Bwanya v Master of the High Court: Right for the Wrong Reasons

"Hold Hands" by Billy Simon is licensed under Creative Commons BY-NC-SA 2.0.

In 2005, in Volks NO v Robinson, the Constitutional Court held that the Maintenance of Surviving Spouses Act 27 of 1990 (MSSA) did not unfairly discriminate on the basis of marital status by excluding persons in permanent life partnerships from claiming maintenance in terms of that Act.

So tendentious were the arguments in Volks that the decision was rightly described as one of the Court’s worst and as betraying a latent moral conservatism. It is therefore cause for celebration that, on 31 December 2021, in Bwanya v Master of the High Court, the Court (in effect) overruled Volks and held that both the MSSA and Intestate Succession Act 81 of 1987 (ISA) are unconstitutional to the extent that they do not confer benefits on persons in permanent life partnerships who have undertaken reciprocal duties of support to their partners.

There is much in Bwanya deserving of analysis and praise. My present concern, however, is a limited one: this being the first time that the Court has expressly refused to follow one of its previous decisions, on what basis did it do so, and what are the implications for the doctrine of precedent? 

Precedent and Volks

The doctrine of precedent provides that a court is bound by its previous decisions unless satisfied that they are clearly wrong. We might therefore have expected that, in Bwanya, the Court would either have held that Volks was clearly wrong or would have been constrained to follow it. But alas, said Madlanga J for the majority—

“[m]uch as I am convinced that the Volks decision was wrong, I am unable to make the jump and conclude that it was clearly wrong. Does this mean I must reach the same conclusion as Volks? I think not.”

His reasoning was this. Volks rested on two central arguments— the so-called “choice argument” and what we might call the “extension of legal duty argument”. In terms of the former, the MSSA does not unfairly discriminate, because persons choose to marry, and thus choose whether or not to accept the benefits and burdens of marriage, including those of the MSSA.

For Madlanga J, this argument rested on a factual premise: that persons always genuinely choose whether to marry or not.  Since the doctrine of precedent does not apply in respect of factual findings, Madlanga J held that he was not constrained to follow the “choice argument”.

This, however, was a misconstrual of that argument.  As Ngcobo J (as he then was) explained in Volks—

“It is also true that [a couple] may not marry because one of the parties does not want to get married. Should the law then step in and impose the legal consequences of marriage in these circumstances? To do so in my view would undermine the right freely to marry and the nature of the agreement inherent in a marriage. Indeed it would amount to the imposition of the will of one party upon the other. This is equally unacceptable.”

Volks therefore did not endorse the factual premise that persons always have a choice whether or not to marry. Instead, as Ngcobo J made clear, the Volks majority took precisely the opposite as a premise – some persons have no choice to marry because their partners refuse to do so – and held that the MSSA passed constitutional muster because it preserved the choice of the refusing partner. We might well disagree with that reasoning but it is plainly normative rather than factual, and could not therefore be ignored on the basis that it was a mere finding of fact.

The Bwanya majority’s further reason for departing from Volks concerned the “extension of legal duty argument”. In terms of that argument, an invariable consequence of marriage is that, by operation of law, married partners bear a legal duty of support to each other during the subsistence of the marriage. The same is not so in respect of permanent life partnerships where duties of support (to the extent that they exist) arise by agreement. For the Volks majority, this difference had an important implication: it was fair for the legislature to limit the freedom of testation of a married person by requiring such person’s estate to maintain the surviving partner, because this was merely to extend a duty which arose by operation of law. Conversely, it was fair to not similarly limit the freedom of testation of persons in permanent life partners, because to do so would amount to more than the extension of a legal duty which arose by operation of law.

Madlanga J held that he was entitled to depart from this reasoning because, since Volks, the common law had been developed in Paixão v Road Accident Fund. There, the Supreme Court of Appeal held that a person in a permanent life partnership can claim in terms of the dependant’s action and can therefore claim loss of support from a wrongdoer responsible for the death of his or her partner.

