Evolving Crisis and Changing Networks: NGOs’ Stakeholder Engagement on the Global Refugee Crisis

Abstract

Nongovernmental organizations (NGOs) increasingly utilize social media for strategic stakeholder engagement. This study proposes a network-oriented theoretical framework to understand how NGOs’ engagement with complex networks of stakeholders on the global refugee issue varies as the issue moves from low to high public attention stages. We draw from research on multistakeholder issue networks and issue niche theory and analyze a large-scale Twitter data set containing tweets from hundreds of organizations from more than 30 countries. This cross-national, longitudinal study tracks issue evolution and NGOs’ tie formation patterns among themselves and with complex stakeholders (i.e., government and media) as public attention to the refugee issue increases. The results of our exponential random graph models (ERGMs) show how cross-sector stakeholders interact dynamically and how different issue identities position NGOs uniquely in issue niches as the issue evolves. We also find that organizations’ country-level homophily influences tie formation. Theoretical and practical implications are discussed.

Keywords

multistakeholder issue network issue niche theory stakeholder engagement network evolution refugee crisis

Today, millions of nonprofit and nongovernmental organizations (NGOs hereafter) leverage social media tools in their daily operations (Campbell & Lambright, 2019; Guo & Saxton, 2018; Saxton & Waters, 2014). Prior works have documented how advocacy organizations use social media to gain public attention and mobilize social change (Guo & Saxton, 2014, 2018). Others show how human service organizations use social media to serve members (Campbell & Lambright, 2019) or fundraise (Bhati & McDonnell, 2020). Other studies assert that social media like Twitter are critical for managing NGOs’ complex relationships with stakeholders (Svensson et al., 2015; Xu & Saxton, 2019).

Previous studies suggest that the predominant approach to examining NGOs’ relationship building on social media focuses on how a few or a group of NGOs deploy stakeholder engagement strategies and tactics and examine the antecedents or outcomes of such communicative activities (Bhati & McDonnell, 2020; Campbell & Lambright, 2019). When it comes to complex social issues such as wicked problems,¹ NGOs often need to communicate and manage relationships with a community of stakeholders, who are interconnected through complex networks (Saffer, 2019; Saffer et al., 2018). We argue in this study that a network-oriented perspective is necessary to account for the complexities of how NGOs manage stakeholder relationships around issues with far-reaching social consequences.

In this study, we draw upon the multistakeholder network framework and issue niche theory to examine how an NGO community builds complex stakeholder relationships around the global refugee crisis on social media. The global refugee crisis is an example of a wicked problem that draws stakeholders from multiple social sectors. Although the refugee migration to Europe captivated international attention, the crisis has evolved to become a global issue. Refugees currently living in Europe only account for 17% of the global refugee population, whereas the situation is much drier in regions such as Africa (30%), Middle East (15%), and North Africa (26%) (United Nations High Commissioner for Refugees [UNHCR], 2021). An issue of this scale is impossible to be handled single-handedly by any government or organization; it requires the involvement of multistakeholder issue networks.

Multistakeholder issue networks form when institutions such as governments, corporations, IGOs (international governmental organizations), and NGOs connect through information exchange, project collaboration, and joint membership in advocacy coalitions around an issue (Roloff, 2008a, 2008b). To understand how stakeholders form networks in multistakeholder issue niches, we further introduce issue niche theory (Yang, 2020), which is a recent extension of the organization ecology theory (Aldrich & Ruef, 2006), to explain how issue evolution affects NGOs’ engagement with stakeholders. In this study, we argue that the multistakeholder issue network framework highlights the interconnectedness of issue stakeholders, whereas issue niche theory provides an explanation of how issue evolution changes relationships among issue stakeholders. Together, this network-oriented theoretical approach allows us to understand how the interorganizational relationships among issue stakeholders evolve and change as an issue evolves. Moreover, this approach helps policy makers and NGOs better anticipate the growth and decay of cross-sector networks and relationship building opportunities, and design policies and advocacy campaigns accordingly.

This study makes several contributions to the nonprofit and voluntary research literature. First, our study proposes and tests a theoretical framework to help understand NGOs’ complex stakeholder relationship networks on social media. Second, we compare how relationship building occurred in a multistakeholder issue network at low and high public attention issue stages (Heath & Palenchar, 2008). The longitudinal design reveals nuanced dynamics that would otherwise be unobservable if the refugee issue is conceived as a static event. Our findings indicate that issue evolution powerfully changes the patterns of relationship building for specialist and generalist NGOs and introduces new ways of considering contingency factors that shape tie formation opportunities in the process of niche partitioning (Margolin et al., 2015).

The rest of the article is organized as follows. First, we review the multistakeholder issue network framework and identify its connections with the issue niche theory. We then shift attention to issue niche theory and derive our hypotheses to guide further analysis. We then report findings from exponential random graph models (ERGMs) based on an analysis of interorganizational networks extracted from a large Twitter data set (2,413,155 tweets collected between 2013 and 2016). The theoretical and practical implications are also discussed.

Multistakeholder Issue Networks and Wicked Global Issues

A growing number of recent studies have examined NGOs’ interorganizational networks (Doerfel & Taylor, 2017; Margolin et al., 2015). These studies have explored the ways through which NGOs’ relationship networks are structured or what network positions are advantageous to NGOs. Most studies focus exclusively on the NGOs’ community as if their communication occurs in a social vacuum. Yet, often the issues these NGOs attempt to address require organizational responses from other sectors such as government and business.

Recognizing the interdependence among organizational actors across sectors is a critical development in the network approach to studying interorganizational networks and governance structures (Klijn et al., 2010). Previous network studies have assumed that organizations such as governments, NGOs, or firms “can solve the problem alone or that they can impose their solution on other actors” (Klijn & Koppenjan, 2000, p. 139). Increasingly, scholars, as well as actors from civil society and business sectors, have recognized that organizations face and must address shared challenges despite various obstacles (McGuire & Agranoff, 2011; Smith, 2005, 2008). Businesses in particular have become sensitive to stakeholder demands on complex issues; subsequently, they are increasingly engaging other types of organizations as a means of cooperation (Moog et al., 2015). Trends such as these have given rise to interorganizational networks that form around issues.

Issue-focused networks are quite distinct but come in many different forms that vary in the deliberateness of interactions. As Klijn and colleagues describe, issue-focused networks

emerge out of a necessity to interact and are, on the one hand, consciously planned in the sense that actors deliberately interact and attempt to structure these interactions with organizations and rules, but on the other hand, are also unplanned as a result of coincidental interactions and strategies and previously created rules. (p. 1063)

For instance, networked governance structures such as multistakeholder partnerships (Bäckstrand, 2006), multistakeholder regulation (O’Rourke, 2006), or multistakeholder initiatives (Moog et al., 2015) form from cross-sector interorganizational relationships where network members voluntarily address pressing political and/or social issue(s) (Le Ber & Branzei, 2010). Even advocacy coalitions can lean toward the deliberative end with their intention of achieving policy goals or outcome. Unplanned networks like those that unfold in discussion on social media, as studied here, can be less calculated and without particular goals.

Multistakeholder issue networks can be broadly conceptualized as a set of “actors from civil society, business and governmental institutions [that] come together in order to find a common approach to an issue that affects them all” (Roloff, 2008a, p. 238). According to Roloff (2008b), any group or individual who can affect or is affected by the solution to the problem being addressed by the network are stakeholders. Rather than take interest in the governance processes or the network management outcome, a multistakeholder issue network provides a lens for understanding how various actors are stakeholders to an issue.

In the case of the global refugee crisis, as the refugee population grew and the public attention on this issue gradually intensified between 2013 and 2016, many government agencies scrambled to address the humanitarian crisis while being pressured by rising nationalism movements to take measured actions. Many NGOs, especially refugee and human rights NGOs, were motivated by their mission to alleviate the humanitarian disasters and were working to advocate and mobilize resources to improve refugee lives. In addition to government agencies and NGOs, another major stakeholder group is media, who are driven by public attitudes to cover the issue and are eager to bring firsthand information to audiences.

