Navigating networks – to make a difference: the support base composition of local change agents in Amsterdam, Birmingham, Glasgow and Copenhagen

ABSTRACT

As the attention for collaborative governance is on the rise emphasizing bottom-up approaches in urban planning and regeneration, there is an increasing interest in local people able to enact ‘place-based’ leadership and change. While several studies have characterized these individuals, other parts of the scholarly literature show that they seldom work alone but often in collaboration with others. Whereas various urban scholars have highlighted formal relations as important in the support bases of local change agents, few have studied the informal site of these support bases. In this article, we study the networks that function as support bases of 40 local change agents in neighbourhoods across Amsterdam, Birmingham, Copenhagen and Glasgow. More specifically, we describe with whom these change agents collaborate. By applying a theoretical framework that distinguish actors according to dimensions of sectors (public, private) and proximity (local, regional, national) we assess the diversity (heterogeneity) and the similarity (homogeneity) of their support bases. Our findings demonstrate that most local change agents rely on informal trust-based relationships and homogenous support bases. As such, this research empirically unfolds the strengths and vulnerabilities of local change agents’ support bases, which contains key insights to improve current neighbourhood work in our cities.

KEYWORDS:

• Local change agents
• collaborative capacity
• neighbourhood
• place-based leadership
• support base

Introduction

Within current European urban policies, there is growing acknowledgement of the importance of involving residents and local stakeholders in order to implement and ‘catalyse’ positive spirals to tackle social exclusion, segregation and physical decay in deprived neighbourhoods (European Commission 2021; UN Habitat 2020). This has led to a scholarly interest in local and civic initiatives, ‘place-based’ leadership and change agents, those people able to motivate and mobilize local actors to make a difference, in places where they operate (Vallance, Tewdwr-Jones, and Kempton 2019; Durose et al. 2019; Pesch, Spekkink, and Quist 2019). In that regard, numerous studies have been published on persons able to make things happen working ‘in-between’ governments, private sector, civil society and NGO’s. For example, they have been characterized as ‘everyday makers’ (Bang and Sørensen 2003) or ‘everyday fixers’ (Hendriks and Tops 2005), ‘collaborative capacity builders’ (Weber and Khademian 2008), ‘boundary spanners’ (Van Meerkerk 2014; Williams 2002), ‘social and civic entrepreneurs’ (Durose 2011; Mens et al. 2021), ‘exemplary practitioners’ (van Hulst, de Graaf, and van den Brink 2012), ‘street level policy entrepreneurs’ (Arnold 2015) and ‘smart urban intermediaries’ (Durose et al. 2019). Other parts of the literature criticize the focus on individuals and argue that the isolated hero or ‘hero-innovator’ rarely exists (Hendriks and Tops 2005). These scholars claim that conducting change in relation to complex or ‘wicked’ issues (Pesch and Vermaas 2020) is seldom done in isolation or stand-alone actions (Van Bueren, Klijn, and Koppenjan 2003). Instead, they claim that those who enact change most often do so in collaboration with people in different types of networks (Koppenjan and Klijn 2004; Wegner and Verschoore 2021). Some scholars conceptualize these ways of operating as ‘distributed heroism’, or as ‘embedded’ and ‘distributed forms of agency’ (Garud and Karnøe 2003; Meijer 2014). While there is an increasing scholarly attention on collaborative capacities and collaborative action, especially in formal or contractual networks (Bel and Sebő 2021; Le Galès 2002) only few have studied the informal support bases of change makers who make things happen (Dempwolf and Lyles 2012; Shin 2021; Uittenbroek et al. 2019; Weber and Khademian 2008). Since local change makers rely on both formal and informal resources, we will specifically also shed a light on this informal terrain.

Within the literature of social capital and network studies we find a number of studies, that point to the potential of networks to address complex problems in urban contexts at different levels (Lang and Novy 2014; Agger and Jensen 2015; Siciliano, Carr, and Hugg 2021). Yet most of these studies, tend to focus on formal networks and relations, and thereby are at risk at overlooking what recent strands in the research literature describe as ‘politics with a small p’ or ‘urban everyday politics’ (Beveridge and Koch 2019). These are activities or actions, that often take part in informal support bases, disconnected from formal institutions of urban governance but that can make a difference and have a huge positive influence in their locality. The aim of this article, is thus to obtain a better understanding of local change makers, that we define as ‘people that are able to enact ‘place-based leadership’ by setting things in motion through combining relationships and resources in their locality (Durose et al. 2016, 2019, 2021). More specifically, we intent to study the support base of these change agents by looking at both formal but in particular informal relations, with whom they collaborate at different levels.

In doing so, we contribute to the scholarly field, by combining different streams of literature. On the one hand, we know from studies on urban practitioners and social capital that embedded forms of agency often are determined by either heterogenous or homogeneous resources, actors and networks (Osborne, Baldwin, and Thomsen 2016). Some scholars point to the importance of access to different type of networks where certain urban practitioners are known for ‘crossing boundaries and connecting persons’ (Honig 2006), for ‘participating in different arenas’ (Weerts and Sandmann 2010) and for ‘engaging in dialogues with people from different organizations’ (van Meerkerk and Edelenbos 2018). By working with heterogenous relations, consisting of different types of actors, they create ‘bridging’ forms of social capital (Putnam 1995) that provide crucial access to new information to construct ‘vital coalitions’ to ‘get things done in the neighbourhood’ (Siddiki, Kim, and Leach 2017; Siciliano, Carr, and Hugg 2021). However, on the other hand, other network studies show that ‘birds of feather flock together’, or in other words ‘similarity breeds connection’ (McPherson, Smith-Lovin, and Cook 2001). This is understood as relations with those who are similar measured by similarity in socio-economic status, cognition or resources (Musso and Weare 2015; Siciliano, Wang, and Medina 2021; Shin 2021) termed as ‘bonding social capital’ (Putnam 2000). Scholars point out that such relations function as a social safety net which generates reliance on trust and on risk avoidance that together constitute a support base for future actions (Agger and Jensen 2015). This tendency to form homogenous relations is identified as the homophily principle and an important determinant of network structures (Gerber, Henry, and Lubell 2013).

