Making Demands on Government: Theorizing Determinants of Backyard Residents ’ Collective Action in Cape Town, South Africa

: Informality is growing with Africa ’ s rapid urbanization. Much like residents of other types of informal housing, backyard dwellers face overall poor living conditions and political marginalization. However, backyard residents are in an ambiguous legal area and have been far less politically active and organized to pursue their rights to adequate housing. Using a qualitative case study of backyard residents in three Cape Town neighborhoods, Harris, Scheba, and Rice bridge theories of infrastruc-turalcitizenshipandcollectiveactiontoshedlightonhowinformalitymayunderminecollectiveaction,andtheyidentifyfourfactorsin ﬂ uencing collective action.


Introduction
In May 2018, backyard residents of a neighborhood in Cape Town, South Africa, protested their lack of adequate housing by blockading and thus shutting down the M5 freeway, an important artery connecting the southern suburbs to the city's urban core.The predominantly poor insurgent citizens (Holston 2009) were demanding government action to improve their access to dignified housing and services, voicing concerns over their poor living conditions and increasing rental, water, and electricity costs.Since then, "backyarders" have organized a few more protests in Cape Town and other cities in South Africa (Maregele 2019;IOL 2019;Hendricks & Washinyira 2019).While these incidences of community mobilization have succeeded in bringing the plight of backyard tenants into the public focus, organized collective action-especially that which goes beyond protests toward actively engaging with government-remains rare among backyarders.Compared to residents of other informal settlements, backyard residents-the largest tenant population in South Africa-have been far less politically active and collectively organized to improve their specific living situation (Isandla 2021).In this article, we investigate why backyard residents do not collectively organize at higher rates to demand improvements to their current living conditions compared to other poor and informally housed residents.
We investigate this puzzle and develop a theory to better understand the conditions under which collective action among backyard residents is more or less likely to occur.In doing so, we aim to advance knowledge on collective action among informal tenants more generally, given that they are a growing population group in African cities and elsewhere in the world.Informally constructed secondary dwellings on formal plots are common in countries as diverse as Chile, Haiti, India, Mexico, Peru, Brazil, Thailand, and Australia (Scheba & Turok 2020;Shrestha et al. 2021;Baqai & Ward 2020;UN Habitat 2003).Due to the housing shortages prevalent in most African, and indeed southern, cities, poor residents commonly rent informally constructed dwellings from small-scale landlords in low-income neighborhoods, as observed in urban Ghana, Kenya, and Nigeria (Mwau & Sverdlik 2020;Melzer et al. 2018;UN Habitat 2003;Arku et al. 2012).These dwellings are frequently hidden, sometimes sharing the urban space with other tenants and/or landlords.Often occupying a legal grey area, the residents of these spaces face considerable constraints in engaging with the government either individually or collectively, which in turn affects their social, economic, and political capacities to improve their wellbeing.Their dependence on landlords and informality likely lowers their agency and limits their ability to reshape their situation to their advantage in ways that might make their place in the urban space more permanent and desirable (e.g., as Paller [2019a] argues in the Ghanaian context).Like the backyarders in Cape Town, others renting in informality across the continent live in the forgotten shadows of urban politics and policies.
Bringing together literature on informal housing, infrastructural citizenship (Lemanski 2019) and collective action (e.g., Ostrom 1990), this article analyzes how these groups mobilize themselves to collectively engage with the state over housing and infrastructure provision.Drawing on qualitative fieldwork in three neighborhoods in Cape Town-including the neighborhood mentioned above, referred to as S3 below-we identify the key factors influencing collective action: diversity among backyarders, the promise of a house, the degree to which a backyarder occupies a legal grey area, and the presence of a charismatic leader.Our study builds on Jeffrey Paller's (2019b) work, which finds that a more accountable politics is possible when citizens 1) can speak and are listened to, 2) share bonds of respect with their representatives, and 3) come together to make demands or express needs.While we observe that all such conditions are lacking for backyard residents in our case of interest, we focus and elaborate on the latter condition.We believe that our findings are relevant to understanding collective action among other types of marginalized populations living in informality and/or other forms of legal ambiguity, such as immigrants, racial or ethnic minorities, and informal workers. 1 We also build on past work that finds that, often, those in the most need are the very same citizens who are the least empowered to make demands on the government (Carlitz 2019).In short, this article uses the case of backyard tenants in South Africa to advance scholarship on infrastructural citizenship, collective action, and informality.

