Landscape and Urban Planning Place-keeping in action: Evaluating the capacity of green space partnerships in England

nine cross-sector partnerships in Shefﬁeld, Hackney, and Stockton-on-Tees. a qualitative research approach, a framework for partnership capacity based on interrelated factors, including capital, commitment, skill base, motivation, communication and political inﬂuence. The ﬁndings that partnership capacity goes beyond these it can be inﬂuenced by the political and historical legacy of a given place and the speciﬁc nature and context of place-keeping tasks. While ﬁndings show that partnerships work positively in practice, there are a number of barriers to community groups managing green spaces independently of local authorities, occurring at different scales including individual, group, partnership and the wider context. Without sustained resources and ongoing public sector support, the effectiveness of place-keeping partnerships is called into question.


Introduction
Over the last couple of decades, community involvement in green space provision, design, management and decision-making has risen up the political agenda. This is illustrated in the ongoing shift from (local) government green space management to a community empowerment through a decentralised redistribution of power; the encouragement of greater social responsibility; and, the creation of an enabling and accountable state. This marked a shift from top-down micro-management to a flexible and locally responsive approach, made manifest through the introduction of the 2011 Localism Act. Through Area Panels and Community Assemblies (the lowest rungs of government), the Act provided significant community rights regarding government expenditure on local service provision and delivery, including budget allocations for parks. However, since 2011 the responsibilities associated with these rights have become unclear as (top-down) central government-led local authority budget cuts continue, including the abolition of Area Panels, Community Assemblies and ongoing reductions in park staff numbers.
The responsibility for parks, which were often land bequeathed to a town/city from original landowners and/or philanthropists, and other green spaces mostly lies with local authorities in England (Conway, 1991). As a non-statutory service, funding for parks and green spaces has long been adversely hit by budget cuts, and increasingly maintenance services are contracted out to nonpublic sector organisations in efforts to reduce costs. Alongside this fragility of funding is a historically strong and active involvement of communities and non-state actors in green space management (Jones, 2002). With political will driving forward distinctly local agendas, but without accompanying funding, the need to understand how public-community green space partnerships function in practice is timely and relevant. In this way, this paper takes a practice-oriented approach to understanding green space partnerships 'in action' in three parts of England. Generally speaking, green space practices tend to be state-dominated as the local authority has responsibility as landowner and/or manager with funding primarily from local and national taxation. Decisions about how general revenue budgets are allocated are taken by local councillors (CABE Space, 2006). The paper aims to understand better the capacity of cross-sector partnerships within this wider policy context by applying a conceptual framework of partnership capacity in relation to long-term and sustained green space management, or 'place-keeping' which builds on existing empirical research (Dempsey, Smith, & Burton, 2014). The paper will apply this practical framework to a number of existing cross-sector partnerships in nine green spaces to examine the nature and extent of their capacity for place-keeping. This will involve a qualitative exploration of the extent to which stakeholders within partnerships can withstand and undertake specific responsibilities and the ensuing challenges (after Macmillan & Townsend, 2006). The paper provides a timely examination of place-keeping in practice which can help professionals and academics understand the challenges faced by partnerships on the ground. This paper builds on work conducted by the EU-funded project MP4: Making Places Profitable, Public and Private Open Spaces (2008)(2009)(2010)(2011)(2012)(2013) which explored examples of place-keeping in northern Europe . The project aimed to establish and examine the overlapping dimensions of place-keeping, which are partnership, governance, funding, evaluation, policy, and design/maintenance. It became clear that more investigation was needed of these dimensions in specific political, social, economic and environmental settings, which was outside the scope of MP4. This paper is therefore a pilot study focusing on partnerships in a small number of sites in England.

