1 - Politics or Principle?

Summary

This chapter focuses on the so-called politics behind nil compensation against the background of some recent judicial developments, which arguably show a conservative trend in awarding compensation that deviates substantially from market value. The chapter questions why the narrative in favour of nil compensation is so dominant if it is argued that it is already legally possible to expropriate for very little compensation. Considering the Msiza judgments in the Land Claims Court and the Supreme Court of Appeal, the chapter asks: Do we need to rethink the space that nil compensation occupies in our legal framework? The fact that expropriation is essentially compensation-based, coupled with the difficulty that courts have to determine compensation that is not (always) related to market value, suggests that it may be necessary for us to engage more directly with the idea of nil compensation in a much more open, honest and principled manner. There is enough evidence to show that this option is not only politically driven but in fact legally necessary.

Introduction

It has been argued that the idea behind nil compensation for expropriation is essentially political (Dugard, 2019: 137). The political dimension is driven, in part, by a particular narrative that is fundamentally based on the assumption that providing no compensation for expropriation will pave the way for large-scale, rapid and much-needed land reform in South Africa.¹ It is certainly no secret that in the context of land redistribution, as a sub-programme of land reform in South Africa, expropriation has not been used effectively as a tool to ensure more equitable (re)distribution of land. A number of reasons can potentially be advanced for this state of affairs – some of which are not necessarily linked to the compensation question (Hall, 2014: 659). For instance, the policies and laws to ensure land redistribution are not always clear enough to sufficiently ensure the reallocation of property rights in South Africa (Walker, 2009: 472; Kirsten & Sihlobo, 2021; Kotzé & Pienaar, 2021: 295–98; see further DRDLR, 2017). Questions connected to the issues mentioned above relate to the beneficiaries of land redistribution and the type of rights that should be established in terms of the land redistribution programme. In some respects, there is also a lack of political will to ensure that expropriation is a serious option to effect land redistribution (Dugard, 2019: 158). Not all of these alleged reasons for the slow pace of land redistribution are necessarily linked to compensation. However, there are also claims that compensation potentially stands in the way of expropriation for land redistribution purposes. The argument in favour of nil compensation speaks directly to these claims. In this regard, I would like to argue that we should not underestimate a principled approach to nil compensation and the potential it has to unlock the hand of the state to ensure that land reform is speeded up. A more principled approach in either legislation or policy may also be required to provide the necessary guidance to courts on when nil compensation is a serious option – if at all.

Lest I be misunderstood, let me say at the outset that I have, on a number of occasions in the last couple of years, joined in on the argument that it is not legally necessary to amend section 25 to achieve land reform in South Africa because of the numerous possibilities that are locked up within a progressive interpretation of section 25, and on the assumption that the tools and mechanisms that are currently in place, or could potentially be developed, are actually used. So, section 25 itself is not necessarily the problem. In Rakgase (para. 5.4.1),² the court remarked that ‘[s]ince the birth of democracy in our country in 1994, land reform, despite it being a Constitutional imperative, has been slow and frustratingly so’. Consequently, Pienaar warns that ‘if we are to avert systemic failure in the context of land reform, a concerted effort needs to be made to ensure that the programme is “pursued conscientiously and meticulously”’ (Pienaar, 2020: 546).

For land reform to work effectively, we need a legal framework that allows for it, but we also need a capable and proactive state and, very importantly, we need courts that are willing to assume the responsibility of interpreting section 25(3) in such a manner that compensation is not a factor that stands in the way of land reform. However, we are now at a point where various concrete suggestions are, or have been, on the table in terms of expropriation laws in South Africa. For instance, we have the suggestions that were made in the Constitution Eighteenth Amendment Bill 18-2021 (as tabled in August 2021), which sought to provide the authority for nil compensation to be paid in instances where property is expropriated to ensure land reform, although this Bill was rejected by the National Assembly on 7 December 2021. We also have the Draft Expropriation Bill B23-2020, which is still on the table. Given these examples and the problems we see in determining compensation for expropriation (especially by courts), I would like to posit that this is an opportune time to reflect on whether our legal framework should make room for nil compensation in some form and where and how such accommodation should be made.

