"I realised it weren't about spending the money. It's about doing something together:" the role of money in a community empowerment initiative and the implications for health and wellbeing.

Community initiatives aiming to reduce health inequalities are increasingly common in health policy. Though diverse many such initiatives aim to support residents of disadvantaged places to exercise greater collective control over decisions/actions that affect their lives - which research suggests is an important determinant of health - and some seek to achieve this by giving residents control over a budget. Informed by theoretical work in which community capabilities for collective control are conceptualised as different forms of power, and applying a relational lens, this paper presents findings on the potential role of money as a mechanism to enhance these capabilities from an on-going evaluation of a major place-based initiative being implemented in 150 neighbourhoods across England:The Big Local (BL). The research involved semi-structured interviews with 116 diverse stakeholders, including residents and participant observation in a diverse sample of 10 BL areas. We took a thematic constant comparative approach to the analysis of data from across the sites. The findings suggest that the money enabled the development of capabilities for collective control in these communities primarily by enhancing connectivity amongst residents and with external stakeholders. However, residents had to engage in significant 'relational work' to achieve these benefits and tensions around the money could hinder communities' 'power to act'. Greater social connectivity has been shown to directly affect individual and population health by increasing social cohesion and reducing loneliness. Additionally, supporting enhanced collective control of residents in these disadvantaged communities has the potential to improve population health and reduce health inequalities.


Introduction
Community empowerment as a route to greater health equity is enshrined in foundational health promotion/public health statements (WHO, 1997;WHO, 1986). Definitions vary but we define community empowerment as processes through which communities of interest or place develop the capabilities they need to exercise greater collective control over decisions and actions impacting on their lives and health. Initiatives aiming to enhance individual or community empowerment are supported by a growing body of research demonstrating that 'control over one's destiny' (Syme, 1989) is a fundamental determinant of health, and lack of control could be a significant cause of health inequalities. Community empowerment is thus now integral to the Global Sustainable Development Goals and many local, national and international strategies for social and health development (e.g. WHO EURO, 2013;UN Economic and Social Council, 2019).
Local place-based initiatives designed to be 'empowering' are diverse but some seek to enhance collective control over decisions and actions by giving community members control over a budget. Informed by theoretical work in which community capabilities for collective https://doi.org/10. 1016/j.socscimed.2020.113176 Received in revised form 2 March 2020; Accepted 25 June 2020 control are conceptualised as different forms of power -'power within'; 'power with' and 'power to' -this paper considers the role of money as a mechanism to enhance these capabilities presenting findings from an on-going evaluation of a major place-based initiative being implemented in 150 neighbourhoods across England.
The paper starts with a brief overview of research on the relationship between collective control and health and on the role of money in community-based initiatives. It then describes the Big Local (BL) initiative and the evaluation design and theoretical frameworks shaping the analysis. The findings are then presented followed by a concluding discussion.

