Introduction

‘Expropriation without compensation’ (EWC) is a politically potent and simultaneously ambiguous term. It is politically potent not despite but precisely because of its ambiguity, in that it signals a radical departure from a land property regime that is patently illegitimate and unjust while obscuring how it is to be changed. It centres exclusively on the acquisition of land – thus on the nexus between the state and landowners – rather than on the distributive agenda – and thus on the nexus between the state and landless citizens. In this way, the EWC narrative sidesteps foundational questions of who should get which land, on what terms, for what purposes, where, and any wider agrarian reform agenda. These questions, which I have summarised as ‘who, what, where, how, why’, constitute the real politics of land reform and have been the focus of intense political negotiation, public debate and policy deliberation since the start of the political transition over three decades ago (Hall, Reference Hall2010, Reference Hall, Cousins and Walker2015). Their scale and complexity have frequently been contracted into ‘the land question’.

The promotion of EWC as the purported silver bullet to ‘the land question’ is relatively recent, arising from mounting factionalism within the ruling party after its 52nd National Conference at Polokwane in 2007, the emergence of an opposition party to the left of it in Parliament in 2014 and culminating in a parliamentary motion on EWC in 2018. This, in turn, birthed two distinct and overlapping processes driven by the legislature and executive, respectively: a parliamentary constitutional review committee to review and propose whether amendment was justified, and a Presidential Advisory Panel on Land Reform and Agriculture (PAPLRA) to advise on precisely the same matter, as well as a broader spectrum of policy questions. I characterise these processes as a form of ‘political theatre’ in that, while they addressed real issues, they were acted out performatively, for purposes and in ways quite distinct from what was being claimed, and primarily as a substitute for action rather than a precondition for it (Death, Reference Death2011). In this sense, the EWC debate that dominated much political discourse for a four-year period, while failing to yield any tangible outcome, legally or practically, arguably produced distinct political benefits. As a concept, EWC deftly obscures profound political differences and has been used to refer to a spectrum of measures ranging from the nationalisation of all land (as in the Economic Freedom Fighters’ (EFF) initial formulation) to state custodianship (in the EFF’s later formulation) or to selective case-by-case property acquisitions (as proposed by the African National Congress (ANC)).

The well-known story of how EWC moved to the centre stage of South African politics in 2018 is outlined in the Introduction to this volume, and here I rehearse only its headline features. Following the EFF’s motion in Parliament to review the property clause, the ANC amended the motion by adding several caveats about the rule of law and national food security, and together they mustered a majority and passed it on 27 February 2018. Parliament then set up a Constitutional Review Committee (CRC), which conducted mass public hearings around the country and received over 600,000 written submissions – more than at any time since the constitutional consultations in 1995. The bulk of written submissions objected to any constitutional amendment, while the overwhelming majority of those attending the hearings supported amendment – this distinction largely correlating with race in a predictable manner. The Committee, reporting in November, concluded in favour of amendment ‘to make explicit that which is implicit’, namely that EWC is permissible (CRC, 2018: 34). EWC, it affirmed, is a justifiable objective, which is already provided for, but amendment should nonetheless be pursued.

In response, Parliament mandated a new parliamentary ad hoc committee on section 25 (the property clause) to propose and consult on such an amendment. This inter-party committee, predictably hampered by the continued deadlock, took more than two years to propose new wording. The two-stage parliamentary process proved effective in separating the in-principle debate on EWC from its content: hearings provided a cathartic public spectacle as a foil to the political theatre among political parties in Parliament, and the technical work of developing an amendment which followed was out of public purview and drew less interest, allowing the political momentum to recede. A further bifurcation was achieved by establishing the parallel PAPLRA (see section on PAPLRA below). It all ended with a whimper, rather than a bang, with the amendment bill failing to secure the required two-thirds of the votes in the National Assembly on 7 December 2021 – four years after the EWC debate had emerged as part of the king-making deals across ruling party factions which had secured Cyril Ramaphosa’s presidency at the ANC’s 54th elective conference in December 2017.

Expropriation is one means by which land can be acquired to be made available for redistribution or restitution. Even once the legal framework is fixed, and policy is developed to determine when expropriation should take place and how compensation is to be addressed in diverse situations, the questions as to who this will help, and how, remain. The popular imagination has been seized by a polarised debate between those blaming the 1996 Constitution of the Republic of South Africa (the Constitution) for the failures of land reform, and those defending it (and incorrectly interpreting it as requiring compensation). But there was a third position, which several academics, lawyers and activists tried to develop and communicate throughout this process. This chapter sets out this third position: that the ‘property clause’ provides a powerful mandate for transformation and an under-developed right of equitable access to land for landless citizens. In this sense, the question of EWC occluded the question of redistribution. Now that the constitutional amendment bill has failed and this red herring lies dormant for now, a more productive conversation could address as its central concern the question of the ways by which and terms on which people can gain access to land.

