On the 18th Amendment to the Constitution, some tentative notes

I thought it worthwhile to jot a few preliminary ideas down about the second draft Eighteenth Amendment to the Constitution. This is a fairly lengthy blog – I have tried to make headings so that it is easy to navigate if you want to skip some of the wanderings. If you just want a clause by clause analysis, click here. If you want to read about custodianship, click here.

I will share our submission again once it is done.

The language of our Constitution

I want to start off with a short discussion on the text of our Constitution. Our Bill of Rights is written in “plain language”, consciously steering away from over legalese and detailed formulations. This is to ensure that the Constitution is flexible and adaptable to endure.

Constitutions are also “final” documents, in the sense that amending them (and especially the Bill of Rights) are long and complex processes. The current amendment is evident of this. So, it is not cast in stone, Constitutions are meant to endure.

Section 25 is no different.

The making of a constitutional amendment

When making a Constitution, politicians say that they “speak for the people”. This is also true for the current amendment.

In one of the meetings, for instance, the chairperson stated that

this Ad Hoc Committee amending Section 25, has no room for winners and losers amongst it. Commenting on failure to “give and take”, it will require all political parties represented in Parliament to put the interest of South Africa and her people first.”

In a previous meeting, a lot of emphases was laid on participation:

(The members) have also agreed that democracy is both representative and participatory, therefore as public representatives they do not have a monopoly on wisdom. The Constitution enjoins the Committee to conduct public hearings, to ensure that the people speak for themselves.”

The idea is that a Constitutional text must capture the popular sentiment. But because constitutional making often involves various parties with various interests, this “speaking for the people” will not be a unitary voice but a bargaining-and-compromise voice.

What typically happens is that politicians state that the particular content is what the people want. Vox populi, whether real or perceived, becomes the justification for the suggested content.

It is then left up to lawyers to draft the specific wording, ensuring that it complies with drafting practices.

The people drafting often also advise what is possible or not – always mindful that it is up to the drafting body (an Ad Hoc committee in this case), to accept the advice.

Community of interpreters

Härble speaks of an “open community of interpreters”, where (translated) “(a)ll organs of State, all public powers and all citizens and groups are potentially involved in the process of constitutional interpretation. There is no fixed number of constitutional interpreters!”

So, let’s look at the options proposed, and how it fits into the bigger legal picture. As always, you might be able to glean my preference through the writing. I am by no means someone who professes that section 25 has only one meaning. Language allows for more than one meaning, and we have seen different interpretations of section 25 throughout this process.

That being said, I disagree with the chairperson of the Ad Hoc committee who, on numerous occasions, declared that “if you put a hundred lawyers in the room you will get hundred opinions”. In S v Zuma, Kentridge stated:

”Nor, I am equally sure, did the learned judge intend to suggest that we should neglect the language of the Constitution. While we must always be conscious of the values underlying the Constitution, it is nonetheless our task to interpret a written instrument. I am well aware of the fallacy of supposing that general language must have a single “objective” meaning. Nor is it easy to avoid the influence of one’s personal intellectual and moral preconceptions. But it cannot be too strongly stressed that the Constitution does not mean whatever we might wish it to mean.”

Indeed. Language allows for different interpretations, but this does not mean there is free reign. It merely encourages discussion and debate about the meaning of provisions. And in this process, we give meaning to the Constitution.

This, in my opinion, has been the benefit of the amendment process, regardless of the outcome.

Specific provisions

With that, let’s have a look at the specific provisions.


In the past couple of decades, preambles have become important in the process of interpreting provisions. Preambles often give a background to or provide why the legislation is promulgated, and as such, can help with its interpretation.

It has previously been pointed out that the first paragraph is problematic. For one, it suggests that the slow pace of land reform is only viewed as problematic by the dispossessed when it is a widely held opinion. This has also been pointed out in many judgments – Rakgase, Mwelase, and District Six Committee judgments. Instead, the Preamble should speak to all people in South Africa, as a successful and orderly land reform programme that brings real change to people is in the interest of all South Africans.

The second paragraph mimics the committee’s mandate: amending section 25 “to make explicit what is implicit…so that nil compensation is explicitly stated as legitimate option for land reform”.

There seems to be little consensus on what is currently “implicit” in section 25, concerning compensation for expropriation. On my interpretation, what is implicit, is that in any expropriation, an amount of compensation that is just and equitable under the circumstances is payable. However, in some cases, the circumstances may indicate that the just and equitable amount of compensation is R0 (or nil), just as, in other cases, it may be market value, below market value, or even above market value.

This position – in other words, that compensation is always payable for expropriation but may under appropriate circumstances be nil – is the point of departure for the current interpretation of section 25 as well as any amendment thereto.