Exploring partnership capacity
Partnership in place-keeping describes an association of two or more partners with shared responsibility for the long-term management of a place (Barnes et al., 2008;Burton & Mathers, 2014).
Partnerships may be informal, based on a mutual understanding of roles and responsibilities, or formal, based on written agreements and contracts. According to Burton and Mathers (2014), partnership is a contested term as in practice partnerships may not demonstrate genuine working together, but be 'little more than rhetoric' (Carnwell & Carson, 2008, p. 4). Partnership is related to the concept of governance which supposes that government does not work in isolation but through relations with civil society and non-governmental sectors, including the community. Governance in place-keeping describes these interactions, defining their roles and responsibilities in relation to the management of a place . It has already been highlighted that political interventions (such as the Localism Bill) directly relate partnership capacity together with voluntarism as a mechanism for effective partnerships. This has led to a marked and active engagement with the voluntary sector. According to Milligan and Conradson (2006, p. 2), this is increasingly viewed by the state as 'an attractive intermediate organisational form in relation to the somewhat tired state-market dichotomy'. The growing interest in volunteering has been underpinned by 'the debates around active citizenship, governance and neocommunitarism' (Osborne, 2010, p. 3). Of particular interest in this paper is how citizenship can be oriented around place, and how and why such citizenship develops more strongly in some places than others, contributing to potential differences in capacity. It has been claimed that voluntary organisations and community groups often cannot function wholly independently of the local state (Milligan & Conradson, 2006). It is also argued that 'community participation in public service provision is not necessarily an emancipatory claiming of rights by citizens' but is rather a process passing on 'state responsibilities to civil society' (Rosol, 2012, p. 240). In this way, it is important to be mindful of the political context within which place-keeping is occurring.
The formation of a community group may be influenced by the extent of deprivation in a given neighbourhood. Chanan argues that 'disadvantage impedes participation' indicating that those living in deprived areas begin at a weaker position in comparison to residents in other areas (2003, p. 6). This may manifest itself as imbalances where certain (e.g. middle-class) groups can better act on their needs and communicate their demands (Rosol, 2012), perhaps with easier access to funding and (political) support. Stipulations in urban regeneration (or place-making; Dempsey & Burton, 2012) programmes in deprived areas, often require engagement and participation of residents in decision-making processes, to help strengthen social capital (Carpenter, 2006). In her examination of community gardens in Berlin, Rosol claims that "the starting point of the new interest in volunteering is the lack of funding for the parks maintenance" (Rosol, 2010, p. 557) illustrating how a lack of funding and attempts to mobilise communities can go hand in hand. Community engagement requirements in deprived areas (where social capital may be weak) are often managed as part of structured intervention from local authority-led partnerships (Chanan, 2003) -e.g. England's New Deal for Communities programme at the turn of the millennium. This highlights specific contextual characteristics for this paper in relation to spatial and inequitable disparities in funding, supporting governance structures and policy implementation (after Rosol, 2012). So for example, while an individual's propensity to volunteer can underpin the extent of wider community participation, the nature of the latter will also depend on the wider socio-economic and socio-demographic context. Insofar as is possible, we will explore the extent to which context has a bearing on partnership capacity in place-keeping in practice. For the purposes of this research, we define partnership capacity as the degree to which cross-sector partnerships are able to develop and deliver its aims and to withstand and respond to internal and external changes affecting place-keeping in practice. While the focus here is on the capacity of partnerships to carry out place-keeping in practice, it should also be noted that our conceptualisation of capacity may have resonance for partnerships engaged in other kinds of activities, not necessarily oriented around green and open spaces.

Emerging themes of partnership capacity in place-keeping
This section outlines the emerging partnership capacity themes which emerged from empirical research project findings (MP4) and form the analytical framework for this paper (Table 1).
MP4 findings reflected the importance of partnerships in place-keeping activities  and identified a move away from the traditional state-centred model approach to place-keeping, where local authorities deliver place-keeping with minimal input from others, towards alternative approaches. This includes a shift towards user-centred models such as partnerships between local authorities, communities and charitable trusts where members share interests in the quality of the space for users, and market-centred models (e.g. public-private partnerships) where partners' motivation is largely profit-driven (Dempsey & Burton, 2012). Whilst it is suggested that different combinations of the three models could deliver effective open space management (De Magalhães & Carmona, 2009), MP4 demonstrated that partnerships which involve community partners can bring benefits that individual partners alone cannot achieve (Burton & Mathers, 2014), some of which are discussed below.
Place-keeping partnerships can attract additional capital to the place-keeping of a green space through fundraising through events organisation and gaining access to funding streams unavailable to local authorities. These are likely to become increasingly important in light of shrinking green space local authority budgets. It has already been highlighted that state funding is often directed to deprived neighbourhoods through discrete initiatives with budget requirements (Burgess, Hall, Mawson, & Pearce, 2001). This can have mixed effects on partnership capacity in terms of performance over time and legacy (after Carpenter, 2006;Skidmore, Bound, & Lownsbrough, 2006). Furthermore, funding sought for place-keeping is often oriented towards place-making activities, i.e. involving the creation or 're-making' of a place, for example a new playground or heritage restoration project. Securing revenue funding for long-term management is more difficult, putting pressure on partnerships to engage in place-making, rather than placekeeping, activities (Kreutz et al., 2014). This is further compounded by dwindling state funding provision (certainly in England) for The motivational drive of the partner's involvement in the site in terms of social, environmental or political interest.

Skill base
The skill base of the partner in terms of relevant training, professional background, familiarity with the site that contributes to the development and management of the site.