This chapter aims to focus on the politics behind nil compensation against the background of some recent judicial developments, which arguably show a conservative trend in awarding compensation that deviates substantially from market value. More specifically, I am interested in the following questions: Why is the narrative in favour of nil compensation so dominant if it is argued that it is already legally possible to expropriate for very little compensation? Stated differently, is there a need for greater clarity about the specific instances where nil compensation is a viable option? I think these are important questions as we move forward with the debate around compensation for expropriation. I hope to provide some thoughts on nil compensation for expropriation in light of the Msiza judgments in the Land Claims Court (LCC) and the Supreme Court of Appeal (SCA) (for a critical discussion of the judgment, see Du Plessis, 2019), and I will ask: Do we need to rethink the space that nil compensation occupies in our legal framework?

Are the Calls for Nil Compensation Legally Justified?

Introduction

When considering whether our legal framework should make room for nil compensation, it is valuable to consider the extent to which nil compensation is possible (or not) under the current framing of section 25. It is difficult to conceive of situations where nominal or very little compensation would realistically be possible. The problem is we do not see many examples in the cases that have been presented to courts. In fact, barring some outliers like Du Toit that are clearly not reasoned or argued very well,³ what we do see are courts really struggling to provide compensation below market value and, in fact, moving towards market value, as I will show in the discussion of Msiza LCC.⁴ Moreover, what we also see is the state either not expropriating for land reform purposes or offering exorbitant compensation – even above market value. This makes one wonder whether the call for nil compensation to be provided for explicitly in legislation or in the Constitution of the Republic of South Africa, 1996 (Constitution) is not more legally necessary than we initially anticipated.

Let me illustrate by way of Msiza LCC. The facts of the judgment can briefly be described as follows: Mr Msiza was a labour tenant on a farm situated in the district of Middelburg, Mpumalanga Province (Rondebosch). In 2004, Mr Msiza was awarded a part of the farm under section 16 of the Land Reform (Labour Tenants) Act 3 of 1996 (Labour Tenants Act). In the earlier judgment on the merits of the case, Moloto J found that Mr Msiza was a ‘labour tenant’ for the purposes of the Act and was therefore entitled to a specific portion of the land. The landowners sought compensation from the state for the part of the property expropriated in favour of Mr Msiza, but the parties were unable to agree on an appropriate amount of compensation. Consequently, the LCC had to decide the appropriate amount according to section 16(1)(a) and (b) of the Labour Tenants Act (Msiza LCC, para. 3). The owners wanted market value according to the development potential of the land, which increased from R1,800,000 (if viewed in terms of agricultural use) to R4,300,000 (if the property was valued according to the township that could be developed on the land).

Section 23 of the Labour Tenants Act authorises the court to determine compensation and states that an owner ‘shall be entitled to just and equitable compensation as prescribed by the Constitution’. Therefore, the Labour Tenants Act ensures that compensation is just and equitable as section 25 of the Constitution prescribes. When compensation is determined for purposes of section 23 of the Labour Tenants Act, section 25 should therefore be central to calculating such compensation.

The LCC began by setting out the legal position for assessing and determining just and equitable compensation in terms of the Constitution. The court acknowledged that ‘[t]he award of land to the applicant by this court in its 2004 judgment is an act of expropriation’ (para. 3) and questioned whether the requirements for expropriation were complied with. It disposed relatively quickly of the requirement of the law of general application (para. 11) and proceeded to discuss the public interest/public purpose requirement (paras. 12–15). Having accepted that both these requirements were complied with, the court questioned whether the requirement of just and equitable compensation was met. As was mentioned earlier, the 2004 decision entitled the owner to compensation, but the amount of the compensation was disputed in the LCC (para. 16). The landowners insisted that just and equitable compensation in the particular case was compensation at market value (para. 29). In this regard, the court held that:

I must dispense with this argument at this early stage. Market value is not the basis for the determination of compensation under s 25 of the Constitution where property or land has been acquired by the state in a compulsory fashion. The departure point for the determination of compensation is justice and equity. Market value is simply one of the considerations to be borne in mind when a court assesses just and equitable compensation. It is not correct to submit, as was done on behalf of the landowners, that the jurisprudence of this court installed market value as the pre-eminent consideration.

