Government and Civil Society Organizations: Close But Comfortable? Lessons from Creating the Dutch “Strategic Partnerships for Lobby and Advocacy”

Motivation. Governments commonly support the advocacy role of civil society organizations (CSOs) in development, but studies argue that close linkages between government and CSOs are problematic. The Netherlands&#8217; Dialogue and Dissent policy programme brings together the Netherlands Ministry of Foreign Affairs (NMFA) and 25 (alliances of) CSOs as advocacy partners. The programme is set up from an awareness of the challenging nature of such collaboration. Purpose. We contribute to understanding the ways in which issues with donor&#8211;CSO collaboration can be engaged in donor policies. This article addresses two questions: To what extent and how does this programme confront and overcome the challenges of close collaboration between government and CSOs? What practical lessons can be learned?. Approach. We conducted 33 (group) interviews with CSOs and policy officers involved with the Dialogue and Dissent programme, exploring their understandings, expectations and strategies as partners in the programme. We also analysed CSO programmes, policy documents and publicly available information. Findings. In Dialogue and Dissent, space for dissent, flexibility and relative equality between government and CSOs positively address mutuality and institutional pressures. Challenges remain as estimated strategic significance, diverging capacities and risks to autonomy work against mutuality. Certain challenges are engaged with, but we identified no strategies countering pressures that stem from managerialism within the NMFA, external political pressures and conflicting government objectives. While the programme counters tendencies towards institutionalization of CSOs as insiders, some important challenges to public engagement identified in the literature remain insufficiently addressed. Conclusions. Conditions built into policy can address challenges identified in the literature. However, challenges remain that are rooted in wider organizational and political realities. Lessons for practice are: (1) the advocacy role of CSOs can be advanced by building in certain formal conditions and advancing these informally; (2) flexibility allows for collaboration to develop as government and CSOs negotiate their roles, cognizant of pressures that get in the way; (3) challenges rooted beyond the support policy and government agency involved in the collaboration will constrict the power of policy design; (4) public engagement as a foundation for CSOs&#8217; advocacy roles deserves much more attention.


|
identity, expertise, experience and networks, as well as respect for each other's independent roles and responsibilities" while also "identify [ing] opportunities for joint, complementary action to effectively advocate change and influence policy" (Government of the Netherlands, 2014, p. 2).
Overall, donors and academics recognize the advocacy role of CSOs in development, based on their potential to contribute to development through advocacy. In the context of development, advocacy's overall aim is to combat the structural causes of poverty and injustice. CSO advocacy is, therefore, a tool to fight the causes of poverty or injustice and influence structural change by challenging existing power structures. Advocacy includes "outsider" and "insider" strategies ranging from public campaigning to lobbying. This goes beyond influencing policy; it aims to make sustainable contextual changes through, for example, awareness-raising, legal actions and public education, as well as building networks, relationships and capacity (Barrett et al., 2016). It also includes engagement in structural collaboration with state and private actors, as is commonly found in multi-stakeholder processes and governance. This conceptualization moves away from understanding advocacy as linear efforts to influence decision-makers.
It is widely accepted that, through advocacy, CSOs can help set development agendas by calling attention to certain situations, influencing the definition of issues and promoting their prominence in public debate. CSOs can help to capacitate people in vulnerable positions by articulating their needs and demands. CSOs can also insert evidence, testimony and propositions into policy processes, and collaborate across policy levels, offering some measure of transnational democracy. This is much needed when issues, interests and political processes cross state borders (see e.g. Barrett et al., 2016;Green, 2016;Keck & Sikkink, 2014;NORAD, 2016;SIDA, 2013). Numerous state donors, including the US, the UK, Sweden, Norway, Denmark, the Netherlands and Austria, support civil society advocacy (see e.g. Arensman et al., 2015;DFID, 2014;NORAD, 2016;SIDA, 2013). Typically, CSOs apply for donor funding through programmes that include advocacy as a strategy.
Although donor support for CSO advocacy is widespread, the NMFA's approach is unusual (Banks et al., 2015) in making CSOs' advocacy role the focus of an extensive policy programme. Several considerations were behind this choice. First, as ministry staff members informed us, the choice was motivated by the felt need to work strategically with a budget that had shrunk substantively in the wake of the economic crisis following the crash of 2008. Advocacy was a niche that few donors had occupied. Second, there was an assumption that advocacy could potentially yield high impact with small investments. Third, the focus was very much influenced by the personal views of Lilianne Ploumen, the Minister for Foreign Trade and Development Cooperation at the time. It is likely that the choice for advocacy partly stemmed from her belonging to the Social Democrat party, with a long history of affinity with advocacy and social action, and her own past leadership roles in women's rights organizations. 2 With the Dialogue and Dissent programme, the NMFA not only supports advocacy, but also takes as a given that the advocacy role of CSOs can be advanced through close collaboration with the government of the Netherlands. This effort to contribute to transformative political processes by supporting autonomous civil society action through NMFA-CSO partnerships runs counter to the critical literature on such collaborations. In the Dialogue and Dissent programme, the NMFA, 25 largely | 731 VAN WESSEL Et AL. The Dialogue and Dissent programme presents the interesting complication of seeking to support the advocacy role of CSOs through close collaboration, with an awareness that donor-CSO relations can work against this role. Seeking this close collaboration can be seen as directly opposed to Banks et al.'s (2015) recommendation that donors "step back." This presents a valuable opportunity to address two key research questions. First, how and to what extent does this programme confront the challenges associated with close government-CSO collaboration? And, second, what can be learned from this case in terms of how donors and CSOs can work together to advance CSOs' advocacy role? This study aimed to contribute to the understanding of how challenges in donor-CSO collaboration can be engaged in donor policies.
We begin with a theoretical framework charting the main challenges of government-CSO collaboration for CSOs' advocacy role. Then, after summarizing the study methods, we introduce the Dialogue and Dissent programme and present the findings, elaborating on how Dialogue and Dissent engages with the identified challenges. A concluding section integrates the main findings and presents implications for practice.

