Abstract
Research suggests that organizations tend to collaborate with others that share similar ascribed status, but focuses less on the role of value homophily. To advance a fuller understanding of how organizations select partners, this study examines the roles of—and relationship between—status and value homophily in interorganizational collaboration (IOC). Specifically, it examines the influence of value homophily (i.e., similar service, staff, and organizational identity) on three forms of status homophily (attribute-based, geography-based, and institutional) in nonprofit collaborative aspirations. Survey data from 141 U.S. faith-based organizations (FBOs) revealed the differential impact of organizational and service religiosity on FBOs’ collaboration preferences. Specifically, findings suggest that organizational religiosity makes FBOs more restrictive, but service religiosity makes FBOs less restrictive in ideal partner selection. The results suggest theoretical contributions to communication research on IOC and social networks, as well as implications for navigating multifaceted organizational identities and cross-sector partnerships.
Keywords
Communication researchers have focused on interorganizational collaboration (IOC) as a means of navigating complex environments and addressing grand challenges such as disasters and social injustice (Cooper & Shumate, 2012; Gray & Purdy, 2018). The culmination of prior IOC literature reveals that organizations tend to collaborate with others similar to themselves (i.e., homophily) to promote trust, minimize uncertainty and risks, ease communication, and improve legitimacy (Atouba, 2019). For instance, nonprofits tend to collaborate with other organizations of the same age (Atouba & Shumate, 2015), organizational type (Lai et al., 2019), issue domain (O’Brien et al., 2019), social mission (Pilny & Shumate, 2012), and funding source (Atouba & Shumate, 2010).
Scholars have identified two broad types of homophily: status and value (see Lazarsfeld & Merton, 1954; McPherson et al., 2001). Prior research has primarily focused on status homophily, suggesting three types (i.e., attribute-based, geography-based, institutional) common in IOC (see Atouba & Shumate, 2015). This attention to status homophily has been useful in exploring how status similarity influences relationship formation, but is less useful in exploring collaborative behavior. Value homophily shapes behavior, and many IOC studies conclude that compatible values and shared beliefs help initiate and facilitate collaborations (for a review, see Popp et al., 2014). In some ways, the difference between the related constructs of status and value homophily may be useful in understanding two other related constructs: networks and collaboration. Whereas networks typically refer to structures, collaboration refers to activity (Gazley & Guo, 2020). A focus on relationships prompted by status homophily and enacted by value homophily allows for further exploration of these partnerships. For instance, we might expect status similarity to drive the formation of collaborative ties, but value similarity to sustain long-term, successful IOC.
Scholars have suggested that organizational identity—central, enduring beliefs about who we are and what we do (Albert & Whetten, 1985; Ashforth et al., 2008)—may influence IOC (Barbour & James, 2015; Koschmann, 2013b; Lewis et al., 2010; Woo, 2019). The present research considers organizations’ aspirations for partner identity similarity as an example of value homophily. Specifically, this research focuses on faith-based organizations (FBOs) and explores value homophily induced by religiosity. We focus on FBOs because they are likely to have complex, nuanced identities, informed by their mission (what they do in service), and religious ideology (who they are as staff members and an organization) (McNamee, 2011; Sass, 2000).
The current research pursues two interrelated goals. First, we explore how organizations envision potential partners. Specifically, we seek to understand what characteristics FBOs ascribe to their ideal collaborative partners and partnerships, as exploring idealized partnerships alongside actual partnerships may offer avenues for better exploring IOC in general. Second, because compatible values and logics are often prioritized in interorganizational work (see Popp et al., 2014), we examine the relationship between value and status homophily in FBOs’ collaborative preferences. In surveying 141 FBOs across the United States, we explored how nuanced identities informs collaborative aspirations, and, specifically, offer implications for how organizations may navigate competing identities or priorities.
This study makes four theoretical contributions to communication research. First, we examine multiple forms of value homophily (i.e., service, staff, and organizational identity) and three types of status homophily (i.e., attribute-based, geography-based, institutional). Second, we contribute to a more nuanced understanding of the social network mechanisms that shape partner selection by exploring the relationship between value and status homophily in IOC. Third, in extending FBO religiosity (Ebaugh et al., 2006) to the IOC context, our findings shed light on the influence of multiple organizational identities in IOC. Together, these findings suggest a fourth contribution by offering preliminary understanding of how organizations envision their ideal partners—a topic that has not yet been explored in IOC or social network research.
The rest of the paper is organized as follows. First, we review research at the intersection of IOC and social networks, highlighting three types of status homophily that significantly shape IOC (Atouba & Shumate, 2015). Drawing on research in organizational identity and faith-based organizing, we then develop hypotheses to understand the interrelationship between value and status homophily. Next, we describe the method and the results of the analysis. Finally, we discuss the theoretical contributions and practical implications of this research.
IOC Explored from a Social Network Perspective: Status and Value Homophily
IOC describes a “set of communicative processes” in which multiple organizations work together to achieve common goals (Keyton et al., 2008, p. 381). Collaboration is a common strategy across nonprofit, business, and government sectors to tackle social challenges at high stakes and deliver public services (Gray & Purdy, 2018). However, IOC takes place in a complex and dynamic environment in which organizations may have overlapping members and multiple collaborative opportunities (Woo, 2021). To navigate environmental complexity, research suggests that homophily—“birds of a feather flock together” (McPherson et al., 2001, p. 415)—significantly influences IOC partner selection. Notably, homophily may reflect status or value.