As these actors interact, they form multistakeholder issue networks that are often complex and not confined to national borders. A few high-profile examples include the UN Global Compact, Forest Stewardship Council (FSC), Alliance for Water Stewardship, and the Roundtable on Sustainable Palm Oil (Dentoni et al., 2018). Such networks create a space to foster global collaborations and dialogues among businesses, civil society, and government agencies, and some have framed it as a novel form of governance structure for organizational alliances (Moog et al., 2015). We argue that multistakeholder issue networks hold great conceptual utility for advancing our understanding of how societies respond to wicked problems (Alford & Head, 2017).

Previously, Doerfel and Taylor (2017) examined the democracy NGOs in Croatian civil society and Shumate and colleagues have studied multistakeholder issue networks of NGOs (O’Brien et al., 2019; Shumate et al., 2005). These studies focus primarily on ties among NGOs/IGOs. Expanding the research scope, Saffer (2019) used the multistakeholder issue network approach to examine NGOs, business, government agencies, and other institutions that had formed around the issue of sustainable sanitation. Findings showed that actors in a multistakeholder issue network may have competing interests, their familiarity with others in the network, or even the common procedures for addressing issues likely vary. Nonetheless, a long line of research suggests multistakeholder issue networks are a common approach for organizations of various types to respond to current and future pressing issues.

Advancing the Multistakeholder Issue Network Approach

Despite the value of the multistakeholder network framework, we recognize three gaps within the current scholarship. First, the multistakeholder issue network assumes the issue at the center of a network is static. In reality, as scholars of issue management have known for decades (Heath & Palenchar, 2008), issues evolve. As such, without properly accounting for the effect of issue evolution, the framework may not fully realize its theoretical potential of placing issues at the center of theorization. Take, for example, Moog and colleagues’ (2015) examination of the FSC. This multistakeholder issue network brought together institutions to set standards in forestry and eco-labeling of timber products. Their cross-sectional study only considered how the issue and the network existed at one moment in time.

Second, prior work using the multistakeholder issue network approach has not considered how media are important stakeholders, especially for controversial issues. In reality, most publics see an issue through the media’s lens. Media can bridge between institutional actors and the public. Unlike the media, NGOs/IGOs, corporations, or government agencies cannot raise awareness about an issue to scale. Even on social media, traditional media accounts still generally have a larger reach than most other types of organizations. The global refugee issue is one such example where the refugee crisis, for most Western audiences, has mostly been an issue that they learn from the news. Furthermore, in a transnational digital environment, media form networks to facilitate the transmission of information across borders, build ideology coalition, and exhibit perspectives and identities (Heft et al., 2020; Knüpfer & Entman, 2018). Media coverage has brought public attention to the issue, constructed the issue through framing, and subsequently influenced the issue evolution.

Third, the multistakeholder issue network approach lacks theoretical utility. In other words, the framework describes unique phenomena without suggesting under what conditions organizations would form relationships. Many multistakeholder issue network studies simply reveal the diversity of connections among organizations (Roloff, 2008a, 2008b). Recognizing the value of the multistakeholder issue network framework along with its limitations, we advance the approach by integrating a theory that addresses the above-mentioned gaps.

An Ecological View on Multistakeholder Issue Network

As discussed earlier, at the center of multistakeholder issue networks are socially significant issues such as climate change or the global refugee crisis (Roloff, 2008a, 2008b; Saffer, 2019; Saffer et al., 2018). To address these issues, actors from civil society, business, and governmental institutions often come together in search of common solutions (Dentoni et al., 2018). Issue networks that form around either climate change or the refugee crisis have porous boundaries—the challenges these issues present are not contained to specific geopolitical or sectoral boundaries. The current study recognizes that issues rise and fall in ways that may have considerable influence on the operational environment of network members.

As stakeholder organizations begin interacting and collaborating around an issue, an organizational community starts to take form. In this study, we draw upon issue niche theory to guide our analysis. Issue niche theory considers stakeholder organizations of an issue as an organizational community (Yang, 2020). According to the theory, an issue niche is co-created by a social issue and stakeholders with stakes in the issue, as “an issue niche is formed when a number of issue stakeholders with various identities and ideologies attach significance to a situation or perceived problem, interpret the nature of the issue, and further mobilize resources to provide issue solutions” (Yang, 2020, p. 43). According to this definition, a multistakeholder issue network is an organizational community working on an overlapping issue niche (Monge et al., 2008). This ecological view recognizes the interdependence of issue stakeholders and also recognizes that their operations and relationships are influenced by the shared issue niche.

The issue niche theory conceptualizes civil society organizations as natural habitants of an issue niche because civil society actors often operate on one or a few issues. NGOs often utilize a range of advocacy tactics (e.g., research, protest, social media campaigns) to advance their perspectives on issues. Moreover, the success of their operations often depend on how well they can attract other issue stakeholders’ attention and mobilize resources and support (Roloff, 2008a). In comparison, corporations or government agencies have a set of core economic or governance functions and they tend to enter an issue niche as a means of strategic issues management (Heath & Palenchar, 2008). Corporations and government agencies would assess the issue’s relevance and impact on their own interest/stakes and choose to work with partners that best suit their needs and interests. As such, the interesting question the issue niche theory attempts to answer is how issue evolution changes corporations or government agencies’ likelihood to collaborate with NGOs.

It is necessary to note that issue niche theory shares similarities and important differences with other theories that explain networks formed around social issues. For example, the advocacy coalition framework (ACF) explains how networks of policy advocacy coalitions interact and how such networks shape policy changes (Sabatier & Weible, 2007). While both issue niche theory and ACF recognize actors’ ability to shape issue or policy development through networked interactions, the focus of issue niche theory is on how niche width and positions condition organizations’ tie formation patterns. In comparison, ACF focuses on how the belief systems of policy elites shape networks and the conditions under which policy-oriented learning across coalitions can occur. Another relevant framework is the concept of epistemic community, which refers to “a network of professionals with recognized expertise and competence in a particular domain and an authoritative claim to policy-relevant knowledge within that domain or issue-area” (Haas, 1992, p. 3). Both issue niche theory and the concept of epistemic community suggest that NGOs or members of epistemic communities could identify and provide interpretation of social issues to influence decision-making. However, the issue niche theory describes how a wide variety of actors come together to define an issue, negotiate solutions, and collaborate to implement such solutions, whereas epistemic communities focus on how such communities arise in response to uncertainties and how information diffusion (rather than cooperation) occur among experts and decision makers. As such, the issue niche theory is uniquely positioned to provide an ecological perspective for understanding how NGOs and other issue stakeholders interact with each other surrounding the global refugee crisis.

Moreover, the theory proposes that two factors, issue niche width, which is defined as the cumulative amount of public attention associated with an issue, and NGOs’ issue identities affect niche partitioning and tie formation among issue stakeholders. Each is further discussed below.

Issue Niche Width

Public attention is one of the most valuable and scarce resources in an information-based economy. It is “the extent to which multiple individuals are exposed to cultural products across space and/or time” (Webster, 2011, p. 45). For advocacy organizations who seek social change, they must first capture a public’s attention before that public can response to calls of actions. Publics’ attention is a prerequisite for meaningful social change. Issue niche theory draws on this understanding of public attention to assert that the width of an issue niche depends on the cumulative amount of public attention it attracts. Yang (2020) explains that the issue niche width “is shaped and reshaped in the process of communication, largely depending on the issue’s evolution and public attention” (p. 44). Thus, any discussion of issue niche width must consider the life cycle of an issue given that each stage can attract varying levels of public attention.