Infrastructural Citizenship and Collective Action
Our approach brings together literature on "infrastructural citizenship" and collective action as a conceptual frame.The concept of "infrastructural citizenship" has been developed by Charlotte Lemanski (2019Lemanski ( , 2020) ) as a broad term and analytical framework "to highlight how citizens' everyday access to, and use of, public infrastructure in the city affect, and are affected by, their citizenship identity and practices" (Lemanski 2020:589).Lemanski coined the term "infrastructural citizenship" to emphasise the critical role of infrastructure (including housing) in shaping, mediating, and revealing state-citizen relationships in cities.The fundamental premise of the concept is that infrastructure, which is a socio-political process as much as it is a physical artefact, is inherently political and therefore provided, maintained, and transformed through political practices.While protests and political organizing are key citizenship practices to influence infrastructure provision and access, "infrastructural citizenship" highlights the crucial role residents' identities, perceptions, and everyday practices play in shaping infrastructural relationships (Lemanski 2019).Importantly, the concept of citizenship has been expanded from a narrow, legal status of rights and responsibilities toward a broader, more dynamic, practice-orientated concept that focuses on the acts of all urban dwellers-ranging from "radical protests" to mundane everyday practices and relationships (Staeheli 2010).
Studying physical infrastructure networks in their socio-geographical contexts allows us to understand better the relationships between citizens and the state.In fact, it renders these power-laden political relations visible at the everyday temporal and spatial level of the household and settlement (Lemanski 2019).While studies have shown how individual citizenship practices play an important role in transforming public infrastructure provision and access (McFarlane & Silver 2017;Caldeira 2017), scholarship on collective action suggests that the ability of communities to come together in order to agitate for their needs is an important tool to promote development of and provide public goods to underserved communities (Magaloni, Díaz-Cayeros, & Ruiz Euler 2019;Baldwin 2016;Ostrom 1990;Habyarimana et al. 2009).Often these communities lack fiscal capital but can use their social capital to leverage the weight of the people's voice to obtain much-needed services and improvements to their livelihoods.However, one key challenge to such collective action is overcoming the free rider problem (Ostrom 1998). 2 Past research has found a number of possible sources of effective collective action that overcome the free riding problem, including mutual dependence facilitated by institutions (Tsai 2007;Harris & Honig 2023), trust (Ostrom 1998), repeated interactions and communication (Axelrod 1984), and networks and social ties (Habyarimana et al. 2009;Miguel & Gugerty 2005).Literature on the role of leaders in collective action suggests they can help overcome the free rider problem by providing a focal point for the community on what actions are expected (Glowacki & von Rueden 2015), they can set and enforce norms of participation (Calvert 1992), and, when trusted, can more effectively gain compliance (Lim et al. 2021).
Bringing these two theories of infrastructural citizenship and collective action together sheds new light on our understanding of collective action in contexts of informality.Infrastructure citizenship holds that actions seen as illegal or informal by the state (e.g., a backyard resident connecting their shack to the main house's electric supply) are acts of "citizenship-in-practice," or ways for backyard residents to assert their membership in and belonging to the city (Lemanski 2020).Thus, we can see collective action among people living in informal backyard dwellings not only as a way to demand services and adequate housing but also as a demand for political inclusion in the urban environment.Backyard residents in South Africa occupy a legal grey area in which the state's role in service provision is not entirely clear, and thus backyard residents must first ensure their citizenship and rights to services as well as their access to needed services.Collective action among backyard residents, from this perspective, is difficult because it has two interrelated goals to meet: 1) political membership in and consideration by the state and 2) access to services.According to infrastructural citizenship, access to services is a way to access full membership in urban societies.Thus, backyard residents need to overcome both the free rider challenge of collective action and the limits that access to services places on their urban citizenship.
Therefore, in our study of those in backyard dwellings (BDs) in Cape Town, we look at both citizenship and free rider challenges to collective action and how they interact with and compound one another.For example, we investigate how the limited and state-perceived illegality of access to services limits a sense of belonging to the city, which can in turn dampen residents' willingness to engage in making demands on government via collective action.Importantly, we consider how living on private versus public land conditions one's sense of citizenship and thus relations with the state.Further, we investigate key limitations to effective collective action, including: social ties/diversity (Miguel & Gugerty 2005;Humphreys & Weinstein 2008;Calvert 1992) and leadership (Rodriguez et al. 2021;Lobo et al. 2016;Kosfeld & Rustagi 2015).And finally, we briefly consider how these citizenship and coordination challenges and/or opportunities interact with and reinforce one another.

Study Context and Design
The right to adequate housing was enshrined in South Africa's democratic constitution of 1996 and has been a key pillar of the country's post-apartheid governance.Numerous policies, programs, and initiatives have been implemented-such as the Reconstruction and Development Programme (RDP) and Breaking New Ground (BNG)-which have provided more than 4.7 million homes.However, the population of homeless and inadequately housed people has mushroomed.The estimated housing backlog across the country is 2.6 million units (Thukwana 2020).Entrenched poverty, unemployment, and exclusion from private markets, coupled with declining delivery rates of public housing, has pushed millions of people into informality.BDs have grown rapidly in the townships of South Africa's larger cities, where asset-rich but income-poor homeowners seize the opportunity to generate income from renters (Scheba & Turok 2020;Gardner & Rubin 2017;Lemanski 2009).As such, these areas accommodate the largest tenant population in the country.The growth of BDs has resulted in the proliferation of makeshift structures, with associated health and safety risks, growing pressure on public infrastructure and services, and exposure to exploitative rental relations (Isandla 2021; Lategan et al. 2020).Despite this growth of BDs, compared to residents in other informal settlements, backyard dwellers have received far less attention in public discourse and policy until recently (Isandla 2021; HSRC 2019).
