Was abolishing the death penalty the right decision? - IRR

Institute asks whether there is a case for the reintroducation of capital punishment

Capital punishment in South Africa: Was abolition the right decision?

Is there a case for South Africa to reintroduce the death penalty?



It is now just over 21 years since the Constitutional Court abolished the death penalty. It did so at a time when opinion polls showed that most South Africans supported capital punishment. Although there is now little public debate on the issue, some people believe such abolition was a luxury South Africa could not afford. Others believe the death sentence should be reinstated and have asked the Institute of Race Relations (IRR) to review the question of capital punishment.

This paper will do that, examining arguments for and against capital punishment, as well as crime trends both in South Africa and elsewhere. It will also review the history of capital punishment in South Africa, the announcements leading up to its abolition, and the reasoning of the Constitutional Court in the decision it handed down in June 1995. In addition, it will raise various broader questions about the criminal justice system in South Africa.


Hangings took place in South Africa prior to the unification of the two Boer republics and the two British colonies in 1910 and they continued after Union. Between 1910 and 1975, altogether 2,740 people were executed, and another 1,100 between 1981 and 1989. This would bring the total to around 4,000, yielding an average of 50 a year over 80 years. The last hanging took place in 1989, following which the then state president, F.W. de Klerk, put a stop to them pending a decision on capital punishment by the Convention for a Democratic South Africa (Codesa). Judges could still impose the death sentence, however, and they continued to do so, but the sentences were not carried out. Although Codesa adopted a comprehensive bill of rights as part of the 1993 Constitution, it did not outlaw capital punishment, leaving that matter to the court. When it issued its prohibition on further executions in 1995, possibly as many as 400 people were waiting on death row.

Until it was abolished, capital punishment had been implemented not only for murder but also for rape, housebreaking and robbery or attempted robbery with aggravating circumstances, sabotage, training abroad to further the aims of communism, kidnapping, terrorism and treason. It had also at one stage been mandatory until courts were authorised to impose lesser sentences if they found “extenuating circumstances”. One of Mr de Klerk’s reforms was to make the death sentence an “option” for “extreme cases” only, thereby giving judges greater discretion and providing for automatic appeals. Some 90% of executions carried out were for murder, but people were also hanged for rape, aggravated robbery, and sabotage. Barend van Niekerk, South Africa’s best-known abolitionist, said in 1967 that 47% of the executions in the world were carried out in South Africa, a figure that was questioned but never refuted.

The decision of the Constitutional Court

The decision of the Constitutional Court outlawing capital punishment was made in the case of State v Makwanyane and Mchunu, which was heard in February 1995, judgement being handed down in June that same year. The court’s decision, written by its president, Arthur Chaskalson (later chief justice), was unanimous. All ten of the other judges gave reasons for their concurring views, with some of them placing emphasis on particular points. Some of these will be discussed below where relevant.

The two accused had been sentenced to death in terms of the Criminal Procedure Act of 1977 on each of four counts of murder. This was a lawful sentence when it was handed down, but the 1993 Constitution had subsequently come into effect. In his judgement, Justice Chaskalson stated that it would have been better if the framers of the Constitution had decided whether or not the death sentence was permissible. It had not done so, with the result that the court now had to decide whether the implementation of the sentence was constitutional.

The judgement summarised the two opposing viewpoints presented to the court: one advanced on behalf of the condemned men, the other on behalf of the state by the attorney general of the Witwatersrand, Klaus von Lieres und Wilkau. The former argued that the death sentence was cruel, inhuman, degrading and an affront to human dignity; that it was inconsistent with the right to life entrenched in the Constitution; that it could not be corrected in the case of error; and that it could not be enforced in a manner that was not arbitrary. The latter contended that the death sentence was recognised as legitimate in many parts of the world; that it was a deterrent to violent crime; that it met society’s need for adequate retribution for heinous offences; and that it was regarded in South Africa as an acceptable form of punishment.

Although the Constitution prohibited “cruel, inhuman or degrading treatment or punishment”, it did not define this so the court had to. It did so by reference to other provisions of the Constitution, which guaranteed the rights to life, dignity and equal protection of the law. Said Justice Chaskalson: “The carrying out of the death sentence destroys life, it annihilates human dignity, elements of arbitrariness are present in its enforcement, and it is irremediable... I am satisfied that in the context of our Constitution the death penalty is indeed a cruel, inhuman and degrading punishment.”

That, however, was not the end of the matter, because the Constitution provided for rights to be overridden. But they could be limited only if the limitations were reasonable, justifiable, necessary, and did not negate the essential content of the rights in question. The value of the death sentence as a deterrent, as a form of retribution, and as a means of preventing further crime, had to be weighed against alternative measures available to the state, particularly life imprisonment. It had not been shown that the death sentence was materially more effective than life imprisonment, with the result that the requirements of the limitations clause were not met. Taking this into account, concluded Justice Chaskalson, “as well as the elements of arbitrariness and the possibility of error in enforcing the death penalty, the clear and convincing case that is required to justify [it] as a penalty for murder has not been made”.

The relevant provisions of the 1977 act, and any other legislation sanctioning capital punishment, were therefore “inconsistent” with the Constitution and so “invalid”. The state was accordingly “forbidden to execute any person already sentenced to death” under any of the laws now declared to be invalid. Such persons should remain in custody pending the imposition upon them of other punishments.

The court did not, however, invalidate the death penalty for treason committed in wartime. No argument was put to it on this issue, so it refrained from expressing any views about it.

Public opinion

Speaking for the state, Mr von Lieres argued that “overwhelming public opinion in favour of the retention of the death penalty is sufficiently well known to be accepted as the true voice of South African society”. Justice Chaskalson said he was prepared to accept this contention. However, he added, “the question before us is not what the majority of South Africans believe a proper sentence for murder should be. It is whether the Constitution allows the sentence”. The Constitution, in other words, outweighed public opinion, which was effectively dismissed by the court (along with the notion of submitting the matter to a referendum).

Justice Chaskalson went on to quote an American judge, Robert Jackson, in a 1943 case:

“The very purpose of a bill of rights is to withdraw certain subjects from the vicissitudes of political controversy, to place them beyond the reach of majorities and officials, and to establish them as legal principles to be applied by the courts. One’s right to life, liberty, and property, to free speech, a free press, freedom of worship and assembly, and other fundamental rights may not be submitted to vote; they depend on the outcome of no elections”.

Other countries

Ellison Kahn, one of South Africa’s leading academic lawyers, pointed out in an article in 1970 that England at one stage had 200 capital offences, including consorting with gypsies. Even children as young as nine years old could be hanged. However, since the early 19th century the trend had been in the direction of abolition.

According to Amnesty International, more than 20,000 people worldwide were under sentence of death in 2015. In that same year 1,634 executions were carried out in 25 countries. Of these executions, 28 took place in the United States, about the same number as in Iraq, Somalia and Egypt respectively. There were 158 in Saudi Arabia, 326 in Pakistan and 977 in Iran. The number in China – “the world’s top executioner” – was not known, but thought by Amnesty International to run into thousands.

Amnesty International reported that the 1,634 known executions in 2015 was the highest number in 25 years. On the other hand, four more countries abolished the death penalty for all crimes, bringing the total to 102. If the countries which had abolished the death penalty in practice although not in law were included in the count, the total came to two thirds of the world.