Based on our previous studies, we know that especially local change agents need both trust-based relations and access to novel information and resources to ‘make a difference’ (Durose et al. 2016, 2021). In this article, we present the empirical and descriptive results of a study of 40 local change agents from Amsterdam, Copenhagen, Glasgow and Birmingham that took part in a comparative European research project.¹ We asked each of these change agents to mention up to 10 of their most important relations in 2019, and we repeated the question two years after in 2021, to check the sustainability of the data. Our findings shed light on how the support bases of local change agents are composed? The composition of these support bases can be perceived as an indicator for the collaborative capacity of local change agents.

The article is structured as follows. In the next Section 2, we introduce our theoretical framework, while Section 3 presents the methodological approach and empirical fieldwork data. This is followed by Section 4 where we present the empirical findings. In Section 5, we present the comparative analysis. Finally, Section 6 contains the conclusion and outlines implications for future research.

Analysing local change makers and their support bases

As noted above, part of the scholarly literature points out that urban practitioners that are considered to make a difference locally are not ‘heroic innovators’ who work ‘alone’ when they enact change or conduct successful projects (van Hulst, de Graaf, and van den Brink 2012; Meijer 2014). Rather, their efforts and contribution to local change and transformation processes are more often found to be the result of ‘team efforts’ where they act as ‘catalysts’ or ‘enablers’ (Sirianni 2009) ‘synthesizers’ or ‘brokers’ providing visions and aligning resources in order to empower collective or collaborative capacity (Igalla, Edelenbos, and van Meerkerk 2020; Steyvers et al. 2008). In order to create collaborative capacity, it has been acknowledged that it is crucial for change makers to build and sustain personal networks (Mens et al. 2021). While the literature informs us about how formal urban practitioners manage formal networks and resources and why it has importance to collaborate, it is less informative how informal (non-contractual based) change agents manage their informal support bases. Following this, we will draw on literatures on change makers and local change agents for empirical fieldwork data about what is already known and use parts of the network literature as it offers some conceptual vocabulary to understand the composition of support bases.

Support base composition: similarity or difference

Although network literature is vast and comprehensive, it contains some core concepts, that are relevant for us, to get a better understanding of the people – and support base, behind local change agents. We are interested in the actual composition that function as a support base of local change agents. Thus the composition of their support base, in the literature measured through its diversity or range of relations (ties) and defined as ‘the extent to which network relations span institutional, organizational, or social boundaries’ (Siciliano, Carr, and Hugg 2021, 13). The composition of those relations can differ between involved actors, or they can be seen as quite similar. This composition is interesting as it indicates something not only about the span of contacts but also about access to resources that can be mobilized in collaborations.

From the empirical literature on networks, we know that networks with a broad variety of actors from different sectors and levels are crucial for their ability to enact changes (Clifford and Tewdwr-Jones 2013; Agger and Sørensen 2018; Forester 2021). Such networks are labelled heterogeneous networks, as they consist of members from different sectors. This diversity is considered to be important enabling innovation, flexibility and ability to address complex problems (Eggers, Singh, and Goldsmith 2009). The literature points out, that having access to a diverse range of actors and resources, is essential for accessing non-redundant and novel information (Reagans and McEvily 2003).

Another useful term in the scholarly literature is the labelling of networks as homogeneous. Such networks consist of most-similar relations, building on strong relations within the group, typically among families or close friends. Such relations are mainly based on trust and can function as an important support base for future actions (Agger and Jensen 2015). An asset of homogeneous networks is that similar and known actors are more likely to ensure, maintain and sustain relationships than a diverse group of actors (Gerber, Henry, and Lubell 2013).

For the purpose of this article, we are interested to what extent local change agents rely on heterogenous or homogeneous relationships. To address this question, we will conceptualize the composition of their support base in terms of homogeneity and heterogeneity. Further, we operationalize the support bases of our local change agents. Here, we are inspired by the literature that differs between sectorial-based affinity and scale proximity. Sectorial affinity (e.g. public, private or NGO) is about to what extend actors are active in comparable sectors. Studies point out that actors from comparable sectors tend to collaborate more than actors from different sectors. Scale proximity (e.g. local, regional, national, global) depicts to what extend actors are located in a similar area or environment (Gerber, Henry, and Lubell 2013). Here, the assumption is that actors operating in different scales will be confronted to varied types of information and are expected to be more likely to collaborate with actors operating at similar than different scales. Taking this scaling into account is particularly relevant for local change agents since their collaborations are contextually embedded (Emirbayer and Mische 1998).

Summing up, we assume that the benefit of working with actors from different scales and sectorial domains, is that one may benefit from getting access to more novel information and resources (Schrama 2019). Therefore, if actors seek to ‘maximize their own influence [it] can therefore be expected to form social ties that add unique skills and resources’ (Nohrstedt and Bodin 2020, 1088). Other studies point out how information exchange is more likely when actors have different organizational backgrounds and that access to diverse networks on different levels help to enhance the performance (Siciliano, Carr, and Hugg 2021).

Nonetheless, the aspect of similarity is also relevant to consider as it helps to build trust which is of particular importance in the mobilization of resources, and thus a critical determinant of collaboration (Gerber, Henry, and Lubell 2013). To sum up, it is about ‘the need of new information and ties that bridge so-called structural holes to disperse new information are the source of added value, while network closure is critical in realizing this added value by improving communication and enhancing trust’ (Schrama 2019, 126).

