Phillip Burring
  BA LLB PG Dip in Bus Man LLM

  Managay Reddi
  Associate Professor and Dean and Head, School of Law, University of KwaZulu-Natal

 46 Volume 4 2013 pp 928
  Download Article in PDF


Artikel 49, Dodelike Geweld en Lesse uit die De Menezes Skietvoorval in die Verenigde Koninkryk

Die Strafproseswysigingswetsontwerp 2010 stel sekere wysigings aan artikel 49 van die Strafproseswet 51 van 1977 voor ten einde die artikel in lyn te bring met die riglyne wat vasgestel is deur die Konstitusionele Hof in Ex parte: The Minister of Safety and Security: In re S v Walters 2002 2 SASV 105 (KH) rakende die gebruik van geweld ten einde ’n inhegtenisneming uit te voer. Hierdie artikel poog om die aandag te vestig op die tekortkominge van die voorgestelde wysigings aan artikel 49, deur gebruik te maak van die De Menezes-voorval wat plaasgevind het in die Verenigde Koninkryk, as ’n voorbeeld om die potensiële probleme met die bewoording van die wysigingswetsontwerp toe te lig. Hierdie voorval, waarin ’n onskuldige man deur die Britse polisie doodgeskiet is, gee aanleiding tot ’n aantal interessante vraagstukke met betrekking tot die vlak van geweld wat deur die polisie gebruik mag word om ’n inhegtenisneming te bewerkstellig, en die impak wat terrorisme - spesifiek selfmoordterrorisme - internationaal gehad het op polisie taktiek. In lyn met die Britse ondervinding, sal dit betoog word dat die voorgestelde wysigings aan artikel 49 nie ’n aanvaarbare oplossing in Suid-Afrika bied vir ‘n soortgelyke terroristebedreiging nie. Hierdie argument is gebaseer op die voorgestelde teks in artikel 49(2)(b) - naamlik dat “the suspect is suspected” (“die verdagte word vermoed”) mag lei tot onskuldige dodings, aangesien die woorde die potensiaal het om die beluitneming, rakende die gebruik van dodelike geweld, van die arresteerder na die verwyderde “supervisor” (“inspekteur”) oor te dra. Verder het die insluiting van die teks, “having committed a crime” (“wat 'n misdryf gepleeg het”), en die uitsluiting van die frase, “future death or grievous bodily harm” (“toekomstige doding of ernstige liggaamlike besering”), die gevolg dat die voorgestelde artikel 49 slegs bruikbaar sal wees in voorvalle waar die verdagte ook ’n verdagte is in ’n vorige misdaad, en sal dit nie van hulp wees vir die arresteerder in gevalle waar die verdagte moontlik ’n misdaad in die toekoms beplan nie.

1 Introduction

The Criminal Procedure Amendment Bill 20101 (the Bill) has the stated objective of aligning section 49 of the Criminal Procedure Act2 (CPA), with the criteria laid down by the Constitutional Court in Ex parte: The Minister of Safety and Security: In re S v Walters.3 More specifically, the Bill seeks to amend section 49(2) by aligning the criteria on the use of lethal force in affecting an arrest, with those tabulated by the Constitutional Court in the Walters case.

The present wording of the current section 49 was introduced in the CPA by the Judicial Matters Second Amendment Act,4 since it was considered at the time (and judiciously so given the 2002 Walters case) that the original provisions of the Act would not pass constitutional muster. Following a series of submissions from civil society, the new wording only came into effect in 2003 - a five-year delay. The primary reason, however, for the delay in the implementation of the 1998 amendment was the uncertainty within the South African Police Services about the interpretation and application of the amended section 49.5

The Court in Walters laid down a set of nine principles to govern the use of force in affecting an arrest.6 Unfortunately, and as mentioned above, the Walters decision was handed down too late to enable the current section 49 (which came into effect in 2003) to be amended in line with these principles. It is for this reason that the current Bill has been tabled before Parliament which, at the time of passing the 1998 Judicial Matters Second Amendment Act, had not had the benefit of the authoritative guidelines furnished by the Constitutional Court.7

The provisions of the proposed amendment to section 49 of the CPA will be discussed in this article, in conjunction with an analysis of the Stockwell One Report by the Independent Police Complaints Commission (IPCC) in the United Kingdom . This report c oncerned an Investigation into the shooting of Jean Charles de Menezes at Stockwell underground station on 22 July 2005.8 The aim of this article is to draw attention to short-comings in the proposed amendment to section 49, by using the De Menezes case as a proxy to illustrate potential difficulties associated with the wording of the amendment. This incident, in which an innocent man (De Menezes) was shot dead by the British police, raised some interesting issues pertaining to the level of force that may be used by police in effecting an arrest, and the impact that terrorism - especially suicide terrorism - has had internationally on police tactics.