Crable and Vibbert (1985) proposed a model of issue life cycle to describe how an issue evolves over distinctive stages, and how each stage associates with different amount of public attention. This model begins by positioning an issue in the potential stage. At the potential stage, a few individuals or groups start to recognize or problematize an issue. As the recognition grows and the problematization becomes more well known, the issue then evolves to the imminent stage where it can gain legitimacy among a wider group of stakeholders. Here the issue width widens. Next, at the current stage, information about the issue circulates through media ecosystems and enters public discourse and conversation. The issue niche widens even further. The issue may further evolve to reach the fourth stage, the critical stage, when publics form their own opinions and demand formal institutions (e.g., legislators) take actions on the issue to address it. Or, at any stage of the life cycle, an issue may go dormant. This can happen when policy or legal decisions on an issue have been made. Issues can also go dormant if public attention is lost and the issue niche contracts. Not all issues go through the five stages, but research has consistently recorded the cyclical nature of issue-related public attention (Nigam & Ocasio, 2010). As such, at the minimum, an issue goes through low and high public attention stages.

The stages of an issue life cycle can also be understood broadly as having high or low public attention. The prominent and critical stages are associated with high levels of public attention. The potential, imminent, and dormant stages are associated with low levels of public attention (Heath & Palenchar, 2008). In this study, we focus on how relationship building patterns differ between low and high public attention stages.

The issue niche theory predicts that as an issue evolves, the associated multistakeholder issue network also transforms dynamically in size and composition (Yang, 2020). Given that issue niche width indicates the amount of resources potentially available for mobilization. No matter how resourceful the multistakeholder issue network is, it often still needs to mobilize resources beyond involved stakeholders (Roloff, 2008a, 2008b). When an issue is at the high public attention stage, the general publics and decision makers are more likely to be exposed to the issue and subsequently face demands to address the issue. Such exposure and heightened public awareness would make resource mobilization much easier. In the United States, for example, within 10 days of the first case of COVID-19 with unknown source of infection was confirmed, the government approved an US$8.3 billion emergency fund. In comparison, when an issue is at the low public attention stage, the issue may not receive much discussion and attention, which may constrain issue stakeholders’ opportunities for resource mobilization.

The organizational community formed around an issue niche is different from a community formed around a resource niche (Yang, 2020). Organizations often enter partnerships either in response to crises, adapt to emergent problems, or in anticipation of potential concerns that may occur (Heath & Palenchar, 2008). In these processes, decision-making is often subject to the influence of stakeholder expectations and public opinion. In other words, corporations and governments not only build alliances with each other or NGOs for resources accesses (Austin & Seitanidi, 2012); they often build and communicate such relationships for symbolic, representational values. They want to show to the general publics or stakeholders that they are taking necessary actions. Prominent issues often also attract more media coverage, and thus put the limelight on corporations or government agencies (Roloff, 2008a, 2008b). This type of symbolic motivation makes issues’ public attention level a highly relevant factor that shapes stakeholders’ decision to build alliances. In short, issue niche width affects the desirability of an issue to business and government sectors. When other things are equal, NGOs working on prominent issues are favored by cross-sector partners than others operating in obscure issues.

Organizational Issue Identity

According to issue niche theory, an NGO derives its issue identity from its core issue-related interests, demands, actions, and sympathetic issue stakeholders (Yang, 2020). Because many social issues contain subissue areas or are closely related to other issues, NGOs that address a wide range of related issues can be classified as generalists, whereas organizations that narrowly focus on one or a few issues are specialists. The mechanism of interorganizational accommodation is the main process that accounts for the formation of issue identity–based niche partitioning (Heaney, 2004). For example, Heaney (2004) studied the identity formation of interest groups that depend on similar funding sources and work on similar political issues. Heaney found that these groups mutually accommodate each other’s identity by avoiding others’ issue niches. The cumulative effect of the mutual accommodation process leads organizations to focus on different issue niches or different aspects of issues. Here, we define our generalists as United Nations (UN) or European Union (EU)-affiliated organizations with broad connections and goals, because they are the most resourceful leaders with multinational reach in the refugee crisis, and specialists as other NGOs with specialized missions in children welfare, refugee resettlement, and so on.

For civil society actors to mobilize resources and gain legitimacy, it is important for both generalist and specialist NGOs to project distinctive identities (Hannan et al., 2007). For generalists, the uniqueness of their identities comes from their reputation as the representatives of an overall issue niche. To maintain its identity as a leader in an issue niche, a generalist is motivated to promote the overall importance of an issue and to connect with a broad range of organizations (Shumate et al., 2005). Throughout the network formation process, when an issue is at the low public attention stage and few actors are building ties, generalists may be the default choice for tie formation due to their reputation and visibility. As the issue continues to evolve, generalists may still reap the benefits of existing ties thanks in part to the mechanism of network inertia² (Sydow et al., 2009). As such, we propose that generalists would enjoy relatively more advantageous network positions regardless of issue stages.

Hypothesis 1 (H1): Within the same issue niche, at both high and low public attention stages, generalists are more likely to build connections across subissue areas than specialists.

In contrast, specialists appeal to stakeholders of more narrowly defined subissue areas. For specialists, they construct their identities through appearing as the most relevant experts on a narrow niche. They tend to be smaller organizations with limited resources. Therefore, diverse ties with many NGOs from other subissue areas may either be irrelevant or dilute the messages about their identities (Heaney, 2004). At the low public attention stage, specialists are unlikely to enjoy much tie formation opportunities due to their obscureness. At the high public attention stages, specialists are likely to mainly collaborate with a few organizations in relevant fields because such ties help them construct unique identities as subissue experts and promote the visibility of their specific subissue areas, which can directly benefit specialists’ resource mobilization. Hence, we propose:

Hypothesis 2 (H2): When the issue is at the high public attention stage, specialists are more likely to build ties within subissue areas.

In addition, as a multistakeholder issue network often includes stakeholder organizations from multiple sectors such as businesses, media, and governments, it is also necessary to consider how organization’s issue identity affects their cross-sector tie formation. The issue niche theory explains how issue evolution changes the tie formation opportunities for civil society actors. Specifically, at the low public attention stage, when stakeholders have yet to form their specific strategies on how to deal with an emerging issue, stakeholders are more likely to work in collaborations with well-known NGOs such as generalists (Shumate et al., 2005). Previous research also suggests that when corporations or governments look for reliable partners to demonstrate their action on a social issue, generalists with long tenure, good reputation, and institutionalized hierarchical structure are often perceived as safe choices (Guo & Acar, 2005). The reputations of generalists may also help to increase the visibility of such collaborations (Selsky & Parker, 2005). However, at the high public attention stage, after an issue has stayed in the public sphere for a period of time, publics are relatively well informed about an issue and have formed their own opinions on how the issue needs to be addressed. Meanwhile, corporations or governmental organizations may also have developed their own strategies to address specific, and often narrowly defined, aspects of an issue. At this point, stakeholders may be more motivated to work with specialists in relevant areas to communicate the fact that they are working with best experts. In other words, the unique value of specialists as most relevant experts may become more attractive to stakeholders at this point. Hence, the following two hypotheses are proposed:

Hypothesis 3 (H3): Throughout an issue’s life cycle, NGOs working on the issue are more likely to develop cross-sector collaborations at the high public attention stage, but opportunities for cross-sector collaborations for NGOs are lower at the low public attention stage.

Hypothesis 4 (H4): Within the same issue niche, when the issue is at the low public attention stage, generalists are more likely to build connections with cross-sector partners (H4_a); when the issue is at the high public attention stage, some specialists would experience increased opportunities to build connections with cross-sector partners (H4_b).

