Sport for Social Cohesion: Transferring from the Pitch to the Community?

: European sport policies and programmes have increasingly focused on promoting social cohesion. Often presented as a multi-dimensional concept, social cohesion is considered the ‘glue’ that holds societies together and is seen as essential to addressing common challenges. However, the term remains contested, and it is not always clear how programmes conceptualize or support social cohesion. Thus, this paper explores how three European sport programmes conceptualize and foster social cohesion. Findings are generated from a thematic analysis of interviews, group discussions, observations and documents. The themes developed show how organizations adopt an individual-centerd view of social cohesion, focusing mainly on social relations, tolerance and mutual help. In turn, this translates to an individual-focused practice of social cohesion, emphasizing personal skills, behaviors, and social relations, with the transfer of social cohesion to the broader community left mostly in participants’ hands. Due to a number of systemic barriers, programmes struggle to implement more holistic and structural approaches. As such, if we want to facilitate a move towards more structural or interventionist approaches, we as researchers must play an active role in questioning, challenging, and reshaping the systems that underpin sport-based social interventions.


Introduction
Over the last two decades, social cohesion has gained increasing prominence as both an academic concept and policy objective (Moustakas 2022;Fonseca et al. 2019).The growing prominence of social cohesion has also translated to the field of sport, with numerous policies, programmes and researchers claiming to focus on the concept (Moustakas 2021;Svensson and Woods 2017;Schulenkorf et al. 2016).For instance, about 25% of sport for development (SFD) literature, and 10% of programmes, have been classified under the banner of social cohesion (Schulenkorf et al. 2016;Svensson and Woods 2017).
However, as Raw et al. (2021) note, the way social cohesion has been applied in sport "means that it is often used as a catch-all to describe a broad range of sociological concepts and that this has led to exaggerated claims about how well this area has been researched" (p. 19).Indeed, in one scoping review of 35 articles on sport for social cohesion, around half of the articles did not define the term (Moustakas and Robrade 2022).And when social cohesion is defined, it is often conflated with individual-focused ideas about social capital (Sabbe et al. 2020;Raw et al. 2021;Cubizolles 2015).In other words, though social cohesion is a broad and multi-dimensional concept, its application in literature is often reduced to a focus on the quantity and quality of individual social relationships.
Meanwhile, while the term is contested and debated in the general literature, research has been dominated by top-down, positivist definitions and measurements of social cohesion (Fonseca et al. 2019;Schiefer and van der Noll 2017;Bruhn 2009;Delhey et al. 2018).Together, this has allowed the concept of social cohesion to be taken for granted or, worse, let academics and policymakers impose their understandings on programmes and communities, excluding the voices and experiences of local practitioners and participants (Raw et al. 2021;Bernard 1999).
To unpack the connections between sport and social cohesion, it is essential to explore individual experiences and the practice of social cohesion (Novy et al. 2012;Raw et al. 2021;Sabbe et al. 2019Sabbe et al. , 2020)).Looking at the perspectives of practitioners and participants can help elucidate how social cohesion is understood in practice, how programmes work to support it and the assumptions underlying that practice.Against this background, this paper seeks to answer three related research questions: (1) how are social cohesion and its causes defined within European sport for social cohesion programmes; (2) what practices or activities do these programmes employ to promote social cohesion; and (3) how do programmes support social cohesion in their broader communities?In the following, I present the results of qualitative research carried out with three European sport for social cohesion programmes.Exploring these different contexts can help unearth potential commonalities, shared struggles and differences, thus contributing to ongoing discussions around the meaning and practice of social cohesion in sport.
Moving forward, this paper progresses in four steps.First, I will discuss some of the existing literature focusing on the understanding and experience of social cohesion in sport.Second, I will present the organisations I worked with and the overall methodology.Third, I will present the themes resulting from my analysis and discuss these against the broader social cohesion literature.Finally, I will conclude by suggesting potential research avenues in this area.

Sport and Social Cohesion at the Community Level
The concept of social cohesion has a long and complex intellectual history that dates at least to the 19th century and has since continuously been influenced by the assumptions and boundaries of various scientific disciplines, from political science to sociology, to psychology, to anthropology, to the health sciences (Bruhn 2009;Spaaij 2013;Kearns and Forrest 2000;Taylor and Davis 2018;Fonseca et al. 2019;Schiefer and van der Noll 2017;Dragolov et al. 2016).In turn, this has led social cohesion to be conceptualized with a wide array of dimensions, including shared values, shared experiences, civic participation, mutual help, trust in others, place identification, perception of fairness, social networks, social order, acceptance of diversity, well-being, equality and social mobility (OECD 2011;Bruhn 2009;Fonseca et al. 2019;Schiefer and van der Noll 2017;Jenson 1998;Delhey and Dragolov 2016;Moustakas 2022).
