Introduction

Relations between civil society and illiberal states are often fraught. Though civil society is frequently analyzed as a democratizing force (Putnam, Reference Putnam1993), scholars of non-democratic contexts emphasize how states may circumscribe oppositional civil society organizations (CSOs) while empowering ideologically aligned actors in order to consolidate power (Cavatorta, Reference Cavatorta2009; Hsu, Reference Hsu2010; Jamal, Reference Jamal2007; Lewis, Reference Lewis2013; Wiktorowicz, Reference Wiktorowicz2000). A state may also strategically delegate tasks to nonstate actors—even potentially oppositional groups—to preserve its fiscal and political strength (Brooke, Reference Brooke2019; Cammett & MacLean, Reference Cammett and MacLean2014). Given that civil society in restrictive contexts is often analyzed through a regime alignment-versus-opposition binary (Toepler et al., Reference Toepler, Zimmer, Fröhlich and Obuch2020), it is easy to overlook the fact that many CSOs are more concerned with their day-to-day activities than with staking political positions. Moreover, organizations that work with state institutions do so in a variety of configurations, often with some discretion as to whether and how they engage the state (Stroup, Reference Stroup and Davies2019). The preservation of state capacity may explain why states delegate tasks to CSOs, but it does not explain how or why organizations engage state institutions. This paper asks: How do nonstate organizations carry out their programs in political contexts hostile to civil society activity?

Complementing much of the scholarship on state-civil society relations that prioritizes state action as the object of analysis, this paper focuses on organizational framing, strategizing, and justification of engagement with state institutions in a restrictive context. It examines the field of refugee support in Turkey, which does not break cleanly along political or ideological lines. After the Syrian conflict began in 2011, a range of CSOs in Turkey found themselves in alignment with the ruling Justice and Development Party (AKP) policy of supporting refugees amid broader public opposition to generous refugee policy. The Turkish state’s Temporary Protection Regulation provided Syrian refugees with social rights, including access to free healthcare and education. NGOs, charities, and associations played an important role in facilitating access to these rights as well as providing their own aid and supplementary programming. A variety of CSOs worked with state institutions in project creation or implementation, while many others avoided registering with or gaining project approvals from the state.

In order to explain this variation, the paper draws on interviews conducted in 2018 and 2019 with key personnel across 23 different nonstate organizations providing a range of refugee support services in Istanbul. I find that organizational interactions with the state are dictated by two main criteria: (1) an organization’s capacity, defined by its size, financial resources, and human resources which allow it to reach a large number of beneficiaries (Cammett & Luong, Reference Cammett and Luong2014:194–195), and (2) an organization’s identity, or the labels and codes it adopts and is assigned that make it legible to relevant audiences (Hsu & Hannan, Reference Hsu and Hannan2005). Such signifiers include language used in mission statements and organization materials, the content and goals of organization projects, and organizational type. In the case of refugee reception in Turkey, organizational identities tend to cluster around recognizable categories like rights-based organizations or needs-based humanitarian or charity organizations. Other organizations that adopt ambiguous, non-conforming labels are less easily categorized.

I argue that capacity trumps identity in facilitating state-organization relations where extensive external funding flows in, whereas an organization’s identity is the primary factor in state-organization relations for lower-capacity organizations. High-capacity organizations across a range of identities cooperate with state institutions in order to reach a broad beneficiary base. The degree to which these organizations must compromise their own projects is nevertheless rooted in whether they are perceived as potential antagonists with rights-based agendas or as reliable partners supplementing state policy with humanitarian support. Lower-capacity organizations are more likely to comply with, avoid, or selectively engage state institutions depending on the codes they adopt. Rights-based organizations are strategic in how they engage the state, whereas other organizations distance themselves from state institutions by positioning as neutral humanitarians. More broadly, the paper contends that incorporating analysis of organizational identity into studies of civil society clarifies how assumptions of CSO alignment or opposition with the state are actually constructed, negotiated, and deployed in complex political fields.