We focus on FBOs because they are likely to have complex and nuanced identities, informed by mission and religious motivations. Moreover, research shows that FBOs often seek partners who share similar religious values and identities (Sapat et al., 2019). We next review research in FBOs and partnerships, highlighting the need to study idealized IOC partnerships.
FBOs and IOC: A Focus on Idealized Partnerships
Research on FBOs has increased over the past two decades, a rise that corresponds to a growing body of communication research in the intersection of religion and organizing (e.g., Charoensap-Kelly et al., 2020; McNamee, 2011). FBOs refer to religious organizations that have an “explicit [religious] goal and focus on providing services” (Bielefeld & Cleveland, 2013, p. 444). There are two major types of FBOs: (1) FBOs that primarily offer pastoral care and deity worship (e.g., churches and mosques), and (2) FBOs whose faith is central to their identity, such as universities, hospitals, and charities (see Olufowote & Matusitz, 2016).
FBOs represent a substantial sub-sector of American nonprofits, both in terms of the number of organizations and the resources they control. FBOs are estimated to comprise at least one-third of all nonprofits registered with the Internal Revenue Service (Ressler et al., 2020) and represent a combined economic value of annual revenue at $378 billion (Grim & Grim, 2016). Recent statistics show the dominance of FBOs in philanthropy (Zinsmeister, 2019) and the largest increase in revenue and expenses compared to other types of nonprofits from 2006 to 2016 (National Center for Charitable Statistics, 2020). Among all types of nonprofits, FBOs are typically regarded for their ability to reach underserved and hard-to-reach populations, attributes that make them powerful actors in social service delivery and emergency assistance (Bielefeld & Cleveland, 2013; Kearns et al., 2005; Stritt, 2008). Beyond the United States, FBOs also play pivotal roles in addressing global societal challenges such as natural disasters (Sapat et al., 2019), human trafficking (Koschmann, 2013a), and public health crises (Olufowote, 2016).
FBOs may collaborate with other organizations for the same reasons (e.g., legitimacy, resources, funding, efficiency) that other nonprofits collaborate (Clerkin & Grønbjerg, 2007; Ebaugh et al., 2007). However, FBOs often grapple with various communicative tensions in their collaboration (Koschmann, 2013a; Olufowote, 2016). Despite their extensive involvement in social service delivery, research reveals that FBOs are less likely to partner with secular agencies without a faith-related mission or affiliation (Fu & Cooper, 2021; Kearns et al., 2005).
The research is clear in that FBOs are less likely to collaborate with secular agencies than other types of nonprofits, but less clear as to whether this is an intentional strategy on behalf of FBOs or a consequence of being unable to reach their collaborative aspirations because of environmental or organizational limitations. In this research, we offer several reasons as to why a focus on FBOs’ idealized partnerships—what FBOs aspire to achieve through collaboration—may inform the roles of and, relationship between, value and status homophily.
First, prior research has suggested nonprofits have trouble realizing their collaborative aspirations, hampered by environmental constraints, organizational capacity, and funder mandates (e.g., Cooper & Shumate, 2012). But we suggest that exploring idealized partnerships alongside actual partnerships may offer avenues for better exploring collaboration in general. If, for example, networks serve as “a set of rules and resources actors draw upon in accomplishing communicative interaction” (Corman & Scott, 1994, p. 172), then prior relationships may serve as influential factors on future collaborative behavior rather than just environmental factors. The relationship between actual and idealized collaboration may be particularly relevant to FBOs, which are often expected to provide social services to the local community but may be limited both by their operational capacity as well as religious identity (Fu et al., 2021).
Second, from a social network theory perspective, there is an important distinction between enacted and aspirational collaboration. Because IOC is associated with investment in time, energy, and resources, there exists “a tendency in the nonprofit world to idealize collaboration, maximizing the perceived benefits, and minimizing the potential costs of engaging in resource-sharing activities” (Witesman & Heiss, 2017, p. 1507). The reality of collaboration often falls short of expectation or does not live up to the ideal (Bouman, 2002). For instance, nonprofits may seek partners with higher status and more resources (i.e., resource dependency, Sapat et al., 2019), but such collaborations could not be materialized into a reality, for instance, because of a lack of reciprocity. The differences between enacted and aspirational collaborations are particularly salient for certain types of organizations such as FBOs, who are relatively conservative in IOC intentionally or otherwise constrained by resources or environment.
Last but not least, we suggest that a study of FBOs’ idealized collaboration may shed some light on why we observe certain collaborative patterns. Research suggests that FBOs’ religious identity may impede their collaboration (Clerkin & Grønbjerg, 2007; Ebaugh et al., 2007; Olufowote, 2016). For instance, Koschmann (2013a) suggests that religious identity may create tensions in FBOs’ communication with partners. Howell-Moroney (2009) reveals that the moral stricture of FBO members poses challenges to their collaboration with government agencies. Given that prior research suggests FBOs engage in little collaborative activity compared to their nonprofit peers, we see less need for another study that reports FBOs’ current collaboration. Instead, we propose the study of FBO collaborative preferences and exploration of whether these preferences illuminate collaboration barriers. To that end, we ask two research questions (RQs):
FBOs, Religious Identity, and IOC
Organizational identity describes “a formal and informal point of reference in the representation of the organization” and “a complex of symbolic and material associations with the organization” (Kopaneva & Cheney, 2019, p. 485), reflecting the central, enduring set of beliefs, values, and meanings. Scholars typically consider organizational identity as multifacted, noting that most organizations have multiple identities (Kopaneva & Cheney, 2019; Lammers et al., 2013). Organizational identity is also multi-level, as seen in prior research (e.g., McNamee, 2011) that reveals inherent tensions in members’ individual and organizational identities.