Communication
The communication capability of the partner (both internally and externally) in facilitating the development and management of the site. Political influence The degree and level of influence each partner has in facilitating the development and management of the site.
place-keeping which may have a detrimental impact on the capacity and sustainability of a partnership. Another emergent issue from the findings was how developing effective place-keeping partnerships requires time, resources and long-term commitment from partners. This may manifest itself as paid or unpaid commitment, through volunteering, financial contributions to local organisations (also an example of capital) and/or knowledge sharing (De Magalhães & Carmona, 2009). This might also be manifested formally, through the signed constitution of a partnership or letters of agreement (Burton & Mathers, 2014;Matsudo, 2012). Other examples of commitment can include attendance at meetings and events, and volunteers 'signing up' to collectively derived aims and objectives by taking part in the partnership's activities (Burton & Mathers, 2014).
For such participation to work, there needs to be underlying individual and collective motivation (Mannarini et al., 2010), which may stem from specific issues of interest and/or locality of interest (Barnes et al., 2008). Partner members may not represent all green space users and have different agendas: understanding the nature and existence of motivation is important to identify the expectations held by a partnership, of themselves as individuals/organisations and of other partners. It is often perceived that place-keeping is the local authority's responsibility, which can lead to concerns about liabilities for the open space come with transfer of place-keeping responsibilities from local authorities to community groups; it may also engender a lack of motivation and unwillingness of communities to stay involved beyond the initial, place-making project stage (Burton & Mathers, 2014).
Community involvement permits the local experience of green space to be revealed, unlikely to be uncovered by professionals in isolation, which adds richness to the skill base of place-keeping. The prevalent user-centred model in England puts emphasis on a horizontal, rather than hierarchical, approach permitting formal and informal networks to make use of local knowledge (Mathers, Dempsey, & Burton, 2012). De Magalhães and Carmona (2009) argue that such knowledge and skills are needed if the local context, needs and demands are to be understood fully with a view to the co-production of, here, place-keeping agendas. Bound et al. highlight that such skills can extend to an ability to foster participation particularly where there is 'poor community infrastructure' which is a 'limiting factor in participation' (Bound et al., 2005, p. 7).
Fundamental to all of this is communication, concluded by MP4 researchers to be the linchpin of effective place-keeping partnerships . The context within which a partnership is set and tools of communication to which partners have access may help or hinder effective involvement in place-keeping, and directly affect people's capacity to bring local knowledge and experience to place-keeping process (Van Herzele & Van Woerkum, 2008). Communication between and among partners is therefore crucial to ensure that all aspects of partnership capacity are engaged in by all partners (after Dekker & Van Kempen, 2004). Effective two-way communication is required to inform and engage partners and community members to ensure, for example, early and close involvement in shaping well-informed and suitably context-specific agendas (Bound et al., 2005). The challenges of good communication relate to its nature and direction: one-way and informative communication is insufficient (after Van Marissing et al., 2006) and more inclusive citizen input is required as part of the shift from authority-led decision-making (Castell, 2010).
A final benefit of partnerships identified in the MP4 research was a shift from focus on problems to developing a shared vision for action (Burton & Mathers, 2014). Part of this could be attributed to the influence that partners may have in mobilising political influence. Arnstein's ladder of participation (1969) brings communication together with political influence: moving up the ladder reflects an increase in citizen power in relation, here, to placekeeping decision-making. Thinking more broadly about influence within a partnership, Holt et al. explored the governance of groups involved in urban river corridor regeneration. Findings revealed that while some groups effectively achieved their aims (due to access to local networks and multiple skillsets), they did not represent the local community, with tendencies of self-selection and closed membership (Holt et al., 2012). Achieving true representation is an issue discussed elsewhere: for example, studies into Finnish collaborative urban forestry planning (Sipilä & Tyrväinen, 2005) found numbers of participants in partnership were often limited with agendas driven by a minority of vocal individuals rather than collective decision-making.