Aligned to the British experience, it will be argued that the proposed amendment to section 49 does not provide a satisfactory solution to a possible similar terrorist threat in South Africa. This argument is based on the submission that the proposed text in section 49(2)(b) - namely that “the suspect is suspected”9 may lead to innocent deaths, as these words have the potential to transfer decision-making on the use of lethal force from the arrestor to a distant supervisor (as occurred in the de Menezes incident). Furthermore, the proposed inclusion of the expression “having committed a crime”10 and the exclusion of the phrase “future death or grievous bodily harm”,11 render the proposed section 49 useful only in incidents where the arrestee is a suspect in a past crime, and will not assist the arrestor in those instances where the arrestee may be planning a future crime.

2 Section 49, the Law of Arrest, the Purpose and Means of Arrest, and Proposed Amendments

2 1 Section 49 and the Impact of Human Rights Jurisprudence on the Law of Arrest

The disregard for human rights under apartheid led to many well-documented instances of police brutality and killings. It is primarily against this backdrop that a number of fundamental human rights were enshrined in the South African Constitution,12 in particular the rights to human dignity13 and to life. 14

It is trite law that no right is limitless. This is reflected in section 36 of the South African Constitution, which provides for the instances in which rights may be legitimately limited. It is submitted, for example, that any law that permits the killing of another during an act of self-defence would pass constitutional muster, because the attacker’s right to life does not enjoy primacy over the victim’s right to life; in other words, the attacker’s right to life is not limitless.

A similar “balancing of rights” dilemma arises when an arrest is being attempted; the issue here being under what circumstances can deadly force be used to effect an arrest? Do the suspect’s rights take precedence over the rights and interests of society? In the landmark case of Govender v Minister of Safety and Security,15 Olivier JA commented that the state has a duty to preserve the criminal justice system’s effectiveness as a deterrent to crime. According to Olivier JA, a failure by the state to preserve the effectiveness of the criminal justice system will end in lawlessness and a loss of legitimacy of the state itself.16 This suggests that society’s interest in bringing a suspect to justice outweighs the suspect’s rights. However, does this mean that the suspect can be killed during an attempt to arrest him or her?

The court in the Walters case answered this question in the negative, when it stated that the Constitution demands respect for the life, dignity and physical integrity of every individual, and that normally this respect outweighs the disadvantage to the administration of justice in allowing a criminal to escape.17 In reaching this decision, the Constitutional Court in Walters relied heavily on the defining United States decision in

Tennessee v Garner,18 where White J, writing for the majority, found it “constitutionally unreasonable” to use “deadly force to prevent the escape of all felony suspects, whatever the circumstances”.

The Court in Walters also found that the value placed on the rights to life and dignity is important when weighing them against the competing societal interest in promoting the efficient combating of crime. This stance resonates with the observation of O’Regan and Cameron JJ, in their minority judgment in S v Manamela:19

The level of justification required to warrant a limitation upon a right depends on the extent of the limitation. The more invasive the infringement, the more powerful the justification must be.20

Therefore, evidently, the courts’ view is that whilst killing a suspect during an attempted arrest should not be condoned lightly, there are circumstances under which it may be acceptable. This would notably be where the arrestor or a third party is being threatened with death or serious physical harm by the suspect.

2 2 The Purpose and Means of Arrest

In discussing the purpose of arrest, the Constitutional Court in Walters maintained that arrest is not an objective in itself; rather, it is one of various ways to ensure that a suspected criminal appears before court. Therefore, the purpose of arrest is to take the suspect into custody - to be brought before court as soon as possible on a criminal charge. 21

The Constitutional Court in S v Makwanyane22 discussed the object of criminal punishment, in particular as it related to the death penalty. In summary, the court held that punishment should have a deterrent effect, a preventative effect and an element of retribution. The court stated further that “the greatest deterrent to crime is the likelihood that offenders will be apprehended, convicted and punished”,23 and that while the death sentence was an effective preventative measure (as the offender would now be dead), imprisonment would achieve the same goal.24 The court was also of the view that, of the three elements, retribution was the least important of the objectives of punishment. It therefore found that the death penalty was not the only way of expressing society’s outrage at the crime, as a long prison sentence would also achieve this goal.25 In so finding, the Constitutional Court held that the death penalty was not a necessary tool in the fight against crime, and concluded that it was unconstitutional.

