Becoming a Population

Draft – Please do not circulate without author’s permission 1

Becoming a Population: Seeing the State, Being Seen by the State, and the Politics of Eviction in Cape Town Zachary Levenson University of North Carolina, Greensboro

Abstract: States may “see” their populations, but this does not mean that populations are ready-to-hand. Instead, this paper contends that a group’s “population-ness” is the outcome of political struggles over representation, or who speaks for whom. How they resolve questions of representation is in turn impacted by how residents “see” the state. Drawing on extended ethnographic fieldwork and interviews conducted between 2011 and 2019 in Cape Town, South Africa, this paper compares the fate of two large land occupations in which thousands of residents erected shacks on land to which they had no legal claim. In one case, residents were evicted after a judge upheld the municipal government’s claim that the occupiers were on the land illegally; in the other, they won the right to stay put, and the settlement continues to grow even today. This paper argues that these outcomes had little to do with a local state simply ordering evictions from on high, implementing them onto populations below. Instead, based on divergent understandings of what the state wanted from them, participants in each occupation comported themselves quite differently. Where residents thought the municipal government desired individual homeowners, they were read as factionalized opportunists and evicted from the land. But where they envisioned the state as wanting a governable community, they organized the occupation under an elected leadership that functioned as an informal government of sorts. In this latter case, they were perceived as a deserving population that was intelligible to the state as such, and they were ultimately tolerated.

Keywords: eviction · housing · urban informality · representation · recognition · states · collective action · post-apartheid South Africa

Introduction

In 2011, thousands of South Africans in need of a home occupied two stretches of land on Cape Town’s periphery, just before that city’s municipal elections. In one of these occupations, Rivenland, a thousand residents converged on a publicly owned field adjacent to the final stop on a commuter railway. None of the shacks that were erected interfered with the train’s functioning. This was the poorest section of the township, and there were no middle-class homeowners nearby who were worried about property value. Every single participant identified as “Colored,” an apartheid-era racial category distinct from “African,” and this occupation took place in a neighborhood that was itself nearly entirely Colored. Colored people in this township overwhelmingly vote for the city’s governing party, the Democratic Alliance (DA), and so there was no reason for DA-affiliated officials to suspect that the occupation was a political ploy to bring supporters of their chief rivals, the African National Congress (ANC), into DA territory.

Meanwhile, a couple of kilometers down the road in the same township, a few hundred squatters set up shop on a plot of land called Holfield that was split between two absentee landlords: the one a demolition company, the other a property holding concern. Despite initially being smaller than the first occupation, it quickly grew until within a few months, there were 2600 shacks on the field. The majority of the occupiers were black “African,” their first language


being isiXhosa, and they largely moved from adjacent majority black townships into Colored territory. In both of these nearby townships, residents overwhelmingly vote for the ANC, and so it would not have been a stretch to read this occupation as an attempt to dilute DA support in this ward. Moreover, the occupation was quite visible. If Rivenland was located in a far-flung cul-de-sac in the poorest part of the township, Holfield abutted the main thoroughfare along its wealthiest section. And whereas no neighbors challenged the Rivenland squatters, middle-class residents across the street consistently mobilized against the Holfield occupation, demanding that the city oversee their eviction.

Within a year, Rivenland’s shacks were cleared, every one of its residents evicted and left to fend for themselves. But Holfield was a different story: a judge ruled that they could not be legally evicted, and they were granted the right to stay put. Despite their best efforts, no such ruling was forthcoming for Rivenland. How can we make sense of this counterintuitive outcome? This paper argues that this situation is only paradoxical from the perspective of an omniscient state that devises policies and implements them at will, targeting specific populations. Adapting Schmitt’s (1985) term, this is what I call the decisionist state: one in which the local state’s authority has its foundation in its own will. This state is presumed to gaze out from a privileged vantage point over a range of populations, each of these taken as an empirical fact on the ground. In this telling, the state is imagined as active and civil society passive, as if James Scott’s (1998:4-5) “high modernist” state were taken to be representative of states more generally. It remains a calculating state, one that attempts to optimize surplus extraction while minimizing resistance (Scott 2009:40). The state’s twin aims, in other words, are to facilitate capitalist accumulation, whether through supporting investment in key sectors of the economy or else boosting real estate value; and to reproduce its legitimacy by rewarding supporters and marginalizing those who it deems to be threats (cf. Offe 1984).

Read in an urban context, this means that state-led modes of displacement, from land grabs through gentrification, are imagined to originate in a government office from which officials weigh their options in accordance with any number of variables: real estate value, visibility of the occupation, residents’ “fit” into the surrounding area, potential for disruption, and so forth. But the outcomes considered here have little to do with a local state selecting a few key areas for eviction and implementing these onto populations below. Instead, this paper contends, they had everything to do with how residents collectively comported themselves. They may have been “seen” by the state to be members of populations, but this was not an inherent form of group identity. How they were seen had everything to do with how they themselves envisioned the state and therefore, how they imagined that the state would see them.

Where residents anticipated that the municipal government desired individualized homeowners, they were read as factionalized opportunists and evicted from the land. This is what happened in Rivenland. But where they envisioned the state as wanting a governable community, they organized the occupation under an elected leadership that functioned as an informal government of sorts. This was their experience at Holfield. In this latter case, they were perceived as a deserving population that was intelligible to the state as such, and they were ultimately tolerated. This suggests that their collective action does not qualify them as a social movement, insofar as they were not organized around a sustained campaign of claims making on the state (McAdam, Tarrow, and Tilly 2001; Tilly 2004; Tilly and Tarrow 2007). If anything, they were closer to what Asef Bayat (2013) calls a social nonmovement: they lacked a guiding program, directly appropriated land rather than demanding it, and participated in these actions as


an extension of their everyday practices. Most of the squatters in these occupations did not take land to make a political statement but because they needed a place to live.

As members of nonmovements, participants attempted to evade the gaze of the state rather than engage it. They did not mobilize collectively in an attempt to pressure the state, but to exert moral force: they conceived of the state as desiring a certain type of subject, and they comported themselves accordingly. This is the sense in which their seeing the state impacted the way they were seen by the state. Residents’ representations of the state, how they imagined this state and what it wanted from them, determined their organizational form. They were not naturally constituted as populations laying in waiting to be seen by the state; they had to constitute themselves as such.

The first section of this paper considers the prevailing sociological literature on evictions, arguing that it lacks attention to this formative process of self-organization. This is not a problem because of some nebulous moral imperative that requires researchers to consider voices “from below”; it remains a problem because, as we have already seen, the decisionist model of the state cannot account for eviction outcomes. The second section draws on Partha Chatterjee’s (2004, 2011) attempt to bypass this decisionism by turning to the relationship between populations and the state, though he treats populations as natural features of the political landscape. As an alternative, I explain how struggles on the terrain of everyday life – what I call struggles over representation, or who speaks for whom – are inextricable from vertical struggles for recognition, or how residents collectively project themselves to the state. The third and fourth sections then work through the two empirical cases, the first on Rivenland and the second on Holfield, treating the dialectic of representation and recognition in both cases. These sections reveal how occupiers’ respective perceptions of the state affect how they organize themselves, whether as a series as in Rivenland or as a group as in Holfield, which in turn led to divergent outcomes. The conclusion demonstrates the implications of this divergence, namely, that like everybody else, the urban poor organize themselves on the terrain of civil society.

Beyond the Decisionist Model Why do municipal governments evict squatters from land occupations? The simplest answer is that local states have inextricably connected economic and political aims: they want to cut social spending costs while managing unruly surplus populations. But neither of these explanations is immediately applicable to South African cities, where two things tend to be true. First, squatters frequently build shacks on public land, which would actually save the government money. Even when squatting on private property removes land from the market (Levenson 2017a; Starecheski 2016), it serves as a form of self-provisioning (Holston 2008; Simone 2004), meaning that if the government defines this as adequate housing, they no longer need to provide alternative accomodations in accordance with their constitutional mandate to do so (Huchzermeyer 2001, 2003; Levenson 2018). Indeed, as the economist Hernando de Soto (2000) has argued, Southern governments should welcome shanties on their land and even provide residents with title deeds as a means of integrating them into capital markets (Atuahene 2006; King 2003). Second, evictions in no sense dissipate the political rage of the precariously housed. While they may be deployed as a means of partisan retaliation (Auerbach 2016; Chance 2018; Levenson 2017b; Roy 2003; Weinstein 2013), the removal of militant squatters has done little to extinguish their political demands. The relocation of land occupiers to state-run encampments