Methodology: site selection
To explore cross-sector partnership capacity in this study, a number of green space sites in England were selected from three different urban areas: the city of Sheffield, the market town of Stockton-on-Tees and the London borough of Hackney. These sites were selected in part through contacts made through MP4: two of these sites (Firth Park and Sheaf Valley Park) were case studies in the MP4 project, but the capacity of partnerships was not examined. The other sites were identified through MP4 dissemination events and contacts made through non-academic project partners. Partnership capacity had therefore not been examined in any of these sites before this study. The main site selection criterion was being an urban green space of any size with an associated cross-sector partnership. By urban green space here we mean parks, gardens and green space with amenity, recreational and leisure opportunities (after DCLG, 2002).
In Sheffield, four green space sites were selected (Table 2) in differing geographic and demographic areas, whose sites provide contrasts of scale, nature and establishment. The Friends of Firth Park (FP) have been centrally involved in the park's regeneration since their formation in 1999. FP is a Victorian park where the group works in partnership with the local authority in applying for funding bids which has been instrumental in securing Green Flag status, the national benchmark of a good quality park (Greenhalgh & Parsons, 2004). Sheaf Valley Park (SVP) is a large, transitional green space in the centre of Sheffield which was regenerated approx. 5 years ago. The small Friends Group that developed from the regeneration process has limited the scope of activities to date but actively contributed to the place-making process of urban regeneration. The Friends of the Porter Valley (PV) group was constituted in 1995, has over 470 members, with focus on river corridor regeneration including improvement of natural, historic and archaeological features, and work in partnership with many local organisations. The Friends of Millhouses Park (MP) were constituted in 1991 and have attracted more funding than other Sheffield Friends Groups of parks in their fundraising activities as an independent charitable organisation. In partnership with the local authority, they have transformed a number of areas, creating a sensory garden, water play area, fish pass and tree avenue.
In Hackney, three green space sites were selected (Table 3). Clissold Park (CP) User Group was established in 1994 and has over 110 contacts on their current mailing list. Working in partnership with Hackney Parks Department helped them secure substantial funding to restore Grade II-listed Clissold House and make improvements to the play facilities within this large-scale Victorian Park. The activities of Clapton Square (CS) User Group focus on generating a safe and positive environment. In the late 1990s, the Georgian square was a 'no go area' according to CS due to anti-social behaviour, inspiring the local community to act and improvement the environment, celebrated with a launch event in 1996. Renewed anti-social behaviour in recent years has stimulated a new period of group  Active members are older with gender mix but ethnic minority groups not well-represented Active members are 40-50; younger and more members needed Note: The management responsibility for these sites as public open spaces is held by the local authority as landowner but, as is happening across the country, certain tasks (such as tree management on highways in Sheffield) have been 'contracted out' to non-public sector partners. A partnership agreement was being drawn up between Hackney Borough Council and Robin Hood Community Gardens "because the group are very physically active" (HGSM1). In Stockton, two green space sites were selected (Table 4). The Ropner Park (RP) Friends Group developed in 2002 as part of the funding process for park regeneration. The group's activities focus on increasing the number of park users, events organisation, running the park's café as a not-for-profit enterprise and raising funds for a rose garden. Newham Grange Park (NGP) had deteriorated since the 1980s and the Friends Group emerged in the 2000s with the aim of regenerating the park. Regeneration began in 2004 and the group have been involved in masterplan development with the local authority as well as undertaking practical work. Current activities include events organisation, practical maintenance and seeking out sponsorship.
The nine sites vary in size of green spaces in their respective settlements. The membership of the Friends Groups varies but all have very small numbers of active members in relation to the resident population served by the particular green space. In relation to the make up of all the Groups, there is missing representation of the local community, namely young people (particularly male) and ethnic minority groups. In terms of deprivation, Sheffield's FP is in the 10% most deprived neighbourhoods in England (Rae, 2011) with a deprivation score of 44.44 -where 87.8 indicates the highest level of deprivation for a neighbourhood and 0.53 is the lowest (measuring deprivation in all 32,482 English neighbourhood areas (DCLG, 2010)). SVP is in the 30% most deprived neighbourhoods in England (deprivation score of 27.27) while the other two sites are much less deprived (with deprivation scores for PV between 3.42 and 11.40, and 6.85 for MP). The three sites in Hackney all fall within the 20% most deprived neighbourhoods in England with deprivation scores of 42.02 and 44.30, with CS in the 10% most deprived with deprivation score of 51.13. In Stockton-on-Tees, NGP is located in an area within the 44% most overall deprived wards nationally (Stockton Borough Council, 2011) whereas RP is less deprived (with a deprivation score of 9.23 compared to NPG's score of 15.10). Active members are over 60 and female.
Note: The management responsibility for these sites as public open spaces is held by the local authority as landowner but, as is happening across the country, certain tasks (such as tree management on highways in Sheffield) have been 'contracted out' to non-public sector partners. In Stockton, there is a transfer policy being implemented to transfer the management of parks to the community and let them run them. According to StGSM1, this is "full with complications". 3.1. Contextual information for study sites 3.1.1. Sheffield, South Yorkshire The city has had its 20-year Green and Open Spaces Strategy in place since (Sheffield City Council, 2009). The council's Parks & Countryside Department is responsible for Sheffield's public green spaces and employed seven green space managers (GSMs) at the time of this study who were linked geographically via the (now defunct) Community Assembly (CA) structure. The seven CAs covered different areas of Sheffield, and had some decision-making powers in local areas and limited budgets to spend in the community. GSMs are responsible for capital project delivery and green space management and work with local stakeholders. There are currently over 80 Friends Groups registered with Sheffield City Council undertaking a range of activities including site improvement, practical conservation, fundraising and events (Sheffield Community Partnerships, 2011). Close working partnerships have developed with community groups. However, differentiation in prioritisation of green spaces across city, budget cuts and devolution of set budgets to areas (not based on social indicators) has led to an inequality of support from GSMs as some areas are not as highly prioritised by the local community (discussed in Section 4.1).

Hackney, London
There are sixty-three green spaces in Hackney, overseen by the Parks Department team split into grounds maintenance (two GSMs) and development (three GSMs). The development team is involved in improvement of sites and community engagement. The 2012 council restructure brought the department under a new directorate, with parks playing a key role in the council's 'Health and Wellbeing' strategy and the subsequent focus on health, volunteering and youth. Hackney has had a parks strategy entitled 'Social Spaces' since 2008 (London Borough of Hackney, 2008) and there are currently twenty-seven Friends Groups.

Stockton-on-Tees
There are thirty-seven parks and green spaces in Stocktonon-Tees including nature reserves. The town has had a green infrastructure strategy since 2008 (Stockton Borough Council, 2011). There is also a community engagement strategy which has emerged from the current government's Big Society agenda underpinning the council's activities across the borough and in local communities (Stockton Borough Council, 2010). One way in which these two strategies have dovetailed is through the creation of local nature reserves which involved the establishment of associated Friends Groups as part of a Stockton-wide network. However, the central network no longer functions, but local groups continue to operate.