Actors on the Stage

Who were the protagonists in the EWC debate? The EFF and the Democratic Alliance (DA), for their own respective political reasons, consistently equated section 25 with a requirement of compensation for expropriation. On that, they agreed. The DA considered any deviation from a compensation-based approach to expropriation, and therefore any amendment, as tinkering with the terms of political settlement reached in the 1990s and vociferously opposed it (Democratic Alliance, 2018). A spectrum of business interests and think tanks, such as the Institute for Race Relations and Centre for Development and Enterprise, expressed similar sentiments, warning of mass disinvestment and financial collapse. Yet in my numerous encounters with international financiers, South African banks, foreign embassies and trade missions during this period, my distinct impression was that they could absorb possible losses associated with the forcible confiscation of specific parcels of land, which could be factored into risk calculations, and therefore absorbed. Rather, their concern was about predictability: what they sought was clarity as to when compensation would be paid, when it would not and how this would be determined. This is the clarity that the ANC refused to provide.

The EFF, for its part, depicted the property clause as being the primary impediment to meaningful transformation of land relations and the culprit for the perpetuation of the legacy of colonial theft – a political manoeuvre that, paradoxically, allowed the ANC to abdicate from responsibility for its own failure to use the constitutional provisions to expropriate in the interests of land reform. It called for nationalisation as a decisive reversal of colonial land theft – the term ‘expropriation’ being misleading, as, in fact, what it meant is confiscation of all property, later clarified as a form of state custodianship. In its view, the Constitution has a presumption of compensation: ‘The Constitution as it is currently written does not allow either for expropriation without compensation or for the narrow, piece-meal expropriations advocated by liberals’ (EFF, 2018: 15). In their interpretation that the property clause presumes, even assures, compensation, then, the DA and EFF agreed. In contrast, the ANC’s perspective can only be described as incoherent. Combining some Constitution-blaming with investor-reassuring, its responses to the evolving debate could be described as ‘Talk EFF, walk DA’.

Actors in this public discourse not only took different positions on EWC but also meant different things by it. The ostensible convergence, when the ANC and EFF voted together in Parliament in February 2018, was illusory; they were not agreed on what EWC means, let alone how it would be applied. The ANC resolution at its 2017 conference declared that EWC should be ‘one of the key mechanisms available’ to government (ANC, 2017: para. 15). It did not say that the property clause would need to be amended to achieve this. It had not called for any change to the Constitution, nor did it say why EWC was needed or what problem it would solve. It insisted on caveats such as that land reform must not damage agricultural production or food security and specified unused and under-utilised land, or land held speculatively or indebted, as targets for EWC (ANC, 2017: paras. 16–17) but omitted any mention of expropriating productive farms or urban land.

The ‘property clause’ is a mandate for transformation. By this, I mean that it affirms the rights of individuals, communities and society at large to the transformation of property relations through constitutionally mandated land reform and related reforms. This is not to say that it is sacrosanct or could not be improved. Indeed, a compromise was certainly struck over property rights in the 1990s, but this is most starkly evident not in the new Constitution, but rather in the interim Constitution of the Republic of South Africa, Act 200 of 1993. It is worth returning to these moments where the parameters for the status of (inherited) property rights, and the terms of their taking, were established. It was in this early period of negotiations that the ANC rapidly shifted course and abandoned its (always ambivalent) promise of nationalisation. Questions of land justice were framed on the one hand as being central to transitional justice and simultaneously as a shared responsibility of the whole society, with the costs to be carried by the wealthy, including property owners, via the tax system and the national fiscus.

The Compromise in the Interim Constitution

The status of property rights under a new dispensation was among the most contentious issues in negotiating the terms of political transition in the 1990s. The 1993 ‘interim’ Constitution (Act 200 of 1993) protected existing property rights through the time of the first democratic election while providing for land restitution, paving the way for a Restitution of Land Rights Act shortly after the first democratic elections. In July 1991, the ANC’s National Conference adopted guidelines developed by its Land Commission which rejected any constitutional protection of property rights (Klug, Reference Klug2000: 127; see also Klug, Chapter 11, this volume). Yet a mere few months later, when formal negotiations commenced, the ANC had abandoned nationalisation as a policy goal, acceded to compensation being paid for expropriated property and considered options to fund a programme that would now involve substantial capital costs to the state (ANC, 1992; Hall, Reference Hall2010). A proposed tax on landowners to fund a ‘compensation account’ earmarked for land reform purposes was quickly abandoned (Klug, Reference Klug2000: 128).