Drawing on these different perspectives, we will analyse our data by sectorial affinity and scale proximity to understand the heterogeneity or homogeneity of their support bases. We argue that knowledge about these support bases can show the potential of (un)known access points. This has relevance to understand better with whom local change makers engage in public issues (Edelenbos, van Meerkerk, and Schenk 2018)

We operationalize sectorial (dis)similarity by distinguishing professional sectors in which actors can be affiliated. This includes private (banks, funds, investment agencies), public (governmental bodies and services), NGO (third sector, community sector)² and other actors (such as friends, family and partners). We operationalize scale (dis)similarity by distinguishing on which level or scale actors are mainly active and participating. This is primarily based on the location of actors, next to their span and scale of action. This includes local (community, neighbourhood), urban (city-wide, urban region), national (country-wide), international and other (domestic) scales. Table 1 includes in detail how we conceptualize this nexus of sectorial affinity and scale proximity, operationalized as:

Table 1. Operationalization.

Support base composition
Heterogeneity/Homogeneity	Sectorial affinity	Public
		Private
		NGO
		Other
	Scale proximity	Local
		Urban
		National
		International
		Other

Based on our readings of the scholarly literature, we have formulated two theoretical expectations:

1. if local change makers are more driven by novelty, innovation and diverse information their support bases will consist of diverse actors (different scales and sectors) (heterophily thesis);

2. if local change makers are more driven by trust and risk reduction, their support bases will consist of similar actors (similar scales and sectors) (homophily thesis)

Methods

Our cross-national qualitative research design was developed to gain insights into the practices of local change agents in four North-Western European cities. Based on a comparative data selection protocol, we built up data along the lines of a shared and comparative design. The selection process was carried out in a series of steps. First, our aim was to select neighbourhoods recognized as: ‘diverse, challenging, and vibrant, able to adapt and active’ (Durose et al. 2019, 11). We assumed this as an important neighbourhood context in which local change makers can ‘provide services, connectivity and knowledge to other citizens to create new spaces of action’ (Blom Hansen and Verkaaik 2009; Wagenaar and Healey 2015; Beveridge and Koch 2019), particularly in areas where such practices and resources might be needed the most but are not always self-evident. We used this sensitizing definition to identify four anchor neighbourhood in Birmingham, Glasgow, Copenhagen and Amsterdam. Second, we drew upon the insights of co-operation partners (national and local organizations knowledgeable in the field of urban governance and regeneration). This step allowed us to draw on local knowledge and networks of partners in each national context. Third, we studied secondary data sets, e.g. national and local measures of deprivation and socio-economic compositions for each city. This step provided a base for comparison and enabled us to identify neighbourhoods in each city that in part reflected our sensitizing definition (de Wilde and Franssen 2016; Nio, Reijndorp, and Veldhuis 2008; Birmingham City Council 2019; Meijer et al. 2013; Glasgow Centre for Population Health 2016). Fourth, to complement and check data, we also travelled to the localities. At these sites, we walked around the neighbourhood, called into places, talked to people and took photos. This created an additional and shared understanding of the challenges in the area, adding local knowledge and contacts. Based on this, four extensive City Profiles were completed by each local team which included detailed background information about socio-economic activities, demographics, housing, (un)employment, community cohesion, deprivation and political-administrative information of each anchor neighbourhood to create a shared understanding of the local contexts. This information is summarized in Table A1 (Appendix). Based on that information, we identified neighbourhoods where urban development had been targeted, for example, through government-led interventions, often related to socio-economic disadvantage, or a reputation for community action. This informed our selection of Balsall Heath (Birmingham), Govan (Glasgow), Nieuw-West (Amsterdam) and Nord Vest (Copenhagen) as our four anchor neighbourhoods. These neighbourhoods were selected as relevant contexts for our study, meeting our criteria for the selection of our cases.

Finally, the fifth step was to identify local change makers within these four neighbourhoods. Since we did not select change makers based on their formal organizational attachments or clear institutional affiliations, such as front-line workers or street-level bureaucrats (Verhoeven and Van Bochove 2018; Tonkens and Verhoeven 2019; Petriglieri, Ashford, and Wrzesniewski 2018), our sample of local change agents could not be based on their professional contract or legal status but on the informal significance of what they do. Previous research (Durose, Justice, and Skelcher 2015), follows other researchers (Lipsky 1980; Maynard-Moody and Musheno 2003), by using the sensitizing concept of ‘local practitioners who are making a difference’. We followed this ‘sensitising’ definition, adding that local change makers needed to be people who have a link to the neighbourhood, are recognized as ‘making a difference’ in that neighbourhood and are actively involved in challenging issues. To what extent people are recognized and reputed as ‘making a difference’ needed to be based on local recognition and qualification of ‘significant others’ in the neighbourhood (Durose et al. 2016; Durose et al. 2021). Moreover, people need to ‘have their feet in the dirt’ or have a ‘messy, conflicted, dirty-hands experience’ (Forester 1999, 8). By those definitions, we went in the field, meeting with local institutions (such as the municipal and borough government, national think tanks and knowledge institutes). In parallel, by multiple local cycling and walking tours, ‘being there’ and by talking to numerous initiatives we collected a big database of initiatives, organizations and persons. By snowballing, asking at the close of each conversation, for suggestions of other potentials people to talk to Petriglieri, Ashford, and Wrzesniewski (2018), we created a database of >200 people.

Based on that database, we conducted more than 20 explorative conversations in each site with a wide range of residents, experts, volunteers and professionals. These conversations were an explorative step to get to know people and assess which individuals could meet our criteria. These conversations included questions about the person’s background, current interests, role in the area, and how they perceived ‘making a difference’. In the further selection process of our local cohorts, we aimed to ensure diversity in both position (including, for example, active citizens, community leaders, elected representatives, front-line workers, social entrepreneurs and artists), next to the time and character of someone’s relationship to the neighbourhood, for example, living or being professionally engaged there. In doing so, we aimed to create diversity in their local reputation, diversity in position, relationship to the neighbourhood and length of active engagement. On the basis of those comparatively agreed guidelines, we recruited a cohort that was gender balanced and reflected the neighbourhoods in terms of ethnic diversity (see Table A2 and A3 in the Appendix). The cohort was more ethnically diverse in Birmingham and Amsterdam than in Glasgow and Copenhagen, reflecting the demographic profile of those cities. Finally, we could only select people who are (professionally, psychically and mentally) willing and able to take part in such type of research. Based on that selection, they were invited in the pool of participants for our research. We expected dropouts, therefore we initially aimed for a cohort of 10+ persons in each city.

