CHAPTER 3
SAFETY AND SECURITY

---

Published in Monograph No 88, November 2003

Criminal Justice in Review 2001 / 2002


Ted Leggett, Antoinette Louw
Martin Schönteich and Makubetse Sekhonyane

The broad objectives of the South African Police Service (SAPS) are to:

In order to achieve these outcomes, the South African government employs one of the single largest policing bodies in the world, with a strong central concentration of authority under a single National Commissioner. Beneath this Commissioner is a hierarchy of supervisory structures descending from the national level to the nine provinces, 42 ‘policing areas’, and about 1,200 stations. While the Commissioner reports directly to the President, the Department of Safety and Security oversees SAPS operations and includes the civilian Secretariat for Safety and Security, which provides advice to the minister in the exercise of his functions.⁵

According to its 2001/02 annual report, the values held by the SAPS are to:

This broad mission and these rather vague aspirations need to be tied down to some specific performance indicators, but defining what the police do in numeric terms has confounded police departments around the world.

What was promised?

This question is not as straightforward as it may seem, as the police, like other government departments, has at least two distinct constituencies: the general public on the one hand, and monitoring agencies, both inside and outside government, on the other.

For the public, the issue is crime—they want less of it. Given that “to prevent crime” is at the head of the SAPS’ constitutional mandate, and is not mentioned in the mandates of any of the other government agency, it is not surprising that many of the public pronouncements by the police leadership involve reducing crime levels. These statements cite the crime statistics of the SAPS as though these are true reflections of the real crime situation, which, as discussed above, is a highly problematic assumption. In the end, this means the primary pledge of the SAPS to the public has been to reduce the number of crimes it records.⁷

For monitoring agencies, the situation is more complex. While recorded crime is noted, also important are more detailed indicators of performance. In the process of acquiring funds from parliament, the Department of Safety and Security promised to monitor and report on the indicators of performance listed in Table 1.

Table 1: Safety and Security outputs and indicators, 2002/038
Sub-programme	Outputs	Service delivery indicators
Visible policing	Levels of safety and security. Control firearms. Efficient detention management. Police visibility.	Crime levels per capita. Crime prevention operations in high crime areas. Reduced level of firearm crime. Holding arrestees securely. Implementation of sector policing. Multi-disciplinary initiatives to combat crimes against women and children.
Border policing	Arrests & seizures at border posts.	Arrests for possessing narcotics, illegal firearms & stolen goods. Seizure of stolen goods and contraband.
Public order policing	Crowds managed appropriately.	No. of persons arrested for public disorder incidents.
General investigations	Investigation of general crime.	No. of cases to court in relation to reported crimes.
Organised crime	Policing of criminal organisations.	No. of cases related to organised crime syndicates in relation to cases investigated.
Commercial crime	Investigation of commercial crime activities.	No. of cases to court for fraud and forgeries in relation to cases investigated.
Criminal Record Centre	Criminal record reports.	Automation of Criminal Record Centre.
Forensic science laboratory	Evidence provided.	Turnaround time of requests for evidence.
Crime intelligence	Analyses of crime patterns, threats, profiles & linkages.	No. of intelligence products. No. of National Crime Threat Analyses reports compiled.

Source: National Teasury, 2002

How did the SAPS intend on achieving the goal of reducing recorded crime while maximising performance in terms of their budget vote indicators? In April 2000, the Department of Safety and Security launched a three-year plan, the National Crime Combating Strategy (NCCS). The NCCS is “a multi-disciplinary approach that focuses managerial, human and logistical resources on ‘hot spot’ areas where crime is disproportionately high”.⁹ The department identified some 145 stations (about one station out of ten), which account for more than 50% of serious crime recorded by the police.

Evaluating the NCCS is difficult, given that no public document has been issued describing the strategy in any detail.¹⁰ It can be gleaned from other documents, however, that the NCCS identifies four key strategic priorities.¹¹

First, the strategy prioritises the combating of organised crime, focusing on crimes relating to drugs, firearms trafficking, vehicle theft and hijacking, corrupt public officials and organised commercial crime.