Although the global trend is towards abolition, there is no reason why South Africa should follow it. The arguments about capital punishment in South Africa must stand on their own merits. This does not, however, exclude taking account of decisions and arguments in courts in other countries, as our courts normally do in any event in all kinds of trials. The Makwanyane judgement is liberally interspersed with references to other courts and other judges.

Crime statistics

It is difficult to compare crime figures before and after the Makhanyane judgement because figures for many earlier years exclude some of the homelands. Figures for the period 1994/1995 to 2014/2015 have, however, been published annually by the South African Police Service and are reproduced and analysed by the IRR in its annual South Africa Survey. The police’s crime figures are based on crimes reported to them.

They indicate that aggravated robbery (previously a hanging offence) has increased in frequency in that 21-year period. The aggravated robbery rate has thus risen from 219 per 100,000 of population to 239 (by 9%). The number of aggravated robberies taking place each day has risen from 232 to 354 (by 52%).

Rape was previously also a hanging offence, but trends here are more difficult to identify as rape is no longer classified on its own. Since 2007 it has been lumped together with other “sexual offences”, whose incidence since then has dropped. After rising for a few years, reported sexual offences against women and children have dropped since 2010, although it is too early to say whether this is a steady downward trend.

Turning to murder, police figures show that South Africa’s murder rate has dropped from 67 per 100,000 in l994/1995 to 33 (by 51%) in 2014/2015. The average number of murders committed each day has dropped from 71 to 49 (by 31%). The downward trend in murder figures has, however, been questioned by various institutions, among them the IRR. Moreover, Interpol has suggested that the murder rate is substantially higher than the figure reported by the police. Making use of data on “non-natural deaths”, the South African Medical Research Council has suggested likewise. Official figures on such deaths, not all of which arise from murder, nevertheless show a downward trend. The actual picture of murder trends since the abolition of the death penalty is therefore far from clear.

What is clear is that South Africa, according to comparative data published by the United Nations, still has a very high murder rate of 32 per 100,000, much the same as Colombia, although lower than El Salvador, Honduras, and Jamaica. Our murder rate is nevertheless very much higher than those of countries that include Australia, the Czech Republic, Denmark, France, Germany, the United Kingdom, Ireland, Italy, Portugal, Spain and Switzerland (all between 0.7 and 1.3 per 100,000). Figures for other countries include Mexico at 18.9, the United States at 3.8, India at 3.3, Hungary at 2.7 and Morocco at 1.3. Among this last group of countries, both the United States and India apply the death penalty, but Hungary has abolished it.

Other countries also show declining murder rates, according to a recent article in the British weekly The Spectator by Andrew Taylor, a crime writer. Homicide rates throughout the industrialised world have declined. Falls in Britain, the United States and Canada are particularly marked: these were countries where homicide rates increased in the 1960s and 1970s. Mr Taylor is hard put to explain the downward trend, however. He finds no correlation between national wealth and crime figures. During the Great Depression of the 1930s unemployment rose to 25%, but the crime rate in many British cities went down. When Britain became wealthier in the 1960s, crime started to rise. Since the beginning of the Great Recession in 2008, however, murder rates have continued their downward trend. The United Nations reported in 2011 that the murder rate had fallen worldwide during the recession.

While increases in prison populations (in countries that include the United Kingdom and Australia) are sometimes seen as reasons for declines in crime rates, in Canada and the Netherlands large numbers of prisoners have been released without surges in murder rates. Imprisonment in New York has dropped in the last decade by 26%, and crime has fallen by 28%. The British homicide rate has dropped despite substantial cuts in police budgets. Some analysts believe the decline in homicide in the United Kingdom is the result of the fact that the population is ageing, so that a smaller proportion is accounted for by supposedly crime-prone youth. But London is experiencing a drop in the crime rate at the same time as there has been an increase in the number of 18- to 24-year-old young men. Countries with strict gun laws are said to have fewer homicides, but the rate at which Americans have gunned one another down has halved over the past two decades, even though gun sales have risen since President Barack Obama entered the White House.

After considering various theories, Mr Taylor concludes that the answer to the question “Why has murder dropped?” is “We don’t really know”. Gary Bell QC argues in the same Spectator, however, that technology may be the answer: vehicles can be tracked, as can cellphones, while DNA testing is now available. Says Mr Bell: “It has never been harder to get away with murder.” Unfortunately, this comment cannot apply to South Africa. Even if it is true that the murder rate has come down, as police figures suggest, it is still extremely high. Moreover, as we shall see, the Constitutional Court pointed out that most violent crime in this country goes unpunished and that, rather than the method of punishment, is the fundamental problem.

Since abolition, there has been a 24% increase in the prison population. Although there has been a 56% increase in crime arrests since 2002/2003, the proportion of arrests resulting in convictions has dropped by 43%. It therefore seems likely that the main reason for the increase in the prison population is not that more people are being incarcerated, but that prisoners are staying in jail for much longer. Between 1995 and 2012 the number of prisoners serving sentences of more than 10 years rose from 1,620 to 43,632, an increase of 2,593%. The number of life sentences has risen from 433 in 1995/1996 to 12,658 in 2013/2014, an increase of 2,823%.

Most of these sentences would have been imposed under the Criminal Procedure Amendment Act of 1997, which provides that courts should impose life sentences for “planned and premeditated” murder, rape and certain other crimes, unless they find “substantial and compelling circumstances” to impose lesser sentences. Minimum sentences of shorter periods were stipulated for other crimes in the absence of such circumstances justifying shorter periods. In line with this thinking, the judge in the murder trial of Oscar Pistorius handed down a lesser sentence in July 2016 than the 15-year minimum provided for, after she found that the mitigating factors outweighed the aggravating factors in his crime.


This second part of the paper will discuss various relevant issues in more detail, drawing both on the judgement of the Constitutional Court and on other arguments.

“Cruel and unusual”

There are two issues here. One is the method of execution itself. The other is the period leading up to it.

In his inaugural lecture in 1990 as professor of public law at the University of Natal, Durban, Professor George Devenish quoted a former South African hangman as having said it sometimes took 15 minutes for the hanged prisoner to die. In support of its contention that capital punishment was “barbaric”, the American Civil Liberties Union (ACLU) in 2012 described various methods by which execution was carried out, including hanging, firing squad, electrocution, gassing, and lethal injection. It gave several explicit examples of how executions were sometimes bungled, in different ways, leading to prolonged deaths.

The Constitutional Court said that the mental anguish of convicted persons awaiting execution was well documented. It pointed out that the death sentence was not unconstitutional in India, Zimbabwe, and Jamaica. However, sentences in these countries had been set aside on grounds of mental anguish. Justice Chaskalson quoted a statement by six of the seven judges of the California supreme court in 1992 to the effect that the cruelty lay not only in the execution itself and the pain incident thereto, but also in the dehumanising effects of the lengthy imprisonment prior to the execution. During this period judicial and administrative procedures essential to due process were carried out.

One of the concurring judges, Sydney Kentridge, pointed out that the mental agony must be present even when the time between sentence and execution was measured in months or weeks rather than years. Another concurring judge, John Didcott, quoted the chief justice of Zimbabwe, Anthony Gubbay, as well as various American judges, on the “primal terror” of the condemned man awaiting execution. These assessments had not been disputed, nor was there any reason to believe they were exaggerated or inaccurate. They sufficed to convince Justice Didcott that “every sentence of death may be stamped as intrinsically cruel, inhuman, and degrading”.