(para. 29)

Interestingly, the court emphasised further that market value would be used as an entry level for determining compensation because it is the most tangible in the list of factors in section 25(3) (para. 30). Therefore, market value should be used as a starting point in determining just and equitable compensation. A two-step approach would need to be followed. First, market value would have to be determined, after which the court would have to assess whether other factors justified adjusting the market value upwards or downwards. In this regard, the court was at pains to emphasise that the two-step approach did not mean that market value was the standard for determining compensation. Compensation must always be determined according to the standard of justice and equity (para. 30). This is especially true in light of the pre-constitutional position, where market value was the central (most important) consideration in terms of section 12 of the Expropriation Act 63 of 1975. The Constitution drew a line through the primacy of market value by allowing for a number of factors to determine just and equitable compensation. Very importantly, no hierarchy exists in relation to the factors, and a balance must be struck between the landowner and the public interest (para. 32).

In Msiza LCC, several factors justified a downward adjustment of market value because market value would not (according to the court) reflect just and equitable compensation in terms of section 25 of the Constitution. However, the court emphasised the point made earlier in Du Toit that market value is not the single most important element when it comes to determining compensation for purposes of section 25(3). The LCC awarded compensation at R1,500,000, which was R300,000 less than the market value (assessed according to the value of agricultural land at R1,800,000, which the government was willing to pay for the land awarded to Mr Msiza) (para. 82). Although it is not entirely clear how the factors translated into the exact amount of R300,000, the court purportedly arrived at the reduced amount after considering the factors listed in section 25(3) (paras. 48–76).

In the end, the court provided its reasons for awarding compensation below market value. These included: the difference between the amount paid for the whole property and the market value claimed for the portion of land awarded to Mr Msiza; the fact that the landowners had made no investments in the land; the current use of the property had not changed in fifteen years; the landowners purchased the property with full knowledge of the claim made by Mr Msiza; the claim for the portion of the land succeeded in 2004, after which the landowners were precluded from using that portion of the land; the purpose of the expropriation was land reform, and the landowners should not be able to claim extravagant amounts from the state in this regard; the Msiza family had resided and worked on the land and in line with the objects of the Labour Tenants Act the award of the land serves to compensate labour tenants who worked on the land in exchange for the right to reside there (para. 80).

The SCA’s decision in Msiza SCA is an appeal against the LCC judgment as outlined earlier.⁵ The main thrust of the appellants’ appeal was that the LCC had miscalculated the amount of compensation in line with the use of the property as agricultural land instead of its potential future use for development. Moreover, they argued that the amount of compensation had been incorrectly reduced simply because Mr Msiza was a labour tenant (Msiza SCA, para. 1). More specifically, the appellants asserted that the reduction of the amount of compensation for land reform purposes was arbitrary. As this chapter focuses mainly on identifying whether a more principled approach to nil compensation, specifically in legislation, is favourable, the first argument is not of specific interest here. Therefore, the focus will not be on how the court determined whether market value should be assessed in terms of agricultural or residential property, but rather on how courts are navigating the issue of determining compensation at below market value.

The SCA began its analysis by considering the extent of the land and the labour tenancy agreement to contextualise the determination of compensation for the land that was expropriated. Regarding the amount of land expropriated, the court highlighted that the entire property consisted of 352 hectares, of which just under 46 hectares had been awarded to Mr Msiza (para. 2). The labour tenancy agreement in favour of Mr Msiza (and his family) had been concluded in terms of the Native Service Contract Act 24 of 1932, and it was clear that the family had exercised the right since at least 1936.

The court set out the wording of section 23(1) of the Labour Tenants Act to essentially emphasise the link between determining compensation under the Act and ‘just and equitable’ compensation in line with the Constitution (Msiza SCA, paras. 7–8). It identified what should be taken into account in determining ‘just and equitable’ compensation for the purposes of sections 25(2) and 25(3) of the Constitution. Having regard of these constitutional provisions, the SCA considered the judgment in Du Toit, where the Constitutional Court reiterated the general principles relating to the requirement of just and equitable compensation. As a starting point, the Constitution provides the appropriate standard even in cases where legislation – such as the Labour Tenants Act (as in Msiza SCA, paras. 11–12) or the Expropriation Act (as in Du Toit, para. 26) – applies. Therefore, the first step is to consider the list of factors in section 25(3), even if there is direct legislation that regulates the specific type of expropriation in the case, which includes compensation provisions of its own.