The multi-disciplinary and contested nature of the term has spurred numerous researchers and policymakers to attempt to summarize or (re)define the term.These efforts have led to either maximalist, all-encompassing definitions, or more narrow conceptualizations.On the maximalist side, Fonseca et al. (2019) proposed a model that incorporates ideas of well-being, belonging, social participation, tolerance and equal opportunities.In this view, all elements representing or contributing to social cohesion are mapped in a framework at the individual, community and institutional levels.Likewise, many prominent policy definitions take similar views, integrating a wide range of dimensions, including inequality, well-being and social mobility (OECD 2011; Council of Europe 2010).These expansive definitions have, however, been critiqued for including dimensions that may be better characterised as antecedents or consequences of social cohesion (Friedkin 2004;Moustakas 2022;Chan et al. 2006;Schiefer and van der Noll 2017).
The goal here, however, is not to adjudicate the definitions or conceptualizations of social cohesion.Suffice it to say, despite the variety of definitions and understandings, at a minimum, literature on social cohesion revolves around three core components: a sense of identity or belonging, social relations and orientation towards the common good (Schiefer and van der Noll 2017;Dragolov et al. 2016).Namely, social relations refer to the quality, tolerance and trust within different social networks.A sense of belonging denotes feelings of attachment or identity towards a social or geographic entity.Orientation towards the common good connects to ideas of mutual help, the feelings of responsibility towards others and an acceptance of the social order (cf.Schiefer and van der Noll 2017).Likewise, literature around sport and social cohesion tends to coalesce around social relations, a sense of belonging, and an orientation towards the common good (Moustakas and Robrade 2022).
More critically, the debates on social cohesion are dominated by top-down and positivist approaches to defining and measuring the term.Discussions on the appropriate definition of the term are primarily located within the policy and academic spheres (Schiefer and van der Noll 2017;Chan et al. 2006;Fonseca et al. 2019), while measurement tends to rest on the positivist application of survey and statistical tools (Delhey et al. 2018;Delhey and Dragolov 2016;Langer et al. 2017).
Within sport-related literature, many scholars use more constructionist or critical qualitative approaches to explore the logic, assumptions and practices embedded within specific programmes (e.g., van der Veken et al. 2021;Trejo et al. 2018).However, there has been relatively little literature grappling with the understanding and practice of social cohesion within sport for social change programmes (Sabbe et al. 2020;Raw et al. 2021).And this, despite social cohesion being presented as a crucial organizing concept within programmes (Svensson and Woods 2017), literature (Schulenkorf et al. 2016) and policy (Moustakas 2021).Thus, the following analysis aims to summarize and contextualize the relevant sport literature exploring social cohesion within different contexts.
Indeed, there is a small but growing body of work exploring the understanding, experience and practice of social cohesion, in the sport context.Here, notions of social relations, tolerance, the acceptance of diversity and a sense of belonging are omnipresent, with many studies or organizations focusing on minority or migrant groups (Cockburn 2017;Kelly 2011;Raw et al. 2021;Sabbe et al. 2019Sabbe et al. , 2020;;Meir and Fletcher 2019).For example, many programmes focus explicitly on helping diverse adults and young people forge social relations and build trust (Kelly 2011;Cockburn 2017;Fehsenfeld 2015).
In the above literature, there is a recognition of some of the structural aspects of social cohesion, such as socioeconomic inequality or relations to public institutions, and the need to improve the overall conditions of participants.Still, many organizations and individual practitioners struggle to negotiate these more structural aspects.In an account of their struggles implementing a Participatory Action Research (PAR) approach within a sport for social cohesion programme, Meir and Fletcher (2019) document how unclear programme goals and a lack of support for the participants' day-to-day struggles inhibited their efforts.Likewise, within programme activities, practitioners face challenges integrating structural components such as inequality or systemic racism, often focusing instead on a narrower set of individual-focused approaches and social cohesion goals (Sabbe et al. 2020;Flensner et al. 2020).
Because of this, many programmes have the effect of placing high amounts of responsibility on the shoulders of their predominantly 'vulnerable' target groups.Sabbe et al. (2020) note this explicitly as a theme in their research, whereby many practitioners believe that "participants can transcend their living circumstances, as long as they continuously engage themselves in activities" (p.266).Likewise, Cockburn (2017) notes how minority youth are often put in the position of intermediaries between adult members of their ethnic group and the majority group.In a cumulative sense, these individual-focused programmes risk de-emphasizing structural factors and instead highlight perceived individual deficits (Kelly 2011).