(Chance 2015; Levenson 2017b, 2018; Ranslem 2015) has typically been a source of, rather than solution to, mass mobilizations and sustained protests (SWP Dwellers 2011). More broadly, the most comprehensive analysis to date of demonstrations in South Africa lists housing and evictions as the second largest catalyst of new protest actions (Alexander, Runciman, and Ngwane 2014), coming in just behind deficient service delivery. Evictions then are clearly inefficient instruments of pacification or control; their effect reveals that they actually augment what we might assume they would diminish. If evictions do not reduce social spending or pacify squatters’ rage, a large body of recent scholarship suggests a third function of evictions, namely, that they constitute a strategy of rendering the city “world class.” In the most general sense, this is a reference to a “transnational urban system” (Sassen 2001:xxi; cf. Massey 2007) in which cities compete with one another in order to become privileged sites of investment, producer services, and corporate control functions (Begg 1999; Gordon 1999; Harvey 1989; McDonald 2008; Weinstein forthcoming). Evictions play a role insofar as they represent governmental attempts to emulate existing urban nodes in the global financial circuit by invisibilizing poverty to create a desired image of the city. Whether pressured by local elites (Anjaria 2009; Baviskar 2006; Bhan 2016; Ghertner 2015), growth coalitions (Andranovich, Burbank, and Heying 2002; Berrisford and Kihato 2006; Hiller 2000; Wei 2012; Xu and Yeh 2009), or international real estate developers (Dasgupta 2003; Goldman 2011; Searle 2014; Weinstein 2014), municipal governments are viewed in this formulation as calculating actors who make eviction decisions in the name of boosting their image. Analysts of world-class city-making typically maintain that visibility (Davis 2011; Dovey and King 2011; Ghertner 2011a, 2011b, 2015; Huchzermeyer 2011; Pithouse 2006) is the key variable separating settlements targeted for eviction from those that are officially sanctioned.

If this were indeed the case, we would expect those populations most visible to the state – those “seen” by the state – to be the first targeted for eviction. But in Cape Town, this was not the pattern I observed. In the two cases considered in this paper, Holfield and Rivenland, Holfield abutted a major thoroughfare and was located across the street from a middle class neighborhood whose residents continually mobilized to have the settlement cleared. This mobilization, which articulated its opposition in racialized terms, drew further attention to Holfield, with local politicians holding public forums to diffuse the tensions. But Holfield was ultimately tolerated, with a judge granting them the right to stay put. Meanwhile, Rivenland was located at the end of a cul-de-sac in the poorest section of the same township, and no neighbors demanded their relocation. This land was already possessed by the municipality, meaning that no private property owners could demand residents’ eviction as they did in the case of Holfield. Yet a judge ruled that the Rivenland squatters were “opportunists” and proceeded to order their eviction, and the entire settlement was cleared.

The same problem exists if we view these cases through the lens of the gentrification literature. An emerging body of work transposes Smith’s (1996) classic rent-gap hypothesis onto Southern urban contexts (Atkinson and Bridge 2005; Butler 2007; Lees, Shin, and López-Morales 2015, 2016; Lees, Slater, and Wyly 2008; Slater 2017; Smith 2002). In this formulation, governments do not select real estate for its present value, but rather for its developmental potentiality. In other words, what matters is not the absolute value of the land, but the difference between its currently capitalized ground rent and the rent it would fetch if developed under its “highest and best use” (Smith 1996:62). South Africanists’ insistence that evictions are driven by gentrification (e.g. Bond 2000; Çelik 2011; Gibson 2011; Pithouse 2008) remains a theoretical proposition rather than one supported by empirical case studies. None of the cleared occupations


I observed were located on land earmarked for development. Indeed, more than six years after the eviction of the Rivenland squatters, the land remains largely vacant, just as trash-strewn and peripherally located as it was at the time it was initially occupied. Becoming a Population

What all four of these explanations – austerity, control, aesthetics, and gentrification –

have in common is that they assume a decisionist model, i.e. one in which the local state is a rational actor that realizes its will through policy implementation over a range of populations. In this conception, populations are naturalized as features of the social landscape, among which the state selects some for eviction while tolerating others. But from the perspective of the decisionist state’s gaze, the two cases considered here do not add up. The one we would expect to be evicted by any of these four metrics gained the right to stay put, whereas the one we would expect to be tolerated was forcibly eradicated.

Yet this is only a paradox if we start from the assumption that populations as such always already exist in forms legible to the state. In this formulation, a scarcity of government resources means that Cape Town’s Department of Human Settlements does not retain the capacity to evict all new land occupations simultaneously. Instead, it needs to prioritize some over others, and to do this it constitutes a comparative system that treats each land occupation as a “case” (Foucault 1995:190-1) or what I refer to as a “population.” As an object of knowledge, it is evaluated in accordance with a set of predetermined criteria, namely, those of the four literatures discussed above: cost savings, pacification effects, aesthetics, and rent gap.

Political scientist Partha Chatterjee (2004, 2010, 2011) adapts this Foucauldian account to a postcolonial context. In postcolonial democracies, he argues, marginalized populations are excluded from civil society insofar as they do not relate “to the state through the mutual recognition of legally enforceable rights” (Chatterjee 2011:13-4). Instead, they constitute part of political society “where governmental agencies dealt not with citizens but with populations to deliver specific benefits or services through a process of political negotiation” (ibid.). Members of political society are never recognized as rights-bearing citizens, but instead as part of larger units he terms “populations,” a clear nod to Foucault’s (2007) analysis of groups only formed in relation to being governed. These populations do not themselves possess rights; benefits and services are delivered as part of a political calculus. Because democratization means that postcolonial states are suddenly accountable to their populations, they can longer indiscriminately relegate surplus populations to socio-spatial oblivion. Instead, the municipal state must appear benevolent in order to reproduce its own legitimacy, and thus “authorities make a political calculation of costs and benefits” in each individual case (Chatterjee 2011:13-4).

While the politics of engagement Chatterjee describes captures something essential about the post-apartheid state, he too hastily assume that informal settlements are automatically constituted as populations legible – visible – to the state. This is but another variety of Scott’s (1998) omniscient state, decisionist insofar as a self-enclosed sovereign entity projects policies onto objectively existing populations below. But what if populations are not timeless identities linked to spatial forms (neighborhood, settlement, etc.), but are instead the outcome of residents’ capacity for self-organization? It is the contention of this paper that how (and indeed whether) populations are seen by the state is inextricable from how members of these populations see the state.


What Chatterjee’s concept of political society does is to capture the moment in which the state sees populations. But since the postcolonial governments he discusses operate under conditions of material scarcity, they need to make tough decisions about service delivery: which populations should be prioritized, and which should be deferred? And since they also operate with a limited capacity of implementation, even undesirable policies like evictions entail a process of selection: which populations should be cleared first, and conversely, which should be tolerated? What this does is to foster competition among various settlements, creating a situation in which they struggle against one another to define themselves as the most legitimate population in the eyes of the local state. If Chatterjee finds populations ready-to-hand, the competitive dimension implied in his account suggests that residents struggle to hail the local state. Even if they do this, as Chatterjee suggests, through the mediation of a population, they still need to agree upon the parameters of this political form. Is their best bet to emulate rights-bearing citizens and potentially designate – or even elect – a representative, or even a representative body? Or should they reject such a politics as one that could be viewed as threatening to the state, which might view informal delegation as a challenge to its monopoly over legitimate representation? This question of course turns on how residents themselves view the state, and more broadly, how they conceptualize it. Based on residents’ countless experiences with land occupations in their own neighborhoods, they develop distinct understandings of how the local government perceives these settlements and indeed reads them as populations. How they project themselves to the state is inseparable from how they decide to represent themselves.

This means that Chatterjee’s account of states recognizing populations does not take into account the struggles over representation within each settlement. In this paper, I contrast two very different struggles over representation. In both cases, how residents viewed the state impacted how they decided to represent themselves politically, which in turn affected how they were recognized by that state. In Rivenland, occupiers saw a state that wanted residents to comport themselves as aspiring property owners. They were therefore preoccupied with protecting their own individual plots, rather than trying to represent themselves as a collective with shared goals undertaking a common project. The latter situation characterized the occupation in Holfield. Residents saw a state that wanted them to organize themselves collectively, and they did so. If the Rivenland occupation approximates what Sartre (2004) termed a series – a collection of people all acting in relation to a common object, the land – then the Holfield occupation is clearly what he called a fused group: a collective that not only acts in simultaneity, as in the case of a series, but also remains unified toward a collective project.