Back in the early 1990s, as the details of the interim Constitution and Bill of Rights were debated (and again four years later in 1995 when debating the final Constitution), divisions between the ANC and the National Party (NP) government centred on whether expropriation of property should be allowed for public purposes only or also in the wider public interest; whether or not compensation would be set at market value and whether it would be defined by a court; and the status of the right to restitution of property rights (Chaskalson, Reference Chaskalson1995; Klug, Reference Klug2000: 132–33). The NP argued for unconditional protection of all private property, while the liberal and business-aligned Democratic Party (DP) agreed with the ANC that this should be tempered with the right to expropriate property not only for public purposes (such as infrastructure) but also in the public interest, including for land reform, subject to the payment of compensation as determined by a court (Klug, Reference Klug2000: 127).

By February 1993, the ANC had acceded to the payment of ‘just compensation’ for land acquired by the state based on an equitable balance of public and private interests and subject to legal review – and therefore to the principle that the costs of land reform would be largely carried by the state. Unlike its earlier guidelines, the ANC’s proposals for the interim Constitution drew no distinction between personal and corporate property. The debate on property rights at the Convention for a Democratic South Africa (CODESA) focused on the mines and industry, together with land, a conflation of forms of property that again proved convenient in 1995 for the farming lobby, which was less influential in the negotiations than big business (DolnyFootnote ¹ interview, 2005; own observation). As Mandela told businessmen in London, ‘[w]e have issued an investment code which provides there will be no expropriation of property or investments. Foreign investors will be able to repatriate dividends and profits’ (Kimber, 1994, cited in Hall, Reference Hall2010: 153).

As part of its ‘Back to the Land’ campaign in June 1993, the National Land Committee (NLC) supported a protest march of 500 rural community representatives, drawn from all provinces, at the CODESA negotiations, demanding the removal of the property rights from the draft interim Constitution and the confirmation of a right to restitution. This did not work. Agreement on the property clause on 26 October 1993 was the last item on which the interim Constitution turned, in lengthy and late-night debates described as both fierce and chaotic (Chaskalson, Reference Chaskalson1995). The final agreed text set out an explicit right to property – a positive freedom for landowners to acquire it, hold it and dispose of it. No such right exists within the final Constitution, which instead opens with only a negative property right to govern the way in which those holding property may be deprived of it.

In all the contestation between rural community groups and the ANC, the focus was squarely on the right to restitution and securing the tenure of farm workers and labour tenants (Klug, Reference Klug2000: 133). Throughout this period, the constituency for wider land redistribution was not clear. While landlessness was a popular topic at workshops and central to the rhetoric of land non-governmental organisations (NGOs), it lacked a clearly defined constituency. The NGOs worked with communities that had been forcibly removed or threatened with forced removals in the preceding period. From 1993 onwards, as the restitution process was separated from the development of policies for land redistribution, the claims process was taken forward by the Centre for Applied Legal Studies (CALS), the Legal Resources Centre (LRC) and the Land and Agricultural Policy Centre (LAPC)Footnote ² in a technical approach that favoured the input of lawyers. Dolny (interview, 2005) observed that, by the early 1990s and with the start of formal negotiations, ‘this was a legal liberation struggle’. Its corollary, of course, was that the shaping of land redistribution policy and agricultural policy, in particular, became the domain of agricultural economists, largely the ‘verligte’ (enlightened) academics working for state institutions who favoured small-scale farming options but within a deregulated and liberalised economy and were contracted to the World Bank (Hall, Reference Hall2010). This bifurcation between lawyers and economists continued to beset land reform.

The Mandate for Transformation

The ‘final’ Constitution (Act 108 of 1996) mandated land reform, setting out rights for the dispossessed, landless and those with insecure tenure. In relation to each property right – the right to access on an equitable basis; the right to restitution; and the right to secure tenure – the duty-bearer is the state (s. 25(5), (6) and (7)). The state is empowered to expropriate property, including land, in the public interest as well as for public purposes (s. 25(2)), and a special status is given to those expropriations carried out in pursuit of ‘the nation’s commitment to land reform, and to reforms to bring about equitable access to all South Africa’s natural resources’ (s. 25(4)). Expropriation is subject to the payment of ‘just and equitable’ compensation, taking into account five criteria, of which market value is one, the others being the history of its acquisition, past state subsidies in its development, its current use and the purpose of the expropriation (s. 25(3)).

As in the negotiations at CODESA, the question of property rights again became one of the ‘unresolvable lightning rods’ in the Constitutional Assembly (Klug, Reference Klug2000: 134). At stake was whether property – already regulated through statute and common law – should be given constitutional protection in a Bill of Rights and, if so, how the powers of the state to enact land reform and the rights of citizens to claim land rights would be balanced against the rights of existing property owners. When the NP failed to garner support for any form of group rights (such as a veto for whites or other minorities on certain kinds of legislation), it focused on the defence of individual rights, specifically property rights, as the last bastion of white protection against the redistribution of wealth (Klug, Reference Klug2000).