None of the judges dealt explicitly with the additional anguish likely to be suffered by an innocent person awaiting execution. As we shall see below, however, innocent people are known to have been executed in countries that include Australia, China, Ireland, South West Africa/Namibia, Taiwan, the United Kingdom and the United States.


The Constitutional Court did not address the question of race in any detail, although, as we shall see, poorer people, most of whom are black, may be at a disadvantage in murder and other trials. Professor Devenish quotes another South African law professor and abolitionist, John Dugard, as having said that of 2,740 persons executed between 1910 and 1975, fewer than 100 were white. He also quotes Professor van Niekerk as stating that 288 white people were convicted of raping black people between 1947 and 1966 but that not one of them had been sentenced to death.

On the other hand, 844 black people had been convicted of raping white people over the same period and 122 had been executed. These stark numerical differences suggest that race played a role in whether or not rapists across the colour line were executed. CR Swart is quoted as saying that when he was minister of justice (in the 1950s) not a single reprieve had been granted to a black man sentenced to death for the rape of a white woman. Possible injustices arising from the granting of reprieves are discussed below.

In an article in the South African Law Journal in 1969, Professor van Niekerk reported on a questionnaire he had sent to 158 advocates, nearly all of whom had been involved in trials where the death penalty was imposed. Almost half believed it was meted out on a differential basis to different races and of those, 41% believed that such differentiation was “conscious and deliberate”. Professor Kahn observed that although it was sometimes claimed that no white person had ever been hanged for the murder of a black person, there were at least six cases on record since Union, three of them since 1949.

The history of capital punishment in South Africa cannot be divorced from the apartheid system, which meant that judgement in the past was imposed on black people by white people. In 1994 some 98% of the judges on South Africa’s superior courts were white, a proportion which has now dropped to 36%. Africans constitute 44%, and coloured and Indian judges 10% each. The multi-racial character of the courts may give them greater credibility, but that in itself is no guarantee that racial prejudice one way or the other will no longer be a factor in judicial decisions. However, even if race could be entirely eliminated as a factor in such decisions, this would not remove other objections to the death penalty.


One of the most famous of all British judges, Alfred (“Tom”) Denning, who died at the age of 100 in 1999, said that sentencing people to death “didn’t worry [him] in the least” as it was the most appropriate penalty for murder. Testifying before a royal commission on capital punishment which sat between 1949 and 1953, he said:

“In order to maintain respect for the law is essential that the punishment inflicted for grave crimes should adequately reflect the revulsion felt by the great majority of citizens for them. It is a mistake to consider the object of punishment as being deterrent or reformative or preventive and nothing else. The truth is that some crimes are so outrageous that society insists on adequate punishment because the wrongdoer deserves it irrespective of whether it is a deterrent or not.”

Lord Denning’s statement no doubt echoes the views of a great many people in this and other countries, especially when they think of the extreme cruelty and wanton brutality that accompanies so many crimes, including murder, rape and armed robbery, not to mention the irreplaceable loss of a loved one. They no doubt hold these views even though Lord Denning later changed his mind and wrote: “Is it right for us, as a society, to do a thing – hang a man – which none of us individually would be prepared to do or even witness? The answer is no, not in a civilised society.”

Justice Chaskalson said in the Makhanyane case that punishment must to some extent be commensurate with the offence, but that we didn’t put out the eyes of people who had blinded others in assaults. “We have long outgrown the literal application of the biblical injunction of ‘an eye for an eye and a tooth for a tooth’.” Capital punishment was not the only way of expressing families’ and society’s moral outrage, because a long prison sentence could also do this. Retribution could not be accorded the same weight under our Constitution as the rights to life and dignity, which were the most important of all the rights. And in the long run, he added, “more lives may be saved through the inculcation of a rights culture than through the execution of murderers”. Ismail Mohamed, another concurring judge (and a later chief justice) observed that one did not rape rapists or burn the houses of arsonists.


Several of the judges referred to ubuntu, which Yvonne Mokgoro defined as “humaneness, personhood, and morality”. The state in many ways set the standard for society’s moral values. If it punished killing by killing, “it sanctions vengeance by law”. Another of the concurring judges (also a later chief justice), Pius Langa, explained that the Constitution did not allow us to “kill criminals simply to get even with them” any more than it allowed us to “kill in cold blood in order to deter others from killing”.

Justice Langa further noted that the emphasis he placed on the right to life was influenced by the history of the past decades, during which the value of life and dignity had been demeaned. Heinous crimes were the antithesis of ubuntu, but so was treatment that was “cruel, inhuman, or degrading”. Also citing ubuntu, Justice Chaskalson clarified that the state had to demonstrate respect for life and dignity in everything it did, including the way in which it punished criminals. “This is not achieved by objectifying murderers and putting them to death to serve as an example to others in the expectation that they might possibly be deterred thereby.”


In his concurring judgement, Justice Mohamed said that the death sentence ruled out “the slightest possibility” that the offender might be reformed or rehabilitated. Another concurring judge, Tholie Madala, noted that in many cases, despite the heinousness and brutality of their offence, offenders were capable of rehabilitation during long periods of imprisonment with treatment and training so that they ceased to be a danger to society. Punishment, he added, must be a means to the end of rehabilitation, and not an end in itself. Justice Madala quoted Potter Stewart, an American judge, as having said in 1972 that the death penalty was “unique in its rejection of rehabilitation of a convict as a basic purpose of criminal justice”.

The question of rehabilitation was not discussed by the Constitutional Court in any depth, nor did it address the question of how often rehabilitation was successful. Moreover, many people would agree with Lord Denning that retribution in itself is a legitimate aim of punishment. They would probably also disagree with Justice Madala’s position that rehabilitation should be the primary objective. They might indeed argue that the death sentence is justified as a form of retribution even in the absence of proof of its deterrent effect. This would not, however, overcome the objections of arbitrariness and error, both of which are dealt with in more detail below.


In the Makwanyane case, the Constitutional Court ruled that the death penalty was not a uniquely effective deterrent and could not therefore be justified. Most studies conducted in various countries have come to the same conclusion. The court also observed that the greatest deterrent to crime was the likelihood of apprehension, conviction and punishment, but that these were lacking. Although the court gave no figures, police figures show that only about 14% of the 20 most serious reported crimes result in convictions. As explained below, the proportion of reported murders that had previously resulted in executions is very much lower.

According to Justice Chaskalson, Mr von Lieres had argued that the death penalty had a greater deterrent effect than life imprisonment. He had, however, accepted that there was no proof of this. Mr von Lieres had also argued that the matter could not be proved because only those who were not deterred entered the statistics. The number who were deterred could not be known. Justice Chaskalson accepted this contention, but then went on to say that the doubt which existed about deterrence must weigh heavily against it. Quoting the chief justice of California in a 1972 case, he said: “A punishment as extreme and as irrevocable as death cannot be predicated upon speculation as to what the deterrent effect might be”.