Cape Town, the site of this research, has experienced significant growth of backyard dwellings over the last decades (Turok & Borel-Saladin 2016).According to official statistical data, the number of informal backyard dwellings in Cape Town increased by 256 percent, from 21,780 in 1996 to 77,630 in 2016 (City of Cape Town 2021).The growth of BDs has almost exclusively occurred in townships of the city's urban periphery-the Cape Flats (Scheba et al. 2021).The growth of backyarding has resulted in a "re-informalisation" of formally established low-cost housing settlements (Robins 2002), which were built either during apartheid or as part of the government's massive housing program which was implemented after 1994.Cape Town has seen more backyarder-driven protests and collective action than other cities in South Africa. 3The peak of this collective action occurred in August 2019 under the leadership of the "Gatvol" Capetonian movement, when hundreds of backyarders from several communities (including our three case study sites) simultaneously demonstrated in public, shutdown key roads by burning tires and rubble, and gathered collectively at the civic center to deliver a memorandum to the mayor of Cape Town (Hendricks & Washinyira 2019).Besides raising awareness, these protest actions had limited impacts on policy and the lives of backyarders.Less than a year later, the Gatvol Capetonian movement was back in the media spotlight, announcing another "shutdown," as "nothing has been done to improve the lives of backyarders" (Mitchley 2020:1).
Collective protest actions have attracted the attention of the government, but more sustained and engaged citizenship actions will likely be needed to shift policy and practice.To this end, we interrogate collective action in three "backyarder" neighborhoods in Cape Town, which will be briefly described below.The three neighborhoods where empirical data were collected are all located in the Cape Flats area and are largely representative of lower-income settlements that have experienced a substantial presence of backyard dwellings.We selected three neighborhoods that varied in their backyard density, socioeconomics, and (observable) effective collective action. 4We conducted three focus groups (one in each neighborhood, an hour and a half each) and between two and four interviews (ten interviews in total) and eight site visits, in which we went to selected focus group respondents' BDs in order to understand their living conditions and efforts for improvement.The age range of our participants was between 18 and 65, with an average age of 35.In terms of racial demographics, the participants from S1 were almost exclusively Black, 5 while the participants in the other two locations were almost exclusively Coloured.Both the focus groups and interviews were roughly equally divided between men and women.In total, the research involved 44 respondents across the three communities (26 of whom were women, 59 percent). 6While the empirical focus of this paper comes from these focus groups, interviews, and site visits, this research is also informed by the authors' longstanding relationships with these communities and research into BDs in Cape Town over the past eight years.We describe each community in turn, and Table 1 summarizes the key characteristics of each.
Site 1 is 25 kilometers southeast of Cape Town's city center and next door to its international airport.It sits at an important public transportation junction, which makes it appealing to young professionals who need cheap accommodations and an easy commute into the city center, and has led to a remarkably high concentration of private investments in backyard dwellings.In fact, our focus group participants in site 1 suggested that it is difficult to find a plot in site 1 that does not have at least one backyard dwelling.S1 is also racially and ethnically diverse, made up of Black and Coloured residents.While backyarders have organized loosely to support each other and agitate for their rights, the community organization is weak and, likely due to the high density and mobility of the residences, social cohesion is quite low.Backyard dwellings are mainly located on private land of state-subsidized houses.
Site 2, a neighborhood located a few kilometers south of S1, represents a more established and homogeneous backyard housing context; it is less dynamic and poorer than S1, with more ethnic homogeneity and a majority Coloured population.Unlike S1, where we've seen a growing The final study location was Site 3, which is directly south of central Cape Town (and West of S2).Site 3 represents the most socioeconomically diverse context of the three and is moderately racially diverse (less diverse than S1 but more than S2), but it is a majority Coloured area.As was the case with S2, many backyard tenants are related to or friends of the main households, but there is also an important share of commercial rental activity taking place in the sector.The backyard dwellers are very well organized, led by a charismatic, energetic man.The shacks in S3, as in S2, are on a mix of private and council land.Unlike the other locations, S3 has been part of a municipal pilot program for shacks on council-owned land, in which the Cape Town Municipality has provided basic infrastructure specifically for backyard shacks, including toilets, water, and electricity facilities/access.S3, we find below, is the site that has the conditions best suited for successful collective action.
These three cases present a diverse set of contexts that provide a rich setting in which to explore backyard residents' infrastructural citizenship practices, especially collective action, to make demands on government for better living conditions.These similar low-income contexts, which vary in terms of socioeconomics, racial demographics, density of backyard housing, and backyarder organization, provide sufficient variation to engender confidence in their representativeness of such communities in the larger Cape Town municipality.