Religiosity—the extent to which an organization is guided by religious values and principles—is a defining characteristic of FBOs. FBOs differ in how they incorporate religion or express their religious identity, as prior typologies suggest (see Bielefeld & Cleveland, 2013). Of these, the measure developed by Ebaugh et al. (2006, 2007) stands out for several reasons. First, it’s a quantitative measure as opposed to a typology drawn from qualitative data (e.g., Sider & Unruh, 2004). Second, rather than it being a measure developed to look at FBOs in isolation, this measure was developed in a collaborative context of faith-based social service coalitions comprised of congregations and FBOs. Third, consistent with a multidimensional view of organizational identity, this measure reveals three distinct categories of religious identity.
Service religiosity describes the way FBOs interact with clients using religious elements (e.g., prayer). For some FBOs, religiosity may be a requirement for personnel. Staff religiosity refers to the hiring practices, socialization, or staff actions that suggest religiosity among FBO personnel. Organizational religiosity describes the extent to which organizations publicly present religious identity. The distinction between service, staff, and organization religiosity echoes McNamee’s (2011) finding that members of FBOs usually negotiate identities at both the individual and organizational level. Similarly, Sass (2000) suggests that organizational spirituality may include individual, collective, and organizational spirituality.
Although religion has been used in homophily studies as an individual, demographic characteristic alongside other attributes such as race or gender (e.g., Lee et al., 2019), we suggest religion is better understood as a belief or attitude that guides organizational behavior. This is suggested by the measures developed by Ebaugh et al. (2006, 2007) that recognize religiosity not only in terms of staff beliefs, for example, but also organizational values and practices.
We posit that expectations for value similarity in organizational identities with ideal partners would affect their preferences for the other three types of status homophily. Specifically, if the religiosity between would-be collaborators is substantially different, the demands for other types of homophily become higher. Thus, if focal FBOs and their ideal collaboration partners are similar with respect to religious identity, FBOs may be less picky in other respects. In other words, FBOs prioritize shared religious identity over other ascribed status factors (e.g., attributes, geography, funders) in collaboration partners. This is because religiosity reflects organizational identity and values, whereas other types of status homophily reflect those aspects of an organization over which actors have less control. Based on this past review, we predict that:
Method
Sample and Procedure
Our research operates at the organizational level of analysis. Thus, we relied on organizational leaders as key informants for their organizations’ structure, identity, and collaboration activity (Atouba, 2019; Lai et al., 2019). Nonprofit executive directors or their equivalents participated in a survey (online or via mail as requested by the organization) between September 2016 and September 2017. Participants received a $20 gift card.
We mailed paper surveys to 2,116 FBOs in the National Taxonomy of Exempt Entities X religion-related category from the 2016 Urban Institute Nonprofit Database above the revenue of USD $100,000. We excluded religious congregations. Among the 1,974 FBOs with valid addresses, 395 returned surveys. However, 35 were blank surveys, 48 respondents indicated they did not engage in any social services, 29 considered themselves to be secular organizations, and 78 failed one of the two attention check questions. Among the remaining 205 responses, 141 were usable because their responses to actual and idealized collaboration questions were complete and they had engaged in collaborative activity in the past 3 years.
To assess nonresponse bias, we compared the characteristics (e.g., revenue, total assets, income) of the responding organizations with those of the nonresponding organizations. Results indicated that our final sample of 141 nonprofit organizations appears to be representative of the population of FBOs above the revenue of $100,000 in the United States.
On average, respondents had been working 8.54 years in their organization (SD = 7.53), indicating that participants were knowledgeable about their organization. Most respondents were executive directors (n = 99, 70.21%), though some identified as other C-suite executive, board president, or other leaders. Most of the responding organizations were Christian organizations (n = 107, 77.54%), which reflects the majority of U.S. FBOs (see Bielefeld & Cleveland, 2013). All but five states (Delaware, Hawaii, Maine, Nevada, Rhode Island) were represented in the sample, with Illinois (n = 12, 8.51%), Ohio (n = 11, 7.80%), California (n = 8, 5.67%), New York (n = 8, 5.67%), and Texas (n = 7, 4.96%) most frequently represented.
Dependent Variables (DVs)
During the survey, we asked each respondent to think about an ideal collaborative partnership with another organization. Our prompt was: Imagine that you could have any other organization partner with you in relation to your organization’s social service program. Regardless of whether it is feasible in reality, please think of the characteristics of an organization would be most excited to work with. Using this imagined collaborative partnership, please answer the following questions.
Attribute-based homophily
We asked each organizational leader to indicate the level of organizational revenue of their ideal collaboration partner (1 = less than $100,000, 2 = $100,000–$250,000, 3 = $250,000–$500,000, 4 = $500,000–$1 million, 5 = $1–$10 million, 6 = $10–$50 million, 7 = $10–$100 million, 8 = more than $100 million). Using the same scale, we also asked each respondent to report their organization’s revenue. We then compared the level of revenue of responding organizations and their ideal collaboration partner’s. If the level of revenue was the same, we coded this variable as 1; otherwise, we coded it as 0. We used a binary variable to remain consistent with the DV in testing H2 and H3 (see below).
Geography-based homophily
We asked each respondent to indicate whether they would like to be located in the same geographic location with their ideal partner in which they provide social services (1 = same geographic location, 0 = different geographic location).
Institutional homophily
We asked each respondent to indicate whether they would like their ideal partner to have the same funder with their organization (1) or not (0).