Justice Chaskalson also dismissed the argument that the death sentence ensured that the criminal would never again commit murder. Life imprisonment would serve the same purpose. As for Mr von Lieres’s contention that there had been a substantial increase in violent crime in the five years during which the death sentence was not enforced [1989-1995], Justice Chaskalson said it remained a lawful punishment. Moreover, it had been imposed by the courts on 243 persons even though their sentences had not been implemented. A decision to terminate the moratorium could have been made at any time. In any event, he stated, bare statistics about rising crime on their own “proved nothing, other than that we are living in a violent society in which most crime goes unpunished – something we all know”.

In his concurring judgement, Acting Justice Kentridge observed that the “most impressive” of Mr von Lieres’s arguments was that the awfulness of the death penalty must in its nature deter some would-be murderers. In the face of the appalling murder rates in this country, he had said, we could not afford to relinquish any possible weapon in the fight against violent crime. This, said Justice Kentridge, was a powerful argument. However, “it relies essentially on the mere possibility that the death sentence may deter some murderers. That is not a sufficient justification for the continued existence of such an extreme punishment”.

Johann Kriegler, another of the judges, referred to the case of Furman v Georgia in 1972 in which the American federal supreme court outlawed the death penalty. One of the American judges, Thurgood Marshall, had reviewed almost every piece of Anglo-American evidence for and against capital punishment. In the course of what Justice Marshall had himself described as a “long and tedious journey”, he had made the crucial finding that 200 years of research had established “that capital punishment serves no purpose that life imprisonment could not serve equally well.” Since then a great deal more had been written in support of both the abolitionist and the retention schools, wrote Justice Kriegler, but: “No empirical study, no statistical exercise and no theoretical analysis has been able to demonstrate that capital punishment has any deterrent force greater than that of a really heavy sentence of imprisonment. That is the ineluctable conclusion to be drawn from the mass of data so thoroughly canvassed in the written and oral arguments presented to us”. It therefore “simply cannot be reasonable to sanction judicial killing”.

Justice Didcott also weighed in on deterrence. Statistics that provided a comparison between countries performing executions and those that do not, and records of countries which previously executed criminals but had stopped doing so, “have always turned out to be inconclusive”. Protagonists of capital punishment bore the burden of showing that it had a uniquely deterrent effect. “Nothing less is expected from them in any event when human lives are at stake, lives which may not continue to be destroyed on the mere possibility that some good [might] come of it. In that they have failed.”

Referring to the United States, Justice Mahomed held that there was no proof that crime was higher in states which had no death penalty than in those which did. Nor was there convincing proof that crime had risen in states after they had abolished the death penalty. He further observed that there was no empirical evidence or research in South Africa or abroad that a criminal would decide to commit an offence at the risk of receiving a long term of imprisonment but not if he risked the death sentence.

Professor Kahn wrote in 1970 that law enforcement agencies frequently used the argument that execution had a greater deterrent effect than any other form of acceptable punishment. However, he said, “investigation after investigation in country after country has failed to show that an overall deterrent result has followed”.

According to Professor van Niekerk, the report of the Royal Commission on Capital Punishment published by the British government in 1953 was “the most profound official study of every facet of the problem of the death penalty ever made anywhere in the world.” It had reached the conclusion that there was no indication that capital punishment served any substantial purpose as a deterrent to capital crime. This report had led to the provisional abolition of the death penalty in Great Britain in 1965 and its final abolition in 1969.

Professor Devenish referred to a study published by David Dolinko in an American journal of law and criminology in 1986 in which Professor Dolinko had concluded, after analysing a large body of empirical studies, that there was no scientific evidence that the application of the death penalty reduced the incidence of murder. Two years later the United Nations had concluded after a survey of research findings that there was no scientific proof that executions had a greater deterrent effect than life imprisonment. The main exception to these studies was one by Isaac Erlich published in 1975 in the American Economic Review, in which Professor Erlich argued that “each execution saves between seven and eight victims who would have been murdered by others if there were no execution”, although Professor Devenish noted that his work had been heavily criticised.

According to a report by the American Federal Bureau of Investigation (FBI) for 2014, southern states accounted for 80% of American convictions but they had the highest murder rate (5.5 per 100,000 of population). States in the northeast accounted for 1% of all executions and had the lowest murder rate (3.3). The findings were consistent with those of earlier years.

Perhaps the last word on the subject of deterrence should go to Albert Pierrepoint, a well-known British hangman. He executed more than 400 people, including ordinary and war criminals, saboteurs and spies, as well as the son of one of Winston Churchill’s wartime cabinet ministers, sentenced to death for high treason:

“If deaths were a deterrent, I might be expected to know. It is I who have faced them at the last, young lads and girls, working men, grandmothers. I have been amazed to see the courage with which they take that walk into the unknown. It did not deter them then, and it had not deterred them when they committed what they were convicted for. All the men and women whom I faced at that final moment convinced me that in what I have done I have not prevented a single murder.”

A risk worth taking

In arguing that the death sentence was not a deterrent, Justice Mahomed said it was common cause in argument that 60% to 70% of offenders who committed serious crimes were not apprehended at all and that a substantial proportion of those who were apprehended were never convicted. “The risk is therefore worth taking.” This was “not because the death penalty would, in the perception of the offender, not be imposed, but because no punishment is likely to result at all.”

The risks of being hanged appear to have diminished over time. Analysing figures from 1967 to 1968, Professor Kahn reported in his 1970 article that a murder in South Africa resulted in a hanging in only two or three cases out of 100. So even though South Africa accounted for a large proportion of the world’s executions, a murderer stood a “remote and chancy prospect of paying for his crime with his life”. About a quarter of a century later, in the period between 1990 and 1995, as we shall see below, only 0.36% of murder trials resulted in confirmed death sentences. Since no more than half of reported murders get to trial anyway, the proportion of murders resulting in death sentences is around 0.18%.


Section 8 of the 1993 (interim) Constitution stated that “every person shall have the right to equality before the law and to equal protection of the law”. Punishment, explained Justice Chaskalson, had to meet this requirement as well as the various others enumerated above. Despite what the Constitutional Court said, supporters of capital punishment might argue that murderers forfeit the rights to life and dignity. They might also argue that an “eye for an eye” is justified in cases of premeditated murder or, in American terminology, murder in the first degree. But, as we shall see below, the requirements of the equality clause (which are repeated in Section 9 of the 1996 (final) Constitution) are extremely difficult, if not impossible, to meet.

Extenuating circumstances

As long as the only sentence for capital crimes, including murder, was death, judges and juries faced the problem that they had no authority to inflict punishment short of execution. They would sometimes solve the problem by acquitting a person they knew to be guilty. To mitigate the harshness of this rule, the American state of Pennsylvania divided murder into “degrees” in 1794. First-degree murder – “wilful, deliberate, and premeditated”- would carry the death penalty, but second-degree murder would not. Some juries nevertheless refused to convict first-degree murderers because they did not want to impose a mandatory death sentence. Courts in the United States were then given discretion, and by 1963 all death-penalty states employed discretionary sentencing. Even so, as we shall see below, problems still surround the exercise of discretion by American courts.

In South Africa, the death sentence was mandatory until 1935, when courts were given the power to find “extenuating circumstances”. These were not defined in the legislation, but were described by a judge in a 1938 case as including all those facts associated with the crime which served to diminish the moral, though not the legal, guilt of the accused person. Such factors, according to a standard legal textbook, could include anything from provocation through intoxication to the influence of witchcraft. According to Professor Kahn, the number of reprieves granted to persons sentenced to death for murder dropped after it became possible for courts to avoid the death sentence where they could find extenuating circumstances. Where they could not, they were obliged to impose the death sentence. Instances arose where some judges suspected others of using (or abusing) discretion to avoid handing down death sentences. The problems surrounding the necessary exercise of discretion contribute to the wider problem of “arbitrariness” in the administration of the criminal justice system.