Method

Data Collection

We purchased tweets from DiscoverText, a Twitter analytics tool (Shulman, 2011), between 2013 to 2016. During this time, the influx of refugees in Europe triggered the largest international refugee crisis since World War II. The crisis attracted much media attention in intervals. Using LexisNexis, we gathered how frequently news articles about the refugee crisis appeared in major world publications for each year studied. The frequencies include the date of publication and total number of articles published each day. This allowed us to see specific intervals within each year where the number of published news stories about the refugee was above or below the mean. We operationalized a “news spike” as an interval in which the number of articles published was two standard deviations above the average for that year. The news spikes were then used to identify specific date ranges to gather tweets. We gather tweets posted one day before and one day following a news spike to allow for lead and lag times.

From 2013 to 2015, 1,365,932 tweets from 41 news spikes were collected. Among the 3 years, 2014 had more tweets than 2013 and 2015, showing a recurring tendency of issue evolution. By comparison, in 2016 alone, 2,413,155 tweets from 52 news spikes were collected. Based on the stark difference in coverage, 2013 to 2015 is classified as the low public attention stage, and 2016 is classified as the high public attention stage. We used the retweet relations from the two periods to build networks because in comparison with follower–followee relations or mention relations, retweets are more likely to reflect message-motivated communication (Saffer et al., 2019; Saxton & Waters, 2014; Yang & Saffer, 2020), and therefore capture the dynamic aspect of discourse context. When two organizations communicate with each other through a refugee-related tweet, they often discuss projects, partnerships, shared values, or solidarity. As such, retweet relationships can capture a broad range of partnerships among organizations that range from formal to informal and were communicated openly on social media.

We first identified NGOs, media, and government agencies within the data set. Assisted by keywords³ in Twitter accounts’ bio descriptions, we identified and manually verified 200 NGOs,⁴ 114 media outlets, and 75 government agencies. We searched these accounts in the data set and built two networks at the low and high attention stages. An organization was included in the networks if it had a retweet relation with other organizations.

The retweet network at the low attention stage consisted of 125 NGOs, 61 media, and 51 government agencies from 18 countries, and the retweet network at the high attention stage consisted of 169 NGOs, 88 media, and 43 government agencies from 30 countries. Two researchers read through the retweet relations among these organizations and teased out those connecting the same organizations (e.g., a headquarter organization retweet a subsidiary organization), because including such within-organization ties would result in biased evaluation. Both networks were dichotomized on the median of tie weights, which was one in both cases. Descriptive statistics of the two directed networks are shown in Table 1.⁵ In addition to the two multistakeholder networks, we also contracted two NGOs-only networks to test H1_a–b.

Table 1.

Descriptive Statistics of the Multistakeholder Issue Networks.

	Low public attention stage	High public attention stage
Node types	237	300
NGO/IGO	125	169
Media	61	88
Government agency	51	43
Overall attributes
Countries	18	30
Edges	368	388
Density	0.0066	0.0043
Average degree	3.101	2.587
Reciprocity	0.098	0.052
Transitivity	0.072	0.064

Note. NGO = nongovernmental organization; IGO = international governmental organization.

Measures

Independent variables

Organization type

A categorical variable of organization type is included as a nodal attribute to indicate whether this node is an NGO, a media outlet, or a government agency.

NGO issue identity

We retrieved NGOs’ Twitter bio descriptions and used Latent Dirichlet Allocation (LDA) models to identify issue categories emerging from the bio text. Based on the computational results, two researchers built a codebook of five specialist areas: refugee, children, humanitarian, research, and others. We also identified one generalist: UN- and EU-affiliated organizations. Guided by this codebook (Table 2), we read through the bio description of NGOs and coded each organization into one of the six issue identities. Two researchers checked 20% of the coding and reached an intercoder reliability of 0.72 (Cohen’s kappa).

Table 2.

Codebook for NGOs’ Subissue Areas and Associated Keywords.

Issue areas	No. of NGOs/IGOs	Keywords
Generalist	61	UN, EU
Refugee	43	refugee, asylum, immigrants, seek, legal
Children	27	children, youth, girl, education
Humanitarian	51	human, woman, charity, family, food, relief, build, service, community
Research	2	research, think tank
Others	16	leader, alliance, crowdfunding, democracy

Note. NGO = nongovernmental organization; IGO = international governmental organization.

Country of origin

As the refugee crisis is an issue of international concern, and the actors involved are from different countries, a categorical variable of country origin for all organizations are included. UN and EU affiliates are coded as UN and EU, respectively.

Control variables

As the sample of organizations are from different countries, it was necessary to ensure the effect we observed could be attributed to issue niche theory variables, not organizations’ countries-of-origin. Different countries had different stakes and policies regarding the refugee issue. Thus, we controlled for countries’ social, economic, political, and refugee factors.

Country’s level of democracy

The level of democracy was measured by the political system index based on Marshall et al.’s (2014) Polity IV data. Nations are evaluated on a scale between −10 and 10, and higher values are indicative of “more democratic.” Values for UN affiliates are world averages, and values for EU affiliates are averages of EU member countries. Values for the low attention stage are 3-year averages (M = 4.76, SD = 6.80), and values for the high attention stage are measures for 2016 (M = 4.45, SD = 6.89).

Country’s refugee per capita

Data on country-level refugee per capita were collected from the UN Refugee Agency’s statistics. This variable measured the number of refugees per 1,000 population. The average for low attention stage (M = 21.18, SD =58.80) is slightly lower than the high attention stage (M = 22.42, SD = 57.85).

Country’s unemployment rate

The unemployment rate of each country was collected from the World Bank Development Indicators database, with M = 7.33 (SD = 5.26) for the low attention stage, and M = 7.45 (SD = 5.82) for the high attention stage.

Country’s economic inequality

Gini coefficient was used to measure a country’s economic disparity (Dorfman, 1979). Gini coefficient of each country was collected from the World Bank Development Indicators database, with M = 36.73 (SD = 7.65) for the low attention stage, and M = 36.82 (SD = 7.68) for the high attention stage.

Country’s human rights performance

Freedom House’s Freedom in the World is an annual global report on political rights and civil liberties. Countries are measured on a scale between 0 and 100, with M = 66.55 (SD = 31.09) for the low attention stage, and M = 66.45 (SD = 30.56) for the high attention stage.

Model Specification

To examine the structural patterns in the multistakeholder issue networks, we performed ERGM, which uses Markov Chain Monte Carlo maximum likelihood estimation. ERGM is an inferential technique that examines both endogenous structural properties of networks and the effects of exogenous covariates. The objective is to compare the different effects of exogenous covariates while accounting for the structural dependency of networks at different public attention stages.

Data analysis was performed by the ergm package in R (Hunter et al., 2008). The structural effects included edges, defined by the presence of a tie; mutual, defined as the reciprocity of directed ties; gwidegree and gwodegree (geometrically weighted indegree and outdegree distribution), which account for the skewed degree distribution; and gwesp (geometrically weighted edgewise shared partner distribution), which account for local clustering. For nodal attributes, nodeifactor and nodeofactor add dummy statistics for each unique value of a categorical nodal attribute for indegree and outdegree effects. These ERGM terms test H1 and H4 to examine whether NGOs with different issue identities are more likely to receive or send ties. For homophily effects, nodematch tests within-category tie formation. If positively significant in H2, NGOs were more likely to form ties within subissue areas at the high public attention stage. Nodemix tests the probability of cross-category tie formation. This ERGM term by default adds each possible pairing of attribute values, and by specifying one pairing of attribute value as the baseline, each network statistic can test whether a specific pairing is more likely than the baseline. This term mainly seeks to test the cross-type interaction among NGOs, media, and government agencies specified in H3. Nodeicov and nodeocov for continuous variables at the country level are included as controls.

Results

Models 1 to 4 (Table 3) show the parameters of converged ERGM models of the two NGOs-only networks, and Models 5 to 12 (Tables 4 and 5) examine the formation of multistakeholder networks. All models reported have satisfactory goodness of fit (Figure 1).

Table 3.

ERGM Parameters for the NGOs-Only Networks.