The Project and the Organisations
The three organizations highlighted in this study are, first and foremost, united by their shared participation in a European Social Cohesion Project.Launched at the start of 2021, the project brings together partners from the NGO, university and advocacy sectors to promote social cohesion in diverse settings and support practitioners delivering sport for social cohesion programmes.Recognizing current debates and gaps around social cohesion, the project set out to explicitly explore and understand the definition of social cohesion embedded in the respective communities.Supporting that goal, the project implements a living lab approach (see, e.g., Galway et al. 2022; European Network of Living Labs 2021) to directly engage programme participants, generate an understanding of the elements that promote social cohesion in a sport setting and to co-create relevant tools to allow for the exploration, understanding and improvement of social cohesion outcomes.Formally, living labs are understood as "user-centred, open innovation ecosystems based on systematic user co-creation approach, integrating research and innovation processes in real-life communities and settings" (European Network of Living Labs 2021).This living lab approach also provided a valuable, participatory framework to engage practitioners and participants in the research documented here.
Three organizations from Ireland, Germany and the Netherlands are the focus of this paper.All three have a longstanding involvement in delivering sport for social cohesion activities in their respective urban communities.The Irish organization, which has been in place for over 20 years, focuses on combatting racism and providing opportunities to various groups across their city, including working-class youth, refugees and immigrants.To do so, they host a range of weekly sport sessions, regular festival-type events, and thematic workshops in public and community facilities.In the Netherlands, the organization has worked with municipal and non-governmental stakeholders for over three years and attempts to reach diverse groups, including immigrant parents and their children, through regular sport, cultural and artistic activities.These activities vary according to the partners or stakeholders involved and reach a wide range of target groups to foster familiarity and social encounters.Finally, the German organization operates directly in local primary schools and is integrated into the regular curricula.Concretely, this means that students in these schools participate in the programme once a week for all four years of their primary education.Through this, they use modified sports and games to support interaction between the children as well as develop emotional and intercultural competences.
These organizations were chosen as they not only work on a common objective but also represent the diversity of target groups and settings.In particular, these organizations work across the community, club and school settings that are inherent to many European sport for social cohesion programmes, or sport for development projects more broadly (Moustakas et al. 2021;Svensson and Woods 2017).In addition, these organizations are embedded in numerous pan-European projects and initiatives and exchange regularly with programmes, NGOs, higher education and youth.This engagement gives the organizations a chance to influence and be influenced, suggesting that their views and approaches relative to social cohesion did not emerge in isolation but through iterative and longstanding interaction with the larger European SFD ecosystem.As such, the targeted organizations are likely to provide a suitable basis for developing comparable or transferable findings.

Design and Research Background
This study is based on a mix of interviews, focus group discussions, documents and observations that explore how the three European sport programmes conceptualize and act to support social cohesion in their communities.The qualitative design of this study and explicit focus on social cohesion were chosen to center the perspectives and experiences within the programmes.Thus, I situate this research somewhere between the constructionist and transformative paradigms.In other words, for this topic, I understand social reality and knowledge as perspective and context-dependent, and I also recognize that cultural, historical and political forces influence our knowledge and reality (Mertens 2007;Chilisa 2020;Braun and Clarke 2022).Having said that, I hesitate to fully align with the transformative paradigm, as the interactive link between the researcher and participants in defining the problem was not fully realized (Mertens 2007).
My background differs from many of those involved in the programmes and this research.As a white, male, cisgender, heterosexual Canadian immigrant living and working in Germany, I come from a position of relative privilege and affluence.Still, my experiences moving within and between countries as a child and young adult-sometimes by choice, sometimes not-influence my perspectives on this topic.I have felt varying levels of belonging, trust and identity within the often changing (sport) communities I inhabited, giving this topic a somewhat personal resonance.Professionally, I have also worked on the design, implementation and evaluation of sport for development programmes, at the NGO and academic levels.Combined, these experiences influence my perspectives on this topic and, hopefully, provide me with insight to ask pertinent questions and explore relevant topics.

Data Collection
For interviews, participants were recruited from individuals and stakeholders involved, either now or in the past, with the three targeted organizations.These individuals were targeted as they could provide a range of insights into the targeted organizations, including from the perspective of managers, coaches and (former) participants.Previous professional connections, as well as connections through the project, facilitated many of these contacts.In other words, a mix of convenience and snowball sampling approaches were used.In general, participants were either approached at events, meetings, via e-mail or through other contacts.This allowed me to reach various participants and stakeholders but also restricted my ability to reach certain groups in certain locations (e.g., programme participants in the Netherlands).Before each interview, I explained the general purpose of the research to the participants and assured them that their statements would remain anonymous.Verbal and/or written consent was obtained for all interviews, and the participating organizations provided written support for this research beforehand.Ethics approval was also obtained from my university for this research.When logistically feasible, interviews or discussions were recorded with a digital recorder, and participant approval was obtained to do so.
