Between clients and bureaucrats: An ambivalent position of NGOs in the social inclusion agenda in Czech statutory cities

Abstract

This paper examines the landscape of policy work conducted by NGOs with respect to the social inclusion agenda. Based on a qualitative case study of integration policy in a Czech city, the paper focuses on the relations between collaborative and critical policy work of NGOs. In this case, while the collaborative position is mainly justified by apolitical expertise, long-term professional experience and compliance with official standards of social work, the latter calls upon community-based knowledge and political participation. We argue that despite indisputable long-term benefits of collaborative policy work it includes risks of paternalism, accountability deficit and exclusiveness. These risks become more significant with increasing shared understanding and mutual interdependence. In this situation there is a room for the episodic external critical capacity to challenge the governance structure and enforce the accountability of collaborative networks.

1 Introduction

The central theme we explore in this article is the idea that we can distinguish between two fundamental regimes of NGOs working in the social inclusion area. The first one, which is typical for institutionalized and mostly service-based organizations using neutral professional knowledge, which establishes their authority to be engaged in policy disputes. In relation to clients, their goal is to reach a state where clients will be able to live without any further assistance. The second regime is typical of activist groups using their expert knowledge to assist citizens in exercising their own agency in defending their interests in policy debates. Their ultimate goal is to increase citizens’ policy capacity and to mobilize them to defend their interest without any assistance. We hypothesize that the way different NGOs perceive their relationship to their constituencies determines the character of their policy work. Our main concern is with how these two regimes interface with each other, and with the policy work of bureaucrats and local citizens.

The mushrooming of informal groups emphasizing the advocate's role and mobilizing citizens can be considered as a new phenomenon in the post-socialist context which was previously described predominantly in terms of expert driven ‘transactional activism’ (Císar, 2013; Petrova & Tarrow, 2007). In transactional activism, NGOs’ legitimacy does not result from active membership but from epistemic sources — the scientific knowledge, expertise and experience they produce and use in public discourse (O’Neill, 2001, according to Buth, 2011). Transactional NGOs have been also heavily involved in processes of collaborative planning with public authorities. In contrast to transactional activism, newly emerging activists groups are formed by people with diverse backgrounds and derive their legitimacy directly from citizens’ mobilization. They are also more critical towards politicians and policy bureaucrats and they can be characterized in terms of social movements.

The case study from which we draw these points involves a local controversy which erupted in 2012 between two NGOs representing these different styles of policy work. Their “public war” arose from a housing crisis in one of the largest impoverished Czech areas — Predlice in Usti nad Labem. Studying this controversial case provides us with different accounts of how NGO policy workers see their jobs, construct their policies and engage with stakeholders. Particularly, we intend to describe how NGOs might, at this level, be considered to have expertise and how their expertise is constructed. In addition, we discuss the differences between the character of policy work employed in the formal and informal processes of influencing public policies and the differences between strategies and accounts used in different regimes of NGOs’ policy work, focusing on how they justify their policy positions, choose their policy strategies and exert their policy influence. We are interested in how policy workers, in our case those from NGOs, get a place at the table, how the question is framed, what discourse is accepted as valid, and how this work relates to the outcome at any point in time. Generally speaking, we are interested how collaborative venues for policy work might be related to alternatives styles of NGOs’ policy work such as informal protests and what is the relation between collaborative and critical capacity of NGOs. Especially the role of shared understanding and constructing of target population is taken in consideration.

2 The policy work of NGOs

Drawing on the narrative of “governance” (Offe, 2008; Rhodes, 1996), we consider policy as an ongoing dynamic and interactive process between actors — governmental policy workers, non-governmental actors, or business associations — with diverging interests in formulating, promoting, and achieving common goals by mobilizing, exchanging, and deploying a range of ideas, rules, and resources (Torfing, Peters, Pierre, & Sorenson, 2012: 3). There is obviously no single account of policy work. On the contrary, there are many different activities and ways to make sense of such activities that can be understood as policy work. Policy process participants are supposed to have overlapping agendas, different interpretations of the problem and different levels of concern about its resolution (Colebatch, Hoppe, & Noordegraaf, 2010: 228).