Importantly we did not preselect the local change makers on the basis of their informal work, their sectorial affiliations or the scale in which they gained local reputation. Therefore, this includes local change makers active in urban gardening, social welfare services, community empowerment and women’s emancipation on the local scale of the neighbourhood, but sometimes also being active on the urban or even national scale. More importantly, these change makers are recognized by others ‘by virtue of their reputation, skills and imputed relations provide services, connectivity and knowledge’ to other citizens (Blom Hansen and Verkaaik 2009, 16). We recognize that the local change makers that we study have different aims and work-related content, but by reputational methods, we primarily intended to include those that have similar gradients of local credibility in how they operate. In the research project, we aimed to empirically understand how they developed, sustained and adapted their work practices. In this article, we descriptively aim to understand with whom they developed, sustained and adapted their work practices.

Qualitative data collection: interviews and shadowing

We shadowed and interviewed 40 local change agents across the four neighbourhoods over a 30-month period. Each change agent was shadowed individually (McDonald 2005) at two different time points, where the researcher observed and recounted practices, and documented conversations by ‘running commentaries’ and ‘jotting notes’ (Emerson, Fretz and Shaw 2011). For this article, we primarily rely on the two rounds of interviews and the shadowing notes. Separately, at the start and the end of the shadowing phase, we conducted semi-structured interviews. We comparatively have asked questions about what it means to them to make a difference next to what enables helps and hinders them in trying to make a difference. The first round of interviews and network mapping was conducted in the winter of 2017, the second round of interviews was conducted in the spring of 2019. In this second stage, we were able to duplicate questions and check whether answers were still ‘accurate, comprehensive and consistent’ (Bevir and Rhodes 2006, 184). Our interpretive epistemology encouraged us to work abductively (Schwartz-Shea and Yanow 2013), and follow leads (Charmaz 2014, 25). Our data were coded locally in the language of each site according to a shared codebook. We used memos to share and deepen analysis on emerging themes across sites. Our coding and memos informed an iterative process of ‘member checking’ (Erlandson et al. 1993) with research participants. Furthermore, our data set consists of four local workshops in each site (total 16) and four international workshops (in Glasgow, Lisbon, Krakow and Copenhagen) where we brought change agents together to reflect on emerging insights from and during the research. This resulted in a data set including notes on around 640 h of shadowing, 80 reflective conversations, 20 interactive workshops and a total of 80 interviews.

Identifying the support bases

To map support base compositions, we included questions in our interviews by which participants needed to qualitatively list up to 10 actors they collaborate with. This methodology closely follows the research of Naurin and Lindahl (2010). Depending on the order in which they spontaneously mentioned co-operation partners, researchers included their answers. The assumption behind this approach is that actors co-operate most often with the ones that come first to mind. This resonates with what is known as the widely used measurement of perceived social capital (Williams 2006; Shin 2021). As such, the support bases are not studied from ‘the outside in’ but from the ‘inside-out’ (Bridge 2002). This ‘egocentric’ approach follows relationships from individual respondents to their social contacts. It is the egocentric approach that has characterized this study and resonates with other neighbourhood research (Bridge 2002), while there is a growing acknowledgment in the added value of combining qualitative interviews, stakeholder analysis and social network analysis (Lienert, Schnetzer, and Ingold 2013).

Following the procedure of Naurin and Lindahl (2010), we asked respondents with whom they collaborate. This has resulted in data sets of more than 400 actors (nodes) that have been mentioned by all 40 change agents. During the data processing, we distinguished separate stages of data analysis:

1. We analysed all the (2 × 40) 80 practitioner interviews of all four sites on the network questions. Based on the qualitative answers, we conducted four nodes lists with all actors, one Excel file for each of the site.

2. We separated the answers provided in the first round of interviews (2017) and the second round of interviews (2019), which resulted in two Excel files per site (eight in total).

3. When we completed, all local researchers completed and checked these lists on the attribution of sectorial and scale labels to increase intercoder reliability. This relates to an informed process of ‘peer de-briefing’ within the research team (Erlandson et al., 1993).

4. Next to this, we contextualized and checked the primary sources of the data (interview data) with the shadowing notes to create an internal data-check on actor networks.

5. Finally, these lists were analysed by the software programme Gephi, which enabled to visualize and standardize a reliable procedure to quantitatively compute the composition of support bases.

Results: neighbourhood specificities

In this section, we first present the neighbourhood site specificities and elaborate on how local change agents reflect on their relations, in a comparative perspective. Finally, this builds up to the analysis of sectorial affinity and scale proximity to conclude about the overall support base of local change makers.

Amsterdam Nieuw-West

Our mapping of relations in Amsterdam leaves us with 118 different types of relations. The majority of these relations are mainly related to NGO (civil society sector) locally. What is clear is that in the Dutch data, the municipality plays a key role. Being referred to the central municipality (gemeente) or the decentral level (stadsdeel, gebiedsteam), the municipality and its neighbourhood teams are strongly present in the support bases of all participants. It also contains a wide range of public and NGO-related actors ranging from arts and culture-based organizations, schools and media.

It shows relationships with debate centres, museums, welfare organizations, theatres, local restaurants and higher educational organizations. In terms of geographic spread, most actors and organizations mentioned are local or urban, based in Amsterdam or based in the neighbourhood of Nieuw-West. Only few relations are active outside Amsterdam on the national or international scale.