Independent Variables
To quantify the amount of difference in their religiosity, we focus on the absolute difference in an FBO and their ideal partner’s degree of religiosity. Similarly, social network research often uses the absolute difference in continuous variables between a partnership dyad to test homophily theory in Exponential Random Graph Modeling (see Handcock et al., 2003).
Difference in service religiosity
This variable evaluated the absolute differences in an FBO’s and its ideal collaboration partner’s degree of service religiosity. We first asked each respondent to evaluate the degree of their organization’s religious commitments in their social services using the 10 items developed by Ebaugh et al. (2006, 2007). It was evaluated on a scale of 1 to 5 where 1 = not important at all, 5 = one of the most important (M = 3.08, SD = 1.00, α = .90). In the last section of the survey, we asked respondents to evaluate the preferred degree of their ideal collaboration partners’ service religiosity using these 10 items (M = 2.68, SD = 1.15, α = .95). Sample items included “distribute religious materials to clients,” “pray with groups of clients,” and “programs require religious conversion.” We then calculated the absolute difference in the responding organization’s and their ideal collaboration partner’s service religiosity.
Difference in staff religiosity
Using the same approach, we first asked each respondent to evaluate the degree of their organization’s religious commitments in their staff members using the five items developed by Ebaugh and colleagues (M = 3.74, SD = 0.96, α = .78). Sample items included “pray at staff meetings,” “favor religious job candidate,” and “put religious principles into action.” Later, we asked respondents to evaluate the preferred degree of their ideal partners’ staff religiosity using the same five items (M = 3.53, SD = 1.13, α = .88). We then calculated the absolute difference in each FBO’s and their ideal partner’s staff religiosity.
Difference in organizational religiosity
Using the three items developed by Ebaugh et al., we asked each respondent to evaluate the degree of their organization’s religious commitments in organizational identity and mission (M = 3.06, SD = 1.10, α = .60). Sample items included “religiously explicit mission statement” and “sacred images in public spaces.” Although the internal reliability was marginally acceptable for small scales with only three items (Robinson et al., 1999), the three items are expected to hold together because they signal the symbolic aspect of religiosity to the external audience. Later, we asked respondents to evaluate the degree of their ideal partner’s organizational religiosity (M = 2.42, SD = 1.13, α = .75). We then calculated the absolute difference in the focal organization’s and their ideal partner’s organizational religiosity.
Control Variables
We controlled for several organizational attributes that influence IOC (see Table 1 for measures). First, we accounted for the number of full- and part-time employees (i.e., employee size) using an ordinal 5-point variable. Second, we measured revenue (in USD) with an ordinal 8-point variable. Third, organizational age (logged) was a continuous variable indicating the number of years since the organization’s founding. Fourth, percentage of government funding (logged), from 0% to 100%, was controlled because this funding drives FBOs’ collaboration with other organizations, in particular government agencies (Clerkin & Grønbjerg, 2007).
Descriptive Statistics and Correlation Matrix.
Note. Employee size: 1 = fewer than 10, 2 = 11 to 50, 3 = 51 to 100, 4 = 101 to 500, 5 = more than 500.
Revenue: 1 = <$100,000, 2 = $100,000 to $250,000, 3 = $250,000 to $500,000, 4 = $500,000 to $1 million, 5 = $1 to $10 million, 6 = $10 to $50 million, 7 = $10 to $100 million, 8 = more than $100 million.
p < .05. **p < .01.
Finally, we used Kearns et al.’s (2005) 13 categories of social services and asked each respondent to indicate whether their organization worked in each of the categories (respondents can select all that apply). The seven most popular areas within our sample were public or societal benefit (n = 62, 43.97%), youth development (n = 53, 37.59%), human services (n = 48, 34.04%), civil rights, social action, or advocacy (n = 41, 29.08%), food, agriculture, or nutrition (n = 42, 29.79%), community improvement or capacity building (n = 42, 31.91%), and mental health or crisis intervention (n = 33, 23.40%). Other six categories had limited representation. We thus created eight binary variables, including other, to indicate an FBO’s social mission.
This research is based on the presumption that the patterns of organizations’ prior and current collaboration experience would influence how they envision their ideal collaboration (Atouba, 2019; Gazley, 2010). For instance, a prior negative experience might lead an organization to try to avoid collaborators or collaborative relationships with similar characteristics in the future. We controlled for two aspects of existing collaboration that may shape ideal collaboration preference: (1) average tie strength and (2) collaboration type diversity.
Average tie strength
Using the name generator approach, we asked participants to name up to 10 current or past collaboration partners and indicate the tie strength with each named partner using Hansen’s (1999) two-item scale, comprised of frequency of communication and relational closeness. For frequency of communication, we asked respondents to identify how frequently people in their organization interacted with their partners on average over the past 3 years. For relational closeness, we asked each respondent to identify how close these working relationships were, where 1 = distant, 4 = somewhat loose, and 7 = very close. We then calculated the average tie strength, based on an average of the frequency of communication and emotional closeness (r = .64, p < .001), across all collaboration partners, for each organization. Average tie strength was controlled for because strong and weak ties may create different experience for organizations, hence influencing their choice of ideal partners (Popp et al., 2014).
Collaboration type diversity
Because IOCs vary in their goals, resources, and communication (Fu et al., 2019) and different types of collaboration may bring unique benefits and challenges to the focal organization (Powell et al., 1996), we used collaboration type diversity as a control variable that may influence their choice of idealized partners. For each existing partner listed using the name generator approach, we defined four types of collaboration (Snavely & Tracy, 2000) and asked respondents to indicate the type of collaboration for each listed partnership. We then calculated the number of unique types of collaborations for each organization (min = 1, max = 4). Table 1 presents the descriptive statistics and pairwise correlations of all the study variables and the operationalization of control variables.