Keen (2006) lists activities which can be associated with NGO policy work: (1) monitoring government; (2) researching issues relevant to a policy field; (3) making pre-budget submissions; (4) preparing submissions to government inquiries; (5) liaising with political representatives and their staff; (6) being available for consultation; (7) sitting on government committees and task forces; (8) influencing legislation; (9) networking with relevant policy communities; (10) disseminating information; (11) ensuring visibility, credibility and legitimacy of the organization; (12) encouraging public debate.

In the modern state, crucial sources of policy-making innovations come from professional communities as key fora in developing and testing knowledge, setting standards and steering behaviour (Dunleavy & O’Leary, 1987: 302). Over the last two decades, the expansion of new forms of engagement between state and citizen might be viewed as evidence of a new form of collaborative policymaking (Innes & Booher, 2003); collaborative governance (Ansell & Gash, 2008; Donahue & Zeckhauser, 2011; O’Flynn & Wanna, 2008); or interactive governance (Torfing et al., 2012). Collaborative models of policy work deal both with diversity and interdependence because of needs to be inclusive and to explore interdependence in the search of solutions (Innes & Booher, 2003: 54). In their model of collaborative governance, Ansell and Gash (2008) distinguish: (1) starting conditions (including resource asymmetries, incentives for participation and history of cooperation); (2) collaborative process (face-to-face dialogue, trust building, commitment to process, shared understanding, intermediate outcomes); (3) institutional design (participatory inclusiveness, forum exclusiveness, clear ground rules, process transparency); (4) facilitative leadership and (5) outcomes. Public authorities’ interest in initiating collaboration with NGOs can be summarized in three propositions: collaboration can encourage trust, collaboration can unlock distinct competencies of other sectors, and collaboration can deliver a transformative approach to service improvement (Entwistle & Martin, 2005). Sirianni (2009) presents eight core principles of collaborative governance: (1) coproduction public goods; (2) mobilization of community assets; (3) sharing professional experience; (4) enabling public deliberation; (5) promoting sustainable partnership; (6) building fields and governance networks strategically; (7) transforming institutional culture; (8) ensuring reciprocal accountability.

In the area of social exclusion, Warin (2002) describes how formalized consultative structure and formal subcontracting resulted in more intensive involvement of NGOs in France in the processes of both policy-making and policy implementation as experts. NGOs tend to promote themselves as mediators to restore the broken links between an excluded group and public authorities in the field of social inclusion. Being involved in such collaboration networks also might contribute to public recognition of social workers’ expertise (Macdonald, 1995).

Alongside working as complements to government in a partnership relationship, Young (2000) distinguishes two other views on the government–NGO relationship: (1) nonprofits operate independently as supplements to government or (2) they are engaged in an adversarial relationship of mutual accountability with government. The NGO–state relation may be determined by: (1) structural context (Clark, 1991; Coston, 1998; Gidron, Kramer, & Salamon, 1992); and (2) strategic interests of both governments and NGOs in relation to goals and strategies of action (Najam, 2000). NGOs and government may seek similar goals with similar strategies (cooperation); dissimilar goals with dissimilar strategies (confrontation); similar goals but preferring dissimilar strategies (complementarity); or may prefer similar strategies but for dissimilar goals (co-optation). Ramanath and Ebrahim (2010) point out that a particular NGO may not limit itself to a single strategy and may employ a more complex and nuanced repertoire of strategies and tactics.

In contrast to collaboration, adversarial critical policy work explicates a dominant meaning in policy content and process, uncovers suppressed or marginalized meanings, identifies “agents of impairment” (Lindblom, 1990) who suppress alternative meanings, evaluates institutions in terms of communicative standards and criticizes expert policy analysis (Dryzek, 2006). According to Minkoff (1997), a critical role in civil society means the production of social capital by providing an infrastructure for collective action, facilitating the development of mediated collective identities that link otherwise marginalized members of society, and shaping public discourse and debate. Critical policy work can be labelled as “contentious politics” that often is the main recourse that marginalized people possess against opponents (Tarrow, 2011).

Several types of failure go hand in hand with a particular type of relation between NGOs and public authorities. Salamon (1987) identified four voluntary failures: (1) philanthropic insufficiency, rooted in NGOs’ limited scale and resources; (2) philanthropic amateurism; (3) philanthropic particularism, reflecting NGOs’ choice of clientele and projects; and (4) philanthropic paternalism, where those who control the most resources are able to control community priorities. Whereas the first two failures can be associated with critical policy work, the later pair with collaborative one.