Turkey’s Political Context

While Turkey has a long history of civil society activity, for decades associational life was largely oriented toward promoting state ideals (Bahçecik & Turhan, Reference Bahçecik and Turhan2022). Civil society activity across the political spectrum increased with Turkey’s gradual turn toward Europe in the 1980s and 1990s (Tocci, Reference Tocci2005). A bid for European Union accession in the early 2000s led to growth in the number of civil society organizations, many with a democratizing thrust (Altan-Olcay & Icduygu, Reference Altan-Olcay and Icduygu2012; Keyman & Icduygu, Reference Keyman and Icduygu2003). During this period, the Turkish government liberalized laws on associations and foundations, loosening the degree of government control over organizations, including funding sources and external partnerships (Bikmen, Reference Bikmen2005). Even so, the Department of Associations and the General Directorate of Foundations remained inconsistent in the ways they monitored civil society organizations, which continued to be subjected to onerous bureaucratic rules (Varon et al., Reference Varon, Yermioğlu Yılmaz, Dereci, Ersen and Balcıoğlu2014). Liberalization of CSO regulation reversed course after the AKP won a parliamentary majority in 2011. Following the Gezi Park protests in 2013, the government cracked down on anti-government activists and associated organizations (Sarfati, Reference Sarfati2017). In 2015, after the predominantly Kurdish People’s Democratic Party (HDP) gained a foothold in parliament, the AKP initiated a military incursion in Kurdish regions and cracked down on Kurdish civil society actors (Aras & Duman, Reference Aras and Duman2019). Soon after, a failed coup against the regime in July 2016 triggered heavy suppression across a broad swath of civil society, including NGOs, schools, media outlets, and vast cadres of civil servants. The coup and ensuing suppression increased skepticism of Western NGOs operating in Turkey (Mellen & Lynch, Reference Mellen and Lynch2017).

Despite this intensifying crackdown and increasing authoritarianism, the number of civil society organizations in Turkey continued to grow (Yabanci, Reference Yabanci2019). Part of this growth can be attributed to the organizations that emerged to support the rising Syrian refugee population within Turkey’s borders (Sunata & Tosun, Reference Sunata and Tosun2019). A variety of organizations constitute the “welfare mix” providing support to Syrian refugees, including international organizations, national NGOs, local grassroots organizations, and religious charities as well as state institutions at the national, provincial, and local levels (Yilmaz, Reference Yilmaz2019). The negotiation of a deal between the European Union and Turkey in 2016 provided six billion euros of funding for humanitarian assistance to Syrians in exchange for securitizing its border to prevent migrants from traveling to Europe. The EU Facility for Refugees has disbursed these funds to projects managed by UN agencies via INGOs, NGOs, and state institutions (Mandıracı, Reference Mandıracı2020). As a result, a bevy of CSOs have continued to operate in the field of refugee support, even as the state has consolidated its refugee reception regime and cracked down on civil society activity.

NGOs and the State

In non-democratic contexts, states may use a tangle of bureaucratic regulations to keep civil society organizations under control while repressing the potential for organizations to emerge autonomously from the state (Wiktorwicz, Reference Wiktorowicz2000). Scholars of the Middle East have argued that civil society can exist as an element of “liberalized autocracy” characterized by state control of opposition with limited space for nonstate organizational activity (Anderson, Reference Anderson2006; Brumberg, Reference Brumberg2002; Cavatorta & Durac, Reference Cavatorta and Durac2011). While a large segment of civil society scholarship focuses on the effect of state-civil society relations on democratization outcomes, there is also a growing literature that examines the nature and forms of state-CSO relations. Scholars of East Asia, for example, identify patterns of regulation, cooptation, negotiation, alliance, marginalization, and avoidance to explain configurations of state-CSO relations. The mechanisms to explain these interactional outcomes, however, vary; research differentially highlights state regulatory reach (Shieh & Schwartz, Reference Shieh, Schwartz, Schwartz and Shieh2009), varying levels of elite competition (Heurlin, Reference Heurlin2010), and organizational opportunity structures and the characteristics of organization employees (Hsu & Jiang, Reference Hsu and Jiang2015) to explain outcomes in state-NGO relations.