Having regard to all the factors listed in section 25(3), the court conceded that market value is usually the one objectively quantifiable factor (Msiza SCA, para. 12; Moloto Community, para. 59).⁶ This reasoning endorses that of the LCC in Msiza LCC and the two-stage approach followed in Du Toit. The court in Du Toit stressed that this approach might not work in all instances, but in most cases it appears to be the most practical. According to the court in Du Toit, this approach can only truly reflect just and equitable compensation if all the factors (where applicable) are accorded equal weight and due consideration (Du Toit, para. 84).

In Msiza, the dispute centred on whether the compensation should be assessed according to the actual use of the property (which was agricultural and valued at R1,800,000) or the development potential of the land (as residential property estimated at R4,000,000). An expert on behalf of the state estimated the current value of the property at R1,800,000 (Msiza SCA, para. 15). Interestingly, according to the ‘Pointe Gourde’ principle, Mr Msiza’s claim for compensation should not be taken into account in determining the market value of the property (Msiza SCA, para. 16). This principle (see Msiza SCA, paras. 18–19 for its origins) applies in the context of determining the amount of compensation for expropriation and is contained in section 12(5)(f) of the Expropriation Act 63 of 1975, which provides that

any enhancement or depreciation, before or after the date of notice, in the value of the property in question which may be due to the purpose for which or in connection with which the property is being expropriated or is to be used, or which is a consequence of any work or act which the state may carry out or perform or already has carried out or performed or intends to carry out or perform in connection with such purpose, shall not be taken into account.

In this respect, the court considered whether ‘a known impediment to the property’s development potential when the property was purchased which ha[s] a direct bearing on the price that a willing buyer in the Trust’s position would have been prepared to pay for the property’ (Msiza SCA, para. 19) should be considered when determining compensation. However, the court relied on the earlier decision in Port Edward v Kay⁷ to conclude that the Pointe Gourde principle does not apply in this case and that the accepted market value of the property should be R1,800,000 (Msiza SCA, para. 20). The ‘Pointe Gourde principle, therefore, does not apply to the present case as the Trust bought the land knowing of the Msiza claim and the presence of the Msiza family on the land’ (Msiza SCA, para. 21; see also Moloto, para. 86). Nonetheless, the question remained whether there were any cogent reasons to reduce the compensation to below market value in this particular case.

The SCA considered the approach adopted in Msiza LCC and the reasons for the LCC deducting R300,000 from the market value of R1,800,000. The court found the reasons for reducing compensation as advanced by the LCC unconvincing (Msiza SCA, para. 20). It held that most of the factors listed by the LCC had, in any event, been accounted for in the determination of the market value of the property (para. 25). There was also no indication that the amount claimed as compensation by the appellants was extravagant or that the state could not pay it. Moreover, the court commented that the R300,000 had been arbitrarily arrived at as there was no indication of its basis, especially since all the factors that the LCC indicated for the deduction were already taken into account in considering market value (para. 25). In the end, the SCA held that R1,800,000 – in other words, market value based on agricultural use of the land without the deduction as indicated by the LCC – constituted just and equitable compensation (para. 28).

Reflection

If one reflects for a moment on the difference between the Land Claims Court’s determination of compensation – where we see some engagement with a reduction of compensation to below market value – and the SCA’s difficulty in accepting this reduction, one is forced to consider the question of when (if at all) an amount below market value (never mind nil compensation) would be a serious option, if not provided for on a more principled basis in legislation.

Even though the court followed the two-step approach in the LCC judgment in Msiza, the principle that compensation for expropriation must be just and equitable – as opposed to market value – seems, at least in theory, to have been seriously considered. The way in which the factors in section 25(3) were considered and applied could therefore be applauded. However, given that the LCC is still focused very heavily on market value in its determination of compensation, it forces one to acknowledge that it will be very difficult to deviate from this standard (Du Plessis, 2015b).

Jeannie van Wyk argues that the two-step approach that focuses on market value and determines the extent to which the amount must be adjusted, as developed in the majority of cases dealing with compensation for expropriation, is not ideal (Van Wyk, 2017: 27). The problem remains the risk of making market value the central consideration, as was the case in the pre-constitutional calculation of compensation for expropriation. That is arguably exactly what happened in the Msiza judgment. Elmien du Plessis, therefore, asserts that in the end, the owner in Msiza received market value compensation as ‘just and equitable’ compensation (Du Plessis, 2019: 217). Therefore, according to Du Plessis, the judgment ‘is a showcase of failed reform with regards [sic] to labour tenants and the state’s inability to transfer the land to the lawful beneficiary due to disagreement about the compensation amount’ (Du Plessis, 2019: 217).