In total, 24 individuals participated in interviews (n = 20) or group discussions (n = 4) between April 2021 and June 2022.These interviews occurred during site visits to the organizations, as well as during project meetings and through scheduled online interviews.As such, interviews took place in a number of settings, including at cafés, offices, community centers, during neighborhood walks or online.These interviews sought to generate conversations that would help unearth how individuals understand and experience the programmes, social cohesion and the perceived links between the programme and social cohesion (cf.Smith and Sparkes 2019).Of note, participants were asked about the goal and structure of the programme, their programme's understanding of social cohesion and the status of social cohesion in their communities.For the German context, a research assistant facilitated or co-facilitated discussions with seven individuals.To ensure consistency and quality, the interview guidelines were reviewed with the RA, and the first two interviews were co-facilitated to provide practice and feedback opportunities.Notes were taken following each interview, and verbatim transcripts were produced for most, though some interviews could not be recorded due to external factors (e.g., loud public spaces, weather conditions).An overview of interview participants is provided in Table 1.
Complementing these interviews, I visited, observed and interacted with the organizations throughout the project, including through meetings ( 5), sport or other activity sessions (20) and site visits (6).My presence as a researcher was always known and explicit, and I oscillated between passive and active observation depending on the setting and occasion.At times, I was largely on the sidelines and passively watched (e.g., a sport workshop, a meeting between local stakeholders), while at other moments, I played a more active role (e.g., playing football, coordinating a project meeting).This participation and observation gave me a first-hand glimpse of how the organization's views on social cohesion translated to their everyday activities and a better sense of their overall approach.In addition, this provided me with an opportunity for numerous smaller, informal interactions with various participants and stakeholders, thus bringing in numerous complementary perspectives.During and following these interactions or observations, I took notes to document the physical environment, participants, exchanges, activities and my personal reflections.
Further observations were also made by a research assistant, who regularly visited the German programme over six months.Finally, I obtained access to several programme documents, including manuals, activity guides, presentations, evaluations and project notes, and these further helped organize and contextualize my analysis.These documents were obtained through searches of organization webpages as well as through communications with interviewees and other programme members.Along with the programme websites themselves, these documents included activity handbooks (two), presentations (six), external communication materials (two), and internal research documents (three).

Data Analysis
Data analysis was carried out through Reflexive Thematic Analysis (RTA) (Braun and Clarke 2022).RTA offered the possibility to develop an analysis that allowed me to give voice to the perspectives and experiences of on-the-ground stakeholders often left out of discussions about social cohesion.At the same time, this approach allowed me to locate and analyze these experiences against existing literature and concepts relating to sport and social cohesion, and identify patterned meanings.
I primarily used the MaxQDA 2022 to organize my data, write memos, develop codes, and generate themes.Throughout, I maintained documents to diarize my thought processes and reflections and tracked all interactions with the respective organizations in a separate table (Nowell et al. 2017).These documents, combined with extensive handwritten notes, form the basis of an extensive audit trail meant to enhance the rigor and trustworthiness of the analysis.Overall, the analysis followed the six steps outlined by Braun and Clarke (2022), though this process is inherently recursive and iterative.
First, I familiarized myself with the dataset, reading and re-reading transcripts, interview notes, observation notes, and programme documents.Throughout, I took memos associated with specific data items and separately compiled thoughts, impressions and reflections related to the entire dataset.
As a second step, I coded the interview transcripts and observation notes.Codes were primarily developed inductively, though they were influenced by existing literature on the conceptualization of social cohesion and the delivery of sport for social cohesion programmes.The codes captured a wide range of semantic (e.g., types of activities delivered by programmes) and latent (e.g., assumptions embedded in certain activities) concepts.The initial coding process generated over 75 codes, and I generated a short definition to accompany each code.However, some of these codes captured minor nuances.As such, before moving to the third step of theme identification, I revised, reviewed and merged codes, to avoid duplication or exceedingly small distinctions.Concretely, that meant reviewing coded segments and code definitions, merging similar codes and adjusting definitions, accordingly.