Copenhagen Nord Vest

The Copenhagen database consists of 113 relations which indicates mainly public and local affiliations. The municipality is a key actor, like people active in area-based initiatives (ABI). ABIs are mostly organized as temporary partnerships in which the central government supports and fund local government projects that include various local private and non-governmental partners representing sponsors as well as users and citizens (Agger and Jensen 2015). In that respect, ABIs consist of varied actors but mainly serve governmental aims and objectives and are mainly publicly funded. This is probably why in Copenhagen actors active in ABIs are well present. Many change agents refer to local and public relationships in Copenhagen such as for example the City Council and ABI next to active citizens, neighbours, community activists, local entrepreneurs and businesses, such as supermarkets and local food enterprises.

Birmingham Balsall Heath

The Birmingham database consists of 99 relations which includes mainly NGO and local affiliations. The municipality (City Council or Councillors) seems to play a more limited role. It also shows a low relevance of private organizations. Instead, in Birmingham change makers rely on third sector organizations, social enterprises, civil society organizations, faith-based organizations, mosques, think tanks, community project workers and regional or local networks. Next to these local and regional attachments, support bases include many national and regional relations such as trust funds, Members of Parliament and the National Council. Birmingham support bases rely on relationships at debate centres, museums, welfare organizations, theatres, local restaurants and higher educational organizations such as Birmingham University, Saheali Trust, Balsall Heath Forum and Ort Café.

Glasgow Govan

The Glasgow database contains 95 relations of mainly NGO, local and national actors. This is an interesting finding, since it shows a limited significance of urban actors in the Govan network. For the support base of change agents in Glasgow, the municipality (City Council, Councillors or local policymakers) does not play a key role, while also a low relevance of private organizations. Instead, Govan support bases strongly rely on national relationships such as National Parliament, Scottish Gas and Oxfam Novib which have strong local chapters within Govan, next to local partners such the local police, schools, Govan community centres, Sunny Govan, next to local trust funds like GalGael.

Results: connecting scales and sectors

Our study shows that across the four sites, local change makers acknowledge the importance of particularly strong local relationships to get things done. It is this local embedding of knowing the right people in the right places which they mainly rely on. In our interviews, local change agents reflected on this element repeatedly, mentioning the importance of knowing, maintaining, and nurturing local relationships. Because all depends on ‘having knowledge about the social map’ (P8) since ‘it’s all about people’s knowledge’ (P6). It demands a specific starting point about the social world and how to approach other people, indicated as: ‘I always start by trusting another person’ (P9). For some this is key about what they are doing, because ‘people trust you very quickly. Otherwise, you’re just one of the professionals’ (P14).

Therefore, trust is an important resource to build upon, to build relationships and to get things done. This becomes evident – as our mapping of relations from the first and second data collection demonstrates a consistency in choice of their 10 most close collaborators. However, trust is not something that can be learnt from books but demands time, investment and experience. This is an important skill that defines their work, because: ‘You sort of know how people work. You know who to avoid, you know who to go to […] You can galvanize people’ (P29). This has reciprocal value because having access to and being able to rely on people’s knowledge can only be done with a certain form of credibility, because

The one that calls me knows: you can trust me. He knows what I can mean for him. He knows that I have that knowledge. He knows that I rely on trust-based relationships, that is crucial […] “especially with people who have less trust in others”. (P10)

And this is specifically crucial because ‘I think sometimes things just happen, because of the contacts’ (P27). Local change makers are aware that this is specific for their work because: ‘Your work has to be the testament that gains you trust and respect’ (P29). It is this reliance on trust, credibility, and reliance, which form the core elements in how and why they can create a difference because: ‘my trust-based networks are unknown for the system’ (P1). But this reliance on trust has a strong performative effect on whom they collaborate with, since one adds that:

I look for people who have a passion. I look for people who are enthusiastic and I look for people who really want to create change. I’ll look for somebody who is very well known in the community, who is very well connected. (P24)

While others mention that ‘I have a group of people around me […] that I trust and build upon’ (P4). In reflecting on their contacts, many local change agents emphasize the importance of having trust-based relationships. Thus, many rely on closely affiliated, known and trusted actors. For example, knowing the local theatre manager enables to easily organize a public meeting while others have local credibility to use technical facilities of others to make an event possible (P1: P5, P6, P14, P15, P19; P20; P35). Some are deliberately managing this credibility, mentioned as: ‘I am very careful with my contacts. For instance, I will never share numbers without consultation. It is all about my reputation, which can fade away easily’ (P5).

On sectorial diversity, let aside the many ‘other’ actors in Table 2 (this can be explained by the fact that many refer to their wife or husband, children, family members, relatives), most local change agents acknowledge the importance of good contacts in private, public and third-sector institutions, such as National Parliaments, political parties, banks, councils, trusts, foundations and funds, the police and City Councils. This can be illustrated:

We’ve got good relationships with local councillors, and our reputation is good. […]. We certainly have a good relationship, because we recognize that sometimes you need political backing, and it's good for the local MPs to know, and value what it is that you’re doing. But it doesn’t shape much of what we do, in terms of the engagement with politicians. (P37)

Such contacts are mainly perceived as instrumental access points that could serve as gatekeepers to ‘other’ additional resources ‘outside’ or beyond the neighbourhood. Most local change agents have access to credible entry points on local, urban or national levels to get something done. Some local change agents have the direct phone number of the mayor (P8; P17; P18) the minister (P7; P19) or the aldermen (P11) next to the local butcher, citizens and community organizations. As stated: ‘People always ask me: hey do you have the phone number of … Then I take my phone and 9 out of 10 times, I also have that number’ (P5). Some even created an advisory council (P6) or a think tank (P4) for their organization including close contacts with regional and national authorities, which provides resources at arm's length. Sometimes this creates an opportunity to rent a piece of real estate (P4), a vacant piece of land and a wooden shack (P1), a building (P37) or derelict site to run a new initiative (P16). Sometimes it is about connecting the dots between certain actors to make a difference, as illustrated:

Table 2. Sectorial allocation of support base per site.