The Msiza judgment shows the difficulty courts have in determining just and equitable compensation for expropriation. The obligation is placed on courts to determine just and equitable compensation in each individual case. The task is exacerbated by the fact that the compensation provisions in the expropriation legislation (the Expropriation Act 63 of 1975) and the compensation provisions in the Constitution are (still) not aligned (Iyer, 2012: 74; Van Wyk, 2017: 25). The calculation of compensation in terms of the 1975 Expropriation Act is of course essentially focused on market value (see section 12 of the Expropriation Act; Van der Walt, 2011: 513). Land reform expropriations add a further dimension to the complicated task of calculating compensation for expropriation (see Du Plessis, 2015a: 369–87; Van Wyk, 2017: 35). To what extent does land reform (alone) justify a (significant) reduction in market value?

The decision in Msiza LCC directly raises the question of determining compensation for a land reform expropriation in terms of section 25(3) of the Constitution. More specifically, the decisions of the LCC and the SCA engage (to some extent) with the question of when we can expect the amount of compensation to be less than market value in terms of section 25. But where does the decision leave South African law in terms of the appropriate determination of compensation for expropriation, specifically expropriations undertaken for land reform purposes, and even a further stretch in terms of opening up debate about the possibility of ever having nil compensation as a serious option in the absence of dedicated legislation aimed at achieving that goal?

Evaluation of the Msiza Judgment: Some Implications for the Determination of Compensation for Expropriation

The calculation of compensation for expropriation as adopted in Msiza LCC seemed sensible and, as stated earlier, could even be commended. The way in which the court engaged with all the relevant factors in section 25(3) is particularly encouraging considering the criticisms often levelled against courts for focusing too much on market value (see specifically Mokgoro J’s comments in Du Toit, para. 36). In this regard, the case is a reminder that the Constitution has allowed for the determination of compensation for expropriation on the basis of just and equitable compensation instead of compensation based on market value. According to Du Plessis, this standard of justice and equity should have a direct bearing on the transformative impact of the expropriation clause in terms of land reform (Du Plessis, 2009: 267).

Du Plessis maintains that courts must be aware of what they are protecting in the process of awarding compensation. Compensation may therefore be a way of ensuring redistributive justice. This will create the possibility of moving away from what Du Plessis calls ‘market value centred’ and ‘scientific’ ways of determining compensation, based on a particular legal culture, towards the calculation of compensation for expropriation that is based on a transformative, constitutional legal culture within expropriation law. She introduces the idea of a ‘transformative interpretation of the compensation requirement in the post-apartheid context’ (Du Plessis, 2009: 271) and concludes that there are various considerations that the just and equitable requirement in relation to compensation requires in the new constitutional dispensation (Du Plessis, 2009: 299–300).

The just and equitable requirement may necessitate an inquiry that a narrow market-driven determination of compensation would disregard. Furthermore, determining the amount of compensation requires a contextualised judgement, which should be sensitive to the facts in the particular case and determining compensation cannot be an abstract analysis (Van der Walt, 2011: 509). This should include consideration of the factors listed in section 25, but courts are not limited to considering only those factors. Courts should, however, give special attention to land reform aspirations (Van der Walt, 2011: 509).

The SCA decision in Msiza highlights that courts essentially still follow a predominantly ‘market value centred’ approach when determining compensation for expropriation and find it difficult to deviate from that standard. Stated differently, when considering the factors (other than market value) in section 25, courts struggle to find adequate justification for reducing the value and almost instinctively revert to market value. This conclusion is especially interesting considering the debates around nil compensation. If the practice is to award market value, even in land reform expropriations, it becomes difficult to accept the theoretical arguments asserting that compensation below market value – never mind nil compensation – is possible within the current constitutional or legislative framework.