“Elements of arbitrariness”

One of the arguments used by Justice Chaskalson in ruling the death penalty unconstitutional was the “elements of arbitrariness” in its imposition and execution. These violate the principle of equality before the law, but are unavoidable because they are built into the system. The same objections apply to the administration of justice in general, but in the case of capital punishment any resulting miscarriages of justice are irreversible. One of the most striking things about the main judgement and the concurring judgements of the Constitutional Court in the Makhanyane case was the frank admissions of arbitrariness, and therefore unfairness, made by many of the judges. In the words of Justice Chaskalson,

“At every stage of the process there is an element of chance. The outcome is dependent on factors such as the way the cases are investigated by the police, the way the cases are presented by the prosecutor, how effectively the accused is defended, the personality and particular attitude to capital punishment of the trial judge and, if the matter goes on appeal, the particular judges who are selected to hear the case.”

In South Africa, as we shall see in more detail below, torture by the police to obtain confessions from accused persons or incriminating statements from witnesses is an ever-present risk. Most accused in capital cases are poor and therefore dependent for their defence on usually inexperienced pro deo counsel, acting without the benefit of instructing attorneys. Even if some accused might be able to pay for their own defence, they are unlikely to be able to afford the best counsel and may have to satisfy themselves with a junior. By contrast, a richer person would be able to engage both senior and junior counsel, as well as the best firms of attorneys. They would further be able to pay for expert witnesses and for private investigations and research. The intolerable result is that a poor person might be convicted of a crime of which a richer person with a better defence might be acquitted.

Another disadvantage faced by poor persons is that pro deo counsel assigned to them might have to communicate with them through an interpreter, with the risk of misunderstandings that might not otherwise apply. Language barriers might also affect communication with the judge. This problem was identified by Agnes Winifred Hoernlé, who served as president of the IRR between 1948 and 1950 and was a member of the Lansdown Commission set up to enquire into the death penalty in the early post-war years. Dr Hoernlé noted that there was much more room for mistakes when there was “a barrier of language between judge and accused and when the workings of the human mind are hidden by great differences in social background and modes of thought”.

The court said that of the thousands of persons tried for murder each year only a very small percentage were sentenced to death. Of those a large number escaped that penalty on appeal. Citing figures provided by the police, Justice Chaskalson explained that since the amendment to section 277 of the Criminal Procedure Act in 1990 giving judges greater discretion, more than 40,000 murder cases had been brought to trial, in which 243 people had been sentenced to death, but of whom only 143 had had their sentences confirmed by the Appellate Division (AD) of the Supreme Court (now the Supreme Court of Appeal).

These figures mean that only 0.36% of murder trials resulted in death sentences which the AD confirmed. Since only about half the murders reported to the police get to trial, that minuscule percentage would be halved if calculated as a proportion of murders committed. Justice Chaskalson quoted six of the seven judges in a Massachusetts case in 1980 as having said, “All murderers are extreme offenders. Fine distinctions, designed to select a very few from the many, are inescapably capricious when applied to murders and murderers.”

The very fact that so few persons sentenced to death are actually executed is no doubt evidence of a humane desire on the part of judges to spare all but a few convicted murderers the death sentence. It nevertheless exacerbates the unfairness suffered by those who are finally hanged. Moreover, according to Justice Chaskalson, “It cannot be gainsaid that poverty, race and chance play roles in the outcome of capital cases and in the final decision as to who should live and who should die.”

In his concurring judgement, Laurie Ackermann quoted an American judge, Harry Blackmun, who had observed in a 1944 case that

“Experience has taught us that the constitutional goal of the elimination of arbitrariness and discrimination from the administration of death can never be achieved without compromising an equally essential component of fundamental fairness – individualised sentencing... If the death penalty cannot be administered consistently and rationally, it must not be administered at all.”

Justice Ackermann himself wrote, “It is virtually impossible except in the case of rigidly circumscribed mandatory sentences – which present other dangers – to avoid elements of arbitrariness.” The process of weighing up mitigating and aggravating factors left such a wide latitude for differences of individual assessment, evaluation, and normative judgement that they were “inescapably arbitrary to a marked degree”. For one person to receive the death sentence where a similarly placed person did not, was cruel to the person receiving it. The imposition of the death penalty was therefore “inevitably arbitrary and unequal”.

Although judges are enjoined to be objective and dispassionate, their own beliefs are bound to have some impact on whether they impose the death sentence or not. Some may be more successful in excluding their own beliefs than others. Some may be more susceptible than others to arguments in mitigation of sentence that the convicted person is genuinely remorseful and capable of rehabilitation. Some may be taken in by successful acting by accused persons supposedly demonstrating remorse, while others may see through it. Some judges may be more susceptible than others to arguments about insanity. Some will be more inclined than others to believe expert evidence by psychiatrists and psychologists. Judges of a liberal bent might be more inclined to take cognisance of factors such as poverty and family breakdown, and to be persuaded by arguments that a particular criminal is a victim of circumstance and therefore that his or her moral guilt is diminished. Those of a more conservative inclination are less likely to believe that poverty or other factors can be regarded as excuses.

Provision for judges to find extenuating circumstances warranting a sentence lesser than death poses problems of its own. This is because of the discretion it necessarily allows, with the result that some judges may allow their personal viewpoints to influence whether they find such circumstances. Although he was among the judges personally opposed to the death sentence, Ramon Leon was attacked for nevertheless imposing it on an unrepentant member of Umkhonto we Sizwe for the murder of two women and three children in 1985 when he was unable to find extenuating circumstances. Judge Leon himself commented, “I know from my own experience that some judges find extenuating circumstances more easily than others. I know judges that impose the death penalty not infrequently, and I know one judge who has been on the Bench for some years who has never passed the death sentence. Should a man’s life depend on the chance of the judge before him?”

Another South African judge, David Curlewis, who himself supported the death penalty, stated that judges who did not impose it when they should do so were not doing their duty. Men’s lives depended on who sat in judgement. Yet another judge, one Justice Snyman, believed execution should be retained as an important deterrent, but he personally dreaded pronouncing sentence of death and was always relieved to find extenuating circumstances.

“Execution,” wrote Professor Kahn, “becomes almost a lottery.”

“Elements of arbitrariness” do not stop with the courts. Governments have the power to issue pardons, exercise clemency, and commute death sentences to imprisonment. Professor van Niekerk pointed out that these decisions might be made by senior officials who then passed on recommendations to the relevant minister. Professor Kahn observed that reasons for reprieves were seldom given. It was therefore impossible to determine whether or not there is consistency in the exercise of executive clemency.

It is also difficult to exclude political considerations. A person might be reprieved or have his death sentence commuted because he has the right political connections, but this may also count against him, as it did when the British government in 1945 declined to stop the execution of the son of a prominent former cabinet minister on the grounds that this would invite accusations of political favouritism.

