The Montserrat’s neighbourhood dream: involving Moroccan residents in a school-based community development process in urban Spain

ABSTRACT Previous research has identified the potential of community-based development processes to reverse inequalities. However, in many cases, constraints arise that hinder the achievement of community goals. This article focuses on the role contributed by a school to overcome these barriers. Specifically, we address the role played by the Mare de Déu de Montserrat School, which is a Learning Community, in the launch and development of a community-based process. Known as The Dream, this process was initiated by a group of unemployed Moroccan residents in an impoverished urban neighbourhood in Spain. Through this process, the neighbours have created a self-managed community garden, which has enabled them to improve their job training and build social networks. Based on the communicative methodology, our qualitative study finds that this school has contributed to the creation of specific conditions that helped to overcome some barriers that had hindered the emergence of community responses. In addition, the contributions of the school have promoted a horizontal and dialogical organization of the process, placing neighbours in central positions in decision-making and leadership roles. This case study provides relevant theoretical and practical implications for the contributions that certain school-based interventions can provide to promote community initiatives in impoverished environments.


Introduction
In this paper, we delve into the potential of a school-based intervention to promote a community-led development process for helping Moroccans face the consequences of long-term unemployment in a locally marginalized and disadvantaged context in urban Spain. Community-based development processes are characterized by a bottom-up approach in which the beneficiaries are actively included in the design and management of projects (Mansuri & Rao, 2004) with the aim of finding solutions to their own problems and necessities. Nonetheless, it has been argued that these processes may encounter barriers that are difficult for the vulnerable collectives involved to surmount (Classen et al., 2008), which in turn thwarts the possibility of success in achieving the community's goals.
The article addresses these challenges. It reports the findings of a longitudinal case study that examines the role of the Mare de Déu de Montserrat Learning Community (LC), a primary school in a disadvantaged neighbourhood in northeastern Spain. LC is an educational project that has demonstrated the capacity to promote the participation of traditionally excluded community members (e.g., minority groups in different European and Latin American countries) and to generate positive social and urban dynamics in impoverished neighbourhoods (García-Carrión, Molina-Luque, & Molina-Roldán, 2018;García Yeste, Lastikka & Petreñas, 2013).
We highlight the role adopted by the neighbourhood's school in the community process known as The Dream, resulting in the creation of a self-managed organic community garden. This initiative was developed by a group of Moroccan unemployed people living in this marginalized neighbourhood. Specifically, we answer the following research questions: (1) What role has the Mare de Déu de Montserrat school played in the launch and development of The Dream?, and (2) How this school has facilitated an horizontal and dialogic organization of the community process?
The Dream is an evidence-based social intervention model (Munté Pascual & De Vicente Zueras, 2012) grounded on metaphors provided by relevant historical figures, such as Martin Luther King (Logsdon & Murrell, 2008) and Paulo Freire (1997), who believed that 'to dream' is a collective starting point on the route towards social transformation. Unlike traditional community development plans in which a social technician leads the intervention as an outside expert on the existing social reality, in The Dream process, the leading role in the development of the community action lies with the residents affected by the social problems. The neighbours themselves are the ones who activate the community initiative and make decisions based on their priorities through a process of open dialogue with professionalsresearchers, policy makers, professionals of social action, administrations, schools and teachers, among othersas they adopt the role of community advisors (Munté Pascual & De Vicente Zueras, 2012). Aligned with the asset-based approach (Mathie & Cunningham, 2003), The Dream process has allowed neighbours to mobilize their assets and empowers them to lead the project of building the community garden. The methodology applied throughout the process focuses on creating an enabling environment in which to recognize their strengths and find solutions in ways that match their needs (Mathie & Cunningham, 2003).
Our research highlights two main aspects that have been conducive for setting up The Dream in this case study. Specifically, we focus on the role adopted by the neighbourhood's school in (1) the launch and development of this community initiative and (2) the promotion of a horizontal and dialogic organization of the process. To this end, in what follows, we first discuss the main theoretical concepts that guided our analysis. Second, we describe the methodological design of the case study carried out based on the communicative methodology (CM) and qualitative data collection techniques. Third, we present the results obtained, which reveal the significant role played by the school in the launch, development and organization of the community process.

Theoretical framework
The idea that local communities and individuals have been actively involved in developmental processes has been a recurrent one throughout the previous and the current century (Classen et al., 2008;Mansuri & Rao, 2004). Community development approaches have been broadly implemented to address different social challenges around the globe. The literature has provided evidence of the impact that community development processes can have in improving a variety of disadvantaged situations, such as income improvement (Pinho, Orlove, & Lubell, 2012), women's empowerment in managing natural resources (Flecha, Soler-Gallart, & Sordé, 2015;Wrigley-Asante, 2014), the rehabilitation of deprived environments (Chaskin, Khare, & Joseph, 2012), community reconstruction following natural disasters (Hawkins & Maurer, 2010) and community building (Ernwein, 2014).
Research on community development has been grounded in diverse social disciplines, such as sociology, social work, economics and urban planning. Consequently, numerous conceptualizations exist in the literature. Stopping short of a discussion on definitions of community development, scholars agree that genuine community development processes are bottom-up initiatives run for and by people who share the same social, health or economic circumstances and who draw on their knowledge of shared experience (Seebohm, Gilchrist, & Morris, 2012). Emerging at the local level or from grassroots efforts, the development initiative supports individuals, households, small groups and communities in generating their own solutions (Woolcock & Narayan, 2000) while also helping them to work with public and volunteer services (Seebohm et al., 2012). Moreover, development initiatives refer to those that actively include beneficiaries in their design and management (Mansuri & Rao, 2004) through the application of participatory principles (Cooke & Kothari, 2004). Several scholars have highlighted the large potential gains from community development processes (Mansuri & Rao, 2004). Such initiatives intend to give voice and choice to the beneficiaries (Mansuri & Rao, 2004) and empower them to make decisions and create solutions to their own problems (Seebohm et al., 2012). Moreover, community processes may increase organizational capacity among disadvantaged groups to facilitate improvements in their situations (Classen et al., 2008;Weinberger & Jütting, 2001). The rationale behind these types of initiatives is to create opportunities for sustainable development and to strengthen the capacity of people as active citizens (Mathie & Cunningham, 2003).
Despite this described potential, research has documented numerous drawbacks that may hamper these interventions (Mansuri & Rao, 2004). In an effort to improve the functioning of community processes, social theories have provided conceptual tools for explaining and comprehending the phenomena. Taking into account the ample theoretical production of the field, we draw on the concepts of (1) participation (Mansuri & Rao, 2004), (2) social capital (Woolcock & Narayan, 2000), (3) power (Cahill, 2008) and (4) the relationships between schools and communities (Shatkin & Gershberg, 2007) to underpin our study.

