1    Introduction

In this report, serious crime ratios for 1999 will be compared with those for 1994, 1995, 1996, 1997 and 1998 respectively in order to analyze whether any improvement in the South African crime situation can be discerned up to the end of 1999 or whether the situation is stabilizing or deteriorating (see footnote 1).

2   The national serious crime picture

Figures relating to the 20 most serious crime tendencies (all showing an annual incidence of more than twenty cases per 100 000 of the population) are compared for 1994, 1995, 1996, 1997, 1998 and 1999 respectively in Table 1. These twenty crime tendencies comprise more than 99.0% of the total serious crime figure accounted for by the 28 serious crime tendencies featured in Annexure A, which contains reported crime figures for the above years, as well as ratios per 100 000 of the population.

The 12 month moving averages pertaining to the 20 crime tendencies featured in Table 1 are presented in Figures 1 - 20. This will assist an understanding of why certain tendencies are interpreted as increasing, decreasing or stabilizing. It has to be emphasized that the period on which deductions are based should be clearly specified. The arrows in Table 1 refer to the whole period of 1994 - 1999.

An analysis of the figures pertaining to the crime tendencies featured in Table 1 indicates that:

Three crime trends, namely murder, attempted murder and arson (accounting for 2.7% of all serious crime featured in Annexure A) evidenced a decrease during 1999.

Eleven crime trends (44.6% of all serious crime featured in Annexure A) showed a stabilization during 1999. These crimes are: rape, common assault, burglary at business premises, stock theft, shoplifting, theft of motor vehicles, theft out of/from motor vehicles, commercial crime, malicious damage to property, drug related crime and driving under the influence of alcohol or drugs. It is necessary to focus attention on a very positive development, namely the stabilization in drug related offenses. Between 1994 and 1998 an almost constant decrease in drug related crime cases was observed. During 1999 an increase occurred, with the result that this trend has now stabilized. Drug related crime is usually not reported to the police, but heavily dependent on police action for detection. Especially as a result of Operation Monozite, more and more people in possession of drugs (either for personal use or trading purposes) were arrested during 1999. Drugs, like illegal firearms, are a very important factor conducive to crime in South Africa. Hopefully a tremendous increase in drug related crime cases will be reported in 2000 because that will actually mean that progress is being made in combating this very important generator of crime.

Six trends (accounting for 51.9% of all serious crime featured in Annexure A) evidenced increases during 1999. These are: robbery with aggravating circumstances, assault GBH, residential burglary, other robbery, other thefts and the illegal possession of firearms/ammunition. The increase in the illegal possession of firearms, which accounts for 0.7% of the volume of crime reflected in Annexure A and 1.3% of those crimes on the increase, should not be seen as a negative development. As in the case of drug related crimes, this may indicate increasing effectiveness of the Firearm Investigation Units in the detection of illegal firearms.

The increase in assault GBH should also not be seen as a negative development. The social fabric related crimes (rape and both common and serious assault) are to a large extent social/domestic in nature. The majority of these crimes occur within the family and/or friendship circle. Internationally, victims find it extremely difficult to report these crimes to the police - precisely because they often involve spouses, parents, children, boyfriends or girlfriends, etc. The South African Government, and specifically the SAPS, have since 1994 launched various initiatives aimed at increasing the reporting of these crimes. Women and children’s rights are being emphasized (workshops were even held in the rural areas to this end) and user-friendly units like the Child Protection Units and Family Violence, Child Protection and Sexual Offenses Units were either created or expanded to make it easier for women and children to report crimes of this nature. The Domestic Violence Act was implemented during December 1999, outlining the duties of police officials in cases of domestic violence and making it compulsory to attend to complaints of domestic violence. On account of these actions by Government to address domestic violence, one could have expected the increases in reported cases of rape and assault GBH between 1994 and 1997 and again as far as both types of assault are concerned during 1999. That the initiatives of the Government deliver the expected results is a positive sign.

If this Semester Report (1/2000), which is in fact an annual report for 1999, is compared to Semester Report 1/1999 dealing with the annual crime figures for 1998, the following emerges: 35.5%, 38.1% and 25.7% of the volume of serious crime were respectively increasing, stabilizing or decreasing during 1998 (see Semester Report 1/99), while 51.9%, 44.6% and 2.7% of serious crime were correspondingly increasing, stabilizing or decreasing during 1999. It is also clear from Table 1 that there has been a 4.7% increase in the incidence of the 20 most common serious crime tendencies since 1998.

Notwithstanding the fact that the majority of the twenty individual crime trends featured in Table 1 are stabilizing, the incidence of crime in the country is still at a very high and definitely unacceptable level.

