A Human Right to Water? The South African Constitutional Court’s Decision in the Mazibuko Case

Written by

Peter Danchin is Associate Professor of Law at The University of Maryland Law School.  His recent articles have been published in the Journal of Law and Religion, the Yale Journal of International Law, and the Harvard International Law Journal. His most recent book United Nations Reform and the New Collective Security (with Horst Fischer) has just been published by Cambridge University Press.  In 1999, he served as a foreign law clerk to Chief Justice Arthur Chaskalson of the Constitutional Court of South Africa.

Is there a human right to sufficient or adequate water?  If so, what is the right’s normative basis, its scope and content, and how might this differ in international law, constitutional law, and the domestic law and policy of States?  These were the questions recently before South Africa’s Constitutional Court in Mazibuko v. City of Johannesburg (also known as “the Phiri case”) decided on 8 October 2009, the country’s first test case on the right to water.

The case is of interest for a number of reasons.  First, it is the most recent precedent in South Africa’s closely watched economic and social rights jurisprudence following in the wake of such decisions as Soobramoney (1998), Grootboom (2001), Treatment Action Campaign (TAC) (2002), Modderklip (2005) and Olivia Road (2008).  Second, it sheds critical light on the debate over whether economic and social rights have minimum legal content or a “minimum core” as posited by the UN Committee on Economic, Social and Cultural Rights in its 1990 General Comment No. 3 on the Nature of States Parties’ Obligations.  And third, it provides a useful case study of both the potential and limits of strategic public interest litigation and the justiciability and enforcement of economic and social rights in the national sphere.


The applicants were five poor residents of Phiri in Soweto, one of the poorest urban areas south west of Johannesburg developed as a black township during the apartheid era.   The respondents were the City of Johannesburg, Johannesburg Water (a company wholly owned by the City), and the national Minister for Water Affairs and Forestry.  Appearing as amicus curiae was the international NGO, Centre on Housing Rights and Evictions (“COHRE”).  Also part of the applicants’ legal team was the Centre for Applied Legal Studies (“CALS”) at the University of Witwatersrand along with other human rights organizations.  Finally, it is important to note that the case arose initially out of the organizing efforts of the Coalition Against Water Privatization (CAWP), a South African social movement which worked in partnership with both CALS and COHRE such that the litigation effort was pursued in tandem with social mobilization in order to build a critical base of support and advocacy for realization of the right to sufficient water.

For many years water had been piped to Phiri via an unmetered and unlimited supply for which a flat fee of R68.40 (about US$9) per month was charged on the basis of a “deemed” monthly consumption of 20 kilolitres of water per household.  In actual fact, however, monthly consumption of water per household in Soweto was 67 kilolitres per month.  Johannesburg Water estimated that up to a third of all water it purchased was supplied to Soweto, but that only one percent of its revenue was generated from there.  Part of the reason for this was water losses resulting from leaking pipes, but the main reason was that many residents of Soweto did not pay the deemed consumption charges.  This resulted in three quarters of all water supplied to Soweto being unaccounted for.

A bit of background may be helfpful here for those less familiar with the history of the struggle against apartheid.   Part of the anti-apartheid campaign in the townships included boycotts and the residents in places like Soweto became accustomed to receiving electricity and water without paying for it.  Thus, the post-apartheid government faced the problem of how to restore municipal “order” to the townships while at the same time complying with the Bill of Rights.  This dilemma may help to explain the rationale for the policy ultimately adopted by the City of a certain amount of free water coupled with a subsidized tariff for water provision.

Thus, in May 2003 Johannesburg Water initiated “Operation Gcin’amanzi” (to save water) to revamp Soweto’s distribution of water services.  Its goals were to reduce water losses and demand, rehabilitate the water network, and improve the rate of payment.  Phiri was selected in February 2004 as the area for first implementation of the project.  Central to the City’s plan was the abandonment of the flat rate policy of deemed consumption charges and the installation of pre-paid meters which would provide each household with 6 kilolitres of free water per month (or 25 litres per person per day) with the requirement that any more than that be paid in advance (“the free basic water policy.”)

The applicants challenged the City’s new plan on two grounds: first, on the basis that the free basic water policy violated section 27(1) of the Constitution which provides that “Everyone has the right to have access to … (b) sufficient food and water”; and second, on the basis that installation of pre-paid water meters in Phiri was unlawful under inter alia the Water Services Act and the City’s Water Services By-laws, was administratively unfair, and was unfairly discriminatory under section 9 of the Constitution.

While it is not possible here to set out the details of the applicants’ arguments, two examples in particular are helpful to get a sense of what was at stake in the case.  First, two yard-tenants of one of the applicants had died in 2005 while attempts were made to put out a shack fire and their pre-paid water meter’s supply automatically disconnected.  Second, while pre-paid meters had been installed in Soweto they had not been applied to other bad debtors (as acknowledged by the City) such as businesses and government institutions which received water on credit through conventional meters.  Examples such as these raised issues of administrative unfairness and discrimination against the poorest members of society for whom this method of credit-control would have the harshest impact.

