On 6 May, the Constitutional Court delivered judgment in the “Phala Phala” case, concerning Parliament’s decision not to hold an impeachment inquiry into allegations that the president had committed a “serious violation of the Constitution or the law” after foreign currency was stolen from his farm.

The court took 18 months to deliver the judgment. Five months earlier, the applicant, the EFF, publicly protested over the delay. Delivering one of three judgments, the Chief Justice apologised: “I must take full responsibility for the delay in producing this judgment concerning an extremely difficult matter of great national importance. I tender my sincere apologies to the parties, my colleagues and fellow South Africans for the inconvenience it has caused.”

The apology suggests the chief justice, Mandisa Maya, was herself the cause of the delay. But delays of this duration are not unusual.

In 2024, then-Chief Justice Raymond Zondo reported that two retired Constitutional Court justices were helping read applications for leave to appeal and prepare memoranda to address backlogs following the court’s expanded jurisdiction in 2013. He acknowledged that the expansion had not been accompanied by increased human resources.

Freedom Under Law (FUL) and other civil society organisations objected. We argued that allowing retired justices, whose terms had ended, to advise on which cases the court should hear was constitutionally impermissible. The arrangement had also emerged through media reports, without prior public consultation.

Zondo abandoned the scheme and called for institutional reform. Yet nothing concrete followed – not from the Department of Justice, Parliament, or the Office of the Chief Justice.

FUL has since published a report examining the court’s structure, processes and procedures, proposing reforms to reduce delays and improve efficiency. This lecture draws extensively on that report.

How we got here

Under the Interim Constitution, the Constitutional Court had the final say on constitutional matters, while the AD (Appellate Division) heard everything else.

The structure of the apex courts was hotly debated. Chief Justice Michael Corbett favoured a single appeal court with a constitutional chamber. Instead, South Africa adopted a specialist, representative Constitutional Court (CC) to interpret and enforce a future democratic Constitution — reflecting distrust of an AD emerging from apartheid, and drawing on international models, particularly Germany’s. Under the final Constitution, the CC’s 11 judges heard all cases together, confined to constitutional matters and deciding all cases en banc.

Dividing authority between two apex courts created difficulties. The CC’s lighter workload, compared with the renamed SCA (Supreme Court of Appeal), encouraged an expansive interpretation of “constitutional matter” broadly and later helped justify extending its jurisdiction.

READ: [Click here]

I suggest, however, that the febrile political atmosphere following Jacob Zuma’s election as ANC president in December 2007 created the real conditions for restructuring. After the SCA reinstated corruption charges against him in January 2009, his supporters increasingly characterised it as “untransformed”. As president, he reconstituted the Judicial Service Commission (JSC), paving the way for four new CC appointments.

The Constitution Seventeenth Amendment Act of 2012 extended the court’s jurisdiction to arguable points of law of general public importance, making it an apex court of general jurisdiction.

In my view, no thought was given to the effect of its workload, the wider appellate system, and the expertise required of its judges.

Whatever one makes of that history, the court’s caseload must now be managed. What follows is not an indictment of any judge or political moment — it is an argument about institutional design.

Analysis of the functioning of the court

FUL’s central concern is that the Constitutional Court’s efficiency has declined, causing inordinate delays.

Between 2010 and 2021, the average time from hearing to delivering judgment almost doubled. Since then, the court has taken about 210 days — seven months — to deliver judgment.

This corresponds with increased caseloads, suggesting that processing new applications contributes to the delay. Yet despite hearing fewer cases and delivering fewer written judgments since 2022, the court’s hand-down times have not improved. Annual applications have remained relatively stable.

This shows up in the sluggishness with which new applications are processed in chambers. At least half the petitions filed in 2024 remained pending for six months or more, most ultimately destined for dismissal by a one-line order. Litigants whose matters proceed to a hearing may wait about 18 months from application to written judgment.

The Phala Phala case illustrates the problem. Its 18-month delay ran from hearing to judgment alone. The EFF protested, and the Chief Justice apologised. Most litigants get neither.

These inefficiencies are the consequence of an ill-considered expansion of jurisdiction. The court is now inundated with unmeritorious generalist appeals instead of public importance matters. Litigants have little to lose by rolling the dice on one final appeal, even where their prospects are poor.

Volume is not the only concern. The competence of judicial appointments and delays in filling permanent vacancies are another. The court’s expanded jurisdiction has compromised its specialist constitutional function – giving the dilemma, as FUL reports it, an existential dimension.

Tellingly, the drafters of the amendment never even changed the courts’ names: the CC is no longer a “constitutional court,” and the SCA is anything but supreme.

The interests of justice test

The court’s use of the “interests of justice” test when deciding whether to grant leave to appeal has compounded its workload. We believe this open-ended test needs a clearer, rules-based approach.

Section 167(6) of the Constitution permits direct appeals “when it is in the interests of justice and with leave of the Constitutional Court”. The court asks two questions: does the appeal engage its jurisdiction, and is it in the interests of justice to hear it? Raising a constitutional issue does not guarantee leave.

The expanded jurisdiction similarly covers arguable points of law of general public importance “which ought to be considered.” Those words allow the court to refuse leave even where such a point arises, as it recognised in Minister of Tourism v AfriForum NPC 2023 (6) BCLR 752 (CC) para 27.