Although crime ratios (crimes reported per 100 000 of the population) indicate the trends discussed above, this does not necessarily mean that the overall incidence or figures of serious crime (i.e. frequencies) are experiencing a corresponding decrease, stabilization or increase. The real frequency increases or decreases over the period 1994 to 1999 and also between 1998 and 1999 are indicated in Table 2.

From Table 2 the following can be deduced:

- There has been a real frequency (i.e. raw figure) increase of 307 077 reported serious crime cases between 1994 and 1999 (that is an average increase of 61 415 cases per year). As far as police planning (with regard to both human and logistical resources) is concerned, this will have to be taken into consideration.

- A very strong upward pressure on many of the crime trends since 1998 is further evident from Table 2. Whilst there was an increase of 307 077 crime cases between 1994 and 1999, an increase of 159 779 reported crime cases occurred between 1998 and 1999 alone (the increase over the last year thus accounted for nearly half the total increase observed over five years). The increase between 1998 and 1999, especially with regard to other robbery (which increased at least 7 times in relation to its expected normal population linked increase rate) and robbery aggravated (which increased at least 4 times more than its normal population growth linked increase rate), provides reason for serious concern. Other notable increases are observed with regard to other thefts (5 times more than the normal population increase), assault GBH (4 times more than the normal population increase) and assault common (5 times more than its normal population growth linked increase).

Among the subcategories of robbery with aggravating circumstances, carjacking, hijacking of trucks, bank robbery and robbery of cash in transit are the subtends which receive the most publicity and usually have the harshest impact on national and international perceptions regarding crime in South Africa. Statistics pertaining to these four subtends have been separately kept since 1996 [see the last page of Annexure A for these statistics]. From Annexure A the following is evident:

- The incidence of bank robbery and the robbery of cash in transit, which received prominent media attention ever since the Bronkhorstspruit robbery of R18 million, remained very stable between 1998 and 1999. It is furthermore evident that Gauteng, KwaZulu-Natal and the North West Province accounted for a combined figure of 70.1% of all cases of bank robbery during 1999, as well as 68.3% of the reported cases of robbery of cash in transit. The stabilization in both these two subtends can be attributed mainly to intensive, intelligence driven investigation and resulting arrests.

- Carjackings and hijackings of trucks stabilized during the last months of 1999 after three years of constant increases. From Annexure A it is clear that Gauteng and KwaZulu-Natal accounted for a combined total of 80.4% of carjackings and 83.5% of hijackings of trucks reported during 1999. Within Gauteng it is especially the Areas Johannesburg, East Rand and Soweto and within KwaZulu-Natal, the Durban Area which are very prone to car and truckjackings.

- It is important to keep in mind that the four crime tendencies discussed above have already been accounted for under the crime category of robbery with aggravating circumstances.

3    A provincial comparison of crime ratios

The provincial crime ratios pertaining to fifteen crime tendencies are shown in Table 3. The provinces are also ranked in the table from the province with the highest ratio regarding any of the tendencies to the province with the lowest ratio. All provinces above the national average are shaded in black. The ratios indicated in Table 3 are based on estimates derived from the 1996 census results. From Table 3 the following is clear:

The Western Cape occupies a first position with regard to housebreaking at both business and residential premises, shoplifting and theft out of/from motor vehicles, as well as other thefts. In the case of murder and common assault, it also occupies a first position. It occupies a second position as far as 6 crime categories are concerned, namely other robbery, theft of motor vehicles, commercial crime, attempted murder, rape and assault GBH. The Western Cape also occupies a third position with regard to robbery with aggravating circumstances (with the latter not at a level above the national average). In the case of stock theft, the Western Cape featured a ratio below the national average.

Gauteng occupies a first position with regard to other robbery, theft of motor vehicles, commercial crime, attempted murder and robbery with aggravating circumstances. It also occupies a second position as far as burglary at residential premises, theft out of/from motor vehicles and other thefts are concerned, a third position with regard to housebreaking at business premises, shoplifting, murder, rape and assault GBH and a fourth position in relation to common assault. Only in the case of stock theft does Gauteng feature a ratio below the national average.

The Northern Cape occupies a first position with regard to stock theft, rape and assault GBH. It occupies a second position as far as housebreaking at business premises, shoplifting and common assault and a third position as far as burglary at residential premises, other robbery, theft out of/from motor vehicles, other theft and commercial crime are concerned. With regard to murder, the Northern Cape occupies a fourth position.

The Free State occupies a second position with regard to stock theft and a third position with regard to common assault. It also occupies four fourth positions, namely as far as shoplifting, other thefts, commercial crime and assault GBH are concerned.