	Low public attention stage(N = 114)		High public attention stage(N = 141)
	Model 1	Model 2	Model 3	Model 4
Structural
Edges	−5.125 (.477)***	−0.031 (2.054)	−5.863 (.551)***	−0.578 (1.796)
Reciprocity	1.159 (.391)**	1.188 (.409)**	1.307 (.409)**	1.296 (.392)***
Gwideg^a	−1.289 (.316)***	−1.192 (.328)***	−2.165 (.272)***	−2.220 (.273)***
Gwodeg^a	0.655 (.331)*	0.612 (.326)	0.640 (.283)*	0.468 (.279)
Gwesp^b	1.007 (.144)***	1.012 (.153)***	1.035 (.158)***	1.045 (.155)***
Nodal attributes
Issue homophily (H2)	0.308 (.232)	0.278 (.235)	0.450 (.172)**	0.446 (.173)*
Generalist indeg (H1)	0.720 (.339)*	0.589 (.411)	1.010 (.437)*	0.803 (.434)
Children indeg	−0.088 (.405)	−0.252 (.429)	0.355 (.450)	0.281 (.447)
Humanitarian indeg	−0.352 (.377)	−0.401 (.396)	0.593 (.434)	0.548 (.427)
Refugee indeg	−0.189 (.361)	−0.101 (.375)	0.567 (.437)	0.574 (.428)
Research indeg	0.210 (.605)	0.429 (.667)	0.558 (.760)	0.653 (.763)
Generalist outdeg	−0.733 (.441)	−0.732 (.501)	−0.354 (.383)	−0.671 (.396)
Children outdeg	0.221 (.458)	0.083 (.469)	0.784 (.369)*	0.673 (.358)
Humanitarian outdeg	0.052 (.419)	−0.138 (.431)	0.007 (.374)	0.015 (.355)
Refugee outdeg	0.645 (.391)	0.642 (.401)	0.622 (.363)	0.564 (.346)
Country homophily	1.285 (.189)***	1.629 (.211)***	1.129 (.149)***	1.397 (.171)***
Control
Democracy indeg		−0.021 (.060)		0.044 (.036)
Democracy outdeg		0.114 (.056)*		0.028 (.060)
Refugee per capita indeg		0.020 (.014)		−0.004 (.005)
Refugee per capita outdeg		−0.017 (.014)		−0.011 (.009)
Unemployment indeg		−0.129 (.077)		0.001 (.034)
Unemployment outdeg		0.052 (.052)		0.050 (.032)
Gini indegree		−0.022 (.024)		−0.037 (.018)*
Gini outdegree		−0.063 (.024)**		−0.044 (.020)*
Human rights indeg		0.001 (.016)		−0.022 (.010)*
Human rights outdeg		−0.028 (.015)		−0.011 (.015)
AIC	1,827	1,824	2,169	2,210
BIC	1,947	2,018	2,298	2,419

Note. Outdegree effect for research issue is not listed because there is no tie. Standard errors are included in parentheses. ERGM = exponential random graph models; AIC = Akaike information criterion; BIC = Bayesian information criterion.

Geometrically weighted indegree/outdegree distribution. Decay parameters are selected based on the best goodness of fit. ^bGeometrically weighted edgewise shared partner distribution.

p < .05.**p < .01. ***p < .001.

Table 4.

ERGM Parameters for the Multistakeholder Issue Networks.

	Low public attention stage(N = 237)		High public attention stage(N = 300)
	Model 5	Model 6	Model 7	Model 8
Structural
Edges	−5.420 (.197)***	−5.459 (.123)***	−6.648 (.287)***	−3.549 (.826)***
Reciprocity	1.251 (.335)***	1.235 (.246)***	0.837 (.282)**	0.657 (.355)
Gwideg	−1.845 (.199)***	−1.572 (.217)***	−1.583 (.176)***	−1.483 (.187)***
Gwodeg	0.734 (.255)**	0.609 (.266)*	0.192 (.206)	0.086 (.209)
Gwesp	1.277 (.118)***	1.313 (.100)***	1.347 (.095)***	1.437 (.119)***
Nodal attributes
NGO-media^a (H3)	0.453 (.209)*	0.424 (.217)	1.526 (.279)***	1.593 (.299)***
Gov-NGO^a (H3)	0.097 (.248)	−0.112 (.248)	0.751 (.329)*	0.749 (.348)*
Media-NGO^a	−1.699 (405)***	−1.818 (.406)***	−1.355 (.505)**	−1.307 (.523)*
NGO-NGO^a	0.686 (.179)***	0.543 (.192)**	1.670 (.270)***	1.597 (.289)***
Homophily country	0.937 (.097)***	1.061 (.101)***	1.124 (.086)***	1.419 (.106)***
Control
Democracy indeg		−0.077 (.024)**		−0.050 (.020)*
Democracy outdeg		0.077 (.024)**		0.106 (.026)***
Refugee per capita indeg		0.002 (.008)		−0.001 (.003)
Refugee per capita outdeg		0.008 (.008)		−0.005 (.006)
Unemployment indeg		−0.032 (.039)		−0.006 (.031)
Unemployment outdeg		0.032 (.032)		0.039 (.022)
Gini indeg		−0.017 (.013)		−0.044 (.013)***
Gini outdeg		0.006 (.014)		−0.020 (.012)
Human rights indeg		0.009 (.007)		−0.002 (.004)
Human rights outdeg		−0.005 (.007)		−0.015 (.005)**
AIC	3,838	3,829	4,385	4,363
BIC	3,963	4,044	4,507	4,580

Note. ERGM = exponential random graph models; AIC = Akaike information criterion; BIC = Bayesian information criterion.

All compared with ties from NGO to government. Only pairings involving NGOs are displayed.

p < .05. **p < .01. ***p < .001.

Table 5.

ERGM Parameters for the Multistakeholder Issue Networks.

	Low public attention stage(N = 237)		High public attention stage(N = 300)
	Model 9	Model 10	Model 11	Model 12
Structural
Edges	−5.971 (.166)***	−4.113 (.109)***	−6.461 (.189)***	−2.698 (.024)***
Reciprocity	1.385 (.278)***	1.427 (.301)***	0.880 (.443)*	0.799 (.310)**
Gwideg	−1.455 (.211)***	−1.438 (.219)***	−1.286 (.189)***	−1.373 (.183)***
Gwodeg	0.539 (.233)*	0.552 (.233)*	0.149 (.214)	0.105 (.206)
Gwesp	1.214 (.100)***	1.196 (.108)***	1.343 (.133)***	1.399 (.097)***
Nodal attributes
Generalist indeg (H4_a)	0.403 (.126)**	0.249 (.175)	0.349 (.128)**	0.132 (.132)
Children indeg	−0.353 (.262)	−0.415 (.265)	−0.630 (.232)**	−0.657 (.180)***
Humanitarian indeg	−0.614 (.205)**	−0.614 (.218)**	−0.349 (.159)*	−0.359 (.132)**
Refugee indeg	−0.343 (.138)*	−0.250 (.141)	−0.369 (.147)*	−0.318 (.119)**
Research indeg	−0.403 (.525)	−0.264 (.532)	−0.942 (.835)	−0.825 (.017)***
Generalist outdeg	−0.001 (.173)	−0.035 (.238)	0.287 (.175)	0.131 (.179)
Children outdeg (H4_b)	0.459 (.273)	0.429 (.277)	1.201 (.193)***	1.163 (.174)***
Humanitarian outdeg (H4_b)	0.378 (.197)	0.381 (.202)	0.536 (.184)**	0.528 (.166)**
Refugee outdeg (H4_b)	1.026 (.163)***	1.124 (.170)***	1.192 (.171)***	1.219 (.154)***
Research outdeg	−1.007 (1.071)	−0.856 (1.057)	−0.599 (1.064)	−0.401 (.014)***
Homophily type	0.651 (.125)***	0.645 (.126)***	0.646 (.125)***	0.646 (.107)***
Homophily country	1.080 (.103)***	1.206 (.114)***	1.291 (.104)***	1.517 (.098)***
Control
Democracy indeg		−0.020 (.029)		−0.022 (.022)
Democracy outdeg		0.011 (.036)		0.057 (.030)
Refugee per capita indeg		0.008 (.009)		0.000 (.003)
Refugee per capita outdeg		−0.004 (.008)		−0.010 (.006)
Unemployment indeg		−0.071 (.045)		0.002 (.031)
Unemployment outdeg		0.053 (.034)		0.046 (.023)*
Gini indeg		−0.028 (.014)		−0.032 (.011)**
Gini outdeg		−0.014 (.014)		−0.039 (.011)***
Human rights indeg		0.002 (.007)		−0.004 (.004)
Human rights outdeg		−0.004 (.007)		−0.015 (.005)**
AIC	3,828	3,832	4,394	4,385
BIC	3,980	4,074	4,553	4,639