Participation
One key aspect of community development processes is the active involvement of local members throughout the project's design and implementation. Thus, participation is a key concept in these types of initiatives. Although it has been recognized that participation can occur at many levels (Arnstein, 1969), the most successful types of participation are those in which local members are placed in the centre of the project's decisionmaking processes (Classen et al., 2008). If this happens, projects can be better designed, and the benefits can be better targeted, which in turn improves the project's efficiency and sustainability (Mansuri & Rao, 2004). Furthermore, when community members are allowed to set-up their own goals based on their needs, the project gains legitimacy and builds support for sustained change (Barma, Huybens, & Viñuela, 2014).
However, abundant research has paid attention to the barriers that hinder participation. Mansuri and Rao (2004) argue that the exercise of voice and choice of the most marginalized populations can come at a high price given the time and cost involved in achieving adequate participation. Moreover, in some cases, participatory approaches are used as an instrument for promoting pragmatic policy interests rather than as a vehicle for real social transformation.
Identifying effective strategies to help vulnerable collectives surmount participation barriers is still a recurrent issue that has been recognized as a main setback in the academic literature (Barma et al., 2014). Nevertheless, several researchers have found particular features that help to overcome those barriers. Weinberger and Jütting (2001) observe that incorporating women into the process, especially in societies in which women have low social status, will increase the overall probability of participation. In a different vein, Classen et al. (2008) report that a commitment to long-term capacity building and empowerment is critical if elite capture is to be avoided. To achieve this, research shows that providing equal access to spaces for decision-making through capacity development, empowerment and the help of external organizations and agents can increase the participation of the most marginal members of the community (Classen et al., 2008). Furthermore, Samanta and Nayak (2015) observe that regular awareness campaigns at the local level may have a positive impact on the participation of socially disadvantaged sectors of the community in the decision-making process.

Social capital
The social capital theory has been widely used for grounding studies on community-led processes (Crespo, Réquier-Desjardins, & Vicente, 2014;Hawkins & Maurer, 2010;Nguyen & Rieger, 2017;Weinberger & Jütting, 2001). There is an agreement in the literature that community development relies on the use of the community's social capital to organize themselves and participate in development processes (Mansuri & Rao, 2004). Social capital literature distinguishes between what Gelderblom (2018) calls 'school of cooperation and school of competition'. In the first of these school, the focus is on explaining how the cooperation for the common good is possible, therefore, social capital has been defined as the norms and networks that facilitate collective action (Woolcock & Narayan, 2000), functioning as a 'glue' that helps individuals and organizations work together (Pretty & Ward, 2001). The second approach considers social capital as a resource that limits some and advantages others. (Gelderblom, 2018). It recognized four types of capitals (cultural, economic, symbolic and social) and these are unequally distributed among different social groups. People with much capital enjoy a great amount of agency while those with little capital find their agency constrained. Individuals compete to enhance their capital. Gelderblom (2018) argues that is possible to unify both theoretical approaches because cooperation and competition entails one another.
Scholars have highlighted several limitations to the concept of social capital due to its ambiguity regarding its definition, measurement, causality and as a policy tool (Inaba, 2013). In community development studies, social capital has been criticized for not being concerned enough with issues of class distinction and power (Mansuri & Rao, 2004;Mathie & Cunningham, 2003). Mansuri and Rao (2004) argue for a more nuanced understanding of social capital as part of the power relations within a social system. According to these authors, social capital must be contextualized because it is embedded within structures of power and can be used either to facilitate collective action or to perpetuate social inequality. Building the capacity for collective action cannot be divorced from a broad awareness of the power structure within which unprivileged groups attempt to cope (Mansuri & Rao, 2004).
Scholars have studied the relationships among social capital, networks and community development processes at the interpersonal and organization levels. Crespo et al. (2014) analyse the positive and negative effects of interpersonal networks on community process outcomes and show how structural properties of networks such as families may affect sources of loyalty, trust and collective commitment but are also a cause of discrimination and clannish behaviour. At the organizational level, scholars have observed that the presence of civic organizations can create conditions that foster social capital among low-income groups (Weinberger & Jütting, 2001). For example, Pinho et al. (2012) showed how the Catholic Church supported the self-organizing processes of fishermen to set-up exploitations rules in the absence of a legal framework in the Brazilian Amazon. Likewise, Classen et al. (2008) illustrate how the Foundation for Participatory Research with Honduran Farmers (FIPAH) has been able to involve the most marginal members of the community to solve agricultural challenges identified by the community itself. In a similar vein, Van Laerhoven (2014) examines the likelihood of citizen participation on local municipal environment councils in Brazil and finds that a combination of public-sector, private-sector and civil-society capabilities is necessary to increase the involvement of non-governmental actors.

