jackson, 2019). Moreover, the key issues addressed in the study—inertial forces and niche processes in the job market—are valid across a wide range of professional contexts. Thus, the theoretical framework presented here offers a useful lens to understand general hiring decisions.

There are several methodological limitations. As ERGMs do not support modeling self-loops, all self-hires were removed prior to model estimation. Additionally, network inertia was not modeled as an endogenous structural signature due to the absence of continuous-time relational data (Leenders et al., 2016). Furthermore, the faculty hiring data in the previous two five-year periods help smooth year-over-year trends, but this analytical strategy is still insufficient to fully capture long-term patterns in the communication job market. Last but not least, while valued ties contain richer structural information than do dichotomous ties, this study relied on binary network data for hypothesis testing. The post hoc analysis modeled the hiring network as valued. As the higher-order endogenous parameters (i.e., gwidegree, gwodegree, gwesp, and gwdsp) are inapplicable to valued ERGMs, they were replaced by the nodeisqrtcovar, nodeosqrtcovar, and transitiveweights parameters to control network centralization and clustering. Also, the edges parameter was changed to the sum parameter (defined as the sum of dyad values) in model specification. The remaining endogenous and exogenous parameters in the valued ERGM were the same as those in the binary model (see Table 2). It turned out that the model was non-identifiable. This issue might arise from the increased computational complexity of the valued ERGM (Pilny & Atouba, 2018).

Conclusion

Human resources are an important asset of organizations. The current study explores ecological influences on hiring decisions in the job market. A social network analysis was performed to investigate faculty hiring patterns in the field of communication over a five-year period. Empirical results from a sample of 81 PhD-granting institutions and 647 hiring events showed that organizational decision-making was significantly constrained by network inertia, niche width, and niche overlap.

Supplemental Material

sj-docx-1-crx-10.1177_00936502211034687 – Supplemental material for Ecological Influences on the Formation of the Hiring Network in the Communication Job Market, 2015 to 2019

Supplemental material, sj-docx-1-crx-10.1177_00936502211034687 for Ecological Influences on the Formation of the Hiring Network in the Communication Job Market, 2015 to 2019 by Yu Xu in Communication Research

Footnotes

Acknowledgements

I am grateful to Editor Jennifer Gibbs and three anonymous reviewers for their comments on earlier drafts of the manuscript.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Yu Xu

Supplemental Material

Supplemental material for this article is available online.

Author Biography

Yu Xu (PhD, University of Southern California, 2019) is an assistant professor in the Medill School of Journalism, Media, Integrated Marketing Communications at Northwestern University. He is also a Faculty Affiliate of the Center for Communication & Public Policy in the School of Communication. His research interests include organizational change, social networks, media industries, sociocultural evolution, and computational social science.

References

Aldrich

H. E.

Ruef

Lippmann

(2020). Organizations evolving. Edward Elgar Publishing.

Barnett

G. A.

Danowski

J. A.

Feeley

T. H.

Stalker

(2010). Measuring quality in communication doctoral education using network analysis of faculty-hiring patterns. Journal of Communication, 60(2), 388–411. https://doi.org/10.1111/j.1460-2466.2010.01487.x

Barnett

G. A.

Feeley

T. H.

(2011). Comparing the NRC and the faculty hiring network methods of ranking doctoral programs in communication. Communication Education, 60(3), 362–370. https://doi.org/10.1080/03634523.2011.558202

Bian

(1997). Bringing strong ties back in: Indirect ties, network bridges, and job searches in China. American Sociological Review, 62(3), 366–385. https://doi.org/10.2307/2657311

Breiger

R. L.

(1974). The duality of persons and groups. Social Forces, 53(2), 181–190. https://doi.org/10.2307/2576011

Brown

A. D.

Starkey

(2000). Organizational identity and learning: A psychodynamic perspective. Academy of Management Review, 25(1), 102–120. https://doi.org/10.2307/259265

Burris

(2004). The academic caste system: Prestige hierarchies in PhD exchange networks. American Sociological Review, 69(2), 239–264. https://doi.org/10.1177/000312240406900205

Burt

(2004). Structural holes and good ideas. American Journal of Sociology, 110(2), 349–399. https://doi.org/10.1086/421787

Clauset

Arbesman

Larremore

D. B.

(2015). Systematic inequality and hierarchy in faculty hiring networks. Science Advances, 1(1), e1400005. https://doi.org/10.1126/sciadv.1400005

10.