At hearings on property rights in August 1995 held by Theme Sub-Committee 6.3, three options compiled by the Working Group on Property Rights were presented and debated: two draft versions of a property rights clause and the option of not including a property clause in the Constitution (own observation). Among those who opposed the inclusion of a property clause at the hearings were academics and lawyers who, drawing on experiences in Canada and New Zealand, argued that embedding property rights in the Constitution would ‘insulate’ property owners from redistribution efforts and so ‘institutionalise or entrench imbalances and injustices in the distribution of property’ (Constitutional Assembly, 1995: 26–55; own observation). The most vociferous objections came from the Pan Africanist Congress, based on its own alternative policy vision, rather than from the ANC (Van der Walt, Reference Van der Walt and McLean1999: 112). Although the debate was framed as one about land and land reform, the provisions would extend to all types of property. Arguments for the constitutional protection of property rights came not only from farmers’ associations and political parties representing white interests (the NP and the DP), but also big business and the mining houses, some of which had brokered ‘talks about talks’ with the ANC in the 1980s (own observation; Klug, Reference Klug2000). As a postgraduate student writing my first thesis on land reform, I attended sessions of Theme Committee 6 and was struck that the vociferous demands for a clause to insulate private property from expropriation came not only from the farming lobby but from mining companies. The insistence on property rights did not turn exclusively or, perhaps, even primarily on questions of agricultural land.

The compromise proposal to the Constitutional Assembly for a way forward was that the Constitution should provide for expropriation for land reform purposes, alongside support for all land reform measures. Whereas the interim Constitution’s limitation of expropriation to public purposes meant that land reform could proceed only with the cooperation of existing landowners, this was not the case with the final Constitution of 1996. What Lahiff (Reference Lahiff2007) has termed a ‘landowner veto’ has persisted in practice due to the policy choice not to expropriate – leading to deadlocks in negotiations over price. This veto is a core feature of market-based land reform as a policy paradigm. And yet, later, while mineral rights were indeed nationalised under the Minerals and Petroleum Resources Development Act 28 of 2002 and, similarly, water was nationalised under the National Water Act 36 of 1998, land was not.

A Right of Equitable Access to Land

The ANC’s proposed constitutional principles included explicit protection of property rights subject to certain limitations, among them provision for the taking of property in the public interest, according to legal prescriptions and subject to payment of compensation. The notion of a right of ‘equitable access to land’ was the least debated of all the clauses. It originated from a small group of ANC-aligned lawyers who argued in favour of a regime of ‘property rights for the property-less’ to counterbalance:

In this way, even among those who had argued against a property clause, the idea was born of using a constitutional rights framework to impose a positive obligation on the state to provide suitable land and housing for the landless and homeless; it would empower them to press their claims, and shape the behaviour of state officials to facilitate a responsive land reform. The ‘property clause’ would grant limited safeguards to existing property owners while mandating transformed property relations between the landed and the landless and between owners and tenants. Agreement between the political parties was reached at midnight on 18 April 1996, and President Mandela signed the final Constitution into law in December of that year.

A right of equitable access to land is not, of course, an unfettered right to land. But it is a right to have the state demonstrate that continued denial of access to land is the necessary outcome of a fair and reasonable policy and implementation mechanisms that weigh up the competing needs of and interests in land. In other words, the state is answerable to the landless. Like the right to restitution, the right of equitable access should be able to trigger an expropriation when the rights of the landless directly contradict and are impeded by the exercise of the property rights of the propertied. Where this is the case, the imperative of equitable access takes precedence. I argue this on the basis of two clauses, read together: first, what I characterise as the ‘national interest’ clause refers to ‘the nation’s commitment to land reform, and to reforms to bring about equitable access to all South Africa’s natural resources’ (s. 25(4)(a)), and second, what I characterise as the ‘override clause’ confirms that ‘[n]o provision of this section may impede the state from taking legislative and other measures to achieve land, water and related reform, in order to redress the results of past racial discrimination’ (s. 25(8)). This latter clause is the oft-ignored trump card, which – should the other provisions impede the state – explicitly exempts land reform from the constraints of any other clauses.

Parliament’s High Level Panel

The question of equitable access to land, and the role of expropriation in making it possible, was considered in the High Level Panel on the Assessment of Key Legislation and the Acceleration of Fundamental Change (HLP, 2017), established by the speakers’ forum of the National Assembly, National Council of Provinces and provincial legislatures, and chaired by former President Kgalema Motlanthe. In our commissioned report on land redistribution for the HLP, Thembela Kepe and I made the point that there was no framework to trigger expropriations in the public interest to make possible access to land on an equitable basis (Kepe & Hall, Reference Kepe and Hall2016). But few expropriations had been attempted even in cases of restitution, where the constitutionally recognised rights of the dispossessed to specific parcels of land could not be realised due to the effective ‘landowner veto’ that the state’s refusal to expropriate ceded to the current owners of claimed properties.

