Community leaders as intermediaries: How everyday practices create and sustain leadership in five informal settlements in Cape Town

Leadership-as-practice

L-A-P marks an emerging concept and research lens within leadership studies, with many synergies with the post-heroic paradigm (Collinson, 2018). Whether L-A-P constitutes a subset of the post-heroic paradigm or a distinct approach is not our concern here (for more on this debate, see Collinson (2018) and Raelin et al. (2018)). But L-A-P is still in the process of conceptual articulation, operationalisation and debate, with scholars acknowledging limited empirical research from different contexts (Kempster and Gregory, 2017; Raelin et al., 2018). Influenced by the so-called practice turn within social and organisation theory (Carroll et al., 2008; Schatzki et al., 2001), L-A-P focuses on how leadership emerges within the continual flow of ‘practice’, where the world is understood as a processual, recurrent accomplishment (Kempster and Gregory, 2017; Simpson, 2016). It also explores leadership as ‘a set of practices’ (Crevani and Endrissat, 2016: 23, 31) through which things ‘get done’ (Raelin, 2016b). This includes not only mundane, everyday, habitual practices (Carroll et al., 2008; Raelin, 2011), but also improvisations that unexpectedly disrupt and reorient the flow of activity (Simpson, 2016). The practice lens, therefore, illuminates how practices shape and are shaped by context (Endrissat and Von Arx, 2013; Fisher and Robbins, 2015). By decentring the individual as the source of agency and action, studies in L-A-P unearth contextualised ‘patterns of connected action’ (Crevani and Endrissat, 2016: 25).

Emphasising the ‘doings’ rather than the ‘doers’ of action (Simpson, 2016), L-A-P is less concerned with identifying patterns of individual practices, which may reinforce the usual assumptions around individual influence and control (Ford, 2016). Raelin (2016b), therefore, describes L-A-P as ‘intrinsically collective’ and a ‘collaborative agency’. It is about what people accomplish together, thereby challenging the hierarchical relation between an active leader and passive followers. At the same time, however, Raelin (2016b: 4) recognises that, depending on the situation, certain individuals may still be important and ‘leadership may emanate from the actions of particular individuals’. As organisation scholar Nicolini (2013: 42) explains, practice theory enables understanding of how individuals are ‘carriers of the system’, not only reproducing it in practice, but also potentially introducing variation and change. There is not a single, unified practice theory to explain this dynamic, however (Nicolini, 2013). With regard to L-A-P, Youngs (in Raelin et al. (2018)), argues against a strict dichotomisation between the individual and collective (or heroic and post-heroic approaches), which may produce an oversimplified understanding of how practice unfolds. (That the heroic and post-heroic approaches make different ontological assumptions is beyond the scope of this article, but for a comprehensive discussion on this point; see Uhl-Bien and Ospina (2012).) The pertinent point for us is twofold. Firstly, according to Crevani and Endrissat (2016), many empirical studies in LAP still focus on formal leaders, including their practices as well as motivations/intentionality, and this is a shortcoming, they argue. Methodologically, we also focus on individual leader practices in this article, although we view these as an entry point for understanding certain patterns of activity through which things ‘get done’, including producing particular leadership roles. Secondly, the role of individual leaders vis-à-vis a ‘collective’ is precisely where the intermediary role of community leaders comes to the fore, potentially illuminating how power – a key point of discussion within L-A-P scholarship – emerges and operates.

On the one hand, L-A-P lends itself to a perspective of collaborative agency embedded in democratic values (Woods, 2016), precisely because it moves away from a leader–follower hierarchy. On the other hand, L-A-P is ‘not of necessity or by definition fair and democratic’ (Woods, 2016: 73). As Raelin (2011) acknowledges, practice can also involve contestation, divergence, exclusion and conflict (see also Crevani et al. (2010)). According to Nicolini (2013: 6):

Practices, in fact, literally put people (and things) in place, and they give (or deny) people the power to do things and to think of themselves in certain ways. As a result, practices and their temporal and spatial ordering (i.e. several practices combined in a particular way) produce and reproduce differences and inequalities.

For Collinson (2018), however, the emphasis on collective agency and leadership co-construction within L-A-P still risks neglecting how unequal power relations may persist, and marks a weakness of L-A-P vis-à-vis Critical Leadership Studies (see also Fletcher (2004)). Similarly, Ford (2016: 223) argues that L-A-P may miss how ‘long established patterns and institutionalised norms’ that shape social relations and marginalise certain voices may be perpetuated in practice.