For Madlanga J, this development meant the “extension of legal duty argument” was no longer sustainable. He thus held that “it can no longer be fitting to distinguish the duty of support existing in the two categories of familial relationships (i.e. marriage relationship and permanent life partnership) purely on the basis that one arises by operation of law and the other by agreement”.

Here too, Madlanga J’s reasoning strains the smell test. First, Paixão was an SCA decision. Were it inconsistent with Volks, as Madlanga J held, it would have been wrongly decided, since the SCA was constrained to follow Volks. Second, in Volks, the distinction between a legal duty of support arising by operation of law, and one arising by agreement, was deemed relevant because of the implications of the MSSA for the freedom of testation. In the context of the dependant’s action, which was the concern of Paixão, freedom of testation is not implicated. In such an action, the claim is made against the wrongdoer and not the estate of the deceased partner. For precisely this reason, the Paixão court held that it was unconstrained by Volks. And this makes sense. Consider the following two questions: should the estate of a deceased partner be obliged to maintain his or her surviving partner? Should a person who wrongfully causes the death of a person, and thereby deprives his or her partner of support, be obliged to remunerate such a person? The normative considerations relevant to each question are plainly different, and there is no inconsistency in providing different answers to each of these questions. Paixão therefore provided no way around Volks.

“No longer, it seems, is a precedent binding unless clearly wrong. A new principle has emerged: a Constitutional Court precedent can lose its binding force if subsequent developments of the common law – even if effected by a lower court -  undercut its rationale.”

A simpler route 

A year before Volks, the Court delivered judgment in Daniels v Campbell. That case asked whether a person married in terms of Muslim rites, who had not solemnised their marriage in terms of the Marriage Act 25 of 1961, could claim in terms of the MSSA and ISA.

Like a permanent life partnership, such a marriage does not give rise to a duty of support by operation of law and, like persons in a permanent life partnership, persons married in terms of Muslim rites can choose to marry in terms of the Marriage Act. The “choice argument” and “extension of legal duty argument” were thus in play. Crucially, however, the Court held that neither argument was relevant in deciding whether a person should be entitled to claim in terms of the MSSA and ISA—

“The central question is not whether the applicant was lawfully married to the deceased, but whether the protection which the Acts intended widows to enjoy should be withheld from relationships such as hers. Put another way, it is not whether it had been open to the applicant to solemnise her marriage under the Marriage Act, but whether, in terms of “common sense and justice” and the values of our Constitution, the objectives of the Acts would best be furthered by including or excluding her from the protection provided.”

The principle that emerged was therefore clear: neither the “choice argument” nor the “extension of legal duty argument” provide a constitutionally sound basis to deprive someone of the benefits of the MSSA and ISA.  By subsequently endorsing those arguments, Volks departed from Daniels (without explanation), and for this reason was clearly wrong. As Meyerson commented, the unmistakable implication of this departure was “that those whose reasons for not entering a valid marriage are not religious are second class citizens and less worthy of respect”. 

Right for the Wrong Reasons

In short, therefore, Bwanya was right for the wrong reasons. For Madlanga J, Daniels was not of assistance because the discrimination in that case was a result of racism, not marital status.  But this was to ignore the true import of Daniels.

Bwanya’s implications for the doctrine of precedent are unclear. The finding that the “choice argument” was one of fact – erroneous though it was – is unlikely to have any great precedential effect. By contrast, the finding that a development of the common law, subsequent to Volks, undermined that decision, is both intriguing and of potentially significant import.  

No longer, it seems, is a precedent binding unless clearly wrong. A new principle has emerged: a Constitutional Court precedent can lose its binding force if subsequent developments of the common law – even if effected by a lower court -  undercut its rationale. Most likely, the Bwanya majority felt that by refusing to declare Volks clearly wrong, and instead following this circuitous route, it was best preserving the doctrine of precedent. The irony, therefore, is that by endorsing this decidedly suspect new principle, the Court has done precisely the opposite.