Note. ERGM = exponential random graph models; AIC = Akaike information criterion; BIC = Bayesian information criterion.

p < .05. **p < .01. ***p < .001.

Figure 1.

Goodness of fit diagnostic plots of Models 2, 4, 6, 8, 10, and 12.

H1 and H2 examine how the tie formation of the NGOs-only network was driven by organization issue identities. In Models 1 to 4, the indegree effect of generalist NGOs was positively significant, Estimate_generalist = 0.720, SE = .339, p < .05 at the low public attention stage, and Estimate_generalist = 1.010, SE = .437, p < .05 at the high public attention stage, although these effects were not significant once county-level covariates were controlled. This result shows that across the issue life cycle, generalists were preferred partners for building cross-subissue relations. H1 is supported without control variables but not supported with control variables.

Moreover, the subissue homophily effect was only positively significant at high public attention stage, Estimate_{issue match} = .446, SE = .173, p < .05. This result shows that specialists were likely to support others with the same subissue identities when public attention was high. H2 is supported with and without control variables.

H3 seeks to examine how cross-sector tie formation of the multistakeholder issue networks differed when public attention increased. As shown in Models 5 to 8, NGOs interacted with media and government in five directed relations, among which the tie from media to NGOs was the least possible, and the homophily among NGOs was the most possible. Although NGOs tended to send more ties to media than to government at the low public attention stage (see Model 5), this effect was not significant when country variables were controlled. However, at high public attention stage, as shown in Models 7 to 8, this effect was much more evident, Estimate_ngo-media = 1.526, SE = .279, p < .001, and when control variables were added, Estimate_ngo-media = 1.593, SE = .299, p < .001. This result shows that NGOs made a proactive effort to build media relations to influence public agenda. Furthermore, at the high attention stage, NGOs also received more ties from government agencies than sending ties to government agencies, which again provides evidence for the pivotal role played by NGOs in the refugee crisis. To sum up, NGOs experienced more opportunities to build cross-sector relations at the high attention stage. H3 is supported.

H4 seeks to test how NGOs’ different issue identities influenced the probability of receiving or sending a retweet in the multistakeholder issue networks. As shown in Models 9 to 12, controlling for within-type ties, generalists were more likely to receive ties throughout the issue life cycle, Estimate_generalist = 0.403, SE = .126, p < .01 for the low public attention stage, and Estimate_generalist = 0.349, SE = .128, p < .01 for the high public attention stage. But after controlling for country-level variables, these effects were not significant. In comparison, throughout the issue life cycle, specialists were less likely to receive ties from the cross-sector community. H4_a is supported without control variables but not supported with control variables.

NGOs specialized in the refugee issue were more likely to send out ties at both issue stages, because their expertise is highly relevant to address the refugee crisis. In comparison, other specialists operating in children and humanitarian issues were not active at the low public attention stage, but these NGOs sent significantly more ties at the high public attention stage. The outdegree effects of children and humanitarian issues were positively significant in Model 12, Estimate_children = 1.163, SE = .174, p < .001 for children issue, and Estimate_humanitarian = .528, SE = .166, p < .01 for humanitarian issue. This result confirms that specialist NGOs had more opportunities for cross-sector relation building at the high attention stage. H4_b is supported with and without control variables.

Discussion

The results highlight the utility of issue niche theory for explaining relationship formation patterns in a multistakeholder issue network that involves international issue stakeholders. We believe a number of key insights warrant further discussion.

Issue Evolution and Tie Formation in Multistakeholder Issue Networks

Previous studies have recognized the importance of social media to NGOs’ operation (Campbell & Lambright, 2019; Guo & Saxton, 2018; Saxton & Waters, 2014) and studies have examined the network structure of multistakeholder issue networks (Doerfel & Taylor, 2017; Margolin et al., 2015; Saffer et al., 2018). Our study expands issue niche theory and multistakeholder issue network research by conceptualizing multistakeholder issue networks as a form of issue niche community. This approach explicitly recognizes that organizations’ stakeholder relationship communication on social media is a form of strategic action. Such strategic actions are influenced by various factors such as organizations’ stakes in an issue, identities, and their positions in an issue niche in relation to each other.

Our study shows that analyzing one or a few NGOs’ social media use as isolated incidents overlooks the interorganizational structure within which such interactions take place. A networked communication context such as social media calls for theoretical frameworks that account for how networks affect strategic communication decisions. For instance, we observed that there are structural tendencies for some NGOs to be considered as preferred partners (significant indegree) and there was also a significant tendency for partner clusters to emerge. Our theoretical framework further provides explanations for why some organizations are preferred, what factors drive the formation of clusters, and how such tie formation patterns may change as the issue evolved over its life cycle.

The issue niche theory reconceptualizes resource niche through the lens of public attention and communication. This theory is especially pertinent at explaining NGOs’ relationship formation on social media. Contrary to the assumption that organizations would use social media to build relationships with stakeholders and partners at will, our analysis showed that when public attention is low, even specialist NGOs working on similar subissue niche rarely interact with each other. Yet, when the issue evolved to a high attention stage, we observed a significant homophily effect. Specialist NGOs seemed eager to build relations with other specialists who share the same issue identity. For generalists such as the UN-related organizations, they are preferred target of tie formation in both public attention stages but they are unwilling to initiate ties at high public attention stage, suggesting a selective tie building strategy.

The explanation derived from the issue niche theory is that when an issue is at a low public attention stage, relationship building opportunities are abundant for generalists. In the case of the global refugee crisis, unlike other more well-known issues such as climate change, during the low public attention stage, many publics around the world have little awareness or knowledge of this emerging issue. At this stage, generalists are most likely to actively build connections to promote the overall salience of the issue; in contrast, specialists would lack opportunities to communicate both within and across sectors, likely due to their lack of visibility. As the issue evolves to attract widespread public attention, more resources may flow into this issue area (Yang, 2020). One unfortunate example is that after the image of a 3-year-old Syrian boy who drowned in the Mediterranean Sea made global headlines, many refugee NGOs reported receiving record donations. Specialists at this stage would be motivated to take actions, build alliances, and attract more attention to their specialty areas. At this stage, specialists’ needs to construct their identity as narrowly focused experts would drive them to prioritize building connections with other specialists in their own subissue. Such connections serve the symbolic value of strengthening the specialists’ image as well-embedded in a specialty area and also help to draw attention to that specialty area, which further directly benefits the causes of specialists.