3 The history of public controversy in Usti nad Labem

Through the discussed theoretical lens, we look at the practice of several NGOs working in the field of social inclusion in the local context. We focus on their activities and self-presentation in terms of positions and policies. We analyze mainly the media coverage they initiate, the expert studies they produce, and official statements of municipality representatives.

The social inclusion agenda that is discussed in this paper represents a field where different levels of government — from the EU to municipal levels — interact with non-governmental actors. We understand the social inclusion agenda as an effort to provide all citizens with equal access to education, housing, health care, employment, social services and safety. In the Czech national context, the social inclusion policies are epitomized with policies seeking to tackle the so-called Roma issue. The biggest responsibility in social inclusion agendas lay with local authorities.¹ Municipalities have independent responsibilities in local housing policy, they also govern the process of social services planning and they have multiple mechanisms to support local NGOs working in the field of social inclusion. We believe that the social inclusion agenda, as a multidimensional and multifactor field, could provide representative material for the analysis of the possible association between expertise, the NGO–government relationship, strategies of NGO policy work and possible failures.

Our case study is focused on Usti nad Labem, a regional centre in Bohemia with almost 94,000 inhabitants; more than 10% are impoverished Roma. One can find a couple of socially excluded industrial areas in the city. Predlice is the largest. It is occupied by roughly 3000 socially excluded Roma who compose approximately 90% of the total population of this suburb (GAC, 2009: 16). In the late 1990s, the industrial city became well known because of the local government's efforts to build a wall and separate the majority from the Roma. It raised an immense wave of criticism on both the national and international levels. Even though the wall was demolished, the Roma issue remained a tense political issue.

In September 2012, in Predlice, a young woman died when a ceiling collapsed in one of the many dilapidated houses in the suburb. The situation was the result of a long-term process that had been highly influenced by the wave of privatizations taking place after the 1989 revolution. The former city management sold some of the houses despite NGOs’ warnings to entrepreneurs and other individuals whose primary intentions were to make money from poor tenants. The legal duty of the owners to ensure the satisfactory condition of the buildings was violated often. According to the local NGOs, the city administration's privatization decision led to increased crime and debt, forced prostitution, and a prevalence of drugs. After the young woman's death, the city became more active in construction inspections and ordered the evacuation houses with the highest risk of collapse. Municipal government convened a series of meetings between town representatives, the Municipal Police, the town's advisers, experts, and NGOs.

There are several NGOs in Usti nad Labem whose primary focus is on social inclusion. Since 1999, there has been a branch of one specific NGO, “People In Need” (PIN) operating there with more than 20 employees and approximately 30 volunteers. PIN is a well-established nationwide NGO with around 60 branches across the Czech Republic and Slovakia. It has pioneered professional standards of social work for almost two decades, and has been very active in policy advocacy on the national level. Alongside social work, PIN has been conducting expert studies and drafting evidence-based briefing comments for governmental bodies on both the national and local level. PIN forms a coalition with several others NGOs in Usti nad Labem. Some of these are also nationwide NGOs; some operate only on the local level. All these organizations offer professional and accredited social services, employ professional staff, usually with a university degree in social work, and are funded by national and local authorities. All of them, more-or-less, formed a policy coalition taking a part in local processes of policy making and policy implementation in order to represent their interests and those of their clients.

In the period of 2009–2013, there were several processes through which some of these NGOs participated in municipal policy planning and implementation. Between November 2009 and May 2010, the new Community Plan of Social Services was being prepared. Community planning of social services is based on the participation of local users, providers and municipal authority or regional office. The process of community planning at the level of municipalities is a precondition of access to most state funding. The plan is developed over a period of several years and, finally, adopted by the city council. Several working groups contribute to the planning process. The group targeting socially excluded minorities’ issues was led by the director of the local branch of PIN and more than 10 other NGOs took part in. Even though community planning is supposed to engage citizens in the planning process, citizens are represented only indirectly, through NGOs. The involved organizations, based upon their professional expertise, interpret and frame the experience of their clients.

