Number and distribution of SAPS members

Map 1: Number of people per SAPS member, 2001

Source: Stats SA and CIAC of SAPS

At the latest audit there were 21,797 detectives in the SAPS.10 According to the Auditor-General’s report for 2000, there were 22,034 detectives in the SAPS in 1998.11 This shows that the number of detectives fluctuates, but not significantly, and has remained in the region of 22,000 for the last four years.

However, according to a former detective with 25 years’ experience, until approximately 10 years ago there were in the region of only 7,000 detectives in the SAPS. Thereafter, many uniformed service members who, for example, investigated traffic offences, were also counted as detectives. This raises the possibility that a significant number of so-called detectives may not necessarily have been fully-fledged detectives. Nevertheless, as the section on training below shows, most of the approximately 22,000 detectives are likely to have received some sort of training over the last four years.

While the approximate number of detectives overall is known, it is less clear how many were based at specialised units. Given that, before restructuring began, there were 503 units and 1,103 police stations, if an even distribution between stations and units is assumed, this implies that about 7,000 detectives were based at units. The SAPS is currently in the process of conducting an audit of exactly how many detectives were based at these units and the number that will be redeployed to station level and to the new units.12

However, detectives generally make up approximately 20% of the total police compliment in a particular province. Similar to police members overall, detectives are not spread uniformly throughout the country. The Eastern Cape has the most detectives, and the Northern Cape the least.

Map 2: Ratio of square kilometers per detective, 2001

Source: Stats SA and CIAC of SAPS

Detectives are sparsely distributed throughout the country (Map 2). The Northern Cape has by far the fewest number of detectives according to the size of the province, with only one detective for every 550 square kilometers. Gauteng, however, is relatively better off with one detective having to cover only three square kilometers.

The figures suggest that in the Northern Cape, the vast distances and sparse population may leave an unmanageable amount of terrain to be covered by each detective. The Northern Province is also struggling with a greater than average kilometer per detective ratio.

On the other hand, the Northern Cape also has the fewest members of the public per detective (Map 3). This implies a lighter case load for detectives in the Northern Cape. By comparison, populous Northern Province has the most persons per detective: it has an unmanageable amount of more than 3,000 people for every one detective. This, combined with the low number of detectives, may make the predicament of detectives in the Northern Province particularly difficult.

Map 3: Ratio of people per detective, 2001

Source: Stats SA and CIAC of SAPS

KwaZulu-Natal, Mpumalanga, and the Western Cape also have relatively large populations served by relatively few detectives. The Eastern Cape, Northern Cape and Free State are relatively better off in respect of the population which must be served by detectives.

Map 4: Category 'A' crimes on hand per detective, 1998

Source: The Auditor-General’s report for Vote 29 on the financial statements of the SAPS and the Secretariat for Safety and Security for the year ended 31 March 2000.

Perhaps a more important indicator than area to be covered and population to be served are the numbers of crimes to be solved per detective (Map 4). Figures of crimes per detective were compiled by the Auditor-General’s Office for 1998.The provinces which had the most category ‘A’ crimes (which include only the most serious crimes such as murder, rape and robbery) on hand per detective as at September 1998 were the Northern Province, North West Province, Mpumalanga, and the Western Cape.

Those provinces with the fewest such crimes on hand as at September 1998 per detective were the Free State, Gauteng, KwaZulu-Natal and Northern Cape.

These figures suggest that Gauteng, for example, which had a relatively low number of crimes on hand, should probably register better performance by its detectives in terms of solving cases than other provinces. Whether that is indeed the case will be explored below. It is possible that the extra detectives deployed in the province are not able to translate this additional capacity into higher rates of solving crimes. The information in the preceding map is summarised in Table 5 below.