The divergence between the two occupations’ strategies of representation – how they projected themselves to the state – were crucial to the ultimate outcomes, though not in the way residents predicted. The serial occupiers of Rivenland were not viewed as compliant property owners in the making, but as impatient opportunists seizing public land. Meanwhile, the collective representation of the Holfield occupiers facilitated their recognition as deserving residents, even if this was not the reason they initially formed an elected body. Yet this body allowed the Holfield occupation to be legible to the state as a population in Chatterjee’s sense, whereas the Rivenland occupation failed to achieve this feat. Rivenland: The Politics of Seriality


In Rivenland, participants did not initially view their occupation as such. Instead, they understood it to be the legal distribution of plots of land by an organization ambiguously related to the state. A group called the Mitchell’s Plain Housing Association (MPHA), named after the township in which the occupation took place, began holding meetings in the area, explicitly appealing to backyarders. Most people who lived in informal housing in Mitchell’s Plain, a solid 15 percent in this particular area, occupied backyard shacks. Sometimes backyarders would find friends or family members who allowed them to erect structures behind their formal houses. Or else they would pay monthly rent to the homeowner in return for access to the land, and if they were lucky, water and electricity.

The official unemployment rate in the neighborhood approached 40 percent, with the real rate substantially higher. Consequently, even the seemingly negligible rent payments were out of reach for many residents. Or else they could afford rent, but homeowners wanted to charge them an extra fee for water or electricity. Or sometimes, even when they did not need to pay, homeowners viewed the arrangement as temporary, even if no alternatives were likely to be available any time soon. This was the situation of Aisha and Muhammad. Aisha had four children from previous marriages, and Muhammad had two. The tiny shack in Aisha’s parents’ backyard could not possibly suffice, and besides, relations with her father were tense; he wanted his yard back.

I asked Aisha why she participated in the Rivenland occupation. “Desperation,” she told me. “We were living in a backyard in a shack that’s probably 1 by 2 – 1 meter by 2 meters.” She could not help but laugh at the absurdity of the situation. “And we literally had nowhere to go and were living there, struggling to survive. I mean, we both had families that at the end of the day, we wanted to settle down with them, whether we were going to rent a house, whether we were going to do whatever. The point was that we were going to make provisions so that at the end of the day our children can come to us, you know, and visit us. Not visit us: come and live with us.”

This is why the MPHA’s meetings in their neighborhood immediately appealed to Aisha and Muhammad. In the months leading up to the occupation, they held gatherings at a local community center and a nearby elementary school. As word spread, there were soon hundreds of people at these meetings. They were always chaired by MPHA members, though they sometimes brought guests with them: elected ward councilors or representatives of the local South African National Civic Organization (SANCO), for example. Most backyarders who attended these meetings were under the impression that the MPHA was related to Cape Town’s Department of Human Settlements, the arm of the municipal government responsible for housing delivery.

What they did not know was that the MPHA was actually a front group for the African National Congress (ANC), the ruling party nationally but in the official opposition in Cape Town since 2006. Mitchell’s Plain residents voted overwhelmingly for their chief rivals, the Democratic Alliance (DA), with ANC supporters few and far between. In general, voting is strongly correlated with race in South Africa, with Cape Town’s so-called “Colored” population tending to support the DA, and its “African” population going for the ANC. Due to the spatial legacy of apartheid, these voting blocs are often adjacent but mutually exclusive. Bordering Mitchell’s Plain to the east, for example, is Khayelitsha. While Mitchell’s Plain is the largest “Colored” township in the country, Khayelitsha is the largest “African” township in Cape Town. When the MPHA brought an ANC ward councilor to a meeting, for example, he was elected in Khayelitsha. And SANCO, despite its nonthreatening name, is actually an ANC-affiliated community organization, and its members in Mitchell’s Plain were openly partisan.


What was an ANC front group doing organizing a land occupation in non-ANC territory? In the run-up to local elections, party operatives often organize land occupations in rival territory to expand their voter base, hoping to pick up an extra ward councilor or two. If the MPHA could successfully establish an occupation, they could frame this victory as a project of the ANC. Even the national ANC had designs on Mitchell’s Plain. Cyril Ramaphosa, currently President of South Africa, visited the neighborhood next to Rivenland while serving as his predecessor Jacob Zuma’s Deputy President. As he told an audience in a community center there in 2013, “We are going to mobilise and form ANC structures in Mitchell’s Plain and solve these problems. We will take back what is ours and we will win Mitchell’s Plain” (Palm 2013:3).

In order to pull off this operation of gerrymandering in reverse, moving people rather than district boundaries, MPHA members could not order a group of residents, even those in desperate need of land, to simply walk onto a field and build. They would lack the trust of participants, and besides, given the precarity of land occupations, people would typically want some guarantee that they would not be evicted as soon as the sun rose the next morning. The MPHA would need to figure out how to stave off the city’s Anti-Land Invasion Unit (ALIU), a relatively autonomous arm of the Department of Human Settlements tasked with monitoring new occupations. While ALIU employees are not legally certified to use force or even remove people’s belongings, they work in conjunction with the South African Police Service (SAPS), which is authorized to do so. The ALIU acts as quickly as its capacity allows, since South African law prohibits evictions once people’s homes are established, even in cases of illegal occupation. While the definition of “established” has changed over time in accordance with new court rulings, the 1998 Prevention of Illegal Eviction from, and Unlawful Occupation of, Land (PIE) Act stipulates that courts must consider the occupiers’ vulnerability, the duration of their tenure, and their likelihood of finding housing elsewhere. For this reason, unless the ALIU and SAPS can act before residents finishing erecting and furnishing their structures, they can prevent an occupation; but if they do not catch them in time, they need to secure a court ruling before they can legally proceed.

So the MPHA held meetings in southeastern Mitchell’s Plain for months until there were hundreds of people at each assembly. Backyarders shared their stories of hardship from the floor, and the MPHA succeeded in building a certain amount of rapport among potential occupiers, but with one major caveat. The organization did not try to transform this empathy into solidarity and facilitate the birth of a collective organization, let alone a social movement. Rather than building a cohesive social movement that could unite in the face of the anticipated state repression, the MPHA represented the occupation as the distribution of plots of land to residents, as if they were homeowners-in-the-making. The group’s generic name gave off the impression that it was working in conjunction with the municipal government. Next door in Khayelitsha, for example, groups like the Khayelitsha Development Forum, while certainly not government entities, often worked closely with the City in public-private partnerships. The presence of ward councilors and SANCO officials at their meetings also did not hurt.

Aisha, the occupier introduced above, shared a copy of the diary she kept throughout the occupation. On its first day, she described its origins. “People moved onto the Rivenland field on the 13th of May 2011,” she wrote. “I was on my way to [the] hospital and came across a meeting of the MPHA. They told us about this land invasion that was going to take place. They didn’t use those words, ‘land invasion.’ They told us we were going to get plots. They gave out numbers, little numbers, with their stamp on it and charge people 10 rand [80 cents] for registering with


them and gave us a plot. They had a book where they put your name and ID number, which they said would then secure your plot. They said we will get the plots that Friday the 13th May, 2011.”

As Aisha’s entry makes clear, the formal order imposed by the MPHA, as well as the involvement of local politicians, made the initial process appear less as a land occupation and more as yet another means of obtaining access to housing. They would register with an administrative body, in this case the MPHA, and then would subsequently receive the equivalent of title deeds: the moral authority to lay claim to a given parcel of land. In a real sense, whether this was a self-appointed committee with ambiguous ties to the ANC or an actual representative of the Department of Human Settlements, residents initially perceived their participation as a legitimate engagement with an arm of the welfare state. In South Africa, the government has historically articulated democratization as a distributional project, with access to land, housing, services, and basic income grants comprising its central tenets. Given that many of the participants had been on the waiting list for a government house for decades, it was not such a stretch to think that state representatives might devise a plan for the meantime. How was anyone supposed to wait decades for a house without illegally occupying land?