Lower Court Decisions

The case was heard at first instance in December 2007 in the Johannesburg High Court in which Judge Tsoka held that (1) the free basic water policy was unconstitutional and the basic minimum should be increased to 50 litres per person per day to comply with section 27(1) of the Constitution; and (2) the pre-paid meters had no basis in law and had been implemented in a procedurally unfair manner.

The City appealed to the Supreme Court of Appeal which on 25 March 2009 upheld the decision of the High Court in substance, but on different grounds.  The Court found that (1) the free basic water policy was in violation of section 27(1) but on the basis of expert evidence the minimum amount of water required by the Constitution was 42 litres per person per day; and (2) while installation of pre-paid meters was unlawful, the declaration of unlawfulness should be suspended for two years to give the City time to bring its water policy in line with the reasonableness requirement of the Constitution.

The applicants then appealed to the Constitutional Court seeking to reinstate the order of the High Court granting 50 litres per person per day and to set aside the suspension of the declaration of invalidity.

Decision of the Constitutional Court

Writing for a unanimous Court, Justice O’Regan held that the applicants’ two grounds of appeal should be dismissed.  Contrary to the findings of both the High Court and Supreme Court of Appeal, the Court found (1) the City’s free basic water policy of 25 litres per person per day to be reasonable under section 27(1) of the Constitution, and (2) the introduction of pre-paid water meters to be lawful, procedurally fair, and not unfairly discriminatory.

In the discussion that follows, I consider these two findings in reverse order focusing mainly on the notion of reasonableness in the Court’s economic and social rights jurisprudence.

1.  Lawfulness of Pre-Paid Meters

In answer to the argument that the pre-paid meters were unfairly discriminatory because the residents of Phiri were not given an option whether they wished to have their water on credit, the Court noted that not only were residents who received water on credit not given an option of pre-paid meters but they were charged more and thus subsidized the supply of water to those on the pre-paid system.  Further, the pre-paid meters were aimed at the legitimate government purpose of reducing severe water losses in Soweto and their introduction and the differentiation the policy drew between categories of people was therefore not irrational for the purposes of section 9 of the Constitution.  And finally, the installation of pre-paid meters did not constitute unfair discrimination as they were installed for a legitimate purpose and were clearly tailored to that purpose (e.g. pre-paid meters had not been introduced in other former apartheid “deemed consumption” areas where the unaccounted for water problem was less acute).  On the issue of emergencies, the Court accepted the City’s argument that household water pressure was insufficient to put out serious household fires and that the risk was being addressed by installation of fire hydrants throughout Soweto and by a household annual emergency allocation which households could apply for under the City’s indigent registration policy.

2.  “Minimum Core” versus Reasonableness

But what is of most interest about Mazibuko are the differences in logic and reasoning between the minimum core and reasonableness approaches.  The two lower courts, apparently borrowing from international human rights jurisprudence, determined a minimum core content of the right in section 27(1)(b) by actually quantifying the amount of water deemed sufficient for a dignified life.  In setting 50 litres per person per day as the minimum to meet basic hygiene and consumption requirements, Judge Tsoka followed COHRE’s amicus brief and adopted for the purposes of section 27 WHO Guidelines on “Domestic Water Quantity, Service Level and Health” and the expert opinion of Dr. Peter Gleick, president of the Pacific Institute for Studies in Development, Environment and Security in Oakland, California.   In setting the constitutional minimum instead at 42 litres per person per day, the Supreme Court of Appeal preferred to follow the expert testimony of a local civil engineer who manages a consultancy company with expertise in the development and transformation of water supply and sanitation in South Africa.

This approach of quantifying a minimum amount of water for the purposes of the right in section 27 stood in tension with the interpretive approach of the Constitutional Court which in both Grootboom and TAC had rejected the notion that economic and social rights contained a minimum core which the state is obliged to provide, the content of which is to be determined by the courts.  Rather, the Court had preferred an interpretation imposing a duty of progressive realization through reasonable legislative and other measures.  Rather than fixing a quantified content, the concept of reasonableness was said to place context at the center of constitutional inquiry.  This was said better to reflect the institutional role of the Court in holding government accountable while leaving to the legislative and executive branches the primary policy-making functions of assessing, calibrating and realizing social and economic rights.  As stated by O’Regan J in Mazibuko [61]:

This is a matter, in the first place, for the legislature and executive, the institutions of government best placed to investigate social conditions in the light of available budgets …. Indeed, it is desirable as a matter of democratic accountability that they should do so for it is their programmes and promises that are subjected to democratic popular choice.