COMFORT?
Collaborations between governments and CSOs tend to be complicated, with important implications for CSOs' advocacy role. Existing work primarily points to three fundamental issues cautioning against close government-CSO collaboration, suggesting, implicitly or explicitly, that these have implications for CSOs' advocacy role.
First, the literature on partnerships between governments and CSOs shows that power imbalances in government-CSO relationships tend to inhibit mutuality-the opportunity to participate and influence equally within a collaboration (Brinkerhoff & Brinkerhoff, 2004). These imbalances are largely rooted in funding relationships and are expressed in moderate rather than radical CSOs receiving support-through self-selection because radical organizations do not apply for funding, and through the domestication of supported CSOs by threatening to withdraw funding. Co-optation is a serious risk (Banks et al., 2015;Bloodgood & Tremblay-Boire, 2016).
Second, institutional pressures linked with government-CSO collaboration inhibit working towards social transformation. A short-term orientation and weak connections with deeper political processes keep action within certain bounds. Donor agencies' administrative procedures and practices can undercut the full expression of partnership principles. Donors tend to see partnerships as instruments to achieve efficiency and effectiveness, rather than social transformation (Banks et al., 2015;Brinkerhoff & Brinkerhoff, 2004).
There is also, however, research that finds that donor-CSO relations may not be as detrimental to civil society's advocacy roles as these three issues suggest. Donor-CSO relations are political and riddled with issues of power. However, as this strand of research argues, this does not mean that power is a one-way street. Dynamics can be more complicated, creating space for strategic navigation on the part of CSOs. Hermansen et al. (2016), Banks et al. (2015, p. 712) and Craig et al. (2004) point out that CSOs may manoeuvre strategically, keeping their own views and interests central to their operations, and that mutual dependency between CSOs and governments can counterbalance funding dependencies. In addition, collaboration may not demand compromise from CSOs when they have congruent interests, viewpoints and approaches with those of the government (Hermansen et al.,. Regarding the institutionalization of CSOs, Hermansen et al. (ibid.) also question simple dichotomies between insider and outsider strategies, stating, for example, that space for critique on the inside can make public action redundant: sharing responsibility for policy can offer policy influence without public dissent. As for institutional pressures linked with government-CSO partnerships, Brinkerhoff et al. (2002, p. 27) argue that partnership can contribute to flexibility. This suggests the need to look more closely into organizations' capacity to navigate the tensions that may arise in collaborations with governments and how this type of collaboration may actually advance CSOs' advocacy role.

| THE DIALOGUE AND DISSENT PROGRAMME
As stipulated in the programme's theory of change (ToC), Dialogue and Dissent seeks to advance a "social transformation logic," as opposed to "managerialism" (Kamstra, 2017). Managerialism frames social problems as technical and to be resolved using management and planning tools. In the NMFA's understanding, this perspective has led to a corresponding development policy that works from a linear view of policy and assumes that development can be managed and controlled by identifying the right variables, setting the right targets and employing the right strategies. In this view, "CSOs mainly have a technical service delivery role in promoting development rather than a political one" (Kamstra, 2017, p. 4) and are viewed instrumentally: they are to complement rather than oppose the efforts of states and donors. With the managerialism approach, CSO-donor relations are short-term, contractual and top-down (Elbers et al., 2014 as discussed in Kamstra, 2017, p. 4).
Dialogue and Dissent's social transformation logic sees underdevelopment and poverty as stemming from social, economic and political processes of exclusion and domination, and understands development as "an inherently diffuse political process of addressing power asymmetries and the claiming of rights" (Schmitz & Mitchell, 2016 as cited in Kamstra, 2017, p. 5). For the NMFA, development centres on the "emancipation and empowerment of people who are left behind because of all kinds of (structural) inequalities," requiring local ownership (Kamstra, 2017, p. 5). The role of CSOs, then, goes beyond service delivery, also "addressing the root causes of poverty by challenging the underlying mechanisms of social, economic and political exclusion" (Banks & Hulme, 2012;Banks et al., 2015, both as cited in Kamstra, 2017. This implies that donors should "complement and strengthen the efforts of their local partners by providing financial, institutional and moral support" (Elbers et al., 2014as cited in Kamstra, 2017. Donors should engage in long-term, open-ended and flexible relationships with CSOs, based on equality and trust, and CSOs should be allowed space to develop and change their agendas autonomously (Kamstra, 2017, p. 6).
Notwithstanding its focus on partnership, Dialogue and Dissent is also-perhaps fundamentally-a funding programme. Dutch CSOs or Dutch CSO-led consortia applied for five years' funding from a total of EUR 1 billion. Beyond broad alignment with the Netherlands' development policy agenda, no directions were given regarding topics or agendas in the policy framework or call for applications. Applicants were requested to submit theories of change and an account of their track records in advocacy. From 65 applications, 25 strategic partners were selected in 2015.
A wide range of themes were addressed in the applications, including sexual and reproductive health and rights; disaster risk reduction (DRR); labour conditions; corporate conduct; security and | 733 VAN WESSEL Et AL. human rights; and ecosystem governance, climate and environment. Although many of the CSOs had track records of constructive "insider" engagement with governments and/or private-sector targets (i.e. consultation, informal discussion and evidence-based policy advice), some had used more confrontational strategies (e.g. campaigning to mobilize public opinion). Oxfam Novib, Amnesty International, Hivos, Cordaid, Friends of the Earth Netherlands, Red Cross Netherlands and the International Union for Conservation of Nature Netherlands are some of the better-known funded applicants.