Collective control as a determinant of health and wellbeing
Evidence is accumulating that collective control by communities over decisions and actions impacting on their lives may be a fundamental determinant of population health. Different pathways linking inequities in 'control' to inequities in health have been proposed . At a community level, for example, living in disadvantaged neighbourhoods can produce a heightened sense of collective threat and powerlessness amongst residents; these, acting as chronic stressors, can lead to distress, manifested as anxiety, anger or depressionrecognised as damaging to mental and physical health (Ross, 2011). Obversely, the exercise of collective control could reduce the health impact of disadvantage if, for example, community action successfully prevents the siting of a toxic waste facility or attracted resources that make the environment safer (De Vos et al., 2009;Popay et al., 2007;Popay, 2010). Additionally, experiential knowledge acquired by people living in difficult social and material conditions can help develop more acceptable, and therefore more effective, ways to address the risks to health they face (Wallerstein, 1992(Wallerstein, , 2002Popay and Williams, 1996;Pickin et al., 2002;Morgan and Popay, 2007;Popay, 2010;Whitehead et al., 2016). Positive health effects from collective action may also arise indirectly if participation fosters a greater sense of connectedness, increased social support and reduced alienation within communities, which could lead to improved mental and physical health (Bernard et al., 2007;Popay, 2010;Oakley et al., 1996;Reblin and Uchino, 2008). Individuals who participate in collective action may also benefit from an improved sense of self-efficacy, which research has linked to better health (Whitehead et al., 2014;Zimmerman and Rappaport, 1988). Finally, involvement in collective action may lead to increased political understanding and engagement. This could potentially contribute to democratic renewal and increase public pressure on politicians to deliver more socially just, equitable policies that could in turn address the social determinants of health inequities.
Research testing these pathways has produced a considerable volume of high-quality empirical evidence demonstrating that the level of control individuals have over personal life circumstances is a significant determinant of their health outcomes Marmot et al., 1997;Marmot, 2005;Orton et al., 2019;Woodall et al., 2010). Though more limited, empirical evidence is also accumulating on the impact of enhanced control at the 'collective' or community level. Studies consistently report stronger evidence of impacts on intermediate social determinants of health and health equity than direct impacts on health (Laverack, 2006;Popay et al., 2007;Popay, 2010;Wallerstein, 2002Wallerstein, , 2006Whitehead et al., 2014Whitehead et al., , 2016. In their review, for example, Woodall et al. (2010) found evidence of impacts on social cohesion and trust, but little evidence of direct impacts on health and well-being outcomes at a community level. The review by Whitehead et al. (2014) identified limited but relatively strong observational and ecological evidence linking increased collective control over decisions to better health. Orton et al. (2016) also found limited but good quality RCT evidence on direct health benefits arising from micro-finance interventions that increased collective control amongst women in South Africa, Peru and Bangladesh.
The research briefly reviewed above supports the argument that enabling disadvantaged communities to gain greater collective control over decisions/actions impacting on their lives could contribute to reducing health inequities. However, there is limited understanding of how to design initiatives to successfully support the development of these capabilities. In this context many different 'experiments' are being implemented including initiatives in which money is a key element of the 'theory of change'. These initiatives have primarily been implemented at the individual level as conditional cash transfers (e.g. paying people to ensure children attend clinics and schools or as an incentive to stop smoking) but they can also be found in community and urban development initiatives (Rawlings et al., 2004;Reynolds et al., 2015). For example, participatory budgeting in the English area regeneration programme New Deal for Communities, provided opportunities for 39 disadvantaged communities to have a direct say on how significant amounts of public money were spent (Batty et al., 2010).
Initiatives that go further and aim to give communities full collective control over how money is spent to improve their neighbourhoods are rare, but a few are emerging. Notable in the UK are the Local Conversations programme delivered by the Peoples Health Trust and the Lottery funded Big Local (BL) programme. Through a health equity lens these latter developments pose an important question: in what ways and through what pathways could the transfer of control over how money is spent in disadvantaged communities 'work' to enhance their collective control over decisions and actions that have potential to positively impact on their lives and their health? This paper addresses this question by exploring the role of money in the BL community empowerment initiative in England.

Theoretical frameworks
Two theoretical frameworks have informed the findings reported in this paper. The first, concerns the conceptualisation of the capabilities communities need to exercise collective control. This framework was set out in detail in a paper from the early stages of our longitudinal evaluation  and then applied to qualitative data to explore the power dynamics operating in BL areas Powell et al., 2020) This Emancipatory Power Framework (EPF) utilises the concepts of 'Power Within', 'Power With' and 'Power To', which have their roots in feminist theory (Allen, 1998(Allen, , 2011Arendt, 1970;Rowlands, 1997;Starhawk, 1987). In our framework, the three concepts of power have been adapted from the individual level to the collective. Here, 'Power Within' refers to collective capabilities internal to a community. 'Power With', refers to the power that emerges when a community acts with other agencies and/or communities in the pursuit of shared ends. 'Power To' refers to the exercise of collective control capabilities to achieve desired ends.
Secondly, we drew on Somers' work on 'relational settings' and 'public narratives' (1994) and on Zelizer's concept of 'relational work'the efforts people make in interpersonal relationships -(2012, p149), to examine the settings and relationships involved in the development of capabilities for collective control in BL communities. Somers (1994, p626) defines a relational setting as "a pattern of relationships among institutions, public narratives, and social practices. As such it is a relational matrix, a social network". While 'public narratives' are "those narratives attached to cultural and institutional formations larger than the single individual … [they] range from the narratives of one's family, to those of the workplace (organizational myths), church, government, and nation" (1994, p619).
This relational lens sharpened our focus on how the BL money triggered relationships in particular settings amongst residents and between residents and external institutions, how these relationships were negotiated, and the meaning the money held in these relationships (including the influence of dominant public narratives/stories about previous experience of place-based interventions). Zelizer's (2012), concept of 'relational work'helped to illuminate how BL residents A. Townsend, et al. S o c ia l S c ie n c e & M e d ic in e 2 6 0 ( 2 0 2 0 ) 1 1 3 1 7 6 sought to 'earmark' money, and to identify legitimate ways to use it, as they negotiated existing and new social relations.

The Big Local initiative
Big Local (BL) is a place-based programme in England, launched in 2012 for at least 10 years and funded by the National Lottery Community Fund. Overseen by a national not-for-profit organisation -Local Trust -the programme awarded 150 relatively disadvantaged neighbourhoods just over £1 million each, for residents to decide how to use the money to make their area "an even better place to live" (Local Trust, 2018). The BL areas were selected on the basis that they had historically 'missed out' on Lottery funding. They have considerable flexibility in the design and delivery of local programmes but they are all required to form a resident-led BL Partnership (initially some areas established a pre-partnership steering group of community stakeholders) to oversee the local programme, involve the wider community in developing and delivering the plan; and review progress over time.
Each BL area has professional support through a BL Representative (Rep) and had to identify a 'Locally Trusted Organisation' (LTO) to manage the budget. Many BL Partnerships pay people to undertake specific tasks (e.g. run engagement events and/or manage projects). While not formally required to do so, the resident-led BL Partnerships can (and typically do) engage with local public, private and/or third sector agencies (e.g. National Health Service organisations and local government) to attain their goals (Local Trust, 2018).

