Crime Quarterly NO. 2 SA CRIME QUARTERLY No 2 NOVEMBER 2002 13 David Bruce, Centre for the Study of Violence and Reconciliation dbruce@csvr.org.za STOPPING COP KILLING Lessons and limitations of South African research One issue that has been of particular concern in South Africa over the last few years has been the high number of violent incidents in which members of the SAPS have been killed. Research has been done to determine the risk profile of police, the circumstances of killings both on and off duty, and the motives of police killers. But there are unanswered questions and further research could contribute to a better understanding of the dangers facing police. Table 1: Killings of members of the SAPS: 1991 - 20011 M irroring the broader escalation of violence in South Africa, killings of police escalated dramatically in the early 1990s, with the highest number, 280, killed in 1993. This continued at rates of over 200 a year until 2000. While according to official figures 2001 saw the lowest number of these killings since 1990, they remain at an unacceptably high level (Table 1). In addition to these incidents there are also many incidents in which police members are injured, or are the target of violence, but escape uninjured. These incidents are frequently not documented. The killings of police have been the subject of a number of research studies over the last few years. These have mostly been linked to the work of a task team established in May 1999 to address the high number of these killings. Trying to determine a risk profile The picture that has emerged is that most police who have been killed have been black police members, killed by black male assailants in predominantly black areas. However, the information that has thus far emerged has not Year Total Number Number number killed killed killed on duty off duty 1991 164 65 99 1992 253 116 137 1993 280 104 176 1994 265 84 181 1995 240 84 156 1996 222 76 146 1997 244 90 154 1998 235 91 144 1999 204 81 123 2000 185 60 125 2001 163 67 96 Total 2,455 918 1,537 (100%) (37%) (63%) clarified whether black members are more vulnerable to killings than white members, or whether these patterns of victimisation simply reflect the broader realities of the racial composition of the SAPS and of deployment of police personnel. While the reports that have been produced generally refer to attacks on and murders of police officials, these terms have not been that clearly defined. While the majority of police killings are deliberate and unlawful, and may therefore be classified as murders, the recorded killings appear to include incidents where police officers have been the antagonists, and were killed in disputes – in bars, or in domestic or recreational disputes, or even while committing crimes. The phrase ‘attacks on police’, also widely used as a general term in the research, is also potentially misleading, as will emerge from the following discussion on the general circumstances of, and motives for, the killings. General circumstances of the killings The studies conducted so far indicate that the killings of police mostly fall into one of the following three types of circumstances2: Killings in response to police intervention These would generally be situations where members of the police are on duty, though it may also include situations where they are off duty and place themselves on duty, or exercise the powers of ordinary civilians to act in private defence and effect citizens’ arrests. What distinguishes these situations is that police members mostly initiate their involvement. Robberies and other attacks These include attacks on members who are on duty and members off duty, both in and out of uniform. The killing may be related to the fact that the victim is a police officer, or it could be targeting a specific individual, or it could simply be a random act of crime. Killings in domestic, recreational or occupational circumstances, relating to a dispute or argument involving a police officer These include situations where police become SA CRIME QUARTERLY No 2 NOVEMBER 2002 14 BRUCE involved in arguments in bars, or with members of their families, or with lovers, spouses, partners or colleagues. While some of these killings may be described as attacks, they also involve confrontations where the person who kills them acts in self- defence, or where both parties contribute to the conflict. What distinguishes these situations is that the police officer is killed as a result of being party to an argument or dispute and the consequent heated emotions, and not as a result of carrying out police duties.3 The studies that have been conducted thus far have helped to illuminate the picture regarding the different circumstances in which police have been killed, but important questions remain. While the studies are not unanimous on this point, analysis of the data contained in some of them suggests that killings in response to police intervention (category 1) constitute the largest of the three categories of killings.4 This would however appear to contradict the fact that the bulk of killings that have occurred (consistently over 60% of killings), have been of off- duty police. If most of the killings occur off duty it does not make sense that they are in response to police intervention, unless police are doing more of their work off duty than on duty. It therefore appears possible that killings in attacks and arguments (categories 2 and 3) are underrepresented in the data in some studies. It is likely that there are significant differences in the pattern of killings, with category 1 (police intervention) probably constituting the bulk of killings on duty, but category 2 and 3 constituting the bulk of killings off duty. However, the research has not as yet provided us with a clear picture of the differences between off-duty and on-duty killings. Motives for the killings Some of the issues relating to the motives for the killings that have emerged from research conducted thus far, include:5 Resisting arrest or preventing identification Reporting on interviews with 26 convicted police killers, Minnaar states that 42% (11 out of 26) SA CRIME QUARTERLY No 2 NOVEMBER 2002 15BRUCE indicated that the killings of police were to “avoid being arrested”. Of this number, 19% (5 out of 26) occurred while police were responding to a crime and were trying to