Analysis
We used binary logistic regression to examine whether FBOs who report greater difference in religiosity than their ideal partners would prefer ideal partners that are similar to them. The variance inflation factors for all the variables ranged from 1.21 to 2.87 for all three logistic regression models, indicating that multicollinearity was not an issue.
Results
Attributes of FBOs’ Idealized Collaboration Partner (RQ1 & RQ2)
Table 2 presents the organizational attributes (RQ1, i.e., who is an ideal partner) and partnership attributes (RQ2, i.e., what is an ideal partnership) of FBOs’ idealized collaboration.
Organizational and Partnership Attributes of FBOs’ Idealized Collaboration.
Organizational attributes
Most FBOs prefer FBO partners (n = 104, 87.50%) rather than congregations or secular organizations. Overall, FBOs prefer relatively younger (i.e., <30 years old) collaborative partners (n = 102, 77.87%). As for organizational size, FBOs prefer organizations with fewer than 50 employees (n = 107, 81.68%) but prefer those that have at least $250,000 yearly revenue (n = 106, 80.90%). FBOs also prefer their partners to be co-located in the same geographic location (n = 94, 69.12%) but with different funders (n = 119, 84.40%). More respondents prefer their partners to provide services at the city (n = 58, 41.13%) and county level (n = 57, 40.43%) than other categories such as state and national levels. Notably, FBOs prefer to have significantly higher service (MFBO = 3.08, Mpartner = 2.68, t = 4.91, p < .001), staff (MFBO = 3.74, Mpartner = 3.53, t = 2.44, p = .008), and organizational (MFBO = 3.06, Mpartner = 2.42, t = 6.97, p < .001) religiosity than their ideal partners.
Ideal partnership attributes
As for the type of collaboration, about half of the sample prefer interorganizational planning (n = 64, 46.38%), followed by organizational integration (n = 51, 36.96%), community planning coalitions (n = 14, 10.14%), and case management (n = 9, 6.52%). The average frequency of communication was 4.44 (SD = 1.67), indicating a preference for relatively frequent communication with ideal partners, approximately 2 to 4 times per month. Similarly, the average emotional closeness was 5.23 (SD = 1.06), indicating a preference for closer relationship. Taken together, these findings suggest that FBOs preferred to foster strong ties with their collaboration partners (M = 4.83, SD = 1.13).
Predicting Status Homophily in Idealized IOC Partnerships (H1-H3)
Our results indicate that similarity in different dimensions of religiosity is related to different types of status homophily in FBOs’ choice of ideal collaboration partner. Table 3 presents the results for predicting attribute-based, geography-based, and institutional homophily, respectively. Figure 1 graphically illustrates the main findings for H1 to H3.
Binary Logistic Regression Models Predicting Attribute-Based, Geography-Based, and Institutional Homophily.
Note. Standard error in parentheses.
p < .05. **p < .01.
Summary of results: Characteristics of FBOs’ ideal collaboration partner.
H1 stated that FBOs reporting greater differences in (a) service, (b) staff, and (c) organizational religiosity than ideal partners prefer to collaborate with organizations that share similar organizational attributes. Contrary to our hypotheses, results indicated FBOs reporting greater difference in organizational religiosity than ideal partner were less likely to collaborate with organizations of the same revenue bracket (B = −0.92, SE = 0.36, p = .01). Thus, H1c was rejected. The odds ratio of the coefficient, calculated by taking exponential, was 0.40, indicating that after controlling for other factors, one unit increase in difference in organizational religiosity yields 60% lower odds of collaborating with organizations of the same revenue bracket. H1a (B = 0.28, SE = 0.39, p = .47) and H1b (B = 0.08, SE = 0.41, p = .84) were not supported. In combination of these non-significant results, H1 was rejected for attribute-based homophily.
According to H2, FBOs reporting greater difference in (a) service, (b) staff, and (c) organizational religiosity than ideal partners prefer to collaborate with organizations located in the same geographic area. Results indicate that if an FBO had greater difference in service religiosity than its ideal partner, it was more likely to seek partners that had similar geographic location (B = 0.97, SE = 0.48, p = .04, odds ratio = 2.64), supporting H2a. This indicates that one unit increase in service religiosity difference had 164% higher odds of collaborating with organizations that had similar geographic location, all other factors constant. H2b (B = −0.41, SE = 0.45, p = .36) and H2c (B = −0.12, SE = 0.36, p = .74) were not supported. Therefore, H2 was partially supported for geography-based homophily.
Finally, H3 posited that FBOs reporting greater difference in (a) service, (b) staff, and (c) organizational religiosity than ideal partners prefer to collaborate with organizations that share the same funder. The results suggested that if an FBO had greater difference in organizational religiosity than its ideal partner, it was less likely to seek organizational partners of the same funder (B = −2.61, SE = 1.18, p = .03, odds ratio = 0.07). This indicates that one unit increase in difference in organizational religiosity had 93% lower odds of collaborating with organizations that had the same funder, holding other factors constant. Hence, H3c was rejected. H3a (B = −1.24, SE = 1.08, p = .25) and H3b (B = −0.01, SE = 0.76, p = .99) were rejected. In combination of the non-significant results of H3a and H3b, H3 was rejected for institutional homophily.