| Partnership
Selected applicants were asked to further develop their programmes, in close consultation with the NMFA and their CSO partners in LICs and LMICs. Here, the NMFA explicitly tried to be "more than a donor." Applicants were responsible for developing their programmes, but the NMFA was to be written into the programmes as a partner with whom activities could be carried out, or who could take an active supporting role. Each applicant was matched with a thematic NMFA department working in the same policy domain. Applicants and the departments then to various degrees consulted with each other regarding the roles departments would play. Some departments were also involved in other decisions such as country selection. Largely, however, the CSOs identified countries and CSO partners in these LICs and LMICs. Opportunities were also sought for potential collaboration with Royal Netherlands Embassies. The process resulted in "partnership agreements" between the CSOs and the NMFA, charting broad initial agreements on objectives, responsibilities, roles and forms of collaboration. Relations and collaborations continued to be explored and developed after this initial stage through seeking out opportunities and matching approaches and capacities.

| METHODS
The NMFA's programme seeks to advance the advocacy role of CSOs in LICs and LMICs, primarily through CSO partners with whom the NMFA interacts directly and with which it has contracts. Thus, this article focuses on the advocacy role of the NMFA's direct, mostly Netherlands-based, CSO partners. We selected 14 CSOs or CSO consortia for semi-structured (group) interviews. We used selection criteria ensuring coverage of the breadth of the 25 CSOs/CSO consortia in terms of thematic focus, "insider" and/ or "outsider" orientation, consortium complexity, the core business of the lead party (service delivery or lobby and advocacy), geographical orientation and length of relationship with the NMFA. In addition, "outliers" (with a particularly sensitive thematic focus or a Southern-based lead CSO) were included to provide the widest possible spectrum of programme expectations. In each CSO or consortium, interviews were conducted with two to four closely involved staff members. We also conducted 12 interviews with single or multiple (up to three) NMFA policy officers serving as so-called liaison persons for the selected CSOs/consortia, and conducted two group discussions with the Civil Society Division. Furthermore, we interviewed staff members at five Royal Netherlands Embassies selected with consideration of geographical spread, as well as the embassies' level of involvement with the partnerships. These interviews took place via phone or Skype and were with one or two staff members.
Considering the timeframe of the programme, we were able to chart partnership expectations and strategies as they developed over the 18 months from the selection of the CSOs/consortia (January 2015), when interactions on partnership started between the NMFA and the CSOs, through the first six months of the programme, which began in January 2016. Interviews were recorded, when possible, and subsequently transcribed.
We also held informal preparatory meetings with staff members of the NMFA's Civil Society Division, the department with the leading role developing and managing the Dialogue and Dissent programme. In addition, we analysed a number of NMFA policy documents, including the policy framework, the ToC supporting the programme, assessment forms, partnership agreements, 25 partnership programme documents, as well as publicly available information illustrating the unfolding of the programme.
The interviews and documents were analysed using Atlas.ti qualitative data analysis software. Throughout this article, the anonymity of all interviewees and organizations involved is preserved because of the sensitivities involved.

| FINDINGS
In this section, we examine how the Dialogue and Dissent programme has engaged with each of the three fundamental issues discussed in the theoretical framework (mutuality, institutional pressures and the institutionalization of CSOs as insiders), and their implications for collaboration.