The difficulty lies in the numerous factors considered: prospects of success, public importance, factual disputes, whether lower courts considered the issue, mootness, costs, and whether the appeal concerns an interim order or the development of the common law.

At times, the court treats one factor as decisive without explaining how it weighs against the others; elsewhere, the same factor carries little weight. Practitioners consequently struggle to advise clients whether an appeal will be heard, regardless of its merits. Litigants seeking merely to delay compliance with a lower court’s order have an incentive to roll the dice, further clogging the court’s roll.

Rule-based reasoning

The interests-of-justice test seeks to ensure that deserving cases are heard and undeserving cases do not consume the court’s time. But the rule of law requires clear, intelligible and predictable laws. No discretion should be so unconstrained as to be arbitrary.

As Lord Thomas Bingham explains, legal certainty enables compliance, facilitates obligations and promotes commercial activity. Judge David Unterhalter made this point too, warning against standards “so porous that a litigant cannot be advised, with any reasonable probability, as to the decision that a court is likely to make”.

Frederick Schauer distinguishes between rule-based and particularistic reasoning. The former treats a rule as sufficient for a decision, even where it produces the “wrong” result in an individual case. The latter treats rules as guides, yielding to an assessment of the particular circumstances.

For litigants, the distinction matters. When comparable applications produce different outcomes and refusals carry no reasons, neither applicants nor their advisers can understand what determines admission to the court.

In TWK, Unterhalter applied the Zweni criteria for appealability and the further rule that a dismissal of an exception is not appealable unless it concerns jurisdiction. The rationale is to preserve appellate capacity and prevent piecemeal litigation by ordinarily hearing only final matters. Applying this rule avoids having to reconsider that rationale in every case. It provides, as he puts it, “tolerable certainty”.

The test ought to remain at the heart of appealability, not be swallowed by a free-floating interests-of-justice standard. We accept that deciding which cases to hear cannot be strictly rule-based. Some balancing and discretion are inevitable.

But I suggest the court’s approach has swung far towards particularism.

Possible solutions

Before considering structural solutions requiring constitutional or legislative amendment, there is much the court could do itself.

The starting point is a more rules-based interests-of-justice test. Certain factors should be decisive, or at least pivotal, with applicants required to justify exceptions.

A moot appeal, for instance, should ordinarily be refused unless the applicant shows grounds such as conflicting lower-court judgments causing commercial uncertainty, or irreparable harm. The court should explain why an exception is justified.

READ: [Click here]

The practice of refusing leave through bare, one-line orders should be reversed. Reasoned judgments mostly explain why leave is granted, producing a more permissive jurisprudence than the court intends. Reasoned refusals would give the test content and help litigants assess their prospects before spending time and money. Clearer practice directives on applications and their assessment would reinforce this.

The court should also impose adverse costs orders on practitioners who disregard its rules and enforce existing limits on the length of applications and annexures.

None of this will matter much if those appointed to the Bench cannot do the job.

An apex court requires competent jurists with demonstrable skills across the full range of law before it. That responsibility lies squarely with the JSC – which, I am bound to say, has been indifferent to the outcomes its own appointments have produced.

Long-term solutions

Long-term reform would require constitutional or legislative change. A comprehensive review of our appellate system is overdue.

The most extensive option is to merge the CC and the SCA, with a constitutional chamber hearing appeals involving decisive constitutional issues and a general chamber hearing everything else. Its size would need careful consideration, and provincial decisions would need competent staffing to handle most ordinary appeals.

A more modest option is to restore the court’s exclusively constitutional jurisdiction. But some commentators dispute whether the 2013 expansion was the principal driver of increased caseloads: the court had already interpreted its constitutional jurisdiction broadly. That is a fair challenge. Restricting jurisdiction on paper may achieve little in practice.

If the existing structure remains, smaller assessment panels could decide compliance, jurisdiction and leave to appeal, with any disagreement triggering referral to the full court. Alternatively, a screening unit of legally trained staff could prepare memoranda, as in Germany. The justices would remain the decision-makers, though this would require additional resources.

The court could also expand to 15 justices, with a smaller assessment panel and a larger quorum for hearing appeals. A properly filled Bench could reduce the effects of vacancies, absenteeism and reliance on acting justices. Yet separate panels risk conflicting decisions, undermining the certainty an apex court should provide.

More law clerks is another option, though it’s unclear whether the caseload justifies the additional cost.

Doing nothing not a defensible option

FUL has no definitive view on which proposals are preferable. Our examination of apex courts in the UK, Germany, Canada and the US offers lessons – some more useful than others.

On one thing, though, I will not pretend to be neutral: doing nothing is no longer a defensible option.

Every year this debate is deferred, litigants wait longer, reasons go unexplained, confidence in our courts diminishes and trust in our constitutional democracy is eroded. The next Phala Phala is already on the roll, and somewhere, a litigant will wait a year and a half only to receive a one-line refusal without reasons.

It is our hope that this contribution will stimulate debate and lead to reform. Significant voices within the judiciary share this view. It is time for our universities’ law faculties to join that endeavour.

– Azhar Cachalia is a retired judge of the SCA, chairperson of FUL and visiting professor at Rhodes University.

This is an edited version of a public lecture delivered at Rhodes University on 26 August 2026. This edited version was compiled by Lance Myburgh, a final-year law student. Watch the full lecture [here]

First published on News24