Discussion
The purpose of this research is two-fold: (1) to explore how organizations envision ideal partners and ideal partnerships in IOC, and (2) to examine the influence of value homophily (i.e., staff, service, and organizational identity; Ebaugh et al., 2006, 2007) on three types of status homophily in organizations’ preference for ideal collaboration partner. That is, if organizations prefer a partner that shares values central to their identity, would their expectations for their ideal collaboration partners’ attribute, geographic, and institutional similarity (Atouba & Shumate, 2015) be less restrictive? Survey data from 141 U.S. FBOs revealed the organizational and partnership attributes of FBOs’ ideal collaboration. Moreover, logistic regression results suggested that FBOs that prefer similar levels of service religiosity with their ideal partners prefer difference in geographic location. By contrast, FBOs that prefer similar levels of organizational religiosity with their ideal partners also prefer similarity in organizational attributes and institutional funder. The sections below unpack each of these findings individually.
Attributes of FBOs’ Idealized Collaboration Partner and Partnerships
Overall, results suggest that FBOs’ observed preferences for collaboration partners are driven by the need to serve their communities—and that differing degrees of religiosity may be a deterrent even among organizations that share a faith-based orientation.
Organizational characteristics
Consistent with previous research, FBOs preferred to collaborate with other faith-based nonprofits as opposed to other agencies (Clerkin & Grønbjerg, 2007; Ebaugh et al., 2007). However, FBOs prefer to have significantly higher service, staff, and organizational religiosity than their ideal collaboration partners’. We suggest that this may because of a perception that increased religiosity creates more barriers to collaboration. The same challenges that religiosity poses to secular partners (Howell-Moroney, 2009; Koschmann, 2013a) may be present among FBOs with varying degrees of religiosity. Additionally, FBOs that are seen as more religious might also be perceived as having fewer financial resources.
Employee size and revenue, two most common indicators of organizational size (Jaskyte, 2013), suggest related—and possibly conflicting—collaboration preferences. FBOs prefer to collaborate with medium- and large-sized organizations who have at least $250,000 yearly income, suggesting financial stability and the availability of resources to share. However, FBOs also prefer partners to have fewer than 50 employees. The presence of financial resources and human resources often go hand-in-hand, but the results suggest an interest in financial resources without the complexity of coordinating with a much larger organization. Although financial resources are often associated with older organizations, FBOs indicated a preference for relatively younger partners. This is perhaps because they are perceived as dynamic; older organizations may exhibit structural inertia with established rules, practices, and procedures that hinder organizational change (Hannan & Freeman, 1984). Together, these findings suggest that FBOs may prefer smaller, nimble organizations that can respond more quickly to community-driven needs or hard-to-reach populations that FBOs typically serve (Kearns et al., 2005).
Overall, FBOs prefer to be co-located in the same geographic location with their ideal collaboration partner. The commonality in socio-economic, socio-cultural, and socio-political environment prompt organizations to confront similar issues and deal with similar stakeholders, which may promote familiarity, common interests, shared knowledge and experience of their locality, and ease of communication using a common language (Atouba & Shumate, 2015). Relatedly, FBOs prefer their collaboration partners to provide services at the city and county level, a finding consistent with previous observation that FBOs predominantly provide services to the local community at the county (Graddy, 2006) and city level (Pipes & Ebaugh, 2002).
Surprisingly, most FBOs prefer to not share the same funder with their ideal partners. Institutional homophily is based on information about partners’ relations, which may prompt organizations to conform to similar practices and external pressures (e.g., funder mandates). Although institutional homophily may help organizations maintain legitimacy and buffer against sanctions and risks (Atouba & Shumate, 2015), FBOs may fear that the same funding source that may create competition among partners. Alternatively, they may hope that these partners would expose them to new funding sources—which may be particularly important if FBOs tend to form partnerships at the city and county level and thus rely on the same pool of local funders.
Partnership characteristics
As for partnership type, most organizations prefer two types of collaboration with their ideal partner: interorganizational planning and organization integration. Compared to the two most prevalent forms of nonprofit collaboration (i.e., case management and community planning, see Snavely & Tracy, 2000), interorganizational planning and organization integration are less common in reality. In part, this is due to the fact that these two forms of IOC have a higher degree of resource sharing and service integration than other forms of IOC. Alternatively, these preferences may indicate a potential survival strategy, or a willingness on behalf of FBOs to work more closely with other agencies to navigate organizational or environmental turbulence. Organizational integration in the form of combining resources is one way for nonprofits to survive in a competitive environment; additionally, for some nonprofits, collaborative activity with a trusted partner is a stepping stone to exploring other forms of organizational restructuring (see Cooper & Maktoufi, 2019).
Finally, results suggested that FBOs prefer to foster relatively strong ties with their collaboration partners. Although weak ties provide opportunities to bridge organizations of different types and thus promote integration in the community (Granovetter, 1973), strong ties foster a sense of closure, trust, and commitment that facilitate resource sharing and goal achievement in an IOC context (Eng et al., 2012). Moreover, strong ties bring bonding social capital, resources that organizations have and can mobilize at their disposal, that are conducive for long-term organizational development (Schneider, 2009). Consequently, FBOs prefer to foster strong ties with their ideal collaboration partner to improve service delivery and address community needs. Having explored findings relevant to FBOs’ idealized collaboration, we now turn to a discussion of the roles of value and status homophily in FBOs’ idealized partnerships.
Relationship between Value and Status Homophily in Idealized IOC
One of the unique contributions of this research is that we examine the degree to which value similarity in religious identity is related to other collaborative preferences. In doing so, we explore whether organizational identity is the most important factor in determining FBOs’ choice of ideal collaboration partners and illuminate the roles of value and status homophily in IOCs.