Power
Community development processes have the explicit objective of reversing power relations in a manner that creates agency and voice for poor people, allowing them to have more control over the development of their community (Mansuri & Rao, 2004). However, research revealed that practitioners working on community development process often fail to recognize power relationships among participants (Cooke & Kothari, 2004). Overlooking power structures within the community may thwart the success of the development process. Elite capture has been widely documented as a particular injurious problem in such initiatives (Fung, 2005;Mansbridge, 1983) and is derived from the existence of asymmetrical power relations among beneficiaries. Elite capture refers to the domination of the decision-making process or to the control of project benefits by better-off groups in the community.
Although power has long been identified as a cornerstone of the sustainability of community development (Bastiaensen, De Herdt, & D'Exelle, 2005;Muñoz, Paredes, & Thorp, 2007;Theesfeld, 2004), only recently have scholars begun to disentangle how it operates in development interventions (Cahill, 2008). Drawing on Allen's (2003) geographical analysis of common theories of power, Cahill (2008) offers illuminated insight into how power works in community-led processes. Following Cahill's analysis, there are three conceptualizations of power: (1) power in things, (2) power through mobilization and (3) power as immanent. In the first conceptualization, power is embedded in particular material or immaterial resources. Thus, the powerful are those who use those resources to dominate others. This vision of power informed development practice through the assumption that people require increased access to resources to improve their own well-being (Cahill, 2008). In the second conceptualization, power through mobilization is defined as an effect produced by social network interactions. Under this perspective, gaining power means transferring decision-making powers to local communities to enable them to more effectively manage their own resources (Cahill, 2008). Finally, in the third conceptualization, power is dispersed throughout a complex web of discourses, practices and relationships that position some subjects as more powerful than others (Cahill, 2008). Underpinning this definition, Cahill argues that people exercise power in multiple ways, rejecting the dualism of the 'empowered' and 'disempowered'. According to this view of power, understanding the strategies people use to negotiate power relations is a first step in comprehending how power works through community-led interventions (Cahill, 2008). To challenge existing power structures, Cahill (2008) suggests creating a conducive environment that enable individuals and groups to come together to reflect on the forms of power and resources they could mobilize. This can be done through (1) reframing local relations, knowledge and resources as potential forms of power for disadvantaged groups, (2) allowing local groups to experiment with power as each member contributes with his or her own knowledge, resources and networks and (3) working on existing power structures to structure power in alternative ways.

Relationship between schools and communities
A wealth of literature links urban schools and their communities (Green, 2015). Traditionally, research on school-community development has focused on ways to engage local communities to improve school-centred outcomes, such as academic achievement or parent involvement (Castro et al., 2015;Jeynes, 2012). However, some scholars argue that schools should engage in more community-wide equity and strengthen local institutions through community development (Green, 2015;Warren, 2005). Keth (1996) suggested a three-tier model to join community development and urban schools. The model places community agency, networks and interest at the centre of the process instead of focusing on deficiencies or needs. Then, linkages between community groups are created while establishing tasks through democratic participation. Moreover, outside experts are invited to support the community efforts.
The research that we present in this article is in line with a number of studies that account for the enormous potential that schools have for promoting community engagement and increasing communities' social capital (Smith, 2015;Warren, 2005), thereby boosting social transformation in the contexts in which they serve. Warren (2005) argues that beyond the educational benefits, community participation in school life can build a political constituency to address inequality among vulnerable minority groups. Several studies on revitalizing marginalized neighbourhoods explore the ways in which grassroots organizing for education reform brings communities and schools together for the dual purpose of rebuilding the community and improving education. For example, Shatkin and Gershberg (2007) argue that parent participation in school activities may improve community development in three ways: (1) by fostering leadership within the community, (2) by increasing awareness of community problems and (3) by opening up schools for a number of functions and services that may benefit disadvantaged communities. Moreover, Nast and Blokland (2014) show that neighbourhood schools can provide spaces for families to meet, get involved in shared issues and exchange resources, thus paving the way for the activation of community initiatives.
Moreover, the potential of education to improve the quality of life of the most vulnerable communities has been documented by research on a significant diversity of contexts worldwide. For instance, community-based high schools in Argentina provided job alternatives for groups at risk of poverty by creating a supportive economic framework by re-establishing factories that had been closed during the economic crisis of 2001 (Oraisón & Pérez, 2009). In the United States, Smith (2015) collected different experiences of non-formal and formal educators who are committed to empowering young people living in impoverished urban regions to identify environmental problems in their communities and work collectively on solutions. Based on an active and dialogical process, these educators aim to prepare young people to preserve, restore and reshape their own neighbourhoods and communities according to residents' expectations.
Scholars have also argued about the important role that the school principal has in fostering community development. (Green, 2015(Green, , 2018. Research indicates that principals play an important role in forging robust connections between schools and their local communities (Green, 2015). Green (2015) observed how principals can support urban school reforms and community development by establishing partnerships with a variety of local organizations, making the school a space in which relations and networks are fostered to share information between the school and the community and in which to act creatively and correctively throughout the process.
In this article, we delve into the role contributed by the Mare de Déu de Montserrat Learning Communities (LCs) in the launch and organization of a community process in an impoverished Spanish neighbourhood. LCs are kindergarten, primary, secondary and adult schools in European and Latin American urban and rural contexts (Racionero-Plaza & Puig, 2017). This research-based project was created in the late 1970s by the CREA research centre of the University of Barcelona (Sanchez, 1999). Since then, a wide range of agents, including universities, NGOs, foundations and governments, have been involved to provide training and support to the staff and families of schools who decide to implement the project. Their distinction is the implementation of a set of successful educational actions (SEAs) 1 aimed to create spaces for dialogic learning with the participation of family and volunteers in different spaces of the school. Among the SEAs implemented are Interactive Groups (Valls & Kyriakides, 2013), an inclusive way to organize the classroom in which volunteers from the community participate during students' curricular activities; family education programmes aimed at families and community members and that are based on their own demands and needs; and spaces for decision-making, such as assemblies and committees in which decisions relevant to school life are made. Furthermore, LC promotes a dialogic leadership based on the ability of those situated in leadership positions, such as school principals or teachers, to work together with community members in an egalitarian fashion with the aim of improving the conditions of the school and the neighbourhood (Redondo-Sama, 2015).
Previous studies have shown the impact that LC has in disadvantaged contexts. This impact includes the activation of processes of desegregation, an increase in social participation, the de-stigmatization of the territory and the regeneration of the school's environment (Aubert, Villarejo, Cabré, & Santos, 2016). For instance, García Yeste, Lastikka and Petreñas (2013) provide examples of LCs in different urban and rural contexts in Spain that have managed to reorganize their local resources (human and material) to improve students' academic results and the quality of life of residents in the areas. Among the cases analysed, the study of García Yeste and her colleagues has provided us with preliminary evidence on the benefits generated by the Mare de Déu de Montserrat LC, the case study that is the focus of this research.
Subsequently, we briefly present the methodological approach. We describe the main features of our case study and explain the data collection and analysis processes used to address the research question.