Analysis and Results
Backyarders face an uphill battle when making demands on the state that could improve their housing and living conditions; before exploring the conditions under which collective action is more likely, we highlight two important barriers that constrain "infrastructural citizenship" practices for all backyard communities.First, the informal nature of most backyard dwellings presents a clear obstacle to claiming citizenship and making demands on the government.Informal backyard dwellings typically transgress land-use planning and building regulations.This state of informality casts a shadow over all efforts to engage government as citizens in demand of better service delivery.Backyard dwellers, who typically live hidden in the background, face uncertainty and risk when exposing themselves to the state.As explored below, the degree of vulnerability varies among backyard residents, and the outcomes of state engagement are less predictable.
Second, and relatedly, high poverty rates, unemployment, and precarious livelihoods make it difficult for backyard dwellers to expend the time and energy needed to practice engaged citizenship and collective action.Many backyard dwellers struggle to make ends meet, and many of them require at least some income to pay for the monthly rental.As a respondent in S1 indicated, "If you come to a community and tell them about the right to protest for instance and the people are not really interested because they've got other issues; unemployment, housing issues, water and sanitation and all of that." The informal status and high poverty levels hinder backyarders from practicing "infrastructural citizenship" in all neighborhoods.However, we find that in addition there are at least four other factors that influence their citizenship practices, especially collective action.In the following analysis, we identify respondents with codes that indicate the respondent's community (S1 = Site 1, S2 = Site 2, S3 = Site 3) and allows readers to see the diversity of respondents we directly quote (the numbers are given for each respondent, for example S3-6 is the sixth respondent in our focus group in site 3).

Backyard Residents: A Socioeconomically Diverse Tenant Group
Backyard residents share a common housing situation-they all live in the backyard of somebody else's home.Aside from this, however, they are a socioeconomically diverse group of people (Isandla 2021).Backyarders consist of young and old residents, South Africans and foreigners, unemployed and employed, and people with different income levels.Some of them are renting accommodation from a landlord, others from their parents or extended family.Some are primarily backyard residents in order to save money to invest in a home of their own elsewhere, while others have nowhere else to go.It is hard to coordinate and unite people from diverse class, racial, and socioeconomic backgrounds around a common claim (Piven & Cloward 2012).These groups not only have different priorities and opportunities, but they also have different demands on their time, which influence their infrastructural citizenship practices.
While employment situations and family obligations were clear themes in our focus group discussions, the different circumstances faced by immigrants and different racial groups in South Africa were not discussed.As such, we likely underestimate how social divisions undermine collective action, which only suggests that social divisions within the backyard community more broadly likely have an even larger impact than we can uncover here.
An important socioeconomic subgroup of backyard tenants is made up of single, young, and employed adults, who may see their current home or neighbourhood as a place of transition (HSRC 2019).They are busy making money to enable the purchase of a house elsewhere.This makes them less likely to engage in collective action to improve their backyard housing situation.Even if they were supportive of collective action, their demanding employment conditions may prevent them from participating.Backyarders who were employed expressed how they work long hours (e.g., working from 4 p.m. to 9 a.m.[S2-5]; or having to leave their homes as early as 5 a.m.[S2-7 and S2-14]).Their jobs, while often poorly compensating, are demanding and time intensive, which leaves little time for much else, including community participation.However, a job is a job, and respondent S1-1 indicated that when people, even leaders of the community organizations, get jobs, they step back from community advocacy work because of constraints on their time: "And then someone else gets a job, …when that person gets a job, they don't have time now to be part of the struggle … when [people find jobs], obviously now people start to back off."Material needs and time-intensive jobs make participation in collective demand-making difficult for some BD residents.
Furthermore, the employment status of BD residents is often shifting and unstable, which has key implications for involvement in community organizing.Our conversations with backyarders suggest that not only does employment prevent people from having time to engage in community efforts, as previous research has indicated, but the instability of employment also 1) makes it difficult for these organizations to have consistent and reliable participation and momentum and 2) prevents residents from putting down roots and becoming part of the community.
Another important socioeconomic group among backyarders consists of families.They may be more interested in improving the physical fabric of their living conditions, given that they often live in cramped spaces (e.g., a family of four in one room) and are less transient.There are real concerns over the health of children and the elderly.However, while needs and desires for change are high among this population, they also lack the time and ability to organize and agitate for change, as they struggle to make ends meet.Importantly, respondent S1-2 highlights an additional challenge for parents: living with children exacts a heavy emotional and mental toll, which can often be debilitating.As one parent indicated, "As a parent you feel like I'm failing my kids.Remind yourself, and say, if only I can give them better or if only I can do that.… I have suffered and I'm still suffering from depression … you can't assist your children, it's just never ending" (S2-4).The pressure to provide and the perceived failing as a parent has clearly weighed on respondent S2-4.Overall, children make living in a backyard dwelling more complicated.As our respondents indicated repeatedly, something as simple as trying to cook or children playing in the yard can bring down the wrath of the landlord: "Our children are not even free" (S1-9).