	Private	Public	NGO	Other
Amsterdam	7%	33%	42%	18%
Copenhagen	10%	41%	21%	28%
Glasgow	4%	30%	42%	24%
Birmingham	1%	22%	49%	18%

we made some relations between a newly imitated pizzeria and the local brewery. So now, the pizzeria gets it beer from there. So, we contribute to create these relations- where people meet each other from different parts of the locality. And in reality, it is that link and that work that in essence is what we do. (P14)

Despite change agents having different sectorial contacts, they also experience having a hard time to gain access to the right places: ‘I don’t have the right connection with daily government […] it is crucial to be known in those networks. We devote a lot of time to that’ (P1). This not only applies to formal access points, but the data set also shows quite an absence on contacts in the private sector, as one reflects: ‘The private sector. Euhm … […] I can’t honestly, right off the top of my head, say that there’s anyone in the private sector that we work closely with’ (P35). More general, this is an overall finding since local change agent support bases consist not more than 10% private actors.

Analysis

In terms of scale proximity, the support base analysis shows that local change agents combine multiple scales, including local, urban, national and sometimes even international (non-)governmental organizations. They understand that being active on multiple scales creates more strategic options to ‘play on the chessboard’ (P1, P20). Many local change agents do not deliberately create such strategic venues, but some do with the instalment of a think-tank or Council of Advice for the purposes of one’s own organization (P4, P6). Others create momentum for community projects in making sure it resonates with national policy objectives for instance by inviting the national Minister, City Councillors, scholars, local artists, aldermen and policymakers to the neighbourhood (P35, P7). Moreover, our data show that local change agents combine various local, urban and national relations. Especially in the cases of Birmingham and Glasgow, they show a strong matrix of multi-level involvement, by national and local contacts. Their support bases include relationships with Members of Parliament, Ministries, National Councils and political parties next to regional authorities, local think tanks and community services.

However, this is seldomly observed in the full cohort. Most relationships are embedded locally with a relative absence of international contacts. Especially in Amsterdam, Birmingham and Copenhagen, local change agents support bases are mainly oriented locally (more than half of the contacts, between 54% and 59%). In Copenhagen, this can be explained by the local significance of area-based initiatives, in Birmingham by local faith-based organizations such as Balsall Heath Church Centre and Clifton Road Mosque, while in Amsterdam the role of the local municipality and decentralized public services plays a role. Glasgow deviates a bit here in which local (39%) but especially national (43%) actors seem to play a key role. Moreover, Table 3 shows limited presence of urban relations in Govan (15%), but this also seems to be the case for Balsall Heath in Birmingham (19%). Therefore, this could be an interesting avenue for future research to investigate if this peculiarity says something about the Scottish case or to what extend this can be considered as an ‘UK effect’ (Batley and Stoker 2016). And despite this case peculiarity, the overall picture still shows a strong reliance on local actors in these neighbourhoods. If we relate this finding back to our cohort as a selection of local change makers, these findings confirm the expectation of scale similarity.

Table 3. Scale allocation of support base per site.

	Local	Urban	National	International	Other
Amsterdam	54%	29%	11%	1%	5%
Copenhagen	59%	25%	16%	X	x
Glasgow	39%	15%	43%	3%	x
Birmingham	58%	19%	22%	1%	x

Next, on sectorial diversity we see, despite some variety, some commonalities in the composition of support bases. Our empirical findings firstly demonstrate a strong overrepresentation of public and NGO relationships. Next to that, or maybe consequentially, mainly all support bases show a relative absence especially in terms of private actors, such as national banks or investment funds but also local actors like supermarkets, start-ups or local entrepreneurs. Some do have good relations with private funds and housing agencies such as for instance the National Lottery, trust funds or consultancy offices, but overall, the presence of these relations is very limited. Private actors consist between 1% and 10% of the neighbourhood cohorts. Since most local change agents face challenges that overstretch public or third-sector boundaries, the absence of private actors is striking. Given the sectorial background of our practitioner cohort being mainly active in or affiliated to the public sector or NGO’s, this also confirms the expectation of similarity, as sectorial similarity, since our data shows an overemphasis on public and NGO-related actors.

Discussion and conclusion

Given our cohort of change agents (see Methods and Appendix), being mainly locally active in and affiliated to public or NGO sectors, our data confirm the homophily thesis. Not only in terms of sectorial but also scale proximity. Moreover, our data show that this similarity can be understood by a reliance on trust, local credibility and risk reduction. It is this reliance on trust-based relationships, which feeds the reputation and credibility of local change agents and feeds-forward new relational opportunities (see also Durose et al. 2021). In other words, the support bases of local change makers are mainly based on similarity. The data show that the support bases of local change agents particularly rely on local, NGO and public actors. Local change agents seem to rely more on trust-based relationships than on relationships that could increase innovation, novel information and new resources. The analysis unfolds the homogeneity of their support bases. Not only in scale but also in sectorial terms. Despite some variances, this is a key comparative finding and a characteristic of most local change agent support bases. As such, this article holds empirical contributions especially to the literature on how urban practitioners, collaborative change makers and local change makers make a difference (Verhoeven and Van Bochove 2018; Tonkens and Verhoeven 2019; Igalla, Edelenbos, and van Meerkerk 2020).

Moreover, showing the homogeneity of their support base goes against many studies addressing the way how some change agents create added value. Some are depicted as known for ‘crossing boundaries and connecting persons’ (Honig 2006) ‘participating in different arenas’ (Weerts and Sandmann 2010) and ‘engaging in dialogues with people from different organizations’ (van Meerkerk and Edelenbos 2018). Our empirical fieldwork questions the span and reach of such differences, by evidencing local change agents get most things done by collaborating with similar actors, confirming that ‘similarity breeds connection’ (McPherson, Smith-Lovin, and Cook 2001; Gerber, Henry, and Lubell 2013). It shows that local change agents prioritize trust and risk reduction instead of novelty, innovation and diverse sources of information. The impact of such relations on the scope and outcomes of initiatives has been reported in other studies (Durose et al. 2019, 2021).