A pertinent question arising from the SCA Msiza judgment is: Is there a missing link between the rhetoric that expropriation below market value is possible and the actual practice playing itself out in courts? More specifically, the SCA decision calls into question the theoretical argument that compensation below market value is ever possible. Msiza LCC certainly purports to be a different approach to the one which singles out market value as the determining factor, especially since the LCC ordered compensation at below market value. The fact that the SCA overturned this decision raises serious doubts regarding the contention that compensation below market value is a serious possibility. The centrality of market value is nothing new and has, of course, been the focus of the courts for decades, before and after the property clause came into effect. Several judgments, even in the constitutional dispensation, have highlighted market value as the starting point in the calculation of compensation for expropriation, making it very difficult to deviate substantially from this standard. In Ash v Department of Land Affairs⁸ (paras. 34–35), the LCC formulated a two-step approach when calculating compensation. The court indicated that it would determine the market value of the property and thereafter subtract from or add to the amount of the market value, as other relevant circumstances may require (paras. 34–35). A similar approach was, of course, adopted in Du Toit (para. 37) – as highlighted earlier. This approach is arguably understandable since the general tendency of courts has been to compensate those expropriated by placing them in the same position they were in but for the expropriation (Du Plessis, 2015b: 1728). This is in line with section 12(1) of the Expropriation Act, which indicates that compensation is determined according to what the property could have been realised in an open market if sold by a willing seller and purchased by a willing buyer. There are several judgments that highlight this point.⁹ Recently, the LCC in Moloto had to decide whether the formula of adding the market value to the current use value and dividing the consolidated value by two would constitute just and equitable compensation for purposes of section 25. The court mentioned that this formula is, in fact, similar to the two-stage approach ordinarily adopted by the courts (paras. 20, 63). Although this is a somewhat different approach, market value still plays a key role in that it is used as the starting point from which compensation for expropriation is determined. In the end, the court in Moloto concluded that ‘[i]n the absence of any other information and satisfactory evidence upon which just and equitable compensation can be assessed, this court is constrained to conclude that market value is, in the circumstances of this case, just and equitable compensation as the landowners contend’ (Moloto, para. 96).

Although recent judgments including Msiza and Moloto tried to indicate that market value should not be the primary focus when it comes to compensation for expropriation, the judgments fail to provide clarity on the question of whether expropriation below market value can be justified in the land reform context, and if so, how such an adjustment from market value should be made within the current legal framework. The cases prove that it is difficult to justify why a reduction in market value is possible, even though it is often argued that the law allows for such a possibility in theory. Ernst Marais made the same argument, submitting that Msiza SCA appears to suggest that a downward adjustment of compensation at market value, purely on the basis of land reform, is impermissible (Marias, 2018).

The difference between the approach to the reduction of compensation in the LCC and the SCA in Msiza indirectly invites a conversation about whether land reform alone is sufficient justification for a significant reduction of market value (even a nominal amount of compensation). Interestingly, in this respect, in Du Toit, a so-called non-land reform case, the Constitutional Court was willing to recognise a significant reduction in the market value of the gravel because it held that the public interest in the building of roads was important for the economy and the improvement of the road system in general (Du Toit, para. 51). This case is clearly an outlier, and the interpretation of the purpose of the expropriation in relation to the determining compensation has been criticised. Van der Walt, for instance, argues that the interpretation of this factor in the calculation of compensation in Du Toit is unconvincing from a practical and economic perspective (Van der Walt, 2011: 514). He goes even further to argue that expropriation for land reform purposes without compensation will, in most instances, be unconstitutional. This is because all the factors have to be considered, and ‘land reform should therefore not on its own imply that compensation is not required’ (Van der Walt, 2011: 518). These arguments made by leading scholars on expropriation law and compensation for expropriation have huge implications for the assertion that expropriation at nil compensation is already possible under the current legal framework. In fact, it negates entirely any possibility that awarding nil compensation for expropriation is possible within the current framing of section 25 or the current Expropriation Act. This is not because it is theoretically impossible, but perhaps because it is difficult to conceive of examples where this would be possible.

As indicated earlier, the question remains: Can the purpose of the expropriation (alone) justify a (significant) reduction in compensation? Du Plessis asserts that courts dealing with this factor in the determination of compensation tend to confuse the requirement of public purpose/public interest and public purpose as a factor in calculating compensation for expropriation (Du Plessis, 2015a: 369–87). She uses the examples of Du Toit and Mhlanganisweni Community to argue that the interpretation of public purpose when determining compensation for expropriation is misconstrued in both cases. In Du Toit, the court’s reasoning is problematic because it would mean that in all cases where the expropriatee has property necessary for the upkeep of national resources (or assets), he can expect compensation that is below market value (even significantly so). The decision in Mhlanganisweni Community is disconcerting because it would mean that where property is expropriated for land reform purposes, it should be treated the same as non-land reform expropriations, with the potential that the state may have to pay full market value for those properties in all instances. She considers both interpretations unfair and confusing – Du Toit because one individual is unduly burdened with the task of paying for the upholding and maintenance of a national asset that should be borne by the general tax-paying public, and Mhlanganisweni Community because ‘in view of the history of the privileged land ownership in South Africa and the constitutional imperative to transform, one should acknowledge that market value cannot be treated as a strict requirement’ (Du Plessis, 2015a: 379).