In short, like the poor elsewhere, backyard parents' struggles to work out short-term solutions to improve living conditions for their children and the debilitating emotional and mental stress of living with children under these conditions limit their ability to actively agitate for change at the community level.While they potentially have the most to gain from community organizing, they lack the time, ability, and resources (emotional, material, and civic) to do so.This is not uncommon for marginalized groups.For example, women in Tanzania are primarily responsible for water provision, have the most to gain from improvements, and yet are less likely than men to act to improve water provision (Carlitz 2019).In addition, those who are employed have less time to dedicate to community activism, and unstable employment prevents consistent engagement with community organizations.While the resulting lack of time and/or energy to engage in collective action is the same, the source of these limitations differs across the employed and parents.

An Escape Hatch: The Promise of a House
In addition to socioeconomic divisions, the government's promise of a house makes many backyard residents, especially those who are eligible for state support, less willing to engage in collective action for in situ improvements because it presents an option to "exit" (Hirschman 1970) the politics of backyard housing.In all our conversations with backyard dwellers, it became clear that the ultimate objective was to live in one's own property.Receiving one's own home-specifically in the form of private property-is viewed as central to becoming a citizen in South Africa (Lemanski 2019).While not all backyard dwellers are eligible for government support, for those who are, their political efforts are targeted toward receiving a free government house.They may join collective action to demand free houses, but they are less interested in agitating for upgrades to backyard dwellings.Because of the declining delivery rates, huge backlogs, and increasingly narrow definition of beneficiaries, many years may be spent "waiting" for a free house (Oldfield & Greyling 2015).Related to the desire for homeownership is the fear that state support for backyard dwellings will disqualify them from receiving a house.Indeed, many respondents feared that receiving assistance from the government (access to water, structural improvements, or other such upgrades to their backyard dwellings) would count as their housing subsidy/benefit and they would thereby forfeit their claim to a real government house.Organizing to demand improvements to their BDs was perceived by many to come at the cost of losing their housing opportunity, and therefore, the safest bet was to focus on obtaining a house.This perception was driven, at least in part, by distrust in the government and the news that soon the government would no longer build homes (Thukwana 2020).While we do not have evidence to suggest that a backyard improvement would replace a housing benefit in practice, this fear is real among our respondents, especially those located in S2: The thing I fear is that if government is going to improve my living conditions, I feel like that is going to be my housing opportunity; they not gonna build me a house … if you give me improvement within the backyard and I'm gonna be there for the rest of my life then that's not much improvement.(S2-4) As this was the prevailing perception in S2 (and to lesser degrees in S1), it is not surprising that such individuals remain focused on the individualized benefit of obtaining a government home.Yes, communities protest for access to government homes, but often this is an individual effort among BD residents.For example, S1-10 recounts her experience of going to the housing department, finding someone willing to sell their RDP home, going back to get the council to seek agreement to buy it on her behalf, filling out paperwork to receive this subsidy, and so on (a 6-month plus process which failed in the end), rather than banding people together to demand homes or improvements to BDs.
The discussion of in situ improvements was more measured in S3, where most backyard dwellings are on council-owned land.The respondents expressed gratitude at the provision of toilets, water access/meters, and electric access/meters to those living in BDs on council land. 7The focus group participants also sought more from the government, such as plastic covering for their roofs.However, they also demanded houses.Much like the other focus groups, the first third of the S3 focus group centered around receiving houses, the mismanagement of the housing list, and competition with neighboring communities to occupy homes being built in the area.Thus, the S3 residents, possibly due to their exposure to the municipality, which seemed to reinforce their urban citizenship, were not concerned about missing out on a housing opportunity due to having received upgrades to their BDs.S3 respondents clearly illustrated a more nuanced understanding of the process as well as a greater level of trust in the government.For S3 residents, upgrades now to BDs are not perceived as substitutes for receiving a house later.
While the promise of a house can stifle demands for immediate improvements to BDs, the actual provision of homes also allows people to exit the backyard community, which has pulled some away from the backyard movement.In fact, respondent S1-1 even suggested that the municipality uses the provision of a house to directly undermine community organization and bad press, though there is no evidence this has occurred in practice.However, the point that receiving a house can undermine community organization is clear: "The problem is that [the government] starts to have this kind of a dividing rule.Someone gets a job, someone gets a house … three people in the leadership gets a house and our common goal was that we must fight for all to get a house.And when the three people get a house so obviously they no longer have interest in that particular struggle."It seems that the escape hatch undermines the organization itself; by cutting off the head, the municipality can destabilize BD collective organizing.Stated in the language of infrastructural citizenship, once an individual's citizenship is formalized by housing, the person exits the BD community and thus participation in its collective actions.This exit is also spatial, given that a backyarder's new home is often far from their current community, which echoes Zachary Levenson's finding that when the SA government provides a "housing opportunity," the recipients find themselves "torn from the networks that were the basis of their strategies of reproduction" and collective demand-making (2018:3219, 2022).The way delivery of homes undermines collective action is further exemplified in S3, where the current local councillor is a former (politically active) backyard resident himself.However, respondents indicated that he was not working on their behalf any longer, despite having received a home in the area: "[Our councillor] don't stand with us, you see-and he is from S3and he was also [previously] a backyarder; but now he stay in a house" (S3-15).