Data collection methods
To explore the six capacity themes in the sites, the research team conducted semi-structured interviews in 2011-2012 with a total of seventeen representatives from two stakeholder categories involved in place-keeping. These were firstly the local authority, as land owner/manager of the relevant green space. The same manager was often responsible for more than one site, and we therefore interviewed six managers: two in London (HGSM1&2), three in Sheffield (SGSM1-3) and one in Stockton (StGSM1). Secondly we interviewed the Friends Group as a community group with specific focus on green space management -one representative from each group was interviewed, except Millhouses and Firth Parks where two were interviewed (totalling eleven). The six capacity themes underpinned the questions asked in the interviews which were designed to allow interviewees to reflect on the capacity of their own partnership as a whole. The researchers incorporated interviewees' own assessment of the capacity of the specific partnership. This naturally has its limitations. Firstly, it was not possible to interview all members of the partnership -for this reason, we asked the green space managers (the generic job description given for the local authority representatives to retain their anonymity) to reflect more broadly on both the partnership and partners involved to allow for an overview of all the partners involved (albeit a limited one). Secondly, it was outside the project scope to gauge opinions from stakeholders (e.g. other community groups) outside the partnership to assess its capacity; and, thirdly, the interviewees worked in very different organisational structures with inherently very different kinds of capacity (e.g. available capital for the Friends Group may be on a much smaller (and less secure) scale than for the council). In this way, exploring the individuals' understanding of capacity was an underlying part of the process alongside developing their assessment of the partnership's capacity. The interviews were semi-structured in nature to allow for flexibility in the order of questions asked (Bryman, 2004) and the broad headline questions are listed in Appendices 1 and 2. Probing questions were asked and varied from interview to interview. Each interview was conducted in compliance with the University's ethics review procedure, lasted around sixty minutes and was audio-recorded (with interviewee permission) to allow the conversation to flow unimpeded. This resulted in a large amount of data analysed using content analysis including coding and comparing and contrasting responses to specific questions by GSMs and Friends Groups, Such analysis methods were employed as objective and systematic techniques to minimise the personal bias of the researchers (after Robson, 2011).
As highlighted in Section 2, the capacity themes were identified from an existing conceptualisation of partnerships. This was designed to provide a starting point to examine in detail how well these themes captured issues around capacity. An approach was taken so as not to lead interviewees to focus on these themes alone, but rather to allow them to discuss capacity according to their own understandings -and to permit any discussion of other factors of partnership capacity. Indeed, the emerging findings corresponded with, and went beyond, the six themes identified in the MP4 research. A limitation of this approach was that characteristics of partnerships, and capacity, cannot be fully understood by studying the self-reported behaviour of a specific set of participants, as identified by the researcher, not the participants. This links to issues of scale and the limitations of examining partnership capacity within a spatial boundary (e.g. a park) (Connolly et al., 2014) particularly in light of Bodin and Crona's (2009) findings that stakeholders in networks interact at different scales beyond the local. The research was designed for participants to focus on how well their partnership achieves the place-keeping aims of long-term green space management on-site. The data will show that factors influencing partnership capacity do occur at different scales. However the researchers were unable to examine scale (beyond the site) comprehensively in this pilot study. Other researchers address this issue by adopting a snowballing interview technique to identify inductively the relevant stakeholders in a network and then interview them (Sandstrom & Carlsson, 2008). In addition, others have conducted longitudinal studies to show how stakeholder networks change over time (Bodin & Prell, 2011) or examine the historical development of partnerships (Connolly et al., 2014). However this was outside the remit and resources of this pilot study.

Results
The findings are discussed according to capacity themes which go beyond those identified in the framework, to provide a picture of current place-keeping capacity of cross-sector partnerships in Sheffield, Hackney and Stockton-on-Tees.

Capacity theme Capital
The three local authorities strongly promoted community involvement amongst their core services to address external funding where possible. This is because aside from small amounts of money raised through membership (typical annual membership fees tend to fall within the range of £2-10), Friends Groups are able to apply for funding for any activities or projects which go beyond those covered by local authority funding. In SVP and RP, the (previously non-existent) Friends Groups were established to secure funding, with GSMs making contacts with members of the community to engage them to form the Group through significant LA support provided in the early stages. A number of the Friends Groups had attracted significant external funding (otherwise inaccessible to the local authority) e.g. FP (Changing Spaces Big Lottery Fund), MP (Community Spaces) and CP (Heritage Lottery Fund). Nearly all groups identified GSMs as essential in developing funding applications. In Hackney, GSMs acknowledged they had limited capacity to do this which sometimes resulted in groups not being able to develop -"we are proactive but are mainly reactionary due to our small team, large remit and resource limitations" (HGSM1). It was highlighted by GSMs that Friends Groups able to secure funding often attracted further investment, but the distribution of successful groups was spatially unequal across the three areas. In Sheffield, this was partly attributed to "differentiation in prioritisation of green spaces by the Community Assemblies. . .in the South West they are the number one priority. . .but in the North or East they may be fourth or less behind education, attainment, young people and social services etc." (SGSM1).

Capacity theme Commitment
Managing expectations was revealed as the key consideration for GSMs to ensure sustained community commitment. For example, Sheffield GSMs were keen for transparency in communicating to groups the impact of diminishing public resources on levels of support. Where previously GSMs had the capacity to commit time to attending Friends Group meetings, helping organise events and funding applications, they were aware of the danger of community frustration, and vulnerability of commitment, should declining officer support go unexplained. In Hackney, Partnership Agreements between Friends Groups and the council were seen by GSMs as a positive way of informing groups of the commitment and support they could expect from the council. In Sheffield, stewardship agreements have been introduced by the council to establish management/maintenance regimes for green spaces jointly with Friends Groups. Other Friends Groups interviewed demonstrated commitment to the protection and development of their sites in different ways. In PV, committee members contributed approximately 10-24 h per week while in RP, among other activities, the café is open over 35 h per week. Being actively involved often required a significant time commitment and many groups were comprised of members of retirement age. This was identified by the groups as a potential threat to continuity if key members are no longer capable of involvement. Individual commitment was found to be related to motivation: the Stockton GSM outlined how the personnel changes in one Friends Group led to a current membership "not interested in doing anything active. . ..different groups [have] different approaches" (StGSM1) (see Section 4.4).