Table 5: Number of detectives and caseloads by province, 1998

Province	Category A crime cases on hand as at September 199813	Average number of category A crime cases per detective14
N Province	164,312	97
North West	142,783	85
Mpumalanga	87,570	79
W Cape	95,661	56
E Cape	242,111	48
KZN	144,173	45
Gauteng	204,017	41
N Cape	24,735	38
Free State	53,150	27
TOTAL	1,158,512	57 / 5315

Performance of the detective service

The performance of the detective service as a whole, of the specialised units in particular, and of individual detectives, is difficult to measure accurately. For example, in the past, the work rate of a particular detective was often measured within SAPS in terms of the number of statements he or she took over a particular period of time. This is a clumsy measure of work rate. In complicated commercial cases for example, it may be far more useful to obtain one 100-page affidavit from one highly reliable witness in an important case, which may take many days, than to gather 100 one-page statements in petty cases. In fact, this kind of performance indicator encouraged detectives to work on simple petty crimes rather than on difficult or important crimes.

Currently, a commonly used performance measure is the clearance rate (Table 6). According to this measure, success of the detective service as a whole is measured by the rate at which cases are cleared—that is, the rate at which cases are closed by the arrest of a suspect, or by the determination that a complaint had been unfounded, or when a complainant withdraws the case. The clearance rate for South Africa as a whole between 1995 and 1998 is illustrated in Table 6.

Table 6: Clearance rate in SA (%)*

Crime**	1995	1996	1997	1998
Murder	61	65	58	59
Robbery	34	31	30	31
Rape	80	76	74	63
Serious property crime	21	20	18	18
Assault	85	80	78	75
Fraud	57	58	54	52
Subtotal serious crimes	52	50	48	46
Less serious crimes	44	43	43	42
TOTAL	49	47	46	45

* The columns of this table are drawn from: Department of Finance, National Expenditure Survey 2000, p 186.

** The categories in this table aggregate a number of crime types. Murder includes attempts, serious property crime includes burglaries of residential and business property and theft of vehicles, robbery includes aggravated robberies, and assault includes assault with the intent to cause grievous bodily harm. The subtotal aggregates these crimes; while the next row adds the other crimes the police classify as being serious (such as thefts, arson, etc) and the total sums all serious crimes.

While some of these rates may appear at first glance to be impressive, it must be borne in mind that when a complainant withdraws a case, this is also counted as a cleared case. The clearance rate is therefore too broad to be a good performance measure.

This is particularly so given that the withdrawal rate in South Africa is high, especially in respect of what the SAPS call ‘social fabric crimes’ which include rape, assault and sexual assault. The figures above do show that the clearance rate for crimes such as rape and assault are high. To give an example of how withdrawals can influence the clearance rate, a recent study on social fabric crime in the Northern Cape found that 62% of all these crimes reported were withdrawn.16 This illustrates the extent to which the clearance rate may give an artificially inflated measure of performance.

The clearance rate should therefore be assessed along with the conviction rate, which is a partial measure of the quality of the dockets that are classified as closed when they are taken to court. The success of a case brought to court is heavily reliant on the quality of the investigative work.

One detective interviewed claimed that he had never seen a bad prosecutor lose a good case that had been well investigated. His point was that if a docket is well prepared by the detective, a good detective should be able to assist even an inexperienced and overburdened prosecutor to convict a guilty person.

The conviction rates detailed in Table 7 relate only to the cases actually brought to court. In other words, this is not the rate of convictions per recorded crime, but rather the number of convictions per case brought to court. While there is some debate about whether this too is an appropriate measure, it does provide a further indicator of performance.

Most striking about this data is that the most serious cases, such as murder, robbery and rape, reflect the lowest conviction rates. It is the less serious crimes that have a higher conviction rate.

This is an important point, given that prior to the recent restructuring, a large proportion of serious crime was investigated by the specialised units, while station level detectives dealt mainly with less serious crime. This suggests that station level detectives may have been more successful in court than the specialised units, at least with respect to less serious crime.

Table 7: Convictions as a ratio of cases brought to court in SA (%)*

Crime**	1995	1996	1997	1998
Murder	15	26	27	28
Robbery	17	25	22	22
Rape	14	19	17	20
Serious property crime	30	39	37	38
Assault	23	37	37	37
Fraud	30	36	41	41
Average conviction rate (serious crime)	23	34	33	34
Less serious crime	45	45	45	44
Average conviction rate) all crimes	33	39	39	39

* This table is drawn from: Department of Finance, National Expenditure Survey 2000, p 176.