Can L-A-P enhance our understanding of how power manifests and is exercised within contextually produced practices? According to Carroll (in Raelin et al. (2018)), there are many layers of power relations to consider. While broader patterns of structural inequality remain significant, L-A-P enables a different perspective on power and ‘looks for power at the point it emerges, becomes visible, or is accomplished by both human and material agents’ (2018: 378). This, she continues, often occurs at the micro level, in subtle dialogical moments and everyday interactions. It is the ‘ambiguous space’ of ‘soft’ power. Simpson (2016: 169) describes it as ‘coactive power’, or ‘power with’, as distinct from power ‘as some “thing” that can be acquired, possessed, and exercised over others’. Power from this perspective is not a property or entity but flows through practices and relations (Simpson, 2016). Still, there seems to be some agreement among scholars that greater attention is needed within L-A-P studies to how power manifests, permeates and operates within contextual practices (Kempster, in Raelin et al. (2018: 374)).

Although the question of power in L-A-P is not the primary focus of this article, it is pertinent to the discussion of community leadership. That community leadership exists outside of organisational hierarchies may seem, on the surface, to lend itself well to a theorisation of L-A-P where notions of collective engagement have been employed. However, the intermediary role of community leaders challenges precisely such a collective, non-hierarchical view of agency. It is to these two aspects of community leadership that we now turn.

Community leadership: from collective action to individual intermediaries

While there are studies of community leadership that focus on individual character traits (Nowell et al., 2016), there are also theorisations of community leadership that resonate with relational, constructionist and practice approaches. Within this subset of literature, community leadership is described as a dynamic and emergent social process (Lauzon, 2017; Martiskainen, 2017), constituted by social contexts, relationships and processes (Schweigert, 2007). This suggests community leadership is not defined by organisational structures and formal leader–follower hierarchies (Martiskainen, 2017). Rather, it is seen as non-hierarchical, involving forms of community organising and collaborative action (Nowell et al., 2016). For Ivory (2008), community leadership involves a web of relations and practices (or ‘networked leadership’) that endures the shifting involvement of individuals and groups. Even in studies that identify specific community leaders, scholars often point towards practices of ‘collective power’, ‘participation’ and ‘collaboration’, as well as the importance of shared knowledge and decision-making (Brail and Kumar, 2017; Ivory, 2008; Lin et al., 2017; Onyx and Leonard, 2011; Welton and Freelon, 2018).

At the same time, studies of community leadership acknowledge individual leaders, whether formal or informal, acting as intermediaries between different stakeholders, interests and institutional systems (Bénit-Gbaffou and Katsaura, 2014; Gambrell, 2016; Ivory, 2008; Martiskainen, 2017; Purdue et al., 2000; Welton and Freelon, 2018). This intermediary role may be practised and felt in a variety of ways. As intermediaries, community leaders may bridge community needs and service provision (Lin et al., 2017), bring collective inputs into government development plans (Brail and Kumar, 2017), translate new information and concepts into local contexts (Martiskainen, 2017) and facilitate network engagements to mobilise resources (Ivory, 2008; Majee et al., 2017). They can do so through a mixture of quite ordinary practices, such as attending meetings of different organisations (Lin et al., 2017). In a study of community leadership in Indigenous communities in Australia, Ivory (2008) describes how such intermediary practices become especially pertinent where major changes occur in the broader social context, such as the arrival of European settlers and missionaries, as well as more recent local government reforms. Despite the collective and egalitarian leadership systems in place, Ivory (2008) observes how individual leaders responded differently, either building new relationships or distancing themselves and their communities from these changes.

In the context of informal settlements in Cape Town, Drivdal (2016) identifies similar practices among community leaders as they get involved in specific activities, some related to the management of municipal services. For instance, she finds that leaders interact with external organisations, participate in meetings and carry out administrative tasks such as distributing information to residents, ‘counting and numbering shacks’ and ‘monitoring the settlements regarding, e.g. waste’ (2016: 283). These tasks often emerge as part of local upgrading projects. Ongoing, everyday activities identified by Drivdal tend to pertain to ‘internal’ matters such as mitigating conflicts between neighbours. Leaders’ ongoing work in service maintenance is not the main focus for Drivdal, however, and it is to these kinds of practices that this article will zoom in.