That May, roughly a thousand residents set up camp on the field across the road from the Rivenland train station. They had each paid a small fee to the MPHA, who instructed them to arrive early that morning to secure their plots. With members of the organization supervising, residents got on their hands knees and begin to define the boundaries of each of their “yards” with wooden stakes and bits of string, mimicking the logic of enclosure. Even if the homes they built were flimsy and the plots small, residents perceived themselves as homeowners in the making, acquiring a sense of autonomy absent to backyarders. In this preliminary phase, they therefore viewed the state as a partner in the occupation. They had of course not planned for a confrontation with law enforcement, nor had they talked about defensive strategies more generally. It therefore came as a total shock when the ALIU and multiple police units arrived on the field the following morning. Given their presumption that the occupation was actually legally sanctioned land distribution, the sight of the police was a rude awakening. “Everybody was looking for the Mitchell’s Plain Housing Association,” Aisha told me. “They were the people who told us everything was going to be okay and we had come with all of our possessions.” The ALIU and the police returned daily until they finally secured a judge’s approval to proclaim the occupation illegal six days later. As Aisha’s diary details, “On Tuesday 17th May, the sheriff of the court said over an intercom that we were there illegally and we were not allowed to be there. They gave us an interdict and gave us 5 minutes to vacate the land. Once again they removed whatever we had. People lost their IDs, their papers, their dentures…That was when we realized that this is illegal, we were not going to get anything. Nobody was going to be able to help us with this. We had been manipulated into the situation we are in now.” “The situation” was one in which the occupiers were not organized into a coherent collectivity, but instead into small groupings of contending petty proprietors. Having too many residents on board was thought to be a liability; instead, they hoped to get lucky by appealing to outside actors like local politicians and charity workers. Maybe this, they thought, would help them secure recognition from the municipal government. While the occupation was proclaimed illegal, the residents were still entitled to their day in court, which would not occur for more than a month. One of the occupiers, a longtime Mitchell’s Plain resident named Biggie, used to be an activist and had contacts with the Legal Resources Center in the Cape Town city center, the largest public interest legal organization in


South Africa. He reached out and secured representation by a pro bono lawyer named Sheldon. Sheldon was not as responsive as some of the residents would have liked, so he contacted another organization called Lawyers for Human Rights, enlisting Marius and William as a second legal team. Marius successfully challenged the interim court interdict, with a hearing scheduled for June 1. This meant that the residents had a couple of weeks to regroup.

In the meantime, they continued to look to outside organizations for advice and material support – a search for recognition in the terms described above. Around this time, Aisha began to describe a new character in her journal named Marina. She was the white, Norwegian-born director of a charity that claimed to help “poor Coloreds.” She was also affiliated to an anti-black political party called the Cape Party, of which her son was a leader and perennial candidate. How could Marina reconcile her support for impoverished “Colored” Capetonians with her affiliation to an explicitly racist political party? Aisha meditated on this question in her diary: “She started an organization on the field called ‘First People First,’” she wrote. “I think the Cape Party has this idea that if the people of the Eastern Cape go home” – by which she meant isiXhosa-speaking black “Africans” – “there will be more jobs, and more houses.” In other words, Marina sought to create a white-Colored alliance against black Africans, who she defined as migrants, drawing on an apartheid-era racist trope that falsely posited the Western Cape’s black population as arriving contemporaneously with European colonists. By contrast, the region’s indigenous population, brown people descended from the Khoikhoi and the San, were claimed as ancestors of the present-day Colored population. By rearticulating Colored identity in indigenist terms, Marina was attempting to marshal the progressive rhetoric of native people’s rights – “First People First” – in the service of creating an alliance against any black participants in the occupation. Since the MPHA’s plan was to bring ANC supporters into Mitchell’s Plain, Marina saw it as her duty to keep them out. Oddly enough, by late June, Marina and members of the MPHA had entered into alignment. As the occupation’s black population began to dwindle, the MPHA focused its energies on aligning with Colored squatters. At an occupation-wide meeting on the last day of the month, residents who were skeptical of the MPHA’s maneuvering called attention to the dangerous of continuing to split the occupation into contending factions of petty proprietors. A few days later, they formed a rival alliance called Residents Unite that sought to break with the exclusivist politics of the MPHA and First People First. The day after their first official meeting, the land occupations physically split into opposing camps. Members of each faction threatened the opposing groups with violence, and in some cases, altercations broke out. A month into the occupation then, there were at least three factions struggling over the right to represent the Rivenland squatters. First People First and the MPHA shared an exclusivist outlook, and as such, they were able to enter into alignment. Both groups’ goal was to reserve their limited gains on the field for those who would consent to be represented by their leadership. Residents Unite, meanwhile, refused this politics of petty proprietorship and instead attempted to maximize its base by extending access to the field to anyone in need of land or housing. These goals proved fundamentally irreconcilable, and their mutual opposition was further entrenched as each group continued to seek recognition from outside parties. The rescheduled court date was postponed again, this time until July 27. While these days were simply procedural from the point of the view of the City and the courts, the squatters were terrified by the indeterminacy of it all. The Mayor’s office repeatedly issued statements pointing out that the May interdict prohibited any new structures from being erected. The City government’s spokesperson Kylie Hatton told a community newspaper, “We are sympathetic


that some people have been waiting for a long time for housing and may be impatient. But the City cannot allow people to illegally occupy vacant land or build informal structures. Illegally invading land may delay or prevent formal housing in areas of invaded land. The Rivenland site has been identified for future housing projects” (Papier 2011). Today, more than eight years after this statement, no housing developments are currently planned for the Rivenland field. But Hatton’s statement does get at one important truth: the municipal government views self-provisioning as a threat to the order required to operate a functioning housing delivery system. From the City’s point of view, ordered homelessness is preferable to disorderly survivalism. Legal decisions and government statements obey a logic that opposes order to opportunism, mapping the former onto unitary organization and the latter onto factionalism. On the morning of the July 27 hearing, Aisha wrote in her diary, “It has been 75 days. We survived the rain, cold, wind and daily harassment of the police. Our structures and other possessions have been confiscated and many nights we have been forced to sleep in the open. But our spirit of defiance remains strong and we are determined not to be moved.” The next morning, a representative of the Anti-Land Invasion Unit showed up on the field and asked for Aisha. He heard her speak at a public forum and decided she was a suitable representative. “He asked me not to allow anyone else onto the field, because according to him, that will only cause problems for us. Our people now have hope again.” The idea was to keep the occupation legible to the City government. This meant both prohibiting newcomers from expanding the occupation, as well as keeping the organization of the existing settlement coherent and structured. But before the High Court issued its final ruling on August 30, the occupation was far from organized in a form legible to the municipal government. Instead, residents’ search for recognition from the local state was impacted by their struggles over representation, which were in turn structured by how occupiers saw the state. Since they initially viewed it as a partner in a redistributive project, they largely themselves as petty proprietors forming small alliances and competing with other occupiers. Whenever outside entities entered the scene, including the MPHA and Marina’s charity, factions jockeyed with each other to align themselves with these organizations, hoping to secure a leg up over other residents. The same might be said for residents’ orientation toward their legal team. Rather than collectively interfacing with their lawyers through established representatives, members of each faction would scramble for the lawyer’s attention. Exasperated, each lawyer would demand a single person with whom to exchange information, leaving it up to the residents to resolve their struggle over representation.

One representative interaction of this sort occurred just before the occupiers’ final day in court. Marius, one of their lawyers, had come to Rivenland to prepare them for the ruling. He was standing in the section of the field run by First People First. Muhammad, a member of Residents Unite, ran over and asked Marius, “Do you represent all of us?” in a patois of English and Afrikaans.

“What’s your name?” Marius responded. “Muhammad Laurie,” he called back. “And is your name on the list?” Marius asked. “A list?” Muhammad thought to himself. “What list?” “No,” he replied, “but I have a

number on my shack.” An employee of the Department of Human Settlements had come and painted a number on Muhammad’s shack – a temporary address of sorts – just as he had on everyone’s structure who had been there since the occupation began. This was part of a standard


process called enumeration in which the municipality formally logs every informal structure in order to keep track of new buildings and verify how long each has been there.

Marius did not seem to care. He repeated his question, this time with clear overtones of frustration. “Is your name on the list?!” About twenty of Marina’s followers were gathered around them. One of them shouted, “No!” on Muhammad’s behalf, almost relishing the fact that he was not on the list. “No!” she repeated. “No,” Muhammad responded quietly, puzzled by the concept of their lawyer representing some of the squatters but not all of them even though they were in identical predicaments. “Then I don’t represent you,” Marius barked, brushing off imaginary dirt from the air in front of him. The factionalism on the field was refracted through their legal representation, intensifying in the process. What began as minor political differences turned into formalizing factions opposing one another. These groupings did not reflect preexisting political divisions, but instead actively constructed them in the process of faction formation. This was part and parcel of an individualized competition for land in which serialized residents formed small factions to defend their limited gains against perceived incursions. This mode of struggling for representation would ultimately lead the High Court judge to dismiss the residents as opportunists “illegally invading land,” to quote the Mayor’s spokesperson. He did not view them as a coherent community that might function as one of the populations in Chatterjee’s political society. When the judge finally issued his ruling1, he upheld the eviction order. He began by reproaching the occupiers as “opportunists” in Afrikaans, lecturing them about how the South African government was trying to secure their futures and that their disorderly composition made this impossible. He went on to condemn the MPHA: “Whether it’s about their egos or the depth of their pockets and their own self-indulgence is not relevant,” the judge insisted. These people “exploit the community…I know that they sit here in this court [today],” he told the audience. He was not pulling any punches. “For their own selfish purposes, they abuse the homeless and the poor…This is criminal on the face of it,” he declared. “Such elements do not belong in an ordered society, who then abuse their own people who are vulnerable to their schemes.”