Second, it focuses on reducing the country’s “unacceptably high levels of serious and violent crimes”.¹² Toward this goal, strategies have been designed to:

Third, the development of strategies to reduce the incidence of “crimes against women and children” is prioritised, and the department also intends to improve the investigation of those crimes that do occur. While crimes against women and children are nowhere defined, they would appear to involve crimes in which women and children are uniquely or disproportionately involved as victims, such as the present legal construction of rape, sexual offences against minors, and perhaps domestic abuse.

Fourth, the strategy prioritises improving service delivery at police stations. This most likely involves conforming to the ideals of the Batho Pele White Paper on Public Service Delivery, which seeks to establish a sense of public accountability within the civil service. This priority is very different in kind to the other three, and difficult to quantify.

The heart of the NCCS, and the facet by which it is best known, is its operational strategy. This strategy adopts a “geographic approach” whereby Crime Combating Task Groups were established to target serious and violent crime in identified high-crime zones. Task groups conduct intelligence-driven operational interventions according to the specific crime threat analysis of each zone.¹³ This is done in addition to the usual policing activities in the affected area. The operational strategy consists of:

The second operational element of the NCCS is an “organised crime approach”, whereby organised crime syndicates are investigated in intelligence-driven operations carried out by task teams under the command of experienced detectives.¹⁴

Available resources

In order to achieve its stated objectives, the Department of Safety and Security receives the largest share of the criminal justice budget. In the 2002/03 budget year, R19.2 billion, or 7% of the non-debt budget, was allocated to the department. This share has been decreasing in recent years, however, and is presently increasing below the rate of inflation. About 28% of the department’s budget was spent on administration, of which just above 68% was spent on personnel (about R3.7 billion).¹⁵ Slightly over three-quarters (77%) of the police budget is spent on personnel generally, which compares favourably with other police departments internationally.

At the end of 2002, the SAPS employed almost 130,000 people, of which about 62% were uniformed members, 17% detectives and 21% civilians. South Africa’s police staff member (functional personnel and civilian staff) to public ratio is comparable with that of many developed and Commonwealth countries.¹⁶ However, compared to more developed countries, South Africa’s police officer to civilian ratio is strained by a number of factors:

Moreover, the issue is not only about the number of police personnel, but also about issues of deployment, supervision, and capacity of the personnel. For example, in England and Wales there is approximately one functional (i.e. uniformed officer and detective) police member for every 409 citizens, compared to one functional police member for every 450 South Africans.¹⁸ While the ratios are not significantly different for either country, there are sharp differences when comparing the rank distribution of English and Welsh, and South African functional members (Figure 6).


Figure 6: Police rank distribution: England & Wales vs South Africa



Source: British Home Office, SAPS Management Services

The English and Welsh force is comprised of 97,337 constables (75%), 18,608 sergeants (14%), 6,035 inspectors (5%), and 7,716 senior staff (6%). At the end of 2002, the SAPS, in contrast, was comprised of 12,191 constables (12%) 18,072 sergeants (18%), 56,930 inspectors (55%), and 15,156 senior staff (15%).

The number of functional SAPS members declined significantly after 1995. Between 1996 and the end of 2002 the number of SAPS officers declined by some 7,400 (or 5%), notwithstanding the fact that the number of SAPS officers increased by over 9,000 during 2002. Moreover, between 1996 and late 2002, the number of civilians employed by the SAPS increased by almost 6,500 (Figure 7).