The court in the Walters case stated that “arrest may never be used to punish a suspect”26, but nevertheless did not rule out the use of force, and indeed deadly force in certain circumstances to achieve the goal of arrest. This creates an anomaly: how is it possible that it is unacceptable to kill a person who has been tried and found guilty of a serious crime against society, but yet it is acceptable to kill a person who, at the time of the attempted arrest, is still only a suspect in a criminal matter, still innocent until proven guilty, and still entitled to a fair trial to prove his guilt beyond a reasonable doubt? The courts have answered this question by requiring an actual threat of violence by the suspect, or a suspicion on reasonable grounds that the suspect has committed a crime involving the infliction or threatened infliction of serious bodily harm - to justify the use of lethal force during an arrest.27

The Constitutional Court in Makwanyana, however, in holding that the death penalty is unconstitutional, laid out a cogent argument that could equally be applied to the use of lethal force in effecting an arrest, when Langa J said the following:

A culture of respect for human life and dignity, based on the values reflected in the Constitution, has to be engendered, and the State must take the lead in displaying these values. In fulfilling this role, the State not only preaches respect for the law and that the killing must stop, but it demonstrates in the best way possible, by example, society’s own regard for human life and dignity by refusing to destroy that of the criminal. Those who are inclined to kill need to be told why it is wrong. The reason surely must be the principle that the value of human life is inestimable, and it is a value which the State must uphold by example as well. 28

The argument follows, that in order to set an example to the citizens of South Africa, the police cannot be seen to be indiscriminately killing people in the process of arresting them. To do so would breach the duty that the state has to engender respect for the law. This would also lead to the situation where a public that legitimately grows frightened of the police service, will be more inclined to resist arrest, which will then in turn potentially lead to an increase in the use of lethal force by the police, and so the cycle will continue.

2 3 The Proposed Amendments to Section 49

While there are a number of notable changes to the content of section 49 in the Bill, this article will be focusing on the following two phrases, which form a part of the proposed amendments to section 49(2)(b): (i) “... the suspect is suspected” and (ii) “... having committed a crime involving the infliction or threatened infliction of serious bodily harm”.

In terms of the current section 49(2) of the Criminal Procedure Act, an arrestor is justified in using deadly force “... only if he or she believes on reasonable grounds ...” that such force is necessary to protect the arrestor or any person assisting him or any other third party from imminent or future death or grievous bodily harm,29 or there is a strong likelihood that if not arrested immediately the suspect will cause imminent or future death or grievous bodily harm,30 or that the offence for which the suspect’s arrest is sought is in progress, is forcible and serious in nature and involves the use of life threatening violence or will likely cause grievous bodily harm.31

It is submitted that the current provisions require the arrestor to have a personal belief that the attempt to arrest the suspect complies with one of the three grounds set out in section 49(2). This suggests a subjective test of the arrestor’s state of mind at the time of the attempt to arrest. However, these provisions also require that the subjective belief must be based on reasonable grounds. This means that there must have been an objective reason for the arrestor’s belief. This requirement is in consonance with an observation made by Bruce, that

it would be unconscionable for police or private persons to regard themselves as having the authority to use lethal force for purposes of arrest without feeling that they have to be able to justify their actions in terms of some standard of reasonableness.32

In other words, it will be for the courts to determine whether the arrestor had a genuine belief that he was acting in compliance with section 49, and that that belief was reasonable and not farcical or speculative.

In terms of the proposed amendment to section 49(2)(b) the “... arrestor may use deadly force only if the suspect is suspected on

reasonable grounds ...”.33 This amendment appears to remove the requirement that the arrestor must have personal knowledge of or belief in the status of the suspect’s involvement in a crime involving death or serious bodily harm. The proposed section 49(2)(b) no longer limits knowledge of the crime involving death or serious bodily harm to the arrestor. It appears that this suspicion can be held by anyone, not only and indeed not necessarily the arrestor, although the arrestor would probably have to prove that his reliance on the veracity of the other person’s belief was reasonable.

This suggests that the arrestor would be entitled to use deadly force based on information told to him at the scene by a remote third party. That information would need to be from a reliable source in order for a court to hold that the arrestor’s actions were reasonable, but the wording of the section does not suggest that the arrestor need be aware of the reasonableness of the grounds pointing to the suspect’s involvement. His subjective interpretation of the facts before him could, in some instances, be superseded by information provided to him by a person who may be far removed from the scene of the arrest.

This scenario could lend itself to abuse or error; the arrestor may be justified in believing that the person informing him (that the suspect is suspected of having committed a serious offence) is well informed and honest. It is submitted, for instance, that a police officer who is informed over the police radio by his commander, that the suspect is wanted for multiple murders, would be entitled to claim the protection of the amended section 49, if the use of deadly force ensued. The fact that the commander’s information subsequently turns out to be incorrect, would not render the arrestor any more liable as there is no subjective requirement for that arrestor to believe that the suspect was indeed the suspect in the multiple murder investigation. This is exactly the issue that arose in London in the De Menezes shooting, which is discussed below.