The judge dismissed the MPHA as a group engaging in “haphazard business” as opposed to “an organization that fights for the rights of backyard dwellers.” It was as if he were criticizing them for failing to form a hegemonic body, pursuing a different approach to their struggle over representation. The squatters’ penchant for individualized rather than collective demands particularly irked the judge. “That piece of land seems to me to be very nice with the sea air blowing over the hill,” he joked. “I want to stay there, so now I’m going to take me a piece of land so I can just sit there. Then it takes the City Council months to get to me, and since I built my place and brought my children, even if I’m brought to court, it’s now too late to evict me. It does not work like that.”

1 All quotations are taken from the ruling in Stad Kaapstad v Onwettige Okkupeerders van Erf 41078 Mitchells Plain en Anders (9855/2011) [2011] ZAWCHC 538 (30 August 2011), available online from the South African Legal Information Institute (http://www.saflii.org/za/cases/ZAWCHC/2011/538.html). While most rulings are in English, the judge prefaced his remarks with the statement, “Ek doen hierdie opmerkings in Afrikaans, want ek weet die meeste van die betrokke persone in die hof is Afrikaanssprekend. [I write these comments in Afrikaans because I know most of the people involved in this case are Afrikaans speakers.]” As such, all translations from Afrikaans in this section are my own.


The competitive haggling among factions produced an image of disorder that particularly irked the judge. The occupiers’ names on the list he received from Sheldon did not match those on the list provided by the Department of Human Settlements officials monitoring the site. This was also one of the issues Aisha described in her journal when she wrote about that day: “They called out the names of the people who should be inside – those on the list of people, which I was on.”

The judge concluded his ruling by insisting that the remedy should not be to reward those who have jumped to the front of the line, but rather to help them inresert themselves onto the housing waiting list in an orderly fashion. Lawyers, he suggested, should provide guidance getting them back on the list in such a way as to ensure “that things run smoothly and you do not have this situation” That was his approach to the victims. The perpetrators, however – those who made the occupation appear legitimate and orderly in the first place – would face possible charges: “The deceivers must be denounced.” He then read the eviction order, giving the occupiers a month to vacate the field. Holfield: Achieving Legibility as a Population

The Rivenland occupiers saw the state as a partner in the occupation, and they comported themselves accordingly: as a group of property holders in the making. As atomized individuals acting out of self-interest, they formed small, mutually exclusive factions with the aim of protecting their claims to the land against occupiers they viewed as posing threats to those claims. The irony was that it was this factionalism that rendered them illegible to the judge as a population in need of housing. Instead, they appeared to him as disorderly opportunists whose immediate demands for inclusion undermined the formal rationality of the government’s housing delivery program. It was in this sense that he dismissed them as “queue jumpers.”

Holfield was a different story entirely. From the outset, the occupation was articulated as a collective project, rather than the centralized distribution of plots to prospective homeowners. Nor was this occupation organized by an outside group, let alone by political operatives who did not even live there. Instead, it was self-organized by a group of informal settlement residents who could no longer stay in their former homes. In February 2012, a few hundred backyarders from the township immediately to the west of Mitchell’s Plain – Philippi – left their homes to seek vacant. Philippi was mostly white-owned farmland under apartheid, but as mobility conrols were lifted in the 1980s, black residents from underdeveloped rural areas came pouring back to cities in search of employment. Some paid rent for the privilege of erecting a shack in a backyard; others sought out land in informal settlements, but these grew overcrowded as migration continued.

Khwezi was one such migrant. Now a bus driver and shop steward for the transport workers’ union, he was once a militant in the ANC’s armed wing. After spending the 1970s smuggling comrades back and forth across the border with Lesotho, he was stationed in Cape Town in 1981. He found a home in another township called Crossroads but soon had to flee the extreme violence that made the neighborhood notorious in the 1980s (Cole 1987). Once he fled, Khwezi’s cousin took his house, and he was forced to find somewhere else to live. He built himself a shack in a newly formed informal settlement in Philippi and lived there until 2012, when the place grew so overcrowded that he decided he could no longer live there.

A Holfield resident named Mncedisi told me a similar story. She was renting a shack in a backyard in Philippi, not far from Khwezi’s old informal settlement. But she remained out of


work, and the meager sums she borrowed from friends and family would not suffice once her landlord raised the rent. Desperate for shelter, she abandoned her backyard shack and built a new place in Holfield.

Both Khwezi and Mncedisi participated in the February 2012 wave of settlement, erecting shacks on an open field along a major thoroughfare connecting Mitchell’s Plain to the city center. From their perspective, it was a large, empty plot of sandy land populated solely by dry brush and a few grassy spots. It was clear no one was using it. Unbeknownst to them, the field was actually two adjacent pieces of private property – one owned by a sand mining business, and the other by a demolition company. The former used the land for dumping but had never actually mined on the plot; the latter was simply an absentee landlord. They also did not worry about the fact that there was a fairly large, well-organized middle-class neighborhood on the other side of the road. This neighborhood was effectively entirely “Colored,” whereas the Philippi occupiers were primary isiXhosa-speaking black “Africans.”

In this first wave, the occupiers cultivated a collective spirit, though it was not quite that of a social movement. None of the occupiers framed their actions in terms of post-apartheid land restitution, let alone decommodification, nor did they talk about making collective demands on the municipal government. In Cape Town, housing-related social movements typically deliver memoranda to government officials at the Civic Center, listing their demands (Salo 2015; Staniland 2008; Willems 2011). But the Holfield occupiers simply wanted to be left alone; rather than making demands on the state, they hoped to evade its gaze altogether. This largely had to do with how they saw the state. Most of the participants in this first wave came from the same perpetually expanding informal settlement in Philippi. The majority of its residents were officially tolerated, but as new shacks emerged in its interstices, the Anti-Land Invasion Unit would try to have them removed. This perpetual tussle with an arm of the municipal state was an entirely different experience from the backyarders who occupied Rivenland. Backyard shacks are rarely policed by the Anti-Land Invasion Unit, and an external organization, the MPHA, framed their occupation as actually involving the state. But in the case of Holfield, residents shared a collective memory of the local government as a ceaseless antagonist.

For this reason, they represented themselves very differently than did the Rivenland occupiers. In the case of Holfield, they sought safety in numbers, even appointing an unofficial representative of the occupation. This was Bongkinkosi, who was with the occupiers in the informal settlement in Philippi from the very beginning. Residents’ selection of this particular individual as their representative can partly be attributed to Bonginkosi’s charisma: he always seemed to have a plan, or at least spoke as if things were proceeding according to plan, and he was particularly skilled as an organizer.

As in most sizable land occupations, there were other people vying for leadership, including, as in Rivenland, front groups for political parties, or even people who explicitly represented major parties. A left-wing ANC front group called the Ses’khona People’s Rights Movement frequented the field in these early stages, as did members of the Economic Freedom Fighters (EFF), a self-proclaimed Marxist-Leninist-Fanonist party that is currently South Africa’s third largest in terms of parliamentary presence. Bonginkosi openly disparaged Ses’khona, as did the members of the occupation committee that would be elected a month and a half later. None of the Ses’khona operatives actually lived on the field. And much like the MPHA in the case of Rivenland, Ses’khona tried to represent itself as facilitating land and housing distribution on behalf of the ANC. Because the MPHA did not characterize its project so flagrantly, it took longer for a group of residents to contest its leadership. But Ses’khona was so


heavy-handed in its approach, constantly referencing the party, that residents voted to expel them from the occupation.

In the case of the EFF, things were more complicated. An occupier who would subsequently be elected to the occupation leadership was a card-carrying party member, and he made an early attempt to secure a lawyer for the residents. The lawyer, also a quite vocal EFF member, offered to represent Holfield residents for free if they would agree to join the party. “But we told him no,” Anna recalled, when we discussed the beginnings of the occupation two years later. “We’ll just pay you.”

I later asked one of the committee members named Ntando, who would succeed Bonginkosi as chairperson, about the EFF presence in the occupation. He told me it was there to be sure, but that residents – and above all, Bonginkosi – were quite clear from the outset: no parties would be involved in the occupation. I asked him if he preferred the EFF’s politics to those of the other parties. “No, none of them,” he replied. “That’s what Ses’khona does. They just try to get more members, but they’re not living here. They don’t know us.” Many of the residents were quite apprehensive about openly aligning with opposition parties in DA territory. They just wanted access to the land, and they certainly did not want perceived partisanship to provoke housing officials or the ALIU any more than the occupation itself already did. They were all too aware that newspapers tended to report occupations as party-orchestrated affairs in the lead-up to local elections. While this is sometimes true, most actual participants are not formally aligned with the organizing party. They may become members through the occupation, but such actions are rarely carried out by developed party cadre.