Figure 7: Changing number of SAPS officers and civillians



Source: SAPS Management Services

At the end of 2001 the SAPS had a fleet of 26,908 vehicles, up marginally from 26,857 in 2000, but down from 28,432 in 1999 and 37,094 in 1995.¹⁹ In other words, the size of the police’s vehicle fleet declined by 28% between 1995 and 2001. Of the vehicles in operation at the end of 2001, 36% had covered distances of more than 160,000 kilometres. In theory, four-fifths of the SAPS vehicle fleet is utilised for operational policing, functional crime prevention and reactive policing, while the other 20% are used for administrative purposes, support services and management.²⁰

Workload

Whatever one thinks about recorded crime as a reliable estimate of real crime, it does provide the best single indicator of the SAPS’ workload. About 2.5 million crimes were recorded by the police during the 2001/02 financial year. This represents about 31 crimes for every uniformed member per year, or one crime every eight working days. If all uniformed police members were assigned to street duty, this ratio would seem very light indeed, but this is not the case.

Turning to detectives, all of who deal with real crime, the ratio is 115 crimes per detective per year, or one every other working day^.21 A significant proportion of crimes receive no, or very little, detective follow up due to lack of evidence or leads, but some require far more than two days’ work.

Looking again at the British example, the police in England and Wales recorded 5.5 million crimes during the 2001/02 financial year: a 45% larger workload for a force just 25% bigger than South Africa’s. However, most (82%) of the crimes recorded in England and Wales were against property.²² Serious violent crimes, some of which may demand more police attention if properly investigated, occur at a higher rate in South Africa, comprising about a third of total recorded crime in the country.

Actual performance

Recorded crime

Although recorded crime can serve as an indicator of the workload experienced by the police, there are serious problems with using it as a police performance indicator, including the following:

Changes in legislation and policy can also affect the rates at which certain crimes are recorded. For example, many crimes are only detected by affirmative police activity, such as crimes involving drugs, illegal guns, and undocumented immigrants. These are precisely the types of arrests that are likely to be made during Operation Crackdown-type strategies. Another example is the Domestic Violence Act, which is likely to dramatically increase the number of assault complaints recorded. If this result is not seen, it calls into question whether the police are taking the legislation seriously.

Despite these obvious problems with using recorded crime as a performance indicator for the police, a substantial portion of the speeches and documents released to the public by the Department of Safety and Security focus on recorded crime, and the present operational strategy has been justified and expanded based on these statistics.

In mid-2001 the moratorium on crime statistics that had imposed in July 2000, was lifted. In a press conference in February 2002, then Minister of Safety and Security, Steve Tshwete, declared that the police had achieved the objective of stabilising crime in half the time allocated by the NCCS.²⁵ This statement refers to the objective articulated in the strategic planning documents of the NCCS that crime in the priority station areas would be stabilised by 2004. While stabilisation is not defined in public documents, the commonsense meaning of this term is that recorded crime levels across offence categories do not get worse over an extended period of time.

In support of this claim, the SAPS’ Crime Information Analysis Centre (CIAC) issued a report entitled The Reported Serious Crime Situation in South Africa for the period January–September 2001.²⁶ This report states that, on a national level, 17 of 20 priority crimes had stabilised between 2000 and 2001. There are several problems with this analysis:

While the number of crimes recorded increased between 2000/01 and 2001/02, the rate of increase slowed down. The 12-month period covered by the 2001/02 financial year experienced a 0.9% increase in recorded crime, compared to the previous 12 months—the lowest year-on-year increase since the actual decline in recorded crime in 1996/97 (Figure 8).

Figure 8: Annual percentage change in number of recorded crimes, 1995/96 - 2001/02



Source: SAPS Crime Information Analysis Centre

In speeches associated with the release of the SAPS’ 2001/02 annual report, Minister of Safety and Security, Charles Nqakula, asserted the success of the NCCS in stabilising crime. He claimed that, as of March 2002, another eight of the 145 priority station areas had been “stabilised” by Operation Tsipa, an intensified version of Operation Crackdown aimed at the third of the stations that had not shown statistical stabilisation despite high density operations.²⁷

Thus, according to the SAPS, the high-density operations it has adopted since April 2000 have stopped the rise in crime. It is likely, however, that this rise was itself a product of increased reporting rather than increased incidence, and the alleged stabilisation is most likely a product of changes in recording practices rather than changes in the real crime situation. In the end, the root causes of crime are not impacted by unsustainable high-density police presence in high crime areas, and any suppression effect may be reversed when these operations are withdrawn.