The community planning group addressing the issue of socially excluded minorities mentioned the expert knowledge and policy capacity of NGOs as among the strengths of the city's policy; on the contrary, the NGOs’ vulnerability was associated with their clients’ conflicts with the city. The involved NGOs were well aware that policy work in the social inclusion area is a matter of constant tiptoeing between the interests of their clients and local government. Community planning allows them to push changes in funding priorities as well as other items on their own agenda, as was in the case of PIN. The planning group also agreed to ask the city to develop the Strategy in co-ordination with the governmental Agency for Social Inclusion (ASI) involved NGOs and relevant public authorities.

According to the community planning process outcomes, the strategy was intended to be developed by the Mayor's Action Group responsible for the planning and implementation of inclusion policies in 2008. Its aim was to tackle burning problems and coordinate the activities of stakeholders. However, the inhabitants of socially excluded localities once again did not participate in this group directly. The meetings took place behind closed doors. The Action Group played an active role in dealing with the Predlice crisis when it was transformed into a “Crisis Action Group”. According to the meetings’ minutes,² the role of PIN in the crisis was mainly to assist the families at risk to find flats that they can use. PIN shared its responsibility with the City Coordinator for Ethnic Minorities. In fact, it is the city's legal obligation to assist families facing an inadequate housing situation to find an acceptable and satisfactory solution. In this case, the fulfilment of this duty was partly transferred to the NGO which supplants the function of the city. For several reasons, however, the efforts of the coordinator and PIN failed in the end and this failure became one of the most burning issues of the whole Predlice crisis.

During the housing crisis in Predlice in 2012, a new NGO — Konexe — stepped into the public arena. The Konexe is a new group formed by local activists including inhabitants of the Roma ghetto but also activists from other parts of the country. Unlike PIN, Konexe did not take part in any of these planning processes (it was established only in 2012) and it heavily criticized the local authorities, especially PIN, for not fulfilling their responsibilities. Konexe was closely connected with two informal civic initiatives, “Hatred is Not a Solution” and “Housing for All” (HFA). Its leader, Miroslav Brož, was also the spokesman of the former initiative. He came from Usti nad Labem and used to be an employee of PIN for several years and one other locally based NGO. During the Predlice crisis, Konexe cooperating with HFA and individual activists, organized blockades and demonstrations, spoke for the Predlice residents, provided these people with consultations and supported them by its own presence. They also made the case publicly visible through many press releases, video reports and articles in which local Roma people got a voice.

4 The social construction of target population and the role of expert knowledge in debating social inclusion

The overall approach of PIN towards social inclusion is rooted in Oscar Lewis's concept of “culture of poverty”, which stresses that long-lasting poverty leads to the acceptance of a specific culture that serves as an adaptive function. Culture of poverty represents an effort to cope with feelings of hopelessness and despair that arise from the improbability of achieving success in terms of the prevailing values and goals (Lewis, 1966: 21). PIN generally rejects the ethnic basis of social inclusion policies — government policy should be primarily aimed at vulnerable, excluded and marginalized people generally, and not specifically on the Roma. PIN recommended an ethnically blind social inclusion policy to the city of Usti and Labem in several documents (PIN, 2006, 2007). The solution of social exclusion is seen primarily in the focus on a social work practice centred on individuals and individual families. PIN argues that it is often not possible to understand Roma as a community because they, in fact, form no community (Moravec, 2005). For this reason community approach cannot be applied in Predlice:

“These days, living in Predlice is rather a temporary crisis solution. People move in and out depending on the lowest prices for housing. That is why I am not able to think about Predlice as a community.” (Jan Cerný (PIN) in Rychetský, Slacálek, A2, 2013)

PIN emphasized that it is, first of all, a provider of social services, urging its clients to take an active approach towards solving their problems. The ultimate goal on the local level is to empower families to be able to handle future problems without any further assistance. On the national level, PIN tries to pursue some systemic changes (with the same ultimate goal) by sitting on several government expert committees, providing comments on many relevant bills and conducting various expert studies, mainly sociological, anthropological, legal and demographical ones. However, even in this context, it has always stressed the apolitical and expert basis of its attitudes. “We are not a political organization”, argues Jan Cerný, PIN's director of social integration programmes, in a media interview. PIN acts mostly as an expert advisor, not policy advocate.