State-civil society dynamics have been examined extensively in the Turkish context as well, using comparable concepts. Studies address the Turkish state’s attempts to coopt civil society and blur the boundaries of state and society (Altan-Olcay & Icduygu, Reference Altan-Olcay and Icduygu2012). This process of cooptation “tames” civil society through what Yabanci (Reference Yabanci2019) calls either appropriation or containment. Studies also account for the ways the state actively refashions civil society around its ideological agenda. In the gender rights field, organizations close to the ruling party are empowered while other groups’ activities are circumscribed (Doyle, Reference Doyle2017, Reference Doyle2018). Often, assumptions of alignment with or opposition to the state are starting points for analysis in explaining outcomes in state-NGO relations. Less attention is paid to the ways these relations are negotiated by organizations (Özler & Obach, Reference Özler and Obach2018; Sarkissian & Özler, Reference Sarkissian and Özler2013).

While political or ideological proximity is a useful heuristic for thinking about state-civil society relations, political favor or disfavor are not the only determinants of these interactions. State-CSO relations may be shaped by the content of specific activities and issues at hand (Bahçecik & Turhan, Reference Bahçecik and Turhan2022). In service provision contexts, the balance between state capacity and organizational capacity determines whether states will control nonstate organizational service delivery or deliver goods jointly with organizations (Brooke, Reference Brooke2019; Cammett & MacLean, Reference Cammett and MacLean2014). An explanation of state-CSO relations must therefore account for organizational capacity and organizational activities without resting on overbroad assumptions of pro-or anti-government orientations.

Organizational Identity: Categorizing Refugee-Supporting Organizations

In Turkey, the state will sometimes “extend its political power through…NGOs” that carry out projects that advance a ruling-party agenda (Eder, Reference Eder2010: 181). Multiple studies of refugee reception in Turkey examine how the state prioritizes Islamically-motivated organizations while closing NGOs it sees as threatening (Aras & Duman, Reference Aras and Duman2019), controlling organizations that are not trusted (Memisoglu & Ilgit, Reference Memisoglu and Ilgit2017), or displacing them with government-created or loyal actors (Danış & Nazlı, Reference Danış and Nazlı2019; Sunata & Tosun, Reference Sunata and Tosun2019). These analyses focus on the religio-political compatibility or antagonism between the state and organizations. Nimer (Reference Nimer2020) broadens this focus to include how the ability to harness external funding can lead otherwise ideologically unaligned organizations to work in cooperation with the state. This point is critical, as it highlights the fact that relationships between state institutions and refugee-supporting organizations are also affected by other actors in the broader global refugee reception field, especially insofar as they shape organizational capacity. Organizations navigate this terrain deliberately, balancing interactions with state institutions, beneficiaries, benefactors, and other stakeholders (Akşak & Dimitrova, Reference Akşak and Dimitrova2021).

Given that external stakeholders determine organizational funding, partnerships, and ultimately survival, organizations adopt coherent identities that can be understood and evaluated by these audiences (Hsu & Hannan, Reference Hsu and Hannan2005; Zuckerman, Reference Zuckerman1999). Organizational identities are partially animated by what members see as central, enduring, and distinctive features of the organization (Albert & Whetten, Reference Albert and Whetten1985). Yet given the importance of audience evaluation, organizations are also mindful of labels and categories that signal specific types of commitments. A categorical distinction commonly made by actors and audiences in the refugee field is that between “rights-based” organizations on the one hand and “needs-based” or Muslim humanitarian organizations on the other (Boşnak, Reference Boşnak2021; Mackreath & Sagnic, Reference Mackreath and Sagnic2017). Yet, these distinctions are not necessarily clear in organizational activities. Moreover, some scholars suggest there is no rights-based movement in refugee support in Turkey; instead, organizations have largely focused on apolitical “service-oriented action” due to the restrictive environment (Paker, Reference Paker2019; Paker & Keyman, Reference Paker, Keyman, Baban and Rygiel2020: 110). The categories of “rights” and “needs” serve as labels that are legible to external stakeholders. Given that the language of rights-based development is a dominant paradigm among Western aid agencies, the label serves as a legitimizing signal to a broad swath of funding agencies (Kindornay et al., Reference Kindornay, Ron and Carpenter2012). In the Turkish context, however, the language of “rights-based” activities is counterposed against the ruling party’s increasingly undemocratic policies. Rather than taking the binaries that characterize refugee organizations at face value, the paper examines how these categories are deployed, combined, and assigned meaning, and what their implications are for relations with the state.