Once this was finished, I moved on to the process of theme development.Here, themes should be understood as coherent patterns "of shared meaning organised around a central concept" (Braun and Clarke 2022, p. 77).To develop these, I reviewed code excerpts and used in-built visual tools (e.g., code maps, code relations) to explore patterns and connections across the data.In particular, I used the MaxMaps function to draw my thematic maps, organize codes, and identify patterns of meaning.In other words, these maps helped clarify the relationship between codes and between themes (Braun andClarke 2006, 2022;Trainor and Bundon 2021).This process fed into the fourth and fifth steps, as I revised my thematic maps, refined my themes, wrote theme summaries and eventually settled on the themes presented here.Finally, the sixth step involved the actual writing of the results, which are presented in the sections below.As this paper focuses specifically on the understanding and practice of social cohesion, I concentrate on three themes related to the understanding and delivery of social cohesion through sport.

Findings
Based on the process described above, I generated three themes that illustrate how the programmes conceptualize and support social cohesion: "together with appreciation", "the skills to come together . . .for some", and "what you do is up to you".Together, these themes illustrate how the programmes adopt an individual-focused understanding and practice of social cohesion, placing significant responsibility for the development and transfer of social cohesion onto so-called vulnerable groups.In that sense, the notion of an individual-focused social cohesion can be described as the central organizing concept of the analysis, and the themes demonstrate how this individual focus translates to the conceptualization, implementation and, perhaps most crucially, the transfer of social cohesion outcomes.In the below section, the main findings, supporting quotes and related analysis are all presented together.As such, this should be understood as a qualitative report that combines results and discussions, and allows the results to be situated in the context of wider research and theory (Braun and Clarke 2022).

Together with Appreciation
This first theme exemplifies the core ideas embedded in the programmes' understanding of social cohesion, and contributes to answering the first research question.Though academic literature often presents social cohesion as a contested term (Fonseca et al. 2019; Schiefer and van der Noll 2017), the responses here show a high level of consistency.Participants did not always spontaneously bring up social cohesion, and not everyone felt completely secure defining the term.Nonetheless, the answers did coalesce around common areas.For the people here, social cohesion can be summarized as peaceful social relations based on an appreciation of diversity that breeds a sense of identity and networks of mutual support.
"A peaceful, appreciative interaction of several groups of our society, from any dimension of the diversity model, with different backgrounds" (Transcript, James, Management, Germany) "When talking about social cohesion, she mentions networks of people that help each other out" (Interview Notes, Vina, Management, Netherlands) "Social cohesion is about people living together, isn't it?About getting on together, about having areas where they can better themselves, isn't it?And learning from each other."(Transcript, Kelly, Management, Ireland) To some extent, this consistency may even suggest that debates about the meaning of social cohesion have been overinflated by policymakers and academics, and these conceptual debates may not resonate with practitioners on the ground.One interesting contrast with many current definitions (cf.Schiefer and van der Noll 2017) is the general aversion to the term "tolerance".Across these responses, individuals explicitly go beyond the notion of tolerance and instead favor ideas of appreciation or respect.Essentially, many view tolerance as having a negative connotation, that people only tolerate things they inherently do not like them: "I hate this word tolerant, I would say more respectful.Because I think tolerant is a negative" (Transcript, Bernard, Management, Ireland).
Despite this relatively consistent and progressive view of social cohesion, a strong individual or micro focus underpins this understanding.This manifests itself in two ways.First, structural meso or macro level items in other definitions of social cohesion, such as inequality, the perceptions of fairness, social mobility or trust in institutions (Schiefer and van der Noll 2017;Delhey and Dragolov 2016;OECD 2011), are mostly absent.Certainly, some participants connect greater social cohesion with notions of greater equality, but this is not the norm: "it's to stop racism and discrimination and bring equality among us" (Transcript, Alan, Former Participant/Coach, Ireland).
Second, the underlying causes of social cohesion, or a lack thereof, are viewed as primarily the result of individual characteristics.In particular, there are strong ideas about discrimination and prejudice being rooted in individual attitudes or behaviors and that these individual characteristics must be addressed.Thus, these organizations do not explicitly take a systemic view of racism or discrimination, whereby these are not merely surface-level phenomena rooted in ignorance or hate but a system driven by the self-interest of particular groups (Feagin 2013;Kendi 2019).Documents often highlight goals such as "intercultural competence" (Handbook, Germany) or "life skills" (Handbook, Ireland).Within interviews, discrimination or racism are likewise perceived as a matter of individual attitudes, behavior and upbringing: "There needs to be more awareness ( . . . ) Like, it all starts with from home.Yeah, like parents."(Transcript, Alan, Former Participant/Coach, Ireland) "[When] I don't like someone else because it's actually something in me." (Transcript, Tina, Management/Coach, Germany) In other words, these illustrate how the programmes define this core element of social cohesion and reveal how they perceive the root causes of social cohesion.Likewise, other issues, such as a perceived lack of social skills, are viewed as a product of individual environments and not systemic issues.As I will discuss next, this individual-focused understanding of social cohesion and its causes translates directly to the type of activities or mechanisms favored by the programmes.