As a point of departure, this position is notably different from “expropriation without compensation”, which in principle means that there is no obligation at all to compensate when property is expropriated. With “expropriation without compensation”, the expropriatee can, in other words, not question the compensation because there is no per se duty to compensate. In our view, this is not implicit in the current section 25.

Any amendment, to make explicit what is implicit, must then be clear that in any expropriation (also expropriation for land reform purposes), there is always a duty to compensate, but that the amount of compensation must be determined on a case-by-case basis, on the standard of justice and equity, and may sometimes be an amount of nil.

This might seem confusing, but in short, nil compensation (R0) is not “without compensation”. “Semantics!”, the owner receiving R0 may say. True, but important semantics. Because R nil/zero still implies an obligation to pay compensation and justify why “just and equitable” compensation, in a particular case, lands on R nil/zero. This is an important accountability mechanism. (There was a discussion about this in the Ad Hoc Committee here. The EFF’s proposal that you can find here states “without compensation”, which is the one point they differ on with the ANC.)

Either way, linking in to the first part of this blog: the language of the Constitution is purposively broad – the details are left for legislation. The Expropriation Bill, when passed, will make “explicit what is implicit” — as legislation should. We should be careful not to legislate through constitutional amendments.

The third paragraph speaks to the amendment addressing the historical wrongs. Paragraph 4 speaks to the amendment “ensuring equitable access to land”. I will deal with it later when I discuss the specific new wording and concepts introduced. Suffice it to say, an amendment alone will not do that. Legislation, policy and implementation by strong, working institutions are essential.

Amending paragraph 25(2)(b)

The only change in this paragraph from the previous version is removing the courts as the only forum that can determine nil compensation. In our last submission, we have supported such a stance. (A copy and paste from that submission).

In practice, this would require that, whenever a state authority contemplates expropriating for nil compensation, the process would have to be halted in order to refer the question whether nil compensation is appropriate to a court for decision. This process, in our view, would be unduly cumbersome and time-consuming. There is no need for the courts to be the only institutions clothed with the authority to make this decision. It should be remembered that the State has the authority to determine that in a particular case the just and equitable amount of compensation is nil, and if dissatisfied, the expropriatee can then challenge that conclusion in the appropriate forum. The expropriator should have the authority to decide administratively that nil compensation is appropriate in a particular case. Should the expropriatee disagree with this, the dispute may always then be taken to court. The courts will, in other words, retain the ultimate authority in case of a dispute to determine whether nil compensation is just and equitable and would not be the initial decision-maker in this respect.

It is crucial that the enabling legislation, the Expropriation Act, have sufficient internal and affordable mechanisms to hold the decision-makers accountable and to ensure that only the disputes that could not be resolved between the parties end up in court. It would be unfair on an expropriatee to have to fork out the costs of litigation where the decision-makers make irrational and unreasonable determinations.


We accordingly recommend that the proposed amended subsection 2(b) be changed in the following way (words in bold and inside brackets should be omitted from the current proposed wording, and words in bold and underlined, we argue should be inserted):

(b) subject to compensation, the amount of which and the time and manner of payment of which have either been agreed to by those affected or decided or approved by a court:

provided that in accordance with subsection (3A) (a court may) where land and any improvements thereon are expropriated for the purposes of land reform, (determine that) the amount of compensation (is) may be nil.

What is meant with “improvements” here is also unclear. In terms of the basic principles of property law, all permanent attachments form part of the land. It is submitted that where land has any improvements (in the sense of permanent attachments), such improvements automatically form part of the land in terms of the principle of attachment (or inaedifiatio, to be more specific).

Amending section 25(3)

Not much has changed here. What is important from a legal perspective is to note what is left in this section, namely, that the compensation standard remains “just and equitable”. What is “just and equitable” will differ from case to case. The problem, and in my opinion where the bulk of the conversation should lie, is how do we calculate “just and equitable” compensation because justice and equity are difficult to quantify.

Adding subsection 3A

This subsection changed from the previous one, in that references to the courts are taken out, and “land reform” was added. This has been done in the Expropriation Bill that is awaiting finalisation in Parliament.

A fair amount of time has been spent in the previous Ad Hoc committee surrounding whether the Constitution should list the circumstances where property can be expropriation at nil compensation or whether this should be left to the legislature to legislate on. Concerns were raised that the amendment to the Constitution requires a 2/3 majority, while the promulgation of legislation only requires 51%.

At this point, it should be reiterated that the Constitution is a framework, and any legislation promulgated by the legislature must comply with the principles and values laid down in the Constitution. The Constitutional standard of “just and equitable” will always be applicable in the calculation of compensation. While a list might serve to signal what property may be subjected to nil compensation, the principle remains flexible and contextual.

