I have just published an ethnographic monograph about mass evictions in contemporary South Africa called Delivery as Dispossession: Land Occupation and Eviction in the Postapartheid City (2022, Oxford University Press). But before I proceed to elaborate its central arguments, I should clarify what I mean by two concepts in the title: delivery and dispossession. Then I can explain what it means to think about delivery as dispossession.

In the book, dispossession refers to the forcible separation of people from their homes. Under apartheid, 3.5 million Black South Africans were dispossessed from their places of residence, relegated either to distant rural areas or remote townships on the outskirts of the city. And delivery? Since the end of apartheid in 1994, South Africa’s first democratic government has delivered roughly 4 million houses to those in need. In this sense, we might provisionally think of dispossession and delivery as antithetical concepts mapping onto each of these periods, with apartheid characterized by racialized dispossession and the post-apartheid period by a remedial project of delivery.

Unfortunately, this neat periodization doesn’t do us any analytic favors. The government’s housing delivery program actually began under apartheid, and it was only in building new housing on the periphery that dispossession could proceed unabated. And today, the post-apartheid government regularly evicts both individuals and, more commonly, entire settlements – typically in the name of protecting its housing delivery program from those who refuse to wait their turn. Both under apartheid and since, delivery and dispossession have gone hand in hand, the one enabling the other. The trick is not to map them onto mutually distinct historical epochs but to understand how the two concepts work together in novel ways in each period.

Delivery as Dispossession makes three interrelated arguments along these lines. First, it tries to understand why a government that stakes its very legitimacy on reversing the damage wrought by centuries of apartheid, segregation, and colonialism evicts new land occupations and dispossesses residents on a regular basis. This book argues that it does so because, by labeling land occupations as disorderly and distinct from other residents who wait patiently, housing officials believe that dispossession enables delivery. This first argument, then, is about how government actors see residents. Most of the literature on mass evictions focuses on land grabs, development projects, and real estate profits, but the logic of evictions in post-apartheid South Africa is notably different: rather than economic, it is political. Under apartheid, the government built many hundreds of thousands of houses to which they could forcibly relocate Black residents living in cities. In other words, delivery enabled dispossession.

After apartheid, the relationship between these two state-initiated processes was reversed: today, housing officials believe that dispossession enables delivery. However, dispossession does not ultimately enable delivery, but instead reproduces the problem anew, driving residents lacking alternative options to occupy land yet again. The crucial point here is that housing officials think that dispossession enables delivery, and this is the primary rationale for evictions today. The post-apartheid government has provided millions of homes to needy residents since 1994, registering countless others on a waiting list. However, the scale of this need vastly outstripped the state’s capacity to deliver, and the housing backlog – those in need of housing – remained constant or actually increased in most major municipalities. This meant that residents had to provide for themselves in the meantime, and they did: by occupying land. But rather than a consequence of the state’s failure to deliver, these new occupations tend to be read by housing officials as its cause. They are evicted, paradoxically, in the name of protecting delivery, even if these evictions accomplish no such thing.

Ultimately, housing officials frame occupations as inherently disorderly, making the case to judges that housing delivery requires residents who wait patiently rather than “jumping the queue” and seizing housing on the spot. But not all occupations are evicted. Under which conditions then do they tend to be read as disorderly? This brings me to the second argument advanced in this book, which moves us from how state actors see residents to how residents see the state. How they see the state impacts how organizers articulate the project of occupation, which shapes the organizational form assumed by its participants. This organizational form, in turn, affects how the state sees them. When, for example, initiators frame an occupation as if it were a social movement, they may do so for a variety of reasons. In one of the cases considered in the book called Siqalo, organizers united residents under a single leadership because they had a history of clashing with government actors, whom they viewed as an obstacle to their project of attaining housing, and they saw unity as essential to resisting state violence.

But in a second case called Kapteinsklip, organizers framed their occupation quite differently: as the legitimate distribution of plots to needy residents. In this case, participants comported themselves as atomized recipients of housing. They participated in the occupation simultaneously, but they did not do so collectively, and their leadership quickly fragmented into competing factions. This contrast is crucial to making sense of eviction outcomes. We might expect a wary municipality to view the collective actors, rather than the atomized squatters, as the more substantial threat. But Delivery as Dispossession demonstrates that the opposite tends to be the case. Interacting with housing officials, lawyers, police, and judges through the mediation of an agreed-upon leadership makes the occupation appear orderly; while fragmenting into competing factions makes the occupation appear disorderly and, therefore, a threat to the order required for delivery. Siqalo was tolerated, but Kapteinsklip was evicted.