The Skills to Come Together . . . for Some
As social cohesion is seen as closely connected to social relations and individual attitudes or behaviors within those social relations, programmes consciously provide opportunities for mutual interactions and the development of the perceived skills needed to promote social cohesion.In other words, as the antecedents of social cohesion are conceptualized on an individual level, activities to promote social cohesion are as well.For one, this means establishing regular sporting activities that target various community members and emphasizing an open atmosphere.These activities seek to bring different groups together and allow participants to form bonds across ethnic or geographic divides.For instance, I actively participated in a weekly open football session hosted by the Irish organization, and I also took part in an activity session for parents and children hosted by the Dutch organization.This, of course, is hardly a new approach within the SFD sphere, as the ideas around using sport as a platform for mixed-group interaction are well established (cf.Schulenkorf and Sherry 2021).And, certainly, for some participants, this had an apparent effect on building up their social networks: "that's pretty much what stands out to me, because, you know, just get to make new friends" (Transcript, Pedro, Coach, Ireland).
Alongside these mixed group activities, the organizations implement a variety of modified sports or games that are used to develop a range of life skills, such as communication skills, emotional competences, or intercultural awareness, considered essential for greater social cohesion.The German organization explicitly integrates games and activities meant to develop intercultural or emotional competences, whereas the Irish and Dutch organizations occasionally use modified sport approaches, including football matches based on the football3 methodology (see Fox et al. 2013): "[the project] uses sport and exercise as pedagogical tools to promote the emotional, social and intercultural skills of the participating children" (Handbook, Germany).
Crucially, these activities do not merely focus on the individual level but mostly on only certain individuals.Interviewees or documents variously described the target groups of their programmes as "immigrants", "refugees", "asylum seekers", "vulnerable", or "socially weak".During my observations, I was similarly struck by the absence of majority group participants in activities.As a result, the interactions promoted within the activities tend to be predominantly 'outgroup-outgroup' oriented.Though these programmes rely on intergroup contact, the contact generated does not align with the ingroup-outgroup mix initially suggested by Allport (1954) but instead focuses on bringing minorities from different groups together: "this was actually brought up by one person ( . . . ) they said that like, the team could also be perceived as grouping those that are different together" (Transcript, Amy, Management, Ireland).This is largely by design.All of the organizations have made conscious choices to work in neighborhoods with particular demographic characteristics, or have designed activities meant to target specific groups that are seen as lacking certain skills or opportunities.The Dutch organization may be the most explicit about this.They work in what they consider highly "diverse" neighborhoods (Presentation, Netherlands) and aim to bring disparate ethnic groups together to form bonds and networks of mutual support: "the neighbourhood is diverse, but bigger ethnic groups don't interact with each other, and smaller groups are left out" (Interview Notes, Charlie and Marie, Management, Netherlands).
On the one hand, this outgroup focus reflects the explicit focus on vulnerable or marginalised groups inherent to many European sport or social cohesion policies (Dobbernack 2014; Moustakas 2021).On the other hand, this may reflect an implicit or explicit recognition that focusing on building relationships between diverse groups, especially amongst outgroups, may seem less threatening to the majority population or institutions, as this divorces them from any need to reflect on, engage with, or take responsibility for issues in their communities (Ahmadi 2018;Nixon 2019).
That is not to say that these activities did not generate improved social relations or skill development among participating individuals.Increasing instances of mutual help and emotional communication were noted throughout observations in Germany.For instance, over time, observations noted the children offering help or support to others, be it with tasks such as cleaning up or assisting distressed peers.Nonetheless, by concentrating almost solely on these "vulnerable groups", programmes may reinforce notions that these groups are primarily responsible for any perceived lack of social cohesion and, in turn, discount the role played by privileged ingroups or institutions (Nixon 2019).

What You Do Is up to You
Embedded in social cohesion are notions of a broader community of people than those involved in an organization or programme.The very nature of the term suggests that it manifests itself not only at the individual or small group level but at the meso or macro levels (Fonseca et al. 2019).Thus, how programme outcomes transfer and manifest themselves in the community could be expected to be a key concern for these organizations.After all, the programmes are relatively explicit about having goals at the meso or macro levels, be it to "contribute to more social cohesion" (Presentation, Netherlands) or "challenge discrimination" throughout Ireland (Presentation, Ireland).However, as their understanding of social cohesion is related to a number of individual skills and attitudes, the development or transfer of social cohesion to the broader community is often left in the hands of individual participants.In other words, individual participants move from being the core focus of programme activities to being ostensibly left to their own devices-the programme is, simply put, no longer part of the equation.Despite this, there is an expectation or hope that participants will carry programme outcomes to their wider communities and foster greater social cohesion.Likewise, the programmes do not take sustained action at the advocacy or policy levels to help change the underlying conditions faced by participants.Essentially, the programmes adopt a 'ripple effect' type approach, whereby a focus on individual beneficiaries is expected to spread out to further layers of society (cf.Sugden 2010).