The opponents from Konexe and related citizens’ initiatives argued that there is no such thing as an apolitical approach — this was instead PIN's attempt to hide its ideology and political preferences behind the discourse of objectivity, expertise and professionalism, define norms and strengthen its powerful position. Konexe rejected the solutions based on expert knowledge and service provision.

“The most essential thing is that the problems of impoverished Romani communities will never be resolved by college-educated gadje³ providing them with social services.” (Brož, M. Open letter to PIN, 2012)

Konexe's conception of social inclusion is completely different. The system (composed of the state, local authorities, established NGOs) is responsible for the current conditions experienced by the poor and it is the system which must be changed, not the poor themselves. It considers the roots of social exclusion to be structural and significantly influenced by the racism of the Czech majority. It blamed the current social inclusion policy (including social services provision) to be extremely paternalistic. It called for the political mobilization of Roma people (“activate their own internal resources and the potential of their community members”) as a part of a “civil society of the socially excluded” (Brož, M. Open Letter to PIN, 2012). Konexe and HFA urged the existing NGOs to encourage impoverished Roma to make their own decisions, to advocate for their rights and to be “the conscience of those in power” (HFA, 2013). Konexe argued for a social policy that would lead to political emancipation of the vulnerable Roma minority. Konexe considered the poor to be their partners, not clients, and adopted this premise into everyday practice:

“Large NGOs consider these people to be their clients. We consider them as equals. We live with them, sleep, cook and organize meetings at which they can express themselves. We respect their will.” (Brož, M. in Angermannová, A. MF Dnes, 2013).

Using Schneider and Ingram's (1997) theory of policy design, PIN constructs its target group as a group whose behaviour the policy seeks to change, by portraying the socially excluded as mainly dependent but not entirely blameless. As part of its defence, PIN blamed the residents for inaction in solving their housing problems. Poverty, especially in the local context, is depicted as an individual problem and individuals or families are seen as objects whose behaviour is to be changed. In contrast, Konexe sees as its main target group the local government and other NGOs. Not the clients, but rather the entire political approach and the process of policy making must be changed. Whereas PIN speaks about dependants as its main target population, Konexe challenges the advantaged, including local government and local NGOs as a main source of the problem. In the case of both NGOs, their social construction of the target group determined the methods of their policy work. Whereas the former promoted its interest in dialogue with the local government, the latter went for an overt conflict with the local governance structure at large. While the latter blamed the former for paternalism and camaraderie with the municipal government, the former accused the latter of amateurism and seeking for media awareness.

5 Revolving doors of local collaboration

Participation in planning groups — the Community Plan of Social Services, the Mayor's Action Group, the Local Partnership — formed a stable learning community where NGOs interact with local government and other public authorities. In terms of Ansell and Gash's model (2008), local NGOs had sufficient organizational structure and skills and knowledge to participate in such collaborative community. A history of successful past cooperation created social capital and high levels of trust producing climate for a long-term collaborative structure.

There was an apparent personal link between NGOs and government in both local and central level social exclusion agenda. The current governmental Agency for Social Inclusion director used to work with PIN and the deputy mayor of Usti nad Labem was responsible for social affairs and a Member of Parliament, and used to work at the local office of PIN. This revolving door between NGOs and local governments fosters the building of trust and the development of shared responsibilities and policy accounts and meanings of problems to be solved. This mechanism is reinforced by the existence of learning communities where policy bureaucrats meet with NGO representatives to discuss matters of mutual concern. Indeed, shared meaning and responsibilities smoothens transition from one sphere to another. On the other hand, it might give rise to critical reactions. The close alliance between city government representatives and PIN was criticized in the Predlice controversy.

In Young's (2000) sense, PIN works along with other NGOs as a complement to government in a partnership relationship. The resulting type of policy work is characterized by the development of horizontal ties with other NGOs and vertical ties between NGOs and public officials. The negotiations between local authorities and NGOs take place within an institutionalized framework of knowledge sharing and joint effort to define and solve emerging problems and challenges. The NGO's representatives appreciated the cooperation with the local government because it made the work of bureaucrats more effective, improved the situation in impoverished suburbs and also changed bureaucrats’ understanding of the source of the problems. In a world where a variety of policy accounts co-exist in the public arena, participation in the policy-making process allows NGOs to influence the discourses of policy makers through mutual learning. On the other hand, participation in learning groups like these has improved the NGOs’ understanding of the work of public bureaucrats, including the limits of their competencies and the institutional setting of public authorities. It also helped the NGOs disseminate their understanding of social problems and the conditions their clients live in. Mutual learning is very often reinforced by the fact that representatives of NGOs share the same educational background in social work with public bureaucrats responsible for the social inclusion agenda. Even negotiation about the form of a contract between a city and an NGO should be understood as another opportunity to influence policy when the NGO explains to municipalities what it is not willing to do for them and why.