Thus, under the reasonableness approach if the government takes no steps to realize economic and social rights, the courts will require action.  If the government’s adopted measures are held to be unreasonable—especially for those most desperately in need as in Grootboom—the courts will review those measures so as to meet the constitutional standard of reasonableness.  And if the government policy has unreasonable limitations or exclusions as in TAC, the courts may order those removed.  Finally, the obligation of progressive realization imposes a duty on the government continually to review its policies. [67]

In applying this test in Mazibuko, O’Regan J rejected the five contentions advanced by the applicants in holding the City’s free basic water policy to be reasonable.  First, in answer to the claim that 6 kilolitres of free water was allocated to rich and poor alike, the Court noted that wealthier consumers who were provided water on credit were in fact charged at higher rates thus cross-subsidizing both the free allocation and those who use less water.  Second, in answer to the claim that the amount was allocated per stand rather than per person, the Court accepted the City’s argument that it was impossible to base its policy on a per person allocation given the continual movement of people and variation in numbers between households at any given time.  Third, in answer to the claim that the City’s policy was based on the misconception that it was not bound to provide a specified amount of free water to citizens per month, the Court reaffirmed that the City “is not under a constitutional obligation to provide any particular amount of free water” and is rather “under a duty to take reasonable measures progressively to realize the achievement of the right.”  Fourth, in answer to the claim that the amount was insufficient for large households, the Court noted that for average size households (3.2 people) the policy would provide approximately 60 litres per person per day (i.e. in excess of the amount claimed by the applicants) and to increase the amount for households with many residents would be expensive and inequitable as it would disproportionately benefit households with fewer residents.  And finally, in answer to the claim that the 6 kilolitre amount was inflexible, the Court found that the City was continually reviewing its policy and investigating ways to ensure that the poorest households were able to gain access to water (e.g. by additional allocations for indigent households).  Indeed, much of this engagement had been spurred by the litigation in the case itself.

The sequence of judgments in the Mazibuko case lends credence to Katherine Young’s argument in her recent article “The Minimum Core of Economic and Social Rights: A Concept in Search of Content,” 33 Yale J. Int’l. L. 113 (2008) that “the minimum core concept will always elude attempts at definition along essentialist, positivist, or even institutionalist lines.”  Jackie Dugard, a Senior Researcher at CALS, by contrast has argued that certain ESC rights such as the rights to water and housing peculiarly lend themselves to a minimum core approach.  But as evidenced by the disagreement between the High Court and Supreme Court of Appeal over the minimum amount of water needed to maintain life and basic hygiene, even a “survival” or “basic needs” approach runs into the problem of arbitrariness.

The notion of reasonableness is said to focus instead on a sociologically-framed and value-based analysis emphasizing the values of human dignity, equality and freedom—the values underlying the South African constitution itself.  In my view, implicit in this approach must be some notion of  a minimum core exerting normative pull towards a standard by which the Court is able to determine whether a particular government policy or action falls below or outside a constitutional right.  But the key to the concept’s power and durability is precisely its flexibility and openness, its unsuitability to the technical tasks of quantification and specification.

However, this in effect may simply trade arbitrariness for indeterminacy—what Alfred Cockrell early in the Court’s tenure unkindly termed “rainbow jurisprudence” (see 12 SAJHR 1 1996).  Does it offend human dignity, equality or freedom, for example, to charge some residents more than others for water, or to require indigent residents to register on a means-tested basis?  How are courts to make such determinations in an objective fashion?  And as Frank Michelman has observed, redistributive policies which impact human dignity in South Africa are likely to differ greatly from what is considered “reasonable” in the constitutional culture of societies such as the United States.

If neither survival nor human dignity can provide an uncontested minimum core, what (if any) other approaches may be available?  A consensus approach to a minimum core of economic and social rights norms is unlikely to be any more successful.  Whose or which consensus is to count as normative?  The ESC rights committee?  The Limburg principles or Maastricht guidelines?  The expert opinion of Peter Gleick or I H Palmer (the civil engineer relied on by the Supreme Court of Appeal)?  What about a more institutional or procedural approach which focuses less on norms and consensus and more on obligations, strategies for rights protection (such as justiciability) and potential remedies?  As Young has argued again, however, this too is unlikely to be effective as the “insurmountable problem for the notion of core obligations is that the particular forms of duties are intrinsically polycentric and cannot be subject to a definitive ranking”.