In terms of the current section 49, the force used in effecting an arrest must be necessary to protect any person from imminent or future death or grievous bodily harm, or there must exist a risk that the suspect will cause imminent or future death or grievous bodily harm if not immediately apprehended. These provisions clearly state that the arrestor may use deadly force to prevent not only a present attack, but also one that may occur shortly thereafter, or even in the future. The wording of this provision echoes that used in the Canadian Criminal Code:

The peace officer or other person using the force believes on reasonable grounds that the force is necessary for the purpose of protecting the peace officer, the person lawfully assisting the peace officer or any other person from imminent or future death or grievous bodily harm. 34

In terms of the proposed amendment to section 49, the arrestor may use deadly force if the suspect is suspected of having committed a crime involving the infliction or threatened infliction of serious bodily harm. As the wording includes a threatened infliction of harm, the suspect need not actually have harmed anyone, but the threat must already have happened. Clearly, the section does not appear to cover a person who has not yet committed a harmful act, or a threat of harm.

This is in keeping with the decisions of the courts in Tennessee v Garner and Walters, both of whom use this wording in their recommendations. Consequently, it is easy to understand why this wording has been used in the proposed amendment to section 49. Writing in support of these courts’ view on this issue, Bruce notes the following:

What the test says in effect is that an individual who has (or is reasonably believed to have) committed a crime involving the infliction or threatened infliction of serious physical harm, has thereby defined him or herself as posing a danger of harm to other people. The test put forward in Garner, Govender and Walters is therefore a test of future danger, and the motivation for the use of lethal force is that of preventing future harm 35

Bruce goes on to discuss why he thinks this is preferable system, when he states:

The judgments also state that - implicitly - a person who is reasonably believed to have committed a crime involving the infliction or threatened (sic) of serious bodily harm, may reasonably be believed to pose such a danger. The judgements therefore go beyond a statement of the principle of ‘future danger’ which provides the motivation for using lethal force to provide a concrete test as to when an individual may be seen to pose such a danger. 36

Therefore, according to Bruce, this provision of section 49 acts as a limitation on the arrestor’s ability to use lethal force based on some speculative notion of future risk. This is commendable. However, with respect, it is submitted that Bruce has missed a beat here. While future harm is covered by the wording of Garner and Walters, it is submitted that the wording of the proposed new section 49 only permits the use of deadly force when the suspect is suspected of future harm based on the fact that the suspect has already committed another crime involving the actual or threatened infliction of harm.

It is submitted further, that this may have an impact on the police reaction to a terror suspect. For instance, if the police are acting on intelligence that a suspect is planning a bombing, they will not be able to use deadly force in effecting the arrest of that person unless the suspect has committed a previous crime involving the infliction of serious bodily harm, or unless a situation of private defence arises which is covered by the common law and by section 49(1) in any event. Therefore, the use of lethal force to affect the arrest of a suspect who has not committed or threatened to inflict serious bodily harm on anyone, will consequently not fall under the provisions of the proposed section 49(2). However, the current section 49(2) does in fact cover this situation, as under the current provisions, the arrestor would be able to rely on the fact that, based on the intelligence received, the arrestee is likely to cause imminent or future death or grievous bodily harm.

3 The De Menezes Killing in the United Kingdom and the IPCC Stockwell One Report

3 1 A Synopsis

On 22 July 2005, shortly after the London public-transport bombings, Jean Charles de Menezes was shot and killed by anti-terrorist police, in the mistaken belief that he was a suicide bomber. The police had been staking out a property in which a suspected terrorist (Hussain Osman) was residing. When de Menezes, a Brazilian migrant who also lived in the same building, left to go to work, the police, in a confused interaction between the surveillance team and headquarters, thought that he was Osman and proceeded to follow him. De Menezes boarded a few buses, alighted at an underground station, paid for his ticket, and proceeded to board a train at the bottom of the escalators. Moments later he was dead from seven bullet wounds to the head.37

3 2 Atmosphere of Fear - Events Leading up to the Killing

The investigation into de Menezes’ killing was conducted by the Independent Police Complaints Commission (IPCC), who compiled their findings into the Stockwell Report.38 The report noted that the incident took place at a particularly sensitive time, as there was “an atmosphere of fear for those living and working in the capital”.39 In fact, more than 3,900 calls had been received by the Anti-Terrorist Hotline in the preceding two weeks.40 Accordingly, the threat level to the United Kingdom had been raised to Level 1 (critical), which states that

Available intelligence and recent events indicate that terrorists with an established capacity are actively planning to attack within a matter of days (up to two weeks). An attack is expected imminently. 41