In summary, the promise of a house focuses BD residents on this goal and outcome, which often leads to individual actions and solutions (even for those with little time).While the BD community has shared challenges and circumstances, if one can get their promised house, one can solve these problems.Therefore, there less community organization around improvements to BDs, which likely perpetuates the poor conditions.Interestingly, the promise of a house, which allows BD residents to exit the community, may increase the self-efficacy of those who have received a house, but further entrenches those who did not in a "self-efficacy trap" (Lieberman & Zhou 2022).While self-efficacy did not arise as a theme in our focus groups and interviews (beyond feelings of helplessness), it may be worth exploring, as it has been found to matter for protest action in Africa (Harris & Hern 2019).
We observed key variations across contexts with regard to housing: S2 residents were overly suspicious of government help due to a fear of losing this housing opportunity, while S1 respondents were simply preoccupied with this promise, and S3 actively sought for both short term improvements to BDs and to agitate for houses. 8It is interesting, though, that the promise of a house seemed to consistently depress the hunger for collective action to improve BD dwellings across all contexts.The promise of a house thus operates as a type of escape hatch that in a way excuses people from collective action.First, it provides a hope for a better living situation, which depresses desires to improve one's current BD, especially if it will constitute someone's "housing opportunity" from the state.Second, when key members of the movement receive a house, it disrupts community coordination and organization.Underlying all of this is the goal of full urban citizenship which is enhanced (possibly even fully achieved) by obtaining a house of one's own.

Legal Standing and Urban Citizenship: BDs on Private versus Public Land
The differences across communities regarding the promise of a house may in part be driven by the type of land on which a BD is built, private versus public.As noted above, BDs in S3 are predominantly, though not fully, on council land, those in S2 are more evenly mixed, and those in S1 are almost exclusively on private land.The S3 and S2 experiences with government and understanding of government upgrades, subsidies, and housing provisions vary, while S1 respondents saw no clear relationship between upgrades and the provision of a house.The government's relationship with public and private land, and as such the degree of citizenship granted to residents on each type of land, may explain some of this variation.In short, we find that BDs on public land occupy a less ambiguous legal area than those on private land, which provides insight into the role of quasi-legality and illegality of one's housing situation on infrastructural citizenship and the ability and/or willingness to engage in collective action. 9 In S3, residents of BDs on council land (for example, a shack behind a council flat) tended to be more willing to ask the city for services, because it is clearer that it is the city's responsibility, while those in S2 were much more skeptical about the (un)intended consequences of government interventions, including stricter control of service consumption, higher costs, and indebtedness.In contrast, backyarders living on private land had low expectations about the government investing in homes or services.This is partly the case because the municipality has consistently argued that they are not legally able to invest in improvements to private property. 10Conversations with BD residents suggest that in contexts where BDs are on private land, residents are less likely to see the government as able to assist them in their current circumstances, and thus they are less likely to make collective demands on government (beyond that of a house, which is often individualized, as noted above).Their lack of full urban citizenship or inclusion limits their expectations that government is able to do anything to assist or help them.
Those living in BDs behind private homes are often left to their own devices and frequently expressed feeling abandoned by the municipality.For example, one resident (S2-1) told us, "Yes I would like it if the government support us with electricity [but the] reason is they not giving us electricity [is] because it's not council property … just because it's private property they can do nothing and I think it's very unfair."Such backyard residents are disempowered by their lack of leverage; because they occupy a BD on private land, they are outside the responsibility of urban government.Respondent S3-11, whose BD is also on private land, indicated that because BDs are technically illegal, "if you build a structure on Friday, by Monday, the neighbors, they phone law enforcement and [municipal] law enforcement come out and give your 7 days or 14 days to [demolish the structure]."Meanwhile, those living on council land claimed that the municipality must provide plastic sheets, water meters, and toilets to their BDs (this was especially the case in S3, where some provision had already taken place).The dynamics at play in S3 are in line with Gabrielle Kruks-Wisner's (2018) findings regarding claim-making in India: those living on public land had greater exposure to the state, which in turn provided them a clearer linkage with the state, and thus potentially more information about the state and its responsibilities, which Kruks-Wisner argues facilitates claim-making.
This difference in formal tenure, which maps onto the degree to which the state is willing to help, creates a much stronger sense of urban citizenship among those in BDs on public land relative to those on private land.Those on private land are much less likely to make demands on government and thus are less likely to engage in collective action.We should therefore expect neighborhoods with more BDs on private land to be less likely to engage in collective demand-making.