KwaZulu-Natal and Mpumalanga both feature 4 crimes above the national average, while the North West and Eastern Cape respectively feature 3 crimes above the national average. The Northern Province does not feature above the national average at all.

4     The crime situation at area level

Crime ratios indicating the incidence of crime in the 42 Police Areas of South Africa are featured in Table 4 (indicating the 10 areas experiencing the highest crime ratios for each crime) and the maps contained in Annexure C (see footnote 2). From Table 4 and the 17 maps contained in Annexure C, the following are evident: (see footnote 3):

Johannesburg, the Eastern Metropolis and Southern Cape are the only three Policing Areas featuring in the top two fifths of the spectrum for murder (which fluctuated between 137 and 11 per 100 000 of the population during 1999 among the 42 police areas).

Johannesburg, the Eastern Metropolis, the Western Metropolis, Soweto, the Midlands and Umfolozi featured the top attempted murder ratios (between 143 - 100 per 100 000 of the population) during 1999 [see map 2]. Other Areas with high ratios are the East Rand and Vaalrand.

With regard to the ratio pertaining to robbery with aggravating circumstances, Johannesburg is the only Area featuring among the top two fifths of the spectrum. However, other Areas featuring significantly high ratios in this regard are the East Rand, Soweto, North Rand and Durban. It is interesting to note that the ratio for robbery with aggravating circumstances fluctuates between 13 and 1 754 per 100 000 of the population among the 42 Police Areas, with Johannesburg topping the list at the latter ratio.

Port Elizabeth, Southern Cape, Johannesburg, Uitenhage, Diamond field, Vaalrand, Soweto, Upper Karoo, Gordonia, the Eastern Metropolis and Southern Free State feature within the top two fifths of the rape spectrum [see map 4]. The rape spectrum fluctuates between 43 and 232 per 100 000 of the population.

Gordonia (with a ratio of 1 997 per 100 000) is the Area featuring the highest incidence of assault GBH (serious assault). Four other Areas, namely Upper Karoo, Diamond field, Uitenhage and Southern Cape, also feature within the top two fifths of the spectrum as far as serious assault is concerned [see map 5].

The Southern Cape, Upper Karoo, Namaqualand, Boland, Karoo, Johannesburg, Diamond field, Uitenhage, Southern Free State and Eastern Metropolis are the Areas featuring the top ratios as far as common assault is concerned, with ratios ranging between 952 and 1 473 per 100 000 of the population [see map 6].

The Police Areas of Johannesburg, Pretoria, the Southern Cape, the Western Metropolis, the Eastern Metropolis and Boland all feature ratios for housebreaking at residential premises which fall within the top three fifths of the spectrum as far as this crime is concerned [see map 7].

The Police Areas of Johannesburg, Southern Cape, Boland and Upper Karoo all feature ratios of housebreaking at business premises which fall within the top two fifths of the spectrum as far as this crime is concerned [see map 8]. Other Areas worth mentioning in this regard are Namaqualand, Karoo, Diamond field, the Western Metropolis, the Eastern Metropolis and Gordonia.

As in the case of robbery with aggravating circumstances, only Area Johannesburg features among the top two fifths of the crime ratio spectrum as far as other robbery (mostly common street robbery) is concerned [see map 9]. The other Areas featuring significant ratios for this crime are Pretoria, the Western Metropolis, Port Elizabeth, Diamond field, the Eastern Metropolis, East Rand and Vaalrand.

The stock theft ratio fluctuates between 1 (Soweto) and 562 (Karoo) [see map 10]. The Police Areas falling within the top two fifths of the stock theft ratio spectrum are the Karoo, Upper Karoo, Drakensberg and Eastern Highveld Areas. Other Areas worth mentioning in this regard are Grahamstown, the Eastern Free State, Southern Free State, Tugela and Gordonia.

As far as the shoplifting ratio is concerned, the fluctuation ranges between 27 (Soweto) and 657 (Johannesburg) [see map 11]. The Police Areas ranked among the top three fifths of the ratio spectrum for this crime are Johannesburg, the Western Metropolis, Southern Cape and Port Elizabeth.

As in the case of robbery with aggravating circumstances and other robbery, only Johannesburg features among the top two fifths of the spectrum pertaining to motor vehicle theft [see map 12]. The overall ratio per Area fluctuates from 32 (Far North) to 2 244 (Johannesburg). The only other areas featuring significantly high ratios for this crime are Pretoria, the East Rand, West Rand and North Rand.