The Communities in Control (CiC) study
The CiC study is a multi-site, mixed-methods longitudinal evaluation of BL being conducted by a collaboration of academics around England. It comprises three phases from 2013 to 2021. It is funded by the National Institute for Health Research (NIHR) and the first two phases were conducted within the NIHR School for Public Health Research.
The findings reported here are based on qualitative data generated during phase 1 between March 2014 and November 2015. This phase aimed to: gain an in-depth understanding of early implementation of the local programmes; identify any impacts on the communities' capabilities for collective control; and explore change processes associated with these. Two waves of fieldwork were conducted in 10 areas across England, selected from the 150 BL areas to reflect diversity in geographical spread and local context. Key elements of the latter were population characteristics, urban/rural, contemporary socio-economic conditions and historical trajectory.
The dataset across the ten field-sites included semi-structured faceto-face interviews with 116 residents and other stakeholders (e.g. BL Reps, workers appointed by residents, officers/elected members from local authorities and staff of voluntary organisations). Initial interviews explored a priori issues, such as impetus for BL activities, as well as specific activities/incidents judged to have potential to illuminate the development of collective control amongst residents. Subsequent interviews followed up significant issues emerging during earlier fieldwork. The interviews were audio-recorded and transcribed verbatim. Other data collection methods included: participatory activities (e.g.walkabouts guided by residents); observation of Partnership meetings and informal conversations recorded in fieldnotes and documentary sources (BL Partnership minutes, website material). A mixture of verbal and written informed consent was obtained for all fieldwork. Ethical approval was granted by Lancaster University Research Ethics Committee (February 3 2014).

Data analysis
Interview transcripts were anonymized, entered into Nvivo 10 and thematically coded using a common framework for ease of retrieval and cross-referencing during more focussed analysis. Initial thematic analysis was 'within site', followed by a comparative analysis across sites. The analysis and interpretation were based on a process of review, refinement and group discussion within the research team, with agreement being reached about a set of general propositions in relation to the cross-site data (Yin, 2009). Analytic memos also informed the process enabling researchers to use the full range of data (Charmaz, 2006;Birks et al., 2008). As key themes developed, the research team formed subgroups to analyse and discuss particular themes in more detail.
The 'money' sub-group applied a power lens and a relational lens to their analysis. Once an initial overall story about the 'role of the money' had been developed AT re-read all the interview transcripts, to check the extent to which the 'story' was similar across all the fieldwork sites. The research team also re-visited observational data to increase the rigour of the 'story.

Findings
In the 10 fieldwork sites during these early years of the intervention the £1 million appeared to make a substantive contribution to the development of 'power within' these communities and to their capabilities to exercise 'power with' others. There were, however, situations in which the money constrained the development of these collective control capabilities and/or delayed residents' 'power to act'. Across the sites it was apparent that residents had to engage in significant relational work in order to achieve the benefits control over the money could engender.