Arrests and seizures

In line with some of the indicators promised in the 2001/02 and 2002/03 budget votes (see Table 1 above), the SAPS’ 2001/02 annual report lists an impressive number of operations, arrests, and seizures that have resulted from the present operational strategy (Tables 2 and 3).²⁸ These figures represent the removal from society of dangerous individuals and contraband, the impact of which should not be trivialised. By removing criminals (many of who are likely recidivists) and the instrumentalities of their offences (such as guns and stolen vehicles) from society, the police performs the crime prevention function most appropriate for its skills. Moreover, the comparison of arrests made between the 2001/02 financial year and the previous one indicates a positive trend toward more serious arrests.

Table 2: SAPS operations, 2001/02
Type of operation	Number of operations
Roadblocks	50,424
Cordon and search	47,694
Vehicles searched	1,735,053
Premises searched	230,346
Persons searched	3,795,533
Vehicle patrols	268,001
Foot patrols	186,950

Source: SAPS 2001/02 Annual Report

Table 3: SAPS arrests, seizures and recovery of stolen goods, 2001/02 and 2000/01
Type of operation	Quantity
	2001/02	2000/01
Arrests: serious crimes	188,931	176,235
Arrests: less serious crimes	202,141	309,316
Vehicles recovered	8,176	15,804
Firearms seized	23,303	18,402
Stolen electronic property recovered	32,830	31,049
Cannabis	495,927,905 kg	N.A.
Mandrax	4,202,835 tablets	N.A.
Cocaine	2,842 kg	N.A.
Heroin	85kg	N.A.

Source: SAPS 2001/02 Annual Report

While the NCCS does seem to have resulted in a net increase in arrests and convictions, it is not clear how many of these arrests would have been made in the absence of the strategy. The 600,000 operational arrests claimed in Table 3 are roughly equal to 80% of all cases referred to court in 2001. It seems likely, therefore, that the lines have become blurred between ‘operational’ arrests and those that occur in the normal course of police business. Comparing 1999 (the last year in which the NCCS was not functional) with 2001, the rate of cases referred to court has increased by 28%, but the rate of withdrawals has increased by 44%. Nonetheless, convictions are up by 30% between 1999 and 2001, which represents a net increase in enforcement productivity.

Seizures represent a mixed indicator of performance, because the presence of contraband in society in the first place can ultimately be blamed on lax law enforcement. In the future, it is hoped that such large seizures will be impossible, because the contraband will no longer be present.

Investigations

An internationally accepted performance indicator for detectives is the rate at which cases are referred to court, also known as the ‘arrest rate’, ‘detection rate’ or ‘clearance rate’. This rate is never 100%, because some cases are affirmatively determined as unfounded, either because the original claim is fraudulent or because further investigation reveals that the facts of the case do not represent a crime.

In addition, some cases have to be withdrawn by the police because witnesses refuse to cooperate and a case cannot be made without them. This number can be minimised by good detective work, however, and ultimately this should count against detective performance. Finally, a certain share of cases will remain ‘undetected’ because, although it is clear that a crime was committed, there is simply not enough evidence to make out a case against any individual. While a sizeable share of cases are always lost in this way, it is again incumbent on detectives to minimise this number. Detectives cannot be allowed to close cases at will without this reflecting on their personal performance.

Since all detectives internationally encounter the same problems of unfounded cases, uncooperative witnesses, and undetectable crimes, referral rates could, in theory, be compared between countries. However, different countries have different definitions of crime detection. In the United Kingdom, for example, to be counted as a ‘detected crime’, there must be: ‘sufficient evidence to charge’, an interview of the offender (not including convicted prisoners), and notification of the victim—a far stricter standard than that found in South Africa. The SAPS pushes the standard lower still, including in its 2001/02 annual report definition of ‘detection rate’ both ‘unfounded’ and ‘withdrawn’ cases. Further, details on each of these respective categories have not yet been released to the public. This makes it impossible to make even the broadest comparisons between national standards for the 2001/02 financial year.