Discrepancies over time in the commuting of sentences in South Africa suggest that the personal views of ministers and officials, or prevailing opinion in judicial, legal or political circles about capital punishment might have influenced such decisions. At one stage, according to Professor van Niekerk, 61% of death sentences for murder were commuted, a proportion which then dropped to 21%. Professor Kahn further pointed out that the commutation rate for women was 93%. This high proportion suggests that the governor general, or whoever was advising him, was more inclined to take pity on women than on men facing the death penalty.

The unavoidable conclusion is that the whole process from first appearance in court to actual execution is arbitrary, which means that it violates the constitutional principle of equality before the law and the rights of all persons to equal protection of the law. Even if the Constitution contained no such provisions, the arbitrariness surrounding the death penalty violates the basic human instinct and sense of natural justice that people should be treated fairly, which includes the view that like crimes should attract like punishments.

According to Justice Didcott, American courts had constantly wrestled with the arbitrariness problem but “by no means to their satisfaction”. He continued, “For such arbitrariness is largely inherent in the nature of the proceedings from start to finish. Similar trouble may be inescapable, to be sure, in cases that are not capital ones. But in those producing sentences of death the arbitrariness is intolerable because of the irreversibility of the punishment. Mistakes discovered afterwards could not be rectified. Notwithstanding “the myth to the contrary”, mistakes “do occur now and then”.

Perhaps the most comprehensive account of the conflict between the death penalty and the right to equality was that written in the Makhanyane case by Justice Mahomed.

“I have no doubt whatever that judges seek conscientiously and sedulously to avoid any impermissible unequal treatment between different accused whom they are required to sentence, but there is an inherent risk of arbitrariness in the process, which makes it impossible to determine and predict which accused person guilty of a capital offence will escape the death penalty and which will not.

“The fault is not of the sentencing court, but in the process itself. The ultimate result depends not on the predictable application of objective criteria but on a vast network of variable factors which include the poverty or affluence of the accused and his ability to afford experienced and skilful counsel and expert testimony; his resources in pursuing potential avenues of investigation, tracing and procuring witnesses, and establishing facts relevant to his defence and credibility; the temperament and sometimes unarticulated but perfectly bona fide values of the sentencing officer and their impact on the weight to be attached to mitigating and aggravating factors; the adequacy of resources which compels the pro deo system to depend substantially on the services of mostly very conscientious but inexperienced and relatively junior counsel; the levels of literacy and communication skills of the different accused in effectively transmitting to counsel the nuances of fact and inference often vital to the probabilities; the level of training and linguistic facilities of busy interpreters; the environmental milieu of the accused and the difference between that and the comparative environment of those who defend, prosecute, and judge him; class, race, gender, and age differences which influence bona fide perceptions relevant to the determination of the ultimate sentence; the energy, skill, and intensity of police investigations in the particular case; and the forensic skills and experience of counsel for the prosecution. There are many other such factors which influence the result and which determine who gets executed and who survives. The result is not susceptible to objective prediction. Some measure of arbitrariness is inherent in the process.

“This truth has caused [Harry]Blackmun, one of the most experienced judges of the United States supreme court, finally to conclude that it ‘is virtually self- evident to me now that no combination of procedural rules and substantial regulations ever can save the death penalty from its inherent constitutional deficiencies. The basic question – does the system accurately and consistently determine which defendants “deserve” to die – cannot be answered in the affirmative’.”

Justice Chaskalson pointed out that some of the “imperfections inherent in criminal trials” could be mitigated by allowing convicted persons to appeal to a higher court. However, he noted, “appeals are decided on the record of the case and on findings made by the trial court. If the evidence on record and the findings made have been influenced by these factors, there may be nothing that can be done about that on appeal”. And, of course, wealthier convicted persons will have more resources to pursue appeals all the way to the Constitutional Court.

Political factors

In terms of section 84(2)(j) of the Constitution, the president is responsible for “pardoning or reprieving offenders and remitting any fines, penalties, or forfeitures”. Unlike courts of law, the president is not required to explain his reasons. This heightens the risk that political, factional, ideological or personal considerations will influence decisions. Such factors are widely believed to play a role in the granting of parole to, or the withholding of parole from, prisoners by the relevant minister. This serves to discredit the parole system. If the death penalty were to be reintroduced, similar abuses cannot be ruled out in decisions by the president and those advising him on whether or not to implement particular death sentences. This added risk of abuse that results in differential treatment and therefore injustice is yet another argument against the death penalty.

The American position

As already explained, judges in the Makwanyane case frequently cited the decisions of American courts on capital punishment. The problem of arbitrariness and inconsistency was the main reason why the Supreme Court of the United States (the federal supreme court) declared the imposition of the death penalty to be cruel and unusual, and therefore unconstitutional, in the case of Furman v Georgia in 1972. Following the decision, 37 American states sought to overcome the arbitrariness objection by enacting laws to guide the discretion of courts in imposing capital sentences.

In Gregg v Georgia in 1976, the same court assessed whether the guidelines introduced by various states had met its requirements. These were that discretion in imposing the death sentence should be exercised according to objective criteria, but that the sentencer (whether judge or jury) should also take into account the position of the individual accused. Where states had failed to fulfil the court’s requirements, capital punishment remained unlawful. Where they had succeeded in satisfying the requirements, the death penalty could once again be lawfully imposed. Executions were resumed in 1977. Since then 1,436 people have been executed (although last year Pennsylvania became the 18th American state to have abolished the death penalty for all crimes).

Like that of South Africa, the American constitution outlaws “cruel and unusual punishments”. But, unlike South Africa, it also expressly permits capital punishment. Whereas the United States court found that clear guidelines could overcome the problem of arbitrariness, the South African court declared the problem to be insoluble.

The United States court also found that retribution could be consistent with human dignity on the grounds that society believed that “certain crimes are themselves so grievous an affront to humanity that the only adequate response may be the penalty of death”. Although the court found that there was “no convincing empirical evidence” either way on deterrence, it could not discount the possibility that for certain “carefully contemplated murders” a possible death sentence “may well enter into the cold calculus that precedes the decision to act”.

Torture, forensics, and plea bargains

The judgement in the Makwanyane case made no explicit reference to torture, but it appears to be widespread. Confessions extracted under torture, as well as evidence obtained through torture, are normally legally inadmissible but there is no guarantee that they cannot slip through, resulting in miscarriages of justice and wrongful convictions leading to wrongful executions.

In 2014/2015 some 3,856 complaints of assault or torture were lodged with the Independent Police Investigating Directorate (IPID) against the police. In that same year, there were 19 criminal convictions of police officers for torture and/or assault. Peter Jordi, a practising attorney attached to the Law Clinic at the University of the Witwatersrand (Wits), told a meeting at the IRR in 2012 that the police used torture on a massive scale. Carolyn Raphaely of the Wits Justice Project reported Professor Jordi as having said: “The police torture people all the time – in their homes, in police cells, in the veld, in cars… Torture is standard police investigation practice. These policemen are serial criminals. They have methods of investigation that are unlawful and for which they could be prosecuted but they never are. Police torture is a daily occurrence where I practise.”

Following the shootings at Marikana in August 2012, IPID stated that at least 194 charges had been laid by Marikana survivors who were allegedly assaulted and tortured while in police custody. After some of the striking miners had been released from police custody, the New Age newspaper reported that many claimed to have been tortured by the police to extract confessions.