For cross-sector relationship formation, again we found that different issue stages would lead to different tie formation patterns. When the public attention to the refugee issue is low, few NGOs actively retweet media. However, as the issue evolves to a high public attention stage, NGOs are significantly more likely to initiate ties to the media. One possible explanation is that NGOs increased their media catching efforts due to the anticipation that the media would be more interested in covering refugee-related news at the height of this crisis. This is because unlike issues such as climate change where most people experience the impact directly. In the case of the refugee crisis, oftentimes NGOs are the main channel through which the rest of the world can learn about conditions in refugee camps. NGOs also became preferred targets for government organizations at the high public attention stage, probably because government organizations need to mobilize the resources and expertise from NGOs to address this unprecedented crisis. In the case of refugee settlement, government agencies often depend on NGOs for distributing essential supplies such as food, clothing, blankets, and tents. They also depend on NGOs for providing critical health care services such as setting up emergency hospitals and visiting patients in camps (UNHCR, 2021).

Moreover, while previous studies often suggest that generalists are preferred partners of cross-sector alliances (Selsky & Parker, 2005; Shumate et al., 2005), the current study reveals a much more nuanced situation. Specifically, we found that when the issue is at the low public attention stage, the generalists received significantly more ties than other organizations. Organizations in several specialist areas, such as organizations in children refugee issue and humanitarian aid issue, become highly active at the high public attention stage. According to issue niche theory, at low public attention stage, cross-sector stakeholders are most likely to work with generalists because these organizations offer higher levels of legitimacy (Guo & Acar, 2005). However, as public attention rises and the complexity of the issue becomes more well known, cross-sector stakeholders are motivated to work with specialists who are the most appropriate experts on issue solutions (Doerfel & Taylor, 2017). Our study thus reveals that a static view on how cross-sector stakeholders build relationships in an evolving issue would miss out important dynamics.

National Context and Multistakeholder Issue Network

In addition to findings related to the issue niche theory, the global nature of this refugee crisis gave us a rare opportunity to examine how organizations’ countries-of-origin influenced tie formation. One striking observation is that across models, the homophily effect for countries-of-origin constantly stayed positive and significant, even when controlled for political, economic, and refugee-related factors. In the process of ERGM model building, several models failed to converge if country homophily was not controlled. This finding suggests that the global response to the refugee issue is largely a domestic effort, and organizations primarily seek out other organizations from the same country to build relations. This observed phenomenon may suggest a tendency of nationalization/localization of a transnational issue. It seems that although the Twittersphere potentially offers a platform for global collaboration and interactions, a range of factors still drive organizations to prefer domestic partnerships over international ones. Future research may further look into the conditions under which international interactions on social media increase or decrease, which may provide valuable insights for resource mobilization and advocacy for international issues.

When organizations seek out international collaborations, in both of the NGOs-only and multistakeholder networks, organizations from highly developed democracies with good human rights protection records tend to be active in terms of initiating ties. However, it is interesting to observe that organizations from countries with high income inequality were unpopular partners. One possible explanation is that countries with high Gini indexes may present an unfavorable context for forming multistakeholder responses to the refugee issue because social inequalities may lower public support for helping vulnerable immigrants.

Limitation and Recommendations for Future Research

Theoretically, multistakeholder issue networks are usually tripartite that involve business, civil society, and government (Roloff, 2008a). In the current research context, possibly due to the sensitivity of the refugee issue, business sectors refrained from joining the multistakeholder issue networks. This fact illustrates how different social issues are likely to involve different actors. To build a comprehensive theory, future research could explore the tripartite relations among business, civil society and government in issues such as the COVID-19 global pandemic to understand how societal institutions with different resource and identity endowments play different roles in multistakeholder issue networks.

Second, the current study focused on the tie formation of the multistakeholder issue networks of the international refugee issue but did not analyze the content of communication. That is, when different societal institutions interact, what do they communicate about and how does the content of communication relate to the change of issue frames? The current study incorporates media outlets as important stakeholders, and future studies could expand the sample and examine how influential media professionals affect issue framing. To summarize, future research could incorporate the semantic dimension (Russell & Reimer, 2020). When societies are challenged by a wicked problem, a diverse range of perspectives, topics, and interpretations are likely to emerge. Future research could explore how meanings and networks co-evolve in multistakeholder issue networks.

Conclusion

For NGOs to effectively utilize social media as a new frontier for stakeholder engagement (Xu & Saxton, 2019), it is critical that we advance new theories that can accommodate the networked communication context. This study integrates the multistakeholder issue network approach (Roloff, 2008a, 2008b; Saffer, 2019) and the issue niche theory (Yang, 2020) to provide a theoretical framework to understand how issue evolution and NGOs’ issue identity affect tie formation in a multistakeholder issue network. In different issue contexts, other factors such as organizations’ ideologies, organizations’ issue response orientations may also shape alliance network structures. As worrisome trends such as climate change continue to worsen, more studies should devote attention to this line of research and continue to offer critical insights on how NGO communities can effectively engage with critical stakeholders to face global challenges ahead.

Footnotes

Appendix

Keywords for Initially Identifying Nonprofits, Media, and Government Agencies.

Organizational type	Keywords
Nonprofits	NGO, non-governmental, nonprofit, not-for-profit, charity, philanthropic, 501(c), refugee, UN, children, poverty, women, UNICEF, UNHCR, mission, donation, advocacy, coalition, foundation, relief, food bank, humanitarian, human right, fundraising, asylum.
Media	media, news, news outlet, newsroom, breaking news, reporting, news network, newspaper, radio, television, journalism, commentary, magazine, journalist, correspondent, coverage.
Government	government, governor, minister, ministry, embassy, diplomat, consulate, parliament, department, state, senate, congress.

Note. UN =United Nations; UNICEF = United Nations Children's Fund; UNHCR = United Nations High Commissioner for Refugees.

Authors’ Note

The authors would like to thank the Editor and two anonymous reviewers who provided valuable suggestions and comments on the earlier version of this manuscript.

Declaration of Conflicting Interests

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

Funding

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

ORCID iDs

Jingyi Sun

Aimei Yang

Notes

Author Biographies

Jingyi Sun (PhD, University of Southern California) is an assistant professor at Stevens Institute of Technology. Her research focuses on the ecological and network dynamics of individuals, groups, and organizations.

Aimei Yang (PhD, University of Oklahoma) is an associate professor at the University of Southern California. Her research focuses on interorganizational networks, stakeholder networks, civil actors’ issue advocacy, and issue networks.

Adam J. Saffer (PhD, University of Oklahoma) is an assistant professor at the University of Minnesota’s Hubbard School of Journalism and Mass Communication. He takes a network perspective to explore the ways individuals, groups, and organizations establish, use, and are influenced by their communication networks.

References

Aldrich

H. E.

Ruef

(2006). Organizations evolving (2nd ed.). SAGE.

Alford

Head

B. W.

(2017). Wicked and less wicked problems: A typology and a contingency framework. Policy and Society, 36(3), 397–413. https://doi.org/10.1080/14494035.2017.1361634

Austin

J. E.

Seitanidi

M. M.

(2012). Collaborative value creation: A review of partnering between nonprofits and businesses: Part I. Value creation spectrum and collaboration stages. Nonprofit and Voluntary Sector Quarterly, 41(5), 726–758.

Bäckstrand

(2006). Multi-stakeholder partnerships for sustainable development: Rethinking legitimacy, accountability and effectiveness. European Environment, 16(5), 290–306.

Bhati

McDonnell

(2020). Success in an online giving day: The role of social media in fundraising. Nonprofit and Voluntary Sector Quarterly, 49(1), 74–92.

Campbell

D. A.

Lambright

K. T.

(2019). Are you out there? Internet presence of nonprofit human service organizations. Nonprofit and Voluntary Sector Quarterly, 48(6), 1296–1311.

Crable

R. E.

Vibbert

S. L.

(1985). Managing issues and influencing public policy. Public Relations Review, 11(2), 3–16.

Dentoni

Bitzer

Schouten

(2018). Harnessing wicked problems in multi-stakeholder partnerships. Journal of Business Ethics, 150(2), 333–356.