On the contrary, Konexe believes that PIN's financial dependence on the municipality and long-lasting close cooperation with politicians and bureaucrats prevent PIN from taking a more critical view towards the city administration and its failures. However, PIN claims this is not the case. Jan Cerný explained PIN's decision not to criticize the municipality publicly as a matter of a pragmatic strategy. He argues their main strategy in Usti nad Labem is to cooperate and negotiate with the local government about particular solutions to existing problems and to provide it with critical feedback based upon their research studies and professional experience. They prefer fostering critical dialogue behind the closed doors of policy networks over public fighting reported in the media and discussions on the on the level of actual cases than the level of abstract principles.

NGOs involved in collaborative policy work prefer an incremental policy of small steps, even though this limited their advocacy efforts. Such policy work can be possible only when policy workers from both the NGOs and the local government share a similar construction of the target group and are willing to learn from each other. It is inevitably associated with the emergence of social service contracting and horizontal sharing of responsibilities across different spheres. It is based upon the existence and official recognition of NGOs’ experience in mediating the needs and interests of their clients. On the other hand, it poses the threat of excluding citizens and causing the NGOs to be identified too strongly with the local government representatives. The stronger policy capacity they have, the more limited their critical capacity is.

6 Local alternatives to the dominant individualistic social inclusion discourse

In Usti nad Labem, the collaborative forum is “the dominant game in town” (Ansell and Gash, 2008), and excluded stakeholders seek alternative venues. For Konexe and activists from HFA, relations with the media became the most important tool for promoting their ideas and interests. They prepared press releases; were interviewed by all leading Czech newspapers; and closely cooperated with Romea.cz, the most popular Romani news server, with which they shared an emphasis on the ethnic dimension. Romea.cz published all their press releases and several articles written by the leader of Konexe himself. Instead of being part of an institutionalized network, Konexe cultivated a network of people and organizations pushing alternatives to the dominant individualistic and ethnically blind social inclusion discourse. Konexe also tried to exert pressure on the local government by extending the conflict to the national level. To raise nationwide public awareness, it also organized a blockade of the Ministry of Labour and Social Affairs in Prague. Konexe asked the other actors, mainly the municipality and “big NGOs” generally, including PIN, to be more accountable. It openly criticized them for their failure and accused PIN of shifting their priorities from helping the poor to maintaining their position of power and funding. Konexe further criticized a democratic deficit of the governance structure, i.e. the fact that the excluded Roma inhabitants of Predlice were not involved properly in the process of solving the crisis.

Konexe argued that the residents themselves know the situation best and they have the right to participate in a solution. Konexe and HFA often gave voice to Predlice inhabitants in their articles. They also made several short video reports in which Predlice Roma spoke about their experiences. The style of reporting advocated on behalf of Roma, rather than questioning their arguments. Konexe also prepared letters and statements together with the Predlice residents that were broadly publicized. It ruled out any kind of mediation process to bridge differences in favour of direct involvement of affected citizens. It did not claim any boundary between the organization and Predlice's inhabitants; it did not mediate their experiences because they were part of the organization. However, it was always the leader of Konexe or other not socially excluded member of Konexe who took the lead in all Konexe's activities. Calling the city to account, Konexe asked for such a solution of the crisis leading to permanent financially sustainable housing for the entire group of the people living in the evacuated house. Last but not least, in a declaration that was read by activists with faces covered by scarves after the last demonstration in front of the residential hotel where the evacuated lived, HFA refused making the poor responsible for their current situation and charged the system with failure and incompetence in a very general way.