Methods
This study has been conducted following the CM of research (Gomez, Puigvert & Flecha, 2011). CM aims to go beyond the diagnosis of situations of inequality by identifying interventions in different social areas that contribute to social change. In the CM, knowledge emerges through egalitarian dialogue between researchers and participants throughout the research process (Melgar, Larena, Ruiz, & Rammel, 2011). Creating situations that are more conducive to egalitarian dialogue involves identifying and avoiding elements that perpetuate hierarchical relationships in the research. Thus, during our study, we have taken into account some of the spatial and interpersonal factors that promote the emergence of power claims. For instance, we invited participants to be the ones who decided on where to conduct the interviews and the focus group, enabling interactions to occur in natural contexts (e.g., the neighbourhood association's headquarters, the community garden or the school). In addition, interviews were conducted with the aim of generating an intersubjective dialogue between researchers and social actors in which the new knowledge emerges from the intersection of scientific knowledge provided by the researcher involved in the dialogue with the knowledge of the situation in real life provided by research participants.

The case study: Montserrat's neighbourhood and its school
The Dream is a community development process initiated by a group of Moroccan residents and supported by the local school. Challenged by deteriorating economic and social situations, these residents made use of their social networks connected through the school to mobilize knowledge and resources aimed at finding solutions to their needs. After a participatory process involving residents and local organizations, they developed the idea of creating a community garden to provide fresh food to residents and to gain labour skills as a first step to combat unemployment. The methodology applied throughout the process is aligned with the asset-based community development approach (ABCD) (Mathie & Cunningham, 2003). A set of methods that have been used in this process resonate with those of the ABCD: (1) collecting community success stories, (2) forming a core steering group, (3) building relationships among local assets for mutually beneficial problem solving and (4) leveraging activities, resources and investments from outside the community (Mathie & Cunningham, 2003).
The case study builds on the participatory process that occurred in 2001 through which the school turned into an LC that aims to improve students' academic achievement. We argue that this participative process had consequences beyond the school, inspiring the local community to mobilize themselves to transform their hardships. In 2012, a new process commenced when several Moroccan neighbours, teachers and volunteers participating in the LC came together to look for solutions to the deterioration of their livelihoods. In this article, we analyse the contribution of community participation in the school in articulating The Dream initiative.
We now describe the main features of the neighbourhood in terms of its history, social composition and current economic situation. Moreover, we explain the process through with the school became an LC, its results and its consequences for the community.
The Montserrat neighbourhood is located in the city of Terrassa in northwest Spain. The Montserrat neighbourhood is on the outskirts in a particularly isolated area separated from the rest of the city by a stream and a road. The neighbourhood was built as part of a protected housing programme in 1955. In its beginnings, it received Spanish migrants from Andalucía. Currently, the neighbourhood takes in non-European workers hailing mainly from Morocco, Senegal and different countries in Latin America. Most of them who immigrated to Terrassa were attracted by the city's industrial growth and job demand in the construction sector. In 2014, the neighbourhood had a population of 1654 residents, of which 689 were non-European immigrants (Municipality of Terrassa, 2014). Therefore, almost half of its residents speak neither of the official languages, Catalan or Spanish, increasing the struggles of the integration process.
Montserrat's residents mostly depend on low-income unskilled jobs or remain in long-term unemployment. The unemployment rate of the neighbourhood's district was 28.5% in 2014 (Municipality of Terrassa, 2014), eight points higher than the national rate (Instituto Nacional de Estadísticas [INE], 2014) and six points higher than the city's rate for the similar period (Institut d'Estadística de Catalunya [IDESCAT], 2014). The active population of the neighbourhood was mainly employed in the building sector, which has been strongly affected by the outburst of the 2008 economic crisis. For instance, statistical data show that unemployment increased in the building sector from 830 individuals in 2007 to 5141 individuals in 2012 (Terrassa City Council, 2017). The number of job applications of the foreign population in Terrassa also increased from 98 in 1998 to 3907 in 2008 and 6597 in 2012. Thus, although the crisis deepened in subsequent years, migrant workers were unable to find new jobs, and unemployment benefits ended, leaving households in a highly vulnerable situation.
Furthermore, our case study analyses the dynamics promoted by the Mare de Déu de Montserrat elementary school. The increase in the immigrant population in the neighbourhood is reflected in the school's composition, which shows that the percentage of foreign children has soared from 11.87% in 2000to 46.08% in 2006(Flecha, 2015. Comparatively, the school received three times more students with a migrant background than other schools in the region. Additionally, according to regional standards, 82.2% of the school's students present specific educational needs derived from disadvantaged socio-economic situations 2 (Terrassa Pedagogical Counseling and Guidance Service, 2017).
To challenge the situation of educational underachievement and absenteeism that this school had historically suffered, in 2001, it embarked on a process of transformation into a LC 3 (Aubert et al., 2016). As a result of the project's implementation, students' educational achievement improved against all odds (Elboj, 2015). For example, students' reading achievement increased from 17% in 2000to 85% during 2006(Flecha, 2015. These observations were carried out in assemblies, working commissions, training sessions and working days and were held in different spaces in the neighbourhood: the school, the square, the community garden and the neighbours' association. The observations were audio recorded with the consent of the attendees, and the researchers took field notes on the most remarkable facts and interactions. We obtained copies of the minutes taken by participants in the assemblies and the working commissions.