| Mutuality
Existing literature on partnerships between governments and CSOs shows that power imbalances in government-CSO relationships tend to inhibit mutuality-the opportunity to participate and influence equally within a collaboration (Brinkerhoff & Brinkerhoff, 2004). To what extent does the programme provide opportunities for CSOs and the NMFA to participate in and influence their collaboration on equal terms? Below, we discuss the ways the set-up of the programme has been conducive to mutuality, challenges to mutuality that emerged, and the way that these have been addressed in the programme. CSO interviewees were mostly very positive about working closely with the government because of potential benefits for their advocacy work. In their understanding, and sometimes also experience, this collaboration could provide access to important actors and arenas, "opening doors." The government's convening power could bring CSOs together with actors impossible for them to engage directly. The NMFA in The Hague, and to some extent the embassies, were also seen as important advocacy targets that were more accessible as partners and open to CSOs' dissenting voices. The NMFA could also be an ally in strengthening messages and broadening their reach. Through partnership with the NMFA, CSOs could learn important lessons, such as government perspectives and the workings of the state in different countries. The embassies could also play an important role by protecting activists and organizations. CSO interviewees also commonly thought synergies could arise within partnerships because of the complementary nature of potential contributions to programmes. They expected these synergies in complementary expertise, complementary networks, and complementary lobby and advocacy roles.
The department liaison persons working with the CSOs generally valued collaboration in the partnership and welcomed the prospect of working with CSOs on an equal basis. Interviewees often stated they were ready to work together, embracing the idea that partnership can contribute to strengthening the capacity of civil society, which they described as important both in its own right and for the departments. The departments could learn important lessons because the CSOs provide access to information "from the ground" and serve as sparring partners for the NMFA, helping to attain inclusiveness | 735 VAN WESSEL Et AL. and bring in issues, ideas and approaches that would otherwise be neglected. Although these interviewees were not necessarily excited about engaging with dissenting CSO voices, these voices were accorded legitimacy (van Wessel et al., 2017). Liaison persons also saw the NMFA as having at least some capacity to support CSOs in their work (e.g. by "opening doors," bringing them together with other policy actors or taking their message to international policy arenas).
Such viewpoints indicate relatively equal relations between the NMFA and the CSOs, potentially contributing to mutuality. Such mutuality was seen as being strengthened through the gradual building of relationships, allowing sufficient time and space to search for common ground and opportunities, and through the use of ToC (instead of logframes) and regular interaction. Aiming to develop activities together means that relatively equal relations and interactions can develop between CSOs and the NMFA. Although many NMFA staff members have a history of more hierarchical, managerial relations with CSOs, they have generally embraced more equal and collaborative relations. Such relations facilitate communication between partners, including confronting differences in relatively open ways. We also found that both NMFA staff and CSOs saw themselves as aligned on many issues and envisaged-and had sometimes also enacted-complementary collaborative roles. Notably, some interviewees from both sides described their collaborations as a continuation of existing practice, as many of the partners already had longstanding collaboration experience with the NMFA. For some, Dialogue and Dissent was partly a continuation of previously existing policy, but challenges to mutuality were also identified in the programme. We will turn now to these.

| Strategic partnership?
Interviewees had diverging and often diffuse understandings of what should constitute partnership. Many CSO interviewees saw important strategic opportunities to harness the NMFA's power to advance CSO objectives. They also found, however, that their liaison persons were often not as ready to commit to shared objectives or responsibilities as they had hoped, and they sometimes felt that the NMFA had no clear ambition for itself in the partnerships: I remember you [speaking to colleague] you wrote that you asked the Ministry: "What are you bringing into the partnership?" Because we had to include in the proposal who was going to do what. And they had said, "You should write down what you expect from us." So little came from that. I think that's peculiar. When you talk about partnership and you are looking together at what everybody can contribute, but then we have to tell the Ministry what we expect from them. Without us knowing whether they can live up to that.
Most department liaison persons did not clearly conceive of the CSOs as potentially strategically useful actors in the context of their own policy work. Rather, the department liaison persons approached collaboration with CSOs primarily in terms of support, and occasionally in terms of risk.
Arguably, this challenge to mutuality was built into the Dialogue and Dissent programme. The programme's starting point was the provision of substantial financial support to CSO programmes based on their ToC and track records, linked to a partnership role for the NMFA. This role was to take shape after selection and be incorporated into programmes primarily devised and owned by the funded CSOs. Mutual relations of shared responsibility, trust and shared objectives were therefore to develop after the selection process. Moreover, the responsibility to take up a partnership role was largely at 736 | VAN WESSEL Et AL. the discretion of NMFA staff members. In some cases, previous collaborations had fostered trusting relations, but these sometimes had to be created from scratch.

| Capacity
Interviewees identified an imbalance in information and ability to commit to the programme as a second challenge to mutuality. The NMFA's role in the partnership was initially explored via interaction between one ministerial staff member, who also had many other tasks, and a large CSO or consortium team. Department liaison persons emphasized their limited time availability. The partnerships were added to their already heavy workload, meaning that they had to limit their involvement: It is a shame; most colleagues at Foreign Affairs have so little time for this. I would like to spend much more time on partnerships, looking into the content and more deeply into what the possibilities are and really explore the advantages of this collaboration.
Capacity was also sometimes a matter of available expertise, as agendas and topics sometimes differed between the CSO programmes and the department or embassy. The NMFA's capacity limitations led to some cases where the CSO-NMFA partnership did not turn out to be as important as the policy framework suggests or many of the CSOs would have liked.

| Autonomy
A third challenge to mutuality is that close association may compromise autonomy. Some NMFA and embassy staff members feared embarrassment from CSOs doing things that could negatively affect the Netherlands' relations with other states or the NMFA's relations with other actors. Partnership could be misunderstood or misrepresented if the involvement of the Dutch government is publicly perceived as a statement of approval or if a CSO falsely claims to articulate a Dutch government position. CSO representatives stated that association with the Dutch government could compromise their own autonomy or that of their partners in LICs and LMICs, who might be challenged for being associated with a foreign government: It's a risk for partners. If you are a grassroots organization and you are financed for 80% by a foreign country, you are less legitimate. In this programme, this problem is more explicit. The Ministry is a partner, not a funder in the background. If people start digging, it becomes easier to claim that the Dutch government is intervening in [country].
Some interviewees from the NMFA were also conscious of this risk. For them, this issue was rooted in the "space for dissent" in the programme. Although they did not challenge this space as such, for them, dissent generally raised the question of how to manage it.
In some cases, to improve the alignment of policy priorities, country selection and strategy in the programme documents, department liaison persons sought to maximize the programmes' compatibility with the departments' work. Other liaison persons did not have this ambition, and some were reluctant because they felt this was inappropriate given their partner role, and CSOs generally found that their autonomy was respected. Attempts to influence the CSOs did, however, sometimes | 737 VAN WESSEL Et AL. become sensitive, and some CSO interviewees described these as invasive efforts to limit their autonomy.