Money contributing to the development of 'power within' and 'power with'
From an early stage the money operated as a catalyst for community participation: "We had that money upfront and that was a hook" (A10int-LGO).
The chance to control £1 million nurtured the development of power within these communities by increasing collective confidence in the communities' power to spark change and the connections, skills and knowledge needed to do this. Community events (e.g. festivals and dog shows) built interest and increased knowledge about BL. Connections were made between residents and local organisationslocal authorities; schools and not-for-profit/community organisations to share ideas. The £1 million worked to "help move things along" (A2-int-BLR) prompting a "coming together and drawing up a vision" (A10-int-BLW). Community relationships were newly established and extended. In some areas the local Partnership emerged out of existing groups but in the majority the £1 million brought together a relatively 'new' group of residents to work together for the first time. On all the BL Partnerships the opportunity to have control over the money for local benefit gave residents a focus for change, and excited them to get involved. As one resident Partnership member, a local councilor, commented: "We just talked to everyone … People were really energised by it, they thought: 'Right, we've got this money, we can change this community'" (A6-int-RPM).
In some areas decision-making processes involving control over how the money was to be spent was extended beyond the residents on the BL Partnership: "We had a participatory budgeting event which captured people's A. Townsend, et al. S o c ia l S c ie n c e & M e d ic in e 2 6 0 ( 2 0 2 0 ) 1 1 3 1 7 6 3 imagination … it also gave them [residents] an opportunity to come together and make really quite significant decisions about who got money and who didn't. So … the ball was completely in their court" (A10-int-BLW).
Significantly, there was widespread recognition that the £1million had more than monetary value; as this resident Partnership member highlighted, it fostered connections and collective identity within BL communities: "To me BL isn't about the money … it's not about the million pound is itit could be £10, whatever, it's about getting the community involved and doing something together … I had a few ideas about what to spend it on but then I realised it weren't about spending the money and that's when my ideas started to change. And instead of voicing my ideas as mineit were always about the village for me" (A8-int-RPM).
From the beginning, in all the areas, local organisations in the public and not-for-profit sectors were attracted to the opportunities the £1million opened up. As an organizational stakeholder on the steering group, set up in one area before the resident led BL Partnership was established, commented: "We got involved … as a key organisation in the community [the million] could be really useful … and of course it fits very much with what we want to do here … we want to connect with these people; we want to be part of the development of this part of town and this community"(A4-int-O).
In this context, the money operated as a mechanism for residents to begin to connect with local agencies and increased their capability to exercise 'Power With' these agencies as equals. As this resident Partnership member explained when asked about the role of the money: "It's enormously important … it gives some level of credibility to what we're doing … you can go to people and say 'Will you sponsor this, will you support this?' and they'll go 'Yes … what's it all about?'… 'We've just got a million pounds worth of Lottery funding that has to be spent in the community.' So they can see the benefit" (A10-int-RPM).
Similarly, this paid worker described the assertive way in which their Partnership approached discussions with other agencies: "We want to invest some money. Who else wants to do it with us?" (A10-int-BLW).
Over time the £1 million provided opportunities for residents to further develop their 'Power With' by extending local connections with a wider range of organisations and in new ways (e.g. A2, A4, A7, A8, A9, A10). As the money enhanced the perceived legitimacy of BL Partnerships, Partnership meetings could be a forum to engage professionals, to deliver their plans. In several areas, professionals were invited to formally present their proposals to Partnerships, e.g. a builder for A8's infrastructure project and an environmental worker for A9's green space project.
Through new connections BL residents also acquired new knowledge and skills. As one Rep commented: "The million pounds is, it's for facilitating the community to come together, building their capacity and assets, strengths, and levering in, using that strength to bring in others" (A2-int-BLR). There were numerous examples of BL Partnerships leveraging in matched funding from external agencies across the areas (e.g. A1, A2, A5, A9, A10). These included local government providing professional support with BL Partnerships providing cash (e.g. A1, A10) and a local college in A7, match funding training courses.
These alliances could shift perspectives on where leadership and control should lie: establishing new relationships and/or re-negotiating the balance of power in existing ones. For example, as was highlighted in observational notes, in A1, funding for a multi-use games area had been suggested during a pre-BL consultation between the local government and young people. The BL Partnership supported the project, contributing more than twice the funding that the local government provided, effectively transferring ownership from the council to the community via the BL Partnership. Notable, was how the decision to make such a sizeable contribution helped the Partnership to realise that they could do 'big things.' Up until then, they'd been allocating funds to small projects and community events. This bigger venture released them from focusing only on smaller initiatives. As the resident chair of the BL Partnership commented: "A new play facility, a multi-games area, places for the kids to go, a complete thing, with some money from the local ward councillors, would be a great thing to do" (A1-int-RPM).
There were instances, however, when geographical, cultural and/or social obstacles limited the ability of the money to catalyze new relationships. In some instances physical boundaries inhibited connections, such as where a main road effectively cut a BL area in two and some people did not identify as BL residents (e.g. A2, A10). In other cases issues around identity operated as barriers to connectivity, when for example, some residents did not see themselves as part of a disadvantaged area (e.g. A1, A10) and hence did not see the money as 'for them'.

Tensions over money: constraints on collective control
The growth of 'Power Within' and 'Power With' in these communities was accompanied by challenges that required residents to engage in significant 'relational work' involving negotiated efforts to establish and maintain new and changing relationships and to remove constraints on residents' ability to work with other agencies. These challenges were seen in most areas and were associated with various factors.

Debilitating public narratives: the history of 'failed' place-based initiatives
Shared public narratives of an area shaped meanings around how far communities could have control of the £1 million. In particular, memories of previous money-based initiatives (e.g. A2, A4, A8) and pre-established alliances (e.g. A7, A9) influenced perceptions of BL and could provoke cynicism As one resident noted: "There's money that comes and goes with all these other initiatives that have come and gone over the years" (A2-int-RPM). The BL Rep for A4 similarly reflected on how past failures manifested as current challenges when attempting to forge new relationships between residents, and with outside agencies, which in turn influenced the level of enthusiasm and ultimately the pace of progress: "It's a lot of trying to build the trust locally … having failed so many times in the past and there is a lot of apathy of 'Oh we've heard it all before … and all the money disappeared.' So it's getting over that" (A4int-BLR).