To varying degrees, programmes place responsibility for the development of social cohesion in the community on the shoulders of their participants.For some, there is a more laissez-faire approach where programmes might provide an initial platform to support the development of social relations or skills as a starting point.Afterwards, practitioners hope for further development to be led by participants outside of the programme context: "[The goal is] to connect with youth to let them see the power of sports (. . . ) and hopefully that they host their own sports activities" (Observation Notes, Netherlands) "However, if individuals take those connections or friendships beyond those events is 'up to them'" (Interview Notes, Line, Coach, Ireland) Others are more explicit about placing responsibility directly on the individual participants.In particular, this is seen within the notion that specific skills or behaviors are essential to personal success and greater social cohesion: "if I start with my emotional competencies, the social competencies will grow, for sure.And I think this is the best approach to foster social cohesion" (Transcript, Tina, Coach/Management, Germany).Such statements not only reflect the idea that the causes of social cohesion reside within individual behaviors or skills, but also show how the responsibility for improved social cohesion is implicitly placed on changed individual behaviors.
For many, this focus on individual development and transfer can be connected to the limitations faced by their programmes.For instance, some in the German programme feel conflicted about the school setting.Though this offers regular contact with the children over four years, it is also a very structured and controlled setting that reduces opportunities for informal contact: "in the programme, I often feel that some [participants] don't have enough time or attention" (Focus Group Notes, Annika, Coach, Germany).Funding and related issues in staff turnover are also a recurring topic and are noted as critical limiting factors for the programmes: "you know, is there more that can be done?Definitely.But again, it comes down to how much money you have, how much you can finance into these projects" (Transcript, Line, Coach, Ireland).

Discussion
Through the three themes developed here, I have shown how the programmes adopt an individual-focused definition and practice of social cohesion, placing significant expectations on individuals to develop social cohesion in their communities while not necessarily providing support outside the programme context.The first theme highlights how the organizations have adopted a fairly consistent view of social cohesion that centers on peaceful social relations and appreciation of diversity that promote a sense of identity and networks of mutual support.This understanding likewise highlights how individual factors are viewed as the main source of that social cohesion.The second theme builds on that and shows how programmes implement practices that aim to develop the individual characteristics seen as essential to social cohesion.This theme also exemplifies how they target groups, like migrants or minorities, who are perceived as lacking these characteristics.Finally, the third theme highlights how programmes expect individuals to take responsibility for social cohesion in their broader communities.Flowing from this analysis, there are both theoretical and research implications worth noting.
Theoretically, these themes hint toward a tension between meso or macro-level concepts, such as inequality or trust in institutions, present in academic or policy conceptualizations of social cohesion and those at the programme level (cf.OECD 2011; Schiefer and van der Noll 2017; Fonseca et al. 2019;Council of Europe 2010).This signals that different conceptualizations exist in practice and perhaps gives further credence to the argument that broad definitions of social cohesion conflate the causes of social cohesion with factors that inhibit or promote social cohesion (Chan et al. 2006).Though there were nuances between individual and programme understandings of social cohesion, these differences were relatively minute and indicate that, in practice, the term is not nearly as contested as within the literature.As such, this highlights the need to continue exploring how local practitioners and participants understand and experience social cohesion within their communities, as opposed to simply taking existing top-down definitions for granted.
More broadly, the findings, along with those in the literature documented above (Flensner et al. 2020;Kelly 2011;Sabbe et al. 2020), show a common focus on individuallevel outcomes and processes.Thus, the "vulnerable groups" targeted are expected to support social cohesion independently, although the programmes seldom acknowledge or challenge the systemic discrimination or inequality that may limit that cohesion in the first place.To an extent, as the interviewees recognize, this is driven by funding, material and structural constraints.Paradoxically, the programmes also engage in many activities aimed at working beyond the individual level, even though they recognize these limitations.Namely, the organizations are well-established in their communities, primarily operate outside of the much-critiqued project-based approach (Lindsey 2017), work with a variety of local stakeholders, including higher education institutions and municipal officials, and are developing an increasingly participatory culture through the implementation of a living lab approach.