| Conflicting government objectives
Conflicting objectives within Dutch government policy constitute a fourth challenge to mutuality. NMFA departments working to advance the economic interests of Dutch business, for example, might be displeased that another department in the same ministry supported CSOs that were challenging trade agreements (e.g. to address economic inequalities or environmental issues involving multinational corporations).
According to both groups of interviewees, collaboration in strategic partnerships can put the NMFA in difficult positions. Pressure may build, especially when conflicting interests in the NMFA are at stake. For example, tensions can appear in relation to another ministry or between NMFA departments. A department liaison person we interviewed described this, discussing how the Dialogue and Dissent programme is sometimes confronted because of other policies where the NMFA is involved: There are colleagues at [department] who get angry. We are partners with [CSO]. "How dare they send a letter against [trade agreement]?," I hear them say.
CSO interviewees sometimes experienced the limits of partnership with the NMFA, including uncertainty among CSOs about the value of partnership when the NMFA is under political pressure. As a CSO director explained: With this form of partnership, with the plan that we have developed and submitted, and the certainly good talks with the Ministry, I still do not feel that the Ministry really takes collective responsibility for the programme. Absolutely not. What will the Ministry say when in Parliament questions are raised about this? And whilst our plan is modest, others go much further with lobbying. As partners, you are not responsible for each other's programmes. But I expect to be politically covered for what we do. The plans have been approved and we are held to implementing them. I really doubt whether that cover will always be there.

| Strategies to strengthen mutuality
For many interviewees, mutuality develops through the growth of mutual understandings, in tandem with exploring possibilities for acting as partners. Learning how to be partners involved seeking common ground in the face of differences, starting from expectations, and responding to new lessons learned about possibilities and limits. A CSO interviewee reflected on how his organization experienced this exploration: We realize our own capacity limits and the limitations of the Ministry. Let us not tie ourselves to a giant Christmas tree of expectations. Let us look more realistically: These are the five themes we bring in. Let us zoom in on where we think there are realistic possibilities and where we, if we really work together, can really get something at the table about which we think, this is advantageous and also manageable for exactly that reason. VAN WESSEL Et AL. Because, if we come up with something and within our organization it is rejected because it doesn't relate to our priorities or whatever, or within the Ministry we have to go against the flow, this is not particularly helpful.

|
In addition to this exploration, to tease out possibilities while accepting limits, those involved sought to advance the desired mutuality by mitigating differences. One strategy was communicative in nature. Civil Society Division staff members took on the role of championing the programme within the NMFA, seeking to build relations between CSOs and departments (referred to jokingly by a staff member of the Division as "relationship therapy") to overcome tensions arising from different viewpoints or a lack of trust.
A second strategy used by both NMFA and CSO staff members consisted of navigating the political risks of close association, for example, by seeking distance from partners when politically expedient. When there are differences in viewpoints or approaches, or on certain issues, the NMFA or the CSOs could choose not to get involved, so, for that particular situation, their association would not be clearly visible. As an embassy staff member explained: With food security, I don't see very contentious issues, but when it comes to land, sexual minorities, sexual and reproductive health, young people, that is going to challenge the government of [country]. We have been asking partners to think about risk analysis. The embassy doesn't want to be caught off guard […] In some situations, being visible as an embassy can be counterproductive.
A related expression of this strategy was keeping each other informed to allow the NMFA to prepare the response to their CSO partners' politically risky action. Department contact persons and embassy staff spoke of directly approaching the CSOs concerned so that these situations come more predictable and manageable. In this way, strategies can be negotiated and responses from the ministry or the relevant embassy to possible backlashes can be carefully prepared.