In response to the threat, London’s Metropolitan Police Service prepared a new strategy called “Operation KRATOS PEOPLE” (KRATOS). This strategy was devised with suicide bombers in mind, due to the extraordinary threats posed by such terrorists - in particular, the fact that when apprehended, the suicide bomber will often detonate a bomb. As this would thereby render the suspect’s arrest futile, and endanger civilians in the bomber’s presence, a unique approach was required outside the usual strategy. A crucial change in policy was encapsulated, as follows:

It may not be appropriate to issue a warning; the shot may be to the head to avoid detonating an explosive device and that decision to shoot may have to be taken on the command of a senior officer who has sufficient information to justify the use of lethal force. 42

The strands of KRATOS were three-fold: subject identification, subject confirmation and neutralisation, and it was noted that neutralisation could include a without notice killing if that was the only available option.43 The operation was headed by Central based at New Scotland Yard, with two surveillance teams on the ground outside the suspect’s building. Also in attendance was a specialist firearms’ department team, codenamed CO19.

3 3 The Sighting and Killing

The surveillance team reported to Central that they had seen a man leaving the building with a “good possible likeness to the subject” Osman.44 There were a few communications between Central and surveillance, as Central tried to obtain a positive identification from the surveillance team. There were mixed reports between the team members of what had transpired, but what was clear was that at no stage was there ever a 100% positive identification of the suspect conveyed by the surveillance team, while it appears that Central did believe that a positive identification had been established. 45

The CO19 firearm team was late to arrive at the underground station; De Menezes had already entered the station. Again, there were mixed reports about the instructions given to CO19, but all agreed that the instruction was to “stop” the suspect. Some CO19 members say they heard “stop at all costs”. In any event, the CO19 team, who had until boarding the tube not seen the suspect, at that stage formed the belief that they were required to stop a suspected suicide bomber. Based on their instructions from Central that a positive identification had been established, and based on the surveillance team’s identification of the suspect on the train, the CO19 members stormed the train, apprehended the suspect, and fired a number of bullets into his head at close range - with the provisions of KRATOS as their guiding tactics.

3 4 The Relevant Law

The IPCC set out the relevant law and its application to the incident.46 The relevant United Kingdom legislation that was applicable under the circumstances, was section 3(1) of the Criminal Law Act, 1967, which states:

A person may use such force as is reasonable in the circumstances in the prevention of crime, or in effecting or assisting in the lawful arrest of offenders or suspected offenders or persons lawfully at large.

The IPCC interpreted this to mean that the test to be applied to a person who relies on the above section, or the common law principles of self-defence, is a subjective one. In other words, the determination to be made is: “what is reasonable in the circumstances as he honestly believes them to be in the defence of himself or another”.47

Whilst the IPCC report went on to say that the “test is entirely subjective; it is the honest perception of the person using force in self-defence which matters”, it is submitted that the test contains an aspect of an objective element. This is because although the courts guard against armchair judgments, they nonetheless assess situations like this to ensure that the force used by the person was reasonable in the circumstances. This is similar to the approach adopted in South Africa, where the courts apply the subjective test to the state of mind of the person, but apply an objective test as a limitation in determining the lengths to which a person can react in any given situation.

3 5 Findings of the IPCC Stockwell One Report

The IPCC report held that none of the CO19 officers could be found guilty of murder or manslaughter. All of the officers honestly believed that the suspect was a suicide bomber who was about to detonate a bomb, honestly believed that the suspect had been correctly identified by Central and surveillance, and honestly believed that the only way to prevent the suspect from detonating a bomb was by shooting him in the head. The report found that no material has been seen or assembled by the IPCC to suggest that this tragedy was the result of any deliberate act designed to endanger the life of any innocent third party, or indeed to kill such an individual. All those involved at both command and operational level, were intent upon protecting the general public from a perceived threat of illegal lethal force.48 The report stated further that “Charlies 2 and 12 [two members of CO19] clearly believed they were acting in self

defence, and had the right in law to use the force they did”.49 However, it is evident that the death of De Menezes was caused mainly because the CO19 operatives were led to believe that De Menezes was a deadly threat, even though they had no direct knowledge of this fact, and even though they had not had an opportunity to conduct their own identification or confirmation of De Menezes’ identity. Central had informed the surveillance team of their mission, and had sent them photographs of the suspect. Accordingly, the surveillance team was dependent on the intelligence coming from Central. The surveillance team then staked out the apartment building and later informed their superiors at Central that the suspect had left the building. Without being able to independently verify this information themselves, Central therefore formed the honest belief that the man heading for the underground station was indeed the suspected terrorist; they relied on the information coming in from trained special operatives on the ground. Based on this information from surveillance, Central then informed the CO19 operatives that this was indeed the man in question, and that they needed them to “stop” him (an ambiguous term in police circles, especially in light of this situation). Consequently, the CO19 operatives formed the reasonable belief that De Menezes was indeed the suspected terrorist.