They stressed that in Usti nad Labem we are not dealing with a local issue but a much more general problem:

“We will be passing on what we learned in Usti and Labem. We will turn up again in places where the powerful least expect us and where people at the bottom of the barrel want to fight for their rights to a dignified life, for the rights of their children, for housing, and against racism.” (HFA, 2013)

PIN, as a part of collaborative governance structure, opposed Konexe's approach as unrealistic and counterproductive. They accused Konexe in the media of (un)intended manipulation with the impoverished residents and questioned the absence of boundaries between the organization's leaders and local inhabitants. However, PIN's general media strategy was strongly defensive and less intensive. Throughout the controversy, it mainly advocated its positions and explained its own activities. PIN also stressed that the situation was far more complex than Konexe described.

Whereas PIN offered an expert solution, Konexe argued for a deliberative one. The citizen initiative pointed out the limits of current institutional framework and criticized its procedures for neglecting Romani inhabitants and exclusiveness of local collaborative fora. Its policy work was not outcome-oriented but rather tackled procedures and pointed out political participation of Roma population: for better outcomes, the social exclusion agenda needs to be democratized and directly involve citizens. The initiative called for increased political participation of Roma, while putting all the responsibility for their current situation on the side of the majority population, those in power. Unlike expert NGOs, the critical policy argumentation was predominantly moral, not rational. While experts spoke about the effectiveness of their policy work, activists spoke about fairness and morality. As in collaborative policy work, we can also speak of a learning community with regard to critical policy work. In this case, however, the learning community was formed by local activists and residents with the support of a non-local central based activist. The goal of this activity was to nationalize the local agenda and put pressure on the local government and local NGOs.

7 Collaborative and critical policy work

PIN and Konexe represent two different conceptions not only in the social inclusion agenda but also in NGOs’ policy work in general, i.e. their role in public policy and cooperation with the municipality/state and the target population of the social inclusion policy. Whereas PIN works more-or-less as a partner with and complement to government, Konexe operates in an adversarial relationship. The former represents a service-providing organization deeply involved in so-called collaborative policy work based on interactions with local authorities and NGOs through negotiations in a relatively institutionalized framework of knowledge sharing and joint definition and solution of emerging problems and challenges. The latter can be associated with critical policy work which challenges a dominant meaning in policy content and process, uncovers suppressed or marginalized identifies. The characteristics of both kinds of policy work can be seen in the table below:

Table

	Collaborative policy work	Critical policy work
Mode of policy work	Participation in governance structure — planning groups, round tables, etc.	Mobilizing the public — media, protest campaigns, citizens’ meetings, watchdog activities
Mode of policy change	Incremental	Disruptive
Type of policy workers	Expert advisors, honest brokers	Policy advocates
Policy-relevant knowledge	Expertise, professional experience	Community-based
Mode of dialogue with local governments	Institutionalized networks	Public criticism
Relation to government	Complementary	Adversarial
Goals and strategies	Cooperation, complementarity	Confrontation
Risk of failure	Paternalism, client policy-making	Amateurism, unrealistic goals

Collaborative NGO policy work corresponds with the trend of governance policy networks and it can be associated with local of transactional activism on the national level (see Císar, 2013; Petrova & Tarrow, 2007). In these policy networks, NGOs can use their broker expertise to mediate between communities and local governments and in turn get credibility for both their clients and public bureaucrats. They participate in institutional platforms such as planning groups and round tables, sharing responsibilities with public bureaucrats. Through those platforms, they can promote incremental changes in favour of their goals, push forward their policies and influence public bureaucrats’ understandings and frames as they may become more responsive to their suggestions. Their activities are sustainable in the long-term. Shared meanings in those communities stem from similar constructions of the policy's target groups, similar education backgrounds, and the revolving door mechanism where professionals migrate between NGOs and public offices. In the local context, all those structures are strongly shaped by the public participation discourse of New Public Management on the national level. For example, in community planning, local governments implement guidelines developed by the Ministry of Labour and Social Affairs according to the best practices of other European states (mainly the UK or the Netherlands).

In terms of NGO–local government relations, one can see complementarity or cooperation in Najeb's sense as the predominant strategies. Collaborative elements of policy work such as consultation and networking prevail. As policy workers, NGOs in this regime act mostly as expert advisers (Hoppe & Jeliazkova, 2006) or honest brokers (Pielke, 2007) and they seek for consensus across different spheres of governance. Due to their involvement in the governance structure and high level of professionalization, there is a risk of philanthropic particularism (when focusing on goals in relation to funding possibilities), and exclusive client politics (when neglecting the needs and attitudes of their clients as a form of paternalism).