To make sense of how community leaders take up their intermediary role, and how being situated ‘in between’ informs their practices, Bénit-Gbaffou and Katsaura (2014) employ the Bourdieusian concept of political capital. Because community leaders lack any ‘fixed position of superiority’ (Schweigert, 2007) available to organisational leaders, they must continually work to obtain legitimacy with their constituencies (‘the bottom’) and with institutions such as government (‘the top’) (2014: 1809). The pressure to gain legitimacy in these two directions provides an interpretive frame for understanding how and why leaders take up their roles in the flow of activity and in the context of these different relationships. According to the authors (2014), leaders’ legitimacy with their community depends largely on their representativeness, or the extent to which they are seen to belong, to be ‘one of us’ and able to speak on behalf of local interests. It also depends on their ability to deliver and resolve individual and/or collective issues. This ability derives from leaders’ local knowledge and political skills (e.g. knowing the rules of the game and being able to articulate themselves in public meetings), as well as their visibility and approachability. Similarly, with the ‘top’, community leaders may gain legitimacy through a show of loyalty and support to the institution (i.e. they are perceived as not too radical), and also as ‘entry points’ into their communities. Bénit-Gbaffou and Katsaura (2014) give examples of leaders who are not only able to mobilise residents for a march (a show of legitimacy with the bottom) but also able to get high-level officials to sign a petition or speak at an event (a show of legitimacy with the top).

Unsurprisingly, navigating these upwards and downwards pressures requires leaders to balance conflicting expectations, loyalties and logics (Gambrell, 2016; Purdue et al., 2000; Welton and Freelon, 2018). This can be especially challenging for leaders who are caught between what Massey (2015: 303), in reference to informal settlements in Cape Town, describes as ‘conflicting governmentalities’ between a technocratic, neoliberal government and the organic systems of communities. Buire (2011) confirms this simultaneous ‘upwards’ and ‘downwards’ pressure on leaders, finding the Cape Town municipal system and daily neighbourhood life constitute two distinct yet ‘cohabiting’ forms of legitimacy. The intermediary role thus requires leaders to act at the intersection of different social systems, norms and patterns of activity. Stated otherwise, through the concept of legitimacy, we can begin to make sense of how leaders navigate these different systems and become intermediaries. Drawing this back to L-A-P, we are also interested to explore how power manifests and operates in this process.

Infrastructure maintenance in informal settlements

Informal settlements in South Africa must be understood in the context of apartheid, which forced the black African majority to the urban periphery while purposefully neglecting service provision to those areas (Von Schnitzler, 2008). With over one million households residing in informal settlements (Housing Development Agency, 2013), urban spatial transformation, including socio-economic integration and the provision of quality infrastructure, remains a major task (Cameron, 2005). In Cape Town alone, approximately 146,000 households live in 437 informal settlement pockets (with only 204 officially recognised) (ismaps.org.za, n.d.). Service management and maintenance, therefore, become especially important for the long-term operation and sustainability of available infrastructure. However, the provision of substandard and ‘emergency’ forms of services for unspecified and long-term duration (McFarlane, 2008) exacerbates not only the need for but also the difficulty of doing maintenance. Challenges are commonplace and include, inter alia, the maintenance of sufficient supply of safe quality water; clean and functioning sanitation facilities; public lighting to ensure safety and safe access to other public services; proper electrical equipment to reduce the risk of fire; and roads and storm water drains to reduce the risk of flooding (McFarlane, 2008; Williams et al., 2018).

That there is a role for leadership in service and infrastructure maintenance is not obvious, however. Leadership scholars have commonly distinguished leadership, understood as the articulation of vision, meaning-making and the accomplishment of change, from management, defined as the ordering of operations and task implementation (Kotter, 1990; Madumo, 2012). But there is emerging recognition that operational, motivational and meaning-making practices intermingle in the policy implementation process (Ospina, 2017). In fact, addressing service delivery challenges in informal settlements is far from a routine technical matter where difficulties can be resolved through engineering or management expertise. Govender (2016), for instance, has detailed how service delivery in South Africa constitutes a wicked problem (Rittel and Webber, 1973), especially given the durable spatial form of the apartheid city. Ongoing delivery, maintenance and repair of infrastructure services are also often the primary way in which citizens experience and interact with local government. Given the limited reach of government in informal areas (Drivdal, 2016), this suggests service maintenance offers an important context for the practice of community leadership.