Despite these efforts, programmes have struggled to develop more structural approaches called for by numerous researchers over the last decade.Many have advocated for structural or transformative approaches that move beyond a micro-level focus to combat the exclusionary mechanisms faced by participants, emphasize social justice and empower individuals to succeed within existing social systems, while also actively working for change (Giulianotti 2011;Hartmann and Kwauk 2011;Sabbe et al. 2021).On a practical level, this can mean having programmes work with groups beyond those deemed "socially vulnerable" populations.As it stands, the constant focus on only marginalised groups reinforces notions that these groups are primarily responsible for their condition (Nixon 2019).Programmes should move away from a practice of social cohesion that singles out supposedly challenging groups for change (Dobbernack 2014;Dierckx et al. 2022).A more structural approach might mean integrating groups associated with privilege or wealth within activities or, as Sanders (2016) contends, taking a more prominent role in the realms of advocacy and policy.
Yet, the convergence of approaches also likely suggests that broader, pan-European systemic factors may be pushing programmes towards similar, micro-level approaches.That means that, as researchers, we must not only thoroughly describe what a structural programme approach could look like but also research, challenge and support systems that can underpin such structural approaches.As a start, that may mean more deeply interrogating existing structures and questioning how sport can challenge or change those structures.In that sense, I very much echo the call from Darnell and Millington (2019), who argue that sport sociology should interrogate and foreground power relations in policy and practice.Along this vein, there are numerous avenues worth pursuing related to sport and social cohesion.Policy and related funding mechanisms often have a narrow focus on pre-defined groups and advocate for primarily neoliberal, technical, and individual-focused mechanisms (Moustakas 2021;Hayhurst 2009).There remains a need to understand how those receiving funding reflect on, adjust to and challenge the assumptions embedded within these funding programmes and associated policies (Hayhurst 2009).Such exploration should not be limited to international or European policy, as has been the case so far (e.g., Hayhurst 2009;Moustakas 2021;Lindsey and Darby 2019), and should look more closely at national and institutional policies and funding mechanisms.
The pedagogy of sport-based social interventions should also be researched and interrogated more closely.Indeed, numerous authors have questioned sport-based social interventions and have proposed more critical curricula or approaches (Spaaij and Jeanes 2013;Meir 2022).Many studies have also explored the development, tensions and power dynamics associated with critical pedagogical approaches (Meir 2022).Yet the content and logic embedded in the growing number of education programmes, online courses (McSweeney et al. 2021) and practitioner manuals (e.g., Scott et al. 2020;Jobse et al. 2019) should likewise be questioned.These educational programmes and materials are increasingly prominent, directly influence the practice of programmes on the ground, and often reproduce neoliberal and individual-focused approaches and understandings.As researchers, we must critically reflect on how such neoliberal approaches have been reproduced at scale through these materials, especially considering that academics are often integrated into developing and delivering such educational content.In short, when we are called upon to contribute to developing policy or educational materials, we must consciously work to combat these neoliberal tendencies and instead focus on concepts of equality, social justice and system change.

Limitations and Conclusions
There are certainly limitations in the above analysis.The voices of participants are largely missing from this research.Yet, their understanding and experience of social cohesion and their programmes are essential to unpack how sport-based social interventions can contribute to social cohesion.Relatedly, the multi-organizational approach used here certainly carries benefits in terms of analysis and comparison, but it inhibits the in-depth immersion inherent to research contained in a single location, especially as it relates to relationships directly with participants.
Despite this, this study supports trends highlighted in existing literature, whereby sport-based programmes take a predominantly individual view on social cohesion and its development.Namely, the programmes focus mainly on components connected to social relations, appreciation of diversity, identity, and mutual help.In turn, this translates to an individual-focused practice of social cohesion, emphasizing individual skills, behaviors and social relations, with the transfer of social cohesion to the broader community left mainly in participants' hands.Though the programmes are taking conscious steps toward more meso or macro-level approaches, numerous systemic factors are likely limiting this transition.
If we want to facilitate a move towards more structural or interventionist approaches to sport-based social interventions, we as researchers must play an active role in questioning, challenging and reshaping the systems that underpin sport-based social interventions.We must also critically (self-)reflect on why the structural approaches we have called for have not materialized.And perhaps most crucially, we cannot limit our work and advocacy to academic circles and must also engage policymakers, practitioners, educators, funders and more.
Funding: Research for this project was supported by the European Commission through an Erasmus+ Project Grant.This paper was further supported by the Internal Research Funds of the German Sport University, grant agreement number L-11-10011-240-031000, and by a Family Support Grant from the German Sport University.

Institutional Review Board Statement:
This study was approved by the ethics committee of the German Sport University under Application Number 134/2020.Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Table 1 .
Overview and description of interview participants.