Three points from the HLP merit emphasis. First, the law is insufficiently developed: the Provision of Land and Assistance Act 126 of 1993 is an insufficient guide to give effect to section 25(5), as it empowers a Minister to acquire and/or allocate land but does not prescribe the rationale or manner in which this must be done. Predating the Constitution and considering redistribution as a matter of ‘largesse’ rather than ‘right’, to use Budlender’s (Reference Budlender1992) terms, it is permissive rather than prescriptive. Second, eligibility, prioritisation and selection are among the parameters left unspecified. To align with the right of access to land on an ‘equitable basis’, legal or policy prescription is needed to indicate:

1. (a) who should benefit;

2. (b) how prioritisation of people should take place; and

3. (c) how rationing of public resources should take place.

Third, the categorisation of applicants needs to be linked to a baseline survey and longitudinal studies to track change over time to show the benefits of land redistribution to people’s livelihoods.

The meaning of section 25(5) has not been interpreted legislatively or judicially. The vast majority of South Africans are eligible for land reform, but very few are actually getting access to land. Based on delivery data to date, and prospects for scaling up, even a very substantially expanded land reform programme would be likely to benefit only a minority. The question, then, of who should be selected as beneficiaries and what they are eligible to get is of central importance. On the one hand, ‘decision-making about who actually gets land through redistribution is opaque’ (HLP, 2017: 212), and on the other, there is growing evidence of elite capture by the wealthy, non-farmers and politically connected (Hall & Kepe, Reference Hall and Kepe2017; SIU, 2018; Mtero et al., Reference Mtero, Gumede and Ramanstima2019).

The legal provisions for making a claim for restitution clarify to whom the state is responsible for the realisation of the right – the claimants – and what manner of state action would constitute adequate realisation of the right – the content of the right. But for land redistribution, the holder of the right is not specified beyond ‘citizens’ and the content of the right of ‘access … on an equitable basis’ is not defined in any way. The HLP, therefore, concluded in favour of a Land Reform Framework Bill (or Redistribution Bill) to spell out the nature of the right of equitable access and to provide the basis for citizens to pursue their claims against the state for access to land as a constitutional right. An ‘indicative draft’ of such a Bill was even appended to the HLP report.Footnote ³ Yet the bulk of the far-reaching implications of the HLP proposals were studiously ignored amid President Zuma’s forced resignation, Ramaphosa’s ascendancy and, just a fortnight into his tenure, the EFF proposed its motion in Parliament to review section 25 of the Constitution, which the ANC supported, amending it with several caveats to placate critics. Four years elapsed without any official process towards further development of this proposal.

The Presidential Advisory Panel on Land Reform and Agriculture

While the Constitutional Review Committee was embroiled in hearings, President Cyril Ramaphosa constituted the PAPLRA. The terms of reference covered largely the same terrain as the HLP yet were broader, including the question of EWC along with agricultural development and farmer support, institutional arrangements and financing. The composition of the Panel was a feat of political engineering, bringing together ten South Africans: five women, five men; seven black, three white; two lawyers; two presidents of national farmer associations; two individual leading farmers (one young black woman and one older white man); two agricultural economists (both men); and two interdisciplinary social scientists (both women).Footnote ⁴ In many respects it must be considered a success in establishing a process that would hold at bay the political demands in an election year and keep some momentum within the presidency, as opposed to the parliamentary process, at a time when the EWC issue served as a proxy for internal factional battles within the ANC.

The formation of the PAPLRA followed a frantic period from March to August 2018, during which a sequence of internal ANC processes attempted to enlist allies from outside the party – including groups traditionally excluded and cast as government critics. A set of related strategies, both explicitly articulated and implied, emerged during this period among a network of land activists, academics and lawyers. The first was to support the call for the state to override the interests of property owners for the benefit of land reform by using its expropriation powers. The second was to affirm the principle that compensation should not be presumed and, indeed, that no compensation is an acceptable principle. The third was to draw attention to the scope already available within the Constitution – and, therefore, to the centrality of the political choices made by the state rather than the Constitution itself being the binding constraint. Fourth was to offer the option of a ‘compensation spectrum’ as a way of operationalising the ‘just and equitable’ requirement, recognising that people affected by expropriations are not, and will not always be, wealthy or privileged – indeed the state regularly and mostly expropriates the property rights of poor people, even if these are not rights of private ownership. Fifth was to use this rare opening of space into party and government processes to foreground other neglected and urgent issues, including tenure rights for farm workers and former labour tenants, communal tenure, restitution claims and communal property associations, and agrarian reform. What also became clear was that the message carrier mattered. The ANC clearly wanted academics and lawyers who would affirm the first two points, not least to quell the reaction in the international media and among investors and to give credence to its position that EWC could be done responsibly and was respectable in global terms, while also wanting authoritative, especially black, voices to undermine the EFF’s position.