The Heights

The five informal settlements where the research took place, known collectively as The Heights, are located in a diverse ward in the southern region of the Cape Town metro. The settlements are concentrated in an area of the ward predominated by formal housing (18% of dwellings are informal or backyard shacks). The average household size is five people, and unemployment is approximately 42%. According to the community leaders, the settlements were established in the 1970s. Despite the provision of government-subsidised housing in 2005, people from other areas ‘invaded’ and occupied most of the dwellings. Residents struggled for many years to eventually receive basic services of communal taps, toilets and electricity. The settlements vary in size: the smallest comprises 41 dwellings compressed in one street block, while the largest comprises 2000 dwellings across a much larger area.

According to data provided by the city, service levels across the five settlements vary drastically. At the time of the research, some areas reportedly have one communal tap for every three households, with the worst areas having only one tap per 56 households. The city’s target ratios are one tap for every 25 households, and one toilet for every five households (Storey, 2014). For electricity, prepaid meters are available in most households, although a few areas remain without any electricity. Many other challenges persist, with electricity infrastructure regularly failing or being vandalised. There are also major issues with gangsterism, crime and drug addiction. It is in this context that we explore the practices of the 10 community leaders.

Reporting service issues: accessing the city

One of the main tasks of community leaders is to report service issues to the city. These include specific issues, such as when streetlights go out, electricity boxes catch fire or water pipes burst. The city has established a notification system that allows residents to phone or send a text to a central call centre that directs each complaint to the appropriate department. Although intended to depersonalise access to the city, in informal areas it is often through direct communication between officials and local leaders that the system actually becomes effective.

In informal settlements, systems and practices around reporting service faults bring up issues related to local political representation. Although the notification system provides residents and leaders with an alternative to going through the ward councillor, their formally elected political representative, it does not mitigate their frustrations with him, and leaders generally perceive him as unreliable (Workshop 1, 16 March 2015). Such perceptions are not unique to these communities, and research on informal settlements across Cape Town has found citizens are largely dissatisfied, if not distrustful, of local representatives (Thompson and Nleya, 2010). In this case, however, the leaders explain their frustrations in terms of the councillor’s ineffectiveness in reporting problems, which results in residents asking the leaders to report service faults on their behalf. In the excerpt below, two leaders and one of the facilitators try to identify why the city’s response to issues in one settlement is much better than in another:

Facilitator: So, are you saying that you get a better response from the city in Overcome than you are having in Military? And why do you think that is?

In this exchange, the leaders challenge the legitimacy of the ward councillor, despite his formal leadership position, on the basis of a pragmatic need and what they perceive as his lack of support for the community. He does not ‘belong’ in the way that they do because he fails to ‘deliver’ by relaying service faults to the city or following up to make sure the problem gets fixed.

Conversely, the community leaders’ legitimacy with their constituencies emerges through their efforts to monitor for problems, their knowledge of the notification system and their willingness to call ‘with our air time’ (Workshop 1, 16 March 2015). This is evident in the following quote where a leader explains her frustrations with having elections in their settlements every year:

Leader B: Can you see what’s happening? The ward councillor comes in and then he say, okay you have to have an election. And now all of a sudden there’s an election and now she comes on board; I’m out as the chairperson […] At the end of the day this lady doesn’t know anything, she doesn’t have the numbers. Now at the end of the day our community is suffering. That is what happens. (Workshop 1, 16 March 2015)

Leadership is, thus, constructed in the positive image of the leaders who are knowledgeable because they have the number to the call centre. While this number is posted on refuse containers throughout the settlements, the community leaders use this point again and again to show the failures of the councillor and to claim their own legitimacy and commitment to the community.

Although it is possible that the leaders may exaggerate the extent to which residents come to them to report problems (perhaps to gain legitimacy with the researchers), Drivdal (2016) found a similar situation in other settlements regarding the reporting of crime and vandalism. It could, therefore, be that residents go to their local leaders rather than use the notification system precisely because leaders have historically played such an intermediary role in their communities and are by default closer and more accessible. The notification system represents a new governance mechanism that has yet to replace existing and embedded social practices and relations. If this is the case, the city’s communication of this mechanism might not yet be sufficiently effective to transform reporting practices (for similar findings, see Masafu (2015)).