Dimmick

(2003). Media competition and coexistence: The theory of the niche. Lawrence Erlbaum.

11.

Gondal

(2018). Duality of departmental specializations and PhD exchange: A Weberian analysis of status in interaction using multilevel exponential random graph models (mERGM). Social Networks, 55, 202–212. https://doi.org/10.1016/j.socnet.2018.07.005

12.

Granovetter

M. S.

(1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360–1380. https://doi.org/10.1086/225469

13.

Hannan

M. T.

Carroll

G. R.

Pólos

(2003). The organizational niche. Sociological Theory, 21(4), 309–340. https://doi.org/10.1046/j.1467-9558.2003.00192.x

14.

Hannan

M. T.

Freeman

(1984). Structural inertia and organizational change. American Sociological Review, 49(2), 149–164. https://doi.org/10.2307/2095567

15.

Hannan

M. T.

Freeman

(1989). Organizational ecology. Harvard University Press.

16.

Headworth

Freese

(2016). Credential privilege or cumulative advantage? Prestige, productivity, and placement in the academic sociology job market. Social Forces, 94(3), 1257–1282. https://doi.org/10.1093/sf/sov102

17.

Hsu

(2006). Jacks of all trades and masters of none: Audiences’ reactions to spanning genres in feature film production. Administrative Science Quarterly, 51(3), 420–450. https://doi.org/10.2189/asqu.51.3.420

18.

Hsu

Hannan

M. T.

(2005). Identities, genres, and organizational forms. Organization Science, 16(5), 474–490. https://doi.org/10.1287/orsc.1050.0151

19.

Jackson

M. O.

(2019). The human network: How your social position determines your power, beliefs, and behaviors. Pantheon Books.

20.

Kawa

N. C.

Michelangeli

J. A. C.

Clark

J. L.

Ginsberg

McCarty

(2019). The social network of US academic anthropology and its inequalities. American Anthropologist, 121(1), 14–29. https://doi.org/10.1111/aman.13158

21.

Kim

T. Y.

Swaminathan

(2006). Framing interorganizational network change: A network inertia perspective. Academy of Management Review, 31(3), 704–720. https://doi.org/10.5465/amr.2006.21318926

22.

Koschmann

M. A.

(2013). The communicative constitution of collective identity in interorganizational collaboration. Management Communication Quarterly, 27(1), 61–89. https://doi.org/10.1177/0893318912449314

23.

Lee

Monge

P. R.

(2011). The coevolution of multiplex communication networks in organizational communities. Journal of Communication, 61(4), 758–799. https://doi.org/10.1111/j.1460-2466.2011.01566.x

24.

Leenders

R. T. A. J.

Contractor

N. S.

DeChurch

L. A.

(2016). Once upon a time: Understanding team processes as relational event networks. Organizational Psychology Review, 6(1), 92–115. https://doi.org/10.1177/2041386615578312

25.

Lusher

Koskinen

Robbins

(2013). Exponential random graph models for social networks: Theory, methods, and applications. Cambridge University Press.

26.

Mai

Liu

González-Bailón

(2015). Network effects in the academic market: Mechanisms for hiring and placing PhDs in communication (2007–2014). Journal of Communication, 65(3), 558–583. https://doi.org/10.1111/jcom.12158

27.

March

J. G.

(1991). Exploration and exploitation in organizational learning. Organization Science, 2(1), 71–87. https://doi.org/10.1287/orsc.2.1.71

28.

Margolin

Lee

Shen

Weber

Monge

Fulk

(2015). Normative influences on network structure in the evolution of the Children’s Rights NGO network, 1977–2004. Communication Research, 42(1), 30–59. https://doi.org/10.1177/0093650212463731

29.