Money distracting from 'genuine' community action
"It's a distraction" (A6-int-O). Participants from several areas felt that the £1million risked distracting residents' from the collective pursuit of common goals. Some paid workers discussed how framing the BL initiative as having £1 million to spend, undermined a community ethos: "Telling people [about] the £1 million … I'm not sure that's a good strategy personally because it's always this thing about money" (A8-int-BLW). Similarly, this BL Rep expresses how, for some stakeholders, the emphasis on the money had led to a relative lack of focus on forming, developing and supporting effective community networks. "A few people are saying to me money is almost a distraction … the million almost needs to be put aside for a bit … we need to look at the community first'" (A4-int-BLR). A view shared by this non-resident local government officer: "Although the money has brought them [community members] together it doesn't necessarily mean it's the right conduit to drive them [residents] forward together because, from my perspective, people have ad i fferent interest if there's money on the table … Sometimes that money is a driver when a group's not quite ready for it. It can sometimes take over what would naturally develop or expand within a group …" (A9-int-LGO).
Though concerns about the potential distraction of the money was 4 more likely to be expressed by non-residents, some residents who had been working to improve their neighbourhood prior to the arrival of BL felt that the £1 million had undermined collective action: "the Partnership … It's not organic. It's artificial … The million pounds is … a red herring …. preventing you doing what you can do" (A8-int-RPM). This resident, described effective, small scale community improvements pre-BL that were undertaken with very limited funding by skilled and experienced residents. Another resident in A3 contrasted BL with the community's recent participation in the production of their Neighbourhood Plan noting work already done, which could continue without the £1 million.

Tensions over how the money should be used
Differing geographies and diverging understandings about legitimate uses for the money could provoke disagreements and risk fragmenting social relationships amongst residents and with external agencies. In some BL areas (e.g. A2, A10) tensions arose when different sub-areas identified competing priorities. In one case, confusion about the boundary of the BL area, which had been extended from one housing estate to include a number of more affluent streets, caused disagreements about who had legitimate claims on the £1 million (A6). In other areas, some participants suggested that BL Partnership members were driven by personal interests or pet projects and questioned particular claims on the money: "There was always that tension … money to be used for … activities that were already started off, like the gardening club, the luncheon club. And they just saw it as a pot of money they might be able to draw on …" (A6-int-RLC).
There were also different opinions in some areas about the legitimacy of investing money in and/or working with local businesses though many residents recognised that economic development was an important aim. For example, residents in A4 were initially very clear that local shopkeepers should not participate in the initiative, pushing them out of the steering group, although this position softened over time.

Constraining residents' ability to work with other agencies
BL was implemented as the budgets of public and third sector agencies were being significantly cut by the policy of austerity introduced by central government after the 2008 financial crisis. In this context, participants in all areas expressed some distrust of the motives of external agencies. One BL Rep for example, reported concerns that the million pounds was attracting some parties "who were blatantly chasing the money" (A4-int-BLR). A Rep in another area expressed these concerns in vivid language: "One of the things with Big Local nationally … is that: 'Oh a million'. The predators move in. You know 'us' in public services who are being cut to ribbons gosh we can have some of that. 'Yeah we'll deliver what you want but it'll cost you £30,000 rather than £3000' (A10-int-BLR).
Uniformly, participants expressed a desire to honour one of the principles underpinning the BL initiative: that the money should not replace local government funding responsibilities. As a BL worker managing community consultation noted: "The responsibility for providing for young people and creating opportunities for them, fits squarely with the local authority and with employers and other organisations" (A1-int-BLW). In some areas BL Partnerships sought to create distance from potential collaborators in order to protect their 'ownership' of the money. For example in A4 negative feelings about the local governments' previous involvement in the area meant that initially at least, there was almost no contact between the Partnership and the local council.
There was evidence that appeared to justify such caution. In A8, BL funds were used to support provision of a youth worker when redundancies happened in local government posts, whilst in A6 the BL Partnership was funding youth provision that had been cut. These circumstances could lead to a complete breakdown of relationships. For example, as reported in observational notes: the Partnership in A10 was negotiating with the local council over a small disused green space. At an informal meeting, they were presented with an invoice for the cost of fencing the area that the residents had agreed to maintain in exchange for the council ensuring it was safe for public use by fencing it. The residents declined to pay the invoice. The project was shelved and the residents were left feeling disappointed and duped and trust was lost.