It is clear, therefore, that not only was the CO19 knowledge of the suspect’s status not first-degree, it was arguably fourth-degree. Yet the investigation finding was that this was an acceptable leap of faith to make, based on the flow of information along the chain of command. Essentially, it could be argued that de Menezes was shot by CO19 operatives, not because they suspected him of anything, but because the suspect was suspected of something which was reported to the operatives.

4 Application of the de Menezes Killing to the Proposed Section 49 Amendment

It is not difficult to accept the De Menezes findings in the Stockwell One Report, based on the fact that the officers involved in the killing did subjectively believe that the Brazilian was a suicide bomber, and was capable of immediately detonating a bomb if confronted. It is conceded that, in practice, this situation would possibly be covered by section 49(2)(a) of both the current version of the CPA (“the force is immediately necessary for the purposes of protecting ... any other person from imminent death or grievous bodily harm”), and the proposed amendment to the section (“the suspect poses a threat of serious violence to the arrestor or any other person”).

The De Menezes incident, however, does illustrate the dangers inherent in the removal of the phrase “he or she believes on reasonable grounds” from the current section 49, and its replacement with the phrase “the suspect is suspected on reasonable grounds”. It is submitted that the content of the proposed phrase lends itself to the errors that culminated in the death of De Menezes. In that instance, the CO19 operatives used deadly force solely based on Central’s suspicion of the suspect, rather than their own suspicion. Under the wording of the current section 49, the operatives could argue that the reasonable grounds - on which they based their belief - were that Central had positively identified the suspect as the suicide bomber. It would be reasonable for the operatives to believe that their superiors and surveillance team had correctly identified the suspect. The operatives would only have to show that their faith in the instructions from their superiors was reasonable. Once they had done that, any error that occurred thereafter would be exonerated by section 49.

Under the proposed section 49, however, it would be possible to argue that there is no need for the operatives to form an opinion on the instructions emanating from their superiors. This is not specified by the new wording. All that is required is that the suspect is suspected by someone, not necessarily the arrestor, of having committed a violent crime, and the use of deadly force would then be permissible if no other means is available to arrest the suspect. It is submitted that this could possibly lead to a number of civilian deaths, where the arrestor may have relied on information from headquarters, that may be intentionally, negligently or mistakenly incorrect.

To take this argument further, if the officer at headquarters is relying on information from the field officer about the identity of a suspect, without actually being able to see that person, and the field officer in turn is acting in accordance with information provided by the officer at headquarters, then it is conceivable that a blind-leading-the-blind situation could arise, and an innocent man may be shot dead as a result. There can never be a complete mutual agreement between these two hypothetical officers on the identity of the suspect, because the officer at headquarters is not there to have sight of the suspect. Therefore, he would be reliant on the field officer’s information to him, that the suspect is indeed the suspect they are chasing, at which point, acting on this verification, the officer at headquarters will form the suspicion that this is indeed the suspect who is suspected in respect of a violent crime, and he will then instruct the field officer accordingly. The field officer, oblivious to the possibility that there may be an identity error, will now be entitled to act under the protection of the proposed new section 49(2)(b), and may resort to the use of lethal force if necessary. It is only when the officer from headquarters arrives on the scene that the identity error will become apparent, but of course by then it would be too late for the deceased “suspect”.

As Bruce notes, it is probably for this reason that certain jurisdictions in the United States, namely Kansas City and Atlanta, require “police officers to have direct personal knowledge that someone had committed a crime”.50 Although this would be an ideal situation to emulate, the practicality of such a policy in crime-ridden South Africa, with its policing problems, would have to be assessed. Ideally, a police officer should not be able to shoot a person simply because he (the police officer) has been told over the police radio that that person is suspected of having committed a crime of threatened violence. Ideally, it should only be the police officer who actually witnessed the alleged crime, who should be permitted to take the decision to use lethal force in attempting to apprehend a suspect. Arguably, to kill in any other situation, would amount to a trial, conviction and sentence, based solely on hearsay evidence, often in a split second, and without the benefit of a lengthy trial. Such a scenario must surely be contrary to the role of the state as laid out in the Makwanyane case - that the state should not only preach respect for the law and that killing must stop, but by example should demonstrate regard for human life and dignity by refusing to destroy that of the criminal. 51