What then is the way forward in developing a coherent approach to the normative basis and substantive content of economic and social rights?  This is far from obvious and serious dialogue and deliberation on these questions has only just begun.  What seems important to consider is what is at stake in advocating and contesting the minimum core content of a right such as the right to sufficient water.  In the earlier cases of Grootboom and TAC and again in Mazibuko, the amicus curiae based their submissions on the notion of an internationally-recognized minimum core in an effort to focus the judicial inquiry less on progressive realization and more on the state’s negative obligation to respect rights with substantive content and to reverse the onus of proof so as to require the state to prove that it has taken “reasonable legislative and other measures, within its available resources, to achieve the progressive realization of the right”.  The question for the future is how courts are to respond to claims of this kind.

The South African test of reasonableness is clearly one avenue of response, but reasonableness folds within it a multitude of factors and considerations involving the defining, demarcating and ranking of divergent values and interests which are arguably better disaggregated and more transparently and carefully analyzed.  The emerging area of benchmarks and indicators provides one possible way of approaching this task.  Another is to extend and refine limitations- and balancing-style analysis in the domain of economic and social rights.  Much thought and work, however, remains to be done on these important issues.

Concluding Remarks

The significance of Mazibuko lies in three areas.  First, consistent with the earlier economic and social rights decisions of the Court, the case provides the intriguing next chapter in the domestic implementation of international human rights norms in South African law.  The applicants and COHRE were successful in persuading the two lower courts to incorporate international standards in their interpretation of section 27(1).  The Constitutional Court, by contrast, has signaled that while international law is relevant and helpful for constitutional analysis (as the Constitution itself requires) it does not intend to adopt the minimum core approach but rather will develop its more flexible reasonableness doctrine in an effort to forge a distinctly South African attitude to the justiciability of economic and social rights.

Second, although the final outcome in Mazibuko has no doubt disappointed rights advocates, the case provides an interesting case study of (a) how ESC rights can be claimed, contested and progressively realized if only indirectly as evidenced by the City’s continual review and adjustment of its water policies in response to ongoing public interest litigation; (b) how government itself and specific aspects of public policy can be held accountable via a constitutional culture of justification; and (c) how participatory and deliberative democracy is ultimately deepened by the constitutional contestation of economic and social rights.  Of particular interest in this regard is how both international and local human rights NGOs such as COHRE and CALS were able to partner with a social movement such as CAWP to use strategic litigation in conjunction with strong social mobilization and advocacy in an effort to secure recognition and implementation of an economic and social right.

Finally, and perhaps controversially, one may query the extent to which the Court’s caution and restraint in delimiting its adjudicatory role in South Africa’s constitutional democracy and its receptiveness to the City’s arguments (by contrast to the more rights-protective approach of the lower courts) was influenced by the political turmoil resulting from the Hlophe and Zuma sagas which have buffeted the institutional legitimacy of the Court itself and focused renewed attention on the “counter-majoritarian” nature of the Court’s constitutional task, especially in the areas of equitable distribution of power and resources. 

Print Friendly, PDF & Email


No tags available

Leave a Comment

Comments for this post are closed


Bret Thiele says

January 19, 2010

The article provides an interesting analysis of the case, although misses to some degree the nuance of COHRE's amicus curiae intervention (as we argued that a minimum core should be used in determining reasonableness rather than in place of the reasonableness standard).

Jackie Dugard says

April 17, 2010

This article misses many important points about the Mazibuko case and judgment. I will focus on only three here. First, the Constitutional Court misunderstood the applicants as arguing for a minimum core approach to the right to water. They did not. Rather, the applicants pursued the approach established by the Constitutional Court in Grootboom (in its rejection of the minimum core content approach, as being too inflexible), which is that rights and obligations can only be established in context. This is precisely what the applicants did in Mazibuko: they asked the Court to determine the reasonableness of the City's Free Basic Water policy in the context of a high-density urban township with waterborne sanitation and no alternative water or sanitation sources. The Court, however, cast this as a minimum core content argument. And, displaying an extraordinary degree of deference, found the City's Free Basic Water policy to "fall within the bounds of reasonableness", which appears to me to be a worrying retreat from the standard of reasonableness and of inquiry set in Grootboom. Second, contrary to the findings of both the High Court and the Supreme Court of Appeal, the Con Court found the City's interpretation of the by-laws as allowing the installation of prepayment meters to be "textually permissable", which seems to be a new form of highly deferent legal interpretation. Third, in dismissing the applicants arguments that prepayment meters amount to unfair discrimination based on race - because, despite proven debt across the City, prepayment meters have only been installed in poor black areas - the Court said that the applicants had not proven that prepayment meters were installed in ALL black areas. This is nonsensical and goes against all its previous equality decisions. It would mean that, for example, if I allege that my dismissal on the grounds of my sexual orientation (a listed ground in the Constitution) amounted to unfair discrimination, I would have to prove that my employer had dismissed all other gay employees in the organisation. In South Africa, there is growing concern about the Mazibuko judgment and the Court's apparent retreat from enforcing socio-economic rights. The subsequent judgment of Nokotyana is another case that raises similar concers.