NGOs operating in the regime of critical policy work question the possibility of apolitical expert advice, considering it just another form of ideology. They prefer the emotional language of personal stories and video footage over rational expert reports. Unlike the outcomes-oriented policy work of the previous regime, they challenge procedural issues in favour of active engagement of citizens. They organize public meetings, mobilize public through media and organize blockades and protests. They disseminate information through the network of their followers and like-minded activists; at the same time, they are able to broaden the scope of conflict to the national level to increase societal pressure on the local government. Their activities have inevitably an episodic character.

In terms of NGO–local government relations, confrontation is the main strategy to encourage public debate. As policy workers, NGOs in this regime act mostly as policy advocates (Hoppe & Jeliazkova, 2006). They tend to polarize public debates at the expense of consensus building. Due to their fluidity and informality, there is a risk of philanthropic insufficiency (due to their limited scale and resources), and philanthropic amateurism. As their goals are very often found unrealistic by other political actors, allies from local government and wider public are difficult to find. On the other hand, their media work can put established networks under pressure and prompt them to adjust their goals and change their discourses in order to avoid a negative public campaign.

8 Conclusion

In the local context of social exclusion policies, we identified two functions of NGOs’ policy work: (1) to be an honest broker between public bureaucrats and their clients; and (2) to be a policy advocate. While the former justifies NGOs’ positions mainly by apolitical expertise, long-term professional experience and compliance with official standards of social work, the latter calls upon community-based knowledge and political participation. Based on the different priorities and goals, they choose different strategies to exert their policy influence — long-term participation in governance structures based on the expert knowledge on one side, and confrontation, media work and political empowerment of the excluded minority on the other.

In the context of New Public Management which stressed contracting social services through NGOs and sharing competencies between NGOs and local governments, clear boundaries of responsibility might be lost. Closer cooperation between NGOs and governments enabled NGOs to enhance their financial sustainability and policy capacity and create conditions for incremental change. However, this constellation is also prone to an accountability deficit (Bovens, 2007). Experts from NGOs that were supposed to question governance structures and possibly criticize them publicly for failures became inseparable parts thereof. Even if the form of the relationship between NGOs and government refers to partnership, the positions of the partners remain highly unequal. NGOs involved in collaborative governance structures are, inevitably, losing part of their critical capacity, as they consider public criticism incompatible with their position of partners and expert advisors. They keep any form of criticism behind the closed doors of advisory bodies and planning committees. Consequently, the government may still receive their feedback; however, the public and the public discussion miss out on any solid explanation and justification of the conduct of governance, as they are not publicly forced by NGOs. We argue that despite indisputable long-term benefits of collaborative policy work it includes risks of paternalism, accountability deficit and exclusiveness. These risks become more significant with the increased shared understanding and mutual interdependence.

The role of public advocates and accountability enforcers is overtaken by informal citizen initiatives and emerging grassroots NGOs. Their criticism is aimed not only at the old government structures but rather at the governance hierarchies and networks as a whole. Their rather episodic activities, limited scope of responsibility and professional standards, and high alarming potential enable them to bring the otherwise excluded issues back onto the agenda. Within the short-term period spent on the case discussed in this article, we could already see some effects the policy advocates had over the public debate and the work of established NGOs. However, the long-lasting and supra-regional effect of these local episodes can only be determined through further study.

Acknowledgement

The work on this article was done under the research project supported by the Czech Science Foundation ‘Social Dynamics Acceleration: Uncertainty, Hierarchical/Flat Governance Structures, and History Interiorization’ (grant no. P404/11/2098).

Notes

1 Generally speaking, Swianiewicz (2013) includes the Czech Republic among relatively decentralized CEE states alongside with Estonia and Latvia. The common feature in this category is the combination of a wide scope of functions provided by local governments with a negligible role of locally controlled taxes.

2 The meeting minutes were not published, however, the City Coordinator for Ethnic Minorities provided us with them on the basis of our demand.

3 In Romani, gadje is the expression for non-Romani person.