Capacity theme Skill base
Working in partnership, GSMs and Friends Groups have many complementary skills. Whilst GSMs are paid professionals who understand the complexities of project management, tendering, design and consultation, their time is spread thinly over many groups and sites. Friends Groups often commit a greater amount of time than could be achieved by GSMs alone to the focused development of a site of which they have rich local knowledge and regularly use. The interviews highlighted how Friends Groups are highly effective, skilled and motivated to organise events as a means of local engagement and raising funds inaccessible to the council (e.g. RP, CP and MP) as many group members had professional backgrounds. However, when addressing the potential maintenance gap created by diminished council capacity to undertake manual work on-site, some Friends Groups showed little interest in developing this skillset (e.g. FP, SVP and CS). Most perceived this to be the council's responsibility, and the demographics of many groups suggested a limited capacity to undertake manual work, considered to be "too much work. . . we have lives!" (FP). There are also organisational difficulties involving groups in manual work. In Sheffield, groups pay for their own insurance, while in Hackney, groups were insured by the local authority to undertake manual work. This requires annual financial commitment from the local authority with the benefit of increasing its capacity for day-to-day maintenance. Of those groups directly involved in manual work (MP, PV, CP and RHCG), RHCG was the most actively involved in site clearing and building, maintenance and developments. This group's high capacity for manual work resulted from the Chair's professional landscape background, motivation, younger age of the members and the large partnership network with other community organisations. An enhanced partnership agreement was in the process of being drawn up with the local authority so RHCG could take on further on-site responsibilities.

Capacity theme Motivation
Some GSMs discussed their personal motivation for involvement in green spaces, which exceeded their paid responsibility. They were dedicated to improving the environment as a means to improving residents' quality of life. Friends Groups had different motivation for getting involved. For example, the involvement of many older female individuals was largely driven by social reasons, as Friends Group provide opportunities to meet new people and organise community events (FP, NGP and PV). A challenge mentioned by all groups was how to motivate members to take on administrative, organisational or practical responsibilities. In RP, there were difficulties in persuading people to be on the committee which often resulted in the time-consuming tasks being left to a small number of committed individuals. Cyclical trends in involvement and motivation were identified across most groups. All groups (regardless of membership demographic) were motivated by the desire to see environmental improvement occurring in their local green space. Many groups had experienced a noticeable decline in green space quality (MP, FP, CP, NGP and CS), which incited them to act. In CS, ongoing anti-social behaviour (ASB) had made life 'increasingly unpleasant for residents' (described as evidence of drug use, drinking alcohol in the park and vandalism). After the Friends Group formed, facilities and user experience greatly improved alongside a decrease in ASB. Having reached this point, some group members lost interest in staying involved. With reduced capacity to sustain and develop the space, equipment and facilities gradually declined through vandalism, and ASB increased according to FCS. This prompted another period of activity for the group of clearing up broken bottles, this time focusing on events including annual festivals and monthly teas (FCS).

Capacity theme Communication
Communicative capacity varied greatly between Friends Groups and local authorities. Sheffield's GSMs acknowledged a lack of communication internally and externally (echoed by Friends Groups in their comments about GSM capacity) which they attributed to reducing GSM numbers. In terms of external communication to the wider public, there was also considered to be a need to disseminate widely the contribution of Friends Groups. In Hackney, GSMs proactively promoted this community engagement work via newsletter, website and online forum, considered key capacity building tools to develop and sustain involvement, sharing information between groups, councillors and other stakeholders. However, sometimes such tools were limited, e.g. newsletter print runs were considered too short for RP. Most Friends Groups employed their own communication methods to promote their activities and inform members. These included formal and informal face-to-face meetings, minutes, telephone calls and notices in the park (e.g. FP, RP and NGP), whilst others focused on maintaining an online (websites and social networking) and media presence (local, regional and national press) (e.g. MP and CP). Events were identified as a key profile-raising tool to attract new members (PV), and ranged from informal site walks and open days (RHCG and NGP) to more high-profile annual festivals (FP, MP and CS). Friends Groups and GSMs relied upon their internal and external networks to advertise their presence. Groups with large existing networks had a greater capacity to draw on resources and develop further networks and sustain membership more easily than others.

Capacity theme Political Influence
In the context of national politics, the top-down drive has shifted power through devolution to focus on localism. Therefore the potential political influence of community groups is greater than ever before. While GSMs considered the political influence of Friends Groups to be variable, most community groups interviewed felt they had relatively strong political influence at a local level. Some cited partnership with other interest-led organisations. To what extent these wider groups have increased political influence remains to be seen.