Doing relational work: negotiating tensions and transforming relationships
Tensions in relationships between groups of residents and with external organisations were evident in all 10 areas. However, as this resident illustrates, there was also a widespread recognition of the need for the relational work required in "establishing, maintaining, negotiating, transforming ... interpersonal relations" (Zelizer, 2012:149).
"You've got to be cautious, and you've got to be accommodating. But you sometimes don't want to be. But you have to work with people … We have had councilors [elected officials] attend meetings … it's generally because they want to suggest where money could be used. And I always feel defensive straightaway. But no, at the back of my heart I do know that yes, work sensibly and use funding properly" (A2-int-RPM).
BL was seen by some to have the potential to mend fractured relationships deeply entrenched over many years in shared public narratives of a place. In A8, a mining community with a history of social cohesion and community activism, which had experienced high levels of job loss in recent decades, a representative from the Locally Trusted Organisation expressed the: "dis-engagement from decision-making over the years … [residents are] very sceptical that it (£1 million) will just get hived off. And that is quite [strong] I think within an established community. The older established communities are sort of very difficult to break …. Big Local's an opportunity to change that …"(A8-int-BLW).
Likewise, in A4, there were suggestions that BL could right the perceived wrongs of the past, by using the money to fund collaborative work between residents and local agencies. In this and other areas the tensions provoked by the money and the subsequent relational work required to resolve them, were seen as an almost inevitable part of the BL process. In A6, for example, a LGO reflected on potentially positive impacts of the tense relational dynamics triggered by the money: "I was a bit worried about the conflict it was creating. I didn't think that was good for people's health and wellbeing. And people feeling exasperated and walking out … I just worried about that from the community engagement perspective and the council's perspective. … But … maybe that's a process they needed to go through … Because … it was very pioneering" (A6-int-LGO).
Some local organisations also understood the need for relational work: to adopt different approaches to negotiate new relationships with BL communities. In A6, for example, a youth charity worker described how his organisation found ways to resist being seen as 'chasing the money' when the award of one million pounds was announced and recognised the shift towards greater community control that the million offered: "We backed off a little bit … I've … re-engaged with it [BL Partnership] in the last six months … because … it was a bit like vultures around a carcass … a million pound … eyes light up … So … we have to fund our work but we don't want to be just like dipping into all different places just to get the money … we want to do things that are benefitting and empowering the community. Which is exactly what this is about" (A6-int-O).

Relational work in BL Partnerships: delaying resident-led decisions and action
BL Partnerships were the local governance spaces with final A. Townsend, et al. S o c ia l S c ie n c e & M e d ic in e 2 6 0 ( 2 0 2 0 ) 1 1 3 1 7 6 5 collective control of how the £1 million was spent. This decisionmaking process required significant relational work amongst Partnership members, a majority of whom were residents, as the money was 'earmarked' for what was considered legitimate purposes. Observational notes showed how, during meetings, all the Partnerships expressed a strong commitment to accountability and responsibility: to be seen to be 'doing the right thing' with the money. But 'getting it right' meant different things to different Partnership members. These conflicting perspectives were apparent in three areas in particular: the governance of the money: the balance between immediate small spends and longer-term larger investment; and the balance between direct and indirect benefits to the community. As we discuss below, navigating the complex terrain between divergent views in these three areas, involved considerable relational work which could make collective decisionmaking processes lengthy, with many areas struggling to meet their initial spending timelines.

Getting the governance right
Participants from several areas described lengthy timeframes between announcement of the £1 million and seeing impacts of its investment, in their communities. Some areas established particularly transparent, but time-consuming processes to demonstrate legitimate decision-making (e.g. A4, A10). In A10 an audit group met regularly, discussed funding applications from community members and reported back to Partnership meetings. Then if no consensus was reached, community members could be requested to submit an amended proposal.
Less commonly, external governance procedures were perceived to create unnecessary delays, as one resident noted: "Every now and again … he (Rep) puts another obstacle in our way … rules and regulations … Sometimes it feels like you've got this money … like a big carrot … and they keep moving it higher … and you have to … jump through that hoop … another hoop … He's like St Peter. And the Big Local are like God …. Because he's like their representative … we've been sitting on this money for the last two years … and nothing's happened yet (A1-int-RPM).
A lack of tangible signs that the money was being spent to benefit communities was a source of discontent amongst some residents, prompting more relational work to manage expectations as this LGO worker highlighted: "It [BL] was sort of sold quite early on as: 'Oh you've got the money you can do what you like.' Well obviously you can't, can you? And that can sometimes be a false expectation for people, then I think, so you have to manage that" (A10-int-LGO).