| Institutional pressures working against transformation
The second issue identified in the literature cautioning against donor-CSO collaboration is that institutional pressures, such as planning, reporting and results requirements, may curb the advancement of social transformation. Below, we discuss the way the NMFA has sought to address these pressures in the set-up of the programme, and the limits to this given the broader context in which it operates.
The NMFA, aware of the potential problem of institutional pressures, consciously sought to build flexibility into their policy's design. The NMFA works with theories of change (ToCs), rather than log frames, to specify plans and to assess programmes and their results. ToC is increasingly used as an approach to shaping, monitoring and adapting interventions in relation to the complexities of change processes and contexts (Stein & Valters, 2012). The different conceptualizations of this approach agree that ToC offers a vision for grounding action, providing space for learning-based development and adjustment. This is exactly what Dialogue and Dissent offers, through ToC but also by other means. The programme also offers CSOs the freedom to define their own results and shape their own evaluation; an external end-evaluation is compulsory, but its development and execution is left to the CSOs. Moreover, a year-long inception phase at the start of the programmes, with a strong role for partner CSOs in LICs and LMICs, was set up to guarantee the programmes' bottom-up development. CSO interviewees were encouraged by this flexibility: | 739 VAN WESSEL Et AL. I see a tremendous effort to be consistent in the approach, having it impact on reporting and the pushing back of reflexes at the Ministry. I have real respect for that. Something has really changed. Whether we will manage to keep that up in the sphere of accountability I don't know. Internally, we try to contribute in a way, saying gee, there is much more space, potentially, to reflect on one's work.
However, the Dialogue and Dissent programme is implemented in a context that is not necessarily conducive to its transformational approach. First, managerial approaches are evident within the NMFA in terms of the roles and expectations of at least some involved thematic departments and staff members, in terms of both understandings of the programme and concrete practices. For some NMFA staff members, working with ToC led to uncertainty regarding engaging with the programme and envisaging their own roles and responsibilities. Their verifying role, which fits more with a managerial perspective, was especially difficult to reconcile. Liaison persons often interpreted their own roles and responsibilities in terms of safeguarding proper implementation and concern for effectiveness, rather than sharing responsibility with the CSOs. Relatedly, they saw themselves faced with many questions concerning the programme's effectiveness: can ToC work? Can our partnership work? What will the impact be? Can this contribute to our thematic goals? What can we say when parliament asks us to account for the money spent? They often interpreted open-endedness as problematic uncertainty rather than as providing the necessary space for CSOs.
Many CSO interviewees challenged the combination of partner and managerial roles in a single liaison person, asking, for example, how open and honest they could be with each other when these roles are combined: There is a tension between the role of the subsidy provider and the partner. On the one hand, controlling conditions for subsidy, and that it is all accounted for, and on the other hand they have to think along in a positive way. Such roles do not necessarily go together. In the collaboration until now, I sometimes see that. Not doing much at the time of the few-pager [document charting collaboration plans]. And then you submit something and it is assessed like a sort of application for subsidy, while it was meant to be a negotiation document for equal partners to look at together [to determine] what makes sense to do and what not.
The challenge of shaping the relationship between the CSOs and the NMFA such that the NMFA's verifying and partnership roles do not clash is likely to be more prominent when strategic commitment and collaboration between these entities is more intense. However, given our findings on respect for CSO autonomy, such tensions are not likely to impede CSOs' advocacy work.
The force of managerialism stems partly from established organizational norms and practices within the NMFA, but it is also rooted in the political context in which the NMFA finds itself. Regardless of transformational ambitions, the NMFA faces political pressure to demonstrate results. Liaison persons are highly aware of this, and their concerns about effectiveness are partly explained by this reality, which is particularly pressing because of the largely intangible and uncertain nature of advocacy results .
Such dynamics are beyond the control of the Civil Society Division responsible for shaping and executing Dialogue and Dissent. In line with this, we did not find evidence of formal or informal strategies to counter the challenges identified above.

|
VAN WESSEL Et AL.

| Becoming "insiders"
The third issue identified in the literature is the institutionalization of CSOs as "insiders" in policy processes, to the detriment of their engagement with and representation of support bases and publics. Whereas the previous two issues were explicitly understood and addressed by the policy programme and the actors involved, this third issue was only partly addressed. Below, we discuss how this issue emerges in the structure of the policy, the limited ways in which the NMFA has addressed it, and challenges that we identify when it comes to addressing this issue further.

| Space for dissent
The risk of co-optation was addressed by formally granting and informally championing (internally) CSOs' freedom to disagree and target the Dutch government. This went so far that some MFA staff started to refer to one of the leading staff members of the Civil Society Division as "Mister Dissent," an identification this same staff member jokingly maintained was wrong, in a meeting with us. Rather, as he explained: "there can be no dialogue without dissent." In his framing, different viewpoints were a normal part of constructive interaction. Indeed, in their interactions with other ministerial staff, Civil Society Division staff members and some department liaison persons framed dissent as contributing to good policy and described the government of the Netherlands as unusual and commendable in "organizing its own opposition." Indeed, since the start of Dialogue and Dissent, several partners have campaigned against or otherwise opposed the Dutch government. Examples include a campaign where strategic partner Both ENDS addressed the activities of FMO, a Dutch public-private development bank (Both ENDS, 2018) and the opposition of strategic partner Centre for Research on Multinational Corporations (SOMO) to the Dutch Agreement on Sustainable Garments and Textile (SOMO, 2018). Although many partnerships work mainly through dialogue, space for dissent exists. However, some CSO interviewees sensed that the space for dialogue with the NMFA in the programme can push CSOs away from dissent. Constructive relations stimulated by the programme have their own role to play here. As one CSO interviewee stated: It's important to point out the great importance of personal relations within the NMFA for our advocacy work. It's essential to have good access points. A disadvantage can be that by these kinds of contacts you are forced always to go for dialogue, while sometimes it can be better not to do that. Sometimes you can achieve more by dissent and by focusing on the differences.
Such tensions may make it difficult for CSOs to combine insider and outsider strategies. This can be seen, for example, in multi-stakeholder processes involving the NMFA, the Dutch private sector and CSOs. In some cases, CSOs may simultaneously engage in dialogue and exert pressure, which may complicate relations. As a CSO interviewee said: You can tell it is a bit of a struggle, especially for the civil servants. There is this tendency that you can't sit at the table and protest at the gates of [companies] at the same time. That isn't fair. They find that unpredictable, or worse, unreliable.