** The categories in this table aggregate a number of crime types. Murder includes attempts, serious property crime includes burglaries of residential and business property and theft of vehicles, robbery includes aggravated robberies, and assault includes assault with the intent to cause grievous bodily harm. The subtotal aggregates these rates, while the next row adds the other crimes the police classify as less and the total sums all crimes.

An even more revealing performance indicator is that of convictions per recorded crime (Table 8). This measures the ratio of the number of crimes reported in a particular year to the number of convictions obtained. In other words, all the crimes which occurred in a particular year are measured up against all the convictions in that same year. The specific crimes and their convictions do not necessarily relate to each other. In this sense, this is a somewhat artificial measure, as the convictions counted are not in respect of the actual crimes reported. This is necessarily so because some crimes take many years to be processed through the system. A crime committed in 1998 for example may only be convicted in 2000.

Again, the most notable aspect of Table 8 is that serious crimes, with the exception of murder, have a far worse ratio of convictions to reported crimes than less serious crimes. Since serious crimes are generally investigated by specialised units, this would tend to indicate that specialised units may have under-performed in the periods under consideration. Nevertheless, it is true that serious crimes may be more difficult to prosecute.

Table 8: Convictions calculated as a ratio of the number of recorded crimes in SA (%)*

Crime**	1995	1996	1997	1998
Murder	8	10	11	11
Robbery	3	4	4	4
Rape	7	9	8	8
Serious property crime	4	5	5	5
Assault	8	12	11	11
Fraud	6	8	9	8
Subtotal serious crimes	6	8	8	8
Less serious crimes	12	12	12	12
TOTAL	8	10	10	9

NOTE: This table looks only at crimes reported in a year and convictions in a year. The convictions do not necessarily relate to the crimes reported in that year, as many cases take years to finalise.

* This table is taken from: Department of Finance, National Expenditure Survey 2000, p 177.

** See note to table above.

While Table 8 reflects an average conviction rate in the region of 10%, detectives that were interviewed estimated that if one were to follow each reported crime though to its conclusion, the rate of conviction would be more in the region of 5%. However in the absence of an integrated criminal justice information system capable of following a case from the reporting stage through to its conclusion, there is no means of substantiating this estimate. It is however encouraging that a pilot project of such a system is operational in Port Elizabeth at present.17

An attempt will be made in this section to measure the relative performance of detectives in the various provinces. The next few tables consider the numbers of certain types of crimes occurring in each province in 2000 with respect to the number of detectives in those provinces, as well as the number of guilty verdicts for those crimes. The crimes selected for analysis are murder, robbery with aggravating circumstances (which includes hijackings), rape and housebreaking.

An attempt is made to measure the relative performances of the detective service in the provinces, given the crime rate for a particular type of crime, the number of detectives, and the number of guilty verdicts in relation to the number of detectives in a province.

Murder, as the most serious crime, is considered first. Table 9 shows the various figures in respect of recorded murder and murder convictions in the various provinces.

Table 9: Number of recorded murders and guilty verdicts, per detective, 2000

Province	Number of detec-tives	Number of murders	Murders per detective	Murder cases going to court	Guilty verdicts for murder cases	Guilty verdicts as % of cases to court	Guilty verdicts as % of reported murder cases	Guilty verdicts per detective
Gauteng	5,320	4,913	0.92	1,721	716	42%	15%	0.13
KZN	3,816	5,771	1.51	2,389	690	29%	12%	0.18
Mpumalanga	1,071	1,033	0.96	407	118	29%	11%	0.11
N Province	1,539	808	0.53	377	149	40%	18%	0.10
N Cape	659	485	0.74	394	222	56%	46%	0.34
North West	1,697	1,101	0.65	555	221	40%	20%	0.13
E Cape	3,061	3,538	1.16	2,386	795	33%	22%	0.26
Free State	1,685	933	0.55	486	180	37%	19%	0.11
W Cape	2,058	3,413	1.66	1,981	916	46%	27%	0.45
TOTAL*	21,797	21,995	1.00	10,696	4,007	37%	18%	0.20

*Includes 891 detectives based at national headquarters not accounted for elsewhere.