Getting the balance right: community benefits vis-a-vis spending wisely for sustainability
In several BL Partnerships agreement on specific spending was hard won, despite having broadly shared priority areas. Disagreements often reflected schisms between Partnership members about spending approaches: "I have a number of plans that are costed and ready to go and in my view address the priorities that we identified with the consultation … that approach hasn't gone down that well with some of the others … who want to spend a bit longer talking about things rather than doing anything …" (A3-int-RPM).
In A4 the steering group attempted to balance the need to be seen to be spending the money 'wisely', and the expectations of residents asking why the money was not being spent on tangible benefits. Hence they decided to spend on a high profile project for a 'quick gain'; the painting of a mural on the side of the building where BL meetings took place.
For most BL Partnerships working with external agencies (power with) was the key to 'sustainable' spending: "We're not spending the money as quickly as probably expected and it's because we're looking for who else wants to work with us, who else wants to invest" (A10-int-BLW). But the relational work required to build and maintain optimum relationships with local agencies was time consuming. For example, one ex-chair of a Partnership expressed ambivalence around working with the local council to ensure long term gains but recognised the need to do so and highlighted communication as key: "A million pounds is not a lot of money stretched over 10 years but if we know what the council's plans are or we can have an influence on what plans the council put into place … then in terms of long term plans we might be able to achieve a lot more. But at the same time also keeping in mind that we don't want the council to think BL is going to replace anything that they're going to withdraw. So, I think it's so important to have the communication" (A6-int-RPM).
A participant from the same area highlighted how maximising impact was dependent on using the £1 million creatively on structures and processes that supported sustainability: "It's not a lot of money … it's got its own logic to it … you start something and then it creates more and more and more activity. Because a million pounds isn't a lot, so it's got to be about creating an ethos and a structure that allows things to keep going …. " (A6-int-RLC).

Getting the balance right: direct and indirect benefits
There were mixed views about using some of the £1 million for dayto-day running of BL as opposed to projects with direct benefits for the community. Some areas hired professional expertise early, ensuring ongoing support for their work (e.g. A1, A7, A8, A9). In A1 Partnership members saw specialist support as an investment and commissioned a community development organisation to help them design and deliver the initial community consultation; a youth work organisation to consult with young people and paid for a local government officer one day per week to co-ordinate BL activities. In contrast, A5 were reluctant to finance anything not considered to be directly beneficial for the community, while in A10 resident members of the BL partnership volunteered to undertake everyday tasks and administrative duties as a cost saving exercise. However, this created problems. As one resident explained, she had left the Partnership because their reluctance to pay for professional support had placed an unacceptable burden on volunteers.
Over time more areas recognised that 'buying-in' professional help would extend administrative capacity, sustain day-to-day management and reduce the volunteering burden on residents (e.g. A2, A6, A7, A8, A10). But employing workers brought its own challenges. In A4, observations showed that the paid worker found it impossible to manage conflicting priorities amongst steering group members and resigned. In general, areas appeared to be more likely to pay for support if expenditure was perceived as an investment and route to sustainability. This was very clearly expressed as a priority in some areas. For example, as observational data showed, in A10 money was used to hire professionals (in the short term) to support resident volunteers to set up a job club. The aim was, that the residents would gain experience and skills, to equip them with the ability to run the job club themselves, while also laying the groundwork to ensure the initiative lasted beyond the 10 years of funding.