Findings: Organizational Capacity and Organizational Identity

Organizational features are summarized in the Table in the Appendix A. Organizational capacity is a broad concept, but it can be operationalized using concrete indicators of an organization’s ability to carry out projects to reach many beneficiaries. Resource aggregation is a key element of an organization’s capacity to mobilize (McCarthy & Zald, Reference McCarthy and Zald1977). Moreover, organizations must channel resources effectively. Measures of organizational capacity include the age and size of an organization, how formalized or institutionalized it is, and the amount and type of funding and political resources it can harness (Eisinger, Reference Eisinger2002; Suarez & Marshall, Reference Suarez and Marshall2014). In this paper, organizational capacity is approximated using a combination of key indicators: geographic scope of refugee support operations, number of employees, how long an organization has been operational, and external funding sources.Footnote ²

Funding sources for refugee projects are the most important indicator, given that they enhance organizational capacity relative to state institutions. Time of establishment and number of employees are used as a proxy for organizational institutionalization and professionalization. I categorize organizations with fewer than 50 employees, operations that do not expand beyond Istanbul, and low external funding coming from donations or small grants as low capacity. High-capacity organizations are more variable: some work only in Istanbul but still bring in a variety of international funds (for example, Organization 21). Organizations that can employ more than 50 employees and work in multiple provinces are categorized as high capacity. For organizations that span both low and high categories across indicators, I arbitrate based on the breadth of their operations in comparison with other clearly low- or high-capacity organizations. A few organizations still fall in an ambiguous space, which I categorize as medium capacity and address briefly in the conclusion.

Organizational identity combines three indicators: type of organization, type of activities, and, most importantly, key self-descriptors. Organization type indicates (1) if an organization is managed by foreigners, refugees, or Turkish personnel and (2) whether and how an organization is locally registered. Whether an organization appears foreign or local matters both for how it situates itself in local political contestations as well as how state institutions perceive it. Self-categorizations are key descriptors drawn directly from interview data where employees answered questions about organizational goals and purpose, as well as from mission and values statement published on organizational websites and in yearly activity reports. Finally, a summary of activities animates the self-categorizations and provides more details about the organizations examined in the paper. It is important to note that many organizations feature indicators that span both rights and needs identities, falling in a neutral service-oriented space. In these cases, audiences may categorize organizations based on selected features. As a result, a non-conforming organization may be categorized as “rights-based” or “needs-based” depending on how the state fixates on specific salient features.

High-Capacity Organizations

This paper finds that when organizations have high capacity, they will partner with state institutions to achieve shared goals in refugee support. High-capacity organizations take on a role as agents of state institutions for projects where their goals are aligned, however narrowly (Shapiro, Reference Shapiro2005). Organizational identities matter insofar as how they cooperate with state institutions and external funders. Organizations that coalesce around refugee protection activities either self-categorize as or are categorized by the state as “rights-based”; these organizations, which harness ample INGO and Western NGO funds, make programmatic compromises in order to jointly implement projects with state institutions. While their rights-based agendas help harness international funding, they make ruling-party actors wary of possible anti-government claims-making. Those, on the other hand, that frame their projects around needs, humanitarianism, and charity, and receive funds from international Islamic organizations signal greater alignment with the state objective of providing aid to vulnerable Muslim refugees (Memisoglu & Ilgit, Reference Memisoglu and Ilgit2017). These organizations enter collaborative arrangements with state institutions. There are two main exceptions: first, when an organization spans rights and needs codes, it has more difficulty justifying itself to the state as a reliable partner. Second, if a high-capacity organization is already fully enmeshed with the state, its identity matters primarily as a signal to external actors rather than domestic institutions, leading to the existence of collaborative rights-based organizations.