Community Leadership
A final influential factor regarding collective action is community leadership.Even from our limited case studies, it is evident that successful organization and demand-making on government depend on strong leadership.Our focus group participants from S1 talked about a long-trusted and supportive leader, who nonetheless had not been able to productively rally backyard residents.The reasons for that were not entirely clear.Some suggested that he was not overly charismatic, while others viewed external factors to be more influential.The backyarder community in S1 is especially socioeconomically diverse, with large groups of young professionals and foreigners, who mostly live on private land.In such a situation, the backyard residents had little time to organize for improvements to their BDs, and even if they did have time, the focus of citizenship action would be on claiming their own homes, given that the council was not able to invest in private land.
The backyard residents in S2 were led by a motivated, long-term activist and community leader who had developed a personal relationship with many backyard dwellers.She and others not only agitate for improvements from government but work together to help each other with daily life challenges.For example, the S2 community established the "Wake up a Child" initiative to help neighbors get their children to school.This unique child-focused effort in S2 is likely a result of this leader being the only female leader in our three research sites.More generally, this leader was able to activate 60 backyard households; while this was a small fraction, those involved were highly engaged.However, from focus group discussions and an interview with this leader, it was apparent that many backyard residents lacked trust among each other (the exact source of mistrust was unclear), which limited the appeal of the organization and the organizer.
The S3 backyard residents were led by a trusted, strong leader.He was largely seen as the only person looking after the backyard community, and this perception likely came from his success in providing improvements to backyard residents' lives.If a backyarder wanted something done, they could get it from this leader.From our research, it also seemed that no one was excluded (or excluded themselves) from aligning with this leader, and trust in him was high.In fact, unlike the others, the S3 focus group largely became a discussion of this leader's importance in the community.This community leader did not only agitate for and lead collective protest action, but he also assisted backyarders with mundane, everyday practices to claim their citizenship (such as resolving disputes, filling out forms, sending emails, or claiming support from the government).This community leader offered important support to his fellow residents, which they were not able to obtain from their locally elected councillors.
Community leadership clearly sets S3 apart from the others, and it is the site that has a more successful history of collective action among backyard residents.It is also important to note that community organizing was likely easier in S3 because most BDs are on public land and the residents were moderately socioeconomically homogenous.

Community Profiles and a Theory of Collective Action
The community profiles that emerge can help us develop a theory of collective action among backyard residents.First, it is clear that S1 is in the worst position to successfully engage in collective action because leadership is weakest (though still present), and most BDs are on private land, which suggests that these individuals also have little hope that the city will help them.As such, it is not surprising that backyard residents in S1 have not banded together to send petitions to government or protest for improvements to their BDs and livelihoods.S2 is in a moderately good position regarding the potential for collective action.This community has moderately strong leadership and a mixture of private and public land, but division within the community, the limited reach of the leader, and suspicions regarding government provision to the BD and its implications for receiving a house all limit the desire and ability of this community to make collective demands on government to improve the lives of backyard residents.
Of the communities we worked in, S3 seemed to be situated the most advantageously regarding collective action for improvement in backyard housing.This community has strong leadership, and its residents tend to reside on public land.As such, it is less surprising that S3 backyarders not only make more demands on government via protest, community meetings, meetings with government officials, and petitions to government, but they have also seen the government meet some of their demands.While backyard residents in S3 still face important challenges, they have seen improved access to and meters for water and electricity and the provision of toilet facilities, which improve their infrastructural citizenship.Further, while we conducted the research, the residents had an ongoing, organized campaign (and an imminent meeting) to receive plastic sheet coverings for their roofs.Nevertheless, the dream of homeownership looms large also in S3.Homeownership remains the ultimate goal of citizenship practices there as in all other backyarder communities in Cape Town.
Our analysis has shown that social divisions, an escape hatch (in the form of the promise of a house), the legal tenure status of residents (whether it is public or private land on which BDs are constructed), and community leadership are key factors that influence the ability of communities to collectively organize.Further, the legal status of BDs conditions backyard residents' urban citizenship and whether or not the escape hatch deters collective action; when BDs are on public land, citizens feel more a part of the city because the municipality sees them as deserving of services, and improved services are not perceived as fulfilling the resident's housing opportunity, both of which make involvement in collection action more likely.
Our theory of collective action among backyarders, which incorporates infrastructure citizenship, is in line with past research on collective actions which emphasized the important role of social ties, shared identities, and leadership.However, we improve upon these theories by uncovering the important roles of tenure and the presence of an escape hatch, which are both rooted in infrastructural citizenship.We propose that, for backyard resident collective action, infrastructural citizenship conditions the effects of conventional collective action determinants.When backyard residents are (or at least perceive themselves to be) more fully integrated into the city, they are more likely to make collective claims on the government.More specifically, when the legal status of their dwelling is relatively more legal, they are more willing to make demands on government, are more open to receiving a larger range of improvements to their BDs (rather than focusing completely on exiting the backyard in a way that prevents improvements to their current living situation), and leadership can more effectively deliver improvements in a way that encourages citizens to participate in further engagement.