Discussion
There is a long-standing debate globally regarding the relative merits of programmes that target resources at issues identified by funders (e.g. Brazil's Bolsa Familia conditional cash transfer programme (Shei et al., 2014) and those that give communities of interest and/or place some measure of control over how funds are spent to address local needs (e.g. the EU Community-led local development approach to fund allocation, European Commission, 2018). A recent review of health inequities in England (Marmot et.al. 2020:10) concludes that when community approaches are empowering they can be "central to efforts to reduce health inequalities" increasing collective control which has a A. Townsend, et al. S o c ia l S c ie n c e & M e d ic in e 2 6 0 ( 2 0 2 0 ) 1 1 3 1 7 6 "positive influence on health." (2020:139). However, evidence on the relative effectiveness of giving disadvantaged communities influence over how funds are to be invested to improve their lives has been argued to be "incomplete and results are open to interpretation" (Van Domelen, 2007, pii). More recently, Reynolds et al. (2015, p1) has shown there is considerable diversity in the type and extent of influence over resources communities are given in policy initiatives and there is very little evidence on the precise role of control over money in pathways to positive benefits. The findings reported here add to this limited evidence base illuminating how giving control over money to communities bearing the brunt of social inequities can operate to support the development of the capabilitiesunderstood as different forms of powerthey require to exercise greater collective control over the social determinants of health and hence act as a potential mechanism to reduce health inequities. Our findings illustrate how the £1 million given to these 10 BL areas acted as a catalyst in reshaping, rebalancing and extending relationships amongst residents and between residents and local agencies.
As BL residents came together to identify common concerns and interests and share knowledge and skills, they gained greater confidence in their ability to act collectively so their 'power within' grew. Controlling how the £1 million was to be spent also provided credibility to resident-led BL Partnerships, enhancing their capability to develop 'power with' others so encouraging them to enter into, build on and negotiate relationships with external agencies, sometimes shifting the power balance. In all areas growing 'power within' and 'power with' was associated with greater 'power to' act, as residents became more assertive about taking control over how money was to be spent.
As Reynolds and colleagues note, however: "'community' cannot be interpreted merely as a setting or recipient of such an intervention, but something constructed and negotiated through the flow of money itself'" ( Reynolds et al, 2015, p88). The social connections the money drove amongst residents and with local agencies, and the positive impacts these had on BL communities' 'capabilities' for collective control, did not come easily. In all areas the £1million created tensions, which were often significant, and BL residents had to engage in complex and often time consuming relational work to overcome these. Like Cornish and Ghosh (2007) revealed in their community led project, participating required work to change relationships between the community and more powerful external agencies (p496). Our findings highlight how: BL communities balanced caution with accommodation when negotiating with cash strapped local councils; learnt to re-build trust in previously fractured relationships and attempted to re-calibrate wellestablished divisions of control and power. The role of trust in initiatives has been emphasised by others (Cornwall, 2008), holding symbolic value (Renedo and Marston, 2015) that influences the dynamics and outcomes of community participation. This relational work involved residents "establishing, maintaining, negotiating, transforming, and terminating interpersonal relations" (Zelizer, 2012,p149), to ensure that the money operated effectively. Resonating with Campbell and Cornish (2010), who recognised relationship-building as key for community mobilisation, applying a relational lens to our data, revealed how the precise nature of relational work was shaped by diverse relational settings -the 'pattern of relationships among institutions, public narratives, and social practices' (Somers, 1994,p626) -operating within and across these 10 areas. Key properties of these settings included the peculiarities of local geographical boundaries, the diversity of cultural understandings about legitimate uses for the money, negative public narratives about previous community initiatives and significant reductions in public expenditure on local services resulting from central government's austerity policies.
Our findings also reveal ambivalence in the relational work undertaken. Residents saw opportunities to forge new relationships with external agencies albeit recognising the risks. On the one hand, influenced by dominant public narratives, they were wary of working with agencies that were perceived to have 'behaved badly' in the past or which they considered to be desperate for funding to continue to deliver services. On the other hand residents recognised that working with others would ultimately increase the impact and sustainability of the £1 million. Negotiating new ways of working together meant residents risked being (or feeling) duped, and members of organisations risked being seen as disingenuous 'vultures', attracted by the money. In this context, both residents and the staff of local organisations needed to negotiate to re-establish trust and (re) build viable and meaningful relationships in particular settings. However, in some circumstances resident-led Partnerships felt they had to protect the money from other parties (e.g. councils with cuts to budgets). In these situations residents exercised their 'power to' withdraw from negotiations -to shelve some projects (e.g. A10 failed negotiations for the green space) -giving up some opportunities to exercise 'power with' in the short term.

Study limitations
The findings draw on data generated (2014-2015 in a diversity sample of 10 BL areas, in order to gain an understanding of the first three years of this 10 + year initiative. Our qualitative approach allowed a detailed investigation, across these areas, generating an extensive data-set that provided insights into the role of the money in these early stages of the programme. Our analytical strategy was to present the results across areas, while attending to any divergent themes by making constant comparisons between areas. Though there were some differences associated with local context, the relational dynamics identified were present in all ten areas. We cannot say, at this stage, how 'representative' these areas are of all 150 BL neighbourhoods. In later phases of the study we have conducted indepth fieldwork in an additional five areas and are looking explicitly for areas which diverge from the general patterns described here. Additionally, our research reports a snapshot early in a 10 year plus initiative. We are currently tracing the role of the money over the longer term to investigate how the relational dynamcis identified evolve over time and how the role of the money changes. As in all qualitative research, our engagement with participants in the fieldwork sites and their knowledge of our research may have influenced responses during interviews; residents may have been sensitized to 'progress' relating to expenditure timelines; other stakeholders may have prioritised the importance of collaborating with residents. The extensive observational work provides a measure of triangulation.

Conclusion
Community-led approaches to delivering social and/or health improvements are increasingly common in public health and in other policy fields. Whilst few of these initiatives would give residents complete control over a substantial sum of money, as does the BL programme, many involve the transfer to community members of greater collective influence over how resources or assets (financial and otherwise) are used to improve the conditions in which they live. The findings presented here have implications for the design of these initiatives that will help maximise the positive impact (and reduce the risk of negative impacts) of much more modest money/asset based community initiatives.
Whilst we do not know whether the amount of money was significant, our findings suggest that giving communities 'complete' collective control of budgetsno matter the size-could still be impactful as positive benefits derive from both the symbolic and the purchasing value of money. They also point to the need for local initiatives to understand and plan for the scale and nature of the relational work involved in achieving positive benefits and how this may vary across the relational setting in which such initiatives are to be implementedto the history of the area and previous area-based initiatives, to the nature and quality of existing relationships amongst residents and with A. Townsend, et al. S o c ia l S c ie n c e & M e d ic in e 2 6 0 ( 2 0 2 0 ) 1 1 3 1 7 6 external agencies as well as to the impact on these relationships of the wider political or policy agenda. Integrating these understandings into the design of community-based initiatives will increase their potential to improve population health and reduce health inequalities.