An interesting question to ask could be - what would have happened if De Menezes had not been suspected of carrying a bomb, and instead the police had received a tip-off that he was involved in planning an act of terror? Due to the absence of any weapon or bomb on his person, he would not have posed a threat of serious violence or death to the arrestor, and the provisions of section 49(2)(a) would therefore not have been applicable. Furthermore, the provisions of the proposed section 49(2)(b) would also not be applicable, as it would not have been possible to suspect him of “having committed a crime”, as he has not yet committed any offence of which we are aware. The proposed section 49(2)(b) quite clearly stipulates that the crime must already have been committed. Furthermore, the section clearly omits all reference to words such as “future”, “imminent” or “likelihood”. Therefore, based on the wording of the proposed new section, the arrestor would be obliged to let the suspect escape. While it must be conceded that this argument has merit, as the state’s power to use lethal force to prevent possible future events that may never transpire, certainly does need to be curtailed; a suspected terrorist surely cannot simply be allowed to resist arrest and flee. Imagine the public’s outrage if that suspect is, shortly thereafter, successful in staging a bombing. This would certainly help to foster the scenario envisaged by the court in the Govender case, that a failure by the state to preserve the effectiveness of the criminal justice system, will end in lawlessness and a loss of legitimacy of the state itself.52 It is submitted that this is a potential issue that will present the arrestor with yet another dilemma when deciding whether or not he is acting in accordance with the proposed new section 49(2).

5 Conclusion

As South African courts continue to strive to achieve a balance between the interests of the state, suspects and the public, the Constitutional Court’s decisions over the last decade or so have forced a rewriting of section 49 of the CPA. The current section 49 of the CPA is a fairly wordy piece of legislation that was drafted prior to judgment in the two landmark South African cases dealing with the issue of the use of force in effecting arrests, to wit Govender and Walters. Consequently, the proposed amendment to section 49 has been drafted with the aim to amend section 49(2) by

aligning the words in the proviso, that sets out the criteria as to when deadly force may be used in order to effect the arrest of a suspect, with the criteria tabulated by the Constitutional Court in the Walters case.

In effect, the wording of the proposed section 49(2)(b) has been taken almost directly from the eighth guideline laid down by the Constitutional Court in Walters,53 which in turn borrowed from the wording in the Tennessee v Garner decision in the United States. It is submitted that this proposed wording is unsatisfactory, and does not achieve the goal of making perfectly clear the circumstances under which an arrestor may use force in effecting an arrest, for two main reasons. First, the words “the suspect is suspected” may lead to too many innocent deaths due to the reliance by field officers on intelligence passed down to them by remote supervisors. It was this very scenario that led to the tragic death of de Menezes on the London underground in London, in 2005, when a number of field operatives - without at any stage pausing to form a subjective opinion of their own as to the suspect’s status - used lethal force on De Menezes, based solely on the information received from their superiors. This scenario, and the subsequent investigation, occurred within the highly trained, well-equipped and well-monitored London Metropolitan Police Service. It is submitted that there is an even greater likelihood of such an occurrence taking place in South Africa, due to the stressed, under-equipped status of our Police Service. It is therefore submitted that the proposed amendment should be further amended, to require an element of personal belief on the part of the arrestor, in much the same way as the current section 49 does.

Secondly, the words “having committed a crime” may place too great a limitation on the arrestor, as he will only be able to take action against someone who has already committed an offence, while being unable to use lethal force against someone who is reasonably suspected of planning a future offence. It is submitted that this may place an anti-terrorist police operative in a quandary, as they lose precious time trying to determine whether or not the suspect is a suspect referred to in the proposed amendment to section 49. This is not an ideal situation, as it could lead to the escape of a dangerous suspect, or to the death of the deliberating arrestor. It is therefore submitted that the proposed amendment to section 49 be further amended to include situations where the suspect is suspected of planning a future crime, but subject to a caveat on such action, such as the requirement that the suspicion must be based “on reasonable grounds”, for example. This would assist the arrestor in making a speedy decision in a potentially life-threatening situation, and would prevent a suspect from escaping, and to continue potentially life-threatening planning.


1. GG 33619 2010-10-07; Criminal Procedure Amendment Bill B39B-2010.

2. 51 of 1977.

3. 2002 2 SACR 105 (CC). See “Memorandum on the Objects of the Criminal Procedure Amendment Bill, 2010” 1.11.

4. 22 of 1998.

5. Minutes dated 2012-06-20 of a briefing by the Department of Justice to the Parliamentary Monitoring Group on the Criminal Procedure Amendment Bill [B39B-2010] 1.

6. The nine guiding principles are: “(a) The purpose of arrest is to bring before the court for trial persons suspected of having committed offences. (b) Arrest is not the only means of achieving this purpose, nor always the best. (c) Arrest may never be used to punish a suspect. (d) Where arrest is called for, force may be used only where it is necessary in order to carry out the arrest. (e) Where force is necessary, only the least degree of force reasonably necessary to carry out the arrest may be used. (f) In deciding what degree of force is both reasonable and necessary, all the circumstances must be taken into account, including the threat of violence the suspect poses to the arrestor or others, and the nature and circumstances of the offence the suspect is suspected of having committed, the force being proportional in all these circumstances. (g) Shooting a suspect solely in order to carry out an arrest is permitted in very limited circumstances only. (h) Ordinarily such shooting is not permitted unless the suspect poses a threat of violence to the arrestor or others or is suspected on reasonable grounds of having committed a crime involving the infliction or threatened infliction of serious bodily harm and there are no other reasonable means of carrying out the arrest, whether at that time or later. (i) These limitations in no way detract from the rights of an arrestor attempting to carry out an arrest to kill a suspect in self-defence or in defence of any other person.” Idem par 54.