There were almost two murders for every detective in 2000 in the Western Cape. The Northern Province is relatively fortunate to have had almost two detectives for every murder. However, the Northern Province did not appear to translate this relative advantage into a higher conviction rate (calculated on the basis of guilty verdicts as a percentage of reported murders). In fact, the Northern Province had an 18% conviction rate while the Western Cape had a 27% conviction rate.

At first glance the Northern Cape appears to have an astoundingly high rate of convictions as a proportion of cases going to court, and of reported crime, of 56% and 46% respectively, which would seem to indicate that its detective service is performing well. But in order to properly assess how well detectives are performing, one needs to take into account the number of guilty verdicts per detective, and not just the conviction rate. In this way, one accounts in some way for the case load experienced by a detective. The Northern Cape has a lower than average case load, which may to some extent account for its relatively good conviction rate with respect to murder.

Map 5: Guilty verdicts per detective for murder cases, 2000

The number of guilty verdicts per detective illustrated in Map 5 shows the Western Cape detective service to be far outperforming other provinces in convicting murderers, with 0.45 guilty verdicts per detective with respect to murder, while the Northern Province, Mpumalanga and the Free State are lagging behind the other provinces.

Murders are generally a highly prioritised crime. In addition, murder is often investigated by a specialised unit. Therefore this map to some extent reflects on the performance of the specialised units dealing with murder in the various provinces.

The performance of detectives with respect to robbery with aggravating circumstances is considered next (Table 10). This crime category includes crimes like hijacking and bank robbery. Gauteng has over ten robberies for every detective per year, while the Northern Cape is relatively fortunate to have almost two detectives per robbery.

The Eastern Cape is the best performer on both the guilty verdicts as a percentage of cases going to court (40%), as well as guilty verdicts as a percentage of reported robberies, at almost 9%. The good performance of the Eastern Cape also extends to the guilty verdicts per detective, illustrated in Map 6. Gauteng, the Western Cape and KwaZulu-Natal are also better performers with more guilty verdicts per detective than the other provinces.

Table 10: Robberies, robbery convictions and convictions per detective, 2000

Province	Detectives	Robberies	Robberies per detective	Robbery cases going to court	Guilty verdicts for robbery cases	Guilty verdicts as % of cases going to court	Guilty verdicts as % of reported robbery cases	Guilty verdicts per detective
Gauteng	5,320	52,017	9.77	4,809	955	20%	1.84%	0.18
KZN	3,816	23,635	6.19	3,277	598	18%	2.53%	0.16
Mpumalanga	1,071	5,289	4.94	674	99	15%	1.87%	0.09
N Province	1,539	3,162	2.05	630	116	18%	3.67%	0.08
N Cape	659	430	0.65	140	32	23%	7.44%	0.05
North West	1,697	5,569	3.28	823	171	21%	3.07%	0.10
E Cape	3,061	7,640	2.50	1,677	670	40%	8.77%	0.22
Free State	1,685	2,052	1.22	329	69	21%	3.36%	0.04
W Cape	2,058	10,685	5.19	1,439	346	24%	3.24%	0.17
TOTAL*	21,797	110,479	5.07	13,798	3,056	22%	2.77%	0.14

*Includes 891 detectives based at national headquarters not accounted for elsewhere.

Map 6: Guilty verdicts per detective for robbery cases, 2000

The Western Cape has more than 40 housebreakings per detective per year, while the Eastern Cape has only eight per detective (Table 11). Yet the Western Cape also has by far the highest number of guilty verdicts per detective with respect to housebreakings as illustrated in Map 7. Gauteng has by far the worst ratio of housebreaking guilty verdicts per detective. The Northern Cape has the best conviction rate with respect to housebreakings.