While the referral rate in South Africa varies between provinces, so do provincial workloads. While the data for the 2001/02 financial year have not been release, looking at the number of cases referred to court per detective in 2000 gives some indicator of relative detective performance between provinces (Figure 9).²⁹


Figure 9: Number of cases referred to court per detective in 2000, by province



Source: SAPS Crime Information Analysis Centre

The number of cases referred to court has increased significantly since 1996, from 529,000 cases in that year to 782,000 in 2001, and 1.1 million in 2002—an increase of 110% between 1996 and 2002 (Figure 10).³⁰ The significant increase in the number of cases referred to court from 2000 onwards is partly a consequence of the large number of arrests resulting from the high-density operations of the National Crime Combating Strategy.

Figure 10: Number of cases referred to court by police, 1996 - 2002



Source: SAPS Crime Information Analysis Centre, National Prosecuting Authority

Negative indicators of performance

There are also readily available negative indicators, such as complaints against the police, especially allegations of corruption and abuse, and deaths in custody. Complaints against the police have risen dramatically from 1,999 in 1997/98 to 5,675 in 2001/02, an increase of 184%.³¹ This increase is to be expected, however, as more members of the public gain confidence in their ability to challenge police action successfully. In a sense, a high level of complaints at this stage in the country’s history may even be a positive indicator.

In addition, other negative indicators may suggest improvements in police behaviour. Deaths in custody and as a result of police action were lower in the period April 2001 to March 2002, compared to the same period for the previous year. In 2001/02, 585 people died in police custody or as a result of police action, compared to 687 in 2000/01, a decline of 15%.³²

According to the Independent Complaints Directorate or ICD (a statutory body with the responsibility of investigating police conduct and offences committed by members of the police service), of the 214 people who died in police custody in 2001/02, 76 died as a result of suicide, 75 from natural causes, 28 from injuries sustained prior to custody, 25 from injuries sustained in custody, and ten as a result of possible police negligence (Figure 11).³³


Figure 11: Deaths as a result of police action & in police custody, 1997/98 - 2001/02




Source: Independent Complaints Directorate

During 2001/02 the ICD investigated a further 371 deaths as a result of police action. Deaths as a result of police action during the course of an arrest declined from 168 in 2000/01 to 160 in 2001/02. Deaths as a result of police action committed while the victims were allegedly perpetrating a crime declined from 117 in 2000/01 to 80 in 2001/02—a decline of 32%. During 2001, 592 SAPS officers were charged with corruption (down from 1,048 in 2000), and 138 officers were convicted of a crime (down from 193 in 2000).³⁴

Issues to watch

Summary

It is difficult to evaluate police performance. The goals of the SAPS are often vague and information about police performance is not always made public. Claims that crime levels have stabilised do not withstand careful scrutiny. Any variation in the number of crimes recorded cannot be ascribed primarily to police action. Obsessing about the level of recorded crime does not serve the ultimate aim of the SAPS: to provide an effective, efficient, and equitable law enforcement service.

It is therefore imperative that the police and the public refrain from using recorded crime as a primary performance indicator for the SAPS. It creates perverse incentives for the police to under-record crime or to structure operations in a way that serves only to provide cosmetic statistics. The police require performance indicators that measure what they actually do, and which are simple, inexpensive to gather, and do not provide perverse incentives. They need indicators that are in their control and for which they can be held exclusively accountable.

Until this information is available, it is impossible to gauge whether the claims of politicians reflect the reality of performance. The public will continue to see the proclamations issued from parliamentary podiums as empty rhetoric, inconsistent with their own experience. There is a need for real performance measurement, both to account to the public and for the police to assure continuous improvement. The rise in convictions and decline in police killings in recent years suggests that there is good news to tell.