We find that when FBOs found ideal partners that share similar service religiosity, they would not demand similar geographic location, hence homophily in other organizational status or attributes is no longer a salient barrier to their collaboration. This finding suggests that FBOs prioritize shared service religiosity over other ascribed attributes in collaboration partners. This finding complements previous research which suggests that FBOs with more explicit religious policies and practices regarding client and staff interactions are less likely to engage in cross-sector collaboration (Ebaugh et al., 2007; Fu et al., 2021). In short, when similar service identity is established, other similarities in status matter to a lesser extent.
Contrary to our hypotheses, we find that FBOs that demand similar organizational religiosity—the extent to which an FBO is faith-based and how they present themselves publicly—in ideal partners also demand similar revenue bracket and same funder. Rather than forming a common basis of agreement that enables partners to diverge on other attributes, organizational religiosity is the common ground on which FBOs seek to build other similarities. This particularly restrictive approach suggests that FBOs may find it difficult to find desirable partners, providing one explanation for their tendency not to collaborate (Clerkin & Grønbjerg, 2007; Fu & Cooper, 2021). Organizational leaders may be more protective of their organizational identity or perceive that their mission does not benefit from collaborative input. This surprising finding suggests that organizational religiosity—public image and symbolic communication—“rules it all” in IOC in the sense that it limits future collaboration by placing greater restrictions on potential partners.
The differences between service and organizational religiosity in shaping FBOs’ partner preference is consistent with a multidimensional view of organizational identity (e.g., Kopaneva & Cheney, 2019; Lammers et al., 2013). While service identity is more about an organization’s substantive activity core to its social mission, organizational identity is about the symbolic communication to external audiences (Ebaugh et al., 2006). For instance, Ebaugh et al. (2006) found that higher organizational religiosity is related to more explicitly religious names of FBO coalitions, indicating the symbolic nature of organizational religiosity. Our findings indicating FBOs that are selective regarding their ideal partners’ expressive religious identity are also selective in organizational attributes and institutional funder suggest that FBOs seek reassurance of partner compatibility—and are mindful of what these partnerships publicly represent.
However, we did not find a relationship between staff religiosity and status homophily. Prior research suggests that religious beliefs and practices of FBO personnel may motivate FBOs’ formation of cross-sector partnerships (Fu et al., 2021), and that faith-based partnerships are heavily based in personal networks (Clerkin & Grønbjerg, 2007). However, staff religiosity may hamper FBOs’ cross-sector collaboration (Ebaugh et al., 2006, 2007), suggesting that religious individuals in these organizations knowingly or unknowingly avoid working with others who do not share their religious beliefs. Consequently, faith-based programs may promote homogeneity more than they do to encourage diversity (Wuthnow, 2004). In sum, our findings suggest a more complex picture of how staff identity influences status homophily in IOC.
In summary, organizational and service religiosity represent “who are we” and “how do we do our work,” respectively. Hence, organizational and service identity matter with respect to the formation of IOC. In contrast, the degree to which employees (i.e., staff identity) are religious matter to a lesser extent. Taken together, our findings highlight the influence of organizational- and service-identity induced value homophily on status homophily in IOC.
It is also possible that value and status homophily are related to common exogenous factors and shaped by external environment. Moreover, organizations operating within the same environment—as FBOs tend to do (Graddy, 2006; Pipes & Ebaugh, 2002)—may share similar characteristics. For instance, nonprofits working in the same or adjacent resource-deprived communities with relatively homogeneous religious beliefs, all have low revenue, and rely on one another to help their target audience. Resource dependency may also play a role if FBOs are largely bounded by a geographic environment (Cooper & Cooper, 2021). Similarly, a bona fide network perspective of IOC (Cooper & Shumate, 2012) suggests that IOCs are embedded in a complex environment that enables some configurations and constrains others.
Theoretical Contributions
This study makes four theoretical contributions to communication research. First, in examining the roles of three types of value homophily (i.e., service, staff, and organizational identity) and three types of status homophily (i.e., attribute-based, geography-based, institutional), this research provides a more nuanced understanding of the social network mechanisms that influence collaborative choices in nonprofit organizations generally and FBOs more specifically. Our application of Atouba and Shumate’s (2015) homophily framework contributes to a growing body of research on the antecedents to IOC (Gazley & Guo, 2020).
Second, in studying the relationship between status and value homophily, this research helps gain a fuller understanding of partner selection in IOC. Because homophily is associated with trust and communication effectiveness (Atouba, 2019), we anticipated that organizations with greater differences in identities might seek collaborative partners who were similar to them in other ways. However, this turned out not to always be the case. In considering partners who differ in fundamental ways such as service identity, an FBO may indicate that other differences in ascribed attributes are relatively non-significant. By contrast, an FBO that is selective in terms of organizational identity may be selective with regards to other ascribed attributes, suggesting a “picky” organization with several criteria for would-be partners. The differential impact of service and organizational identity indicates the symbolic versus substantive aspects of organizational identity. Although this research represents a specific IOC context, our findings regarding idealized collaboration echoes prior research that organizations make collaborative decisions in a complex and dynamic environment, often with future collaborative opportunities in mind (Woo, 2021).