Map 7: Guilty verdicts per detective for housebreaking cases, 2000




Table 11: Housebreakings, housebreaking convictions and housebreaking guilty verdicts per detective, 2000

Province	Detectives	House-breaking (residential & business)	House-breaking per detective	House-breaking cases going to court	Guilty verdicts for house-breaking cases	Guilty verdicts as % of cases going to court	Guilty verdicts as % of reported house-breaking	Guilty verdicts per detective
Gauteng	5,320	104,446	19.6	8,462	2,380	28%	2.28	0.45
KZN	3,816	63,254	16.6	8,534	2,936	34%	4.64	0.77
Mpuma-langa	1,071	27,755	25.9	3,025	790	26%	2.85	0.74
Northern Province	1,539	23,370	15.2	3,790	1,157	31%	4.95	0.75
N Cape	659	10,130	15.4	2,600	1,296	50%	12.80	1.97
North West	1,697	27,107	16.0	3,781	1,271	34%	4.69	0.75
E Cape	3,061	44,656	14.6	10,010	3,868	39%	8.66	1.26
Free State	1,685	24,982	14.8	3,435	1,299	38%	5.20	0.77
W Cape	2,058	68,857	33.46	10,928	5,055	46%	7.34	2.46
TOTAL*	21,797	394,557	18.1	54,565	20,052	37%	5.08	0.92

*Includes 891 detectives based at national headquarters not accounted for elsewhere.

Table 12: Rapes, rape convictions and rape guilty verdicts per detective, 2000

Province	Detectives	Rapes	Rape per detective	Rape cases going to court	Rape guilty verdicts	Guilty verdicts as % of cases going to court	Guilty verdicts as % of reported rapes	Guilty verdicts per detective
Gauteng	5,320	12,378	2.33	4,693	675	14%	5,45%	0.13
KZN	3,816	9,419	2.47	4,475	565	13%	6.00%
Mpuma-langa	1,071	3,559	3.32	984	130	13%	3.65%	0.12
Northern Province	1,539	4,172	2.71	1,829	349	19%	8.37%	0.23
N Cape	659	1,539	2.34	921	203	22%	13.19%	0.13
North West	1,697	4,739	2.79	1,959	491	25%	10.36%	0.29
E Cape	3,061	6,923	2.26	3,712	734	20%	10.60%	0.24
Free State	1,685	4,495	2.07	1,665	212	13%	6.07%	0.13
W Cape	2,058	6,784	3.30	3,757	660	18%	9.73%	0.32
TOTAL*	21,797	53,008	2.43	23,995	4,019	17%	7.58%	0.18

The Western Cape has over three rape cases per detective per year (Table 12), while the Eastern Cape has just over one rape per detective. The Northern Cape has the highest rape conviction rate, and Mpumalanga the lowest.

However, the Western Cape by far outperforms the other provinces on the rape guilty verdicts per detective illustrated in Map 8, while Mpumalanga is the worst performer.

Map 8: Guilty verdicts per detective for rape cases, 2000

The last map (Map 9) summarises the previous four maps and totals the number of guilty verdicts for murder, robberies, housebreaking and rape per detective.

Map 9: Guilty verdicts for murder, robbery, housebreaking and rape per detective, 2000

As can be seen from Map 9, the Western Cape is the best performer for the sum of these crime types with almost four guilty verdicts per detective per year. Gauteng and the Free State are the worst performers for these four crime types, with only under one guilty verdict per detective.

Factors affecting performance

Based on the available statistics, better performance has been recorded for less serious crimes than for serious crimes. Performance overall is poor, however. Performance varies from province to province, as illustrated by the discussion above. There are a number of factors relating to the work of a detective that may impinge upon performance of individual detectives, and thus, ultimately, the detective service as a whole

Career paths

The career path to becoming a detective is variable. The detectives interviewed described different career paths. Generally, before becoming a detective it is first necessary to join a uniform branch of the SAPS and work as a uniformed officer at station level for approximately five years. However, the stipulated minimum is one year at a uniform branch.18

A uniform member must then apply to become a detective at station level. The new detective may complete some detective’s courses; however these are not obligatory before becoming a detective and may in fact be completed some time after being made a detective.19

After another approximately five years, a detective can then apply to join a specialised unit. This career path as described is not fixed. Detectives believed there was often a measure of luck involved. Whether a detective is appointed to where he or she wishes to go often seemed to depend on contacts made and relationships established prior to making the application.