Data collection and analysis
Second, the focus group and the interviews were conducted in places selected by the participants: the neighbourhood association's headquarters, the community garden and the school. The focus group was audio recorded and aimed to capture emergent interactions among the participants and to contrast different points of view about the initiative. For this reason, the sample for the focus group was intentionally selected using the criteria of significance and diversity in the points of view of the contributors to the project (See Table 1).
The data allowed us to obtain complex insight into the reality of the personal bonding and social networks developed throughout the process and relevant information.
Moreover, to deepen specific aspects, such as the school's role and leadership typologies, two open-ended interviews were conducted with key informants. On the one hand, Elena, the school's principal, was interviewed with the aim of collecting information on the role of the school in The Dream process and on the strategies that boosted the activation of the initiative. On the other hand, Bahir, a member of the organic community garden and a seminal participant of the initiative, was interviewed with the aim of examining his perceptions of the connection between his participation as a school volunteer and his later involvement in the organic community garden. All personal information has been anonymized by applying pseudonyms, codification and carefully safeguarding data privacy. Moreover, all data collected have been transcribed for further analysis.
Data analysis was carried out following the CM approach. The implementation of this approach has implied, on the one hand, that the interpretations obtained by the researchers from the data analysis have been contrasted and were agreed on by the research participants in subsequent meetings.
On the other hand, an analysis matrix was designed to include three categories along the two dimensions related to the CM: the 'exclusionary' dimension and the 'transformative' dimension. These two dimensions are considered in each of the three established categories extracted from the literature review: (1) school and learning, (2) dialogic organization and (3) leadership. Next, we outline each category according to the two dimensions.
School and learning. This category includes information on the school's role in the launch and organization of The Dream. Furthermore, participants' learning processes facilitated by the school have been included. The exclusionary dimension accounts for elements that hinder neighbour participation in school and community initiatives, He is one of the leaders of the community garden. He started participating in the community garden owing to the dissemination performed by the school, where his three children attend and where his wife participates in literacy programs. Abdel Male, 41 years old. Born in Morocco. He has lived in the neighbourhood for 14 years. He is married and has two children. He has worked on public infrastructure maintenance and the construction sector. He had been unemployed for 7 years when the focus group was conducted.
He is a member of the community garden. He became involved when seeking labour advice at the Neighbourhood Association, where he was invited to participate in the project.

Elena
Female, 52 years old. Born in Spain. She is married and has one child. She has been the principal of the Learning Community since 2001.
Since 2012, she has been one of the promoters of Montserrat's Dream. She had collaborated with the launch and organization of the community garden since its beginnings. Máximo Male, 64 years old. President of the Neighbourhood Association. He has lived in the neighbourhood since 1957. He migrated with his mother and three siblings from a rural region in Andalucía (Southern Spain) when he was a child.
He became involved in the launch and development of Montserrat's Dream and the community garden through the existing bond between the school and the Neighbourhood Association.

Marcos
Male, 40 years old. Coordinator of a non-profit cooperative for job placement located in the city of Terrassa.
He became involved in the community garden through his collaboration with the school. He contributes to the community garden with technical support and material resources at different times.
whereas the transformative dimension accounts for factors that enable them to overcome participation barriers.
Dialogic organization. This category refers to the role contributed by the school during the community process for creating spaces for dialogue and promoting conditions for communicative equality (Fung, 2005;Melgar et al., 2011). Thus, in this category, special attention has been paid to those strategies that contribute to including the voices of Moroccan neighbours and allowing the creation of horizontal relationships. The exclusionary aspect accounts for contextual constraints and attitudes that thwart participation, communicative equality and the creation of communitarian bonding. In contrast, the transformative aspect includes strategies and attitudes that promote understanding, consensus, respect, trust-building and the establishment of common goals.
Leadership. This category includes factors that have influenced the types of leadership (Andrews et al., 2010;Thorkildsen, Kaulio, & Ekman, 2015) identified through The Dream and focuses in particular on the dialogic nature of the leadership processes (Redondo-Sama, 2016a, 2016b. The exclusionary dimension accounts for leadership practices that hamper the fulfilment of the community's goals, whereas the transformative dimension refers to a leadership approach that generates favourable ground for community participation.