When the occupation was not met with resistance, others from Philippi moved in. The ALIU can only intervene on private property when the landowners file for an eviction injunction, but it took them many months to do so. Once a few hundred shacks had gone up, others from the informal settlement in Philippi joined them, as did a number of backyarders from that township. Then squatters facing eviction in Strandfontein, another township immediately south of Philippi and west of Mitchell’s Plain, began to join the occupation. This brought in a number of Afrikaans speakers, who quickly established a “Colored” corner of the occupation. By the end of the year, there were at least 6000 inhabitants living in Holfield.

This corner grew once Bonginkosi led a march some five miles to the Rivenland field as it was facing eviction. This was immediately after the final ruling, when the residents had a month to find alternative accommodation. Some two dozen Holfield occupiers urged the Rivenland residents to move with them back to Holfield, explaining that there was safety in numbers, and that there were already hundreds of shacks up. Besides, unlike in Rivenland, the ALIU and police had not been a daily thorn in their side. Certainly police harassed residents, but they could not legally remove them from the field. “We’re not going anywhere unless you’re going with us. Move with us!” Bonginkosi told the residents. I could not help by think about the stark contrast in relation to the exclusivist politics I had observed in Rivenland. This was a very different approach to the struggle over representation. A few dozen evicted residents followed the contingent back to Holfield. Over the course of the Rivenland struggle, some of the residents, especially those affiliated with Residents Unite, had made middle-class activist contacts at a Cape Town-based non-governmental organization. A few of these activists came with their cars and borrowed trucks – bakkies, as they call them in Afrikaans – to help transport the residents’ belongings. At this point, the ALIU had repossessed most of their building materials, but when they arrived at Holfield, they received a warm welcome. They were shown into the larger shack that functioned as a crèche, and told to rest. Residents brought them food and began to move


them into shacks while they built their own. Most of them identified as “Colored” and so moved into this section of the occupation, though it depended where there was space. Many of them lived among amaXhosa residents. While they were initially apprehensive, due to both linguistic and cultural barriers, the hospitality they received more than made up for it. For the first few months, representative power was concentrated in the hands of Bonginkosi alone. He was in direct communication with the occupiers’ pro bono lawyer, who they had secured through the Legal Resources Center, and he seemed to have amicable relations with representatives of the Department of Human Settlements whenever they would stop by Rivenland. Over time, however, residents grew skeptical of his authoritarian tendencies. Khwezi told me that at the height of Bonginkosi’s power, residents would line up outside his shack waiting to talk to him, and the queue would stretch around the corner. It began as a way for residents to access their lawyer, or at the very least figure out what was going on with their case, and it quickly became a venue for mediation in interpersonal disputes, as well as informal governance of the settlement more broadly. Beyond these individualized sessions, he would periodically brief the community, but often this was to a crowd of hundreds. Many people wanted personalized information, or at the very least, assurance that eviction was not immanent. Or else they would urge Bonginkosi to negotiate some concession from the municipality. Over time, Bonginkosi grew less active in settlement-level politics. He accepted a gig from the City coordinating a toilet-cleaning operation in Khayelitsha, the black “African” township immediately east of Mitchell’s Plain. The idea was that he would secure jobs for Holfield residents in return for political support. Residents demanded an alternative, but they also feared splitting the population into rival factions. Their solution was to elect a twelve-person committee that included both Bonginkosi and his chief rival, a middle-aged man who went by Ntando, introduced above. It was evenly split between men and women, though when I attended their meetings, the men would invariably dominate discussion. But the inclusion of both Bonginkosi and Ntando, as well as other more neutral members, eased the transition to a more representative body. Residents’ demands for this newly democratized committee appeared to have less to do with abstract principles of democracy than with the practical necessity of accessing information about the status of their court case. Just as in Rivenland, court dates proliferated. Given that Bonginkosi stopped reporting back to residents with any frequency, no one seemed to know what was going on. Sheldon, the lawyer from the Legal Resources Center, was involved in too many cases and no longer prioritized Rivenland. The EFF lawyer refused to represent them pro bono once they turned down his offer of joining his party. With these options exhausted, residents managed to secure another advocate named Paul through the Legal Resources Center. Paul worked with Stuart, co-founder and executive director of the Socio-Economic Rights Institute in Johannesburg. SERI is easily the most knowledgeable legal entity in the country on questions regarding housing and evictions, having won major Constitutional Court cases in Durban, Johannesburg, and elsewhere. Wilson’s work was at the center of these victories. But more important than any abstract conception of expertise was Paul and Stuart’s combined ability to represent the residents as a community – as a coherent population really – rather than as opportunists, queue jumpers, or other self-interested individuals whose presence in Holfield was represented as fleeting and nomadic. The municipal government had urged the owners of the two adjacent plots upon which Rivenland was built to apply for evictions so that the High Court could issue interdicts. This was in February, immediately after the first wave of Philippi squatters set up shop. At the time, the


City’s advocate would subsequently argue, “the City had the resources to accommodate 100 households under its Emergency Housing Programme.”2 In other words, when there were only 300 or so squatters on the land, the government could find them alternative accommodation. This was feasible. But with the population now roughly twenty times that number, “the demand for housing under the City’s Emergency Housing Programme outstrips its supply.”

The City’s advocates argued that the case was “unprecedented” — they used that word throughout their appeal — insofar as they were dealing with the question of relocating six thousand people. Because the housing supply was overwhelmed by demand, they maintained, “the City is not in a position to immediately provide emergency housing to all persons in need thereof.” Besides, the City warned the landowners repeatedly, and drawing on precedent, the primary responsibility for protecting their property was theirs. “We reiterate, this is unprecedented!”

Throughout the hearing, the City represented the squatters as “opportunists,” to use their advocates’ word. This fit with the City’s larger attempt to represent all land occupiers as free riders in search of a quick buck rather than as homeless people in need. “It is submitted that the legal position is that opportunists should not be enabled to gain preference over those who have been waiting for housing, patiently, according to legally prescribed procedures,” they wrote. But where were they to go? Even the City’s advocates acknowledged the situation of many of the Rivenland squatters: “It is apparent from the questionnaires that form part of the record in this litigation that the vast majority of residents have settled on the property either because they could not afford the rental of their previous homes or because they wanted their own homes or because they were forced to move out of their previous homes” — but these same residents are tarred as “opportunists.” This is despite the fact that “the residents assert in terms that they do not wish to bump anyone off the housing list” and that they “do not assert a claim for formal, permanent housing.” So what is the nature of this opportunism?

Ultimately, it boiled down less to a specific instance and more to the way the government imagined the logic of land occupations. When the participants petitioned the City and demanded that it purchase the plots, the City’s advocates rejected this as unfeasible and above all, immoral. “Private land is vulnerable to unlawful occupation,” they argued, as if there were not countless occupations on public land in the same township. “Accordingly, if this Court were to order that as a result of the unlawful occupation of the properties…the City must purchase or lease this land, this would have the unfortunate consequence of all private land being under threat of occupation by unlawful occupiers….Indeed, this would actively encourage people to invade private land, and in certain circumstances enable land owners to encourage/permit unlawful occupation with the sure knowledge that the City will be ordered to acquire the land.” This is textbook moral hazard, with the City assuming that accommodating the squatters would both incentivize them to occupy private property everywhere, knowing full well that there were no legal consequences for doing so; and incentivize landowners to tolerate (or even encourage) occupation, as this would force the City to purchase their holdings.

“Land invasion is inimical to the systematic provision of adequate housing on a planned basis,” they continued. Occupiers are invariably “opportunists [who] should not be enabled to gain preference over those who have been waiting for housing, patiently, according to legally prescribed procedures.” This unmanageable opportunism is counterposed to orderly subjects of

2 The City’s argument in Lyton Props and Robert Ross v Occupiers of isiQalo and City of Cape Town, from May 20, 2013, is available in full on the SERI website (http://www.seri-sa.org/images/isiQalo_City_Heads_FINAL.pdf). All further references to statements by the City’s advocates are to be found in this document.


redistributive democracy – “those who have been waiting for housing, patiently.” The advocates continued: “For this reason, the residents should not be permitted to claim permanent housing, ahead of anyone else in a queue.”