Promotion of detectives

Promotion is integral to the career path of a detective. Promotion is the traditional way of rewarding good work in the detective service. As a result, however, detectives are often promoted outside of their competencies. This is because excellent investigators often make poor managers, and promotion often entails a greater managerial role. Furthermore, promotion may also entail promotion away from a detective’s network of informants and contacts, as promotion often means transferal to another geographical area.

However the SAPS is currently exploring a system whereby good detectives will be given the option of being rewarded with a pay-rise but not a change in rank; detectives are thereby given the incentive to do good work and their skills as investigators are also retained.20 This kind of system will require a whole-scale revision of the remuneration system of the SAPS, which is largely rank-based.

It is envisaged that the current system, which applies to high ranking officers, of performance-based remuneration will be cascaded down to the lower ranking officers. However for this to be implemented effectively will require a solid performance measurement system and as we have seen in the section on performance above, there are currently few performance indicators regularly measured in the detective service.

To have an accurate measure of conviction rates, for example, will require the entire criminal justice system to be integrated electronically, so that a particular docket can be tracked all the way from the complainant first reporting the matter at station level through to the court case and further into the prison system.

Currently, conviction rates are measured by the artificial ratio of convictions to dockets opened in a particular year. This is because the courts operate on an entirely paper-based system and there is no integrated computerised system following the progress of a particular matter.

This integrated system is still in the developmental phase. A pilot project is operational in Port Elizabeth. Once a national system is in place, it will be possible to obtain a true conviction rate, and as a result, follow the performance of individual detectives. It will then be possible accurately to track which detectives have the best track records, and reward them with promotion accordingly. In the absence of such an objective measure of performance, there is a risk that pay-promotion may fall victim to favoritism.

Training of detectives

Training is not obligatory for station level detectives; basic courses are usually completed after a member has become a detective. Detectives said they did not get any special training as a matter of course before joining a unit, but may get specialised training during the course of their work with a unit. Usually most of what a good detective knows is learnt ‘on the job’.

Detectives said that in the past the detective’s courses were far too theoretical. Where good theory was taught, new detectives were often unable to implement this knowledge because of the hierarchical structure of the police and other bureaucratic problems (see chapter 2).

A new detective academy, the South African Police Detective Academy, was opened in Pretoria in October 1997. It was envisaged as a Detective Academy ‘for Africa’. Both the Southern African Development Community (SADC) and the Southern African Regional Police Chiefs Organisation (SARPCO) were involved in its establishment.21 The training curricula were prepared with the help of the policing agencies of the United States, the United Kingdom, the Federal Republic of Germany and the Netherlands. The US Federal Bureau of Investigations (FBI) also made available lecturers and other means of support in its opening phases. 22

A training academy also exists in Paarl, which caters mainly for middle management and ranking officers’ courses. The academy at Graaff-Reinet has been scaled down, and is concerned mainly with human resource management.

Most detectives said the courses have improved greatly in recent times and the training now involves mock cases and is much more useful and practical. Some excellent training, especially on commercial crime investigation, has been developed, according to those who have experienced some of this training.

However, the problem is that there are few experienced people to implement this training. More and more good people are leaving the special units and detective service—some estimated an almost 100% turnover of people at superintendent level (see the discussion on the skills drain below).

A further problem identified by a former detective is that sending an officer on a course was a good way for a commanding officer to remove someone with whom he or she did not get on. The consequence of this is that certain people went on courses unrelated to their particular work.

Another problem that is said to have occurred is that detectives are often sent year after year on the same courses. Indeed, some viewed going on course as a kind of ‘holiday’ and did not complain that they were not learning anything new. Others were known as ‘professional course go-ers’ and these people rarely implemented their knowledge. The perception is that they were sent to simply fill the quotas.