Other emerging themes
Other related themes emerged from the data which may shed light on partnership capacity as an inter-related concept. Groups identified their unrepresentativeness of the communities in which the green spaces were located, in terms of demographics of (most members are white, female and over 60) and also interests (e.g. often not dog walkers, footballers or bowlers). Almost all the groups found it difficult to attract younger people, particularly teenage boys and people from different ethnic backgrounds. To represent better their communities, it was suggested (Hackney GSM) that outreach work (e.g. door-knocking) was effective but outside the stakeholders' capacity. It was highlighted by NGP that longstanding teaching expertise in their group could be harnessed with local schools to help start a junior section of the Friends Group. In addition, young people have been hanging around the local supermarket near NGP adversely affecting trade: one possible solution (presented by the Friends Group) was the potential for the supermarket to sponsor a skateboard facility for teenagers in the park. These are just two examples emerging from the interviews where cross-theme capacity might be increased.
Another theme is the extent to which groups work with other Friends Groups. All groups have a good working relationship with the council's GSM(s) and some have developed partnerships with other organisations (e.g. community groups and Wildlife Trusts). But this does not necessarily extend to other Friends Groups, despite the cross-area forums (Section 4.6). For example, RP, CP and FP do not cooperate with other Friends Groups and SVP made a point of stating that involvement should be kept local 'to those who will remain committed' to the Group and local community. In Sheffield, a sense of competition prevailed between groups when one group 'took funding' that another was applying for, while in Hackney, RHCG claimed that 'though local groups might be competing for same funding, it doesn't feel like a competition as they all want to improve the neighbourhood'.
The political and historical legacy is another theme affecting partnership capacity. In Hackney, poor public sector management in the early 2000s led to mistrust of the local authority which still remains according to interviewees. It manifests itself in: (1) groups' independence from the council (CP); (2) perceptions that the council a. doesn't listen and, b. has de-skilled tasks so they no longer require (e.g. horticultural) expertise and can be contracted out to non-specialist private contractors (e.g. tree management in CS) (3) partnerships 'dropping off' when achievements are made (RHCG).
In Stockton, the previous RP Chair set up the park café as a commercial entity which 'was never meant to be run that way' leading to strained relations with the council which the current Chair works to improve. In Sheffield, MP described their biggest barrier as a small group of GSMs who 'still see the park as purely council owned/involved' making changes 'without any consultation'. Elsewhere in Sheffield, relations with the local authority have been strained (SVP) where the Friends Group felt that community needs (particularly disabled members) were not considered in the park's re-design. Sheffield GSMs noted that the short-term approach taken by policy-makers to funding and locating GSMs throughout the city made it difficult to sustain (or make viable) local authority support for the long-term plans of Friends Groups. It was also highlighted that the reduced resourcing and capacity of Sheffield GSMs mean little support for new Friends Groups suggesting that they "are set up to fail" (SGSM1).

Understanding partnership capacity in practice: support and barriers
This section provides a discussion of the findings at different levels, to show that partnership capacity is not something that is wholly dependent on, or occurs within, the partnership itself. This is followed by a discussion of the implications of the findings in relation to the assumptions that community groups are willing and able to take on place-keeping activities which historically have been delivered by the public sector.
The findings show that the partnership capacity of local authorities and Friends Groups to deliver place-keeping varies greatly due to different factors. It is useful to consider that these factors as they occur at different levels (listed here with examples from the findings): • Individual partners: e.g. the specific reasons that individuals are involved; the specific skills individuals bring; their ability to carry out specific activities. . . • Individual groups/departments: e.g. Friends Groups' representativeness of the wider population; the manifestation of the shared commitment to protecting their green space (e.g. group constitution); Council department's degree of communication with external parties. . . • Collective of same sector partners: e.g. ability and appetite to build up political influence by connecting with other Friends Groups/similar organisations; the sense of competition between same sector partners (for funding). . . • Cross-sector partnership: e.g. scope for further financial resources not available when working as separate groups; Friends' Groups taking on specific maintenance activities that LAs no longer can; formalised shared commitments (signed agreements); strains on partnership due to historical mistrust (of the LA) and misaligned aims. . . • Wider contextual influences: e.g. decreasing amount of funding available for green spaces; long-standing mistrust of the LA; propensity of residents/communities to engage in place-keeping; the prioritisation (or not) of green spaces by local decisionmakers. . .
In this way, the findings demonstrate how partnership capacity is influenced by these inter-dependent factors, ranging from wider contextual influences to partnership-specific issues. This initial research adds to the body of knowledge about partnerships by highlighting these different types of factors but aspects such as relationships between individuals in an organisation as well as those between partner organisations need to be explored further. In addition, proposed actions such as junior membership of Friends Groups, the involvement of the private sector and groups working in partnership with other community groups point to a need for further research to examine how they might contribute effectively to place-keeping in practice.
That these factors occur at different levels has implications for political expectations of devolved governance regarding green spaces. While some community groups include members with transferable skills and bring with them a wealth of local and site-specific knowledge, others lack the expertise or inclination to undertake place-keeping activities. Groups remain reliant on the local authority for support to sustain site quality as they are unwilling to take on specific tasks of, for example, grass cutting, tree management and litter removal. This challenges the assumption that community groups can and will take on place-keeping activities. Such an assumption is reliant on groups having requisite commitment and involvement from members (although it is not clear what 'requisite' might mean in this context). Some Groups interviewed (e.g. MP, CP and PV) described themselves as stable in terms of commitment and capacity for involvement, but the Groups consider this a fragile and temporal situation. Central to all Groups is a reliance on a minority of members to sustain group momentum, putting pressure on a small number of people and presenting wider challenges for inclusive participation and representation.
Alongside the shift towards localism discussed earlier, collaborative approaches to planning have created opportunities for communities to take further control in decision-making (Healey, 1997), through a democratic approach to challenge the dominance of the professional (Irwin, 2006;Cohn, 2008). The findings reflect this to some extent but as the partnerships examined do not currently use, or have access to, methods to draw in underrepresented groups in the wider community, devolving power may result in favouring the interests of the prevailing few (after Tewdwr- Jones & Allmendinger, 1998). Almost all the groups, in both affluent and poorer areas, acknowledged difficulties in recruiting broad community representation and ensuring continuity of membership. Moreover, while local authorities are keen to work with Friends Groups, these may not always be the most effective conduits for the local area, e.g. other community associations might be stronger and perhaps more representative, e.g. tenants and residents associations or neighbourhood forums.
For devolved governance to permit communities to take further control in decision-making, it is clear that there is a need to understand communities' motivation and interest. Stand-alone projects were identified as a successful means through which community involvement could be initiated. For example, the regeneration of Clissold House at CP is an example of place-making which attracted volunteers and opportunities for funding. A continuing challenge for place-keeping is how to sustain motivation and interest in longer-term place-keeping beyond the 'end' of such place-making tasks. Ongoing two-way communication and continuity with GSM was valued by the Friends Groups across the sites and proved highly effective in encouraging and sustaining volunteer involvement and community interest (by promoting events and the green spaces according to Hackney GSMs), and to explore and secure ongoing revenue funding where possible.
The nature of place-keeping activities is also important: some groups are hesitant, under-skilled and/or unprepared to take on (further) management responsibilities. To help address this, resourcing (as in Hackney) is required to provide training, tools, and (where required) machinery and liability insurance. As community groups evolve and develop skills, they can attract further investment and funding (e.g. MP). However, this may not happen for all groups. The sustainability and capacity of groups, like the communities of which they are a part, is linked to their internal organisation and their ability to create and maintain partnerships through external networks (Berg & Nycander, 1997;Chaskin, 2001). Those partnerships studied with effective partnership capacity tend to call on large networks of partners for support and political backing (e.g. MP and PV). Partnerships with very few partners, e.g. between a Friends Group and the local authority only, may find it difficult to sustain momentum if local authority support diminishes or disappears (e.g. FP) -an issue of resonance for partnerships elsewhere (Molin & Konijnendijk van den Bosch, 2014).
Finally, it cannot be ignored that lack of public sector funding and greater community involvement are two inextricably linked drivers. As Section 4.1 highlighted, Friends Groups can access funding to which public sector is exempt. As local authority budgets continue to reduce, it seems apparent that together these two themes provide important impetus for non-public sector partners to be involved in green space management, and perhaps where funding will be sourced in the future.