From the outset, as a Panel, we conducted a reality check: in contrast with the proclaimed importance of land reform, redistribution had ground almost to a halt, making all proposals moot. Expropriation had been effected in respect of twenty-seven land claims – but not in redistribution. The Office of the Valuer-General (OVG) explained that its adopted practice was to take an average between ‘market value’ and ‘current use value’ as the level of compensation, ignoring the other constitutionally specified criteria.

In our first allocation of work packages in the Panel, we allocated matters of land acquisition, beneficiary selection and land allocation to Advocate Tembeka Ngcukaitobi and myself. Together we wrote an initial outline position which was debated within the Panel, later refined and presented at colloquia held in December 2018 and March 2019, and the gist of it incorporated into the final report (PAPLRA, 2019). We suggested a compensation formula on a sliding scale across four categories that would require definition as part of a spectrum of ‘just and equitable’ compensation: zero compensation, partial compensation, market-related compensation and above-market compensation. We emphasised that the purpose of EWC was not primarily to speed up land reform – indeed, as we conceded, it would likely be slower, at least initially. And while it may have the advantage of limiting the cost of acquiring land, it would probably not be entirely free. In our view, the decision not to compensate for expropriated properties was a matter of principle; the question was, therefore, when this principle should apply – a consideration on which the courts had not had the opportunity to rule. Du Plessis’ contribution to the Panel, which we summarised, clarified that a deprivation of property – via nationalisation or state custodianship – would be distinct from an expropriation and would not attract compensation (PAPLRA, 2019: 73–75). The requirement of compensation was, therefore, only a consideration if EWC were enacted on a case-by-case basis rather than through a systemic change in property regime. Ironically, this meant that despite being the primary proponent, the EFF’s proposals did not require a constitutional amendment.

Our proposal within the Panel was for an expedited, primarily administrative process, with recourse to the courts and with the rapid development of a body of jurisprudence as a guide. The extent of compensation is not the only consideration. Section 25(3) requires not only that the amount of compensation is just and equitable but also that ‘the time and manner of payment must be just and equitable’. In Haffejee,Footnote ⁵ the Constitutional Court held this does not mean that compensation must be determined and paid prior to expropriation. Yet the determination of the amount could occur after expropriation. In Latin America, government bonds offered a mechanism to provide some compensation while deferring the cost (Cliffe, Reference Cliffe2007). The anticipated delays pending court challenges would otherwise stymie reform and incentivise landowners to drag out disputes. Expropriation can, though, proceed separately from the determination of compensation, which in turn would be ‘bifurcated’: initially, compensation should be administratively determined by a state valuer by applying a formula defined in policy and, if the property owner was dissatisfied with the compensation, this could be appealed and a court could review and approve or set aside the compensation – but without stopping the expropriation from proceeding.

The PAPLRA proposed ten circumstances where EWC may be considered: abandoned land; hopelessly indebted land; land held purely for speculative purposes and a clarification of what constitutes ‘speculative purposes’; unutilised land held by state entities; land obtained through criminal activity; land already occupied and used by labour tenants and former labour tenants; informal settlement areas; inner-city buildings with absentee landlords; land donations; and farm equity schemes (where the state has purchased equity and no or limited benefits have been derived by worker shareholders). These were more expansive than the categories in the draft Expropriation Bill and would be subject to review and determination by the courts. The two white commercial farmers on the Panel, AgriSA President Dan Kriek and Nick Serfontein, distanced themselves from all our recommendations on expropriation and compensation, among other matters.

The majority view of the Panel was that the Constitution is ‘compensation-based’ in that its provisions entail a presumption that compensation of some kind must be paid. At the same time, during our deliberations, Parliament voted to amend the Constitution and appointed a committee to develop a proposal to amend section 25. Cognisant of the political imperative of an amendment ‘to make explicit that which is implicit’, and drawing on Du Plessis (in PAPLRA, 2019), we offered a suggestion for a constitutional amendment to clarify that compensation does not need to be paid in each case, and insisted that framework legislation is needed to guide compensation, among other matters. We proposed this addition to the property clause: ‘(c) Parliament must enact legislation determining instances that warrant expropriation without compensation for purposes of land reform envisaged in section 25(8).’

In contrast to outward appearances, EWC was not the most controversial or difficult matter debated within the Panel, or at the consultations we convened to engage more broadly with people. The main sticking points were the purpose of land reform, the nature of agrarian reform and the class agenda of land reform. In our consultations, most contested were the powers and roles of traditional authorities in general in land administration and specifically the status of the Ingonyama Trust. (On this see Mnisi Weeks, Chapter 7, this volume.) Skirmishes also emerged over the rights and entitlements to land of farm workers and dwellers, climate change and the notion of a ‘just transition’ towards a low-carbon future.