Adding subsection 4A

This is where things get, well, interesting. It is unclear where this suggestion comes from, and I cannot remember a discussion about it in the meetings.

“Common heritage”, in international law, is a principle that is often used in the context of natural resources that should be shielded from (commercial) exploitation because of the value that it holds (also for future generations).

The Freedom Charter stated that “The national wealth of our country, the heritage of all South Africans, shall be restored to the people”, but it referred to wealth. A similar statement was not included under land.

The Mineral and Petroleum Resources Development Act that states that “(m)ineral and petroleum resources are the common heritage of all the people of South Africa”, and then, importantly, continues to state that “the State is the custodian thereof for the benefit of all South Africans”. In other words – because minerals are common heritage, it should be under state custodianship – similar to the thinking in international law.

However, the proposed amendment to section 25(5) only refer to “certain land”, and it can perhaps cheekily be asked, if in that case, this section should not also refer to “certain land” as common heritage.

This clause was presumably added to get the EFF on board for the amendment. The EFF suggested that similar wording be added to section 25(4), namely:

”Land is a natural resource and the common heritage, which belongs to the people as a whole, under the custodianship of the democratic State.”

Note, however, that the EFF makes a clear link between “common heritage” and “state custodianship”.

Amending section 25(5)

This is a new proposed amendment. Section 25(5) orders the State to redistribute land. Despite this clear instruction, we currently lack a comprehensive “Redistribution Act”. The policies that guide redistribution are muddled, often inaccessible to the public. Policies, in any event, does not create strong rights and lacks strong accountability mechanisms.

EFF proposal:

”The State must take reasonable legislative and other measures which enables state custodianship and for citizens to gain access to land on an equitable basis.”

The reference to “certain land” signals that it does not envision placing all land under custodianship as the EFF proposes. And this is the heart of the deadlock.

Let’s get to the custodianship issue.

State custodianship was introduced into South African law through the National Water Act 36 of 1998. The concept is not new. This was followed by the National Environmental Management Act (NEMA), the Mineral and Petroleum Resources Act (MPRDA), the Biodiversity Act, the Integrated Coastal Management Act. These are all pieces of legislation that deal with natural resources.

Custodianship of “certain land”, namely agricultural land, has been tried before in the form of the Draft Preservation and Development of Agricultural Land Framework Bill of 2015. Clause 3(1) states that “agricultural land is the common heritage of all the people of South Africa and the Department is the custodian thereof for the benefit of all South Africans”. State custodianship was removed in the final version that is now back in parliament.

But what does this mean? In terms of this construct, the State must act as “trustee” or “custodian” of the particular resource. American scholar Sax explains, with reference to public trusteeship which is similar, that the underlying idea of this construct is that “certain interest are so particularly the gifts of nature’s bounty that they ought to be reserved for the whole of the populace”.

The State is not the owner, as it would be in the case of nationalisation. The State cannot alienate the resources. The State does not benefit from the resources. The holder of a mineral right, for instance, uses the right for their own benefit, even if they are not the owner.

But, the State is strictly bound to regulate the resources in terms of the legislation that bestows this trusteeship, or custodianship, on it. This is typically done through permits or licenses.

This concept is thus relatively novel as a statutory construct in South African law. Still, some scholars make the argument that this is not a completely novel concept in and is comparable to customary law (where water, for instance, has been part of the commons and, in that sense not privately owned) and Roman law (the idea of res nostro patrimonio, as Justinian would have it), land under the Ingonyama Trust, or even land under the South African Development Trust during Apartheid.

So, what does this mean for property rights? It removes certain specific forms of property, as clarified in legislation, from the realm of private property and places it in the arena of public property. But it stops short of nationalisation (although some scholars question this).

Case law, such as the HTF Developers case, stated that the effect of environmental laws is that landowners could not use their land as they wish but had to use the land in such as way so that future generations may also benefit from it.

The AgriSA case, dealing with old order minerals rights, made clear that in terms of the MPRDA, the State does not acquire the minerals. In other words, unlike with nationalisation, the State is not the owner of the resource.

When the State places a resource under its custodianship, compensation is not due as it is not an expropriation (as per AgriSA) since it is a deprivation.

But that does not mean that the State can place all kinds of resources under state custodianship to bypass the duty to pay compensation. These powers are restricted, and the question of when it will be unconstitutional is still determined.

Such an institutional change – in other words changing the resource from private property to public property – will be permissible if, in the words of late André van der Walt, such a scheme is based on Constitutional authority, is rationally legitimate and procedurally just.