Doerfel

M. L.

Taylor

(2017). The story of collective action: The emergence of ideological leaders, collective action network leaders, and cross-sector network partners in civil society. Journal of Communication, 67(6), 920–943. https://doi.org/10.1111/jcom.12340

10.

Dorfman

(1979). A formula for the Gini coefficient. The Review of Economics and Statistics, 61(1), 146–149.

11.

Guo

Acar

(2005). Understanding collaboration among nonprofit organizations: Combining resource dependency, institutional, and network perspectives. Nonprofit and Voluntary Sector Quarterly, 34(3), 340–361.

12.

Guo

Saxton

G. D.

(2014). Tweeting social change: How social media are changing nonprofit advocacy. Nonprofit and Voluntary Sector Quarterly, 43(1), 57–79.

13.

Guo

Saxton

G. D.

(2018). Speaking and being heard: How nonprofit advocacy organizations gain attention on social media. Nonprofit and Voluntary Sector Quarterly, 47, 5–26.

14.

Haas

P. M.

(1992). Introduction: Epistemic communities and international policy coordination. International Organization, 46(1), 1–35.

15.

Hannan

M. T.

Pólos

Carroll

G. R.

(2007). Logics of organization theory: Audiences, codes, and ecologies. Princeton University Press.

16.

Heaney

M. T.

(2004). Outside the issue niche: The multidimensionality of interest group identity. American Politics Research, 32(6), 611–651.

17.

Heath

R. L.

Palenchar

M. J.

(2008). Strategic issues management: Organizations and public policy challenges. SAGE.

18.

Heft

Mayerhöffer

Reinhardt

Knüpfer

(2020). Beyond Breitbart: Comparing right-wing digital news infrastructures in six western democracies. Policy & Internet, 12(1), 20–45.

19.

Hunter

D. R.

Handcock

M. S.

Butts

C. T.

Goodreau

S. M.

Morris

(2008). ERGM: A package to fit, simulate and diagnose exponential-family models for networks. Journal of Statistical Software, 24(3), nihpa54860.

20.

Klijn

E. H.

Koppenjan

J. F.

(2000). Public management and policy networks: Foundations of a network approach to governance. Public Management an International Journal of Research and Theory, 2(2), 135–158.

21.

Klijn

E. H.

Steijn

Edelenbos

(2010). The impact of network management on outcomes in governance networks. Public Administration, 88(4), 1063–1082.

22.

Knüpfer

C. B.

Entman

R. M.

(2018). Framing conflicts in digital and transnational media environments. Media, War & Conflict, 11(4), 476–488.

23.

Le Ber

M. J.

Branzei

(2010). Towards a critical theory of value creation in cross–sector partnerships. Organization, 17(5), 599–629.

24.

Margolin

D. B.

Shen

Lee

Weber

M. S.

Fulk

Monge

(2015). Normative influences on network structure in the evolution of the children’s rights NGO network, 1977–2004. Communication Research, 42(1), 30–59.

25.

Marshall

M. G.

Gurr

T. R.

Jaggers

(2014). Polity IV project: Political regime characteristics and transitions, 1800–2013. Center for Systemic Peace.

26.

McGuire

Agranoff

(2011). The limitations of public management networks. Public Administration, 89(2), 265–284.

27.

Monge

Heiss

B. M.

Margolin

D. B.

(2008). Communication network evolution in organizational communities. Communication Theory, 18(4), 449–477.

28.

Moog

Spicer

Böhm

(2015). The politics of multi–stakeholder initiatives: The crisis of the Forest Stewardship Council. Journal of Business Ethics, 128(3), 469–493.

29.

Nigam

Ocasio

(2010). Event attention, environmental sensemaking, and change in institutional logics: An inductive analysis of the effects of public attention to Clinton's health care reform initiative. Organization Science, 21(4), 823–841.

30.

O’Brien

N. F.

Pilny

Atouba

Y. C.

Shumate

Fulk

J. L.

Monge

P. R.

(2019). How does NGO partnering change over time? A longitudinal examination of factors that influence NGO partner selection. Nonprofit and Voluntary Sector Quarterly, 48(6), 1229–1249.

31.

O’Rourke

(2006). Multistakeholder regulation: Privatizing or socializing global labor standards? World Development, 34(5), 899–918.

32.

Roloff

(2008a). Learning from multi-stakeholder networks: Issue-focused stakeholder management. Journal of Business Ethics, 82(1), 233–250.

33.

Roloff

(2008b). A life cycle model of multi-stakeholder networks. Business Ethics: A European Review, 17(3), 311–325.

34.

Russell

Reimer

(2020). Attribute degree centrality and attribute tie strength as criteria of argument quality. Communication Monographs, 87, 336–358. https://doi.org/10.1080/03637751.2020.1724312

35.

Sabatier

P. A.

Weible

C. M.

(2007). The advocacy coalition framework. Theories of the Policy Process, 2, 189–220.

36.

Saffer

A. J.

(2019). Fostering social capital in an international multi-stakeholder issue network. Public Relations Review, 45(2), 282–296. https://doi.org/10.1016/j.pubrev.2019.02.004

37.

Saffer

A. J.

Yang

Morehouse

(2019). It takes a village: A social network approach to NGOs’ international public engagement. American Behavioral Scientist, 63(12), 1708–1727. https://doi.org/10.1177/0002764219835265

38.

Saffer

A. J.

Yang

Taylor

(2018). Reconsidering power in multistakeholder relationship management. Management Communication Quarterly, 32(1), 121–139. https://doi.org/10.1177/0893318917700510

39.

Saxton

G. D.

Waters

R. D.

(2014). What do stakeholders like on Facebook? Examining public reactions to nonprofit organizations’ informational, promotional, and community-building messages. Journal of Public Relations Research, 26(3), 280–299.

40.

Selsky

J. W.

Parker

(2005). Cross-sector partnerships to address social issues: Challenges to theory and practice. Journal of Management, 31(6), 849–873.

41.

Shulman

(2011, June). DiscoverText: Software training to unlock the power of text. Proceedings of the 12th Annual International Digital Government Research Conference: Digital Government Innovation in Challenging Times, ACM, New York, NY, United States.

42.

Shumate

Fulk

Monge

(2005). Predictors of the international HIV–AIDS INGO network over time. Human Communication Research, 31(4), 482–510.

43.

Smith

(2005). Globalizations and transitional social movement organizations. In Daus

McAdam

Scott

W. R.

Zald

(Eds.), Social movements and organizational theory (pp. 226–248). Cambridge University Press.

44.

Smith

(2008). Social movements for global democracy. John Hopkins University Press.

45.

Svensson

P. G.

Mahoney

T. Q.

Hambrick

M. E.

(2015). Twitter as a communication tool for nonprofits: A study of sport-for-development organizations. Nonprofit and Voluntary Sector Quarterly, 44(6), 1086–1106.

46.

Sydow

Schreyögg

Koch

(2009). Organizational path dependence: Opening the black box. Academy of Management Review, 34(4), 689–709.

47.

United Nations High Commissioner for Refugees. (2021). NGOs: Our right arm. https://www.unhcr.org/en-us/publications/refugeemag/3b53fd8b4/refugees-magazine-issue-97-ngos-unhcr-ngos-right-arm.html

48.

Webster

J. G.

(2011). The duality of media: A structurational theory of public attention. Communication Theory, 21(1), 43–66.

49.

Saxton

G. D.

(2019). Does stakeholder engagement pay off on social media? A social capital perspective. Nonprofit and Voluntary Sector Quarterly, 48(1), 28–49.

50.

Yang

(2020). The issue niche theory of nongovernmental and nonprofit organizations’ interorganizational network ecology. Communication Theory, 30(1), 41–63.

51.

Yang

Saffer

A. J.

(2020). Standing out in a networked communication context: Toward a network contingency model of public attention. New Media & Society, 23, 2902–2925. https://doi.org/10.1177/1461444820939445