Insights from Cases on Redistribution and Compensation

An irony throughout the EWC debate has been the privileging of the question of acquisition over that of redistribution. Acquisition might be a precondition for redistribution, but the state has consistently shown that it is more proficient at the former than the latter. Whether acquiring land through the market or compulsorily, the state has been unwilling and/or unable to redistribute it effectively, in a coherent manner and with secure tenure rights. A range of situations have been documented: redistribution beneficiaries who have been unable to acquire either title or long-term leases to their land, even a decade or more after acquiring permission to occupy (Hall & Kepe, Reference Hall and Kepe2017); the poor capacity of the state to collect rents owed by its tenants on state-owned land (Auditor-General reports); and many others (Mtero et al., Reference Mtero, Gumede and Ramanstima2019). An illustrative example is Rakgase,Footnote ⁶ in which the High Court found that the state had failed to comply with the Constitution by not converting the tenuous land rights (under a long-term lease) of a black farmer, David Rakgase, into ownership when it was able to do so and he qualified for a grant; this, it found, constituted a breach of a constitutional obligation. This failure of administrative justice flags the limitations of state trusteeship in the absence of capacity to redistribute rights to citizens. Trusteeship was patently no guarantee of access to land on an equitable basis, as required in section 25(5).

If completing the work of redistribution involves securing tenure, the initiation of redistribution requires serious engagement with demand for land; between the two sits the question of how to get the land. The PAPLRA’s report urged that the most urgent needs for land be prioritised, to resolve the outstanding land restitution claims, to give the land that the state has and to identify much more effectively privately owned land needed for redistribution. Resolving the chaotic, conflictual and insecure tenure arrangements on redistributed land requires either state capacity to manage leaseholds or an exploration of alternative models. The state’s ongoing attempts to extract rents contradict the proclaimed intention of EWC: under the current leasehold model, beneficiaries are to pay rent to the state for fifty years, before being given an option to purchase the ‘redistributed’ land. So, even if the state were to acquire it without paying compensation, this does not translate into its being given out for free. The ‘market-based’ approach remains in the broader conception of land reform, as being about redistribution for production for the market and payment of a rent. The irony is that some of the political formations that promote state trusteeship have also been fighting against the state for evicting people.

One labour tenant case illustrates both the legal possibility of using the constitutional parameters to drive down compensation for land reform – though whether this could extend to no (or ‘nil’) compensation was not tested. Acting in Msiza at the Land Claims Court,Footnote ⁷ Ngcukaitobi set compensation at a level 16.6 per cent lower than the estimated market value, using the criteria of section 25(3). Mr Msiza and his family had, from 1936, resided on and farmed land in KwaZulu-Natal, but never owned it. Owners had come and gone, taking with them the rising value of the land over the intervening sixty years, and even after Msiza’s claim was lodged, the property had again changed hands. As Ngcukaitobi observed,Footnote ⁸ the appreciation in the value of the land should have accrued to Mr Msiza and his family. Had the state better executed its obligations, they would have been the owners by the time the Trust had acquired the farm, several years after the claim. And further, had the LCC been deploying the constitutional criteria of ‘just and equitable’ compensation as set out in section 25(3) over the past twenty years or so, and determining compensation on this basis in diverse cases, a body of jurisprudence would by now be established.

Overturned in the Supreme Court of Appeal,Footnote ⁹ the time allowed for a further appeal to the Constitutional Court elapsed, and the opportunity to test ‘just and equitable’ compensation in court was lost. Inadequate precedents exist precisely because, by choosing not to use these powers or to test the constitutional requirement of ‘just and equitable’ compensation, the state has over time created a situation in which the meaning of this phrase remains uncertain. When compensation should be set below market price and under what circumstances it could be zero has never been determined by a court, except in Msiza. Deputy Chief Justice Dikgang Moseneke admitted in 2018 that it was a matter of disappointment for him and others on the bench that no cases had been brought to test these provisions.

Beyond the Caricature: Land and Inherited Privilege

Certain contradictions and paradoxes arise, some of which are distinctly uncomfortable. First is the question of the land as a signifier of privilege as opposed to being the repository of privilege. While it is arguably the most potent signifier, ‘land’ is not the main repository of wealth or privilege. Substantial wealth is of course still bound up in land, including but not primarily agricultural land, but much of the wealth built up and accumulated on the land, from conquest, through settler colonialism, to segregation and grand apartheid, has since migrated off the land. The land is no longer the repository of all the wealth that it produced. Yet, unlike Bantu education, cheap labour and the suffering of separated families, the land remains a relic, a physical remnant of what was lost and can conceivably be restored. The equation of landed property now with inherited privilege is both correct but also incomplete, given the extent to which the capital accumulated through cheap land and labour is now held in urban residential property, in financial instruments, on the stock market, in global financial markets and in intergenerational investments in education, rather than in the land.