Lower-Capacity Organizations

Most refugee-supporting organizations do not have broad reach and extensive external funding. Smaller organizations tend to have a targeted set of programs that they implement in the field. They must still strategically navigate interactions with state institutions, from registration processes to the implementation of projects. These organizations are not of practical use to state institutions for the externalization of provision tasks, given their smaller budgets and limited personnel. As such, the power imbalance between state and organization is starker, and the point of encounter with state institutions is centered on applying for permissions and program protocols. Lower-capacity organizations either selectively engage/avoid state institutions or embrace a neutral identity as justification for operating autonomously. Positioning as rights-based or community/aid oriented is centrally important to their interactions with the state.

Selective Engagement and Strategic Avoidance among Rights-Based Organizations

Organizations that identify as rights-based tread cautiously in their interactions with state institutions. Unlike their larger counterparts, small rights-based organizations offer limited benefit to expanding state capacity to reach refugee populations. Without this benefit, their function as conduits for claims-making is a political nuisance. Invoking clear categories, one administrator of a small organization described difficulties in running programs in the post-coup environment: “Rights-based programs with a human rights perspective have been especially affected in the current situation. Our capacity to work in the public sphere has shrunk a lot at this point…So, we have to find new ways to carry out programs…we’ve started doing online programs to limit risk.”Footnote ⁴ Instead of dealing with state agencies, this organization adapted its educational program for teachers working with refugee students in a way that would protect the organization and its program participants from state intervention (Organization 18). This avoidance also allowed the organization to make adversarial claims that may have been stymied under more direct regulation. The organization maintained a strong political motivation in its daily work: “Our main focus is democratization; we could say [democratic] transformation is our basic aim…we deal with various topics including transparency, participation, and rights.” In this sense, the organization’s avoidance of state protocols may have reduced the reach of activities the organization could carry out, but it also helped minimize the moderating effects that regulation can have on pursuing organization goals.

Even so, smaller organizations that identify as rights-based organizations were not completely avoidant of the state. Rather, they maneuvered local institutions to find those with which they could partner while remaining wary of antagonistic institutions. Another self-categorized rights organization explained that they advocate for refugees who are marginalized by existing policies, including Kurdish Syrians and Doms.Footnote ⁵ They noted the challenges of carrying out programs. “[T]he protocols are so tough. There are some who get them but they’re mostly large organizations and even for them the requirements are really limiting.” They lamented that given the state’s distrust of rights-based organizations, their funding has dried up. Yet they find ways to gain institutional support for programs by creating strategic state partnerships.

We can’t work with our own municipality because they come and question us, they’re AKP. But we can work with [an opposition-run] municipality….We’re able to form really nice partnerships with the municipalities that are open to a rights-based approach but unfortunately the one we are located in is not one of those (Organization 14).

By keeping actions geographically targeted, the organization was able to enhance its capacity without compromising its commitment to advocate on behalf of marginalized members of the community.

Other organizations, in contrast, emphasized a neutral stance as community-oriented humanitarian organizations while minimizing their capacity in order to bypass bureaucratic challenges. One manager described their organization:

It’s not registered, it’s a volunteer-run space. We did try to register earlier…And then we submitted our paper at the time just before the coup happened. So then things got stuck and a lot of people who were on the board either backed off or left, you know, because the situation was so fragile…And because we were not thinking of expanding or going beyond what we were doing at that point, like getting funding from large grant or funding agencies, we didn’t really think that it made much of a difference if we are registered or unregistered” (Organization 12).

The organization leadership decided that small workshops and in-kind aid distribution in the community were sufficiently neutral. They kept their operations modest while partnering with organizations that are officially registered for any larger-scale outreach. Moreover, they were able to continue their operations quite comfortably even with the knowledge of street-level bureaucrats. “The municipality and the muhtars Footnote ⁶ know about us. The municipality also used to send us stuff to distribute, but they never bothered us.” They developed a neutral needs-based orientation as the key to their unobtrusive existence in the neighborhood.