It is not known to what extent these problems have been remedied with the establishment of the new academy. Higher ranking detectives said this problem had not been completely eradicated.

Quantitative statistics on training in the detective service show that in the four-year period since the opening of the new detective academy, from October 1997 to September 2001, as many as 31,107 detectives received some sort of training. This implies that some detectives received more than one kind of training over this period, as the latest figures show there to be 21,797 detectives in the detective service. However, of the 31,107, as many as 17,684 were trained on the ‘Detective Learning Programme’, a basic course. More specific training accounted for the training of 11,299 detectives only in the period under consideration.

Moreover, a significant portion (10%) of training in this group was in general computing skills. A further portion (15%) was in commercial crime related disciplines. This may be because of the highly specialised skills required for commercial crime investigation. A further significant portion (16%) was in management related courses.

Work load of detectives

Detectives are overloaded, as the figures in the section on performance above makes clear. Many detectives said overloading was exacerbated by mismanagement, and that the vast numbers of cases reported did not alone cause the problem. Instead of teaming up to tackle cases on a case-by-case basis, each detective generally receives an unmanageable load to tackle on his or her own. This is no longer the situation in the two new specialised units, where ‘task teams’ are now set up to handle specific problems.

Another example of a management problem given by a detective was with respect to the commercial crime unit. Members of the public had a tendency to load the police with small cases in order to further their own civil claims. The management of the unit did not dissuade these complainants and instead attempted to deal with these myriad small claims. One detective suggested that teams of detectives should have been assigned to major cases, and minor complainants encouraged first to gather more evidence substantiating a charge before approaching the police.

Detectives said that at the most, they could properly deal with ten cases at any one time. However, even those from specialised units mentioned having to handle 40, to 50 and even 60 matters at any one time.

Statistics provided by the Auditor-General (Table 13) show the case load of category ‘A’ crimes per detective to reach as many as 97 per detective in 1998 in the Northern Province, for example. Category ‘A’ crimes include only the most serious crimes such as murder and rape.

Table 13: Detective case loads, by province, 1998

Province	Category A crime cases on hand as at September 1998	Average number of category A crime cases per detective	Number of detectives
N Province	164,312	97	1,694
North West	142,783	85	1,680
Mpumalanga	87,570	79	1,108
W Cape	95,661	56	1,708
E Cape	242,111	48	5,044
KZN	144,173	45	3,204
Gauteng	204,017	41	4,976
N Cape	24,735	38	651
Free State	53,150	27	1,969
TOTAL	1,158,512	57 / 5323	22,034

Source: The Auditor-General’s report for Vote 29 on the financial statements of the SAPS and the Secretariat for Safety and Security for the year ended 31 March 2000. Performance Audits of Human Resource Management and the Management of the Central Firearms Register.

The consequence of such overloading is that the investigation is administered rather than solved. The case becomes so old that it is eventually dropped as the complainant, witnesses, and other persons involved simply lose interest. Evidence may also be lost and witnesses’ memories may fail them. Withdrawals of charges and prosecutors declining to prosecute are the likely result.

Skills drain

Detectives mentioned a significant skills drain. Members have left the detective service for similar employment with private companies, such as forensic auditing companies; for posts at academic institutions; and for posts with the Directorate of Special Operations (DSO). Some have also left to take up employment completely unrelated to their previous work.

Reasons given by detectives for leaving the service related mostly to their inability to perform their jobs adequately such as:

Surprisingly few mentioned low pay as a factor. Indeed most said given the various benefits available to members of the SAPS, they were not that much better off now that they had left the service for private or other employment.

Many who had left the detective service were intensely proud of having been detectives and would always considered themselves to be detectives first and foremost. However, these respondents in particular felt the low prestige attached to being a detective in the SAPS today was a significant factor in their leaving the service.

The skills drain becomes a further problem with which detectives who remain in the SAPS have to contend. The loss of skills means fewer mentors for new detectives; fewer possibilities for teamwork; and greater loads on experienced detectives. These problems have contributed to the less than optimal performance of the detective service as a whole.