A second angle is that, while most property owners are white South Africans, the vast majority of the nearly 5 million white South Africans do not own agricultural land – which is the focus of the call for EWC. The majority of the approximately 30,000 commercial farming units are owned by South Africans, overwhelmingly white, yet even by generous calculations, these landowners likely constitute less than 0.6 per cent of the white population. This also raises the question of the beneficiaries of colonialism and apartheid and how those who benefited should contribute to transformation and redress.

Third is a temporal dimension, given the lapse of so much time not only between dispossession and a constitutional democracy but also between a constitutional democracy and the realisation of rights to redistribution, restitution and security of tenure. Between 1995 and 2008, over 5 per cent of agricultural land was transacted through the market annually (Aliber, Reference Aliber2009: 13). And given the significant restructuring in the sector through the 1990s, triggered by deregulation, drought and trade liberalisation, the majority of private property owners holding agricultural land parcels by 2018 had not owned these properties in 1994. As a corollary, the vast majority of white South Africans who had owned farms in the early 1990s no longer did so. The very limited intergenerational transfers of commercial farms, the active land market and rising property prices during the late 1990s and especially the early 2000s, prior to the crisis of 2008, meant that a certain generation of white farmers were able to sell properties acquired and developed with state support, realising the full improved prices for these, and invest the proceeds elsewhere.

In view of this, a striking silence in the debate about the property clause is any discussion of the expropriation of property other than land. Of all the forms of property that underpin inequality and injustice, land has been singled out, although even the category or categories of land that should be subject to EWC have been rather limited in the arguments of many of the protagonists. The EFF, for instance, started by calling for the nationalisation of all land, but later clarified that EWC should apply to farmland only and not to residential land, after an outcry from within its support base. Expropriation of intellectual property or financial assets have been largely undiscussed.

Further, the determination of compensation has to date been highly discriminatory, with an asymmetry between the calculation of compensation to the dispossessed and compensation to the possessors. Land restitution claimants opting for cash compensation have not received the current market value but rather a historical value inflated to current values using a consumer price index. Almost invariably, this discounts the difference between general inflation and property values; it also ignores the forgone opportunities of having been denied property rights for the intervening period. To the extent that compensation is treated differentially, then, it demonstrates a system of affirmative action in favour of the current possessors of property rights over those dispossessed in the past, and in favour of private title holders over those holding informal property rights.

Conclusions

The insistence on a constitutional amendment and blaming the Constitution for the failures of land reform runs far deeper than any literal reading of the law. Instead, it is a political act – in part political theatre, in part restorative dignity-claiming. The call for the decolonisation of landed property relations deserves wider debate, as does what it means to dismantle the hierarchy between property owners and the landless. Missing in the debate so far has been the crucial question of the commons and of defending and expanding access to land as common property, alongside or instead of redistribution within the private ownership or private leasehold model. Attention to the property clause is productive only insofar as it spurs on this broader debate, rather than fixating on the Constitution as the site of resolving the intransigence of the state.

The core problem is not the state’s powers to acquire land – which are well established, if seldom used – but rather the ability of citizens to claim access to it on an equitable basis. Any battle that targets the state’s expropriation powers in isolation from the question of the onward redistribution of the land that the state acquires in this manner is likely to yield a Pyrrhic victory. Rather, the ‘redistribution’ clause in section 25 provides the basis for precisely this: a framework to guide how citizens can make claims on the state and in whose interests the state should expropriate. While a ‘redistribution bill’ and ‘area-based planning’ for ‘inclusive people-driven land reform’ and other models and guides have proliferated, missing among these initiatives are organised formations of landless and land-poor people making their own plans for which land and where should be redistributed to whom, for what and on what terms. A right to land for the landless, on an equitable basis, in both urban and rural areas, and for diverse purposes and with flexible tenure, throws open more useful and more liberatory possibilities than the convergence of neoliberalism and authoritarian populism that we have seen to date; it could serve as a strategic focal point for socio-legal activism, urban and rural social movements, human rights lawyers, NGOs and allies in academia.

The conditions for such struggles to find purchase and for claims to have traction are many and will doubtless continue to be debated and tested in practice. They will stand a better chance if, in the interim, the state is claim-ready with a new Expropriation Act, a revived and capacitated Land Court and a Compensation Policy that operationalises the criteria in section 25(3) of the Constitution. But such measures should not be confused with moving ahead with land reform; they are just yet more mechanisms, illusory until used. While the state steadfastly refuses to instigate significant (let alone pro-poor) land reform ‘from above’, such legal and policy reforms cannot be seen as advancing land reform in themselves. However, they could serve as frameworks against which political movements, communities, families and individuals might at times assert their claims ‘from below’.