7. Memorandum 1.6.

8. See: Stockwell One Report title page at (accessed 2013-12-17).

9. At page four of the Bill. The proposed amendment to s 49(2)(b) reads as follows: “If any arrestor attempts to arrest a suspect and that suspect resists the attempt, or flees, or resists the attempt and flees, when it is clear that an attempt to arrest him or her is being made, and the suspect cannot be arrested without the use of force, the arrestor may, in order to effect the arrest, use such force as may be reasonably necessary and proportional in the circumstances to overcome the resistance or to prevent the suspect from fleeing but in addition to the requirement that the force must be reasonably necessary and proportional in the circumstances, the arrestor may use deadly force if - the suspect is suspected on reasonable grounds of having committed a crime involving the infliction or threatened infliction of serious bodily harm and there are no reasonable means of effecting the arrest, whether at that time or later.” [Words in bold are the proposed amendment to s 49(2)(b)].

10. Ibid.

11. S 49(2)(a), (b) CPA.

12. See Ch 2 Constitution of the Republic of South Africa, 1996.

13. S 10 Constitution.

14. S 11 Constitution.

15. 2001 4 SA 273 (SCA).

16. Idem par 12.

17. 2002 2 SACR 105 (CC) 129 par 44.

18. Tennessee v Garner 85 L Ed 2d 1. Here the court stated the following: “It is not better that all felony suspects die than that they escape. Where the suspect poses no immediate threat to the officer and no threat to others, the harm resulting from failing to apprehend him does not justify the use of deadly force to do so. It is no doubt unfortunate when a suspect who is in sight escapes, but the fact that the police arrive a little late or are a little slower on foot does not always justify killing the suspect. A police officer may not seize an unarmed, nondangerous (sic) suspect by shooting him dead. The Tennessee statute is unconstitutional insofar as it authorises the use of deadly force against such fleeing suspects. It is not, however, unconstitutional on its face. Where the officer has probable cause to believe that the suspect poses a threat of serious physical harm, either to the officer or to others, it is not constitutionally unreasonable to prevent escape by using deadly force. Thus, if the suspect threatens the officer with a weapon or there is probable cause to believe that he has committed a crime involving the infliction or threatened infliction of serious physical harm, deadly force may be used if necessary to prevent escape, and if, where feasible, some warning has been given.”

19. 2000 1 SACR 414 (CC).

20. 442 par 69.

21. 2002 2 SACR 105 (CC) par 49.

22. S v Makwanyane 1995 2 SACR 1 (CC).

23. 50 par 122.

24. 52 par 128.

25. Par 129.

26. 2002 2 SACR 105 (CC) 134 par 54.

27. Ibid.

28. 1995 2 SACR 1 (CC) 85 par 222.

29. S 49 (2)(a) CPA.

30. S 49(2)(b) CPA.

31. S 49(2)(c) CPA.

32. Bruce “Killing and the Constitution - Arrest and the use of lethal force” 2003 SAJHR 430 442.

33. The full provision of s 49(2)(b) is as follows: “the suspect is suspected on reasonable grounds of having committed a crime involving the infliction or threatened infliction of serious bodily harm and there was no other reasonable means of effecting the arrest, whether at that time or later.”

34. Canadian Criminal Code Ch C-46 s 25(4)(d).

35. Bruce 2003 SAJHR 430 445.

36. Ibid 448.

37. Stockwell One Report.

38. Ibid.

39. Idem 4.6.

40. Idem 4.11.

41. Idem 4.7.

42. Idem 9.6.

43. Idem 9.12.

44. Idem 12.27.

45. Ibid.

46. Idem 19.3.

47. Idem 19.4. Beckford v The Queen [1988] AC 130.

48. Stockwell One Report 20.1.

49. Idem 20.74.

50. Bruce 2003 SAJHR 430 453 (quoting from Sherman “Reducing Police Gun Use: Critical Events, Administrative Policy, and Organisational Change” in Control in the Police Organisation (Punch) (1983) 108.

51. 1995 2 SACR 1 (CC) 85 par 222.

52. 2001 4 SACR 273 (SCA).

53. 2002 2 SACR 105 (CC) 134 par 54.