While the call centre’s ‘direct line’ to the city may overcome some of the tensions between the councillor and community leaders, repair of infrastructure is not only simply a matter of reporting but also of responding. In this regard, community leaders play a central role in providing access to the settlements for officials and technicians.

Contacting and meeting officials: accessing the community

Officials who participated in the project raised concerns about their ability to respond to service faults due to logistical and social issues. This included residents not having addresses, or moving their shacks and taking their addresses with them, both of which makes it difficult for technicians to locate reported problems (Workshop 3, 17 March 2015). Among the social challenges, gang violence and the safety risks this poses for officials were a major concern, with some areas considered too dangerous for officials to enter. In instances of major electricity outages, a police escort has been used (Workshop 3, 17 March 2015). Such conditions have direct implications for the repair of infrastructure, as well as for the emergence of an important leadership practice. Once a problem has been reported, officials rely on local leaders to communicate when it is safe to enter, and then to meet and assist them in locating the problem. As one official explained:

In that area, there’s two ladies there who deal directly with us. They will phone us directly and say, ‘Don’t come in, there is a gang war taking place so we’ll let you know when you can come in’. Sometimes we need to just wait for them to get back to us. […] But if they’re not available, then we can’t just go into the area (Workshop 3, 17 March 2015).

From the perspective of the leaders, this is one of their responsibilities: to ensure officials do not get hurt or robbed not only because ‘they are scared’ and ‘they are only human’ but also because otherwise the problem does not get fixed (Workshop 1, 16 March 2015).

In this way, the leaders function as the ‘eyes and ears’ of the city, although they are not simply responding to instructions from the city or performing tasks for the city. Rather, their assistance to the city has become part of retaining their legitimacy with their communities. In other words, the leaders’ ability to deliver to their communities evidently also depends on the city’s ability to respond – to physically show up and fix the problem. At the same time, officials’ ability to respond depends to some degree on the leaders’ legitimacy in their communities in the first place. The leaders can take officials to specific service faults and provide some security for technicians because they ‘belong’; they have established relationships and networks with residents and thus have knowledge of what is going on. By meeting officials and helping them address problems, the leaders can simultaneously showcase their local ‘power’ and relevance, and their access to and ‘power’ with the city, thereby reinforcing their legitimacy with both the city and their communities.

Preventing vandalism: protecting services

A major issue in the five informal settlements, recognised by both leaders and officials who work in the areas, is vandalism and general ‘misuse’ of infrastructure. In one of the city workshops, vandalism was noted as a major concern, with all infrastructure being continually misused and/or broken (Workshop 3, 17 March 2015). These concerns were further reflected in officials’ views that communities should ‘take ownership’ of the infrastructure so that ‘it becomes their taps, it becomes their water, it becomes their infrastructure’ (Workshop 3, 17 March 2015).

The community leaders expressed similar frustrations with a lack of ownership by residents. They described how they address vandalism by phoning and shouting at specific culprits for damaging taps or wasting water, or even phoning the police in cases of theft. These practices can be understood as part of the internal mediating role of leaders, also noted by Drivdal (2016) in her analysis. These are not formally instituted practices but are rather ad hoc and often depend on residents putting pressure on the leaders, or leaders taking notice of certain individuals and behaviours. These practices are likely to generate legitimacy with both ‘the bottom’ and ‘the top’ as residents and officials see the leaders taking care of the shared infrastructure and resources. Indeed, officials who participated in the project reported improved views of the community leaders once they heard leaders share their concerns and try to act on vandalism (Workshop 4, 6 April 2015).

Defending illegal connections: accessing services

Efforts to prevent and address vandalism seem, however, to contrast directly with how community leaders sometimes ‘manage’ illegal connections. In the following exchange, two leaders discuss the fact that water taps are communal and thus erected in the street, yet it is possible for someone to move it into their yard:

Leader A: If a single house can have maybe a R50 or whatever, you can have the tap in your house, to make it convenient for yourself. Because we don’t know how long we’re still gonna live like this. […] Now what we did there, we bought the pipes, we bought our fittings and everything, and take it from the road and...

In this dialogue, the leaders justify illegal connections (some even reported doing it themselves), through claims to individual responsibility for one’s dignity and comfort. This suggests a different kind of legitimacy and sense of community at play, one centred on ‘coming up for each other’. What comes to the fore here is the fact that leaders share with residents the conditions of living in informality – they understand the local reality and the uncertainty of any future housing and infrastructure development. The leaders must, therefore, navigate their belonging to and residing in their areas, and the limits of what they are able to deliver.