Conclusion
In African cities and many other urban areas worldwide, there is a growing population of informal tenants.In South Africa, backyard residents constitute the fastest growing housing segment and largest tenant group in the country.Despite the urgent need to improve the material conditions in the sector, backyard residents-compared to residents in other informal settlements-have been far less active collectively to claim their "infrastructural citizenship" (Lemanski 2020).While acknowledging the geographic and empirical limits of our data, we were able to develop a theory to better understand the conditions under which collective action is more or less likely to occur.
As might be expected, the informal status and high poverty levels among backyard residents constrain citizenship practices, especially collective action, to claim better housing and service delivery.The informal status of backyarders' homes makes them vulnerable as citizens, creating uncertainty and risk of exposure when engaging the state.The high levels of poverty coupled with the need to pay monthly rentals (or make other in-kind contributions) to the main household cause many backyard dwellers to struggle to make ends meet.In their struggle for survival, employment and income become top priorities, which leaves little time and energy for citizenship actions.Beyond these common limitations to collective action, we find that socio-economic diversity, an escape hatch, legal standing, and community leadership are additional factors conditioning collective action.
While backyard residents, or informal tenants more generally, may share a common housing situation, they can differ considerably in terms of their socioeconomic situation, racial or ethnic background, and interests.Overcoming this diversity is a key challenge to fostering collective action.This appears to be made even more complicated by the promise, and sometimes the actual provision, of government support for private property.Backyard residents and community leaders struggle first and foremost for their own property as opposed to seeking government intervention to upgrade the sector (not dissimilar to the barriers to successful movements explored by Frances Piven and Richard Cloward [2012]).While the promise of one's own home serves as an escape hatch, residents living on public land who have strong leadership appear to be better placed to engage government collectively over infrastructure provision.Thus, our theory suggests that the grey legal area BDs occupy not only limits the willingness of BD residents to make demands on government but also limits their expectations of governmental response.Importantly, our backyard residents vary in their degree of legal status: those on public land enjoyed a relatively more legalized status (though not fully under law).This suggests that we should see more collective action as groups of citizens' circumstances move upward along the continuum between illegal and fully legal.Further, we observed collective action to be critical to improving living conditions and gaining political recognition over time in the localities (S3 and to some extent S2) that had succeeded in collectively organizing.As such, there is likely an endogenous feedback loop: if some collective action can start, it can change political and/or legal standing, and that can in turn encourage further collective action and thus more change.While this relationship clearly illustrates the barriers to collective action among communities of individuals who occupy quasi-legal and illegal spaces, it may also help explain the evolution of protest movements and lead us to explore more fully the emergence of and importance of tipping points.While the path-dependent nature of protests (especially with regard to how past gains in the form of legalization can be resources for current or future efforts) has been acknowledged elsewhere with regard to LGBTQþ protests (Ghaziani et al. 2016:177), we bring this to additional communities and call on studies of collective action to consider the evolutionary dynamics of collective action.
Another key variable found to condition the success of collective action among those renting in informality is the presence of charismatic leadership.Charismatic leaders in two of our three case study neighborhoods have managed to rally support behind collective action, which occurred not only in the form of protests and community mobilization but also as part of more mundane infrastructural citizenship practices of supporting each other with everyday life struggles, providing access to information, filling out forms to claim government support, and assisting in resolving conflicts between landlords and tenants.In creating a bridge between citizens and the local municipality, community leaders filled a glaring vacuum caused by the generally poor performance of local councillors.Importantly, local community leaders can play a key role in the realization of infrastructural citizenship which has been largely overlooked.Future research would do well to investigate the ways in which different types of leaders and leadership qualities facilitate infrastructural citizenship.
Our findings further improve our understanding of service delivery or "valence protests" (Harris & Hern 2019) in South Africa.Many studies focus on citizens and their socioeconomic circumstances and/or political alignment as drivers of protest and other movements.However, our study suggests that the legal standing of an individual's housing (and potentially other) situations conditions one's relationship to the state, which in turn plays a key role in facilitating individual and collective action.These state-society relations as well as community governance systems are the contexts in which we study collective action and as such can be overlooked or taken for granted; our study follows others in the call to consider variations in structural and local governance factors that may facilitate or hinder collective action (Lust 2022) and full realization of one's infrastructural citizenship (Lemanski 2020).
Our article has discussed some of the challenges and factors influencing collective action among backyarders, but further research is required to test our theory systematically and to understand under what conditions backyard residents, and informal tenants more generally, can improve their housing and living conditions through political action.

Table 1 .
Characteristics of the Research Neighborhoodscommercialization of backyard rental arrangements, family and personal relationships continue to play an important role in S2.Many backyard tenants are related to or friends with landlords.Most of them are eligible for their own state-subsidized housing and have been on the "waiting list" for many years.The backyard group organization consists of a fairly strong, tight-knit group of friends, family, and neighbors helping each other to survive.Some of the members have also been active players in housing movements across the city.The BDs in S2 are not overly robust and are located on a mix of private and council-owned lands.