An alternative approach to the role of public purpose as a factor in determining compensation may be to distinguish ‘run-of-the-mill’ or ‘business-as-usual’ expropriations and land reform expropriations (Du Plessis, 2015a: 380). In non-land reform expropriations, the payment of market value may reflect just and equitable compensation as market value may strike the most appropriate balance between the interests of the public and the landowner affected by the expropriation. This is if there are no other factors that nonetheless justify a downward adjustment of market value in these instances. In land reform expropriations, where there may be other considerations at play, and the protection of existing property rights must be assessed in light of the promotion of social justice and transformation, a different interpretation of public purpose may be required when calculating just and equitable compensation. Reconciling the opposing claims in a just and equitable manner may require a more contextual, balancing approach that is sensitive to the task of promoting the spirit, purpose and object of the Bill of Rights (Du Plessis, 2015a: 387). A downward adjustment may be more appropriate in the latter case than the former – as is, in fact, illustrated by the LCC in Msiza. This may be one approach to determining when compensation below market value would be justified. Another (or perhaps supplementary) approach would be to provide guidance in legislation on more specific instances where compensation below market value is plausible. Either way, what is clear is that courts need some more direction in this regard; otherwise, we may continue to see a natural inclination towards market value compensation.

The Suggested Way Forward: A New Expropriation Bill?

Du Plessis points out that it is time for the legislature to deal with compensation for expropriation in a pertinent manner. She notes that ‘[t]he legislature can do this by making sure it provides clear guidelines on the calculation of just and equitable compensation, rather than a mere “copy and paste” of Section 25(3)’ (Du Plessis, 2015a: 387; and see Du Plessis, 2014). I would agree with Du Plessis and take it a step further. A new Expropriation Bill could potentially provide greater clarity regarding compensation for expropriation in the land reform context. It could do so by providing more indication of how the different factors relate to one another, especially if the Bill is to provide further guidance to courts regarding the relative importance of the factors listed in section 25(3) when it comes to calculating the amount of compensation in land reform expropriations. I think the Bill could even do more than that. It could guide the courts in establishing when nil compensation should be a viable option.

There are relatively few instances in which I envisage that nil compensation can be awarded, especially given the tendency of courts to compensate individuals for their loss experienced as a result of the expropriation, as highlighted in the chapter thus far. However, we do see some examples. Clause 12 of the latest Expropriation Bill aims at replacing section 12 of the 1975 Expropriation Act. The Bill is not perfect, but it does lay down the principles that must be adhered to when determining compensation, and in this respect it is certainly more aligned with the Constitution than the existing legislation. The Bill makes it clear that the compensation standard is just and equitable and not market value, thereby bringing it in line with the Constitution to a much greater extent than the current Expropriation Act does. Of particular interest for the purposes of this chapter are the examples listed in clause 12(3), which indicate the instances where nil compensation is plausible. It is important to note that there is still some discretion in terms of the Bill to determine when it may be just and equitable for nil compensation to be paid. Therefore, clause 12(3) is peremptory but not exhaustive. This provision leaves the discretion to the expropriating authority to determine whether the compensation will be nil. Since the expropriating authority is left with a discretion, I would argue that it may be even more helpful to have guidelines on how such a discretion must be exercised.

Clause 12(3) of the Expropriation Bill is relevant when thinking about instances where nil compensation may be applicable and reads as follows:

1. (3) It may be just and equitable for nil compensation to be paid where land is expropriated in the public interest, having regard to all relevant circumstances, including but not limited to –

   1. (a) where the land is not being used and the owner’s main purpose is not to develop the land or use it to generate income, but to benefit from appreciation of its market value;

   2. (b) where an organ of state holds land that it is not using for its core functions and is not reasonably likely to require the land for its future activities in that regard, and the organ of state acquired the land for no consideration;

   3. (c) notwithstanding registration of ownership in terms of the Deeds Registries Act, 1937 (Act No. 47 of 1937), where an owner has abandoned the land by failing to exercise control over it;

   4. (d) where the market value of the land is equivalent to, or less than, the present value of direct state investment or subsidy in the acquisition and beneficial capital improvement of the land; and

   5. (e) when the nature or condition of the property poses a health, safety or physical risk to persons or other property.