It is not clear from the data how prevalent these practices are, or to what extent they might be a source of tension and struggle between residents. It is possible, for example that some leaders might privately take a different stand regarding illegal connections, or that some residents receive preferential treatment. Indeed, it may even be a mechanism for local patronage where the leaders defend those who in turn support them, and possibly report on those who do not. But these practices illuminate some of the contradictions and challenges with which leaders and residents of informal settlements must live.

Community leadership as practice: a tenuous, continuous and contextual practice

The four practices illuminate a certain tension at the centre of community leaders’ intermediary role, also picked up by Bénit-Gbaffou and Katsaura (2014). Practices that produce leadership can complement or contradict one another. Certain practices undertaken in a specific time and place can gain legitimacy for leaders with both the community and the city as their interests and objectives intersect and they find mutual benefit in a particular outcome (such as when service faults are addressed). Some practices may benefit or ‘empower’ only the community or the city, and thereby reinforce legitimacy with the one while undermining it with the other. Alternatively, leaders may do things that strengthen their legitimacy with the city, and do other things that simultaneously weaken or undermine that legitimacy (e.g. addressing vandalism and defending illegal connections). A similar tension applies to their relations with their communities in how they prevent vandalism but allow illegal connections. This suggests leaders’ intermediary roles remain tenuous since they depend on such relational practices with residents and the officials for the ongoing sustenance of their leadership. Stated otherwise, they are never in a ‘fixed position of superiority’ (Schweigert, 2007) or in a single, static power relation, and there is no clear, concrete agenda they must follow in either direction.

This aligns with the extant community leadership research that finds leaders employ a mix of oppositional and collaborative strategies (Oldfield and Stokke, 2007), fulfilling their intermediary role either as champions, opponents or pragmatists (Purdue et al., 2000). We would add that the choice of how they do so is itself a pragmatic one which they continually make. It also resonates with theorisations of L-A-P as produced by the non-linear flow of activity involving ongoing divergences and conflicts (Crevani et al., 2010). Alvesson and Jonsson’s (2018: 51–55) study of leadership in middle manager work similarly depicts leadership practices as multifaceted and ‘episode driven’. It is thus less about a particular set of competencies or behaviours than it is about an ‘appreciation of context’, with resultant adaptive practices that are complex, iterative and even contradictory (Tourish, 2014: 93).

Despite the distinct features of community leadership and community contexts which, we believe, contribute to these dynamics (and discuss in more detail below), we cannot definitively conclude from the evidence that this is unique to the practice of community leadership. There are potential similarities, as noted above, between community leaders as intermediaries and middle managers in organisations (Kempster and Gregory, 2017), or leaders of cross-sectoral or inter-organisational collaboration (Connelly, 2007; Crosby and Bryson, 2010). Middle managers and leaders of collaborative initiatives are also situated in an ‘in between’ space where they must navigate different institutional systems, practices and agendas.

There are still two important ways in which community leadership is likely to differ, however. First is the fact that middle managers and collaborative leaders are likely to be appointed. They would thus be able to rely on their positional power to fulfil their intermediary roles. Although they must also navigate between different interests, they would ultimately have to account to a particular ‘entity’, whether this is their organisation’s or the collaboration’s purpose, or the dictates of top management. Community leaders, on the other hand, are not appointed but rather elected or emerge informally as leaders. It is precisely for this reason that they must work to continually maintain their legitimacy as leaders through pragmatic and relational work. This is complicated by the fact that local contexts, particularly informal settlements, are themselves highly fluid, with shifting members (residents) and agendas. For community leaders, the work is never done, and the questions around ‘leadership for what’ and ‘for whom’ are never fully answered.

Community leaders as intermediaries: the social construction of relations of dependence

Although there may not be a specific organisational context shaping community leaders’ practices, their practices do emerge within the broader context where informality and municipal systems and practices intersect. For informal settlements, limited access to communal service infrastructure means service issues are prominent in daily life. The leaders’ practices can therefore be understood against a more or less stable background of community practices through which people share basic services. This includes social practices around reporting service faults, but also vandalising infrastructure or making illegal service connections.