Maurer

Ebers

(2006). Dynamics of social capital and their performance implications: Lessons from biotechnology start-ups. Administrative Science Quarterly, 51(2), 262–292. https://doi.org/10.2189/asqu.51.2.262

30.

McPherson

(1983). An ecology of affiliation. American Sociological Review, 48(4), 519–532. https://dx-doi-org-s.web.bisu.edu.cn/10.2307/2117719

31.

Monge

P. R.

Contractor

N. S.

(2003). Theories of communication networks. Oxford University Press.

32.

National Communication Association. (2019). 2018-2019 academic job listings in communication report. Retrieved May 15, 2020, from https://www.natcom.org/sites/default/files/publications/2019NCA_JobListings_Report.pdf

33.

National Science Foundation. (n.d.). Survey of earned doctorates. Retrieved April 20, 2020, from https://www.nsf.gov/statistics/srvydoctorates/#tabs-2

34.

Pallais

Sands

E. G.

(2016). Why the referential treatment? Evidence from field experiments on referrals. Journal of Political Economy, 124(6), 1793–1828. https://doi.org/10.1086/688850

35.

Piercy

C. W.

Lee

S. K.

(2019). A typology of job search sources: Exploring the changing nature of job search networks. New Media & Society, 21(6), 1173–1191. https://doi.org/10.1177/1461444818808071

36.

Pilny

Atouba

(2018). Modeling valued organizational communication networks using exponential random graph models. Management Communication Quarterly, 32(2), 250–264. https://doi.org/10.1177/0893318917737179

37.

Scolari

C. A.

(2013). Media evolution: Emergence, dominance, survival and extinction in the media ecology. International Journal of Communication, 7, 1418–1441.

38.

Scott

W. R.

(2013). Institutions and organizations: Ideas, interests, and identities. Sage Publications.

39.

Shi

Dokshin

F. A.

Genkin

Brashears

M. E.

(2017). A member saved is a member earned? The recruitment-retention trade-off and organizational strategies for membership growth. American Journal of Sociology, 82(2), 407–434. https://doi.org/10.1177/0003122417693616

40.

Shumate

Lipp

(2008). Connective collection action only: An examination of the hyperlink network structure of an NGO issue network. Journal of Computer-Mediated Communication, 14(1), 178–201. https://doi.org/10.1111/j.1083-6101.2008.01436.x

41.

Sohn

M-W.

(2001). Distance and cosine measures of niche overlap. Social Networks, 23(2), 141–165. https://doi.org/10.1016/S0378-8733(01)00039-9

42.

Stieglitz

Knudsen

Becker

M. C.

(2016). Adaptation and inertia in dynamic environments. Strategic Management Journal, 37(9), 1854–1864. https://doi.org/10.1002/smj.2433

43.

Van den Bulte

Wuyts

. (2007). Social networks and marketing. Marketing Science Institute.

44.

Vergne

J. P.

Wry

(2014). Categorizing categorization research: Review, integration, and future directions. Journal of Management Studies, 51(1), 56–94. https://doi.org/10.1111/joms.12044

45.

Wallerstein

(1974). The modern world-system. Academic Press.

46.

Weber

M. S.

(2017). Unseen disruptions and the emergence of new organizations. Communication Theory, 27(1), 92–113. https://doi.org/10.1111/comt.12105

47.

Weber

M. S.

Monge

P. R.

(2017). Industries in turmoil: Driving transformation during periods of disruption. Communication Research, 44(2), 147–176. https://doi.org/10.1177/0093650213514601

48.

(2021). Evolution of audience duplication networks among social networking sites: Exploring the influences of preferential attachment, audience size, and niche width. New Media & Society. https://doi.org/10.1177/1461444821993048

49.

Fulk

Monge

(2021). Ecological influences on network tie dissolution in the evolution of affiliation networks in the International Communication Association, 2009–2015. Human Communication Research, 47(1), 75–103. https://doi.org/10.1093/hcr/hqaa009

50.

Yang

(2020). The issue niche theory of nongovernmental and nonprofit organizations’ interorganizational network ecology. Communication Theory, 30(1), 41–63. https://doi.org/10.1093/ct/qtz014

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.48 MB