apprehend the suspects – for instance during or just immediately after a robbery or vehicle hijacking.6 These are mostly incidents that occur during normal police duties, but the report also includes a member who was executed by colleagues to prevent exposure of their corrupt activities, and a member who returned home during a house break-in to find that one of the perpetrators was known to him. Robbery One question that research has tried to answer is the extent to which robberies of firearms have motivated attacks and, related to this, whether it might be in the best interest of the police to disarm, even when on duty. One report, for instance, indicates that firearms were stolen from the deceased members in 114 (15%) of the 754 cases.7 But while firearms are sometimes stolen when police are killed, it is often difficult to clarify whether this is the primary motive for the killing or whether firearms were simply taken opportunistically afterwards. Of the 26 convicted police killers, for instance, two said that they attacked police to steal the service pistol, while another indicated that “the victim was robbed, his car hijacked and they also stole his firearm before killing him”. Self-defence Some of the statements by convicted police killers indicate that they acted in the belief that it was necessary to kill in order to protect their own lives. In one of the cases that Minnaar classifies as “resisting arrest”, the convicted killer implies that the killing would not have occurred if the police had followed formal procedure by identifying themselves. He states that “I thought they were robbers from the township”.8 Another killer suggests that police may worsen the level of violence against them by the manner in which they enter into situations: The police shoot too quickly as they arrive [at the scene of a crime]. They just begin shooting. The police die from their mistakes. If they shoot I will shoot back. On the other hand one of the killers indicates that his actions were motivated by a perception that the police would not act in a lawful manner: I was scared that he would shoot me because he asked me to come closer. A recent newspaper story also suggested that some suspects fear that they will be summarily killed by the police. An alleged armed robber indicated that he was willing to surrender to the police but that he feared for his life if he did so.9 Discussing this issue in the US, Geller and Scott have also stated that: another possible explanation for long-term downward trends in shootings of police ... is the possibility that at least some potential police assailants have been made less afraid over the years that the police will administer ‘street justice’ with a nightstick or lethal weapons and hence are less likely to shoot an officer to protect themselves. Reasonable or not, there can be little doubt that some of those who assault police do so in the belief that the officers are looking for an excuse to harm them and will do so unless prevented.10 Hatred and anger Another factor that has been considered by some researchers is that the killings may be related to residual hostility on the part of community members towards members of the police, partly linked to the role played by the police in enforcing apartheid. But while this may be a contributing factor, research has not shown that this is a prominent motive for police killings. Thus while two of the 26 police killers interviewed indicated that the killings were related to political motives (some were involved in killings prior to 1994), none of the respondents appear to have placed any major emphasis on hostility towards the police. In the one case where negative attitudes to police were mentioned, the SA CRIME QUARTERLY No 2 NOVEMBER 2002 16 BRUCE interviewee suggested that this might be motivated by the manner in which police handle suspects: .....when they arrest suspects they must not hit the suspects or cause them injury. And it can happen that when they arrest that suspect after a time and hit him and he is outside on the street again .... he sees the police as the enemy.11 Police who resort to violence unlawfully may be doing their colleagues a disservice, as they may at a later point in time encounter a suspect who believes that he will be ill-treated or killed if he surrenders. Arguments and alcohol As indicated, one of the three major categories of killings has been that of arguments or disputes in domestic or recreational circumstances. In Nel and Conradie’s 1998 report, off duty killings included ‘arguments’ in 29% of cases, and ‘love triangles’ in 6% of cases. Of those killed 16% “had to some degree been intoxicated when the murders occurred”.12 One of the risk factors for police may be their own alcohol use, or their proximity to police who become unstable from alcohol. Nel and Conradie point out that in 9% of the 385 cases in their sample “perpetrators were indicated as being other SAPS members”. Minnaar indicates that out of 1,364 cases of police killings during 1994 to 1999, the perpetrator was a colleague in 81 of those cases (6% of the total).13 In these circumstances the factor of self-defence features once again as a possible motive for some killings. One of the police killers interviewed, a former policeman, said that he had acted in self- defence in an attack in his own house by an intoxicated police member.14 The killings of police and police safety The evidence that has emerged is that the bulk of the killings of SAPS members have either been linked to attempts to evade arrest or in other ways resist police intervention, and to attacks, sometimes for the purpose of robbery. The primary explanation for these deaths may be found in the general societal factors currently contributing to high levels of crime and violence in South Africa. This means that the key components of the current police safety strategy, namely education, tactical training, improving the equipment provided to police, and building community relations, are appropriate. Tactical training may also attempt to address what may be a problem of over-hasty actions by police, unnecessarily escalating the levels of violence in some confrontations. Reckless and irresponsible behaviour by SAPS members off duty should also be examined in more detail as a factor contributing to the killings of police. In this regard it needs to be stated