These instances are not without criticism but may be the starting point when considering possibilities where nil compensation is envisaged.¹⁰ Clause 12(4) is particularly interesting if one considers some of the issues I mentioned in relation to the Msiza judgment. Clause 12(4) states that when a court or arbitrator determines the amount of compensation in section 23 of the Labour Tenants Act, it may be just and equitable for nil compensation to be paid, having regard to all relevant circumstances. A number of questions arise: First, this section brings claims of labour tenants under the purview or possibility of nil compensation. However, given the difficulty portrayed by courts in even reducing market value, never mind ordering nominal or nil compensation, it is not clear exactly how this provision is going to take us further in terms of assisting courts to deviate from market value. Second, questions may arise about whether a principled or default approach in favour of nil compensation in the context of labour tenants is even the best example or category. In this regard, it is not evident why this group of claimants (namely labour tenants) are included when other groups of claimants, such as restitution claimants, are specifically not included.

These questions, together with those one can equally raise about some of the other categories listed in clause 12(3), are not irrelevant, but they arise only when we are willing to acknowledge that it is necessary to have the conversation about nil compensation in the first place. The point that I would therefore like to make is this: If we are willing to open up a conversation about instances where nil compensation is a possible or valid option, we need to potentially think about the following:

1. (i) Why do we need to recognise a principled approach to nil compensation?

2. (ii) How will we demarcate instances or provide categories suited for nil compensation on a more principled basis?

3. (iii) Should we leave an open-ended discretion, or formulate guidelines that are more specific, like all the instances that are currently listed in clause 12(3)(a)–(e) of the suggested Bill?

This chapter has highlighted at least one reason why it may be important for us to have a conversation about instances where nil compensation should be a more principled possibility. First, expropriation assumes compensation. Arguably, whenever you are in the realm of expropriation, there is an assumption of the obligation to pay (an amount of) compensation. The obligation to pay compensation, which ordinarily goes hand in hand with expropriation, is also why there were conceptual difficulties with introducing notions like custodianship (as distinguished from trusteeship or nationalisation) within the realm of expropriation law in 2021. While there is authority to concede that, on the one hand, compensation is not a prerequisite for expropriation in the technical sense of what comes first, and in a legal sense of recognising that expropriation has occurred even though compensation has not been determined or paid, we cannot get away from the fact that compensation is an integral part of expropriation. In the absence of any obligation to pay compensation, one would arguably not be talking about an expropriation but another form of limitation/interference with property rights. We see, for instance, in the Final Report of the Presidential Advisory Panel on Land Reform and Agriculture (PAPLRA) that ‘[t]he words “subject to compensation” and the presence of the word “amount” denote that compensation is indivisible from expropriation’ (PAPLRA, 2019: 71). Compensation can therefore be a stumbling block to the full enjoyment of the benefits of expropriation, especially in the land reform context. We see this unfold in the Msiza judgment.

The Final Report of the Advisory Panel went on to mention that section 25 is a compensation-based clause and that it is ‘highly unlikely and improbable that there could be a plethora of circumstances that would lead to nil compensation’ (PAPLRA, 2019: 72). The presence of a clause dedicated to nil compensation would therefore provide clarity on instances where despite the obligation to pay compensation for expropriation, there may be instances of nil rand compensation. Those instances can then be justified and demarcated more clearly, and we should stop trying to insist that it is already theoretically possible when legally it is unlikely. At this stage of the developments in this area of the law, it is no longer controversial. I think that is one of the reasons we have seen various permutations of nil compensation in a number of Bills over the last couple of years (including, for instance, the Bills aimed at amending expropriation legislation and, of course, the various Bills aimed at amending section 25 of the Constitution), all of which contain varied provisions with possibilities for nil compensation.

The fact that expropriation is essentially compensation-based, coupled with the difficulty that courts have in determining compensation that is not (always) related to market value, suggests that it may be necessary for us to engage more directly with the idea of nil compensation in a much more open, honest and principled manner. I think there is enough evidence to show that this option is not only politically driven but, in fact, legally necessary.