Paul and Stuart, acting on behalf of the residents, challenged the incoherence of the City’s argument3. First, the very notion that they are opportunists is belied by the fact that the City admits that they are largely homeless. Whereas in Kapteinsklip the City was able to successfully represent the squatters as opportunists vying with one another for plots of land, in the case of Siqalo it failed to do so. If the City argued that the spontaneity of land occupations threatened the functioning of the delivery apparatus, the squatters’ lawyers responded by questioning the inflexibility of the City’s plans: “The failure of a municipality to plan for or foresee the possibility of the eviction of a large number of poor people is no excuse for refusing to formulate a rational plan to provide alternative accommodation, once the possibility of an eviction and consequent homelessness is drawn to its attention.” It is undoubtedly frustrating to those overseeing housing delivery when unanticipated externalities threaten the systems’ very functioning. Yet this is to maintain formal rationality as an end itself, forgetting that the entire reason the City even has a housing policy is to accommodate those in need. If the Rivenland occupiers were not in fact opportunists, as the City alleged, but a community of 6000 people rendered homeless who lived on a roadside field as a last resort, the City’s job would be to recalibrate its plan in such a way as to accommodate them, or at the very least tolerate them. The alternative would be to condemn the occupiers to perpetual housing limbo. “Regrettably,” they argued, “that is exactly the stance taken by the City in this application.”

On June 3, 2013, the judge ruled in favor of the occupiers, but with some qualifications. “The City of Cape Town and the two landowners agree,” he wrote, “that the consequences of an eviction at this stage will render the majority of the occupants homeless.”4 He discussed them as a population, never as individual opportunists. “Judging from the expert reports filed, [they] have settled to the extent that there are now some 1800 structures, including crèche’s [sic] and spaza shops5 on the land.” He never refers to contending factions or internal strife, as was the case in Kapteinsklip, but always discusses them en bloc. For now, the squatters were safe. Discussion and Conclusion This then is how Rivenland was designated for eviction, whereas Holfield was tolerated, ultimately growing to roughly 18,000 residents in 6000 shacks. Now seven years since Rivenland was cleared, the municipal government currently has no plans to develop the land. Indeed, none of the standard explanations for evictions hold any water in this comparison. We might have expected Holfield to be evicted given that it was located on private property, whereas Rivenland was on public land. Likewise, we might have expected that mostly “African” squatters living in a “Colored” area might raise the state’s hackles, especially given the sustained campaign by Colored middle-class residents to have them removed; by contrast, there was no

3 Their May 13, 2013 argument in defense of the Siqalo occupiers in Lyton Props and Robert Ross v Occupiers of isiQalo and City of Cape Town can be found on the SERI website (http://www.seri-sa.org/images/RobertRoss_ 1stRespondentsHeads.pdf). Subsequent references to the squatters’ advocates’ defense correspond to this document. 4 The ruling in Lyton Props and Robert Ross v Occupiers of isiQalo and City of Cape Town is available in full on the SERI website (http://www.seri-sa.org/images/isiqalo_judgment_3Jun13.pdf). All quotations in this paragraph are from this ruling. 5 Informal convenience stores.


such mobilization in Rivenland. And Holfield was exceedingly visible, located along a major road, whereas Rivenland was in a substantially poorer and certainly less visible location. But of course it was Rivenland that was evicted and Holfield that gained the right to stay put. These sorts of blanket generalizations are what I have critiqued as decisionist explanations, taking as a given a coherent, self-conscious, high modernist state with its own set of interests that acts “upon” populations. What they fail to explain is how these populations are constituted as such in the first place. And this process of constitution is deeply impacted by how squatters in each respective occupation see the state. In Rivenland, occupiers initially viewed the state as a partner in taking the land, which they did not comprehend as “taking” in any sense. They understood the MPHA to be an arm of the Department of Human Settlements distributing individual plots to each family unit. They therefore comported themselves as a Sartrean series: as participants in the same action, but without participating collectively. Instead, they competed with one another for access to “the state,” which turned out to be nothing more than a political party’s front group. In Holfield, meanwhile, occupiers were deeply fearful of, if not hostile toward, the state. Unlike their backyarder counterparts in Rivenland, these squatters had many years of experience battling over the legitimacy of their informal settlements. They therefore viewed the state with quite a bit of suspicion. As a result, they organized themselves into what Sartre describes as a fused group, essentially a collectivity with a commonly agreed upon set of interests. Rather than dividing themselves into small factions aligned with various outside actors, they self-organized and elected a single committee to represent them. They were therefore recognized as a genuine population in need, whereas the Rivenland occupiers were dismissed as opportunists jockeying for handouts. The key point is that only one of the two occupations was read by the state as a “population”; the other was not legible to it as such. This poses a direct challenge to Chatterjee’s assumption that the informally housed in Southern cities negotiate with the state as populations, or what he calls “political society,” rather than as individual, rights-bearing citizens, his “civil society.” These populations, as I demonstrate here, are not always ready-to-hand, but are instead the outcome of struggles over representation within each settlement, which in turn depend upon residents’ view of the state. “Population” then is not our point of departure, but an outcome of sorts, one that in turn affects eviction decisions made by the state. The point is that the state does not simply gaze out over a landscape of pre-existing populations, but that the “population-ness” of these groups varies in relation to how residents see the state. This is why this landscape is so uneven, with visible population peaks obscuring serial valleys. It is particularly bewildering that Chatterjee misses this point given the Gramscian provenance of his key concepts, political and civil society. For Gramsci (1971), engagement with the state’s formal institutional channels (political society) is inextricable from collective organization beyond its purview (civil society). This is precisely what we observed in both cases considered here. In Rivenland, civil society was fragmented into what I called a series, whereas in Holfield it was unified in the form of an elected representative committee. How residents decided who speaks for whom – what I called the struggle over representation – cannot be separated from how residents project themselves to the state, or what I called the struggle over recognition. If states see populations, but “population-ness” is itself the outcome of struggles dependent upon residents’ perceptions of the state, we cannot separate political from civil society as Chatterjee does.


What this separation does is to strip the urban poor of their potentiality for collective political decision-making. It naturalizes them as passive populations without ever explaining how they became populations in the first place. This is precisely the point of this paper, namely, to demonstrate how subproletarian politics is not, pace Chatterjee, beyond the realm of civil society. It is within civil society that poor people’s representational struggles, just like those of the working and middle classes, are resolved. Many poor people in Southern cities may live in neighborhoods that are both socially and spatially distinct from those of the rest of society, but this does not mean that they are stripped of all representational agency, the lifeblood of civil society. To adapt Helen Safa’s (1970) notorious critique of Oscar Lewis, the poor are like everyone else, Partha! References Alexander, Peter, Carin Runciman, and Trevor Ngwane. 2014. Media briefing: Community

protests. Paper presented at the Social Change Research Unit, University of Johannesburg, South Africa, 12 February 2014.

Andranovich, Greg, Matthew J. Burbank, and Charles H. Heying. 2002. Olympic cities: Lessons learned from mega-event politics. Journal of Urban Affairs 23 (2): 113–131.

Anjaria, Jonathan Shapiro. 2009. Guardians of the bourgeois city: Citizenship, public space, and middle-class activism in Mumbai. City and Community 8 (4): 391–406.

Atkinson, Rowland, and Gary Bridge, eds. 2005. Gentrification in a global context: The new urban colonialism. London: Routledge.

Atuahene, Bernadette. 2006. Land titling: A mode of privatization with the potential to deepen democracy. Saint Louis University Law Journal 50 (3): 761–82.

Auerbach, Adam M. 2016. Clients and communities: The political economy of party network organization and development in India’s urban slums. World Politics 68 (1): 111–48.

Baviskar, Amita. 2006. Demolishing Delhi: World class city in the making. Mute 2 (3): 88–95. Bayat, Asef. 2013. Life as politics: How ordinary people change the Middle East. Stanford:

Stanford University Press. Begg, Iain. 1999. Cities and competitiveness. Urban Studies 36 (5/6): 795–809. Berrisford, Stephen and Michael Kihato. 2006. The role of planning law in evictions in sub-

Saharan Africa. South African Review of Sociology 37 (1): 20–34. Bhan, Gautam. 2016. In the public’s interest: Evictions, citizenship, and inequality in

contemporary Delhi. Athens: University of Georgia Press. Bond, Patrick. 2000. Cities of gold, townships of coal: Essays on South Africa’s new urban

crisis. Trenton: Africa World Press. Butler, Tim. 2007. For gentrification? Environment and Planning A 39 (1): 162–81. Çelik, Ercüment. 2010. Street traders: A bridge between trade unions and social movements in

contemporary South Africa. Freiburg: Nomos. Chance, Kerry Ryan. 2015. Transitory citizens: Contentious housing practices in contemporary

South Africa. Social Analysis 59 (3): 62–84. Chance, Kerry Ryan. 2018. Living politics in South Africa’s urban shacklands. Chicago:

University of Chicago Press. Chatterjee, Partha. 2004. The politics of the governed: Reflections on popular politics in most

of the world. New York: Columbia University Press.