Findings
In what follows, we present the main findings. To respond to the research question, we look into the role played by the school in the launch and development of The Dream in the Montserrat neighbourhood and the achievement of one of its main goals: the creation of a self-managed urban garden run by a group of unemployed Moroccan neighbours. The Dream began in 2012 with the aim of preventing the deterioration of the living conditions of the Moroccan neighbours. A starting point was a call for a general assembly in which a large number of neighbours, teachers and school staff, school volunteers, social agents and civil servants were invited. The assembly aimed to collect the neighbours' demands and priorities related to their quality of life. As a result, assembly attendees identified the need to create job opportunities and training as an immediate priority. A training session was organized to discuss research-based actions that had an effect on the inclusion of vulnerable communities in the labour market. Several actions were discussed, among which the model of successful cooperatives stood out (Redondo, Santa Cruz, & Rotger, 2011). This model became an inspiration for participants who wanted to pursue similar action by setting up an organic garden in the neighbourhood.
Participants in The Dream created the community garden with a double objective. On the one hand, in the medium term, they saw it as a strategy to acquire new job skills and to face the isolation and loss of social networks that they were experiencing given their prolonged unemployment situation. On the other hand, in the long term, they contemplated the garden as a first step towards the creation of a labour insertion cooperative. In this study, we have collected results in relation to the first objective, which is linked to the acquisition of work skills and the improvement of social networks by participants in the garden.
Next, we first address how the school has contributed to overcoming some of the limitations that made it difficult for Moroccan neighbours to articulate a community response to their disadvantaged situation. Second, we focus on the role played by the school during the development of The Dream and highlight its contribution to the maintenance of a horizontal and dialogic organization of the community process.
The role of the school in the process of launching the dream The data obtained from the fieldwork suggest that Moroccan neighbours perceived the existence of specific barriers that made it difficult to articulate community responses to cope with the consequences of long-term unemployment. Among these obstacles, they highlighted a lack of knowledge and economic resources, as well as aspects that limited their participation in public spaces (e.g., low educational levels and difficulties speaking or reading in Spanish). To face these constraints, the school staff was able to mobilize its previous social network to provide quality knowledge on which to articulate the community's action and to promote the participation of Moroccan neighbours in the process.
Mobilizing the school's social network to launch the community process Data obtained reveal that the transformation of the Mare de Déu de Montserrat Elementary School into an LC in 2001 promoted the creation of a strong social network among different agents, such as neighbours, professionals from diverse fields and volunteers participating in several school spaces. This involvement had an impact beyond the school itself, creating bonds among Moroccan neighbours and different agents present in the neighbourhood. We have observed how this school-centred network also had a positive impact on both the launch of The Dream and the achievement of some of its goals.
This collaborative network has extended spaces for sharing concerns about the increase in unemployment and poverty in the neighbourhood and has supposed that the starting point goes beyond a diagnosis of the situation to finding possible solutions. Consequently, the significant diversity of human resources present in the LC has helped neighbours to obtain specialized knowledge and quality information on how to articulate the community process. For instance, the following excerpt from the school's principal offers relevant insights into how the interactions established in the school context have made it possible to base the community process on scientific knowledge. Specifically, the principal mentions one of the conversations established with Noa, a researcher and university teacher who is involved as a volunteer in family education programmes in the school.
Talking to one of the volunteers who comes here from the university to participate in the school, we identified that one possibility to improve people's lives would be to ask them what they would like their neighbourhood to be like and to dream (Elena, school's principal).
Thus, the presence of Noa in the school provided relevant information on researchbased community processes such as The Dream. Obtaining this information was essential to identifying the type of process that offered previous evidence of social impact in highly disadvantaged contexts and thus offered greater chances of success.

Increasing knowledge among Moroccan neighbours
The acquisition of specialized knowledge and support through the social network created by the school has not only made possible the launch of The Dream process but also has made viable the development of the community garden. The case of Laura, a biologist who was contacted during the first phases of The Dream by a school volunteer, exemplifies this dynamic. Laura offered free guidance and training to participants in the urban garden on organic agriculture. Data indicated that the manner in which this training was carried out contributed to reducing some cultural prejudices and gender stereotypes that came into play in the relationship between the group of participants, which was made up entirely of middle-aged Moroccan men, and Laura, a young European woman. Although some group members were reluctant to accept Laura's arguments during the first sessions, this attitude was transformed as she became involved in some assemblies, worked in the garden on an equal footing and provided needed technical knowledge, rather than adopting an attitude of academic superiority.
The social network provided by the LC allowed participants in The Dream to improve their knowledge and to perceive the benefits derived from the knowledge gained throughout their involvement in the community process. In this regard, an unemployed Moroccan neighbour highlighted that the training provided by Laura has allowed him and his family to improve their dietary habits and raise awareness among other community members.
[Since getting involved in the garden] I became more conscious on the health problems chemical products cause to people, at the local and at the global level [. . .]. Thus, I decided to try to do my bit and help raise awareness among the people I know. The great advantage we have here at the community garden is that we eat healthy and we change the mentality of the next generations, of our children, because change starts at the bottom (Mohamed, Moroccan neighbour).

Involving Moroccan residents in the community process
We have identified that one of the difficulties perceived by the promoter group 4 was the necessity of involving the most disadvantaged residents in the community process from the initial stages. The data obtained suggest that the participation that has flourished and has been consolidated at the school contributed to challenging this limitation. For instance, Moroccan women participating in the family education programmes provided by this LC have obtained information on The Dream in the school, a fact that has allowed them to take an active role in its dissemination. Thus, these women have prompted their relatives and acquaintances to join the process. In this regard, an unemployed Moroccan neighbour stated that the information provided by his wife, who participates in the family education programmes, was essential for him to become one of the promoters of The Dream and the community garden: The information [on the project] came to me through my wife [. . .]. And since there wasn't work and I was overwhelmed from going from the TEA to the state agency, from here to there in vain, I thought, alright, I'll spend some time doing what I like, which is to work in the garden and with plants. And then I got in touch with the school principal (Said, Moroccan neighbour).
The role of the school in ensuring a horizontal and dialogic organization One main challenge in the process was to assure that The Dream was carried out based on neighbours' real needs and decisions and not on the professionals' or privileged community members' needs. We have identified that the culture of participation and organization that existed previously in the LC has influenced the dialogic and horizontal functioning identified in The Dream. The Mare de Déu de Montserrat school, in its long trajectory as an LC, often held assemblies in which Moroccan neighbours participate in decision-making processes regarding relevant aspects of the school's life. In addition, Moroccan neighbours were used to working together with teachers and volunteers in the school's mixed committees. Thus, participants have highlighted that the strategies used to foster participation in school decision-making (e.g., assemblies and mixed committees) were applied to The Dream. For instance, eight general assemblies were conducted from the launch of the process in June 2012 to the public event in March 2014 of allocation of the land for the community garden. Moreover, mixed committees have met regularly, taking on board neighbours, volunteers and school staff to work on the design, organization and achievement of the community garden.
During the communicative observations of the assemblies and committees, we have acknowledged that the organizational dynamics promoted by the LC have contributed to reducing some elements that hinder the horizontal and dialogic organization of the community processes. Among these are racist behaviours, power claims or a weak sense of community belonging.