that the carrying of firearms by SAPS members also aggravates the problem of violence against the police. As was mentioned before, it may be that SAPS members are targeted by offenders intending to steal their firearms. The fact that the police are armed also increases the possibility of a violent response in confrontational situations. Confrontations can occur in any circumstance, be it occupational, domestic, or recreational. While in the South African context there is not an argument for the general disarmament of the police, there is a need to pay greater attention to questions concerning the possession of firearms by SAPS members – particularly where they show signs of emotional instability and are prone to alcohol abuse. Furthermore there may be value in providing greater clarity to SAPS members around the responsibility to place themselves on duty when they are officially off duty, and to what extent this necessitates that they carry firearms when off duty.15 While police-community relations have improved in the last decade, hostility towards the police is likely to persist among many of those who inhabit the criminal sub-stratum of South African society. In so far as members of the police service act brutally and unlawfully in dealing with suspects, they may feed into hostility towards the police in general, as well as contribute to beliefs that surrendering to the police is risky. SA CRIME QUARTERLY No 2 NOVEMBER 2002 17BRUCE Finally, further research, using the insights of studies that have already been conducted, improved sampling techniques, and greater definitional clarity may contribute to a better understanding of the dangers facing police. Endnotes 1 There are frequent variations in different reports in the statistics for police killings, although the figures broadly follow the trends outlined in the table. The statistics for 1991-1993 are from A Minnaar, An analysis of attacks on and murder of members of the South African Police Service: Searching for Preventative and Protective Strategies, Paper presented to the Xth International Symposium on Victimology, Montreal, Canada, 6-11 August 2000. Statistics for 1994-1999 are from the SAPS Crime Information Centre as quoted in S Masuku, Most at Risk: murder of police officials, Nedbank ISS Crime Index, Vol 4, 2000. For 2001, the statistics are from Inspector E Nel at the CIAC, telephonic communication, 11 June 2001. 2 In addition to deaths which fall into the three categories identified, statistics provided by the SAPS in 1997 indicated that 88 SAPS members were killed in accidents involving firearms in the period January 1994 - June 1997. (Fax from SAPS CIMC, 11 August 1997). While examining accidental killings is also important to understanding police safety, it appears that the data on killings of police mostly exclude accidental killings. A fifth identifiable category is that of ‘killing of SAPS members arising from pre-meditated criminal activities to which they are party’ though existing studies suggestt that these account for only a small percentage of killings. 3 As is usual with analytical definitions there is not a watertight distinction between these three categories. For example the distinction between ‘attacks’ and ‘police intervention’ is not watertight as in some police interventions (e.g. police search of an apparently empty house) the suspect/killer may retain the element of surprise. 4 It should be noted that the descriptive categories used in existing studies are not clearly defined or distinguished from each other and so it is often not possible to know whether killings are best categorised in categories 1, 2, or 3. 5 See A Minnaar, Opportunity Knocks in the Heat of the Moment: An analysis of perpetrators’ profile and their own perceptions of attacks on and killing of police officials, Institute for Human Rights and Criminal Justice Studies Occasional Paper, Technikon SA, undated. Cases from Minnaar’s study which are not discussed in this discussion of motive include attacks on police to release prisoners (these could arguably be included in the resisting arrest category), for political reasons (some cases dated to the period prior to 1994) and 2 cases involved attacks by police on colleagues. In at least three cases however the respondents motives are unspecified and they are simply indicated as having been involved in ‘direct attacks on police (unknown motives)’ (Minnaar, pp. 7 - 8). 6 A Minnaar, undated, op cit. 7 Ibid. 8 Ibid. 9 Robber tells all, Sowetan, 4 October, 2002. 10 W Geller & S Scott, Deadly Force: What We Know, Police Executive Research Forum, Washington DC, 1992. 11 A Minnaar, undated, op cit. 12 E Nel & H Conradie, The Murder of Police Officials in South Africa: 1994-1997, Pretoria, South African Police Service, 1998. 13 A Minnaar, 2000, op cit. 14 A Minnaar, undated, op cit. 15 Note the remark from Geller and Scott, op cit, that ‘police frequently are at a tactical disadvantage while off duty, being out of radio communication with other officers and usually not having the opportunity to plan a course of action as they travel to a suspected crime scene. Nor do off-duty officers have the same perceived legitimacy to intervene that on-duty officers have when they are summoned to a disturbance by a civilian complaint. ... Moreover, off-duty officers who happen to be socialising in a tavern when armed robbers enter may be in no condition to outdraw the holdup men, yet the presumed imperative to take police action may prompt an ill-advised confrontation”. Other source documents H Conradie, Surviving an attack, Servamus, April 1995, pp 12-14. H Conradie, The Victims of Attacks on Police Officials in South Africa, Crime Research in South Africa 1(1), Department of Criminology, Pretoria, University of South Africa, 2000. H Conradie, The Modus Operandi of Police Killers in Four Provinces of South Africa, Crime Research in South Africa 2(1), Department of Criminology, Pretoria, University of South Africa, 2001. Crime Information Analysis Centre, A Criminological Perspective on the Attacks on Police Officials in Four Provinces of South Africa, Unpublished report, South African Police Service, September 2000. A Minnaar, An Analysis of the Murder of Members of the South African Police Service: 1994-1997, Institute for Human Rights and Criminal Justice Studies Occasional Paper No 5, Technikon SA, 1998.