Johannesburg once again features the highest ratio of theft out of/from motor vehicles, with the only other Police Area falling within the top two fifths of the ratio spectrum for this crime being the Western Metropolis [see map 13]. Other areas with significantly high ratios for this category of theft are Pretoria, the Eastern Metropolis, Port Elizabeth, East London, Durban, Boland and Diamond field.

The ratio for other theft fluctuates between 298 (Umtata) and 4 901 (Johannesburg). The only Police Areas falling within the top three fifths of the ratio spectrum for this crime are Johannesburg and Pretoria [see map 14].

As with regard to a number of property related crimes already discussed, Johannesburg is the only Area falling within the top three fifths of the ratio spectrum as far as commercial crime is concerned [see map 15]. The commercial crime ratio is also high in Pretoria and the West Rand.

From the above information a number of deductions can be made which may shed some light on the dynamics of, and possible explanations for, the phenomenon of crime. It would be of great benefit to further investigate the tentative deductions to arrive at definite explanations. Without explanations for the incidence of crime, prevention remains very difficult and even impossible.

It appears as if some of the violent crimes (murder and attempted murder) and crimes related to the social fabric (both categories of assault and rape) have a higher incidence in -

(a) Areas experiencing a high and rapid rate of urbanization, with concomitant social problems such as urban unemployment, informal settlements which are difficult to police, high levels of relative deprivation and expectations, the displacement of rural support and value structures by peer group pressures (e.g. the phenomenon of gangs), and

(b) Rural areas historically associated with the phenomenon of alcohol abuse (inter alia as a result of the historic "tot system"). The association between alcohol and drug abuse and crimes like murder, rape and assault is also supported by Figures 1, 4, 5 and 6 of the national serious crime picture. All those figures - and especially the assault figures - reach a peak during the December festive season and then a second peak during the long weekends of April.   Examples of category (a) above are the Police Areas of the Eastern Metropolis (Western Cape), Johannesburg, Vaalrand, Soweto (Gauteng) and Port Elizabeth (Eastern Cape), while examples pertaining to category (b) are the Police Areas of Gordonia, Diamond field, Upper Karoo (Northern Cape), Karoo, Uitenhage (Eastern Cape), Southern Cape and Boland (Western Cape).

In the case of attempted murder, the above pattern of crime in rural areas being associated largely with alcohol abuse is to some extent contradicted. There appears to be a much stronger association between attempted murder and Areas where faction/political/ideological conflicts occur, such as the Midlands, Umzimkulu and Durban and Areas where integrant and gang/PAGAD conflicts occur, such as the Eastern and Western Metropolis (the Cape Town Metropolitan Area). Attempted murder is also high in many Gauteng Areas such as Johannesburg, Soweto and the East Rand, where armed robbery (especially hijackings and robbery at residential premises) is at its highest levels. Most of these conflicts/crimes are waged or committed with firearms, whereas assaults (both serious and common) usually involve the use of other physical violence and/or other instruments like knives and broken bottles.

It is interesting to note that malicious damage to property follows much the same distribution pattern than both common and serious assault [see map 16]. It appears as if an association between arson and faction/clan conflict in Umzimkulu, Umtata and Umfolozi and arson and deliberate veld fires in the Southern Free State, Drakensberg, etc. exists. It may also be that an association between assault and arson (e.g. Southern Cape and Diamond field) could be discerned [see map 17].

An evaluation of property related crimes is complicated by the fact that the population basis upon which the ratios pertaining to these crimes is based, is still the human population and not the property involved (e.g. vehicles in the case of vehicle theft) as it actually should be. Notwithstanding the above, the maps illustrating the incidence of property related crime seem to reveal two patterns:

- Property related crimes associated to a larger extent with organized crime (i.e. syndicates involved in crimes like vehicle theft, robbery with aggravating circumstances - including bank robbery, robbery of cash in transit and hijackings - and commercial crime) appear to be more prevalent in Johannesburg and the Areas surrounding it in Gauteng. Other Areas worth mentioning in respect of these kinds of property related crime are the Eastern and the Western Metropolis (together forming the greater Cape Town area), Durban and Port Elizabeth. Those involved in organized crime would logically concentrate on the areas where

(i) targets and recruits are available and
(ii) a market for stolen goods exists.

The metropolitan areas, and especially the greater Johannesburg area, represent such ideal settings.