Allegations are of course only allegations. The small number of convictions in relation to the large number of complaints could be evidence that there are numerous complaints that have no foundation. It could also, however, indicate lack of zeal on the part of police and prosecutors in pursuing complaints, and the difficulty complainants face in producing evidence, especially if they themselves are still in police custody. It is also possible that torture occurs on a far greater scale than suggested by the number of complaints about it that are formally lodged with IPID, since, despite its name and claimed independence of the police, IPID still falls under the minister of police.

In 2011 and 2015 the IRR published two reports, themselves based on numerous press reports, showing that the police were involved in serious and violent crime including murder, armed robbery, rape, burglary, theft and torture on a substantial scale. The IRR reports also found that significant numbers of police officers remained in the employ of the police even when they had been convicted of serious crimes. The conclusion the IRR reached was that police criminality was not a case of “isolated incidents”, but a “pattern of criminal behaviour”. Policemen who are themselves capable of committing the most violent crimes are unlikely to be too concerned about the use of torture or assault to extract confessions and/or other evidence.

All of the above suggests that there is substantial risk that the police will act unlawfully in securing evidence for prosecutions that may result in wrongful convictions and executions. Wrongful convictions and sentences may also arise from police practices that fall short of assault and torture and which may be motivated by nothing more malicious than the desire to demonstrate success in apprehending and convicting criminals.

Although no similar study is known to have been conducted in South Africa, the American Department of Justice and the FBI formally acknowledged in 2015 that nearly every examiner in an FBI forensic squad overstated forensic hair matches for two decades before 2000. A high total of 26 of 28 examiners testifying to hair matches in 268 trials in which 32 death sentences were imposed were found to have overstated the matches and 95% of overstatements favoured the prosecution.

Another American study found that half the cases in which people were wrongly sentenced to death hinged on false testimony by informants, typically criminals rewarded with lighter punishments. Many convictions were reached through plea bargains in which the defendant agreed to plead guilty in return for leniency. But many depended on co-operating witnesses who, in exchange for lighter sentences, often told prosecutors what they wanted to hear.

“The possibility of error”

Justice Chaskalson referred in the Makwanyane case to the “possibility of error”, which could not be excluded. Justice Kentridge added that error could neither be wholly excluded, “nor, of course, repaired.” Professor van Niekerk quoted a judge by the name of J.D. Cloete as having said that it had sometimes been on his conscience that he had sentenced people to death, but that he had been obliged to do so because the law required it. Even the most experienced judges could make grave mistakes in imposing capital punishment, Judge Cloete added.

Professor van Niekerk reported that 78% of the 158 advocates in his survey suggested that there was no absolute guarantee against an innocent person being hanged. Professor Kahn referred to a number of cases in South West Africa (later Namibia) in which people had been executed, subsequent to which the supposed victim had reappeared. He also quoted (with considerable scepticism, however) a South African judge who had written in 1939 that 10% of the Africans convicted of murder in South Africa were innocent, but that they had been convicted because they did not understand procedure. Some of those convicted may have been guilty of manslaughter, but they were not guilty of murder.

The Reverend Dr Henri P. Junod, director of the Penal Reform League established by the IRR in 1939, was convinced mistakes had been made. Having accompanied 800 black people to the gallows in South Africa, Dr Junod, who was later an honorary life member of the IRR, reported:

“We have had cases where, in front of the gallows, we have known that the first crown [state] witness was the one who should have been where the convicted person was, facing death. We could naturally not prove it but the situation was made perfectly clear by long weeks and months with the accused, and by continual denials up to the last moment, when nothing more was to be gained by adhering to a false story.”

Newly available DNA evidence has permitted the exoneration and release of more than 17 death-row inmates since 1992 in the United States. However, such evidence is available in only a fraction of capital cases. A law professor in Michigan conducted a study in which he determined that at least 4% of people on death row were, and are, likely innocent. Professor Devenish cited a report by two American jurists in the Stanford Law Review in 1987 to the effect that 343 people had been wrongly convicted of capital crimes since 1900, and that 25 of these had been executed.

According to a more recent report by the Death Penalty Information Centre in Washington D.C., between 1973 and 1999 an average of three persons sentenced to death have been exonerated each year with evidence of their innocence becoming available. Between 2000 and 2011, there have been five exonerations a year. Since 1973 a total of 156 exonerations have been granted.

Sometimes miscarriages of justice might never be exposed, however. For example, in the state of Virginia in 1988, DNA evidence that might have exonerated a person executed the previous year was destroyed after the prosecution claimed that “it would be shouted from the rooftops that Virginia executed an innocent man.” It can hardly be gainsaid that those involved in administering the criminal justice system have an interest in covering up rather than admitting their mistakes. Where mistakes are discovered, it may often be thanks to the efforts of those outside the criminal justice system. The pardon posthumously granted to Timothy Evans arose in part from the work of a journalist, Ludovic Kennedy, who suggested that the real murderer was John Christie.

As noted above, Justice Chaskalson acknowledged that “the possibility of error will be present in any system of justice”. “Unjust imprisonment is a great wrong, but if it is discovered, the prisoner can be released and compensated. But the killing of an innocent person is irremediable.” It is also worth noting that most of the wrongful executions listed above took place in countries known to adhere to procedures designed to ensure fair trials.

Happy endings

There have recently been a few cases in South Africa where people wrongly convicted of murder were subsequently released after spending many years in prison for crimes they did not commit. One such person was Thembekile Molaudzi, who spent 11 years in prison for the murder of a policeman after being convicted on the basis of false statements made out of court by other accused which his defence did not challenge. After a long struggle through various courts, Mr Molaudzi was eventually released in 2015 on the orders of the Constitutional Court, but this might never have happened without the help of a prison warder who believed his claim to be innocent and put him in touch with the Wits Justice Project.

A few months before Mr Molaudzi’s release, Boswell Mhlongo and Alfred Nkosi were also released, also having been wrongly convicted of the same crime. Mr Mhlongo had the help of a woman he met via online dating from his prison cell whose own father had once been wrongfully arrested and who helped take up his case (and whom he later married). He obtained additional assistance from a fellow inmate who was a law student. Mr Mhlongo tried three times to commit suicide. On the third attempt, when he was in a coma, doctors wanted to switch off the machines but his family refused, and he eventually recovered. He said: “I felt bad when I woke up. I’d spent five years unsuccessfully trying to get my trial transcripts which I needed to appeal my case and could see no reason to live.”

Carolyn Raphaely of the Wits Justice Project commented that neither the Department of Justice and Correctional Services nor the National Prosecuting Authority maintained statistics on wrongful convictions, but that anecdotal evidence suggested there were many similar people, particularly the indigent, behind bars. Similar to Mr Mhlongo, she wrote, they “mostly accept their fate”.

Referring to the wrongful convictions, Rudolph Jansen, chairman of the pro bono committee of the Pretoria Society of Advocates, which appointed advocate Donrich Jordaan to assist in the case, observed that: “Ultimately the judge and the three judges on the full bench of appeal [of the High Court] must take some responsibility”.