In the midst of these practices, the city has introduced a new reporting system that reorients the ‘flow of activity’ (Kempster and Gregory, 2017), and through which individual leaders and officials (and the relations between them) gain importance and influence. The introduction of the call centre and notification number in particular has shifted power in the system away from the ward councillor’s structural or ‘positional power’ (Bass, 1990), towards individual residents and community leaders. This is an intentional aim of the system – to provide equal access for residents across the city. But many informal settlement residents continue to rely on local leaders to report on their behalf. The system and practices around reporting thus reinforce the relevance of individual leaders, and the power asymmetries between leaders and residents as leaders can control these information flows (McCauley et al., 2015). But these reporting practices also reflect socio-historical patterns where local leaders have been expected to take on such intermediary roles in the past (Drivdal, 2016).

This practice and relation of dependence are not simply an outcome of leaders’ intentional efforts. They are also outcomes of the realities of living in informality, and the inadequacies of the reporting system in such contexts: the call number is not well-known; providing the precise location of a service issue can be difficult; and there are no guarantees that officials will be able to enter and/or find the issue. Managing these challenges in order to ‘get things done’ is also not only achieved through the practices of individual leaders. Residents report to local leaders (as part of established patterns of behaviour in informal settlements). Leaders bring their local knowledge and understanding of how the municipal systems work, as well as their contacts with officials. And officials ‘play their part’ by trusting and responding to local leaders and bringing their technical knowledge.

Through the lens of L-A-P, community leaders’ intermediary practices are thus part and parcel of a broader array of intersecting practices. Leaders emerge as intermediaries, but as part of a broader kind of collective agency around service maintenance goals. Although not quite a non-hierarchical collaboration where efforts neatly coalesce and people act either democratically or in unison, it is also not simply a product of power ‘from the top’, or even ‘from the middle’. Power permeates and shifts through these practices, while the practices give expression to and reinforce existing relations of dependence and interdependence.

This dynamic manifests, for instance, in the recursive relation between the agency and legitimacy of the leaders and that of officials produced in this context. Reporting problems and meeting officials have emerged as opportune ways for the leaders to deliver, gain legitimacy and, thereby, maintain their roles as intermediaries. Yet the ability of both leaders and officials to address service issues – an objective they arguably share – is intertwined and depends on each other’s ability and legitimacy. Leaders depend in their reporting practices on the response from officials, and officials depend on community leaders in meeting their service maintenance goals.

Intermediary practices: pragmatic improvisations to find what works

In the informal settlements in our research, the leaders’ practices in the management of services display a particularly supportive role insofar as they help both residents and the city in local governance processes. They assist residents to ‘penetrate’ local government, which is often seen as distant and cumbersome. They also assist officials to ‘penetrate’ the community, usually perceived as amorphous and disorganised (Bénit-Gbaffou and Katsaura, 2014). In doing so, however, their practices also maintain the distance between the city and community, again (re)producing the dependencies that sustain their roles as intermediaries.

At the same time, however, local leaders remain situated in and shaped by broader inequalities that define the relation between informal settlement communities, on the one hand, and the state, on the other. In this sense, the four practices illuminate how community leadership emerges in the context of broader governance systems, where leaders confront the limits of their ability to deliver. They must navigate the difficulties of belonging, in the face of their inability to secure ‘real’ transformation. In response, leaders work to improvise and find ‘what works’. Some of their practices support the functioning of the formal system. Contacting and meeting officials, for instance, although an informal and emergent practice, has also become integral to the effective operation of the city’s maintenance system. Their practices around making and supporting illegal connections, on the other hand, implicitly challenge the formal governance system. Although beyond the scope of this article, illegal service connections have been used by social activists and communities across South Africa to challenge service delivery policies like full cost-recovery (Everatt et al., 2010). Such acts have been described as ‘officially illegal yet morally appropriate’ acts of citizenship in the struggle for recognition in post-apartheid South Africa (Everatt et al., 2010: 224). The practices among local leaders in our research could, therefore, again be understood as part of broader, ‘collective’ forms of agency, emerging on the margins of what is possible for community leaders working as intermediaries within the formal system.

Taken together, community leadership practices involve the pragmatic working out of local possibilities, which can have the effect of mitigating one power differential while creating or reproducing another. These practices involve ongoing, pragmatic improvisations that emerge through relational and collective patterns of interaction. The analysis further suggests L-A-P can both reinforce and challenge power differences and established ways of doing things, thus encompassing the dynamic interplay of power relations at multiple levels.