| CSOs' engagement with publics
A second element of insider status that puts CSOs' advocacy role at risk is the government tendency to approach CSOs as "proxy publics" that can "stand in" for society (Lang, 2012, pp. 7-8). This can contribute to a situation where CSOs value their "insider" place at the table highly, with mobilizing public support potentially becoming redundant or even risky. An analysis of information including NMFA and CSO policy documents shed light on how this issue emerges in the programme. Indeed, the Dialogue and Dissent policy framework considers CSOs to be proxy publics: CSOs are the voice of citizens at local, national and international level. They can help make government more accountable to citizens and increase its legitimacy. In doing so, they contribute to greater social cohesion, stronger and more open democracies, a better response to environmental problems, a better business climate, more opportunities for all and less inequality. (Government of the Netherlands, 2014, p. 1) However, the policy framework also points out that an important part of civil society's legitimacy rests in its capacity to represent constituents. Within the Civil Society Division, there is awareness that assumptions of representation should be questioned. The Dialogue and Dissent ToC establishes preconditions for CSOs, which "need to be locally rooted, strong, legitimate and autonomous to perform advocacy roles," acknowledging that "this is however not always the case" and that donor support often inhibits CSOs' legitimacy, autonomy and embeddedness (Kamstra, 2017, p. 17). The policy framework also states that CSOs' applications would be assessed on their social support base and "inclusiveness." Here, the meanings of these notions are left open to interpretation. In the application assessment form (NMFA, n.d.), however, "societal support base" is defined as whether and how local partners were involved with preparing, planning and implementing interventions, and how the CSO accounts for its work to stakeholders, and "inclusiveness" is defined as the extent to which the applicants advance inclusiveness in their work through gender mainstreaming, efforts to empower women-focusing on improving opportunities and access to basic services among vulnerable and marginalized groups-and results realized in these areas. Thus, despite awareness within the NMFA that inclusiveness and societal support can be problematic, engagement with these issues is explicit but circumscribed; it is addressed in terms of engagement with partner organizations and with undefined "stakeholders." Furthermore, the dynamics between Dutch CSOs (or partner organizations) and their support bases were not incorporated into the policy framework, which does not question, assess or encourage CSOs' legitimacy as "the voice of citizens at local, national and international level" (Government of the Netherlands, 2014, p. 1). Therefore, whether and how CSOs actually mobilize, synthesize and amplify citizens' voices is not assessed or addressed. Rather, the programme appears to be rooted in the assumption that people's voices will be strengthened through conducting advocacy addressing inequalities, vulnerabilities and injustices faced by disadvantaged people in LICs and LMICs, and through building advocacy capacity in these countries. This may be plausible, but it is not a given. Previous research has shown that advocacy CSOs' communications with local people can be poor, and advocacy can harm the interests of people advocates claim to represent (see, for example, Seay, 2015). Communication and accountability between CSOs in the global North and South, and between CSOs and the people potentially affected by their work, are essential to ensure the strengthening of people's voices.
Nor does the policy framework address implications of donor relations for engagement with publics. Department liaison persons were often concerned about dissent, and both they and the 742 | VAN WESSEL Et AL. CSOs saw the most potential for the partnerships when there was alignment, creating possibilities to work as allies or at least engage primarily through dialogue rather than confrontation. This arguably discourages public engagement as a means of influencing, setting the stage for constructive co-operation between the NMFA and CSOs without needing to involve publics. However, having autonomy to choose their thematic focus, strategies (including dissent) and partners provides space for CSOs to engage publics to achieve their political goals. Internal, confidential documents and the practices of some CSOs in the programme reveal that some CSOs do engage publics as an important part of their strategizing. For example, Friends of the Earth Netherlands conducted a campaign to mobilize Dutch people to influence the government regarding the Comprehensive Economic and Trade Agreement in 2017-2018 (Milieudefensie, 2018). In cases of alignment, the NMFA and CSOs can also engage publics together, as happened with the early 2017 "She Decides" campaign to counter Donald Trump's global gag rule banning funding for international NGOs performing or providing information about abortion. This ongoing campaign, started by the NMFA in alliance with CSO strategic partner Rutgers, has mobilized states, organizations and citizens (She Decides, n.d.). Lang (2012) states that donor policy can encourage or require CSOs to engage with "publics," concluding that "CSOs tend to increase engagement with, and mobilization of, publics under two conditions: (1) when they face a critical juncture because access to political institutions is blocked or endangered by challenges to their legitimacy, and (2) when governments actively encourage, promote and fund CSO public outreach" (Lang, 2012, p. 209). Under the first condition, the focus on advancing inclusiveness, often in conditions of high inequality, may encourage public engagement. However, limited and shrinking civic space (CIVICUS, 2019) accompanied by risks to individuals and organizations, may stand in the way. Any policy would have to develop tailor-made approaches in light of such conditions. Under the second condition, policy encouragement might be worth exploring. Some CSOs funded under the Dialogue and Dissent programme conduct public outreach, which can lead to confrontation with the Netherlands government, other states or private-sector actors. Two important issues that Lang does not discuss are relevant here. First, CSOs have highly diverging organizational forms and do not easily fit into a predefined mould for public engagement. Policy, going beyond funding public engagement, has to address this. Much of CSOs' legitimacy may be based, for example, on their expert knowledge of life and risks under conditions of vulnerability, rather than the representation of constituents. For example, Partners for Resilience (PfR) is a Dialogue and Dissent partner centring its role on expert knowledge (Partners for Resilience, n.d.). Whereas expertise-based CSOs need to work on a basis of accountability to the people they say they work for, their sources of legitimacy differ from those of CSOs working on the basis of mobilized public opinion or a stable support base of private donors or members supporting an organization's humanist, faith-based or issue-based focus. A second issue is how encouragement to engage with publics can avoid harming CSOs' autonomy. Such encouragement would need to consider CSOs' specific roles and sources of legitimacy, and the associated forms of accountability and interaction-corresponding to CSOs' understanding of their own roles and identities. Questions of public engagement are important, but they have no straightforward or easily implementable answers.