Also, if the effect of such a change disproportionately lands on one individual or a small group of people, without providing for some form of compensation (damages or some other form of compensation), it will (at least in the current wording of s 25) be unconstitutional.

What is important to note here is that placing “certain land” under custodianship is a policy decision and one that is already possible within the current wording of the Constitution subject to the provisos listed above. In conversation with Johan Lorenzen he pointed out that the insertion of 4A will probably facilitate such a move, which makes sense to me.

Interestingly, the language of the Constitution does not specify on entrench a specific economic system – this is something that is negotiated politically. This is learned from Albie Sachs in his book:

Perhaps comparatively, Ethiopia serves as an example that I will not go into detail now, but if you are curious, have a look at section 40 of the Constitution of the Federal Democratic Republic of Ethiopia.

The details of what the state has in mind with custodianship is not clear. Also, the vagueness of “certain land” certainly doesn’t help give us direction of what government envisions.

While legally permissible, concerns have been raised on how this will work in practice. Implementing state custodianship successfully might require trust in government to do land reform, and strong institutions to drive it, something that is government needs to work on.

Redistribution only through custodianship?

Another problem with this section is the way it is phrased. Reading the phrase, it seems to limit redistribution only happening through a state custodianship model, which is problematic.

The ANC’s policy is broader than that, however. While they have been slow to hand out title deeds where possible, their policy has always been a mix of ownership, lease, state land and “communal” land. But, the opinion of hon member Xaba (ANC) of what constitute “custodianship” is also questionable.

Again, the Constitution provides enough of a framework for this to be done in legislation.

Will this solve the problem of inflated prices paid for land?

Assume that the amendment still requires that “just and equitable” compensation that may in some instance be R0, the question remains if this will solve the problem. There are many stories of owners asking inflated prices for their land. This is morally reprehensible, but the fact is that most of the land was transferred by contract (negotiations), and not by expropriation. An owner will have less room for manoeuvre in the case of Expropriation. In the case of expropriation, if the price the owner wants differ from the valuer, the State can still expropriate (the owner retains the right to dispute the amount).

My guess is that there will initially be a fair amount of court cases, and the courts will have to develop jurisprudence on what is just and equitable (similar to eviction law). Hopefully, the promulgation of the new Expropriation Bill will facilitate the development of this jurisprudence that is in line with the constitutional standard of “just and equitable”.

In my opinion, the large part of problem can also be solved by distinguishing between “value” and “compensation” and the entities that decide on this. I set out my initial thoughts about it (here), and then supervised a LLM student who expanded on the idea (here). It is done in other jurisdictions.

The argument, in short, is that value is something that valuers to. Compensation requires issues of “justice” and “equity” to come into play. And while valuers can make the decision-makers aware of factors that might influence the compensation amount, it should be left to decision-makers who can exercise their discretion based on certain principles, to determine compensation.

Developing principles to determine “just and equitable” compensation is something that, for instance, courts (and perhaps other entities so authorised) do.

I have, in that respect, grave reservations about the approach in the MalaMala case in the Land Claims Court. In MalaMala the Land Claims Commission suggested acquiring the land for R741million. Anything more, the Commission said, would make the settling of the restitution claim by giving back the land “not feasible”.

The Minister refused this.

The owner’s valuer valued it at R989million.

After hearing all the valuations, the court was prepared to accept that R791million will be the lowest.

The State eventually settled for R1,1billion outside court in a covert settlement agreement. It is therefore unclear how they arrived at that amount.

The claimants would have argued in the Constitutional Court that the compensation for the land should be determined by adjusted the initial land price by the ABSA house index, and that this would be “just and equitable”. There were many factors that would support such an argument. This would be a figure around R450 million. Did the State not settle two weeks before the Constitutional Court hearing (on the worst possible scenario price), we would have had better guidance from the Constitutional Court on how to approach such a matter.

The choice of the State to settle is still unexplained. In a sense, they drank their own poison.

IF there is an amendment, what then?

It is unlikely that the amendment will go through. Both the EFF and the DA indicated that they will not support this amendment. The DA will, in fact, not support any amendment.

In terms of section 74 of the Constitution, if the National Assembly adopts the Bill with a two-third majority, the new formulation forms part of the Constitution. It cannot be tested for its constitutionality, except on procedural grounds. We, therefore, recommend that once the Bill is passed in terms of section 74, the President (in terms of section 79) should send the Bill to the Constitutional Court to decide on its constitutionality before assenting.

Still, I do think the conversation that this process started should be had. It was a catalyst for many other things, one can argue, and for that, it was necessary. But the challenge still lies in grafting and promulgating suitable legislation where needed, amending existing legislation where there are gaps, and ensuring that they are implemented effectively in well-operating institutions.