The limitations of our research are also significant, since for example the operationalization and conceptualization of difference and similarity can be contested. Homogeneity can be measured along multiple dimensions, such as political stances, practitioners’ personal attributes such as age and gender. Generalizations are hard in this respect and with this article, we do not aim to generalize our findings to all local change makers. But what we do see throughout the four cities and support bases of the local change agents we studied is a strong reliance and emphasis on sectorial and scale similarity. This has a potential for future empirical research.

This insight holds clear societal relevance, since this strength of similarity can also be quite vulnerable. Particularly in times of austerity and budget cuts, there is a risk when key trusted actors drop-out, or when novel information is not available, this could have a strong impact on the collapse of community relationships, neighbourhood resources and collaborative capacity. This vulnerability speaks from all support bases, being quite unbalanced. For example, the relative absence of private, international and non-local actors illustrates the potential for local change agents to improve their collaborative capacity. This relates to what has been identified as the ignorance of the private sector in many collaborative efforts (Scott and Carter 2019). This can be improved by creating more heterogenous and complementary relationships, especially considering the private sector, or actors on the national and international scale. Probably this is not easily being done but given the importance of pluriform relationships for sustaining their efforts, this cannot be underestimated.

It could provide a starting point for future research, since a different selection of cases or selecting more on private sector agents would have led to different results and could unfold how we better understand and improve this private sector ignorance related to collaborative work. It could also be a starting point for local or national policymakers to improve supporting policies for collaborative capacity. For example, in developing infrastructures or venues where local change agents could meet, exchange and interact with more dissimilar contacts to build up relationships and resources that could boost their situation, develop their resources and improve the sustainability of their support base. This could improve the work of local change agents and their collaborative capacity in the urban neighbourhoods of our cities.

Acknowledgements

The authors thank co-operation partners and participants in Balsall Heath and Sparkbrook, (Birmingham) Govan (Glasgow), Nieuw-West (Amsterdam) and Nord Vest (Copenhagen) for their insights and invaluable contributions to this research. The authors acknowledge the invaluable contributions of our project team members Merlijn van Hulst, Oliver Escobar, Catherine Durose, Alison Gilchrist and James Henderson. We are also grateful for the constructive reviews provided by our anonymous referees.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The research underpinning this article was supported by the Joint Programme Initiative Urban Europe/ ERA-NET Cofund Smart Urban Futures programme, ‘Smart Urban Intermediaries: Trans-European Research, Learning, and Action’ (438-16-405). National funders were NWO (the Netherlands), Innovation Fund Denmark (Denmark) and ESRC (UK).

Notes

1 ‘Smart urban intermediaries – connecting people changing communities’ (JPI Urban Europe).

2 Because of the overlap of community and third sector, we use the term NGO to indicate the broader range of organizations that can be understood as such. Appendix includes definitions of the distinguished sectors.

3 Defined as: working in/for public and governmental organizations and services.

4 Defined as: working in/for profit- or commercial organizations or activities.

5 Defined as: working in/for non-profit or community organizations and activities.

6 Defined as: working in/for organizations which are neither public nor private sector. It includes voluntary and community organizations, social enterprises and co-operatives. Because of the overlap of community and third sector, we use the term NGO to indicate the broader range of organizations that can be understood as such.

7 To acknowledge country-specific approaches on registering and/or labelling ethnic categories, we use multiple labels here, sensitive to local situations and differences.

Appendix

Table A1. City and neighbourhood profiles.

City	Amsterdam	Birmingham	Copenhagen	Glasgow
	Around 850,000 inhabitants and a ‘super-diverse’ population; Spatial segregation is structured around socio-economic status, nationality and ethnicity	Britain’s second city with around one million inhabitants. ‘Super-diverse’ population with a growing proportion of young people. Pockets of deprivation and areas loosely segregated by ethnicity	Around 600,000 inhabitants, 25% of the inhabitants of Copenhagen has a non-Danish ethnic background. Share of poor people in Copenhagen increased, with 20% of less affluent Danish citizens living in Copenhagen	Around 500,000 inhabitants. Glasgow residents have a lower life expectancy and poorer health than people from similarly deprived parts of other UK cities, known as ‘Glasgow effect’
Neighbourhood	Osdorp and Slotermeer-Slotervaart	Sparkbrook/Balsall Heath	North-west	Govan
Inhabitants	Approximately 15,000	Approximately 25,000	Approximately 15,000	Approximately 6000
Urban position	Located outside the city ring, part of the Western Garden City structure	Located as two adjacent neighbourhoods on edge of city centre	Located north-west from the city centre	Located on the south bank of the River Clyde
Social mix	Almost 62% of the inhabitants has an ethnic minority background. Mix of old working class and newcomers ‘The Western Garden Cities are mostly negative in the news […] In a social perspective, the Garden Cities often branded as ghetto. But its variety is very large’ (Nio, Reijndorp, and Veldhuis 2008).	Diverse ethnicities, place of incoming migration and settlement, almost 67% has an Asian background	Almost 35% of the inhabitants has an ethnic minority background. Mix of old working class and industrial area, attractive to young students, artistic and creative citizens due to relatively low housing costs	Almost 16% of the inhabitants has an ethnic minority background. Site of former shipbuilding industry, although most of it has been shut down. Almost 9% of social enterprises located in Govan
Socio-economic indicators	It contains ‘above-average amount of inhabitants in a vulnerable socio-economic situation. […] Tensions between groups are therefore sometimes higher than in other parts of Amsterdam (Municipality Amsterdam 2017, 3).	Most deprived area of Birmingham, with high rates of unemployment, low educational attainment and skills, poor health, characterized for high crime levels	Area in Copenhagen with the lowest mean income	Govan is one of the ‘5% most deprived areas in Scotland’ (Scottish Index of Multiple Deprivation 2004)

Table A2. Cohort overview.