Third, in extending FBO religiosity (Ebaugh et al., 2006, 2007) to the IOC context, our findings illustrate the influence of multiple organizational identities in IOC. Although “religion” is used often with respect to studies of homophily among individuals (e.g., Lee et al., 2019), we have expanded upon it here to refer to organizational identities among FBOs. Extending communication research that focuses on the influence of organizational and collective identity on IOC (e.g., Barbour & James, 2015; Koschmann, 2013b; Lewis et al., 2010; Woo, 2019), our research incorporates a nuanced understanding of FBO religious identity, highlighting the multidimensional, multi-level nature of organizational identity for IOC. Our findings suggest the question of who we are and what we do as an organization (Ashforth et al., 2008), as expressed through service and organizational identity, influenced FBOs’ preference for ideal collaboration partners. This suggests that future research within communication on IOC and faith-based organizing may benefit from a better understanding of organizational and service identity.
Fourth, to the best of our knowledge, this research is the first to study idealized IOC. In asking respondents to identify the attributes of their ideal partners and partnerships, this study helps inform scholarship on the antecedents and barriers to IOC. Research suggests nonprofits have trouble realizing their collaborative aspirations, hampered by environmental constraints, organizational capacity, and funder mandates (Cooper & Shumate, 2012). Because networks function as rules that inform how actors interact with others (Corman & Scott, 1994), we argue that exploring idealized partnerships may offer avenues for better exploring IOC in general.
These findings also offer insights for practice as they suggest how organizational leaders continually negotiate and evaluate multi-faceted identities alongside other attributes as they pursue collaborative opportunities. Findings may be useful to organizational leaders as well as funders as they seek to determine why some organizations are viewed as more desirable partners.
Limitations and Future Research
Although this research makes strides in further exploring the nuances of organizational identity and status homophily in shaping IOC, it is hampered by several limitations. Although we have followed the best practices in survey research to enhance response rate (e.g., sending follow-up surveys and reminder emails, giving participants an option to fill out the survey online), it is a relatively small sample with a relatively low response rate. This may be due to the multiple skip logic questions that disqualified many respondents. This may also be due to the response fatigue associated with name-generator network survey research (Perry et al., 2018). We anticipate that further research that builds upon our study will seek a larger sample.
We achieved less religious diversity among our sample of FBOs than we anticipated in that Christian organizations dominated the sample. While Christianity is the most common religion in the United States (Charoensap-Kelly et al., 2020), it by no means reflects the religious values or orientation of all FBOs. This characteristic of our sample limits generalizability to FBOs affiliated with other faiths. Further research should seek a more diverse sample. Additionally, due to the small number of items, the reliability of the organizational religiosity scale for responding organizations was marginally satisfactory.
This research provides a number of opportunities for future research. First, we focused on FBOs and their religious identities because of FBOs’ important role in the U.S. service delivery and the opportunity to explore the relationship between status and values homophily. Future research might apply this framework to organizations with other salient identities. We derived categories of staff, organizational, and service identity from measures of religiosity (e.g. Ebaugh et al., 2006, 2007), but given that organizational identity is inherently multi-level (McNamee, 2011) and multi-faceted (Lammers et al., 2013), we suggest that other types of organizations—and indeed, the researchers that study—may find that distinctions between staff, organizational, and service identities are useful or inform the development of additional measures of organizational identity for most organizations. Questions of prioritizing value or status similarity may be relevant to other mission driven organizations, including political or partisan organizations, organizations that serve particular populations or provide niche services, or organizations in which staff often identify with clients in terms of lived experiences.
Second, we only examined one measure within each when there are many possible measures in each type of homophily. Measures of attribute-based, geography-based, and institutional homophily could be further specified in future research. As opposed to using binary variables, future research may also ask respondents to indicate the extent to which they would like their ideal collaboration partners to be similar to them in various organizational attributes (e.g., age, employee size), geographic locations, or institutional relations (e.g., legitimacy status).
Third, future research may integrate multiple social network theories, such as homophily and resource dependency (Sapat et al., 2019), to further understand the mechanisms that shape collaborative partner selection in actual and idealized IOC.
Finally, this research suggests where collaborative aspirations meet reality. Our findings offer insight into how organizational attributes (e.g., social mission, age) and existing partnership attributes (i.e., diversity, strength of ties) influence an organization’s collaborative aspirations, areas that may be explored in further research to fully understand IOC. The correlation between existing partnership tie strength with ideal partnership tie strength is .44 (p < .001), the correlation between existing partnership tie strength with ideal partnership type diversity is .25 (p < .001), and the correlation between existing collaboration type diversity and ideal tie strength is .34 (p < .001). A positive relationship between existing collaboration type diversity and ideal tie strength, for instance, may be explained by the paradox of network embeddedness (Uzzi, 1997). Because actors accomplish their network aspirations via communicative activity (Corman & Scott, 1994), prior relationships may serve as influential factors on future collaborative behavior. Future research may provide a more nuanced understanding of how existing partner attribute (e.g., geographic location, revenue) and partnership attributes (e.g., length of relationship, scope of services) shape ideal partner and partnership envisioning.
Conclusion
In this study, we sought to understand (1) how organizations envision potential partners and partnerships, and (2) the roles of value and status homophily in FBOs’ idealized collaboration. This research provides a more nuanced understanding of service, staff, and organizational identity within the IOC context. It also sheds light on the relationship between value and status homophily in IOC, suggesting that service identity similarity is the most important factor shaping IOC. Our results thus suggest a more complex understanding of the antecedents to IOC than prior research typically suggests. Future research may integrate research on social networks and organizational identity to more rigorously examine the antecedents to IOC.
Footnotes
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 disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was generously supported by the National Science Foundation (SES-1264417) and the faculty start-up funding from the Rutgers University School of Communication and Information.