This is not to suggest some one-to-one correspondence between organizational form and eviction outcome. After all, we do need to understand the conditions under which one organizational form prevails over another. If this form is shaped by the way organizers articulate their collective project, we need to ask how these articulations come into being in the first place. This book argues that leaders’ social visions are informed by their respective understandings of the state. In the case of Siqalo, leaders articulated an occupation as a collective action because they viewed the state as an antagonist and mobilized accordingly. In Kapteinsklip, participants viewed the state as a partner in delivery rather than an agent of dispossession. In both cases, how residents saw the state affected how they mobilized, which shaped how they were seen by the state. States may “see” populations, but residents can shape how they themselves are “seen.” Yet their ability does not exist in a vacuum; it is itself informed by how organizers “see” the state. As should be evident, the state’s vision, much like residents’ vision, emerges through complex engagements, interactions, and struggles with antagonists. How the state sees populations, in other words, should not be our constant point of departure, but an outcome in its own right.

The law is the primary terrain upon which these struggles play out. South Africa’s 1996 Constitution was among the first in the world to guarantee “access to adequate housing” and freedom from “arbitrary evictions.” But in practice, this is incorrect: the Constitution only guarantees access to housing insofar as the state can “achieve the progressive realisation of this right” by “tak[ing] reasonable…measures, within its available resources.” While this might seem like splitting hairs, it is an enormously consequential point. The determination as to when the local state has the responsibility to provide housing, as well as when it can legally evict people, is made in the courtroom, and often Province-level High Court judges have the final say: each eviction requires “an order of court…after considering all the relevant circumstances.” This means that how occupiers collectively appear in the courtroom shapes judges’ decision-making. As the later chapters of this book demonstrate, when an occupation appears orderly, with a functioning self-government and little to no visible factionalism, judges are more likely to view it as a deserving population in need of housing. But when an occupation appears fractured into contending factions, judges are more likely to read its participants as greedy opportunists undeserving of toleration.

This connection is not straightforward to occupiers, whose collective appearances are not generated intentionally. Rather, they are unintended consequences of strategies designed for very different ends. In neither case did occupiers mobilize to communicate with government actors; in both, they hoped to fly below the state’s radar. But regardless, both occupations quickly found themselves in dialogue with the state, and more precisely, with legal representatives on the floor of the courtroom. This is the book’s third argument: while certainly progressive in guaranteeing socio-economic rights, the Constitutional guarantee to housing nevertheless forces occupiers onto the government’s terrain. This is a textbook instance of hegemony, which Peter Thomas defines as “the process by means of which social forces are integrated into the political power of an existing state”: occupiers’ interest in securing official toleration is integrated into property’s interest in adjudicating all land disputes in the courts. Of course, when occupations begin, participants are rarely concerned with toleration as such; they are typically far more interested in eluding the state altogether and straightforwardly taking the land that they need. Nevertheless, in nearly every case, residents find themselves arguing for their moral legitimacy in a courtroom.

As the historian Barbara Fields succinctly puts it, “Exercising rule means being able to shape the terrain.” And shaping the terrain is precisely what the 1996 Constitution did: it turns out that both the right to delivery and freedom from dispossession are qualified, and it is up to the courts to adjudicate. All housing struggles therefore inevitably find themselves squabbling over this or that policy through legal intermediaries rather than straightforwardly appropriating land and defending it directly. Constitutional rights then are a double-edged sword. Surely the right to housing is progressive, but it absolutely, to borrow the Comaroffs’ turn of phrase, “judicializes” politics, shifting all substantive debates onto a terrain shaped and controlled by government actors. And this terrain is, ultimately, about language: “forms and languages of protest or resistance,” the late anthropologist William Roseberry insisted, “must adopt the forms and languages of domination in order to be registered or heard.” But as soon as they translate their demands into this obligatory idiom, they find themselves no longer in protest or resistance, but in dialogue.

The relational theory of the state developed in Delivery as Dispossession is transportable far beyond South Africa, or even cases of housing delivery. It is a theory that insists that the state is not an airtight institutional space with the power to rule, but rather a site of ongoing contestation. On the one hand, this means that mass pressure impacts policy outcomes, and we therefore need to move beyond the notion that the state is fully autonomous, or that the actions of marginalized people do not impact governmental decisions. On the other, it means that popular struggles are often absorbed into the state, containing them in the process. Those who find themselves in dialogue with government actors often never meant to engage on this terrain. This is precisely what hegemony entails.

Zachary Levenson is an Assistant Professor of Sociology at the University of North Carolina, Greensboro and a Senior Research Associate in Sociology at the University of Johannesburg. Beginning in September, he will be a Donald D. Harrington Fellow in the Department of Sociology at the University of Texas at Austin.

This post is adapted from the preface to his book Delivery as Dispossession: Land Occupation and Eviction in the Postapartheid City, released last week on Oxford University Press.