Chatterjee, Partha. 2010. Empire and nation: Selected essays. New York: Columbia University Press.

Chatterjee, Partha. 2011. Lineages of political society: Studies in postcolonial democracy. New York: Columbia University Press.

Cole, Josette. 1987. Crossroads: The politics of reform and repression, 1976-1986. Johannesburg: Ravan.

Dasgupta, Keya. 2003. Evictions in Calcutta: Creating the spaces of “modernity.” City 4, 31–43. Davis, Lisa Kim. 2011. International events and mass evictions: A longer view. International

Journal of Urban and Regional Research 35 (3): 582–99. de Soto, Hernando. 2000. The mystery of capital: Why capitalism triumphs in the West and fails

everywhere else. New York: Basic. Dovey, Kim and Ross King. 2011. Forms of informality: Morphology and visibility of informal

settlements. Built Environment 37 (1): 11–29. Foucault, Michel. 1995. Discipline and punish: The birth of the prison. New York: Vintage. Foucault, Michel. 2007. Security, territory, population: Lectures at the Collège de France,

1977—1978. New York: Picador. Ghertner, D. Asher. 2015. Rule by aesthetics: World-class city making in Delhi. Oxford: Oxford

University Press. Ghertner, D. Asher. 2011a. Rule by aesthetics: World-class city making in Delhi. In Worlding

cities: Asian experiments and the art of being global, ed. A. Roy and A. Ong, 279–306. Malden: Blackwell.

Ghertner, D. Asher. 2011b. Green evictions: environmental discourses of a “slum-free” Delhi. In Global political ecology, ed. R. Peet, P. Robbins, and M. Watts, 145–65.

Gibson, Nigel. 2011. Fanonian practices in South Africa: From Steve Biko to Abahlali baseMjondolo. Durban: University of KwaZulu-Natal Press.

Goldman, Michael. 2011. Speculative urbanism and the making of the next world city. International Journal of Urban and Regional Research 35 (3): 555–81

Gordon, Ian. 1999. Internationalisation and urban competition. Urban Studies 36 (5/6): 1001–16. Harvey, David. 1989. From urban managerialism to entrepreneurialism: The transformation in

urban governance in late capitalism. Geografiska Annaler 71B (1): 3–17. Hiller, Harry H. 2000. Mega-events, urban boosterism and growth strategies: An analysis of the

objectives and legitimations of the Cape Town 2004 Olympic bid. International Journal of Urban and Regional Research 24 (2): 449–58.

Holston, James. 2008. Insurgent citizenship: Disjunctions of democracy and modernity in Brazil. Princeton: Princeton University Press.

Huchzermeyer, Marie. 2001. Housing for the poor? Negotiated housing policy in South Africa. Habitat International 25 (3): 303–31.

Huchzermeyer, Marie. 2003. Housing rights in South Africa: Invasions, evictions, the media, and the courts in the cases of Grootboom, Alexandra, and Bredell. Urban Forum 14 (1): 80–107.

Huchzermeyer, Marie. 2011. Cities with ‘slums’: From informal settlement eradication to a right to the city in Africa. Cape Town: University of Cape Town Press.

Loretta Lees, Hyun Bang Shin, and Ernesto López-Morales, eds. 2015. Global gentrifications: Uneven development and displacement. Bristol: Policy Press.

Lees, Loretta, Hyun Bang Shin, and Ernesto López-Morales. 2016. Planetary gentrification Cambridge: Polity.


Lees, Loretta, Tom Slater, and Elvin Wyly. 2008. Gentrification. London: Routledge. King, Winter. 2003. Illegal settlements and the impact of titling programs. Harvard International

Law Journal 44 (2): 433–72. Levenson, Zachary. 2017a. Social movements beyond incorporation: The case of the Housing

Assembly in post-apartheid Cape Town. In Global resistance in Southern perspective: The politics of protest in South Africa’s contentious democracy, ed. M. Paret, Carin Runciman, and L. Sinwell, 89–104 in. London: Routledge.

Levenson, Zachary. 2017b. Precarious welfare states: Urban struggles over housing delivery in post-apartheid Cape Town. International Sociology 32 (4): 474–92.

Levenson, Zachary. 2018. The road to TRAs is paved with good intentions: Dispossession through delivery in post-apartheid cape Town. Urban Studies 55 (14): 3218–33.

Massey, Doreen. 2007. World city. Cambridge: Polity. McAdam, Doug, Sidney Tarrow, and Charles Tilly. 2001. Dynamics of contention. Cambridge:

Cambridge University Press. McDonald, David A. 2008. World city syndrome: Neoliberalism and inequality in Cape Town.

London: Routledge. Offe, Claus. 1984. Contradictions of the welfare state. Cambridge: MIT Press. Palm, Kaylynn. 2013. ANC moves in for ’Plain. Plainsman, July 31, 1, 3. Papier, Melissa. 2011. Voter turnout strong among protests. People’s Post Mitchell’s Plain, 24

May: 6. Pithouse, Richard. 2006. Our struggle is thought, on the ground, running: The University of

Abahlali baseMjondolo. Centre for Civil Society Research Report 41. Durban: University of KwaZulu-Natal.

Pithouse, Richard. 2008. A politics of the poor: Shack dwellers’ struggles in Durban. Journal of Asian and African Studies 43 (1): 63–94.

Ranslem, Duncan. 2015. “Temporary” relocation: Spaces of contradiction in South African law. International Journal of Law in the Built Environment 7 (1): 55–71.

Roy, Ananya. 2003. City requiem, Calcultta: Gender and the politics of poverty. Minneapolis: University of Minnesota Press.

Safa, Helen. 1970. The poor are like everyone else, Oscar. Psychology Today 4 (4): 26–32. Salo, Ken E. 2015. Local practices and trans-local solidarities: Reflections on anti-eviction

practices on the Cape Flats, South Africa, and in South Side Chicago. In Cities and inequalities in a global and neoliberal world, ed. F. Miraftab, D. Wilson, and K. Salo, 217–26.

Sartre, Jean-Paul. 2004. Critique of dialectical reason, volume one. New York: Verso. Sassen, Saskia. 2001. The global city: New York, London, Tokyo. Princeton: Princeton University

Press. Schmitt, Carl. 1985. Political theology: Four chapters on the concept of sovereignty. Chicago:

University of Chicago Press. Scott, James C. 2009. The art of not being governed: An anarchist history of upland Southeast

Asia. New Haven: Yale University Press. Scott, James C. 1998. Seeing like a state: How certain schemes to improve the human condition

have failed. New Haven: Yale University Press. Searle, Llerena Guiu. 2014. Conflict and commensuration: Contested market making in India's

private real estate development sector. International Journal of Urban and Regional Research 38 (1): 60–78.


Simone, Abdoumaliq. 2004. For the city yet to come: African life in four cities. Durham: Duke University Press.

Slater, Tom. 2017. Planetary rent gaps. Antipode 49 (S1): 114–37. Smith, Neil. 1996. The new urban frontier: Gentrification and the revanchist city. London:

Routledge. Smith, Neil. 2002. New globalism, New urbanism: Gentrification as global urban strategy.

Antipode 34 (3): 427–50. Staniland, Luke. 2008. “They know me, I will not get any job”: public participation, patronage,

and the sedation of civil society in a Capetonian township. Transformation 66/67: 34–60. Starecheski, Amy. 2016. Ours to lose: When squatters become homeowners in New York City.

Chicago: University of Chicago Press. Symphony Way Pavement Dwellers. 2011. No land! No house! No vote! Voices from Symphony

Way. Cape Town: Pambazuka. Tilly, Charles. 2004. Social movements, 1768–2004. Boulder: Paradigm. Tilly, Charles and Sidney Tarrow. 2007. Contentious politics. Boulder: Paradigm. Wei, Ye Hua Dennis. 2012. Restructuring for growth in urban China: Transitional institutions,

urban development, and spatial transformation. Habitat International 36 (3): 396–405. Weinstein, Liza. 2013. Demolition and dispossession: Toward an understanding of state violence

in millennial Mumbai. Studies in Comparative Interational Development 48 (3): 285–307.

Weinstein, Liza. 2014. The durable slum: Dharavi and the right to stay put in globalizing Mumbai. Minneapolis: University of Minnesota Press.

Weinstein, Liza. Forthcoming. Evictions: Reconceptualizing housing insecurity from the Global South. City and Community.

Willems, Wendy. 2011. Social movement media, post-apartheid (South Africa). In Encyclopedia of social movement media, ed. J. Downing, 492–95. London: Sage.

Xu, Jiang, and Anthony Yeh. 2009. Decoding urban land governance: state reconstruction in contemporary Chinese cities. Urban Studies, 46 (3): 559–581.

Documents

Becoming a Population