- Property related crimes to a lesser extent associated with organized (syndicate) crime and more often associated with individual criminals such as housebreaking at residential and business premises, common/other robbery, shoplifting, other theft and stock theft, are more widely distributed throughout the country. The explanation for the broader and also more rural distribution of these crimes is that the single criminal tends more often to be motivated by need rather than by greed, whereas in the case of organized crime greed seems to be the prominent motivating factor. Those stealing as a result of need are not restricted to the metropolitan areas, but are distributed throughout the country. The latter type of criminal tends to steal anything capable of satisfying his/her needs, while organized criminals only tend to operate where a ready workforce (so-called runners) and markets are available.
The above arguments are supported by research (particularly docket analysis, as well as multi-variate analysis) done by the CIMC (now renamed the CIAC) since 1996.

5    The crime situation at station level

As stated in previous quarterly reports, it will be difficult (and even impossible) to compare stations (precincts) with one another without an idea regarding the population served by each station in order to calculate the relevant crime ratios. It has also been stated in previous quarterly reports that a synchronization of station boundaries and census boundaries (EA’s) is necessary before the calculation of station ratios can be done. Synchronization has not been achieved to date and the CIAC, in cooperation with the HSRC and Statistics South Africa, is now looking at an alternative option of estimated station populations based on a best fit model. In the meantime, people are always interested in what happens in the crime field within the smallest geographical police area possible. For that reason an extra report containing all official raw serious crime figures for stations was loaded onto the SAPS Web site and provided in hard copy format to CIMC (now CIAC) offices. The same will apply in the case of this report, but an extra Annexure is also added. This Annexure D will contain a weighted crime frequency for 1999 for every station per province. This will clearly indicate those stations with a weighted crime frequency within the national upper quartile (all the stations in a given province among the top 25% in South Africa as far as the incidence of serious crime is concerned), as well as those between the upper quartile and the middle quartile (top 50%) and then those featuring in the mean and lower quartile.

The above division at least gives a broad indication of crime flash points in SA, without proper consideration of population sizes. The weighted frequency for each station was calculated in the following way:

If all the stations in South Africa are ordered from that with the highest crime weight calculated according to the above formula (which is CR Swart Square - 43 238) to the lowest (which is Owerland in Area Umtata and Klaarstroom in the Southern Cape - both 0) and the top 25% of these stations are then identified as real problem/priority stations, the following emerges [see Annexure D for all the detail per province]:

From Table 6 it is clear that between 93.5% (Gauteng) and 43.1% (Northern Province) and in most cases more than 50.0% of each of the nine provinces’ serious crime weight can be addressed by focusing on the top 25% of the problem/priority stations in South Africa. Even more interesting is the fact that by addressing crime in 25.3% of the KwaZulu-Natal precincts, 66.8% of the serious crime weight in that province will be addressed and by addressing crime in 4.8% of the Northern Cape precincts, 43.6% of the serious crime weight in that province will be addressed.

---

Footnotes:

1. In the following sections of this report, crime ratios will be used for discussion dealing with the incidence of crime. In most cases specific crime tendencies are discussed in terms of incidence per 100 000 of the population. These ratios, employing population figures as an equalizer or common denominator, make it possible to compare different crimes with one another. The employment of ratios, instead of frequencies, is common international practice. The advantage of employing ratios is that the growth in population figures over time is neutralized. In this way provision is made to prevent the population growth from causing an artificial increase in crime tendencies which does not reflect a real increase in the incidence of crime. Furthermore, the employment of ratios enables comparisons of the incidence of crime among different communities, notwithstanding differences of population size. It also has to be emphasized that comparisons based on ratios are more correct than comparisons involving frequencies, as ratios indicate the following:

- The probability of becoming the victim of a specific crime. The chances of falling victim to an act of assault with the intent to inflict grievous bodily harm are at least two times higher in the Northern Cape, for example, than in the Western Cape (see Table 3).

- The presence of those factors or combination of factors conducive to the incidence of a specific crime which have to be addressed in order to combat that crime effectively.

It would, however, be incorrect to rely only on ratios as far as planning pertaining to policing is concerned. The distribution of human and logistic resources has to be depended on the frequency of the incidence of crime, while also taking crime ratios into consideration. The largest volumes of crime will logically be associated with the densest concentrations of people - although these areas may not necessarily reflect the highest crime ratios - and will therefore need the largest concentration of human and logistical resources available to the Police. In determining the distribution of Police resources, both the concentration of crime (i.e. frequency) and ratios (i.e. incidence per population), as well as the policeability of the crimes occurring within a given geographic area, should eventually be taken into consideration. (BACK)

2.  Maps 1 - 17 are contained in Annexure C (Crime ratios per 100 000 of the population, related to specific crime categories). The maps have been compiled by the HSRC GIS Unit and are reproduced with their kind permission. (BACK)

3.  The reader should not compare this set of area data with that of previous reports as this set of area ratios are based on the new 1996 Census as adapted for population growth (population figures) for the areas.