The distribution of power
The fieldwork has allowed us to observe how the school staff assumed the role of promoting a horizontal dialogue in which better-off participants' claims were not imposed on the rest. The strategies they have used to facilitate the distribution of power have been included to ensure the presence of translators who support neighbours with less proficient Spanish skills and of moderators. These moderators guarantee the opportunity of all of the participants to intervene and be heard or maintain a strong position to stop racist attitudes or behaviours that favour the elite capture. Our perception coincides with that of the interviewees because during the fieldwork, many of them have identified the mentioned aspects.
In this regard, in one of the communicative observations carried out in a meeting, we witnessed one of the leaders of the neighbourhood association formed by native residents rely on racist prejudices to deny the capacity of Moroccan residents to work and manage the community garden. In addition, he claimed that if the association did not exercise greater control over the garden, the project would fail. This interaction generated tension among the attendees and caused the Moroccan neighbours to not respond in those first moments. However, the school's principal showed that these statements lacked arguments of support and gave the floor to the Moroccan neighbours. In the next quote, the coordinator of a cooperative for work inclusion, who was also present at this meeting, refers to the benefits of the interactions provided by school personnel.
In the role of mediation that the school has made, one thing that was very clear was what the bases of the project were. I remember the first meeting I attended, and it was a tense meeting. I remember it was a tense meeting, but the person who came from the school deep down was quite frank, having very clear ideas and saying, this goes around here. And that starting point helps to clarify much the role of each of the people who participate in the process (Marcos, coordinator of a local cooperative).

Opening dialogic spaces to resolve conflicts and share leadership roles
The data obtained also suggest that school professionals and volunteers supported the Moroccan community members in solving conflicts related to the garden's management. Specifically, they helped the neighbours open dialogic spaces that aimed to resolve tensions and promote leadership roles among Moroccan neighbours.
On the one hand, during the fieldwork, different research participants shared with us their concerns about the emergence of a conflict in the garden. Two of the Moroccan neighbours involved in the garden did not go to work on their assigned shifts and collected fruits and vegetables behind other peoples' back. After a few weeks of increasing tension and suspicion in the garden group, some of its members decided to turn to the school principal for advice. Faced with this situation, the principal offered them a space in the school in which they could hold assemblies and meetings. Furthermore, she put at their disposal school volunteers to help them moderate the meetings, and she offered herself as a facilitator. At these meetings, participants decided to hold compulsory weekly assemblies. Furthermore, to guarantee the validity of and compliance with the agreements, volunteers were required to draw up all agreements in a report and ensure that non-literate members understood and agreed with the decisions.
In this vein, several interviewees highlighted that during The Dream, the school has been perceived as a neutral and accepted space. In addition, they noted the influence that the school context has exerted in moderating some of the conflictive interactions that occurred in the garden. Specifically, as mentioned by the principal, the school context has influenced participants in assemblies to abandon defiant behaviours and adopt consensus-oriented attitudes.
When the points of view are different, the fact of thinking, "I'm in school, I'm with the principal", is very important. Some tensions are blurred because they think: I'm in an educational context, and then I go beyond my personal situation to prioritize what is educational, what is best for everyone. It's not educational that I position myself solely in my point of view, because there're other points of view that maybe I didn't see, but I make the effort. Maybe I don't make the effort in the garden, but I do at that table in the school where we're all sitting together. And in that moment what we all have in common are the school children. All that hardness that was in the garden is blurred, and it's for that educational environment that is breathed. (Elena, school's principal).
In contrast, we have identified during The Dream that the type of leadership exercised by the school professionals and volunteers has facilitated the adoption by several Moroccan neighbours of leadership roles. This leadership modality is in line with the so-called dialogic leadership (Redondo-Sama, 2015). As mentioned in the theoretical section of this article, dialogic leadership is committed to generating dynamics in which the leader position is shared between different participants regardless of their backgrounds. Thus, the school staffs of this LC are far from autocratic leaders and have encouraged Moroccan neighbours to lead community actions.
Several research participants have perceived benefits that derived from the dialogic leadership exercised by the school staff, such as the emergence of a feeling of coresponsibility for community actions or the prevention of patronizing attitudes towards Moroccan neighbours. Nonetheless, the school principal stressed that establishing the foundations of a dialogic leadership entailed a learning process for all depending on their capacity to participate in dialogic spaces from equal positions and through the recognition of Moroccan participants' capacities to contribute to the community process.
We're all seated here [in the assembly], facing this conflict and we say it to our faces, and this is when everything is regulated. However, it's regulated through a shared leadership. It's a horizontal relationship and we base our positions on arguments [. . .]. Then, we learn to listen each other, because we're people that have not been in leadership positions before and haven't made decisions at work. Thus, it's that learning that we have gone through; one that helps us to learn from all of us. (Elena, school's principal).