Conclusions
This paper highlights significant questions for English local authorities facing central government-led devolution, with accompanying expectations that community groups can and will take on green space management responsibility. Evidence from this study suggests that without continuing local authority support, smaller community groups may find it difficult to adapt and may disappear due to lack of internal capacity and external support. To secure long-term community commitment there needs to be sustainability in the cross-sector partnership, and recognition by the local authority of the contribution communities make. As the move towards greater community responsibility continues, underpinned by limited resources, the role played by the public sector may change, potentially from one of implementer to one of facilitator. The research findings suggest that the role of facilitator may involve more support for Friends Groups to develop their capacity to operate independently from local authorities, as found by Chaskin (2001). Part of this could be an initial role for local authorities to facilitate cross-city Friends/community group networks to improve capacity and potentially help reduce inequalities of skills, knowledge and resources across a city (as highlighted in Section 4.6). The findings chime with those elsewhere that groups need to develop external partnerships with other community groups, charitable trusts and local businesses (Jones, 2002). This did happen in the sites studied, but could be further developed. This may address the issue highlighted earlier about Friends Groups not wanting to take on certain roles and responsibilities (e.g. specific maintenance tasks), particularly if (for example) they are already conducted by existing environmental enterprises and charities which could arguably add (expertise-based) value to such management processes. Hence, future research is required to examine how effective network development may be facilitated.
In addition, more knowledge is required about the governance structures in place in partnerships, and the local context. For example, patterns of involvement of partners and their variation according to the socio-demographic make-up of the partnership should be examined to help understand better how decisions are made and to what extent they are sustainable for the long term. This also points to the need for more research exploring the drivers of future change and how management practices may change in response. Such practices may be responses to the current economic drivers which may include a network of Friends Groups across a city, which currently occurs in Birmingham under the umbrella of the Birmingham Open Spaces Forum and nationally under the National Federation of Parks and Green Spaces. Other trends in green space management may be driven by community groups themselves such as (organic) food growing in public places or perhaps addressing unequal distribution of green space partnerships across a city. Such drivers will vary and may/may not lead to the same current 'solution' of community involvement, but may extend to include other partners. This research explored the capacity of partnerships according to just two stakeholders, indicating considerable scope for future research to examine the capacity of other groups using green spaces (e.g. sports groups). This could also involve an exploration of potential as well as existing capacity, which was outside the scope of this research, to shed light on how partnerships evolve over time.
While this research supports the claim that community involvement brings many benefits, it should not be considered as a quick fix in times of economic constraint: rather partnership capacity needs to be developed and supported over the long term to sustain the future of green spaces.