Lower-capacity organizations that may have previously had missions with potentially political valences also neutralized their own operations over time to avoid state intervention and work autonomously. Organization 8, which used to provide services to a marginalized Christian Syrian group, had over time become a more general community center for migrants and locals. After the post-coup crackdown on CSOs, they tread cautiously. “When the [restrictive education] legislation was passed we didn’t do anything for seven to eight months….During that time we were so scared that a government official would show up unannounced and say, ‘What are you doing here?’…they were closing so many organizations at the time, and we knew they wouldn’t feel bad closing us down.” Over time, they gradually implemented small lessons and community-building activities that they did not call “education,” avoiding protocol registration processes they assumed would be unsuccessful. For small organizations, local community-oriented neutrality proved sufficient for continuing operations in peace. The findings of the paper are summarized in Table 1:

Table 1

How organizational identity and organizational capacity shape outcomes in state-CSO relations in refugee support

Capacity	Organizational identity
	Rights-based	Non-conforming	Needs-based
High	Compromised cooperation	Variable forms of cooperation depending on pre-existing relations with state or external actors	Collaboration
Low	Selective engagement/avoidance	Autonomous operations	Autonomous operations

Discussion and Conclusion

Much of the existing research on state-civil society relations in non-democratic contexts in general, and in Turkey in particular, designate organizations as aligned with or opposed to ruling regime policies as a starting point for analysis. In contrast, this paper has focused on a field where state policies and organizational goals may overlap across a complex political terrain. Moreover, in the refugee field, external actors centrally shape state and organizational capacity to carry out projects. Domestic political contestations are just one factor that shapes state-organization relations. For this reason, rather than assuming allegiance or opposition between organizations and the state a priori, this paper has analyzed how organizations situate themselves in relation to the state and funding agencies by invoking a range of labels and categories to represent their work. Their external identification by both international and domestic audiences shapes their relations with the state. Organizational identities must first be invoked and interpreted by relevant actors before they can explain interactional outcomes.

The paper has presented a framework that accounts for organizational identity and capacity to explain how organizations in a restrictive state context variably interact with state institutions. When organizations have high capacity, particularly buttressed by external funding, they will carry out projects in partnership with state institutions across a political spectrum. Organizational identity dictates whether this relationship will be one of compromise with narrow goal alignment or broader collaboration in project generation and implementation. In cases where state-CSO relations are already firmly established, identities are molded to appeal to external funders. For lower-capacity organizations, organizational identity shapes whether and how organizations engage state institutions. The more adversarial the organizational identity to ruling-party policies, the more strategically the organization will navigate permissions and partnerships with state institutions. More neutral-presenting organizations can circumvent registration and other formal bureaucratic procedures, operating autonomously.

It bears emphasis that some organizations do not fit neatly into what is necessarily a simplified framework. Some organizations, for example, have significant capacity in one specific service area due to a large project-based grant from a funding agency (Organizations 9, 15, 22). In these cases, organizations often register and carry out their projects according to state strictures, without invoking language or labels with political valences in the Turkish context. Instead, they draw on notions of community, empowerment, social responsibility, and inclusion. This is particularly true of foreign organizations that register and work locally in Turkey. Their interactions with the state are marked by regulatory oversight, and organizations will often compromise to be able to carry out their specific projects.

As Nimer (Reference Nimer2020) argues, and this paper further underscores, it is critical to consider actors beyond the state when analyzing state-civil society relations. Further research should test this framework across different policy or aid fields where actors outside the country of focus—INGOs, states with development funding, and other funding agencies—may have an interest in improving local conditions. This could include development aid in agriculture, healthcare or education, natural disaster relief, or post-conflict aid. Additionally, this study has focused on only one city. Whether these patterns are consistent across different regions remains a question to be empirically tested. Finally, more fine-grained analysis of funding ledgers could further elucidate and hone the framework by examining variation within high-capacity organizations in terms of how certain types of funding may lead to varied configurations of state-organization interactions. Alignment with and opposition to non-democratic state institutions are critical to shaping state-civil society relations. Yet CSOs operate in complex environments where their activities may not break along clear political cleavages. This qualitative study has sought to shed light on how state-CSO relationships are actively framed, negotiated, compromised, and re-cast by organizations in light of pragmatic considerations to effectively implement programs.