Discussion and conclusion
In this article, we analysed a case study that provides important insights into how vulnerable communities affected by unemployment and poverty can set-up community-based initiatives and build collaborative networks to tackle their difficult living conditions. Concretely, through the analysis of The Dream in the Montserrat neighbourhood, we focused on a community development process carried out by a group of low-qualified and long-term unemployed Moroccan residents in urban Spain. Through a participatory process supported by the local school, a group of residents of this deprived neighbourhood came up with the idea of creating a community garden. This initiative is currently contributing to improving the residents' welfare. Not only has the initiative provided them with free organic fruits and vegetables, but it has also allowed members to gain new knowledge, skills and relationships. Overall, the process resulted in developing organizational capacity (Mansuri & Rao, 2004) while fostering a sense of active citizenship (Mathie & Cunningham, 2003) among local groups.
The case study focused on the role played by the Mare de Déu de Montserrat elementary school, which was created in 2001 as an LC. The school and its local network acted as an enabling environment that facilitated overcoming barriers that have traditionally hindered the success of community-based processes. Concretely, the school provided dialogical spaces through which members of the local community were empowered to express their needs to find solutions to their problems (Mansuri & Rao, 2004). Through an active and dialogical process of participation promoted by the school, the most marginal neighbours of the community were able to exercise alternative forms of power by recognizing and mobilizing their own resources (Cahill, 2008;Cooke & Kothari, 2004). In this sense, our study contributes to the debate of how schools can have a community-wide impact that goes beyond school-oriented outcomes (Flint, 2011;Green, 2015;Warren, 2005). It shows a particular manner of organising and managing schools as a LC based on dialogic leadership and horizontal organization (Aubert et al., 2016) that prompted active participation of the vulnerable members of the community.
First, our results showed how the local school played a crucial role in the launch of The Dream in providing it with a supportive social and institutional network. The transformation of the school into an LC fostered a strong collaborative social network among local agents and institutions, such as with the university or the neighbourhood association. The approach adopted by the school staff allowed them to mobilize this social and institutional network and put it at the services of The Dream. This network of professionals and volunteers provided quality information and support to the Moroccan participants, compensating for their lack of knowledge and experience in setting up and managing a community garden. For instance, we have shown how people involved in the community process gained access to scientific and technical guidance as a result of volunteers collaborating at the school who bridged the gap between the university and the most disadvantaged members of the community. In this regard, we highlighted the importance of the school principal (Green, 2015) in connecting and fostering networks among local community groups and other local organizations. Thus, with this article, we revealed the capacity of this LC to mobilize existing human resources in the local context and use them efficiently to respond to the most urgent needs of vulnerable community members. This result is in line with other studies that show how the existence of a previous social network facilitates the development and sustainability of community-led initiatives (Classen et al., 2008;Pinho et al., 2012;Weinberger & Jütting, 2001).
Second, our study identified successful strategies that facilitated placing the voices and demands of the Moroccan participants at the centre of the community-based process. We observed that the previous democratic and dialogic culture that was consolidated through the LC contributed to The Dream by providing it with open dialogic spaces such as assemblies or mixed committees and with a horizontal and consensus-oriented organization. The school's openness to the problems of its community catalysed the involvement of traditionally excluded groups in the participatory processes (Classen et al., 2008;Mansuri & Rao, 2004), encouraging unemployed Moroccan immigrants to get involved. The school staff and volunteers also played a remarkable role in helping members of the process establish common norms and act creatively and correctively against racist attitudes and imbalances of power (Green, 2015;Woolcock & Narayan, 2000). Furthermore, they supported the Moroccan neighbours in searching for spaces for dialogue to address the problems that threatened the process, such as the emergence of individualistic and selfish behaviours, the lack of selfmanagement skills or a weak sense of belonging. Through these strategies, the school staff was able to put the community's need at the centre of the process while promoting trust, reciprocity and social cohesion. This finding is in line with other studies that show how schools can foster social capital (Flint, 2011).
Third, the case study revealed the importance of promoting a more egalitarian and horizontal dialogue among participants by substituting authoritarian leadership models with so-called dialogic leadership (Redondo-Sama, 2016a, 2016b. This occurred by encouraging the Moroccan neighbours to assume leadership roles throughout the process. Giving the neighbours the opportunity to assume leadership not only prevented the emergence of patronizing attitudes towards them but also challenged traditional power structures between experts and lay participants (Cahill, 2008). The exercise of dialogic leadership permitted each member to contribute his or her own knowledge, resources and networks to the development of the project. This approach to leadership contributed to develop a sense of ownership over the project that is essential for its sustainability. Thus, our study demonstrates the potential of dialogic leadership approaches to allow alternative ways of exercising power in community-based development processes in disadvantaged contexts (Cahill, 2008).
Finally, it should be mentioned that this article provides the first results of an ongoing process. Therefore, further research is required to determine the long-term impact and the transforming capacity of The Dream to make the collective dreams of the Montserrat neighbourhood come true. Notes 1. SEAs are educational actions that have been shown to promote excellent results in terms of improving educational achievement, coexistence and social cohesion, regardless of students' characteristics or the contextual features of the school (Flecha & Soler, 2013). These actions are implemented through the joint intervention of the school staff, family members and volunteers, among which are community members, former students and professionals from various areas. 2. The Government of Catalonia considers students with 'educational needs derived from disadvantaged socio-economic and socio-cultural situations' those who present learning difficulties caused by the interaction between the child's personal characteristics and unfavourable characteristics of the educational environment in which they are enrolled. 3. The implementation of the LC project involves initiating a participatory process that involves school staff and family and community members. During this process, a network of university trainers offers training on the SEAs that articulate the project. Furthermore, decision-making spaces are opened in the school for teachers and community members to make joint decisions about how SEAs can be applied to respond to the needs of their context. 4. A promoter group composed of three Moroccan neighbours, the school principal, the Neighbourhood Association president, two researchers, one biologist and two school volunteers was created in 2012 with the aim of launching The Dream.

Disclosure statement
No potential conflict of interest was reported by the authors.