Fusi Mofokeng and Tshokolo Mokoena were released in 2011, having been incarcerated in 1992, first as awaiting-trial prisoners and then to serve sentences of life imprisonment for murders carried out by an armed unit of the African National Congress. Members of the unit were exonerated in 1998 by the Truth and Reconciliation Commission (TRC), but Messrs Mofokeng and Mokoena declined to apologise and tell the “truth” for a crime in which they had not been involved. Thanks to the Wits Justice Project and a petitions committee of the National Council of Provinces, they were released because they were in any event eligible for parole, although lawyers are still attempting to obtain presidential pardons for them.

“Happy endings” is perhaps not the best phrase to apply to this handful of cases. These men spent substantial parts of their adult lives in prison for crimes they did not commit. Two of them refused an opportunity to secure their release if they had been required to make false confessions to the TRC. Ms Raphaely is no doubt correct in her view that there must be many other innocent people serving long sentences but who are not lucky enough to obtain the kind of assistance that the Wits Law Project, organisations such as Lawyers for Human Rights, and Mr Jordaan were able to provide to others.

Millions of South Africans feel great anger and outrage at all types of crime, especially violent crime, committed against innocent people. Many live fear-ridden lives, the minority who can afford it investing in private security. Amidst the grief that many suffer, there is no doubt enormous frustration and even desperation that crime continues at such high levels.



Although the Constitutional Court did not itself draw the distinction, its findings against the death penalty may be divided into five categories.

The first was that it is indeed a cruel and unusual punishment which violates the rights to life and dignity. Notwithstanding what the court said, however, views as to what is “cruel and unusual” no doubt differ from person to person, as well as from judge to judge. They can also change, as they have throughout the history of crime and punishment. What one community at one time may regard as permissible forms of punishment, others at other times may regard as too lenient: hence the argument sometimes heard that the rights of victims should not outweigh those of suspected criminals.

The court correctly observed that constitutional rights should transcend the vagaries of public opinion, but that does not mean that the views of judges as to what qualifies as cruel and unusual punishment are themselves not influenced by their personal opinions and experiences. In fact, the various concurring judgements in the Makwanyane case revealed differences of opinion as to the weight to be accorded to the various rights being discussed. Moreover, the American federal supreme court, many of whose judges in the two key death penalty cases in the 1970s clearly shared the liberal philosophy of our own Constitutional Court in the Makwanyane case, found that the death penalty was not automatically cruel and unusual.

The second finding was that the death sentence is not a deterrent. Although the court did not state this, the implication of what some of the judges said is that capital punishment is almost irrelevant as a deterrent since apprehension and conviction are lacking and most crime goes unpunished in any event. Yet the very fact that people are concerned and debate and fear a death sentence shows that it might well be a deterrent. There is also the fear that potential victims have in reporting crime because they are afraid of being killed. We also cannot know to what extent even the remote possibility of a punishment as severe as the death penalty may weigh on the criminal mind. This is particularly true when it comes to the perpetration of the most cruel and gratuitously violent premeditated crimes – those crimes most likely to attract such a sentence.

In addition, and despite what the courts have observed, society’s desire for retribution against criminals may override doubts about deterrence. Some communities may also regard the possibility of retribution as more important than the possibility of deterrence or rehabilitation. We must remember that in South Africa the death sentence does apply in practice in many poor and crime-ridden communities where vigilantes have effectively privatised the sentence in answering the community demand for retribution. Reintroducing the death sentence could well reduce the extent of extra-judicial killings. We may further take the view that planned and premeditated murder is so monstrous a crime that the only appropriate punishment is indeed “an eye for an eye and a tooth for a tooth”. As with “cruel and unusual”, these are often matters of personal opinion, and the argument that the death penalty has more to do with retribution than rehabilitation is something that should not, in and of itself, take the death penalty off the table for South Africa.

The fourth finding was about the “arbitrariness” with which the death penalty may be applied and that this may violate the principle of equality before the law. As the judges admitted, our criminal justice system is flawed at its very apex. Judges who are themselves flawed can reduce, but never eliminate, the flaws that characterise the operation of the system from start to finish, from arrest to implementation of sentence. Injustice would arise if judges were denied all discretion in administering punishment to ensure that it fits the crime, but the risk of injustice arises from the very discretion that judges require to administer justice. Yet this is a risk that attaches to all decisions taken by judges and magistrates. It is the reason why they need to be properly qualified and trained and why the most rigorous appeal processes should apply in capital punishment cases.

This brings us to the fifth finding – the possibility of error. The number of innocent people executed will never be known, but it is almost certainly much higher than the figures quoted in this paper. The factors that lead to such miscarriages of justice are all present in the South African judicial system. Technological advances, including DNA testing, may help to reduce the risk of error, as would better police and prosecutorial procedures.

The most rigorous, time-consuming and costly automatic appeal processes could be introduced to lessen the risk. These may also deter prosecutors from seeking the death penalty in all but the most extreme cases. Defendants in such cases could be provided with special resources to mount their defence – far beyond those made available in non-death penalty cases. Even where these precautions are taken, the risk of error might be lessened but could never be completely avoided. It is this fifth objection that is the most difficult to overcome in any argument about whether South Africa should reintroduce the death penalty – and the issue around which any debate about reintroducing the death penalty should centre.

Liberals and conservatives can differ about whether capital punishment is indeed cruel and unusual. They can differ about how much emphasis should be placed on retribution as opposed to rehabilitation. They may also differ as to whether the problem of “arbitrariness” should be overlooked in the cause of punishing crime. Nonetheless, the overriding objection to the death penalty transcends political views and philosophical beliefs about crime and punishment. In the words of Gerald Gardiner, lord chancellor of the United Kingdom from 1964 to 1970: “Human beings who are not infallible ought not to choose a form of punishment which is irreparable.”

Main sources

= The state versus T Makwanyane and M Mchunu (Case No CCT3/94)

= Ellison Kahn: The Death Penalty in South Africa, Journal of Contemporary Roman-Dutch Law, 1970

= George Devenish: The Historical and Jurisprudential Evolution and Background to the Abolition of the Death Penalty in South Africa and its Relationship to Constitutional and Political Reform, South African Journal of Criminal Justice, 1992

= BvD van Niekerk: Hanged by the Neck Until You are Dead

– Some Thoughts on the Application of the Death Penalty in South Africa, South African Law Journal, 1969

= South African Institute of Race Relations: South Africa Survey 2016, Johannesburg, 2016.

= South African Institute of Race Relations: Broken Blue Line, Johannesburg, 2015

= Chris Waldburger: An Indictment of Our Country – An Innocent Man’s 11-year Fight for Justice, Leadership, undated

= Carolyn Raphaely: The Human Cost of Judicial Error,

The Star, 31 March 2016

= Andrew Taylor: “Who killed murder?, The Spectator, 19 March 2016

= Rob McCafferty: Murder in South Africa – A Comparison of Past and Present, first edition, United Christian Action, June 2003

= Ontario Consultants on Religious Tolerance: Executing Innocent People – The Executions of Five Probably Innocent Inmates, printed 23 May 2016

= Wikipedia: Wrongful Execution, printed 23 May 2016; Furman v Georgia, Gregg v Georgia, printed 5 July 2016

= American Civil Liberties Union: The Case Against the Death Penalty, revised edition, 2012, printed 23 May 2016

= Death Penalty Information Centre: Facts About the Death Penalty, 12 May 2016

= Amnesty International: Death Sentences and Executions in 2015, printed 12 July 2016

= Exchanges of emails with Carolyn Raphaely, September 2015, June 2016, July 2016

= Various press reports