Full cohort
Role
Entrepreneur	19%	(8)
Volunteer	10%	(4)
Professional	44%	(18)
Combination of roles	27%	(11)
Gender
Female	46%	(19)
Male	54%	(22)
Age band
21–30	5%	(2)
31–40	25%	(10)
41–50	41%	(17)
51–60	19%	(8)
61–70	10%	(4)
Organization
Public sector^3	17%	(7)
Private sector^4	1%	(1)
NGO – Community sector^5	59%	(24)
NGO – Third sector^6	23%	(9)
Ethnicity^7
Dutch	15%	(6)
Danish	19%	(8)
White British	27%	(11)
Moroccan-Dutch	10%	(4)
Pakistani	10%	(4)
Other	19%	(8)
Active engagement in neighbourhood
0–5 years	19%	(8)
5–10 years	31%	(12)
10–15 years	23%	(9)
15–20 years	10%	(4)
20–25 years	15%	(6)
25–30 years	1%	(1)
>30 years	1%	(1)

Table A3. Full cohort of local change agents.

	Role	Gender	Age band	Position	Organization	Ethnicity	Active engagement
	Amsterdam, Nieuw-West
1	Entrepreneur	Male	41–50	Artist and founder of local community meeting place	Community sector organization	Dutch	5–10 years
2	Entrepreneur	Male	41–50	Director of social enterprise for active citizens	Community sector organization	Moroccan-Dutch	10–15 years
3	Volunteer	Female	61–70	Senior resident and coordinator of organization for seniors	Third sector organization	Dutch	5–10 years
4	Professional	Female	61–70	Director of organization on women’s emancipation	Community sector organization	Dutch	10–15 years
5	Volunteer	Female	31–40	Resident organizing community disco	Third sector organization	Dutch	5–10 years
6	Entrepreneur	Male	41–50	Professional theatre director of community theatre company	Community sector organization	Dutch	10–15 years
7	Professional	Female	31–40	Director of organization on women’s emancipation	Community sector organization	Moroccan-Dutch	5–10 years
8	Entrepreneur	Male	31–40	Young co-founder of youth network	Community sector organization	Moroccan-Dutch	5–10 years
9	Professional	Male	51–60	Welfare and outreach worker	Community sector organization	Greek-Dutch	15–20 years
10	Professional	Male	41–50	Welfare and outreach worker	Community sector organization	Moroccan-Dutch	15–20 years
11	Entrepreneur	Male	51–60	Founder and director of social enterprise	Community sector organization	Dutch	20–25 years
	Copenhagen, Nord Vest
12	Professional	Female	41–50	Outreach worker on green projects	Public sector	Danish	5–10 years
13	Entrepreneur	Female	31–40	Business Owner of local food shop and catering enterprise	Private sector	Danish	0–5 years
14	Entrepreneur/volunteer	Male	41–50	Resident on non-profit and for-profit cultural activities	Third sector	Danish	5–10 years
15	Entrepreneur/volunteer	Male	31–40	Resident outside the neighbourhood on cultural activities	Third sector	Danish/Greenland	5–10 years
16	Professional	Female	31–40	Outreach worker lives in the neighbourhood	Community sector organization	Danish	0–5 years
17	Professional	Female	51–60	Senior resident on local safety issues	Community sector organization	Danish	25–30 years
18	Volunteer	Male	31–40	Resident on local technology lab	Third sector organization	Danish	5–10 years
19	Professional/volunteer	Male	21–30	Local council member active in neighbourhood activities	Public sector	Danish	5–10 years
20	Professional/volunteer	Female	61–70	Senior resident and local council member	Public sector	Danish	10–15 years
	Birmingham, Balsall Heath
21	Professional	Male	41–50	Manager of a local youth centre	Community sector organization	Yemeni	10–15 years
22	Volunteer	Male	41–50	Active in mosque and in welfare initiatives	Community sector organization	Pakistani	0–5 years
23	Professional	Male	41–50	Community worker on neighbourhood and crime prevention programmes	Community sector organization	Pakistani	20–25 years
24	Professional	Male	41–50	Director on crime, safeguarding and sexual health projects	Community sector organization	Asian British	15–20 years
25	Entrepreneur	Female	31–40	Young activist organizing street parties	Third sector organization	Black British	0- 5 years
26	Professional	Male	51–60	Executive of a local community development organization	Community sector organization	White British	10–15 years
27	Professional	Female	41–50	Community development worker on cohesion projects	Community sector organization	Pakistani	10–15 years
28	Professional	Female	51–60	Youth worker and trustee of a local community project	Community sector organization	African Caribbean	10–15 years
29	Professional	Female	41–50	Health development worker running fitness and well-being projects for women	Public sector organization	Pakistani	20–25 years
30	Entrepreneur	Female	41–50	Social entrepreneur on community cohesion in public places	Community sector organization	Middle Eastern	5–10 years
31	Professional/entrepreneur	Female	51–60	Artist, entrepreneur and activist	Third sector organization	White British	20–25 years
32	Professional/entrepreneur	Male	61–70	Political activist and local councillor	Public sector	Scottish/Irish	20–25 years
	Govan, Glasgow
33	Professional	Male	31–40	Coordinates social inclusion and learning hub	Community sector organization	White British	0–5 years
34	Professional/entrepreneur	Female	41–50	Coordinates local community hub	Third sector organization	White British	0–5 years
35	Entrepreneur/volunteer	Male	51–60	Managing Director of community-led development organization and local activist	Community sector organization	White British	>30 years
36	Professional	Female	41–50	Public servant and coordinates local regeneration activities	Public sector	White British	10–15 years
37	Professional/entrepreneur/volunteer	Male	51–60	Outreach worker, local activist and leader	Community sector organization	White British	20–25 years
38	Professional	Female	41–50	Public servant coordinates local community planning activities	Public sector	White British	0–5 years
39	Professional	Male	21–30	Director	Third sector organization	White British	0–5 years
40	Professional/entrepreneur	Male	31–40	Coordinates community enterprise and working hub	Community sector organization	White British	5–10 years
41	Entrepreneur/professional	Female	41–50	Coordinates community participation organization; community advocate	Community sector organization	White British	15–20 years