| Could the issue be addressed?
A fundamental starting point of the programme-CSOs' legitimacy and roles resting in their capacity to represent-presents multiple challenges that are not adequately addressed in the programme, and there are no easy solutions, given the complexities involved.

| DISCUSSION
This article asked: (1) how and to what extent the Dialogue and Dissent programme confronted issues involved with advancing the advocacy role of CSOs through CSO-government collaboration; and (2) what can be learned from this case in terms of how donors and CSOs can work together to advance CSOs' advocacy role. We found that Dialogue and Dissent has the potential to advance CSOs' advocacy role in multiple ways. First, the programme builds in space for dissent, which the NMFA protects formally and informally. Second, the programme builds in space for flexibility and relative equality between the government and CSOs, allowing the adaptation of activities in changing conditions and the development of collaborative relations. Although these features positively address mutuality and institutional pressures, challenges remain. Some of these have been addressed and at least partly overcome, others much less so. Partly this can be explained by the fact that some of the challenges are rooted in wider organizational and political realities, going beyond management of the partnerships and the policy framework. These pressures stem from organizational practices of a managerial nature within the NMFA, external political pressures and conflicting government objectives. There are other challenges that need to be addressed further. This concerns the question of whether CSOs can in fact be understood as "the voice of the people," which is a basic premise of the programme. We also saw that this challenge is highly complicated.
These findings yield several important lessons with wider significance. First, mutuality in government-CSO relations can be advanced by building in formal conditions (flexibility in planning and reporting, and guaranteed space for dissent) and advancing these informally. Second, mutuality takes shape over time, as partners negotiate a range of dimensions, cognisant of the pressures surrounding alignment on topics, agendas and perspectives; capacity in terms of expertise and time; personal and organizational partnership ambitions; visions of contributions to be made; and the management of risk. These findings are in line with recent literature emphasizing organizations' navigation of CSO-government collaboration (Banks et al., 2015;Hermansen, 2017). Third, the policy framework and commitment of actors directly involved are important, but it remains unclear whether these can overcome challenges beyond the government agency responsible for the policy. Clearly, a policy programme, goodwill and efforts to implement it would under any circumstance be hard put to counter such issues effectively. It is, however, important to acknowledge these wider realities and engage with them in policy development and implementation. Fourth, public engagement deserves more attention as a crucial foundation for the advocacy role of CSOs.
More reflexively: questions of autonomy and mutuality re-emerge here in a different form, not discussed in existing research. Involved actors negotiate mutuality. To advance strategic involvement and commitment from the NMFA's side, staff members have been asked to promote synergy between the CSOs' work and their own. Creating such synergy involves finding ways for CSOs' power to be harnessed by the NMFA, in the context of its efforts to influence specific political and policy processes. However, the different roles and responsibilities of the NMFA and CSOs are often likely to require CSOs' adjustment to fit the NMFA's political goals and approaches. The interviews brought out clearly that this is not something CSOs want, and adjusting to the NMFA's objectives is seen as an affront to their understanding of their role.
Another issue involves partnership in the programme mostly developing in the presence of alignment between the NMFA and CSOs. Often, CSOs and the NMFA work together towards the same goals, largely with the NMFA as a supporting ally for the CSOs. Partnership rooted in alignment may reinforce views already present in the NMFA, advancing its existing development agenda. This may be advantageous for this agenda, as synergies can be achieved through aligned and complementary action. This may also strengthen the CSOs' voices on prominent concerns and interests shared with 744 | VAN WESSEL Et AL. the NMFA; CSOs can harness the power of the NMFA. However, although alignment-based partnership can advance mutuality, making partnership a funding requirement may stifle alternative and less-heard voices in the partnerships, despite efforts to maintain CSO autonomy.
Space for dissent is maintained, but many NMFA staff members saw CSOs' dissenting voices, confronting the Dutch state, other states, or private actors, as risky disturbances. Rather than embracing differences in viewpoint as productive, these staff members felt it necessary to navigate differences to prevent damage to the interests and complex diplomatic relations of the NMFA and the embassies. The notion of "partnership" has limited meaning here beyond the acceptance of an autonomous and legitimate role, some mutual co-ordination and perhaps better-than-usual access to policy circles.
Responding to the argument that "civil society may be best nurtured when donors do less" (Edwards, 2011as cited in Banks et al., 2015, our findings indicate that the Dialogue and Dissent programme can support CSOs' advocacy role by creating conditions for collaboration that formally and informally protect space for this role and enable the actors involved to identify and act on opportunities stemming from close collaboration. However, we also find that a programme seeking to advance civil society's autonomous advocacy role may best seek alignment where this is fruitful, without making close collaboration a requirement, because enabling CSOs' advocacy role is contingent on many factors that are only partly under the programme's control. Indeed, partnership in Dialogue and Dissent is contingent. Formally, working with the NMFA towards collective goals is a requirement, but in practice the programme facilitates rather than imposes close collaboration. The shorthand "strategic partnerships," commonly used to refer to the Dialogue and Dissent programme, does not cover the diversity of reality: there are many forms of relations, including the avoidance of relations. Banks et al.'s (2015) proposed distancing seems less necessary than the authors set out for a government donor to advance the advocacy role of civil society, but it is certainly useful as an option.